Southwestern Desert Resources - Amazon Web Services · 2017-05-31 · This book brings together peer-reviewed research from two interagency projects that focused on Southwestern Desert

Southwestern Desert Resources


Edited by William Halvorson, Cecil Schwalbe, and Charles van Riper III

The University of Arizona PressTucson

This book brings together peer-reviewed research from twointeragency projects that focused on Southwestern Desertecosystems and resources. The first is a series of biennialconferences that were held from 1996 through 2006. Thesewere guided and managed by a committee representingArizona Game and Fish Department, National Park Service,The Nature Conservancy, Sky Island Alliance, SonoranInstitute, Universidad de Sonora, University of ArizonaSchool of Natural Resources, USDA Agricultural ResearchService, USDA Natural Resources Conservation Service,U.S. Forest Service, and U.S. Geological Survey. Thesecond is an interagency project that took place from 2000through 2008. This inventory of the National Park Serviceunits in southern Arizona and eastern New Mexico was acoordinated effort of the National Park Service, U.S.Geological Survey, and the University of Arizona, School ofNatural Resources.

The University of Arizona Press© 2010 The Arizona Board of RegentsFirst PrintingAll rights reserved

Manufactured in the United States of America on acid-free,archival-quality paper containing a minimum of 30% post-consumer waste and processed chlorine free.

15 14 13 12 11 10 6 5 4 3 2 1

ISBN 978-0-8165-2817-2Library of Congress Control Number 2010922391

This book was published from camera-ready copy.General layout and editing by the volume editors. Section photos by Cecil Schwalbe. Final design andpreparation by Julie St. John.

CONTENTS

Dedication viiiIntroduction 1

William L. Halvorson

AREA INVENTORIES1. Vascular Plant and Vertebrate Inventory of Casa Grande Ruins National Monument 7

Brian F. Powell, Eric W. Albrecht, Cecilia A. Schmidt, Pamela Anning, and Kathleen Docherty

2. Vascular Plant and Vertebrate Inventory of Tumacácori National Historic Park 23Brian F. Powell, Eric W. Albrecht, Cecilia A. Schmidt, Pamela Anning, and Kathleen Docherty

3. Plant and Vertebrate Inventory of Organ Pipe Cactus National Monument 41Cecilia A. Schmidt and Brian F. Powell

4. Vascular Plant and Vertebrate Inventory of Saguaro National Park, Rincon Mountain 53District

Brian F. Powell and Cecilia A. Schmidt5. Vegetative Characteristics of Oak Savannas in the Southwestern Borderlands Region 71

Peter F. Ffolliott and Gerald J. Gottfried6. Distribution and Conservation Protection of Natural Land Cover in the Sonoran Desert 81

Ecoregion in Arizona, as Described by the Southwest Regional GAP Analysis ProjectKathryn A. Thomas, Keith A. Schulz, and Ken Boykin

7. Vertebrate Species in Desert Caves and Mines — A Comparison Between the 93Chihuahuan and Sonoran Deserts

Thomas R. Strong

STATUS OF RESOURCES8. Implications of Illegal Border Crossing and Drug Trafficking on the Management of 109

Public LandsCraig C. Billington, Randy Gimblett, and Paul R. Krausman

9. The U.S.-Mexico Border and Endangered Species 123Douglas K. Duncan, Erin Fernandez, and Curtis McCasland

vi

10. Adaptive Management of the Grassland-Watershed at Las Cienegas National 131Conservation Area: the Role of Monitoring, Rancher Engagement, and Multi-Stakeholder Advisory Teams

David Gori, Karen Simms, Mac Donaldson, Gitanjali Bodner, and Heather Schussman

11. Ecology and Conservation in the Sonoyta Valley, Arizona and Sonora 143Philip C. Rosen, Cristina Melendez, J. Daren Riedle, Ami C. Pate, and Erin Fernandez

12. Overview of the Lower Colorado River Multi-Species Conservation Program 161William E. Werner

13. Termite Activity on Green Tissues of Saguaro (Carnegiea gigantea) in the 171Sonoran Desert

Alejandro E. Castellanos, Reyna A. Castillo, Adrian Quijada14. Conservation of Amphibians and Reptiles in Northwestern Sonora and 181

Southwestern ArizonaJames C. Rorabaugh

15. Observations on the Status of Aquatic Turtles and the Occurrence of Ranid 205Frogs and Other Aquatic Vertebrates in Northwestern Mexico

Philip C. Rosen and Cristina Melendez16. Challenges to Natural Resource Monitoring in a Small Border Park: Terrestrial 225

Mammals at Coronado National Memorial, Cochise County, ArizonaDon E. Swann, Melanie Bucci, Amy J. Kuenzi, Barbara N. Alberti, and Cecil Schwalbe

17. The Small Mammal Community Associated with Ironwood (Olneya tesota) 241Helí Coronel-Arellano and Carlos A. López-González

18. Finding that 4-star Diner or How Bats Might “Anticipate” Productive Foraging 253Areas

Debbie C. Buecher, Ronnie Sidner, and John L. Koprowski 19. Advancing Large Carnivore Recovery in the American Southwest 263

Tony Povilitis and C. Dustin Becker20. Conservation Genetics of Black Bears in the Sky Islands of Arizona and 275

Northern MexicoCora Varas, Carlos A. López-González, Paul R. Krausman, and Melanie Culver

21. Recovery Efforts for the Sonoran Pronghorn in the United States 283Ryan R. Wilson, Paul R. Krausman, and John R. Morgart

22. Current Status of Mountain Lions and Urban Issues in Tucson, Arizona 293Kerry L. Nicholson, Lisa Haynes, and Paul R. Krausman

vii

SOCIAL AND CULTURAL ISSUES23. Modeling Airborne Mineral Dust: A Mexico – United States Trans-boundary Perspective 303

Dazhong Yin and William A. Sprigg24. Open Space Protection as a Means of Urban Containment: A Case Study from 319

ColoradoDavid Pesnichak

25. Historic Transportation Corridors in the Southwest, 1536 – Present 337Eric Vondy

List of Contributors 345Index 351

viii

DEDICATION — DR. DEAN A. MARTENS

Dean A. Martens was born in Fort Dodge, Iowa to Verona and Gern Martens on June 22, 1954.He received his B.S., M.S., and Ph.D in agronomy from Iowa State University. Dean wasinvolved extensively with soils and agricultural water contaminant problems in Californiabecoming a soil scientist at the National Soil Tilth Laboratory in Ames, Iowa. There he investi-gated erosion, soil stability, and carbon and nitrogen cycling dynamics in agricultural soils. Hewas transferred to Tucson, AZ in 2000 where he worked as a soil scientist with USDA-ARS atthe Southwest Watershed Research Center.

His research efforts were directed to investigations of land use changes in semiarid range-lands on the cycling of carbon and nitrogen and its impact on potential climate change. Theseland-use changes involved the impacts of man and vegetation changes on biogeochemicalchanges in riparian zones, grasslands, and mountainous woodlands. He was also an AssociateAdjunct Professor in the Department of Soil, Water, and Environmental Sciences at the Univer-sity of Arizona. He was an active and supporting member of the Committee that organized the2002 and 2004 Conferences on Research and Resource Management in the Southwestern Deserts.While the Conference Committee was preparing for its sixth conference, Dean passed away aftera brief battle with cancer on November 15, 2005.

Dr. Martens was known for his intense scientific curiosity and boundless energy regardingthe preservation of natural resources. He was a24-year member of both the American Societyof Agronomy and the Soil Science Society ofAmerica. At the time of his death, he was servingas the Division S-3 Associate Editor for the SoilScience Society of America Journal. His scien-tific legacy will live on. His publications havebeen cited in over 300 studies around the world.

He is survived by his wife Ursula anddaughter Sophie.

ix

DEDICATION — ERIC W. ALBRECHT

Eric Wells Albrecht was born September 6, l970 in Eugene, Oregon to Robert and LlewellynWells Albrecht. After graduating from Lane Community College, he moved to Seattle for twoyears before attending The Evergreen State College in Olympia, where he received a Bachelor’sdegree specializing in Natural History in 1992. He developed a particular interest in ornithology,volunteering on a project to band raptors in the Sawtooth Mountain range in Idaho. He alsoworked as a guide at Denali National Park in Alaska.

In 1998, Eric began his career in natural resources after moving to Tucson, Arizona. Therehe assisted his close friend and wildlife graduate student, Brian Powell, in his thesis research onsongbirds in the Sonoran Desert. Eric also assisted other graduate students with research onhawks and the interactions between vegetation, insects, and fire. In 2000, Eric began his owngraduate work at the University of Arizona where his research in the Huachuca Mountainsfocused on fire as a tool to restore populations of grassland birds. On September 20, 2004, atage 34, Eric died unexpectedly of an acute brain infection. He received his Masters of Sciencein Wildlife Conservation and Management posthumously.

From 2002 until his death, Eric was the co-coordinator of the University of Arizona Biolog-ical Inventory and Monitoring program. His work involved the development of monitoringprotocols for the National Park Service Units in the Sonoran Desert Network. His work aided

in the completion of 12 Open-file Reportsfor the U. S. Geological Survey’s SouthwestBiological Science Center.

He was devoted to his family andcommunity, active in his neighborhood asso-ciation in spearheading a community gardenproject. Eric was a passionate conserva-tionist and ecologist, and avid outdoorsmanand nature photographer. His scientificcuriosity was always with him, so much sothat he had a rule when attending presenta-tions: he always asked a question of thepresenter.

Eric is survived by his partner of eightyears, Kathy Moore, and their two childrenElizabeth and Zachary.


Introduction

INTRODUCTION

William L. Halvorson

Over the last few decades, management ofnatural and cultural resources has evolved frommanaging from a position of beliefs tomanaging from a basis of knowledge1. Thoseof us who had responsibilities in federal, state,local and private land management areas oncedid not have much of a response when asked,“How can you manage your land and ecosys-tems without knowing what’s there?”Overcoming the information gap has comeabout through programs like the National ParkService’s Inventory and Monitoring Program(http://science.nature.nps.gov/im/index.cfm)and added emphasis on research on the part ofresource managers, along with national andregional conferences directed at bringingresearchers, managers, and interpreterstogether for exchanging knowledge, ideas,questions, successes, and failures.This book brings together peer-reviewed

research from two interagency projects thatfocused on Southwestern Desert ecosystemsand resources. The first is a series of biennialconferences that were held from 1996 through2006. These were guided and managed by acommittee representing Arizona Game andFish Department, National Park Service, TheNature Conservancy, Sky Island Alliance,Sonoran Institute, Universidad de Sonora,University of Arizona School of NaturalResources, USDA Agricultural ResearchService, USDA Natural Resources Conserva-tion Service, U.S. Forest Service, and U.S.Geological Survey. The coordination of scien-tists within these federal and state agencies andnon-government organizations furthered theresearch and conservation of Southwestern

desert resources much more than could beaccomplished without this cooperation. Thesecond is an interagency project that took placefrom 2000 through 2008. This inventory of theNational Park Service units in southernArizona and eastern New Mexico was a coor-dinated effort of the National Park Service,U.S. Geological Survey and the University ofArizona School of Natural Resources. The Southwestern Deserts stretch from

southeastern California to west Texas and thensouth to central Mexico. The landscape of thisregion is known as basin and range topographywith many hills and mountains rising abovedesert valleys. Where there are well-devel-oped mountains, they are often referred to assky islands; forested islands in a sea of desert.This provides for a uniquely interesting andcomplex ecology. Another issue that thisregion deals with is the international borderthat stretches for hundreds of miles from thecoast of California, through Arizona and NewMexico, east to the Gulf Coast of Texas. Indealing with people moving across this border,governments have also caused difficulties formany animal populations. This book spot-lights individual research projects that increaseour understanding of forces acting on thebiological and cultural resources of this vastregion so that those resources can be managedas effectively and efficiently as possible. Ourintent is to show that collaborative effortsamong federal and state agency, university, andprivate sector researchers working with landmanagers provide better science and bettermanagement than scientists and land managersworking independently.

2 Halvorson

The first part of the book highlights studiesaimed at inventories to discover what speciesand communities are associated with specificland management units. It is surprising to mostthat complete inventories of plants, animals,and communities do not exist for most of thefederal and state land management units. Inmany, only the most obvious species are onany given area’s species list. Even for thoseNational Park Service units at which this majorproject was aimed, major groups like insectsare completely missing. Going beyond a listof species to detail community structure andfunction is likewise missing for many areas.The inventories accomplished for the NationalPark Service was the first step in a long-termmonitoring program. Once the inventorieswere completed and the NPS knew whatresources were found in each managementarea, the steps of describing a conceptualmodel, listing key abiotic and biotic factors inthat model, and noting key vital signs thatshould be monitored could take place. Basedon those steps and understanding the costsinvolved in long-term monitoring, a programwas developed for tracking changes andreporting the status of resources over time.This information can then be provided tomanagers for their decision-making about whatchanges need to be made.The second and largest segment of the book

reports on the status of biological resources ofthe Southwest Desert Region without regardto management areas. We leave it up to themanagers as to how this information can bestbe used within their area of concern. It isimportant to keep in mind that the diversity ofthis region is great because the basin and rangetopography supports ecosystems from hot,lowland deserts to montane coniferous forests.Elevations range from sea level to over 3,261meters (10,700 ft). Resources addressed in thissection range from termites to pronghorn, fromanimals living at sea level to those that inhabitthe tallest mountains. Issues dealt with includethe implications of human border crossers,adaptive management of desert grasslands,conservation plans for the lower ColoradoRiver, and how best to advance the recovery

of endangered species and other largemammals in the borderland desert regions.Mammals whose status is discussed includebats, small mammals in ironwood forests,black bear, pronghorn antelope, and mountainlions. There are chapters on conservation ofthe Sonoyta Valley, monitoring mammals atCoronado National Memorial, the status ofturtles and frogs in Mexico, conservation ofamphibians and reptiles, and termite activityon live saguaro cacti.The final chapters of the book address

social and cultural themes: a technique thattowns can use to protect open space, dustpatterns across the southwestern U.S. andnorthwestern Mexico, and historic transporta-tion corridors.Southwestern Desert Resources is for

researchers, resource managers, and landmanagers, as it shares the similarities inclimate and topography across the region,helps all to understand the status and distribu-tion of species, and discusses many of theissues that make management of resourcesdifficult. All of this information, when sharedbetween all the individuals and groupsworking in the region will increase our collec-tive understanding and help inform oureveryday management decisions. This volumecan serve as a quick guide, a stepping stone ifyou will, to the types of research and manage-ment projects that we need to undertake next2. We are doing a much better job than we did

even twenty years ago. As we look to thefuture we are beginning to understand that wewill need to be managing landscapes that inte-grate human communities with naturalprotected areas and that provide high qualitysustainable habitats for both humans andnatural plants and animals.A work of this nature always involves a

large number of people, the names that aredirectly seen in the book are the tip of theiceberg. The editors give our thanks and appre-ciation to all those for which this book is but asmall portion of all the work being done inmanaging the desert’s natural resources everyday. A special thanks goes to the conferencecommittee for putting on the conferences on

Halvorson 3

research and resource management in theSouthwestern deserts: Acasia Berry, AlejandroCastellanos, Nina Chambers, Doug Duncan,Peter Ffolliott, Brooke Gebow, Jerry Gottfried,David Hodges, Andy Hubbard, Sue Kozacek,Larry Laing, Dean Martens, Joan Scott, FrankToupal, and Dale Turner. We want to thank allof the natural area managers who gave theirsupport and time for the studies represented inthis volume, especially Brian Carey and AlanWhalon of Chiricahua National Monumentand Kathy Davis of Tuzigoot and MontezumaCastle National Monuments. We also appre-ciate the work of the cadre of reviewers whotook on the task so that every chapter in the

book had at least two peer reviews: MikeBarna, Kevin Bonine, T. J. Fontaine, AndyHubbard, Tom Jones, Jeff Lovich, Larry Laing,Bruce Nash, Larry Norris, Carrie Dennett,Mike Sredl, Eric Stitt, Marty Tuegel, andSandy Wolf.

1Halvorson, W.L. and G.E. Davis. 1996.Science and Ecosystem Management in theNational Parks. The University of ArizonaPress, Tucson

2No endorsement is implied by the mentionof commercial products in any of the book’schapters.

Area Inventories

VASCULAR PLANT AND VERTEBRATE INVENTORY OF CASA GRANDE RUINS NATIONAL MONUMENTBrian F. Powell, Eric W. Albrecht, Cecilia A. Schmidt, Pamela Anning, Kathleen Docherty

In 2001 and 2002 we surveyed for vascularplants and vertebrates (amphibians, reptiles,birds, and mammals) at Casa Grande RuinsNational Monument (NM) to document thepresence and in some cases relative abundanceof these species. This was the first compre-hensive biological inventory of the monument.By using repeatable study designs and stan-dardized field techniques, which includedquantified survey effort, we produced inven-tories that can serve as the basis for abiological monitoring program.

Of the National Park Service units in theregion, no other has experienced as muchrecent ecological change as Casa Grande RuinsNM. Once situated near a large and biologi-cally diverse mesquite bosque (forest)associated with the perennially flowing GilaRiver, the monument is now a patch of sparsedesert vegetation surrounded by agricultureand by urban and commercial developmentwhich is rapidly replacing the agriculturalfields as the dominant land use in the area.Roads, highways, and canals directly surroundthe monument. Development, and its associ-ated impacts, has important implications forthe plants and animals that live in the monu-ment. The plant species list is small and thedistribution and number of non-native plantsappears to be increasing. Terrestrial vertebratesare also being impacted by the changing land-scape, which is increasing the isolation of thesepopulations from nearby natural areas andthereby reducing the number of species at themonument. These observations are alarming

and are based on our review of previousstudies, our research in the monument, and ourknowledge of the biogeography and ecologyof the Sonoran Desert. Together, these datasuggest that the monument has lost a signifi-cant portion of its historic complement ofspecies and these changes will likely intensifyas urbanization continues.

Despite isolation of the monument fromnearby natural areas, we recorded noteworthyspecies or observations for all taxonomicgroups:• Plants: night-blooming cereus • Amphibians: high abundance of Couch’s

spadefoot toads• Reptiles: high abundance of long-nosed

snakes• Birds: 10 species of diurnal raptors

including 4 species of falcons • Mammals: American badger

This study was a first step in the process ofcompiling information about the biologicalresources of Casa Grande Ruins NM andsurrounding areas. For complete details of theCasa Grande Ruins NM study see Powell et al.(2006) [http://sbsc.wr.usgs.gov/products/ ofr/].Scientific and common names used throughoutthis chapter are current according to acceptedauthorities for each taxonomic group: Inte-grated Taxonomic Information System (ITIS2001) and the PLANTS Database (USDA2001) for plants, Stebbins (2003) for amphib-

8 Powell, Albrecht, Schmidt, Anning, and Docherty

ians and reptiles; American OrnithologistUnion (AOU; 1998, 2003) for birds; and Bakeret al. (2003) for mammals.

MONUMENT OVERVIEWMonument Area and History

Casa Grande Ruins NM is located in Coolidge,Arizona, approximately 70 km southeast ofPhoenix (Figure 1). The monument currentlyencompasses 191 contiguous ha and managersare proposing to increase the size of the monu-ment by approximately 105 ha (including32 ha adjacent to the current site; NPS 2003a).

Casa Grande Ruins NM was created toprotect the Casa Grande, a four-story adobestructure that was built by the Hohokambetween AD 1200 and 1450 (Clemensen1992). The Hohokam had a sophisticatedculture—they built extensive canals to irrigatecrops and provide water to large communitiesin the vicinity of the Casa Grande. After themysterious departure of the Hohokam inapproximately 1450, the Casa Grande stoodabandoned for over 440 years until, in 1892,the structure and the land surrounding itbecame the first U.S. prehistoric cultural siteto receive federal protection (Clemensen1992). In 1918, Casa Grande Ruins becamepart of the National Park Service system.

Physiography, Geology, and SoilsThe monument is located approximately 1 kmsouth of the Gila River, which now only flowsseasonally. The Pima Lateral canal runsparallel to (and a few meters from) thesouthern boundary of the monument and asmaller irrigation ditch parallels the westboundary (Figure 2). State Highway 87 runsalong the east and north boundaries of themonument.

The monument is situated at approximately430 m above sea level in the Basin and RangePhysiographic Province, which is character-ized by gently sloping valley floors surroundedby mountain ranges. The monument is char-acterized by Quaternary and Tertiary alluvialdeposits (fluvial and lacustrine) from thesurrounding mountain ranges: San Tan Moun-

tains (6 km north), Sacaton (16 km west),Picacho (30 km southeast), and Casa Grande(30 km southwest). The mountains borderingthe valley floor are composed of non-water-bearing Precambrian granite, gneiss, and schist(Van Pelt 1998). All mountain ranges arecurrently isolated from each other by agricul-ture and development. Soil at the monument isCoolidge sandy loam, with caliche two to fourfeet below the surface.

HydrologyThe Gila River is the main water body in the

region, but impoundments upstream from themonument cause the river bed to be dry formost of the year in the reach to the north of themonument. Irrigation canals carry water forcrops, while water for developments comesfrom groundwater pumping (Sprouse et al.2002).

ClimateCasa Grande Ruins NM is located in thesubtropical desert climatic zone of southernArizona which is characterized by heavysummer (monsoon) storms brought about bymoisture coming from the Gulf of Mexico andless intense, frontal storms from the PacificOcean in the winter. The monument receivesan average of 228 mm of precipitation annu-ally. Summers in the area are hot; dailymaximum temperatures from June throughSeptember often exceed 40° C. Winters aremild and temperatures rarely drop belowfreezing (WRCC 2004).

Average annual precipitation totals duringthe course of our study were slightly above thelong-term mean of 228 mm in 2001 (247 mm)but considerably lower than average in 2002(122 mm), one of the driest years on record(WRRC 2004). In the fall of 2000 rainfall wasabove average; this rain may have increasedwinter annual plant seed germination andgrowth prior to our 2001 spring plant surveys.Average annual temperatures during both yearsof our study were 0.9° C above the long-termmean of 20.8° C.

Powell, Albrecht, Schmidt, Anning, and Docherty 9

Figure 1. Location of Casa Grande Ruins National Monument, Arizona.


Figure 2. Aerial photograph of Casa Grande Ruins NM showing it in a patchwork ofcommercial and residential development and agricultural fields (A) and a more detailedimage of the monument’s major features (B).


VegetationThe relatively homogenous vegetationcommunity at Casa Grande Ruins NM is char-acterized as Sonoran desertscrub dominated bycreosote with scattered velvet mesquite, salt-bush and annual grasses and forbs (Reichhardt1992). Shrubs and trees in the vicinity of thevisitor center are irrigated, and the manystanding dead velvet mesquite trees in otherareas of the monument reference a changefrom historic conditions.

Because the area in and around the monu-ment has been intensely used for hundreds ofyears, it is difficult to determine the “natural”vegetation community of the area. Given themonument’s close proximity to the Gila River,coupled with the topographic and soil condi-tions of the site, it is likely that the general areawas once more fully covered with large areasof mesquite, especially before colonization bythe Hohokam. Even since the abandonment ofthe area by the Hohokam, many large mesquitetrees dominated the area, as noted by late 19thcentury visitors (Clemensen 1992). Subsequentcattle grazing probably enabled the increase inwoody shrubs such as creosote, catclaw acacia,and saltbush. However, in the mid to late1930s, the large mesquites at the monumentbegan to die off, apparently due to a drop inthe groundwater related to pumping for agri-cultural irrigation (Judd et al. 1971; Clemensen1992; Nickens 1996; Van Pelt 1998). Thelowering of the water table also likely changedthe soil conditions enough that a shift tookplace from salt bush and catclaw acacia to aproliferation of creosote, which now dominatesat the monument.

Historical Land Use of the Monumentand Surrounding Areas

Clemensen (1992) compiled a detailed historyof Casa Grande Ruins NM and the followinginformation comes from his work. Beginningin the 1870s, settlers began grazing cattle inthe area because of the abundant forage.Grazing continued until 1934 when the monu-ment was fenced to exclude cattle. But it was

agriculture that would become the dominantland use of the area outside the monument, andbeginning in the 1880s settlers began in earnestto clear land in the vicinity of the monument.Water for irrigating crops (fruit trees, grapes,cereal grains, cotton, lettuce, and alfalfa) camefirst from direct diversion of flow from theGila River, and later from above-groundstorage with the construction of nearbyCoolidge Dam in the mid 1920s and ground-water accessed by pumping. In 1925 the townof Coolidge was created and by 1932 themonument was surrounded by agriculturalfields. However, by 1947, agricultural fieldswere being abandoned because of drought anda lowered water table, due in large part topumping of water in excess of recharge rates.Depth-to-water rebounded somewhat by thelate 1990s, in part because of reduced ground-water pumping (Sprouse et al. 2002).

Natural Resource Management IssuesCasa Grande Ruins NM is an isolated patch ofdesert vegetation surrounded by intensivelyaltered land; uses include agriculture, residen-tial and commercial development, and roads.Although it is difficult to quantify the effect ofthese land uses, these (and other) influencesinevitably affect the structure and compositionof plant and animal communities of the monu-ment.

AgricultureAgricultural fields bordering the monument tothe west and north are typical of the dominantland use in the surrounding area. These areasprovide disturbed soils and only marginalspace for other plants to grow, space that istypically occupied by non-native “weedy”plants including redstem stork’s bill, redbrome, Russian thistle, and Johnsongrass. Inaddition, the canals that border the monumentare periodically dredged and the sediment(likely rich in non-native plant seed) isdeposited along the edge of the monumentboundary (Hubbard et al. 2003).


Residential and Commercial DevelopmentCasa Grande Ruins NM is located within theCity of Coolidge. The city’s population (7,786inhabitants in 2000) is increasing rapidly,leading to increased residential and commer-cial development (NPS 2003a, 2003b).Large-scale commercial developments (e.g.,Wal-Mart®) have been built along Highway 87across from the monument. Residential devel-opment abuts the south boundary and isplanned along the west boundary should theproposed monument boundary expansion notbe approved. Impacts of these developmentson the monument’s natural resources mayinclude: (1) an increase in non-native plants,for example the first sighting of a commonplant used in landscaping, crimson fountain-grass, was reported by Halvorson and Guertin(2003); (2) increased trash and run-off of sedi-ment and toxins from vehicles; (3) disruptionof animal movement patterns; and (4)increased harassment and mortality of nativeanimals by free-roaming feral pets (Clarke andPacin 2002).

RoadsCasa Grande Ruins NM is completely encir-cled by roads, most notably Highway 87(Figure 2), the primary highway in the area.Roads act as dispersal corridors for non-nativeplant species, which often thrive in the adja-cent disturbed soils. Roads surrounding themonument likely act as barriers to the flow ofterrestrial wildlife because of direct mortalityand modification of behavior (Trombulak andFrissell 2000; Clark et al. 2001; Tigas et al.2002; Cain et al. 2003).

Groundwater PumpingThe continued pumping of groundwater foragricultural, residential, and commercial usemay threaten existing mesquites on the monu-ment despite the recent (and likely temporary)rise in the level of the groundwater (Sprouseet al. 2002). Groundwater pumping can alsolead to subsidence that threatens the CasaGrande structure (NPS 1998; Richardson2002; Hubbard et al. 2003).

Non-native and Pest SpeciesAwareness of non-native species as a manage-ment issue has increased in recent years;ecologists have ranked this issue as one of themost significant causes of species endanger-ment (Brooks and Pyke 2001). Non-nativeplant species are a significant managementissue at the monument because it is surroundedby roads, agricultural fields, and development,which generally provide ideal conditions forthe dispersal and establishment of some non-native plants. Non-native plants are known toalter ecosystem function and processes(Naeem et al. 1996; D’Antonio and Vitousek1992) and reduce abundance of native species,creating potentially permanent changes inspecies diversity and community composition(Bock et al. 1986; D’Antonio and Vitousek1992; OTA 1993). The Casa Grande and asso-ciated structures provide habitat for manynon-native birds such as house sparrow andEuropean starling, and the adjacent develop-ments provide a source for free-roaming andferal cats and dogs.

In its Integrated Pest Management Plan(IPM; NPS 1997), monument personnel iden-tified a number of wildlife species that arecausing significant damage to the archaeolog-ical ruins in the monument. The IPM plan,along with that by Swann et al. (1994) identi-fied round-tailed ground squirrel, rock pigeon,and house finch as the most important pestspecies.

METHODSPlants

Previous inventoriesThe earliest collecting effort at the monumentwas from 1939 to 1942 when Natt Dodge, theregional naturalist, and Francis Elmore, a parkranger, collected plants from throughout themonument. These specimens (43 species) are atthe University of Arizona Herbarium. Reichhardt(1992) conducted an inventory of plants at themonument in 1987. This work included a list ofplants that she collected, classification of vege-tation communities in the monument, creation ofa checklist of non-ornamental plants, establish-


ment of vegetation plots, mapping of mesquitetrees, and establishment of photo points for usein describing qualitative changes in the vegeta-tion community. Halvorson and Guertin (2003)mapped the distribution of select non-native plantspecies in the monument from the fall of 1999 tothe spring of 2001. Collections of plants from themonument made by additional observers havebeen accessioned to the herbarium at the Univer-sity of Arizona and to the Western ArchaeologicalConservation Center in Tucson. The excellentwork that preceded our effort reduced the fieldwork required for the inventory. Below-averagemonsoon rains in 2002 further limited our effortsbecause most of the species that we hoped torecord are annuals that germinate following rains.

This inventoryIn March 2001 and again in September 2004we conducted 6 person-day “generalbotanizing” surveys at the monument, duringwhich observers walked throughout the monu-ment and opportunistically collected andrecorded plants. In addition to our own results,we present here the first synthesis of findingsfrom past studies and collections. Forsimplicity, we refer to all subspecies and vari-eties (n = 5) as species.

Amphibians and ReptilesPrevious inventoriesTo our knowledge, there has been no inventoryand there is scant research related to amphib-ians and reptiles at Casa Grande Ruins NM,though we located three specimens collectedfrom the monument and know of severalothers collected in the area or region. CharlesConner, a biologist at Organ Pipe CactusNational Monument, has surveyed diurnallizard populations at the monument for severalyears, but to date only a species list has beenproduced.

This inventoryWe surveyed for herpetofauna using fourmethods representing plot-based and moreflexible non-plot-based methods. Plot-based

methods are constrained by time and area, andthus provide data for estimates of relativeabundance that should be unbiased by thesefactors. Random location of these surveys alsoallows inference out to the current monumentboundaries.

Non-plot-based surveys (Crump and Scott1964) allow observers more flexibility inadjusting their search time, intensity, and loca-tion, and this flexibility is important fordetecting rare, elusive, or ephemeral speciesmore likely to be missed using plot-basedsurveys. We used both diurnal and nocturnalsurveys in an effort to detect species withrestricted periods of activity (Ivanyi et al. 2000,Stebbins 2003). We also used pitfall traps(Corn 1994; Gibbons and Semlitsch 1981),road surveys (Rosen and Lowe 1994), coverboards (Fellers and Drost 1994), and incidentalobservations to add species that were other-wise difficult to observe because they are veryrare, have limited periods of activity, or incon-spicuous behavior (Powell et al. 2006).

BirdsPrevious inventoriesTo our knowledge, scant bird research hastaken place at Casa Grande Ruins NM since alimited-scope banding study in the 1930s (Fast1936). Barry (1987) created a checklist for themonument, but no source material exists andso we do not consider it here. There are twoBreeding Bird Survey routes located approxi-mately 5 and 10 km west of the monument(Sauer et al. 2004): “Cactus Forest” transectwas surveyed in 1991, 1993, and 1996–2002;“Coolidge” was surveyed from 1974 to 1985.We found no records of specimens collectedfrom the monument.

This inventoryWe used four field methods: variable circular-plot counts for diurnal breeding birds(Reynolds et al. 1980; Ralph et al. 1995; Buck-land et al. 2001), nocturnal surveys for owlsand nightjars (Colver et al. 1999; Bibby et al.2002), line transects for over-wintering birds(Bibby et al. 2002), and incidental observations


for all birds in all seasons. We concentrated ourprimary survey effort during the breedingseason because bird distribution is relativelyuniform at this time (Bibby et al. 2002). Oursurvey period included peak spring migrationtimes for most species, which added manymigratory species to our list.

We also sampled vegetation around vari-able circular-plot (VCP) survey stations.Vegetation structure and plant species compo-sition are important predictors of bird speciesrichness or the presence of particular species(MacArthur and MacArthur 1961; Rice et al.1984; Strong and Bock 1990; Powell andSteidl 2000). We visited all 12 VCP stationsfour times each in 2001. In 2002 we reducedthe number of stations to 8 (station numbers 1,2, and 6–11) and surveyed each of them fourtimes. Each station was visited for 8 minutes.

We used a modified line-transect method(Bibby et al. 2002) to survey for birds fromOctober to December 2002. Line transectsdiffer from station transects (such as those usedin our VCP surveys) in that an observer recordsbirds seen or heard while the observer walks aline, rather than stands at a series of stations.The transect method is more effective duringthe non-breeding season because bird vocal-izations are less conspicuous and frequent, andtherefore birds tend to be more difficult todetect (Bibby et al. 2002). We established onetransect at the monument, broken into 12sections. Each section was approximately 250m in length. We visited all 12 sections fourtimes in 2002: 24 October, 8 and 25November, and 19 December. The total timespent on each section was 10 minutes.

To survey for owls we broadcastedcommercially available vocalizations (Colveret al. 1999) using a compact disc player andbroadcaster (Bibby et al. 2002), and recordedother nocturnal species (nighthawks and poor-wills) when observed. We established onenocturnal survey transect that bisected themonument along the main entrance road. Thetransect had four stations that were a minimumof 300 m apart. As with other survey methods,we varied observers and direction of travel

along transects and did not survey duringperiods of excessive rain or wind to reducebias. We began surveys approximately 45minutes after sunset. We visited each of thefour nocturnal survey stations three times eachin 2001 and twice each in 2002.

When we were not conducting formalsurveys and encountered a rare species, aspecies in an unusual location, or an individualengaged in a breeding behavior, we recordedUTM coordinates, time of detection, and (ifknown) the sex and age class of the bird. Werecorded all breeding behavior observationsusing the standardized classification system(developed by the North American Ornitho-logical Atlas Committee; NAOAC 1990). Thissystem classifies breeding behavior into oneof nine categories: adult carrying nesting mate-rial, nest building, adult performing distractiondisplay, used nest, fledged young, occupiednest, adult carrying food, adult feeding young,or adult carrying a fecal sac. We madebreeding observations during both standard-ized surveys and incidental observations.

MammalsPrevious inventoriesWe know of no earlier inventory work done formammals. Only two research projects havebeen conducted at the monument, one onground squirrels (Koprowski and Monroe2003) and one on damage to cultural resources(Swann et al. 1994). We located only threemammal specimens previously collected fromthe monument.

This inventoryWe surveyed for mammals using three fieldmethods: live-trapping for rodents and groundsquirrels (primarily nocturnal; herein referredto collectively as small mammals), infrared-triggered photography for medium and largemammals, and incidental observations for allmammals. We also located three mammalspecimens collected from the monument. Withno standing water available, we did not net forbats as it would likely not have been produc-tive.


We selectively placed 8 small-mammaltrapping sites in areas of the monument that wefelt represented slight variations in vegetationcommunity and structure. We avoided thevicinity of the picnic grounds because of thehigh density of round-tailed ground squirrelsin that area and we prioritized the likelihoodof documenting additional species in otherareas. We used infrared-triggered cameras(Trailmaster) to record the presence of mediumand large mammals. Trailmaster cameras havebeen proven to be the most cost-effectivemethod for recording the presence of mediumand large mammal species (Kucera and Barrett1993; Cutler and Swann 1999; Swann et al.2004). We placed cameras in two areas of themonument that we thought would record thehighest number of species; typically these werein areas of dense vegetation. We baited camerasites with a commercial scent lure or cannedcat food.

We recorded UTM coordinates of inci-dental observations made by any of theinventory crew members or the monumentstaff. Finally, we repeatedly checked the CasaGrande and its roof structure for bats.

RESULTS AND DISCUSSIONPlants

We recorded 60 species during our study,including 21 species that had not been previ-ously documented in the monument.Combining data from all studies (Reichhardt1992; Halvorson and Guertin 2003), includingour own and from relevant records in thecollections of two herbaria (University ofArizona and Western Archaeological Conser-vation Center), there have been 127 species ofplants recorded on or adjacent to the monu-ment. This number includes cultivated trees,but not cultivated shrubs and succulents (e.g.,ocotillo) around the visitor center (see Reich-hardt 1992 for an explanation). There havebeen 31 species of non-native plants observedor documented at the monument (24% of totalflora), and of these, nearly 40% (n = 12) aregrasses. The combined results of our inventoryeffort and the surveys of Halvorson and

Guertin (2003) recorded 37 previouslyunrecorded species, 12 of which are non-native). These additional species may indicatea change in the plant community that appearsto have occurred over the last 15 years. Indeed,of the 22 non-native species that Halvorsonand Guertin mapped, 16 were found onlyalong roads and/or the irrigation canal justoutside the boundary, and an additional threespecies were found primarily along the monu-ment’s roads. New species were detectedthroughout their study. There were 52 species,including six non-natives, found by priorstudies but not by our crews or by Halvorsonand Guertin, further suggesting a shift in vege-tation composition and increased non-nativeoccurrence during the last 60 years. The list ofplants that have not been found since 1942includes three species of shrubs (Alkali gold-enbush, fairyduster, and eastern Mojavebuckwheat).

We believe that the combined effort of ourstudy and previous studies and collectionshave recorded virtually all of the perennialplant species that occur at Casa Grande RuinsNM (excluding ornamentals around the visitorcenter). The list of annuals, however, is incom-plete, due in part to the increasing number ofnon-native plants that are becoming estab-lished in the monument (Halvorson andGuertin 2003). Each study at the monument,including ours, has recorded from 11 to 21species that were not reported by any otherefforts. Because most of the new species forthe monument are annual forbs and grasses,these numbers highlight the importance ofsurveying following periods of above- normalprecipitation (as we did in 2001) and to surveyrepeatedly.

Amphibians and ReptilesWe recorded three amphibian and 11 reptilespecies at Casa Grande Ruins NM. Commonside-blotched and western whiptail lizard werethe two most abundant species and togetherthey represented > 75% of all detections acrossall survey methods. We added one new species(Great Plains toad) to the monument list with


the road survey method; in fact we observedall three of the amphibian species recorded byour inventory during one night of road surveyin 2002. Although the pitfall trap did notcontribute additional species to our monumentlist, results from the trap were consistent withother methods and suggest that common side-blotched and western whiptails are among themost common lizards at the monument. Inci-dental detections did not add any species to ourlists, but this method did add records forspecies that were seldom detected by othermethods, notably Couch’s spadefoot, coach-whip, and common kingsnake. We found noanimals underneath coverboards.

It is seems unlikely that we missed severalconspicuous species that we would expect tofind at the monument: zebra-tailed lizard,desert iguana, long-nosed leopard lizard, andsidewinder. Species likely present in the monu-ment that we did not detect include snakes thatare nocturnal and inconspicuous such as:western blind snake, spotted leaf-nosed snake,saddled leaf-nosed snake, glossy snake,western ground snake, western shovel-nosedsnake, and night snake (Stebbins 2003). Webelieve that our inventory detected fewer than90% of the amphibians and reptiles presentsimply because the list has so few names on itthat even adding one is a significantpercentage.

The amphibian and reptile communities atCasa Grande Ruins NM comprise relatively

few species in comparison to what was likelypresent historically or in comparison to whathas been documented in the course of otherrecent herpetofauna inventories in manage-ment units of similar size in southern Arizona(Rosen and Mauz 2001; Powell et al. 2002;Powell et al. 2003; Powell et al. 2006). Thislow species richness likely results from theland uses in the vicinity of the monument anddegradation of the nearby Gila River (i.e., lossof aquatic and riparian resources; McNamee1994; Ingram 2000).

BirdsWe recorded 82 bird species during the twoyears of the study. Seventy-one percent (n =58) of the species that we observed wereneotropical migrants. We observed a numberof species of conservation concern: loggerheadshrike, burrowing owl, peregrine falcon, andferruginous hawk, all of which are considered“Species of Concern” by the U.S. Fish andWildlife Service. Of the 82 bird species on themonument’s list, only three (3.7%) are non-native: rock pigeon, European starling, andhouse sparrow.

We recorded 63 species during VCPsurveys at the monument. The mourning dove,Gambel’s quail, and house sparrow were themost abundant species during this portion ofthe study. We observed 32 species during foursurveys of line transects in the fall of 2002. Themourning dove and great-tailed grackle were

Table 1. Summary results of the vascular plant and vertebrate inventories at Casa Grande Ruins NM,2001 and 2002.

Number of Number of Number of new species Taxonomic group species recorded non-native species added to monument lista

Plants 60 12 21Amphibians and reptiles 14 0 13Birds 82 3 70Mammals 13 2 7Totals 169 17 111

a Species that had not been observed or documented by previous studies.


the most frequently detected species. Werecorded four species during nocturnalsurveys: lesser nighthawk, great horned owl,burrowing owl, and barn owl. Thirty-sixspecies were recorded outside of formalsurveys, including eight observations forspecies that we did not find with any othersurvey type.

All of the most abundant species at CasaGrande Ruins NM are considered human-adapted generalist species in southern Arizona.These species reach high densities in human-dominated landscapes (Mills et al. 1989;Germaine et al. 1998): mourning dove, rockpigeon, Gila woodpecker, cliff swallow, Euro-pean starling, red-winged blackbird,great-tailed grackle, house finch, and housesparrow. Rock pigeon, house finch, and housesparrow regularly use the Casa Grande struc-ture and therefore cause damage throughroosting and nesting (Swann et al. 1994; NPS1997). Based on our complete coverage of themonument for two breeding seasons, webelieve that we recorded all of the species thatpermanently resided or bred in the monument.However, a species-accumulation curvesuggests that we would record additionalmigrant species with further effort.

MammalsThe current list of mammals for Casa GrandeNM consists of 13 species, 7 small mammalsand 6 medium-sized mammals; 2 species(domestic dog and cat) are non-native. Thisstudy added 7 species to the monument’s offi-cial list. Using the live-traps, we recorded allseven species of small mammals. Merriam’skangaroo rat and the Sonoran Desert pocketmouse were found to be the most widespreadand abundant small mammals.

With the Trailmaster cameras we recorded3 species. The most photographed species wasthe cottontail. These animals were most likelydesert cottontails; eastern cottontail is difficultto differentiate in photographs, but is unlikelyto be present in the area near the monument(Hoffmeister 1986). Other mammal speciesdetected with cameras were black-tailed

jackrabbits and the western white-throatedwoodrat.

University of Arizona personnel made inci-dental observations of 5 species during thecourse of the study. Monument personnelreported regular observations of feral cats. Nobats were observed in any part of the monu-ment. We collected two skulls during thecourse of the study, a domestic cat and anAmerican badger.

The majority of our mammal survey efforttargeted small mammals. Based on a speciesaccumulation curve, it appears that we trappedmost of the species that occurred on the monu-ment during the time of the study. A numberof species could be present or may historicallyhave been present at the monument, based onrange maps and published habitat associations,including: little pocket mouse, Bailey’s pocketmouse, cactus mouse, Arizona cotton rat, andthe non-native house mouse (Hoffmeister1986). It is quite likely that these species,particularly cactus mouse and house mouse,would be captured with additional surveyeffort. Also, there are a few species that arewithin range but would require higher densityof vegetation (particularly dense grasses andforbs): Botta’s pocket gopher, silky pocketmouse, banner-tailed kangaroo rat, westernharvest mouse, and hispid cotton rat(Hoffmeister 1986). Based on the descriptionof the vegetation at the monument prior tocattle grazing and mesquite die-off (Clemensen1992), it is likely that these species were oncecommon residents of the monument. Also,Merriam’s mouse, probably once common atthe monument before the die-off of the largemesquite forest, is very restricted to that vege-tation component and therefore unlikely to bepresent now.

We found no large mammals (e.g., moun-tain lion, deer, or bear) during our surveys andno large mammals have been reported fordecades (Clemensen 1992; CGRNM Staff,pers. comm.). Clemensen (1992) also reportsthat fox and bobcat have been seen at themonument. We observed or documented somemedium-size mammals such as badger (skull),


striped skunk (sighting), and feral cats anddogs. Coyotes have been reported in the monu-ment both historically (Clemensen 1992) andrecently (CGRNM Staff, pers. comm.).

One particularly striking change in themammal community has been the use of theCasa Grande by Brazilian (“Mexican”) free-tailed bats. In 1944, monument personnelcounted over 5,000 bats exiting the ruins, butby 1956 bats no longer lived on the monument(Clemensen 1992). This time period coincidedwith the increased use of insecticides,including DDT. More recently, Swann et al.(1994) did not find any bats or sign of bats ontheir inspection of the Casa Grande in the latesummer and fall of 1993 nor did we find anybats during our surveys. Although free-tailedbats have a large foraging range (Best andGeluso 2003), the combination of pesticideuse, subsequent lack of insects in areas adja-cent to the monument, and regional-scalepopulation changes all work to keep the batsfrom returning.

ACKNOWLEDGMENTSThanks to Superintendents Paige Baker andDonald Spencer and their staff at Casa GrandeRuins National Monument (NM) for theirsupport. This project resulted from the collab-oration of many people at the University ofArizona (UA), National Park Service, and U.S.Geological Survey. Larry Norris, NPSResearch Coordinator, spent considerable timeand effort providing clear and timely adminis-trative assistance. Andy Hubbard, NetworkCoordinator of the NPS Sonoran DesertNetwork Inventory and Monitoring program,was a strong advocate for the project.

We thank a core group of dedicated fieldbiologists who collected a wealth of data atCasa Grande Ruins NM: Theresa Dekoker,James MacAdam, and Meg Quinn (plants);Dan Bell, Kevin Bonine, James Borgmeyer,Charles Conner, Dave Prival, Angela Urbon,and Mike Wall (amphibians and reptiles);Gavin Bieber, Chris Kirkpatrick, and GabeMartinez (birds); Neil Perry, Jason Schmidt,and Ronnie Sidner (mammals). We are appre-

ciative of the following people, many of whomnever ventured into the field, but whose workin the office made the field effort moresuccessful: Debbie Angell, Jennifer Brodsky,Valery Catt, Brian Cornelius, Taylor Edwards,Jenny Ferry, Carianne Funicelli, AndyHonaman, Colleen McClain, HeatherMcClaren, Lindsay Norpel, Terri Rice, JillRubio, Brent Sigafus, Taffy Sterpka, PatinaThompson, Jenny Treiber, Cecily Westphal,and Alesha Williams.

Special thanks to Lisa Carder for her yearsof hard work on all aspects of the project.Technical support was graciously given by thefollowing experts: Dan Austin, Michael Cham-berland, Phil Jenkins, and Charlotte and JohnReeder at the UA Herbarium; George Bradleyof the UA herpetology collection; Tom Huelsof the UA ornithology collection; and YarPetryszyn and Melanie Bucci of the UAmammal collection. Review and editingsuggestions were offered by Debbie Angell,Brooke Gebow, Shirley Hoh, Andy Hubbard,Theresa Mau-Crimmins, Phil Rosen, DonSwann, and Carol West.

LITERATURE CITEDAmerican Ornithologists’ Union (AOU). 1998.

Checklist of North American birds, seventhedition. American Ornithologists’ Union andAllen Press Inc., Lawrence, KS.

American Ornithologists’ Union (AOU). 2003.Forty-second supplement to the AmericanOrnithologists’ Union checklist of NorthAmerican birds. Auk 117: 847–858.

Baker, R. J., L. C. Bradley, R. D. Bradley, J. W.Dragoo, M. D. Engstrom, R. S. Hoffmann, C. A.Jones, F. Reid, D. W. Rice, and C. Jones. 2003.Revised checklist of North American mammalsnorth of Mexico, 2003. Occasional Papers of theMuseum of Texas Tech University 229: 1–23.

Barry, P. 1987. Bird checklist of Casa Grande RuinsNational Monument and vicinity. SouthwestParks and Monument Association and UnitedStates Department of the Interior. List found onthe Northern Prairie Wildlife Research Centerhome page: http://www.npwrc.usgs.gov/resource/othrdata/chekbird/r2/casa.htm.

Best, T. L., and K. N. Geluso. 2003. Summerforaging range of Mexican free-tailed bats(Tadarida brasiliensis mexicana) from CarlsbadCavern, New Mexico. Southwestern Naturalist48: 590–596.

Bibby, C. J., N. D. Burgess, and D. A. Hill. 2002.


Bird census techniques. 2nd Edition. AcademicPress, London, England.

Bock, C. E., J. H. Bock, K. L. Jepson, and J.C.Ortega. 1986. Ecological effects of plantingAfrican lovegrasses in Arizona. NationalGeographic Research 2: 456–463.

Brooks, M. L., and D. A. Pyke. 2001. Invasiveplants and fire in the deserts of North America.Tall Timbers Research Station MiscellaneousPublication 11: 1–14.

Buckland, S. T., D. R. Anderson, K. P. Burnham, J.L. Laake, D. L. Borchers and L. Thomas. 2001.Introduction to distance sampling: Estimatingabundance of biological populations. OxfordUniversity Press, London, England.

Cain, A. T., V. R. Touvila, D. G. Hewitt, and M. E.Tewes. 2003. Effects of a highway and mitigationprojects on bobcats in southern Texas. BiologicalConservation 114: 189–197.

Clark, B. K., B. S. Clark, L. A. Johnson, and M. T.Haynie. 2001. Influence of roads on movementsof small mammals. Southwestern Naturalist 46:338–344.

Clarke A. L., and T. Pacin. 2002. Domestic cat“colonies” in natural areas: A growing exoticspecies threat. Natural Areas Journal 22: 154–159.

Clemensen, A. B. 1992. A centennial history of thefirst prehistoric reserve: 1892–1992. United StatesDepartment of the Interior. National Park Service,Washington D.C.

Colver, K. J., D. Stokes, and L. Stokes. 1999.Stokes field guide to bird songs. Time WarnerTrade, New York, NY.

Corn, S. P. 1994. Straight-line drift fences andpitfall traps. In Measuring and monitoringbiodiversity: Standard methods for amphibians,edited by W. R. Heyer, M. A. Donnelly, R. W.McDiarmid, L. C. Hayek, and M. S. Foster, pp.109–117. Smithsonian Institution Press,Washington, D.C.

Crump, M. L., and N. J. Scott. 1994. Visualencounter surveys. In Measuring and monitoringbiodiversity: Standard methods for amphibians,edited by W. R. Heyer, M. A. Donnelly, R. W.McDiarmid, L. C. Hayek, and M. S. Foster, pp.84–92. Smithsonian Institution Press,Washington, D.C.

Cutler, T. L., and D. E. Swann. 1999. Using remotephotography in wildlife ecology: A review.Wildlife Society Bulletin 27: 571–581.

D’Antonio, C. M., and P. M. Vitousek. 1992.Biological invasions by exotic grasses, the grassfire cycle, and global change. Annual Review ofEcology and Systematics 23: 63–87.

Fast, J. E. 1936. Bird banding notes. SouthwesternMonuments. Supplement for February.

Fellers, G. M., and C. A. Drost. 1994. Samplingwith artificial cover. In Measuring and monitoringbiodiversity: Standard methods for amphibians,edited by W. R. Heyer, M. A. Donnelly, R. W.McDiarmid, L. C. Hayek, and M. S. Foster, pp.

146–150. Smithsonian Institution Press,Washington, D.C.

Germaine, S. S., S. S. Rosenstock, R. E.Schweinsburg, and W. S. Richardson. 1998.Relationships among breeding birds, habitat, andresidential development in greater Tucson,Arizona. Ecological Applications 8: 680–691.

Gibbons, J. M., and R. D. Semlitsch. 1981.Terrestrial drift fence with pitfall traps: Aneffective technique for quantitative sampling ofanimal populations. Brimleyana 7: 1–6.

Green, D. M., and M. G. Baker. 2003. Urbanizationimpacts on habitat and bird communities in aSonoran desert ecosystem. Landscape and UrbanPlanning 63: 22–239.

Halvorson, W. L., and P. Guertin. 2003. USGSWeeds in the West project: Status of introducedplants in southern Arizona Parks. U.S. GeologicalSurvey, Southwest Biological Science Center,Sonoran Desert Research Station, School ofNatural Resources, University of Arizona,Tucson, AZ.

Hoffmeister, D. F. 1986. Mammals of Arizona.University of Arizona Press, Tucson, AZ.

Hubbard, J. A., T. M. Mau-Crimmins, B. F. Powell,E. W. Albrecht, N. Chambers, and L. Carder.2003. National Park Service Sonoran DesertNetwork monitoring plan: Phase II. SonoranDesert Network, Tucson, AZ.

Ingram, M. 2000. Reptile and amphibian accounts.In A natural history of the Sonoran Desert, editedby S. J. Phillips and P. W. Comus, pp. 41–50.Arizona-Sonora Desert Museum Press, Tucson.

Integrated Taxonomic Information System (ITIS).2001. Accessed on 20 November 2001.http://www.itis.usda.gov/index.html.

Ivanyi, C., J. Perry, T. R. Van Devender, and H.Lawler. 2000. Reptile and amphibian accounts. InA natural history of the Sonoran Desert, edited byS. J. Phillips and P. W. Comus, pp. 533–585.Arizona-Sonora Desert Museum Press, Tucson.

Judd, B. I., J. M. Laughlin, H. R. Guenther and R.Handegarde. 1971. The lethal decline of mesquiteon the Casa Grande Ruin National Monument.Great Basin Naturalist 31: 152–159.

Koprowski, J. L. and K. Monroe. 2003. Ecology ofround-tailed ground squirrels at Casa GrandeRuins National Monument. Proposal to WesternNational Parks Association.

Kucera, T. E., and R. H. Barrett. 1993. TheTrailmaster camera system for detecting wildlife.Wildlife Society Bulletin 21: 505–508.

MacArthur, R. H., and J. W. MacArthur. 1961. Onbird species diversity. Ecology 42: 594–598.

McNamee, G. 1994. Gila: The life and death of anAmerican river. University of New Mexico Press,Albuquerque, NM.

Mills, G. S., J. B. Dunning, Jr., and J. M. Bates.1989. Effects of urbanization of breeding birdcommunity structure in southwestern deserthabitats. Condor 91: 416–428.


Naeem, S., L. J. Thompson, T. H. Jones, J. H.Lawton, S. P. Lawler, and R. M. Woodfin. 1996.Changing community composition and elevatedCO2. In Carbon dioxide, populations, andcommunities, edited by C. Korner, and F. A.Bazzaz, pp. 93–100. Academic Press, San Diego,CA.

National Park Service (NPS). 1997. Integrated pestmanagement plan. Casa Grande Ruins NationalMonument. U.S. Department of the Interior,Washington, D.C.

National Park Service (NPS). 1998. ResourcesManagement Plan. Casa Grande Ruins NationalMonument. U.S. Department of the Interior,Washington, D.C.

National Park Service (NPS). 2003a. Resourceprotection study environmental assessment /assessment of effect. U.S. Department of theInterior, Coolidge, AZ.

National Park Service (NPS). 2003b. Casa GrandeRuins National Monument resource protectionstudy. U.S. Department of the Interior, Coolidge,AZ.

Nickens, P. 1996. Documentation of fieldwork andassociated activities undertaken at Casa GrandeRuins National Monument, Arizona, March 22–23, 1996. Unpublished report by Batelle-PacificNorthwest National Laboratory.

North American Ornithological Atlas Committee(NAOAC). 1990. Handbook for atlasing NorthAmerican breeding birds, edited by C. Smith.Accessed 13 July 2001 from: http://americanbirding.org/norac/atlascont.htm.

Office of Technology Assessment (OTA). 1993.Harmful non-indigenous species in the UnitedStates. OTA-F-565, U.S. Government PrintingOffice, Washington D.C.

Powell, B. F., E. W. Albrecht, W. L. Halvorson, andK. Docherty. 2003. Biological inventory report forthe Sonoran Desert Network: 2002. AnnualReport No. 2. Sonoran Desert Network InventoryProgram. U.S. Geological Survey, Sonoran DesertField Station and School of Natural Resources,University of Arizona, Tucson, AZ.

Powell, B. F., E. W. Albrecht, C. A. Schmidt, W. L.Halvorson, P. Anning, and K. Docherty. 2006.Vascular plant and vertebrate inventory of CasaGrande Ruins National Monument. USGS Open-File Report 2005-1185. U.S. Geological Survey,Southwest Biological Science Center, SonoranDesert Research Station, University of Arizona,Tucson, AZ.

Powell, B. F., K. Docherty, and W. L. Halvorson.2002. Biological inventory report for the SonoranDesert Network: 2000 and 2001 field seasons.Annual Report No. 1. Sonoran Desert NetworkInventory Program. U.S. Geological Survey,Sonoran Desert Field Station and School ofNatural Resources, University of Arizona,Tucson, AZ.

Powell, B. F., and R. J. Steidl. 2000. Nesting habitat

and reproductive success of southwestern riparianbirds. Condor 102: 823–831.

Ralph, C. J., J. R. Sauer, S. Droege, technicaleditors. 1995. Monitoring bird populations bypoint counts. General Technical Report PSW-GTR-149. Pacific Southwest Research Station,Forest Service, U.S. Department of Agriculture,Albany, CA.

Reichhardt, K. 1992. Natural vegetation of CasaGrande Ruins National Monument, Arizona.Technical Report NPS/WRUA/ NRTR-92/45.University of Arizona, Tucson. CooperativeNational Park Resources Studies Unit.

Reynolds, R. T., J. M. Scott, and R. A. Nussbaum.1980. A variable circular-plot method forestimating bird numbers. Condor 82: 309–313.

Richardson, Jr., J. J. 2002. Legal impediments toutilizing groundwater as a municipal water sourcein karst terrain in the United States.Environmental Geology 42: 532–537.

Rosen, P. C., and C. H. Lowe. 1994. Highwaymortality of snakes in the Sonoran desert ofsouthern Arizona. Biological Conservation 68:143–148.

Rosen, P. C., and K. Mauz. 2001. Biological valuesof the West Branch of the Santa Cruz River, withan outline for a potential reserve; includingpreliminary flora. Document for the SonoranDesert Conservation Plan, Pima County Board ofSupervisors, Tucson. http://www.co.pima.az.us/cmo/sdcp/sdcp2/reports/WB/WestBF.htm#_Toc524108384

Sauer, J. R., J. E. Hines, and J. Fallon. 2004. TheNorth American Breeding Bird Survey, resultsand analysis 1966–2003. Version 2004.1.Patuxtant Wildlife Research Center, Laurel MD.http://www.mbr-pwrc.usgs.gov/bbs/ bbs.html.

Sprouse, T., R. Emanuel, B. Tellman. 2002. Finalreport: Surface water quality monitoring overviewand assessment for the Sonoran Desert Network,National Park Service. Unpublished report. WaterResources Research Center, University ofArizona, Tucson, AZ.

Stebbins, R. C. 2003. A field guide to westernreptiles and amphibians. Third edition. HoughtonMifflin, New York, NY.

Swann, D. E., C. C. Hass, D. C. Dalton, and S. A.Wolf. 2004. Infrared-triggered cameras fordetecting wildlife: An evaluation and review.Wildlife Society Bulletin 32: 1–9.

Swann, D. E., W. W. Shaw, and C. R. Schwalbe.1994. Assessment of animal damages to arche-ological resources at Casa Grande Ruins NationalMonument. Final report to Southern ArizonaGroup. National Park Service. Phoenix, AZ.

Tigas, L. A., D. H. Van Vuren, and R. M. Sauvajot.2002. Behavioral responses of bobcats andcoyotes to habitat fragmentation and corridors inan urban environment. Biological Conservation108: 299–306.

Trombulak, S. C., and C. A. Frissell. 2000. Review


of ecological effects of roads on terrestrial andaquatic communities. Conservation Biology 14:18–30.

United States Department of Agriculture (USDA).2001. The PLANTS Database, Version 3.5(http://plants.usda.gov). National Plant DataCenter, Natural Resources Conservation Service,Baton Rouge, LA.

Van Pelt, D. 1998. Assessing impacts of dropping

water table in Casa Grande Ruins NationalMonument. M.S. Thesis, Department ofHydrology and Water Resources, University ofArizona, Tucson, AZ.

Western Regional Climate Center (WRCC). 2004.Arizona climate summaries for Casa GrandeRuins National Monument. http://www.wrcc.dri.edu/summary/climsmaz.html

VASCULAR PLANT AND VERTEBRATE INVENTORY OF TUMACÁCORI NATIONAL HISTORIC PARKBrian F. Powell, Eric W. Albrecht, Cecilia A. Schmidt, Pamela Anning, Kathleen Docherty

From 2000 to 2003 we surveyed for vascularplants and vertebrates (fish, amphibians,reptiles, birds, and mammals) at TumacácoriNational Historical Park (NHP) to documentpresence of species within the administrativeboundaries of the park’s three units. Becausewe used repeatable study designs and stan-dardized field techniques, these inventoriesserve as the first step in a long-term monitoringprogram. We recorded 591 species at Tumacá-cori NHP, significantly increasing the numberof known species for the park. Species of notein each taxonomic group include: • Plants: Second record in Arizona of

muster John Henry, a non-native speciesthat is ranked a “Class A noxious weed”in California;

• Amphibian: Great Plains narrow-mouthedtoad;

• Reptiles: Eastern fence lizard andSonoran mud turtle;

• Birds: Yellow-billed cuckoo, green king-fisher, and one observation of theendangered southwestern willowflycatcher;

• Fishes: Four native species including animportant population of the endangeredGila topminnow in the TumacácoriChannel;

• Mammals: Black bear and all four speciesof skunk known to occur in Arizona.

We recorded 79 non-native species, manyof which are of management concern,including: Bermudagrass, tamarisk, westernmosquitofish, largemouth bass, bluegill,sunfish, American bullfrog, feral cats and dogs,and cattle. We also noted an abundance ofcrayfish (a non-native invertebrate). We reviewsome of the important non-native species andmake recommendations to remove them or tominimize their impacts on the native biota ofthe park.

Tumacácori NHP possesses high biologicaldiversity of plants, fish, and birds for a park ofits size. This richness is due in part to theecotone between ecological provinces(Madrean and Sonoran), the geographic distri-bution of the three units (23 km separates themost distant units), and their close proximityto the Santa Cruz River. The mesic life zonealong the river, including cottonwood/willowforests and adjacent mesquite bosque at theTumacácori unit, is representative of areas thathave been destroyed or degraded in manyother locations in the region. Additionalelements such as the semi-desert grasslandvegetation community are also related to highspecies richness for some taxonomic groups.

For complete details of the TumacácoriNHP study see Powell et al. (2005)[http://sbsc.wr.usgs.gov/products/ofr/]. Scien-tific and common names used throughout thischapter are current according to acceptedauthorities for each taxonomic group: Inte-grated Taxonomic Information System (ITIS2001) and the PLANTS Database (USDA


2004) for plants; Stebbins (2003) for amphib-ians and reptiles; American OrnithologistUnion (AOU; 1998, 2003) for birds; and Bakeret al. (2003) for mammals.

PARK OVERVIEWPark Area and History

Tumacácori NHP contains three small units:Calabazas, Guevavi and Tumacácori. Tumacá-cori, the main administrative unit, is located on258 hectares at the town of Tumacácori,Arizona. The Calabazas (9 ha) and Guevavi (3ha) units are 15 km and 23 km SSE ofTumacácori, respectively (Figure 1). The unitsof the park lie along the Santa Cruz River; theriver is perennial at Tumacácori and Guevavi.

Tumacácori NHP preserves the remnants ofthree Spanish colonial missions located alongthe upper Santa Cruz River in southernArizona. Originally established in 1908 as amonument under the Antiquities Act, the parkprotected the San Jose de Tumacácori(Tumacácori unit), a Spanish mission foundedin 1691. In 1990, the area was designated aNational Historical Park with the inclusion ofLos Santos Angeles de Guevavi mission(Guevavi unit; founded in 1691) and SanCayetano de Calabazas mission (Calabazasunit; founded in 1756).

Physiography and Geology The Upper Santa Cruz River Valley is locatedin the southern Basin and Range Province ofsoutheastern Arizona and northern Sonora.This terrain of alternating, fault-bounded,linear mountain ranges and sediment-filledbasins began to form in southeastern Arizonaas the result of dominantly east-northeast/west-southwest-directed crustal extension. Themountain ranges to the east of the park (SantaRita, San Cayetano, and Patagonia) consist ofa variety of rocks, including igneous, meta-morphic, volcanic, and sedimentary, rangingin age from Precambrian to Miocene. TheTumacácori and Atascosa Mountains west ofthe park are composed chiefly of Tertiaryvolcanic rocks with the exception of a Jurassicgranitic pluton south of Sopori Wash at the

northern end of the Tumacácori Mountains.The Pajarito Mountains at the southern end ofthe valley, west of Nogales, are composed ofCretaceous volcanics.

Hydrology and SoilsThe three units of Tumacácori NHP are asso-ciated with distinct pockets of reliable water,resulting from basin-fill sediments over rela-tively shallow aquifers that fill quickly afterprecipitation (Sprouse et al. 2002). Perennialflow at the Tumacácori unit is augmented bytreated wastewater discharges from theNogales International Wastewater TreatmentPlant. Basin-fill sediments along the SantaCruz River, north of the City of Nogales toAmado, form three aquifer units: NogalesFormation, Older Alluvium, and Younger Allu-vium (ADWR 1999). In the vicinity of all thepark units the soils are typical of floodplains,alluvial fans, and valley slopes of this semi-desert region; they are deep and well drained,with a high water-holding capacity (NPS1996).

ClimateTumacácori NHP is located within the semi-desert climatic zone of southern Arizona,which is characterized by heavy summer(monsoon) storms brought about by moisturecoming from the Gulf of Mexico and lessintense, frontal storms from the Pacific Oceanin the winter. Approximately half of the annualprecipitation falls from July to September(WRCC 2004). The area’s hot season occursfrom April through October; maximumtemperatures in July often exceed 40° C.Intense surface heating during the day andactive radiant cooling at night can result indaily temperature ranges of 17° to 22° C.Winter temperatures are mild. Prevailingwinds tend to follow the Santa Cruz Valley,blowing downslope (from the south) duringthe night and early morning, and upslope (fromthe north) during the day.

Weather during the three years of this studywas highly variable and atypical. Annual totalprecipitation ranged from slightly greater than


Figure 1. Locations of the three units of Tumacácori NHP in southern Arizona.


average (340 mm) in 2000 (488 mm) and 2001(422 mm) to one of the driest years on recordin 2002 (236 mm). Annual mean temperatureswere above the long-term mean (17.7° C) in2000 and 2002 (18.0° C in both years) andbelow it in 2001 (17.4° C) (WRCC 2004).

VegetationAll three units have vegetation typical of thesemi-desert grassland association (Brown et al.1979; Brown and Lowe 1980). Common speciesinclude velvet mesquite, foothills palo verde andspecies of acacia, wolfberry, and greythorn, aswell as annual and perennial grasses, and forbs(NPS 1996). At the Tumacácori unit, in partic-ular, there are dense stands of mesquite bosqueand gallery riparian vegetation. Velvet mesquite,netleaf hackberry, and Mexican elderberry arecommon in the mesquite bosque areas, whereasin the mesic riparian areas, Fremont cottonwood,Goodding’s willow, tamarisk, and Arizonawalnut form dense and structurally diversestands of vegetation, particularly adjacent tosurface water.

Natural Resource Management IssuesBecause of its location along a river corridor,its proximity to the border with Mexico, andits diversity of biotic communities, Tumacá-cori NHP has many natural resourcemanagement issues that deserve attention.

Adjacent DevelopmentThe boundaries of the Tumacácori unit are nearthe town of Tubac (to the north) and the RioRico development (to the south). Much of theremaining undeveloped land adjacent to thepark is currently used for irrigated agricultureor livestock grazing, but the population of RioRico is expected to increase three-fold by 2025(ADWR 1999) and Tubac is rapidly expandingas well. Similarly, increased residential devel-opment is taking place near the Calabazas unit,which is close to ex-urban sprawl from theCity of Nogales, Arizona. Potential impacts ofresidential development include an increase inthe number and extent of non-native plants,increased runoff of toxins and sediment,

disruption of animal movement patterns,habitat loss and fragmentation, and increasedharassment and mortality of native animals byfree roaming pets, feral dogs and feral cats(Mills et al. 1989; Theobald et al. 1997; Rileyet al. 2003).

Water QualityGiven the park’s location along the Santa CruzRiver, the quantity and quality of surface waterare important concerns at the park (Sprouse etal. 2002; King et al. 1999). Effluent from theNogales International Wastewater TreatmentPlant (NIWTP), located 14 km upstream fromthe Tumacácori unit and across the river fromthe Calabazas unit, has a significant impact onboth water quality and quantity. Treatedeffluent provides perennial surface flow formore than 15 km of the Santa Cruz River in anarea that would otherwise be dry much of theyear.

Countering the benefits from the presenceof the effluent are a number of water qualityproblems that affect park resources. Watersamples from the Calabazas Road Bridge(located between the Tumacácori andCalabazas units) have included twenty-twogroups of parameters that exceeded NationalPark Service Water Resources Divisionscreening criteria (NPS 2001). In addition,dissolved oxygen, pH, chlorine, cyanide,cadmium, copper, lead, mercury, selenium,silver, and zinc exceeded respective U.S. Envi-ronmental Protection Agency (USEPA) criteriafor the protection of freshwater aquatic life(ADEQ 2000; USEPA 2001). Nitrate, arsenic,barium, cadmium and chromium exceededUSEPA drinking water criteria (ADEQ 2000;USEPA 2001), fecal-indicator bacteria concen-trations (total coliform and fecal coliform)exceeded Water Resources Division screeninglimits for freshwater bathing, and turbiditymeasurements exceeded Water ResourcesDivision limits deemed safe for aquatic life(ADEQ 2000; NPS 2001). The ADEQ hascategorized the water in the Santa Cruz Riveras “impaired” due to turbidity at the Guevaviunit and impaired due to fecal-indicator


bacteria concentrations along stretches adja-cent to both the Calabazas and Tumacácoriunits (ADEQ 2000).

Although levels of ammonia decreasedwith distance from the treatment plant, thetoxicity of water upstream of the Tumacácoriunit may dramatically reduce the likelihoodthat additional native aquatic or semi-aquaticanimals will colonize that area from upstreamlocations. Indeed, in an earlier study,researchers noted that despite presence of fivespecies of amphibian upstream from the treat-ment plant, “… no amphibians are foundalong the river from the waste water outfalldownstream for several kilometers” (Drost1998). Upgrades to the treatment plant sched-uled in 2009 will significantly improve thewater quality in the effluent-dominated reachof the river.

Riparian Plant CommunitiesRiparian plant communities in the south-western United States account for less than 1%of the landscape cover (Skagen et al. 1998),yet it is estimated that greater than 50% ofsouthwestern bird species (Knopf and Samson1994) and up to 80% of all wildlife species inthe southwest are dependent on riparian areas(Chaney et al. 1990). Riparian areas in aridregions support high bird species diversity dueto their structural and floristic diversity, givingrise to abundant insects for foraging and largetrees for nesting (Thomas et al. 1979; Lee etal. 1989; Strong and Bock 1990; Powell andSteidl 2000). Riparian vegetation, such ascottonwood, willow, and ash have been foundto decrease levels of heavy metals in water andsoil in addition to decreasing water tempera-tures, providing a source of organic matter,and stabilizing stream banks (Osborne andKovacic 1993; Karpiscak et al. 1996;Karpiscak et al. 2001). The Bureau of LandManagement estimates that less than 20% ofthe western United States’ potential riparianvegetation remains to perform these vital serv-ices (BLM 1994). Such loss highlights theimportance of maintaining these rare riparianplant communities along the Santa Cruz River.

Non-native SpeciesAwareness of non-native species as a manage-ment issue has risen dramatically in recentyears and ecologists have ranked it as one ofthe most significant causes of species endan-germent (Brooks and Pyke 2001). Non-nativeplant species are a significant managementissue at the park, particularly at the Tumacá-cori unit, where invasives such as tamarisk,Bermudagrass, and Russian thistle are wellestablished. Non-native plants are known toalter ecosystem function and processes(Naeem et al. 1996; D’Antonio and Vitousek1992) and reduce the abundance of nativespecies, potentially permanently changingdiversity and species composition (Bock et al.1986; D’Antonio and Vitousek 1992; OTA1993). The Tumacácori unit provides habitatfor non-native vertebrates as well, includingwestern mosquitofish, largemouth bass,bluegill, sunfish, American bullfrog, housesparrow, European starling, cattle, and feralcats and dogs.

Impacts from Undocumented ImmigrantsThe proximity of the park to the U.S.-Mexicoborder results in a number of unique manage-ment issues. The Santa Cruz River provides awell-known and well-used corridor for undoc-umented immigrants traveling north fromMexico, and much of this traffic passesthrough the Tumacácori unit (NPS 2003).Although visitor safety concerns have not yetbecome a significant issue, reported impactsto the park include erosion, compacted soils,and vandalism (NPS 1996). Other nationalparks in the border region have reported firehazards, theft and destruction of historicresources, disruption of wildlife movements(including reduced access to water sources),reduced water quality, and closure of parkattractions due to safety concerns (NPS 2003).To our knowledge only one (recently initiated)study is aimed at quantifying the effects ofimmigrants on animal communities insouthern Arizona (O’Dell 2003; McIntyre andWeeks 2002).


Trash FlowsDuring heavy flooding events in the SantaCruz River watershed, a large quantity of trashwashes downstream, primarily from theNogales (Portrero) Wash that originates inSonora, Mexico. This trash often becomestrapped in a few locations, leading to largeaccumulations. Trash is principally plastics,such as water bottles, but also includesbatteries and tires. Park personnel and volun-teers have done an excellent job of cleaning upa number of these sites in recent years, but theproblem will likely continue.

METHODSPlants

Previous inventoriesThere were two previous plant inventories atthe Tumacácori unit. Mouat et al. (1977)recorded 130 species during seven survey daysin the spring and summer of 1977. A lesscomplete and poorly documented survey wasconducted at the Tumacácori unit in the 1980sor 1990s (Bennett, year unknown). There arealso 50 specimens from the park at the Univer-sity of Arizona (UA) Herbarium. To ourknowledge, there were no prior plant invento-ries at either the Calabazas or Guevavi units.

This inventoryOur surveys included both qualitative andquantitative methods: qualitative “generalbotanizing” surveys in which we opportunis-tically collected and recorded plants, andquantitative plot-based sampling whichincluded three complementary methods to esti-mate abundance, percent cover by species, andspecies composition of all plants in a smallarea.

General botanizing surveys encompassedmost areas within the park during most visits.For modular plots we used a simple randomsampling design (Thompson 1992) to locate17 plots. We also subjectively placed threeplots in community types that we felt were notrepresented by the 17 other plots: one each inan agricultural field, in dense mesquite bosque,and in dense mesic riparian vegetation.

We made 29 general botanizing visits (typi-cally one observer for one-half day) within theTumacácori unit and 14 and 15 visits to theCalabazas and Guevavi units, respectively.Whenever possible we collected one repre-sentative specimen (with reproductivestructures) for each plant species in each parkunit. We also maintained a list of speciesobserved but not collected within each unit.We accessioned mounted specimens into theUA herbarium. Four observers measured vege-tation on 20 plots at the Tumacácori unitduring ten field days from 2 to 17 November2002. Nineteen plots had four modules in a 20x 20 m arrangement and one plot had twomodules in a 20 x 10 m arrangement (78modules).

FishPrevious inventoriesArizona Game and Fish Department personnelcompleted periodic surveys of the Santa CruzRiver, in or near Tumacácori NHP, between1998 and 2001, in 2003, and one survey in theTumacácori Channel in 1999 (Voeltz andBettaso 2003).

This inventoryWe surveyed for fish at two sites at theTumacácori unit: the Santa Cruz River andthe Tumacácori Channel. The channel is anabandoned meander that is watered bygroundwater seepage, maintaining a down-stream connection with the river. We capturedfish using two methods: (1) electrofishing(Dauble and Gray 1980) with a Smith-Rootbackpack unit in both areas (12-B POW setto DC pulse width of 60 Hz, frequency of 6ms, voltage of 300 V; Smith-Root, Inc.,Vancouver, WA); and (2) dipnetting (long-handled dip nets with 4 mm mesh; Daubleand Gray 1980) in shallow-water areas of thechannel. Three to four field personnelsurveyed the channel and river on foursampling periods; once each in the spring andfall of 2001 and 2002.


Amphibians and ReptilesPrevious inventoriesTo our knowledge there were no inventoriesnor any research related to amphibians andreptiles at Tumacácori NHP prior to this study.

This inventoryWe surveyed for herpetofauna in 2001 and2002 using four methods. We used bothdiurnal and nocturnal surveys in an effort todetect species with restricted periods ofactivity (Ivanyi et al. 2000).

For all methods except intensive surveys,we surveyed for herpetofauna in non-randomsites because we wanted to detect as manyspecies as possible. To determine locations forintensive survey plots we used a modifiedsimple-random sampling design (Thompson1992) whereby we used ArcView GIS soft-ware to select points (15) at random in eachpark unit to serve as the southwest corner ofeach plot. We then post-stratified plots(Thompson 1992) by vegetation type; in lieuof a vegetation map for the park we used aerialphotographs to estimate vegetation types andmoved the location of some plots thatappeared to cover more than one vegetationtype into the type representing the majority ofthe area.

In 2001 we used searches constrained byboth time and area to provide a standardizedsurvey method. These surveys are similar tovisual encounter surveys (Crump and Scott1994), but were confined to a 1 ha (100 x 100m) plot. Due to the heterogeneity of vegeta-tion types at the Tumacácori unit, our randomlocations resulted in plots representing each ofthe dominant community types in the park:riparian, mesquite bosque, semi-desert grass-land, and agricultural land. We visited all plotson 24–27 April and most plots again on 9–10September, 2001. We began all 35 surveys(one person-hour each) between 7:30a.m. and12:30p.m. In 2002 we chose not to continueintensive surveys because of the relatively lownumber of species and individuals recorded,and instead focused our efforts on othermethods.

We used the extensive survey method forboth diurnal and nocturnal surveys. To increasethe odds of finding rare animals, we placed 20cover boards (Fellers and Drost 1994) aroundthe festival grounds area at the Tumacácoriunit, and checked these opportunistically byturning the boards. We recorded UTM coordi-nates to define the boundaries of our searcharea or the path we followed during oursurveys.

We spent 168 hours on 52 extensivesurveys between 24 April and 24 September2001, and between 11 July and 25 August2002. Almost 90% of the surveys (n = 46)were initiated during the cooler evening, night-time, or morning hours (5p.m. to 9a.m.). Weoperated one pitfall trap array (with four pitfalltraps and six funnel traps) near the bank of theSanta Cruz for a total of 672 hours between 11July and 24 September 2001 and between 10July and 26 September 2002 (Gibbons andSemlitsch 1981; Corn 1994).

When we encountered uncommon amphib-ians and reptiles outside of formal surveys, werecorded the species, sex and age class (ifknown), time of observation, UTM coordi-nates, and route we were following. Incidentaldetections recorded by other survey crews(e.g., bird crew) were not accompanied byroute descriptions. We collected incidentalobservations between 25 April and 24September 2001, and between 15 May and 25August 2002.

BirdsPrevious inventoriesPrevious bird research at Tumacácori NHPfocused on the yellow-billed cuckoo (Powell2000) and on mist-netting passerines andhummingbirds from 1997 to 2004 (Turner2003).

This inventoryWe surveyed for birds at the Tumacácori unitin 2001 and at all three units in 2002 and 2003.We used four field methods: variable circular-plot (VCP) counts for diurnal breeding-seasonbirds, nocturnal surveys for owls and nightjars


during the breeding season, line-transects forfall and winter-season birds, and incidentalobservations for all birds in all seasons. Weconcentrated most of our survey effort duringthe breeding season because bird distributionis relatively uniform during the breedingseason due to territoriality among birds (Bibbyet al. 2000). This survey timing increases ourprecision in estimating relative abundance andenabled us to document breeding activity. Oursurvey period included peak spring migrationtimes for most species, which added manymigratory species to our list (Reynolds et al.1980; Buckland et al. 1993; Ralph et. al 1995;McGarigal et al. 2000)

We used a modified line-transect method(Bibby et al. 2000) to survey for birds in all unitsfrom November 2002 to January 2003. Linetransects differ from station transects (such asthose used in our VCP surveys) in that anobserver records birds seen or heard while theobserver walks an envisioned transect line, ratherthan by standing at a series of stations. The tran-sect method is more effective during thenon-breeding season because bird vocalizationsare less conspicuous and frequent, and thereforebirds tend to be less visible (Bibby et al. 2000).

To survey for owls we broadcastedcommercially available vocalizations (Colveret al. 1999) using a compact disc player andbroadcaster (Bibby et al. 2000), and recordedother nocturnal species (nighthawks and poor-wills) when heard. We established onenocturnal survey transect along a road or trailin each park unit. The number of stationsvaried from one to three per transect andstations were a minimum of 300 m apart. Aswith other survey methods, we variedobservers and direction of travel along tran-sects and did not survey during periods ofexcessive rain or wind. We began surveysapproximately 45 minutes after sunset (Fullerand Mosher 1987). We did not specificallysurvey for any species listed as threatened orendangered — e.g., the cactus ferruginouspygmy-owl (Glaucidium brasilianumcactorum) because such species requirespecific protocols for surveying.

When we were not conducting formalsurveys and we encountered a species ofinterest, a species in an unusual location, or anindividual displaying breeding behavior, werecorded UTM coordinates, time of detection,and (if known) the sex and age class of thebird. We noted all breeding-behavior observa-tions using a standardized classification system(NAOAC 1990).

MammalsPrevious inventoriesTo our knowledge, there has been no previousmammal research at Tumacácori NHP, andthough some mammal specimens have beencollected from the park, there have been nocomprehensive mammal inventories.

This inventoryWe surveyed for mammals using four fieldmethods: trapping for small mammals,infrared- triggered photography for mediumand large mammals, investigation of roost sitesfor bats, and incidental observations for allmammals.

We trapped small mammals at all threeunits in 2000 and 2001. We used Sherman®

live traps (large, folding aluminum or steel,3 x 3.5 x 9"; H. B. Sherman, Inc., Tallahassee,FL) set in grids (White et al. 1983), with 10 mspacing among traps arranged in configura-tions of five rows and five columns (Calabazasand Guevavi units) or 10 rows and fivecolumns (Tumacácori unit).

We used Trailmaster® cameras (model1500, Goodman and Associates, Inc., Lenexa,KS; Kucera and Barrett 1993) to record thepresence of medium and large mammals at theTumacácori unit only.

As with other taxa, we recorded UTM coor-dinates of incidental mammal sightings.Observers from all field crews (e.g., bird crewas well as mammal crew) recorded mammalsightings and signs such as identifiable tracksor scat, and we took photo vouchers when thesign alone was definitive.

We visited the Tumacácori unit once, on 2October 2001, to search for bats in and around


the Mission structure. We did not mist net batsat the Santa Cruz River because netting ismost efficient when areas of open water arelimited, thereby concentrating foraging batsinto a small area.


We recorded 378 species during generalbotanizing and modular plot surveys in 2000to 2003 (Table 1). We recorded the mostspecies at the Tumacácori unit (293). Werecorded fewer species at the Calabazas (175)and Guevavi units (151). The most commonfamilies were composites (Asteraceae),grasses (Poaceae) and legumes (Fabaceae).More than 82% of the remaining families wererepresented by three or fewer species, a patternconsistent with floras of nearby areas(McLaughlin et al. 2001). We included allsubspecies and/or varieties in our summarystatistics of the number of “species” recorded.

We recorded 67 non-native species in allunits combined. Excluding ornamentals, thepercentage of non-native species was 18% atthe Tumacácori unit (52), 11% at the Guevaviunit (17), and 9% at the Calabazas unit (16).Considering all units, the grass family(Poaceae) had the highest percentage (33) ofnon-native species. Perhaps the most notablenon-native species we documented was themuster John Henry (at the Tumacácori unit);this was the second documentation of itsoccurrence in Arizona.

We searched for two endangered speciesthought to be in the area: Pima pineapplecactus (Coryphantha scheeri var. robustispina)and Huachuca water umbel (Lilaeopsisschaffneriana var. recurva), but did not findeither.

Other researchers listed 46 species at theTumacácori unit that we did not find. Whilethese may still be present and we missed them,alternative explanations include misidentifi-cation (many previous records were notdocumented), local extirpation, and use ofdifferent field methods. Judging from the pres-ence of conspicuous perennial species at both

the Calabazas and Guevavi units but notTumacácori unit (e.g., whitethorn acacia),some local extirpation may have occurred.Three factors likely contributed to our findingtwice the number of species reported byprevious studies (Mouat et al. 1977; Bennettyear unknown): (1) our field effort was morethan twice that of previous studies, (2) oursurvey area was larger, and (3) we likely bene-fited from a winter (2000) that had moreprecipitation than average, resulting in higherspecies richness of annuals during our field-work than during sampling by Mouat et al.(1977).

Once well established, a number of non-native species pose a significant managementproblem. Prominent species include tamarisk,Bermudagrass, Lehmann lovegrass, John-songrass, Russian thistle, London rocket, andyellow sweet clover. Bermudagrass, in partic-ular, is difficult to control and was recorded in85% of the modular plots. A complete descrip-tion of these non-native species as well as theirlife history, threat to native species, and erad-ication method(s) can be obtained fromHalvorson and Guertin (2003).

Additional general botanizing surveys,carried out during wet summers in the expan-sion area of the Tumacácori unit, shouldincrease the species list for annual plants. Adiligent effort to seek out species that havepreviously been reported (Moatt et al. 1977;Bennett, year unknown) but not found duringour surveys, would help confirm possiblechanges to the flora. Additional modular plots,especially in the under-sampled mesquitebosque and dense cottonwood/willow forests,would be an effective tool for long-term moni-toring of vegetation changes.

FishWe recorded eight species (four native, fournon-native) and one hybrid (greensunfish/bluegill) at Tumacácori NHP (Table 1).We recorded all eight species on the firstsampling event in the channel, and Gilatopminnow, longfin dace, and western mosqui-tofish in both sites on all sampling events.


Species richness in the channel was higher thanin the river on all but the last sampling event.

Based on distribution records (Minckley1973) and our experience, we documented allspecies that are thought to occur in the park.The apparent persistence of the federallyendangered Gila topminnow in the Tumacá-cori Channel is perhaps the most importantfinding of our inventory effort. We found Gilatopminnow in all sampling periods and theywere likely present in the channel during theextreme drought of the late spring/earlysummer of 2002 when no surface water waspresent at the river site. It appears that thechannel serves as a refugium during extremedrought events.

The other native species documented werelongfin dace, Sonora sucker and desert sucker.The presence of Sonora and desert suckers forthe first three sampling events in the Tumacá-cori Channel is significant because bothspecies are rare in southern Arizona (Recon2004). There was at least one prior sighting ofthe desert sucker in the channel in June 1999(Voeltz and Bettaso 2003). However, we didnot find either species during our fourthsampling period in the fall of 2002. Thepossible loss of these species from the park isparticularly troubling given that movementback to the park may take considerable time ormay not happen at all.

Once introduced throughout the westernU.S. to control mosquito populations, the

western mosquitofish is thought to be one ofthe major reasons for population declines ofthe Gila topminnow (and other small nativefishes in the southwest) through predation,harassment, and competition (Meffe 1985;Courtenay and Meffe 1989). The persistenceof the Gila topminnow in the presence ofwestern mosquitofish is notable. In addition tolarge numbers of western mosquitofish, werecorded three other non-native sport-fishspecies: largemouth bass, bluegill, and greensunfish.

During most of our surveys, we observedcrayfish (Orconectes virilis), an important non-native invertebrate, especially in the channel.Crayfish are one of the most serious threats tonative aquatic biota because they effectivelycompete with aquatic herbivores, prey onaquatic invertebrates and vertebrates, disruptnormal nutrient cycling, and decrease aquaticmacroinvertebrate diversity (Creed 1994;Fernandez and Rosen 1996). A program toeliminate these crayfish would be beneficial tothe native biota of the Santa Cruz River andthe Tumacácori Channel, but would be logis-tically difficult.

Amphibians and ReptilesUsing intensive surveys, we recorded sevenamphibian and 17 reptile species at Tumacá-cori NHP in 2001 and 2002 (Table 1). Werecorded the most species (n = 22) at theTumacácori unit and the fewest (n = 9) at the

Table 1. Summary results of vascular plant and vertebrate inventories at Tumacácori NHP, 2000–2003.

Number of Number of Number of new species Taxonomic group species recorded non-native species added to park lista

Plants 378 67 168Fish 8 4 3Amphibians and Reptiles 24 1 22Birds 146 3 40Mammals 35 4 33Totals 591 79 266

a Species that had not been observed or documented by previous studies.


Guevavi unit. A single Woodhouse’s toad washeard calling on 24 July 2001; otherwise theonly amphibians heard vocalizing were Amer-ican bullfrogs.

We recorded the most species (n = 7) in thesemi-desert grassland community type and thefewest (n = 2) in the mesquite bosque commu-nity type. Clark’s spiny lizard was the mostwidespread species (recorded at least once inall park units and community types), theSonoran spotted whiptail was the most abun-dant reptile on plots in all community typesexcept in the mesquite bosque where it wasnot recorded. The common lesser earlesslizard, Clark’s spiny lizard, and regal hornedlizard were most abundant in the semi-desertgrassland community type; the eastern fencelizard and the two whiptails were most abun-dant in riparian areas; and the ornate tree lizardand gophersnake were most abundant in thecleared mesquite bosque area.

Relative abundance of all reptiles in theriparian community type was more than 30%higher than in any other community type, yetspecies richness in the riparian communitytype (6 species) was equal to that of the semi-desert grassland community type at theCalabazas unit and only slightly greater thanrichness in the cleared mesquite bosquecommunity type (5 species). The eastern fencelizard and the regal horned lizard were theonly species that we found in only onecommunity type, riparian and semi-desertgrassland, respectively.

We added one new species to the park listby using incidental observations (western boxturtle) and another using the pitfall array(Great Plains narrow-mouthed toad).

Species accumulation curves for amphibiansurveys indicate that we recorded most of thespecies likely to be observed with thesemethods, at least under the environmentalconditions we experienced during our study.In contrast, our reptile list for the park is likelyfar from complete. Based on range maps,known habitat requirements, historic records,and results of a nearby study, an additional 32species may be present, may have been histor-

ically present, or might pass through the parkin the course of movement from nearby areas.

We found American bullfrogs in theTumacácori Channel during both herpetolog-ical and fish surveys. The American bullfrogis native to eastern North America but has beenintroduced throughout the western U.S. forfood production and sport (Stebbins 2003).The American bullfrog is a species of manage-ment concern at Tumacácori NHP becauseboth adults and tadpoles are voracious preda-tors (Kiesecker and Blaustein 1997) and arethought to be partially responsible for thedecline of many native fish species (Minckleyand Deacon 1991), reptiles (Schwalbe andRosen 1988), and amphibians (particularlyother Ranid frogs; Hayes and Jennings 1986;Lawler et al. 1999) in the southwest.

BirdsWe recorded 146 species during the two yearsof the study. Although comparisons amongunits may be biased by unequal survey effort,species richness was highest at the Tumacácoriunit (n = 129), lower at the Calabazas unit(n = 80), and lowest at the Guevavi unit(n = 74). We recorded 50 species at all threeunits and 59 species at only one of the threeunits. All three non-native species that wefound during this study (rock pigeon, Euro-pean starling, and house sparrow) wererecorded at the Tumacácori unit, whereas onlyone (European starling) was recorded at theCalabazas unit and no non-natives wererecorded at the Guevavi unit. Neotropicalmigrant species made up 71% (103) of allspecies recorded.

We recorded 104 species during VCPsurveys in 2001 and 2002. Species richnesswas 57–71 among community types. We found34 species that were unique to a singlecommunity type while an equal number ofspecies were recorded in all community types.The number of species unique to a communitytype was highest for semi-desert grasslands(14) and fewest for the developed area (4).House sparrows were most abundant and wererecorded predominantly in the developed area.


Other abundant species in the developed areawere the vermilion flycatcher and phainopepla.In the mesquite bosque, the yellow-breastedchat, Bell’s vireo, and Lucy’s warbler weremost abundant, in the adjacent riparian area,the yellow-breasted chat, Bewick’s wren, andsong sparrow were the most abundant. At theCalabazas and Guevavi units, representing thesemi-desert grassland community, the Lucy’swarbler and Bewick’s wren were the mostabundant.

We recorded 56 species during line-transectsurveys in the three park units. Species rich-ness was 21–23 among the community types.The most abundant species in each communitytype were: the chipping and white-crownedsparrows in the mesquite bosque, the Europeanstarling and white-crowned sparrow in thedeveloped area, the yellow-rumped warblerand chipping sparrow in the riparian area, andthe chipping sparrow and mourning dove inthe semi-desert grassland.

During nocturnal surveys, we recordedthree species of owls (barn owl, westernscreech owl, elf owl) and one common poor-will. We recorded incidental observations of121 species; 41 species that were not recordedduring another survey type. Based on thespecies accumulation curve, we believe thatwe have recorded at least 90% of the speciesthat breed in and around the park or thatstopover for a significant amount of timeduring the time of the VCP surveys. However,based on the high bird species richness and thediversity of vegetation at the park, we believethat the bird inventory is not complete and islikely missing spring and fall migrants andwinter residents. There are at least 40 speciesthat were not recorded by us, MAPSpersonnel, or other researchers or observers,but which are likely to be recorded at Tumacá-cori NHP with additional survey effort.

Two rare species in particular are note-worthy: yellow-billed cuckoo andsouthwestern willow flycatcher. These arefound in the riparian area at the Tumacácoriunit. This area has been reported to have oneof the highest densities of yellow-billed

cuckoos in the western U.S. (Powell 2000). There are two species that are troublesome:

house sparrows and brown-headed cowbirds.House sparrows in the developed area aroundthe Tumacácori Mission can be a problem forother native bird species and cultural resources.House sparrows nest in cavities or on ledges(Erlich et al. 1988), and are known to beaggressive toward other cavity-nesting species.At Tumacácori NHP these sparrows maydisplace cavity-nesters such as the Bewick’swren and Lucy’s warblers, can damage culturalresources by enlarging existing cracks, andcertainly create a nuisance and distraction viaexcessive nest material and defecation. Brown-headed cowbirds pose a threat to many nativebirds outside of the developed area becausethey are brood parasites and reduce the produc-tivity of host species. Species particularlysusceptible to brown-headed cowbird para-sitism include four abundant Neotropicalmigrants at Tumacácori NHP: Bell’s vireo,song sparrow, yellow-breasted chat, andyellow warbler (Schwietzer et al. 1998;Averill-Murray et al. 1999; Powell and Steidl2000).

MammalsWe trapped 16 species of small mammals inour trap grids. Species richness was highest atCalabazas (n = 12), slightly less at Tumacácori(n = 11), and lowest at Guevavi (n = 9).However, the Guevavi unit had the mostnumber of species (n = 3) not recorded at otherunits (brush mouse, northern pygmy mouse,and northern grasshopper mouse) although justone individual represented each of thesespecies. The desert pocket mouse was the mostabundant rodent at both the Calabazas andGuevavi units and the second most abundantat the Tumacácori unit, where the cactusmouse was the most abundant. The highspecies richness of small mammal communi-ties at the smaller Calabazas and Guevavi unitsis consistent with known patterns of smallmammal species richness in southern Arizona;true semidesert grasslands contain the highestspecies richness of any vegetation community


in southern Arizona (Price 1978; Stamp andOhmart 1979; Hoffmeister 1986; Sureda andMorrison 1999).

Seventy Trailmaster photographs weregood enough that we could identify an animalto genus or species. From these we identified10 species. Eight of 10 species photographedwere represented by five or fewer photo-graphs. The Virginia opossum was the mostfrequently photographed species. Wephotographed all four species of skunks thatoccur in Arizona. Only one species of bat wasrecorded by the study and one large mammal,a black bear, was documented at the park byphotographing definitive tracks adjacent to theSanta Cruz River.

We believe that our mammal inventory wasmost successful in documenting the list ofsmall mammals. We estimate that we docu-mented at least 90% of the rodent specieslikely to occur at the park based on the speciesaccumulation curve; we did not detect anynew species in the last 10 sampling periods.However, for other groups of mammals, thepicture is quite different. Based on a compar-ison of the species we recorded to a list of“possible” species at the park, we believe thatwe recorded approximately 41% of thespecies. Bats make up the bulk of the specieswe did not find; there are a possible 24 in addi-tion to the one that we recorded. It should benoted, however, that not all of the mammalspecies would use the building structures orvegetation for any significant amount of time.Most may simply fly over (bats) or passthrough (e.g., jaguar) the park en route tohabitat elsewhere. Large mammals that wewould expect to find include: gray fox, kit fox,mountain lion, and ringtail (Hoffmeister1986).

The only non-native rodent, the housemouse, was recorded twice at the Calabazasunit, not at all at the Guevavi unit, and 56times at the Tumacácori unit. By using Trail-master photographs and incidental sightings,we documented or observed the non-nativedomestic cat, domestic dog, and cow.Domestic cats can pose a serious problem for

native vertebrates, especially rodents, reptilesand birds, through harassment and predationof nests and individuals (Clarke and Pacin2002). We documented the Virginia opossumin 37% of our infrared photographs at theTumacácori unit. It is likely that opossums arerelatively new to the area (Hoffmeister 1986)and may have several impacts, includingpredation of bird nests (Peterson et al. 2004),competition with other medium-sized, omniv-orous mammals such as the raccoon (Gingeret al. 2003) and transmission of tuberculosis(Mycobacterium bovis) to both wildlife andlivestock (Fitzgerald et al. 2003).

ACKNOWLEDGMENTSThanks to Park Superintendent Ann Rasor,Roy Simpson, and the staff at TumacácoriNational Historical Park (NHP) for theirsupport of this inventory project.

This project resulted from the collaborationof many people at the United States Geolog-ical Survey, the University of Arizona (UA)and the National Park Service. Andy Hubbardat the Sonoran Desert Network Inventory andMonitoring program was a strong advocate forthe project. Larry Norris at the Desert South-west CESU provided clear and timelyadministrative assistance. We are grateful toGeorge Binney, Roy Ross, and Hank Cosperfor allowing us access to their lands.

We thank a core group of dedicated fieldbiologists who persevered, day in and day out,to collect a wealth of data at Tumacácori NHP:Patty Guertin, Sky Jacobs, James MacAdam,Jeff McGovern, Katie Nasser, and Meg Quinn(plants); Miguel Lerma, Michael Olker, AndySchultz, and Thomas Walker (fish); Dan Bell,Kevin Bonine, James Borgmeyer, Dave Prival,and Mike Wall (amphibians and reptiles);Gavin Bieber, Aaron Flesch, Chris Kirkpatrick,and Gabe Martinez (birds); Melanie Bucci,Ryan Gann, Ed Kuklinski, Neil Perry, JasonSchmidt, and Ronnie Sidner (mammals). Weare appreciative of the following people, manyof whom never ventured into the field, butwhose office tasks made the field effortpossible: Debbie Angell, Jennifer Brodsky,


Valery Catt, Brian Cornelius, Jenny Ferry,Carianne Funicelli, Taylor Edwards, AndyHonaman, Colleen McClain, HeatherMcClaren, Lindsay Norpel, Terri Rice, JillRubio, Brent Sigafus, Taffy Sterpka, PatinaThompson, Jenny Treiber, Cecily Westphal,and Alesha Williams.

Special thanks to Lisa Carder for her yearsof hard work on all aspects of the project.Technical support was graciously given by thefollowing experts: Dan Austin, Michael Cham-berland, Phil Jenkins, and Charlette and JohnReeder of the UA herbarium; George Bradleyof the UA herpetology collection; PeterReinthal of the UA fish collection; Tom Huelsof the UA ornithology collection; and YarPetryszyn and Melanie Bucci of the UAmammal collection. Review and editingsuggestions were offered by Debbie Angell,Kevin Bonine, James Feldmann, Denny Fenn,Andy Hubbard, Brooke Gebow, Larry Laing,Theresa Mau-Crimmins, Larry Norris, AnnRasor, Phil Rosen, and Don Swann.

Eric W. Albrecht died soon after this projectwas completed. His vision and leadershipadded greatly to this and other inventoryefforts.


Checklist of North American birds, seventhedition. American Ornithologists’ Union and AllenPress Inc., Lawrence, KS.

American Ornithologists’ Union (AOU). 2003.Forty-second supplement to the AmericanOrnithologists’ Union checklist of North Amer-ican birds. Auk 117: 847–858.

Arizona Department of Environmental Quality(ADEQ). 2000. The status of water quality inArizona Clean Water Act section 305(b) Report2000. EQR–00–03. Phoenix, AZ.

Arizona Department of Water Resources (ADWR).1999. Third management plan for the PhoenixActive Management Area, 2000–2010. Phoenix.

Averill-Murray, A., S. Lynn, and M. L. Morrison.1999. Cowbird parasitism of Arizona Bell’s vireos(Vireo bellii arizonae) in a desert riparian land-scape: Implications for cowbird management andriparian restoration. Studies in Avian Biology 18:109–120.

Baker, R. J., L. C. Bradley, R. D. Bradley, J. W.Dragoo, M. D. Engstrom, R. S. Hoffmann, C. A.Jones, F. Reid, D. W. Rice, and C. Jones. 2003.

Revised checklist of North American mammalsnorth of Mexico, 2003. Occasional Papers of theMuseum of Texas Tech University 229: 1–23.

Bennett, P. Year unknown. Reference list. (A list ofplants of Tumacácori National Historic Park;obtained in 1999). Unpublished; created while anemployee of the U.S. Geological Survey, Biolog-ical Resources Division, Sonoran Desert ResearchStation, University of Arizona, Tucson, AZ.

Bibby, C. J., N. D. Burgess, and D. A. Hill. 2000.Bird census techniques. Academic Press, London,England.

Bock, C. E., J. H. Bock, K. L. Jepson, and J. C.Ortega. 1986. Ecological effects of plantingAfrican lovegrasses in Arizona. NationalGeographic Research 2: 456–463.

Brooks, M. L., and D. A. Pyke. 2001. Invasiveplants and fire in the deserts of North America.Tall Timbers Research Station MiscellaneousPublication 11: 1–14.

Brown, D. E., and C. H. Lowe. 1980. Bioticcommunities of the Southwest: Supplementarymap to biotic communities: Southwestern UnitedStates and northwestern Mexico, edited by D.Brown. University of Utah Press, Salt Lake City,UT.

Brown, D. E., C. H. Lowe, and C. P. Pase. 1979. Adigitized classification system for the bioticcommunities of North America, with community(series) and association examples for the south-west. Journal of Arizona-Nevada Academy ofScience 14: 1–16.

Buckland, S. T., D. R. Anderson, K. P. Burnham,and J. L. Laake. 1993. Distance sampling: Esti-mating abundance of biological populations.Chapman and Hall, New York, NY.

Bureau of Land Management (BLM). 1994. Range-land reform 1994. Draft environmental impactstatement. U.S. Department of the Interior, Wash-ington, D.C.

Chaney, E., W. Elmore, and W. S. Platts. 1990. Live-stock grazing on western riparian areas. NorthwestResource Information Center, Inc., Eagle, ID.

Clarke, A. L., and T. Pacin. 2002. Domestic cat“colonies” in natural areas: A growing exoticspecies threat. Natural Areas Journal 22: 154–159.

Colver, K. J., D. Stokes, and L. Stokes. 1999. Stokesfield guide to bird songs. Time Warner, New York,NY.

Corn, S. P. 1994. Straight-line drift fences and pitfalltraps. In Measuring and monitoring biodiversity:Standard methods for amphibians, edited by W.R. Heyer, M. A. Donnelly, R. W. McDiarmid, L.C. Hayek, and M. S. Foster, pp. 109–117. Smith-sonian Institution Press, Washington, D.C.

Courtenay, W. R., Jr., and G. K. Meffe. 1989. Smallfishes in strange places: A review of introducedpoeciliids. In Ecology and evolution of livebearingfishes (Poeciliidae), edited by G. K. Meffe and F.F. Snelson, Jr., pp. 319– 331. Prentice Hall, Engel-wood Cliffs, NJ.


Creed, R. P. 1994. Direct and indirect effects ofcrayfish grazing in a stream community. Ecology75: 2091–2103.

Crump, M. L., and N. J. Scott. 1994. Visualencounter surveys. In Measuring and monitoringbiodiversity: Standard methods for amphibians,edited by W. R. Heyer, M. A. Donnelly, R. W.McDiarmid, L. C. Hayek, and M. S. Foster, pp.84–92. Smithsonian Institution Press, Wash-ington, D.C.

D’Antonio, C. M., and P. M. Vitousek. 1992.Biological invasions by exotic grasses, the grassfire cycle, and global change. Annual Review ofEcology and Systematics 23: 63–87.

Dauble, D. D., and R. H. Gray. 1980. Comparisonof a small seine and a backpack electroshocker toevaluate nearshore fish population in rivers. TheProgressive Fish-Culturist 42: 93–95.

Drost, C. A. 1998. Survey of amphibians andamphibian habitat on the upper Santa Cruz River,Arizona. Unpublished report to the U.S. Bureauof Reclamation.

Ehrlich, P. R., D. S. Dobkin, and D. Wheye. 1988.The birder’s handbook: A field guide to thenatural history of North American birds. Simonand Shuster, New York, NY.

Fellers, G. M., and C. A. Drost. 1994. Samplingwith artificial cover. In Measuring and monitoringbiodiversity: Standard methods for amphinians,edited by W. R. Heyer, M. A. Donnelly, R. W.McDiarmid, L.C . Hayek, and M. S. Foster, pp.146–150. Smithsonian Institution Press, Wash-ington, D.C.

Fernandez, P. J., and P. C. Rosen. 1996. Finalreport: Effects of the introduced crayfishOrconectes virilis on native aquatic herpetofaunain Arizona. Submitted to Arizona Game and FishDepartment, Heritage Fund, Phoenix, AZ.

Fitzgerald, S. D., L. S. Zwick, K. L. Diegel, D. E.Berry, S. V. Church, J. G. Sikarskie, J. B.Kaneene, and W. M. Reed. 2003. Experimentalaerosol inoculation of Mycobacterium bovis inNorth American opossums (Didelphis virginiana).Journal of Wildlife Diseases 39: 418–423.

Gibbons, J. M., and R. D. Semlitsch. 1981. Terres-trial drift fence with pitfall traps: An effectivetechnique for quantitative sampling of animalpopulations. Brimleyana 7: 1–6.

Ginger, S. M., E. C. Hellgren, M. A. Kasparian, L.P. Levesque, D. M. Engle and D. M. Leslie. 2003.Niche shift by Virginia opossum following reduc-tion of a putatice competitor, the raccoon. Journalof Mammalogy 84: 1279–1291.

Halvorson, W. L. and P. Guertin. 2003. USGSWeeds in the West project: Status of introducedplants in southern Arizona Parks. U.S. GeologicalSurvey, Southwest Biological Science Center,Sonoran Desert Research Station, University ofArizona, Tucson, AZ.

Hayes, M. P., and M. R. Jennings. 1986. Decline ofRanid frog species in western North America: Are

bullfrogs (Rana catesbeiana) responsible? Journalof Herpetology 20: 490–509.

Hoffmeister, D. F. 1986. Mammals of Arizona.University of Arizona Press, Tucson, AZ.

Holdich, D. M., and R. S. Lowery. 1988. Freshwatercrayfish: Biology, management, and exploitation.Timber Press, Portland, OR.

Integrated Taxonomic Information System (ITIS).2004. Accessed on 20 March 2004.http://www.itis.usda.gov/index.html.

Ivanyi, C., J. Perry, T. R. Van Devender, and H.Lawler. 2000. Reptile and amphibian accounts. InA natural history of the Sonoran Desert, edited byS. J. Phillips and P. W. Comus, pp. 533–585.Arizona-Sonora Desert Museum Press, Tucson.

Karpiscak, M. M., C. P. Gerba, P. M. Watt, K. E.Foster, and J. A. Falabi. 1996. Multi-species plantsystems for wastewater quality improvements andhabitat enhancement. Water Science and Tech-nology 33: 231–236.

Karpiscak, M. M., L. R. Whiteaker, J. F. Artiola, andK. E. Foster. 2001. Nutrient and heavy metaluptake and storage in constructed wetland systemsin Arizona. Water Science and Technology 44:455–462.

Kiesecker, J. M., and A. R. Blaustein. 1997. Popu-lation differences in responses of red-legged frogs(Rana aurora) to introduced bullfrogs. Ecology78: 1752–1760.

King, K. A., B. J. Zaun, and A. L. Velasco. 1999.Contaminants as a limiting factor of fish andwildlife populations in the Santa Cruz River,Arizona. U.S. Fish and Wildlife Service Region 2Contaminants Program, Phoenix, AZ.

Knopf, F. L., and F. B. Samson. 1994. Scaleperspectives on avian diversity in western riparianecosystems. Conservation Biology 8: 669–676.

Kucera, T. E., and R. H. Barrett. 1993. The Trail-master camera system for detecting wildlife.Wildlife Society Bulletin 21: 505–508.

Lawler, S. P., D. Dritz., T. Strange, and M. Holyoak.1999. Effects of introduced mosquitofish and bull-frogs on the threatened California red-legged frog.Conservation Biology 13: 613–622.

Lee, L. C., T. A. Muir, and R. R. Johnson. 1989.Riparian ecosystems as essential habitat forraptors in the American West. In Western raptormanagement symposium and workshop, edited byB. G. Pendleton, pp. 15–26. National WildlifeFederation, Washington, D.C.

McGarigal, K., S. Cushman, and S. Stafford. 2000.Multivariate statistics for wildlife and ecologyresearch. Springer-Verlag, New York, NY.

McIntyre, D. L., and J. R. Weeks. 2002. Environ-mental impacts of illegal immigration on theCleveland National Forest in California. TheProfessional Geographer 54: 392–405.

McLaughlin, S. P., E. L. Geiger, and J. E. Bowers.2001. Flora of the Appleton-Whittell ResearchRanch, Northeastern Santa Cruz County, Arizona.Journal of the Arizona-Nevada Academy of


Science 33: 113–131.Meffe, G. K. 1985. Predation and species replace-

ment in southwestern fishes: A case study.Southwestern Naturalist 30: 173 –187.

Mills, G. S., J. B. Dunning, Jr., and J. M. Bates.1989. Effects of urbanization of breeding birdcommunity structure in southwestern desert habi-tats. Condor 91: 416–428.

Minckley, W. L. 1973. Fishes of Arizona. ArizonaGame and Fish Department, Phoenix, AZ.

Minckley, W. L., and J. E. Deacon, eds. 1991. Battleagainst extinction: Native fish management in theAmerican West. University of Arizona Press,Tucson, AZ.

Mouat, D. A., S. J. Walker, and B. D. Treadwell.1977. The Tumacácori Mission National Monu-ment floral inventory and vegetation map project.Office of Arid Lands Studies, Applied RemoteSensing Program, University of Arizona, Tucson,AZ.

Naeem, S., L. J. Thompson, T. H. Jones, J. H.Lawton, S. P Lawler, and R. M. Woodfin. 1996.Changing community composition and elevatedCO2. In Carbon dioxide, populations, and commu-nities, edited by C. Korner, and F. A. Bazzaz, pp.93–100. Academic Press, San Diego, CA.

National Park Service (NPS). 1996. Generalmanagement plan and environmental impact state-ment: Tumacácori National Historical Park. U.S.Dept. of Interior, National Park Service, Wash-ington, D.C.

National Park Service (NPS). 2001. Baseline waterquality data inventory and analysis (BWQDIA),Tumacácori National Historical Park. U.S. Dept.of Interior, Water Resources Division andServicewide Inventory and Monitoring Program,Fort Collins, CO.

National Park Service (NPS). 2003. Resource issuesin southern U.S. border parks from drug traf-ficking and undocumented alien activity.Unpublished report, April 2003. Southern ArizonaOffice, National Park Service, Phoenix, AZ.

North American Ornithological Atlas Committee(NAOAC). 1990. Handbook for atlasing NorthAmerican breeding birds, edited by C. Smith.Accessed 13 July 2001 from: http://american-birding.org/norac/ atlascont.htm.

O’Dell, D. 2003. Effects of trampling on smallmammal, bird, and vegetation communities.Unpublished M.S. thesis proposal. School ofNatural Resources, University of Arizona, Tucson,AZ.

Office of Technology Assessment (OTA). 1993.Harmful non-indigenous species in the UnitedStates. OTA-F-565, U.S. Government PrintingOffice, Washington, D.C.

Osborne, L. L., and D. A. Kovacic. 1993. Riparianvegetated buffer strips in water-quality restorationand stream management. Freshwater Biology 29:243–258.

Peterson, B. L., B. E. Kus, and D. H. Deutschman.

2004. Determining nest predators of the leastBell’s vireo through point counts, tracking stationsand video photography. Journal of FieldOrnithology 75: 89–95.

Powell, B. F. 2000. Results of yellow-billed cuckoosurveys adjacent to Tumacácori National Histor-ical Park in Arizona: A report on the 2000breeding season. U.S. Geological Survey, SonoranDesert Research Station, University of Arizona,Tucson, AZ.

Powell, B. F., E. W. Albrecht, W. L. Halvorson, C.A. Schmidt, P. Anning, and K. Docherty. 2005.Vascular plant and vertebrate inventory ofTumacácori National Historical Park. USGS OpenFile Final Report 2005-1142. U.S. GeologicalSurvey. Southwest Biological Science Center,Sonoran Desert Research Station, University ofArizona, Tucson, AZ.

Powell, B. F., and R. J. Steidl. 2000. Nesting habitatand reproductive success of southwestern riparianbirds. Condor 102: 823–831.

Price, M. V. 1978. The role of microhabitat in struc-turing desert rodent communities. Ecology 59:10–921.

Ralph, C. J., J. R. Sauer, S. Droege, technical eds.1995. Monitoring bird populations by pointcounts. General Technical Report PSW-GTR-149.Pacific Southwest Research Station, ForestService, U.S. Department of Agriculture, Albany,CA.

Recon. 2004. Advance draft, Pima County multiplespecies conservation plan, Pima County, Arizona.Recon report number 3273B.

Reynolds, R. T., J. M. Scott, and R. A. Nussbaum.1980. A variable circular-plot method for esti-mating bird numbers. Condor 82: 309–313.

Riley, S. P. D., R. M. Sauvajot, T. K. Fuller, E. C.York, D. A. Kamradt, C. Bromley, and R. K.Wayne. 2003. Effects of urbanization and habitatfragmentation on bobcats and coyotes in southernCalifornia. Conservation Biology 17: 566–576.

Schwalbe, C. R., and P. C. Rosen. 1988. Prelimi-nary report on effects of bullfrogs on wetlandherpetofauna in southeastern Arizona. In Manage-ment of amphibians, reptiles, and small mammalsin North America, edited by R. C. Szaro, K. E.Severson, and D. R. Patton, pp. 166–173. GeneralTechnical Report RM-166, U.S. Department ofAgriculture, Rocky Mountain Research Station,Fort Collins, CO.

Schweitzer, S. H., D. M. Finch, and D. M. Leslie,Jr. 1998. The brown-headed cowbird and itsriparian dependant hosts in New Mexico. GeneralTechnical Report RMRS-GTR-1. U.S. Depart-ment of Agriculture, Forest Service, RockyMountain Research Station, Fort Collins, CO.

Skagen, S. K., C. P. Melcher, W. H. Howe, and F.L. Knopf. 1998. Comparative use of riparian corri-dors and oases by migrating birds in southeasternArizona. Conservation Biology 12: 896–909.

Sprouse, T., R. Emanuel, B. Tellman. 2002. Final


report: Surface water quality monitoring overviewand assessment for the Sonoran Desert Network,National Park Service. Unpublished report. WaterResources Research Center, University ofArizona, Tucson, AZ.

Stamp, N. E., and R. D. Ohmart. 1979. Rodents ofdesert shrub and riparian woodland habitats of theSonoran desert. Southwestern Naturalist 24: 279–289.


Strong, T. R., and C. E. Bock. 1990. Bird speciesdistribution in riparian habitats in southeasternArizona. Condor 92: 866–885.

Sureda, M., and M. L. Morrison. 1999. Habitatcharacteristics of small mammals in southeasternUtah. Great Basin Naturalist 59: 323–330.

Theobald, D. M., J. R. Miller, and N. T. Hobbs.1997. Estimating the cumulative effects of devel-opment on wildlife habitat. Landscape and UrbanPlanning 39: 25–36.

Thomas, J. W., C. Maser, and J. E. Rodiek. 1979.Wildlife habitats in managed rangelands, theGreat Basin of southeastern Oregon: Riparianzones. General Technical Report PNW-80. U.S.Department of Agriculture, Forest Service,Arcata, CA.

Thompson, S. K. 1992. Sampling. John Wiley andSons, New York, NY.

Turner, W. 2003. Bird banding at TumacácoriNational Historical Park, AZ Draft. Unpublishedreport to the National Park Service, SouthernArizona office, Phoenix, AZ.

United States Department of Agriculture (USDA).2004. The PLANTS Database, Version 3.5 NationalPlant Data Center, Natural Resources ConservationService, Baton Rouge, LA. Accessed 1 February2004: (http://plants.usda.gov).

United States Environmental Protection Agency(USEPA). 2001. National primary drinking waterstandards. EPA 816-F-01-007.

Voeltz, J. B., and R. H. Bettaso. 2003. 2003 Statusof the Gila topminnow and desert pupfish inArizona. Nongame and Endangered WildlifeProgram Technical Report 226. Arizona Game andFish Department, Phoenix, AZ.

Western Regional Climate Center (WRCC). 2004.Arizona Climate Sumaries from TumacácoriNational Historical Park. Accessed 1 March 2004:http://www.wrcc.dri.edu/summary/climsmaz.html.

White, G. C., K. P. Anderson, K. P. Burnham, andD. L. Otis. 1983. Capture- recapture and removalmethods for sampling closed populations. LosAlamos National Laboratory, Los Alamos, NM.

PLANT AND VERTEBRATE INVENTORY OF ORGAN PIPE CACTUS NATIONAL MONUMENT

Cecilia A. Schmidt and Brian F. Powell

One of the goals of the National Park Service(NPS) Inventory and Monitoring Program isproducing up-to-date species lists for plantsand vertebrates in all natural areas managed bythe NPS. In contrast to our work in other parksin the Sonoran Desert Network (e.g., Powellet al. 2005, 2007a), we were able to rely on thefield work done by others to produce thespecies lists for Organ Pipe Cactus NationalMonument (NM). Few natural areas insouthern Arizona have received as muchecological research as Organ Pipe Cactus NM.Bennett et al. (1990) provide an excellentreview of research related to cultural andnatural resources of the monument andsurrounding areas prior to 1981.

The designation of the monument as aUnited Nations Biosphere Reserve in 1976provided important early initiative to scientistsinterested in studying the Sonoran Desert.Biosphere reserves are designated becausethey are thought to represent the mostoutstanding examples of selected ecosystems.Organ Pipe Cactus National Monument’sisolation and geographic location at the centerof the Sonoran Desert provided an excellentlaboratory for research and education. One ofthe first initiatives after the monument’s desig-nation as a Biosphere Reserve was thecompilation of all known information on thenatural and cultural history of the monument.

In 1986, monument staff gathered regionalexperts to help create the first inventory andmonitoring program in the region. Modeledafter the Channel Islands’ Inventory & Moni-

toring Initiative (Davis and Halvorson 1988),the monument’s Sensitive EcosystemsProgram (SEP) was designed to determine: (1)the condition of the monument’s ecosystems,(2) alternatives available for ecosystemmanagement, and (3) the effectiveness ofimplemented action programs. Initially, theSEP program included a broad range of naturalresource studies at the monument, specificallybaseline inventories of plants, songbirds, andnocturnal rodents (Bennett and Kunzman1987). The program was expanded in 1991 toimplement some of the recommended long-term monitoring protocols. Finally, in 1994,the title of the program changed to the Ecolog-ical Monitoring Program (EMP) to reflect achange from the historic focus on sensitivemonument areas to a broader look at theecosystem's many components (NBS 1995).Prior to the initiation of the NPS Inventory andMonitoring Program (NPS 1992; of which theSonoran Desert Network is one program), theOrgan Pipe Cactus NM EMP program was oneof the most extensive ecological research andinventorying and monitoring programs in theNational Park Service. Because of earlyinterest in the monument by ecologists, themonument had a fairly complete list of plantsand vertebrate species, well ahead of otherpark units in southern Arizona.

For complete details of the Organ PipeCactus NM study see Schmidt et al. (2007).Scientific and common names used throughoutthis chapter are current according to acceptedauthorities for each taxonomic group: Stebbins

42 Schmidt and Powell

(2003) for amphibians and reptiles; AmericanOrnithologist Union (AOU; 1998, 2003) forbirds; Baker et al. (2003) for mammals; andRutman (2005) for plants.

MONUMENT OVERVIEWMonument Area and History

Organ Pipe Cactus NM is located near thecenter of the Sonoran Desert, in southwesternArizona on the border with Mexico (Figure 1).The monument was established in 1937 topreserve the largest portion of desert in theUnited States with the park’s namesake, theorgan pipe cactus. Ninety-six percent of themonument is designated wilderness. Themonument was designated as a BiosphereReserve in 1976 by the United Nations Educa-tional, Scientific and Cultural Organization(UNESCO). This designation signifies that themonument contains an outstanding, interna-tionally significant ecosystem. The UNESCOdesignation prompted the NPS to interpret themanagement objectives for the monument aspreserving the monument as a representativeexample of the natural and cultural resourcesof the Sonoran Desert and serving as a naturallaboratory for understanding and managingSonoran Desert ecosystems (NPS 1994a).

At 133,830 ha, it is the largest park unit inthe Sonoran Desert Network, yet it is dwarfedby the major land management unitssurrounding it: Cabeza Prieta National WildlifeRefuge to the west and north, the TohonoO’odham Indian Reservation to the east, andBureau of Land Management land to the north(Figure 1). In Mexico, El Pinacate y GranDesierto de Altar (also a designated UNESCOBiosphere Reserve) borders the monument tothe south.

Archaeological evidence suggests thathumans occupied the monument as far back as12,000 years ago (Rankin 1991). QuitobaquitoSprings has been an active site for settlementin recent history, and it was an importantsource of water for Spanish explorers andmigrants attempting to cross the SonoranDesert (Bennett and Kunzman 1989). Today,there are a number of sites at the monument

that are sacred to the Tohono O’odham. Live-stock grazing was the livelihood for a numberof families who lived in the area, but wasdiscontinued in 1976 (NPS 1997). Prior to thecreation of the monument there werenumerous active mining claims.

Physiography, Geology and SoilsOrgan Pipe Cactus NM is located in the Basinand Range Physiographic Province and itstopography varies from deep alluvial valleysto steep, rugged mountain ranges. Elevation atthe monument is as low as 305 m in the westand extends to 1,465 m at its eastern boundaryatop the Ajo Mountains. Geology of the moun-tains is the result of volcanic flows during theCretaceous, Tertiary and Plio-Pleistoceneperiods. The valleys of the monument areformed of alluvial material originating in themountains and transported down via streamsand sheet-flow (Warren et al. 1981). Soils atthe monument are all classified as aridisols(Chamberlin 1972).

HydrologyThere are no perennial rivers or streams withinthe monument, though there are 11 springs,four with perennial flow. The most prominentspring, Quitobaquito, feeds a large human-made pond. Two of the other perennial springsand five of the intermittent springs occur nearthe Quitobaquito area. There are 60 tinajas(natural depressions in bedrock that holdwater) throughout the monument and they arethe most widespread source of seasonal water.

ClimateOrgan Pipe Cactus NM experiences an annualbimodal pattern of precipitation which is char-acterized by heavy summer (monsoon) stormsbrought about by moisture coming from theGulf of Mexico, and less intense frontalsystems coming from the Pacific Ocean in thewinter. On average, approximately one-half ofthe annual precipitation falls from July throughSeptember (WRCC 2005). The area’s hotseason occurs from May through September;maximum temperatures in July can exceed

Schmidt and Powell 43

40° C. Winter temperatures rarely dip belowfreezing. Average annual precipitation for themonument is 243 mm.

VegetationAccording to Warren et al. (1981), the monu-ment has six plant community types: • Great Basin conifer woodland containing

oneseed juniper–Arizona rosewood mixedshrub association;

• Madrean evergreen forest and woodlandcontaining Ajo Mountain scrub oak mixedshrub association;

• Sonoran riparian woodland containinghoney mesquite riparian woodland associ-ation;

• Interior chaparral containing rock goose-berry–common hoptree mixed scrubassociation;

• Sonoran desertscrub containing 23 associ-ations dominated by creosote bush,burrobush, ragweed, ocotillo, palo verde,jojoba, ironwood, acacia, organ pipe

cactus, and saltbush;• Sonoran interior marshland containing

southern cattail–chairmaker’s bulrushassociation and inland saltgrass–rushassociation. Natural Resource Management Issues

Border Crossings Fifteen years ago, the trespass of drug smug-glers and undocumented immigrants (bordercrossers) across the U.S./Mexico border wasnot considered a natural resource managementissue (NPS 1994a). Today, this issue is one ofthe greater challenges to the ecologicalintegrity of the monument. It is estimated that500 border crossers enter the U.S. through themonument each day and approximately700,000 pounds of drugs are brought throughthe monument each year (NPS 2003a). Bordercrossers pose a serious threat to visitors,employees, and personal property at the monu-ment. They also create a network of trails,leave trash, and damage soils and vegetation.

Figure 1. Location of Organ Pipe Cactus NM in relation to other parks in the Sonoran DesertNetwork of parks.


There is concern that excessive use of themajor springs and water has led to watercontamination (Sprouse et al. 2002).

The movement of illegal vehicles crossinginto the monument and subsequent borderpatrol vehicle pursuits has caused severedamage to fragile desert resources. Respondingto this threat, the NPS constructed a vehiclebarrier across most of the monument’s borderwith Mexico. The barrier will likely preventsome animal movement, but the anticipateddrop in off-road vehicle traffic and associatedimpacts is expected to allow improvements insoil stability and habitat for plants and animals(NPS 2003b).

Vertebrate Mortality along ArizonaHighway Route 85 which runs north/ southand dissects the monumentRoads are a common source of vertebratemortality and they act as barriers to thedispersal of small mammals creating subpop-ulations; deter, disturb and alter movementpatterns of songbirds and medium and largemammals near roads; and pollute soil andwater via runoff. Roads may also attract manyvertebrates, ultimately leading to them beingkilled by vehicles. Herbaceous plant speciesthriving along roadsides from runoff attractgranivorus birds and small mammals and roadsurfaces provide warmth to ectothermicreptiles and amphibians (Oxley et al. 1974;Adams and Geis 1983; Forman and Alexander1998; Trombulak and Frissell 2000; Rosen andLowe 1994, 1996).

Animal Poaching and CollectionOrgan Pipe Cactus NM has several species ofplants and vertebrates that are of interest toillegal collectors and poachers. Many plantsare of value for landscape purposes and manywoody plant species are scavenged by bordercrossers for fire and shelter (NPS 2003a).Many species of reptiles, such as the rosy boa,Gila monster, chuckwalla, sidewinder, tigerrattlesnake, and desert tortoise found in themonument are collected for personal collec-

tions or for the pet trade (Rosen and Lowe1996).

Aircraft Noise Low-flying military aircraft from Luke AirForce Base, law enforcement aircraft from theU.S. Border Patrol, and private aircraft passover the monument often (NPS 1994b). Bothvibrations and noise generated by these aircraftaffect the natural quiet of the monument andmay also affect wildlife in the area. Aircraftoverflights can produce changes in the physi-ology and behavior of some wildlife species(Luz and Smith 1976; Weisenberger et al.1996).

METHODSPlants

The monument’s plant list is the most completeof any large natural area in the desert south-west and is the result of many studies. The firstknown species list for the monument wascreated by McDougall (1945) and was basedon his collections and those of A. A. Nicholsfrom 1939 (Rutman 2005). Other early collec-tions were made by Ora Clark, a scienceteacher in Ajo during the late 1930s and early1940s. Specimens from these collections stillremain at the monument and/or at the Univer-sity of Arizona Herbarium. There were threesubsequent, unpublished lists for the monu-ment (reviewed in Pinkava et al. 1992) butBowers (1980) produced the first annotatedlist. She reported 518 species. Other additionsto the flora included Pinkava et al. (1992) andFelger et al. (1992). Principle works on thenon-native plants of the monument includeFelger (1990) who reviewed plant specimensfrom four herbaria and his own field observa-tions and Halvorson and Guertin (2003) whoprovided location information for 17 speciesof non-native plants. Ruffner Associates (1995)identified 44 species of plants that were in needof monitoring because of “sensitivity to distur-bance” or because they were deemed to beindicators of change.

There have been a number of vegetationstudies in the monument. Steenberg and


Warren (1977) constructed exclosures at foursites to determine the impacts of grazing onplant community structure and composition.These exclosures were resurveyed by Warrenand Anderson (1995). Warren et al. (1981)classified and mapped dominant (perennial)vegetation at the monument. Bowers (1990)reviewed past studies from the region,provided a historical context for vegetationchanges, and used repeat photography to illus-trate some important changes. Brown andWarren (1986) plotted the location and calcu-lated density of riparian vegetation in andaround Quitobaquito Springs and Pond. Parker(1991) established vegetation and environ-mental relationships at 100 sites throughout themonument. Lowe et al. (1995) designed thevegetation monitoring protocol for the EMP.

Rutman (2005) compiled a list of speci-mens in herbaria, at the monument, ArizonaState University (ASU), and the University ofArizona (UA). All specimens were reviewedby experts and those excluded from the monu-ment’s official list were species that wereobviously transitory as no reproducing popu-lations were established. The current list is theresult of dozens of collectors working overapproximately 75 years and represents one ofthe most complete plant lists of any area in theregion.

Rutman (2005) primarily used Flora ofNorth America (FNA 1993) as a taxonomicreferences for her list, but also used W3Trop-icos (MGB 2004), the PLANTS Database(USDA 2004), and individual publications.

FishThe Quitobaquito pupfish (Cyprinodoneremus) is the only species of fish at the monu-ment and it occurs only at QuitobaquitoSprings and Pond and a few sites outside of themonument (Hendrickson and Romero 1989).Listed as endangered in 1986 under the Endan-gered Species Act, the Quitobaquito pupfish isone of several species of pupfish that wereonce found throughout the Gila Riverdrainage, lower Colorado River and Delta, andthe Imperial Valley in California (Miller 1990).

Most of these populations are now extinct,presumably because of habitat destruction(Pearson and Conner 2000).

Monument personnel monitor populationsize annually at Quitobaquito Springs andPond as part of the monitoring program (NPS1998a, 1998b; Tibbitts 1999; Pearson andConner 2000). The pupfish appears to be doingwell at Quitobaquito; in the last 25 years thepopulation has never dipped below 1,800 indi-viduals (Pearson and Conner 2000) and iscurrently thought to consist of approximately8,000 to 10,000 individuals (Douglas et al.2001). A concern to the long-term persistenceof the Quitobaquito pupfish is the potentialintroduction of non-native fish and other verte-brates, invertebrates, and plants (Pearson andConner 2000).

Amphibians and ReptilesLowe (1990) provides an excellent summaryof early amphibians and reptile studies andcollections at Organ Pipe Cactus NM. Morerecently, Rosen and Lowe (1996) conductedan inventory and established a long-termmonitoring program. In this inventory, theyhave created the most definitive, up-to-datespecies list for the monument. We base ouramphibian and reptile species list on this listand on a study of museum specimens(Schmidt et al. 2007). The list by Rosen andLowe (1996) was created using many of theprevious lists, studies, and collections, andover 600 field days of their own research.Most of the species on the list are backed byvoucher specimens located at the monumentand the UA Amphibian and Reptile Collec-tion. Although this is a very complete list,Rosen and Lowe (1996) believe that otherspecies may be found in the monument withadditional work.

The monument has likely experienced lossof species in the last few decades. Two nativespecies, the Mexican spadefoot and yellowmud turtle were previously documented in themonument but are believed to no longer bepresent (Rosen and Lowe 1996). Four non-native species — tiger salamander, American


bullfrog, painted turtle, and pond slider —have also been documented at the monumentbut no longer occur there.

BirdsSteenbergh and Hoy (1963) created the firstspecies list for the monument; it summarizedobservations made by researchers and thegeneral public from 1939 to 1963. Subsequentlists included: Cunningham (1969, 1971), thefirst annotated list by Wilt (1976), and Brownet al. (1985). The most comprehensive anno-tated list was by Groschupf et al. (1988) andlater revised by Tibbitts and Dickson (2005).We based our bird species list on these lists, ona study of museum specimens (Schmidt et al.2007), and on the report by Benson et al.(2001). This list represents one of the mostthoroughly documented bird species lists of anyin the region. Like the lists for plants andherpetofauna, the bird list is an outstandingexample of one built on past efforts with peri-odic updates.

The bird list is one of the most complete listsof its kind in the region. In the 17 years sincethe excellent work by Groschupf et al. (1988),only 11 species were added to the list (Tibbittsand Dickson 2005). This indicates that the birdspecies list is nearly complete. However,because birds are highly mobile animals, it isdifficult to compile a truly complete list, espe-cially for Organ Pipe Cactus NM, which is wellknown for species that seldom enter the U.S.from Mexico. Also, it is likely that even morebirds requiring open water will be found atQuitobaquito Pond because of its proximity tothe Gulf of California.

MammalsMearns (1907) was the first collector at themonument. He and others collected vertebratesat Quitobaquito and others areas aroundSonoyta, Mexico in 1894. Huey (1942) wasthe first to report on the mammals fromthroughout the monument and to document aspecies list. Steenberg and Warren (1977)quantified vegetation characteristics andtrapped small mammals in grazed and

ungrazed areas of the monument to establishthe effects of livestock. Other rodent-trappingefforts included establishment of trapping gridsas part of the monitoring program (Petryszyn1995; NPS 1998a, 1998b; Petterson 1999) andassociated programs (Rosen 2000). Bats arealso well surveyed at the monument. Cockrum(1981) trapped bats in 1979 and 1980 andreviewed past information to create a specieslist of bats for the monument. Petryszyn andCockrum (1990) trapped at Quitobaquito Pondin 1981 and 1982 and created a species list ofother mammals they found there.

The list of the monument’s mammals isbased on Cockrum and Petryszyn (1986), withadditions from the inventory and monitoringprogram (NPS 1998a, 1998b; Pate 1999).Based on the list of species and the many yearsof mammal surveys, most of the mammalspecies that occur at the monument have beenrecorded. There is one species, the feral burro(Equus asinus), which occurred at the monu-ment in the recent past but is no longer present.


There have been 642 species of plants foundat the monument (Table 1), of which 55 (9%)are non-native. Compared to other NPS areasin southern Arizona, the monument’s flora isnot particularly species rich (Bowers 1980).For example, Powell et al. (2007b) foundalmost the same number of species (638) atFort Bowie National Historic Site in south-eastern Arizona, an area < 0.5% the size ofOrgan Pipe Cactus NM. Bowers (1980)provides similar comparisons to other flora insouthern Arizona. However, within the monu-ment there are areas of high species richness,most notably Quitobaquito Springs and Pondand the Ajo Mountains. Bowers (1980) found163 species only in the Ajo Mountains, whichcomprise about 10% of the area of the monu-ment. This high species richness is primarilydue to topographic relief, soil-texture changesand gradients in temperature and rainfall(Bowers 1980; Parker 1991). In addition, anumber of species reach the westernmost limit


of their geographic ranges in the Ajo Moun-tains, including some with distinctly Madreanaffinities (Bowers 1980).

The number of non-native plant speciesrecorded in the monument (n = 55, 9% of allspecies) is low, although slightly higher thanSaguaro National Park (approximately 7%;Powell, et al. 2006, 2007a), which has thelowest percent of non-native species in theSonoran Desert Network. Non-native plantsare a management concern because they alterecosystem function and processes (Naeem etal. 1996; D'Antonio and Vitousek 1992),reduce abundance of native species, and causepotentially permanent changes in diversity andspecies composition (Bock et al. 1986; D'An-tonio and Vitousek 1992; OTA 1993).However, some species have stronger impactson the ecological community than others. Inassessing the potential threat posed by non-native species, it is important to consider thespatial extent of species, particularly thosespecies that have been identified as “invasive”or of management concern. Felger (1990)found 14 species, including red brome andbuffelgrass, to be “thoroughly” invasive andan additional 10 species, including smoothbarley, crimson fountaingrass, and commonsowthistle, that have become established ondisturbed sites.

Amphibians and Reptiles. There are 49 species of amphibians andreptiles that are known to occur at the monu-

ment: five toads, two turtles, 16 lizards, and 26snakes. There were no non-native herpetofaunaspecies found to breed at the monument.Unlike plants, there is a high diversity ofherpetofauna at the monument, related to itsrelatively large size and biophysical variety.Reptiles are well represented at the monument,particularly lizards and snakes. Rosen andLowe (1996) assert that dominant physicalfeatures of the monument’s geology and soilsseparate the lizards and snakes into threecommunities: (1) rock piles, (2) bajadas, and(3) valley-bottom fills. Two “true” desertspecies inhabit only rock piles: commonchuckwalla and speckled rattlesnake. Bycontrast, six species (including: desert hornedlizard, western shovel-nosed snake, andsidewinder) inhabit the valley-bottom fills(containing fine-textured soils) where thevegetation is dominated by creosote bush.Finally, the bajadas contain some species asso-ciated with Arizona Upland SonoranDesertscrub: tree lizard, regal horned lizard,and Sonoran shovel-nosed snake (Rosen andLowe 1996). Another important communitywithin the monument is xeroripariandesertscub along washes, which hosts anumber of species such as the western coral-snake, and common kingsnake. Rosen andLowe (1996) noted that washes become partic-ularly important during droughts when speciesfrom adjacent areas use the washes more thanin times of normal rainfall.

Rosen and Lowe (1996) created a list of

Table 1. Summary results of the vascular plant and vertebrate inventories at Casa Grande Ruins NM.

Taxonomic group Number of species Non-native species

Plants 642 55Fish 1 0Amphibians and reptiles 49 0Birds 285 3Mammals 54 1Totals 1,031 59


species that they considered threatenedbecause of (1) range-wide or local populationdecline, (2) potential for poaching, and/or (3)susceptibility to mortality on Arizona HighwayRoute 85. This list includes the desert tortoise(a federal Species of Concern and an Arizonastate Wildlife Species of Concern) and tigerrattlesnake, species that are targeted by collec-tors. The rosy boa (a federal Species ofConcern and a Sensitive species according toBLM) and the Sonoran shovel-nosed snake (aSensitive species according to the USFS) bothare also targeted by collectors and are oftenkilled on roadways. The Sonoran mud turtle isprobably undergoing a population decline andit is restricted only to Quitobaquito Pond. Thecanyon spotted whiptail, a federal Species ofConcern, has its largest known population inthe Ajo Mountains, which are only partiallywithin the protective borders of the monument.The following species have restricted distri-butions, isolated populations, populationcenters off the monument, or are uncommon:Sonoran green toad, longtailed brush lizard,desert horned lizard, black-necked gartersnake, southwestern black-headed snake,speckled rattlesnake, Sonoran whipsnake, andwestern shovel-nosed snake. The one addi-tional species not listed by Rosen and Lowe(1996) that is federally listed as a Species ofConcern is the common chuckwalla. Theyforesaw no immediate threat to this species.

BirdsThere have been 285 species of birds recordedat the monument. Of these only three are non-native. Organ Pipe Cactus NM has the highestbird diversity of any unit in the Sonoran DesertNetwork. This diversity results from threemain factors. First, the monument has hadextensive surveys and observations over thepast century, which has enabled the monumentto have a near complete species list. Thesecond factor is that many species have theirnorthernmost distribution at the monument(i.e., crested caracara). The third factor deter-mining the diversity of birds at the monumentis the variety of biophysical situations within

the monument, from mixed Sonorandesertscrub in the western flatlands to thejuniper-oak woodland/mixed mountainscrubin the eastern highlands to marsh and openwater (Rosenberg et al. 1991).

Quitobaquito provides an oasis of openwater and marsh in an area otherwise devoidof surface water. This important resourceattracts birds requiring open water and alsohosts many migrants en route to more northernor southern wintering or summering areas.Seventy-three have been documented at Quito-baquito Pond: 21 species of ducks and geese,four species of grebe, seven species of heronand egret, five species of rail, 19 species ofshorebirds, nine species of gulls, and eightother species. Several species that use this openwater and marsh are federally listed as Endan-gered or Species of Concern including thewood stork, brown pelican and white-facedibis. Although extremely rare at the monument,they have been found at Quitobaquito Pond.

Another important resource for birds are thexeroriparian areas along washes such as AlamoCanyon and Growler Wash. Hardy et al. (2004)surveyed ecologically similar areas north of themonument and found that most of the springpassage migrant species preferentially selecteddry washes, and many species used themexclusively. Also, many of the species thatbreed at the monument prefer the xeroriparianwashes compared to upland sites, presumablybecause washes provide cooler microsites andprotection from predators (Parker 1986).

MammalsThe current list of mammals for Organ PipeCactus NM consists of 54 species: 14 bats, 20small, terrestrial mammals (principallyrodents) and 20 medium to large mammals.Also included in this list is one non-nativespecies, the feral dog. Quitobaquito plays animportant role in the high mammal diversity atthe monument by providing the largest sourceof perennial water in the region. Quitobaquitois an important resource for bats that use openwater to hunt for insects and is the only site inthe monument were the desert shrew is located


(Cockrum and Petryszyn 1986). Other impor-tant resources are the night-blooming cacti,including organ pipe and saguaro, whichprovide nectar for the endangered southern(lesser) long-nosed bat.

There are two large mammals that arefound at the monument which are uncommonin Arizona: the Sonoran pronghorn and desertbighorn sheep. The pronghorn is believed tobe an occasional visitor to the monument andthe desert bighorn sheep is found in very smallnumbers in the Diablo, Puerto Blanco and Ajomountains. Both species may occasionally befound at Quitobaquito.

ACKNOWLEDGMENTSThanks to Organ Pipe Cactus NM Superin-tendent Kathy Billings, Chief of ResourcesMary Kralovec and all the staff at Organ PipeCactus National Monument for their support ofour program. This project resulted from thecollaboration of many people at the NationalPark Service, and was facilitated by the DesertSouthwest Cooperative Ecosystem StudiesUnit (CESU). Andy Hubbard and Larry Norris(NPS) provided administrative support for theproject. Additional administrative support wasprovided by Cecily Westphal of the School ofNatural Resources at the University of Arizona.

This report summarizes numerous researchprojects over the last 75 years. More recently,a number of people have been instrumental incollecting data in the monument and helpingus gather data for this report including: DennisCasper, Charles Conner, Peter Holm, AmiPate, Sue Rutman, and Tim Tibbitts. RichardFelger and Sue Rutman have been stalwartkeepers and revisers of the plant list and we aregrateful for their diligence and years of hardwork. Philip Rosen and Charles Lowe deservemuch of the credit for the amphibian andreptile list. Debbie Angell, Pamela Anning, andRyan Reese assisted with data entry and data-base design for this project. We receivedhelpful reviews of the manuscript from PeterHolm, Andy Hubbard, Mary Kralovec, LarryLaing, Theresa Mau-Crimmons, Larry Norris,Sue Rutman, and Tim Tibbitts.

LITERATURE CITEDAdams, L. W., and A. D. Geis. 1983. Effects of

roads on small mammals. The Journal of AppliedEcology 20: 403–415.

American Ornithologists’ Union (AOU). 1998.Checklist of North American birds, seventhedition. American Ornithologists’ Union and AllenPress Inc., Lawrence, KS.


Baker, R. J., L. C. Bradley, R. D. Bradley, J. W.Dragoo, M. D. Engstrom, R. S. Hoffmann, C. A.Jones, F. Reid, D. W. Rice, and C. Jones. 2003.Revised checklist of North American mammalsnorth of Mexico, 2003. Occasional Papers of theMuseum of Texas Tech University 229: 1–23.

Barnett, J. J., and D. Sharrow. 1992. Water resourcesmanagement plan for Organ Pipe Cactus NationalMonument. U.S. Department of the Interior,National Park Service, Western Region, Denver,CO.

Bennett, P. S., R. R. Johnson, and M. M. McCarthy.1990. Assessment of scientific information andactivities at Organ Pipe Cactus National Monu-ment, Biosphere Reserve. Special Report No. 10.Cooperative National Park Resources StudiesUnit, University of Arizona, Tucson, AZ.

Bennett, P. S., and M. R. Kunzmann. 1987. OrganPipe Cactus National Monument, BiosphereReserve, Sensitive Ecosystems Program. SpecialReport No. 7. Cooperative National ParkResources Studies Unit, University of Arizona,Tucson, AZ.

Bennett, P. S., and M. R. Kunzmann. 1989. Ahistory of the Quitobaquito resource managementarea, Organ Pipe Cactus National Monument,Arizona. Technical Report No. 26. CooperativeNational Park Resources Studies Unit, Universityof Arizona, Tucson, AZ.

Benson, L. A., B. Alberti, C. Gordon, A. Whalon,and T. Wood. 2001. Summary report of thesouthern Arizona/Sonora MAPS project for 1997–2000. Unpublished report by the National ParkService, Southern Arizona Office, Phoenix, AZ.

Bock, C. E., J. H. Bock, K. L. Jepson, and J. C.Ortega. 1986. Ecological effects of plantingAfrican lovegrasses in Arizona. NationalGeographic Research 2: 456–463.

Bowers, J. E. 1980. Flora of Organ Pipe CactusNational Monument. Journal of the Arizona-Nevada Academy of Science 15: 1–11, 33–47.

Bowers, J. E. 1990. Vegetation change at OrganPipe Cactus National Monument. In Assessmentof scientific information and activities at OrganPipe Cactus National Monument, BiosphereReserve, edited by P. S Bennett, R. R. Johnson,and M. M. McCarthy, pp. 85–132. Special ReportNo. 10. Cooperative National Park ResourcesStudies Unit, University of Arizona, Tucson, AZ.


Brown, B. 1991. Land use trends surroundingOrgan Pipe Cactus National Monument. TechnicalReport No. 39. Cooperative National ParkResources Studies Unit, University of Arizona,Tucson, AZ.

Brown, B. T., R. R. Johnson, and K. Groschupf.1985. A checklist of birds of Organ Pipe CactusNational Monument. Southwest Parks and Monu-ments Association, Tucson, AZ.

Brown, B. T., and P. L. Warren. 1986. A descriptiveanalysis of woody riparian vegetation at Quito-baquito Springs Oasis, Organ Pipe CactusNational Monument, Arizona. Technical ReportNo. 19. Cooperative National Park ResourcesStudies Unit, University of Arizona, Tucson, AZ.

Carruth, R. L. 1996. Hydrogeology of the Quito-baquito Springs and La Abra Plain Area, OrganPipe Cactus National Monument, Arizona, andSonora, Mexico. Unpublished report by theUnited States Geological Survey, Tucson, AZ.

Chamberlin, E. 1972. Soil survey: A special report;Organ Pipe Cactus National Monument, PimaCounty, Arizona. U.S. Department of Agriculture,Soil Conservation Service, Phoenix, AZ.

Cockrum, E. L. 1981. Bat populations and habitatsat the Organ Pipe Cactus National Monument.Technical Report No. 7. Cooperative NationalPark Resources Studies Unit, University ofArizona, Tucson, AZ.

Cockrum, E. L., editor. 1984. Reference manual:Mammals of the Organ Pipe Cactus NationalMonument. Unpublished report to Organ PipeCactus National Monument, Ajo, AZ.

Cockrum, E. L., and Y. Petryszyn. 1986. Mammalsof the Organ Pipe Cactus National Monument.Special Report No. 5. Cooperative National ParkResources Studies Unit, University of Arizona,Tucson, AZ.

Cunningham, W. 1969. Bird checklist for OrganPipe Cactus National Monument, Arizona.Unpublished list to Organ Pipe Cactus NationalMonument, Ajo, AZ.

Cunningham, W. 1971. Bird checklist for OrganPipe Cactus National Monument, Arizona.Unpublished list to Organ Pipe Cactus NationalMonument, Ajo, AZ.

D'Antonio, C. M., and P. M. Vitousek. 1992.Biological invasions by exotic grasses, the grassfire cycle, and global change. Annual Review ofEcology and Systematics 23: 63–87.

Davis, G. E., and W. L. Halvorson.1988. Inventoryand monitoring of natural resources in ChannelIslands National Park, California. Channel IslandsNational Park, Ventura, CA.

Douglas, M. R., M. E. Douglas, and P.C. Brunner.2001. Population estimates, movements, and sizestructure of the endangered Quitobaquito pupfish,Cyprinodon macularius eremus. SouthwesternNaturalist 46: 141–150.

Felger, R. S. 1990. Non-native plants of Organ PipeCactus National Monument, Arizona. Technical

Report No. 31. Cooperative National ParkResources Studies Unit, University of Arizona,Tucson, AZ.

Felger, R. S., P. L. Warren, L. S. Anderson, and G.P. Nabhan. 1992. Vascular plants of a desert oasis:Flora and ethnobotany of Quitobaquito, OrganPipe Cactus National Monument, Arizona.Proceedings of the San Diego Society of NaturalHistory 8:1–39.

Flora of North America Editorial Committee. 1993.Flora of North America: North of Mexico. OxfordUniversity Press, New York, NY.

Forman, R. T. T., and L. E. Alexander. 1998. Roadsand their major effects. Annual Review of Ecologyand Systematics 29: 207–231.

Groschupf, K. D., B. T. Brown, and R. R. Johnson.1988. An annotated checklist of the birds of OrganPipe Cactus National Monument, Arizona. South-west Parks and Monuments Association, Tucson,AZ.

Halvorson, W. L., and P. Guertin. 2003. USGSWeeds in the West project: Status of introducedplants in southern Arizona parks. USGS, SonoranDesert Research Station, University of Arizona,Tucson, AZ.

Hardy, P. C., D. J. Griffin, A. J. Kuenzi, and M. L.Morrison. 2004. Occurrence and habitat use ofpassage Neotropical migrants in the SonoranDesert. Western North American Naturalist 64:59–71.

Hendrickson, D. A., and A. V. Romero. 1989.Conservation status of desert pupfish, Cyprinodonmacularius, in Mexico and Arizona. Copeia 1989:478–483.

Huey, L. M. 1942. A vertebrate faunal survey of theOrgan Pipe Cactus National Monument, Arizona.Transactions of the San Diego Society of NaturalHistory 32: 353–376.

Lowe, C. H. 1990. A 50-year summary for theherpetofauna of Organ Pipe Cactus NationalMonument, Arizona (1932–1982). In Assessmentof scientific information and activities at OrganPipe Cactus National Monument, BiosphereReserve, edited by P. S. Bennett, R. R. Johnson,and M. M. McCarthy, pp. 63–73. Special ReportNo. 10. Cooperative National Park ResourcesStudies Unit, University of Arizona, Tucson, AZ.

Lowe, C. H., E. B. Wirt, and P. C. Rosen. 1995.Vegetation structure and diversity in naturalcommunities monitoring protocol for the Ecolog-ical Monitoring Program in Organ Pipe CactusNational Monument, Arizona. In Organ PipeCactus National Monument Ecological Moni-toring Program monitoring protocol manual, pp.2-i–2-61. Special Report 11. Cooperative NationalPark Resources Studies Unit, University ofArizona, Tucson, AZ.

Luz, G. A., and J. B. Smith. 1976. Reactions ofpronghorn antelope to helicopter overflight.Journal of Acoustical Society of America 59:1514–1515.


McDougall, W. B. 1945. Check list of the knownplants of Organ Pipe Cactus National Monument.Unpublished list to Organ Pipe Cactus NationalMonument, Ajo, AZ.

Mearns, E. A. 1907. Mammals of the Mexicanboundary of the United States. A descriptive cata-logue of the species of mammals occurring in thatregion; with a general summary of the naturalhistory, and a list of trees. United States NationalMuseum Bulletin 56. Washington, D.C.

Miller, R. R. 1990. Status of the classification of theQuitobaquito pupfish. In Assessment of scientificinformation and activities at Organ Pipe CactusNational Monument, Biosphere Reserve, editedby P. S. Bennett, R. R. Johnson, and M. M.McCarthy, pp. 75–81. Special Report No. 10.Cooperative National Park Resources StudiesUnit, University of Arizona, Tucson, AZ.

Missouri Botanical Gardens (MBG). 2004.W3Trpicos website. http://mobot.mobot.org/W3T/Search/vast.html.

Naeem, S., L. J. Thompson, T. H. Jones, J. H.Lawton, S. P. Lawler, and R. M. Woodfin. 1996.Changing community composition and elevatedCO2. In Carbon dioxide, populations, and commu-nities, edited by C. Korner, and F. A. Bazzaz, pp.93–100. Academic Press, San Diego, CA.

National Biological Service (NBS). 1995. OrganPipe Cactus National Monument ecological moni-toring program monitoring protocol manual.Special Report 11. Cooperative National ParkResources Studies Unit, University of Arizona,Tucson, AZ.

National Park Service (NPS). 1992. NPS-75:Natural resources inventory and monitoring guide-lines. U.S. Department of Interior, Washington,D.C.

National Park Service (NPS). 1994a. Natural andcultural resources management plan for OrganPipe Cactus National Monument. Organ PipeCactus National Monument, Ajo, AZ.

National Park Service (NPS). 1994b. Report oneffects of aircraft overflights on the National Parksystem. Report to Congress. Prepared pursuant toPublic Law 100-91, The National Parks Over-flights Act of 1987.

National Park Service (NPS). 1997. Generalmanagement plan, development concept plan, andenvironmental impact statement. Organ PipeCactus National Monument, Ajo, AZ.

National Park Service (NPS). 1998a. Organ PipeCactus National Monument ecological monitoringprogram annual report 1994. Organ Pipe CactusNational Monument, Ajo, AZ.

National Park Service (NPS). 1998b. Organ PipeCactus National Monument ecological monitoringprogram annual report 1995. Organ Pipe CactusNational Monument, Ajo, AZ.

National Park Service (NPS). 2003a. Resourceissues in southern U.S. border parks from drugtrafficking and undocumented alien activity.

National Park Service, Southern Arizona Office,Phoenix, AZ.

National Park Service (NPS). 2003b. Proposedvehicle barrier environmental assessment. OrganPipe Cactus National Monument/CoronadoNational Memorial. U.S. Department of the Inte-rior, Washington, D.C.

Office of Technology Assessment (OTA). 1993.Harmful non-indigenous species in the UnitedStates. OTA-F-565, U.S. Government PrintingOffice, Washington, D.C.

Oxley, D. J., M. B. Fenton, and G. R. Carmody.1974. The effects of roads on populations of smallmammals. The Journal of Applied Ecology 11:51–59.

Parker, K. C. 1986. Partitioning and foraging spaceand nest sites in a desert shrubland bird commu-nity. The American Midland Naturalist 115:255–267.

Parker, K. C. 1991. Topography, substrate, andvegetation patterns in the northern Sonoran Desert.Journal of Biogeography 18: 151–163.

Pate, A., editor. 1999. Organ Pipe Cactus NationalMonument ecological monitoring program annualreport 1996. National Park Service, Ajo, AZ.

Pearson, G., and C. W. Conner. 2000. The Quito-baquito desert pupfish, an endangered specieswithin Organ Pipe Cactus National Monument:Historical significance and management chal-lenges. Natural Resources Journal 40: 379–410.

Petterson, J. 1999. Nocturnal rodents. In Organ PipeCactus National Monument ecological monitoringprogram annual report 1996, edited by A. Pate,National Park Service, Ajo, AZ.

Petryszyn, Y. 1995. Small nocturnal mammalsmonitoring protocol. In Organ Pipe CactusNational Monument ecological monitoringprogram monitoring protocol manual, pp. 5-i–5-23. Special Report 11. Cooperative National ParkResources Studies Unit, University of Arizona,Tucson, AZ.

Petryszyn, Y., and E. L. Cockrum. 1990. Mammalsof the Quitobaquito Management Area, OrganPipe Cactus National Monument, Arizona. Tech-nical Report No. 36. Cooperative National ParkResources Studies Unit, University of Arizona,Tucson, AZ.

Pinkava, D. J., M. A. Baker, R. A. Johnson, N.Thrushell, G. A. Ruffner, R. S. Felger, and R. K.Van Devender. 1992. Additions, notes, and chro-mosome numbers for flora of vascular plants ofOrgan Pipe Cactus National Monument, Arizona.Journal of the Arizona-Nevada Academy ofScience 24: 13–18.

Powell, B. F., E. W. Albrecht, W. L. Halvorson, C. A.Schmidt, P. Anning, and K. Docherty. 2005.Vascular plant and vertebrate inventory of Tumacá-cori National Historical Park. USGS Open-FileReport 2005-1142. U.S. Geological Survey, South-west Biological Science Center, Sonoran DesertResearch Station, University of Arizona, Tucson,


AZ.Powell., B. F., W. L. Halvorson, and C. A. Schmidt.

2006. Vascular plant and vertebrate inventory ofSaguaro National Park, Rincon Mountain District.USGS Open-File Report 2006-1075. U.S. Geolog-ical Survey, Southwest Biological Science Center,Sonoran Desert Research Station, University ofArizona, Tucson, AZ.

Powell., B. F., W. L. Halvorson, and C. A. Schmidt.2007a. Vascular plant and vertebrate inventory ofSaguaro National Park, Tucson Mountain District.USGS Open-File Report 2007-1296. U.S. Geolog-ical Survey, Southwest Biological Science Center,Sonoran Desert Research Station, University ofArizona, Tucson, AZ.

Powell., B. F., C. A. Schmidt, W. L. Halvorson, andP. Anning. 2007b. Vascular plant and vertebrateinventory of Fort Bowie National Historic Site.USGS Open-File Report 2005-1167. U.S. Geolog-ical Survey, Southwest Biological Science Center,Sonoran Desert Research Station, University ofArizona, Tucson, AZ.

Rankin. A. G. 1991. Archeological survey in OrganPipe Cactus National Monument: 1991 progressreport. Unpublished report to Organ Pipe CactusNational Monument by the Western Archeolog-ical and Conservation Center, National ParkService, Tucson, AZ.

Rosen, P. C. 2000. A monitoring study of vertebratecommunity ecology in the northern SonoranDesert, Arizona. PhD. Dissertation. Departmentof Ecology and Evolutionary Biology, Universityof Arizona, Tucson, AZ.

Rosen, P. C., and C. H. Lowe. 1994. Highwaymortality of snakes in the Sonoran Desert ofsouthern Arizona. Biological Conservation 68:143–148.

Rosen, P. C., and C. H. Lowe. 1996. Ecology of theamphibians and reptiles at Organ Pipe CactusNational Monument, Arizona. Technical ReportNo. 53. Cooperative National Park ResourcesStudies Unit, University of Arizona, Tucson, AZ.

Rosenberg, K. V., R. D. Ohmart, W. C. Hunter, andB. W. Anderson. 1991. Birds of the lowerColorado River Valley. University of ArizonaPress, Tucson, AZ.

Ruffner Associates. 1995. Special-status plantsmonitoring protocol. In Organ Pipe CactusNational Monument ecological monitoringprogram monitoring protocol manual, pp. 1-i–1-39. Special Report 11. Cooperative National ParkResources Studies Unit, University of Arizona,Tucson, AZ.

Rutman, S. 2005. A checklist of vascular plants ofOrgan Pipe Cactus National Monument. Unpub-lished report.

Schmidt, C. A., B. F. Powell, and W. L. Halvorson.2007. Vascular plant and vertebrate inventory ofOrgan Pipe Cactus National Monument. USGSOpen File Report 2006-1076. U.S. Geological

Survey, Southwest Biological Science Center,Sonoran Desert Research Station, University ofArizona, Tucson, AZ.

Sprouse, T., R. Emanuel, and B. Tellman. 2002.Surface water quality monitoring overview andassessment. Unpublished report to the SonoranDesert Network Inventory and MonitoringProgram, National Park Service, Tucson, AZ.


Steenbergh, W. F., and W. E. Hoy. 1963. Checklistof birds of Organ Pipe Cactus National Monu-ment. Unpublished report to Organ Pipe CactusNational Monument, Ajo, AZ.

Steenbergh, W. F., and P. L. Warren. 1977. Prelim-inary ecological investigation of naturalcommunity status at Organ Pipe Cactus NationalMonument. Technical Report No. 3. CooperativeNational Park Resources Studies Unit, Universityof Arizona, Tucson, AZ.

Tibbitts, T. A. 1999. Desert pupfish. In Organ PipeCactus National Monument ecological monitoringprogram annual report 1996, edited by A. Pate.Organ Pipe Cactus National Monument, Ajo, AZ.

Trombulak, S. C., and C. A. Frissell. 2000. Reviewof ecological effects of roads on terrestrial andaquatic communities. Conservation Biology 14:18–30.

United States Department of Agriculture (USDA).2004. The PLANTS Database, Version 3.5(http://plants.usda.gov). National Plant DataCenter, Natural Resources Conservation Service,Baton Rouge, LA.

Warren, P. L., and L. S. Anderson. 1995. Naturalcommunity grazing-recovery monitoring protocol.In Organ Pipe Cactus National Monument ecolog-ical monitoring program monitoring protocolmanual, pp. 3-i –3-28. Special Report 11. Coop-erative National Park Resources Studies Unit,University of Arizona, Tucson, AZ.

Warren, P. L., B. K. Mortenson, B. D. Treadwell, J.E. Bowers, and K. L. Reichhardt. 1981. Vegeta-tion of Organ Pipe Cactus National Monument.Technical Report No. 8. Cooperative NationalPark Resources Studies Unit, University ofArizona, Tucson, AZ.

Weisenberger, M. E., P. R. Krausman, M. C.Wallace, D. W. DeYoung, and O. E. Maughan.1996. Effects of simulated jet aircraft noise onheart rate and behavior of desert ungulates. Journalof Wildlife Management 60: 52–61.

Western Regional Climate Center (WRCC). 2005.Arizona climate summaries from Organ PipeCactus National Monument, Arizona. Accessedon 4 June 2005 from: http://www.wrcc.dri.edu/summary/climsmnm.html

Wilt, R. A. 1976. Birds of Organ Pipe Cactus NationalMonument. Popular Series No. 18. Southwest Parksand Monuments Association, Globe, AZ.

VASCULAR PLANT AND VERTEBRATE INVENTORY OF SAGUARO NATIONAL PARK, RINCON MOUNTAIN DISTRICT

Brian F. Powell and Cecilia A. Schmidt

From 2001 to 2003 we surveyed for vascularplants and vertebrates (amphibians, reptiles,birds, and mammals) at the Rincon MountainDistrict of Saguaro National Park to documentthe presence of species within its boundaries.Park staff also surveyed for medium and largemammals using infrared-triggered camerasfrom 1999 to 2005. Our spatial samplingdesign was ambitious and was one of the firstof its kind in the region to co-locate study sitesfor vegetation and vertebrates using a stratifiedrandom design. We also chose the location ofsome study sites non-randomly in areas thatwe thought would have the highest speciesrichness. Because we used repeatable studydesigns and standardized field methods, theseinventories serve as the first step in a long-term, biological monitoring program for thedistrict.

With the exception of plants, our surveyeffort was the most comprehensive ever under-taken in the Rincon Mountains. We recordeda total of 801 plant and vertebrate species,including 50 species not previously found inthe district, of which five (all plants) are non-native species. Based on a review of ourinventory and past research at the district, therehave been a total of 1,479 species of plants andvertebrates found there. We believe invento-ries for all taxonomic groups are nearlycomplete. In particular, the plant, amphibianand reptile, and mammal species lists are themost complete of any comparably large naturalarea of the “sky island” region of southern

Arizona and adjacent Mexico. For all groups except medium and large

mammals, the low elevation stratum (< 1219 mor 4,000 ft) contained the highest species rich-ness, after accounting for differences in surveyeffort among strata. This is consistent withknown patterns of species richness in the skyisland mountain ranges.

Our review of species lists and park recordsreveals that the district has lost species, partic-ularly plants and mammals, in the past fewdecades. Because of the district’s close prox-imity to the rapidly growing city of Tucson,there are a number of development-relatedthreats that could cause additional loss ordecline in abundance of some species. Inparticular, the increasing groundwaterpumping near Rincon Creek, the most species-rich area in the park, is likely to impact theunique riparian vegetation and animals of thatarea.

See Powell, et al. (2006) for completedetails of the study of Saguaro National Park,Rincon District [http://sbsc.wr.usgs. gov/prod-ucts/ofr/]. Scientific and common names usedthoroughout this chapter are current accordingto accepted authorities for each taxonomicgroup: Taxonomic Information System (ITIS2001) and the PLANTS Database (USDA2004) for plants, Stebbins (2003) for amphib-ians and reptiles, American OrnithologistUnion (AOU; 1998, 2003) for birds, andBaker, et al. (2003) for mammals.

54 Powell and Schmidt

PARK OVERVIEWPark Area and History

Saguaro National Park is located in easternPima County adjacent to Tucson, Arizona(Figure 1). Originally designated as a nationalmonument, the park was created in 1933 topreserve the “exceptional growth” of thesaguaro cactus (NPS 1992). In 1961, the parkwas expanded to include over 9,000 ha(22,239 ac) of the Tucson Mountains (knownas the Tucson Mountain District). The RinconMountain District is the subject of this report.It is 27,233 ha (67,294 ac) in size and isbounded by U.S. Forest Service land to theeast; Forest Service and private land to thenorth; Forest Service, private and state landto the south; and private land to the west(Figure 2). Although created to preservenatural resources, the park is also home tonative American campsites and petroglyphsand contains remnants of early ranching andmining (NPS 1992). Annual visitation to bothdistricts of the park averages approximately700,000 (NPS 2005).

Physiography and GeologySaguaro National Park is located within theBasin and Range Physiographic Province.The district encompasses most of the RinconMountains, one of the region’s prominent“sky island” mountain ranges. Topography atthe district varies from low-elevation desertflats to steep rocky canyons and high-eleva-tion coniferous forest and meadows.Elevation ranges from 814 m (2,670 ft) in thenorthwestern corner of the district to 2,641 m(8,665 ft) at Mica Mountain. The RinconMountains are primarily metamorphic inorigin, with rocks of the Santa CatalinaGroup, a mixture of Pinal Schist, ContinentalGranodiorite, and Wrong Mountain QuartzMonzonite (McColly 1961; Drewes 1977).All components are of Precambrian rockparentage, subsequently deformed and recrys-talized. Sedimentary rocks in the vicinity arelargely Permian limestones of Earp andHorquilla formations (Drewes 1977).

HydrologyThe Rincon Mountain District has severalsources of perennial water: Chimenea,Madrona, Rincon, and Wild Horse Creeks; andDeer Head, Spud Rock, Italian, and ManningCamp Springs. The most prominent hydrologicfeature is Rincon Creek, which drains approx-imately one-half of the district (Sprouse et al.2002).

ClimateSaguaro National Park experiences an annualbimodal pattern of precipitation which is char-acterized by heavy summer (monsoon) stormsbrought about by moisture coming from theGulf of Mexico, and less intense winter frontalsystems coming from the Pacific Ocean. Onaverage, approximately one-half of the annualprecipitation falls from July throughSeptember (WRCC 2005; PCFCD 2005). Thearea’s hot season occurs from April throughOctober; daily maximum temperatures exceed40° C (104° F) at lower elevations and 30° C(86° F) at high elevations. Winter temperaturesdip below freezing and snow is common athigh elevations.

From 2001 to 2003, during the time of mostof our inventory effort, annual precipitationtotals for the high elevation areas were belowthe long-term mean of 69.1 cm (27.2 in).Annual temperatures for high elevationsranged from slightly below to slightly abovethe long-term mean of 8.5° C (47.3° F;PCFCD 2005). Annual precipitation totals forlow elevations ranged from slightly to substan-tially below the long-term mean of 28.6 cm(11.3 in). Annual temperatures for low eleva-tions from 2001 to 2003 were above thelong-term mean of 21.3° C (70.3° F; WRCC2005).

Vegetation and Biotic CommunitiesThe Rincon Mountains are one of the “skyisland” mountain ranges of southeast Arizonaand northern Mexico. Sky islands, so calledbecause the mountains are isolated by “seas”of desert and semi-desert grasslands, are areasof remarkable biological diversity as a result

Powell and Schmidt 55

Figure 1. Location of the two districts of Saguaro National Park in southern Arizona.

of elevational gradients and subsequent differ-ences in precipitation and temperature. Thesemountain ranges extend from subtropical totemperate latitudes, hosting species whose coredistributions are from the Sierra Madre ofMexico and the Rocky Mountains of the

United States and Canada (Warshall 1994). Insouthern Arizona, the sky island mountainranges all have similar biotic communitiesfrom low-elevation Sonoran desertscrub tohigh-elevation conifer forests (Whittaker andNiering 1965).


Sonoran DesertscrubSonoran Desertscrub is found at the lowestelevations and driest areas of the district, on itswest and southern boundaries. The dominantshrubs are velvet mesquite, acacias, paloverdes, and creosote bush. Succulents are ubiq-uitous and include: saguaro, agave, yucca,barrel cactus, pincushion cactus, prickly pear,and cholla. Warm- and cool-season annuals,both native (e.g., woolly plantain) and intro-duced (e.g., red brome) are common followingrainfall.

Southwestern Deciduous Riparian ForestThese forests are found along low-elevationwashes and creeks and are among the mostbiologically unique communities in theSonoran Desert. At the district they are foundalong Rincon Creek and to a lesser extentalong its tributaries. The dominant tree species

are Fremont cottonwood, Arizona sycamore,velvet ash, willow, and netleaf hackberry. Inthe Rincon Mountain District, SonoranDesertscrub bounds these zones.

Semi-desert GrasslandSemi-desert grasslands occur in some middleelevation areas of the district, primarily alongthe northern boundary of the district and on afew areas of Tanque Verde Ridge. The commu-nity is composed of perennial short- andmid-grass species, with most areas invaded byvelvet mesquite.

Oak SavannahThe oak savannah community is found athigher elevations than the semi-desert grass-land community and lower elevations than thepine-oak woodland, and it contains elementsof both communities. It is ecologically similar

Figure 2. Aerial photograph showing major features of Saguaro National Park, RinconMountain District.


to the chaparral communities of centralArizona. In this community there are densestands of manzanita and oak, with a variety ofannual and perennial grasses.

Pine-oak Forest and WoodlandPine-oak forest and woodland is ubiquitous atmid-elevations throughout the Apache High-lands (Bailey 1998; McPherson 1993).Madrean evergreen woodland is characterizedby evergreen oaks with thick sclerophyllousleaves, such as Emory oak, Arizona white oak,and Mexican blue oak. Mexican pinyon pineand alligator juniper are the commongymnosperms. Understory grasses are usuallyabundant. At the higher elevations and indrainages, there is also ponderosa pine.

Coniferous ForestDominated by gymnosperms such as pines andfirs, coniferous forests represent the cold-hardiest biotic community in the district. Inthese communities, ponderosa pine andDouglas fir dominate, with some temperatedeciduous plants intermixing, primarily on thenorth-facing slopes: Gambel oak, quakingaspen, maples, and boxelder. Conifer forests arefire-adapted ecosystems, with natural low-intensity fires occurring every 6 to 15 years(Baisan and Swetnam 1990; Dimmitt 2000).

Natural Resource Management IssuesAdjacent Land DevelopmentIncreased housing development along thewestern and southern boundaries has becomethe most pressing natural resource issue for thedistrict. Sandwiched between both districts ofthe park, the greater Tucson metropolitan areais one of the fastest growing in the UnitedStates. The area currently has an estimatedpopulation of 800,000, a 44% increase over thelast two decades (PAG 2005). The increase inhuman residents brings with it a variety ofnatural resource-related problems includingharassment and predation of native species byferal animals, increased traffic leading toaltered animal movement patterns andmortality, the spread of non-native species,

illegal collections of animals, vandalism,increased water demands, air pollution fromvehicle emissions, and visual intrusions to thenatural landscape (Briggs et al. 1996).

Of immediate concern is the depletion ofgroundwater and its effects on the ecologicallyvaluable Rincon Creek (Baird et al. 2000).There are numerous single-family housingunits being constructed (or planned) directlyadjacent to the district. The proposed RockingK Ranch development anticipates 9,000 resi-dents and has been granted a permit by theArizona Department of Water Resources towithdraw 4,400 acre feet of water per yearfrom the underlying aquifer (Mott 1997).Rincon Creek has the most well-developedstretch of southwestern deciduous riparianforest in the district and it will likely be nega-tively impacted by drawdown of the aquifer.Groundwater drawdown at Tanque VerdeWash has already affected the riparian commu-nity there (Mott 1997).

Non-native Species and Changes to VegetationBuffelgrass, Lehmann lovegrass, red bromeand other non-native grasses, have increasedin the last ten years (Funicelli et al. 2001).The spread of some non-native plants usedfor landscaping, such as crimson fountain-grass, from development bordering thedistrict is also a concern. The invasion of non-native grasses has led to structural changes invegetation, from areas that supported mostlysparse bunchgrasses to areas of uniformgrass. This change in species composition andstructure alters the fire regime of the area bysupporting larger fires and higher firefrequencies, thereby leading to other changesin vegetation composition and structure(Anable et al. 1992). Nowhere are theseeffects more evident than in the SonoranDesertscrub community, which rarely burnedhistorically (Steenbergh and Lowe 1977).Many native plant species, especially succu-lents, are not adapted to short duration,high-intensity fires and therefore die(Schwalbe et al. 1999; Dimmitt 2000).


Wildland FireSince the park began keeping records in 1937,there have been 572 fires in the district(Swantek, et al. 1999; NPS Files), and since1984, park personnel have burned approxi-mately 1,450 ha (3,583 ac) through their activefire management program. Fires play a crucialrole in the middle and high-elevation semi-desert grasslands and forests by depletingdense understory vegetation and downed-woody debris. Even in these fire-adaptedecosystems, however, fire can be devastating,particularly after decades of suppression andsubsequent buildup of fuel loads. In additionto fire-caused loss of species, a number of largefires in the last few decades have led tomassive runoff of sediment and ash, fillingdown-stream perennial pools, and destroyinghabitat of leopard frogs and other aquaticspecies. Despite such problems, the NPS iscommitted to returning natural fire cycles tothe park’s biotic communities

METHODSFocal Points

We used a process in our design that allowedfor all inventories to be conducted partially atthe same locations so that analyses could bemade about communities at specific spots onthe landscape. These co-located sites weredefined using a randomization technique andresulted in the creation of 17 focal points thatwere available for use by all inventory crews(Powell et al. 2006).

PlantsWe located specimens representing 883species at the University of Arizona herbarium.Many of these specimens were collected orreported in Bowers and McLaughlin (1987).Their treatise is the most comprehensive anno-tated flora for the Rincon Mountains, thoughspecies have been added to the list since itspublication. Bowers and McLaughlin (1987)also provide an excellent overview of previousresearch and collecting from the range, theplant communities present, species richnessgradients, and a list of species extirpated from

the range. The Bowers and McLaughlin listwas compiled from work by Bowers (1984)above 1,371 m (4,500 ft) elevation and byCarole Jenkins who collected from 1978 to1982 below 1,371 m. The list was updated in1996 to include the addition of 34 species andthe subtraction of four species due to incorrectidentifications (Fishbein and Bowers 1996).There were floras developed for four desig-nated natural areas of the district: WildhorseCanyon (Rondeau and Van Devender 1992),Chimenea Canyon (Fishbein et al. 1994a), BoxCanyon (Fishbein et al. 1994b), and MadronaCanyon (Fishbein 1995). Halvorson andGebow (2000) compiled these works into asingle volume. Halvorson and Guertin (2003)mapped locations of 27 species of non-nativeplants.

We collected species opportunistically andwhen we thought we had found a species noton the district list. We collected specimensduring 38 days of fieldwork between 10 Apriland 24 September 2001 and 4 and 5 May2002. We collected specimens from 41 loca-tions throughout the district and many of thecollections were made in the course of trav-eling to and from focal points. We usedmodified-Whittaker plots to characterize plantcommunities at the inventory focal points.Each plot was 20 x 50 m (1000 m²) andcontained 13 subplots of three different sizes(Stohlgren et al. 1995). We used the point-intercept method (Bonham 1989) to samplevegetation along 50-m transects located at 13focal points.

Amphibians and ReptilesPrevious inventoriesPrior to this study several species lists had beendeveloped (Black 1982; Doll et al. 1989; Loweand Holm 1991; Swann 2004). Goode et al.(1998) inventoried the district’s ExpansionArea in Rincon Valley and Murray (1996) andSwann (1999) inventoried both the ExpansionArea and the nearby Rocking K Ranch andprovided detailed information for these areas.Most recently, Bonine and Schwalbe (2003)inventoried the Madrona Pools of Chimenea


Creek; their effort was limited to five days inMay.

This inventoryWe surveyed herpetofauna in 2001 and 2002using four field methods: plot-based intensivesurveys, non-plot based extensive surveys,road surveys, and incidental observations. Weused multiple methods to ensure coverageacross a broad range of environmental featuresand to facilitate complete species lists and esti-mates of relative abundance (Powell et al.2006).

Intensive SurveysAt focal points in 2001, we completed 131surveys at 51 subplots located along 17 focal-point transects. In 2002 we discontinuedintensive surveys because of the relatively lownumber of species detected.

Extensive SurveysNon-plot based extensive surveys facilitatedsampling in areas where we expected highspecies richness, abundance, or species notpreviously detected. Typically, we selectedareas for extensive surveys in canyons orriparian areas, and also included ridgelines,cliffs, rock piles, bajadas, summits, or otherphysiographic features. We surveyed inspring (4 April to 24 May) and summer (25June to 20 September) of 2001 and 2002.One, two, or three observers searched 85areas. Total duration of surveys among allobservers ranged from 1.2 to 20.4 hours.Survey effort was roughly three times greaterthan for other methods and focused mainlyduring daylight except at lower elevationswhere we also surveyed during late eveningsand nights.

Road SurveysWe focused mainly on the Cactus Forest LoopDrive and also drove Speedway Boulevardfrom Douglas Spring Trailhead to the inter-section with Tanque Verde Loop Road andCamino Loma Alta from the trailhead to OldSpanish Trail. We recorded each individual

detected by species and whether animals weredead or alive. We surveyed between 29 Apriland 18 August 2001 and between 9 to 14 July2002 during nights and occasionally duringevenings. We conducted 55 road surveystotaling 46.3 hours of effort.

Incidental ObservationsIncidental detections were often recordedbefore or after more formal surveys and weused these sightings to determine speciespresence and richness. We also included inci-dental sightings from other field crews (e.g.,birds).

BirdsPrevious inventoriesThere has been considerable bird research atthe Rincon Mountain District, but no compre-hensive and well-documented inventory hasbeen completed. Monson and Smith (1985)compiled a checklist for both districts of thepark, but there is no documentation of the dataused to create that list. The park has contractedfor periodic raptor surveys (Felley and Corman1993; Berner and Mannan 1992; Bailey 1994;Griscom 2000). Park personnel surveyed threeBreeding Bird Atlas blocks within the district(Short 1996) and those results are reported inCorman and Wise-Guervais (2005). TheTucson Bird Count includes three low-eleva-tion sites in the park, including Rincon Creek(TBC 2005).

This inventoryWe surveyed for birds at the Rincon MountainDistrict from 2001 to 2003. We used four fieldmethods: variable circular-plot (VCP) countsfor diurnal breeding and spring migrant birds,nocturnal surveys for owls and nightjars(breeding season), line transects for diurnalbirds in the non-breeding season, and inci-dental observations for all birds in all seasons.We concentrated our primary survey effort inthe breeding season (Bibby et al. 2002) whichincluded the peak spring migration times formost species, thus adding many migratoryspecies to our list.


VCP SurveysIn 2001, we spent more effort surveying atfocal-point stations (n = 272) than at non-random stations (n = 160). In 2002 wesurveyed exclusively at non-random stations,both repeat-visit (n = 130) and reconnaissance(n = 107).

Line-transect SurveysWe used a modified line-transect method(Bibby et al. 2002) to survey for birds fromNovember 2002 to February 2003. We estab-lished three line transects in the district andsurveyed each transect four times in the winterof 2002 and 2003.

Nocturnal SurveysTo survey for owls we broadcast commerciallyavailable vocalizations (Colver et al. 1999)using a compact disc player and broadcaster(Bibby et al. 2002) and recorded othernocturnal species (nighthawks and poorwills)when observed. We established nine transects.We broadcasted vocalizations of species thatwe suspected, based on habitat and range infor-mation, might be present:• Low elevation: elf, western screech,

burrowing, and barn owls;• Middle elevation: elf, northern pygmy,

flammulated, and whiskered screechowls;

• High elevation: northern pygmy, flammu-lated, northern saw-whet, and whiskeredscreech owls.Although we had the most transects in the

high elevation stratum, we had most (56%) ofour survey effort in the low elevation stratumbecause of greater ease of accessing stations.Incidental and Breeding Observations When we were not conducting formal surveysand we encountered a rare species, a species inan unusual location, or an individual engagedin breeding behavior, we recorded it. Werecorded all breeding observations using thestandardized classification system developed

by the North American Ornithological AtlasCommittee (NAOAC 1990).

MammalsPrevious inventoriesSaguaro National Park has never had acomprehensive survey of its mammals, andsurprisingly little research has been conductedon mammals in the Rincon Mountain Districtconsidering the park’s long history as a unit ofthe National Park System. However, a fewstudies provide valuable information onmammals, particularly Lowell Sumner’s workin the mid-20th Century (Sumner 1951) andRussell Davis and Ronnie Sidner’s survey ofmammals in the high country of the Rinconsin the early 1990s (Davis and Sidner 1992). H.Brown and L. Huey (unpubl. data) madecollecting trips to the Rincons in 1911 and1932, respectively (Davis and Sidner 1992). Inaddition, the park’s administrative records atthe Western Archeological and ConservationCenter contain invaluable files (dating from the1940s and 1950s) on mammal sightings andspecies of concern including the Mexican graywolf and tree squirrels. More recent surveyswere conducted by McCloskey (1980),Duncan (1990), Sidner (1991), Sidner andDavis (1994), Fitzgerald (1996), Lynn (1996),Bucci (2001), Swann (2003), and Sidner(2003).

This inventoryWe surveyed for mammals using five fieldmethods: trapping for small mammals,infrared-triggered photography for mediumand large mammals, netting for bats, pitfalltraps for shrews and pocket gophers, and inci-dental observations for all mammals.

Small MammalsWe trapped small mammals using Shermanlive traps set in grids (White et al. 1983) alongfocal-point transects. The majority of our trap-ping effort in 2001 was at focal-point transects.We trapped for 4,589 trap-nights. We had themost trapping effort in the middle elevationstratum, less in the high elevation stratum, and


the least in the low elevation stratum. In non-random areas, the percentage of the totalnumber of trap nights was 36%, 50%, and 37%for the low, middle, and high elevation strata,respectively.

BatsWe surveyed for bats using two field methods:roost-site visits and netting. For netting, weconcentrated our survey effort in areas thatwere most likely to have bats, mostly riparianareas with surface water present. We did notsurvey for bats near focal points because of thelow probability of success in these areas. Wevisited roosts that were known to have batsbased on historic records or were likely to havebats based on habitat characteristics. We nettedbats at six sites for a total of 13 nights ofnetting in 2001 and four nights of netting in2002. Most of our netting effort was at lowerRincon Creek and at Manning Camp Pond; wenetted at each site for five nights. Deer Creekwas the only site at which we netted on the eastslope of the Rincon Mountains.

Large and Medium MammalsSaguaro National Park initiated a medium andlarge mammal inventory in 1999 (Aslan 2000;Wolf and Swann 2002; Swann et al. 2003a;Swann 2003). We used infrared-triggeredcameras to detect medium and large mammalsat a combination of random and non-randomsites from January 1999 to June 2005. Welocated non-random sites primarily at knownwater sources and animal trails. We chose thelocation of these sites to be in areas that webelieved would have the highest species rich-ness. We placed cameras at 74 non-randomand 40 random sites throughout the district. Weplaced 54% of the camera time in the lowelevation stratum, 28% in the middle, and 18%in the high elevation stratum. The total numberof camera nights at all sites was 3,895.

Pitfall trappingTo survey for shrews and pocket gophers weplaced pitfall traps in moist, north-facingslopes of the Rincon Mountains in 2001. We

placed traps adjacent to a natural feature suchas a fallen log or rock. We placed 10 traps (22May to 24 September) at the North Slope Trailsite, and four traps each at Italian Spring andSpud Rock Spring.


We collected 741 specimens representing 523species from the Rincon Mountain District ofSaguaro National Park (Table 1). We found 39species that had not previously been docu-mented in the district, almost one-half of them(n = 19) during the course of surveying atpoint-intercept and/or modified-Whittakerplots. The list of new species that we foundincluded five non-native species, most notablyAfrican sumac. Native species of note that weadded to the flora included cleftleaf wildhe-liotrope, Arizona dewberry, and Americanblack nightshade.

Based on a thorough review of past studies,floras, and collections located at the Univer-sity of Arizona, there have been a total of 1,170specific and intraspecific taxa documented atthe district, of which 78 are non native (6.7%).Excluding eight species in the UA collectionthat Bowers and McLaughlin (1987) cite aslikely extirpated from the district, there havebeen 1,120 species (1,162 including intraspe-cific taxa) documented since the early 1980s.Of these species, six were thought to be extir-pated by Bowers and McLaughlin (1987) butwere found by other studies: purple scalystem,Lemmon’s hawkweed, alderleaf mountainmahogany, Baltic rush, poverty rush, andcommon barley.

We found 367 species associated with the17 focal points. Approximately 47% of thesespecies we found associated with only a singlefocal point, whereas six species (spidergrass,side-oats grama, plains lovegrass, bullgrass,sacahuista, and skunkbush sumac) were asso-ciated with 10 or more focal points. Theskunkbush sumac was the most widespreadspecies; we found it at 71% of focal points. Werecorded 307 species on 13 modified-Whit-taker plots. The mean number of species per


plot was 60 with the range from 97 species inone of the Sonoran Desertscrub plots to 20species in one of the Conifer Forest plots.

We found 189 species on 17 point-intercepttransects. The mean number of species pertransect was 28.3, with a range from 8 to 43.

The district’s flora is perhaps the mostcomplete of any large natural area in the skyisland region of southeastern Arizona. In ourmany days of collecting, we found 39 previ-ously undocumented species, which representsa 3.3% increase in the flora for the district.Almost one-half of these species were foundduring the course of conducting surveys atfocal points. We also found a number ofspecies on the east slope of the Rincon Moun-tains, and collectively these areas, particularlythose away from hiking trails, are the least-surveyed areas of the district and finding newspecies there is not surprising.

Assessing overall inventory completenessis problematic given the size of the district anddifficulty accessing many areas because ofrough terrain. Due to the fact that much of thedistrict remains unsurveyed, it is possible thatwe and others have not reached the goal ofdocumenting 90% of the plant species for theentire district. However, if we look at inven-tory effort in different areas, the completionestimates are mixed. For example, low eleva-tion, more easily accessed areas almostcertainly have a species list that is close to

completion. We found only three new speciesat or near focal points in the low-elevationstratum, and only one new species in an areanear the loop drive, a highly visited area. Thepark’s monitoring efforts have had similarresults in low-elevation areas; in their 25 long-term monitoring plots (surveyed for sevenyears), park staff have found only 15 newspecies for the district . The flora for the high-elevation areas of the district is similarlycomplete. We found only one species in thearea around Manning Camp, an area that hashad extensive plot-level research related to thefire effects program. That program hasproduced only 30 new species in 15 years ofsurveys of 71 plots (Saguaro National Park,unpubl. data). By contrast, the mid elevationareas are the least surveyed and our resultsreflect this; we found most of our new speciesat focal points in the middle elevation stratumwhere plots were in the most difficult areas ofthe district to reach. Based on this evidence,we suggest that the floras for low and highelevation areas are nearly complete and thatfuture surveys should focus on middle eleva-tion areas, especially the east slope of theRincon Mountains and the north-easternboundary of the district.

Amphibians and ReptilesWe recorded seven amphibian and 39 reptilespecies (Table 1). Reptile species included two

Table 1. Summary of vascular plant and vertebrate inventories at Saguaro National Park,Rincon Mountain District, 1999–2005.

_______UA inventory_______Number of Number of new Number of Total number

species species added non-native of species onTaxonomic group recorded to district list species district list

Plants 523 39 78 1,162Amphibians and Reptiles 46 0 2 56Birds 173 10 3 198Mammals 59 1 3 63Totals 801 50 86 1,479


turtles, 19 lizards, and 18 snakes. Species rich-ness was highest for incidental (n = 43) andextensive surveys (n = 39) and lowest for inten-sive (n = 25) and road surveys (n = 22). Wefound seven species with only a single surveymethod, but all other species were found withtwo or more methods. Road and extensivesurveys each yielded detection of one speciesthat was not detected by using other methods(Great Plains toad, and Great Plains skink,respectively) and incidental surveys yieldeddetection of five species not detected by usingother methods (Mexican spadefoot, canyonspotted whiptail, ring-necked snake, westernground snake, and Mojave rattlesnake). All 25species that we detected during intensivesurveys were detected using other methods,although Madrean alligator lizard was detectedonly during intensive and extensive surveys.

We detected 4,292 individuals during thisstudy, 3,066 during intensive, extensive, androad surveys combined and 1,225 incidentalobservations. Most individuals (n = 1,909)were detected during extensive surveys andfewest (n = 469) were detected during roadsurveys. The number of individuals detectedper unit time was greatest for road surveys(mean = 14.9 individuals/hr) markedly higherthan for extensive (4.1 individuals/hr) or inten-sive (3.6 individuals/hr) surveys. The specieswith the most detections (all methodscombined) was the ornate tree lizard (n = 750).

A review of our inventory effort and otherefforts in the district indicates that the districtsupports a total of nine amphibians and 48reptiles. All but five species have beenconfirmed with a specimen and/or photo-graphic voucher. Our inventory did not resultin detection of any species not alreadyrecorded in the district, although we producedthe first documentation (in the form of spec-imen and photographic voucher) for a numberof species, including the Mojave rattlesnake.

We recorded only two species of amphib-ians (Sonoran desert toad and canyon treefrog)during intensive surveys. During extensivesurveys we recorded five amphibians: Couch’sspadefoot toad, Sonoran Desert Toad, red

spotted toad, canyon tree frog, and lowlandleopard frog. Road surveys resulted in thecollection of four species: Couch’s spadefoottoad, Sonoran Desert Toad, red spotted toad,and Great Plains toad. Only one other specieswas documented during the incidental collec-tion periods, the Mexican spadefoot toad.

Environmental factors that explained patternsof species richness and relative abundancevaried. Snake richness increased with cover ofgrasses whereas lizard richness increased withdecreasing bare ground. Species richness ofsnakes and lizards increased with shrub coverabove 2 m, though influence of shrub cover wasmuch greater for snakes, and richness of lizardsdecreased with tree cover between 0.5 and 2.0m. Relative abundance of all lizard speciescombined declined with increasing cover of bareground. The black-necked garter snake andwestern diamond-backed rattlesnake were themost common snakes and western and moun-tain patch-nosed snakes, Sonoran coral snake,and common kingsnake were the rarest.

Species accumulation curves nearlyreached an asymptote for extensive and inten-sive surveys, suggesting that additional surveyswould have produced few new species. In fact,many species that we found only incidentallyor have been documented few times are so rarethat encountering them is largely a function ofchance. We believe that the goal of a 90%species list has definitely been achieved in thecase of amphibians and reptiles.

BirdsWe made over 15,000 observations of birds andfound 173 species from 2001 to 2003. We found10 species that had not previously been foundin the district including the sulphur-belliedflycatcher, elegant trogon, and pinyon jay. Somehad conservation designations, including thenorthern goshawk, yellow-billed cuckoo,Mexican spotted owl, and buff-breastedflycatcher. Unusual sightings included a nest ofthe sulphur-bellied flycatcher, a singing malebuff-breasted flycatcher, and sightings of thewild turkey, common black hawk, and yellow-breasted chat. We recorded three non-native


species, including the rock pigeon, a newspecies for the district. We recorded the mostspecies during incidental observations (n = 154)and VCP surveys (n = 149), and fewestduring nocturnal surveys (n = 9).

We recorded 143 species at all repeat-visitVCP stations. We found the most species in theRiparian community (n = 102) and fewestspecies in the Conifer Forest community(n = 51). The number of species found in theother three communities (Sonoran Desert-scrub, Oak Savannah, and Pine-oak Woodland)was intermediate. We recorded twelve speciesin all five communities and 39 species in onlya single community. The ash-throatedflycatcher was the most widespread species;we recorded it on 21 of 23 repeat-visit tran-sects. We recorded four other species at > 75%of transects: rufous-crowned sparrow, commonraven, brown-headed cowbird, and white-winged dove. We recorded an additional 22species on > 50% of transects and an equalnumber of species on only a single transect.The white-winged dove had the highest meanfrequency of detection across strata and it wasthe only species for which we recorded anaverage of over one individual per station. Themourning dove and ash-throated flycatcherwere the only other species with relativefrequency of detection estimates > 0.75.

We calculated relative abundance for 120species. The most abundant species for eachcommunity type were:• Riparian: verdin, Lucy’s warbler, and

mourning dove;• Sonoran Desertscrub: black-throated

sparrow, cactus wren, and verdin; • Oak Savannah: Bewick’s wren, rufous-

crowned sparrow, and ash-throatedflycatcher;

• Pine-oak Woodland: Bewick’s wren,spotted towhee, and black-throated graywarbler;

• Conifer Forest: yellow-eyed junco, moun-tain chickadee, and spotted towhee andcordilleran flycatcher.

We found 63 species during line-transectsurveys in the winter of 2002 and 2003including six species that we did not recordduring VCP surveys. We found the mostspecies in the Lower Rincon Creek area(n = 45) on the south side of the district andfewest in the Douglas Springs area (n =31)near the northern boundary.

We observed 154 species during incidentalobservations, including 13 species that we didnot record during other surveys. We made 288observations of 78 species that confirmedbreeding in or near the district; we found 104nests of 48 species. We found two instances ofbrown-headed cowbird parasitism: one blue-gray gnatcatcher feeding a fledgling cowbirdand one Bell’s vireo nest with a cowbird egg.Considering all of the other research and site-specific inventory efforts in the district, we areconfident in concluding that at least 90% of thespecies that regularly occur in the district havebeen recorded.

MammalsWe confirmed a total of 59 species of mammalsin the Rincon Mountain District. This included12 species confirmed through specimens, 32species confirmed through photographs, ninespecies captured for which a voucher specimenpreviously existed, five species confirmedthrough a combination of voucher specimensand photos, and one species confirmed throughreliable observation. One species included inthis total (eastern cottontail) was confirmed byphotographs in appropriate high-elevationhabitat, but requires further documentation. Weconfirmed three species of mammals not previ-ously confirmed for the district: western red bat,fulvous harvest mouse, and Virginia opossum.The latter two species represent significantrange extensions. We observed only one specieslisted by the U.S. Fish and Wildlife Service asendangered, the southern long-nosed bat. Threespecies of non-native animals were docu-mented for the district (feral cat, domestic dog,and domestic cattle) but we do not believe thatany of these species have established feralpopulations in the district.


There have been a total of 66 speciesobserved or documented in the district in thelast few decades based on this and previousstudies. We did not document the presence ofeleven species that were previously docu-mented for the Rincon Mountain District. Wedid not confirm the deer mouse, captured inthe early 1950s near Manning Camp. We didnot confirm the banner-tailed kangaroo rat,previously confirmed by specimen voucher(Hoffmeister 1986), and did not observe anyof the distinctive sign of this very largekangaroo rat. Three species of bats that wedid not observe, the western small-footedmyotis, Yuma myotis, and western pipistrelle,have been confirmed recently (Davis andSidner 1992; Sidner 2003). One species ofrodent (southern grasshopper mouse) is alsopresent; a roadkilled individual found by DonSwann in 1997 was confirmed by YarPetryzyn at the University of Arizonamammal collection. Four species are extir-pated from the district (grizzly bear, jaguar,Mexican gray wolf, and bighorn sheep), anda fifth species (North American porcupine)may be extirpated, though it remains on thespecies list.

Small MammalsWe trapped 544 individual rodents (includingrecaptures) in 2001 and 2002, and documented13 rodent species and three species of diurnalsquirrels. One species, the fulvous harvestmouse (4 captures) was a new species for thedistrict. We did not capture two species thathave been previously documented for thedistrict (the southern grasshopper mouse andbanner-tailed kangaroo rat). In general, rela-tive abundance was higher at both low andhigh elevations than at middle elevations. Wetrapped only one species (rock squirrel) in asingle elevation stratum, and only one species(brush mouse) in all three strata. Theremainder of the species we found in twostrata, either in the low and middle or themiddle and high elevation strata. We trappedno species solely in the middle elevationstratum.

BatsWe confirmed 15 species, including onespecies that was not previously found at thedistrict (western red bat). We observed batsin only one roost site, where 500-1000 cavemyotis and six southern long-nosed bats werefound. This was the only site at which weconfirmed the southern long-nosed bat.Lower Rincon Creek had the highest speciesrichness of any site, and Manning Camp hadthe highest percent netting success and themost individuals captured. We captured fivespecies at Lower Rincon Creek that we didnot capture in any other site and one speciesat Manning Camp Pond that we did notcapture at any other site. At no other site didwe capture species that were not found else-where. The big brown bat was the mostwidespread and abundant species; it wasfound at five of the six sites and in all eleva-tion strata. Big brown bats were captured in80% of the visits to Lower Rincon Creek andManning Camp Pond. The Brazilian free-tailed bat was the next most captured bat; wecaptured 16 individuals at three sites. Of the14 species that we captured at the RinconMountain District, 10 were represented byfour or fewer individuals. Even the cavemyotis, for which we found a roost of > 500individuals, was only represented by a fewindividuals captured by netting.

Medium and Large MammalsIn 3,895 estimated camera nights, 2,939photographs captured at least one mammal,and a total of 3,407 individual mammals thatcould be identified to genus. We photographed27 species, including two non-native species,domestic dog and cattle. We documented onespecies (Virginia opossum) not previouslyreported for the district and a large number ofspecies for which there had previously beenonly observational records. The largestnumber of photographs was of the gray fox(1,018 photos), followed by collared peccary(588 photos), and ringtail (229 photos).


Pitfall TrappingWe trapped eight animals in pitfall traps: sixdesert shrews at the North Slope site, onewestern harvest mouse, and one Botta’s pocketgopher.

SUMMARYWe confirmed a total of 59 species ofmammals in the Rincon Mountain District andfailed to confirm ten species that have beenpreviously documented for the Rincon Moun-tains. Of these ten, four species (grizzly bear,jaguar, Mexican gray wolf, and bighorn sheep)are certainly extirpated from the district andtwo others (deer mice, North American porcu-pine, and banner-tailed kangaroo rat) may beextirpated. We believe that three species of batsand one rodent that were documented in thepast are still present and would be confirmedwith additional effort. Our inventory confirmed93% of mammals known for the district. Thespecies accumulation curves for smallmammal trapping, bats, and infrared-triggeredcameras also suggest that the mammal inven-tory for the district is at or above 90%completion.

ACKNOWLEDGMENTSThanks to Saguaro National Park Superin-tendent Sarah Craighead, Chief of Science andResource Management Meg Weesner, andbiologists Natasha Kline and Don Swann forproviding leadership and administrativesupport for this project. Other park staff whoassisted our project included Matt Daniels,Mark Holden, Bob Lineback, Todd Nelson,Kathy Schon, Mike Ward, and Jim Williams.

Andy Hubbard at the Sonoran DesertNetwork I&M program has been a great advo-cate of our program. He also provided fundsfor Don Swann to work on this report. KathyDavis, Superintendent of Tuzigoot andMontezuma Castle national monuments playedan instrumental role in this project byproviding important early initiative. LarryNorris at the Desert Southwest CESU hasprovided strong support for our program andspent considerable time and effort providing

clear and timely administrative assistance.Matt Goode, Don Swann, and Dale Turnerprovided much of the early planning for thisproject; we are indebted to their vision andwork. We thank a core group of dedicatedfield biologists who collected a wealth of dataat Saguaro National Park: Greta Anderson,Theresa DeKoker, Sky Jacobs, Shawn Lowery,Meg Quinn, Rene Tanner, Dale Turner, andEmily Willard (plants); Dan Bell, KevinBonine, James Borgmeyer, Matt Goode, DavePrival, and Mike Wall (amphibians andreptiles); Eric W. Albrecht, Gavin Bieber,Aaron Flesch, Chris Kirkpatrick, and GabeMartinez (birds); Clare Austin, Eric W.Albrecht, Mike Chehoski, Ryan Gann,Michael Olker, Neil Perry, Jason Schmidt,Ronnie Sidner, Mike Sotak, Albi von Dach,Michael Ward, and Sandy Wolf (mammals).We are appreciative of the following people,many of whom never ventured into the field,but whose work in the office made the fieldeffort successful: Debbie Angell, JenniferBrodsky, Chuck Conrad, Louise Conrad, BrianCornelius, Taylor Edwards, Carianne Funicelli,Marina Hernandez, Colleen McClain, HeatherMcClaren, Lindsay Norpel, Ryan Reese, JillRubio, Brent Sigafus, Taffy Sterpka, JennyTreiber, Zuleika Valdez, Alesha Williams, andErin Zylstra. Pam Anning, Kristen Beaupre,and Matthew Daniels assisted with databasedesign. Pam Anning also provided the mapsfor this report. Additional administrativesupport was provided by Valery Catt, JennyFerry, Andy Honaman, Terri Rice, and espe-cially Cecily Westphal of the School of NaturalResources at the UA. Special thanks to PamAnning, Lisa Carder, and Kathleen Dochertyfor their years of hard work on all aspects ofthe project.

Technical support was graciously providedby the following experts: Dan Austin, MichaelChamberland, Phil Jenkins, and Charlotte andJohn Reeder at the UA herbarium; Tom Huelsof the UA ornithology collection; GeorgeBradley of the UA herpetology collection; andYar Petryszyn and Melanie Bucci of the UAmammal collection. Thanks to Sharon Megdal


and Peter Wierenga, the current and formerdirectors, respectively, of the UA WaterResources Research Center, and all their staff.Thanks to Mau-Crimmins et al. (2005) andSprouse et al. (2002) for use of their back-ground information on the park and AaronFlesch (2001) for use of some of his discussionin the mammal and amphibian sections. Wereceived helpful reviews of earlier versions ofthis report from Danielle Foster, NatashaKline, Jeff Lovich, Theressa Mau-Crimmins,Larry Norris, Cecil Schwalbe, Don Swann, andMeg Wessner.


Checklist of North American birds, seventhedition. American Ornithologists’ Union and AllenPress, Inc., Lawrence, KS.


Anable, M. E., M. P. McClaran, and G. B. Ruyle.1992. Spread of introduced Lehmann's lovegrassEragrostis lehmanniana Nees. in southernArizona, USA. Biological Conservation 61: 181–188.

Aslan, C. 2000. Mammalian predators in the RinconMountain District of Saguaro National Park: Ahigh-elevation survey using remote photography.Unpublished report to University of ArizonaConservation Biology Internship Program andSaguaro National Park, Tucson, AZ.

Bailey, R. G. 1998. Ecoregions: The ecosystemgeography of the oceans and continents. Springer-Verlag, New York, NY.

Bailey, S. J. 1994. Saguaro National Monumentsensitive raptor surveys. Unpublished notes (for12 through 17 April 1994) to Saguaro NationalPark, Tucson, AZ.

Baird, K. J., R. MacNish, and D. P. Guertin. 2000.An evaluation of hydrologic and riparianresources in Saguaro National Park, Tucson,Arizona. Department of Hydrology and WaterResources, Research Laboratory for RiparianStudies, University of Arizona, Tucson, AZ.

Baisan, C. H., and T. W. Swetnam. 1990. Firehistory on a desert mountain range: Rincon Moun-tain Wilderness, Arizona, USA. Canadian Journalof Forestry Research 20: 1559–1569.

Baker, R. J., L. C. Bradley, R. D. Bradley, J. W.Dragoo, M. D. Engstrom, R. S. Hoffmann, C. A.Jones, F. Reid, D. W. Rice, and C. Jones. 2003a.Revised checklist of North American mammalsnorth of Mexico, 2003. Occasional Papers of theMuseum of Texas Tech University 229: 1–23.

Berner, L. R., and R. W. Mannan. 1992. Survey forsensitive raptors in the Rincon Mountains ofSaguaro National Monument, Arizona. Finalreport to Saguaro National Park, Tucson, AZ.

Bibby, C. J., N. D. Burgess, and D. A. Hill. 2002.Bird census techniques. Academic Press, London,England.

Black, K. L. 1982. An annotated checklist of theherpetofauna of Saguaro National Monument’sRincon Mountain Unit. Unpublished report toSaguaro National Park, Tucson, AZ.

Bonham, C. D. 1989. Measurements for terrestrialvegetation. John Wiley and Sons, Inc., New York,NY.

Bonine, K. E., and C. R. Schwalbe. 2003. Herpeto-fauna inventory of the Madrona Pools area. InMadrona pools PULSE study, edited by T.Edwards and D. E. Swann, pp. 44–61. Final reportto Saguaro National Park, Tucson, AZ.

Bowers, J. E. 1984. Woodland and forest flora andvegetation of Saguaro National Monument.Unpublished report to Saguaro National Park,Tucson, AZ.

Bowers, J. E., and S. P. McLaughlin. 1987. Floraand vegetation of the Rincon Mountains, PimaCounty, Arizona. Desert Plants 8: 50–95.

Briggs, M. K., L. Harris, J. Howe, and W.Halvorson. 1996. Using long-term monitoring tounderstand how adjacent land development affectsnatural areas: An example from Saguaro NationalPark, Arizona (USA). Natural Areas Journal 16:354–361.

Bucci, M. 2001. Final report: Small mammal trap-ping, Saguaro National Park Expansion Area andRocking K Ranch. Unpublished report to SaguaroNational Park, Tucson, AZ.

Corman, T. E., and C. Wise-Gervais, editors. 2005.Arizona breeding bird atlas. University of NewMexico Press, Albuquerque,NM.

Davis, R., and R. Sidner. 1992. Mammals of wood-land and forest habitats in the Rincon Mountainsof Saguaro National Monument, Arizona. Tech-nical Report 47. Cooperative Park ResourcesStudy Unit, School of Natural Resources, Univer-sity of Arizona, Tucson, AZ.

Dimmitt, M. A. 2000. Biomes and communities ofthe Sonoran Desert region. In A natural history ofthe Sonoran Desert, edited by S. J. Phillips and P.W. Comus. Arizona-Sonora Desert MuseumPress, Tucson, AZ.

Doll, M. H., R. V. Ellis, R. L. Hayes, R. Sidner, H.McCrystal, R. Davis, and R. L. Hall. 1989. Achecklist of the herpetofauna and mammals ofSaguaro National Monument. Southwest Parksand Monuments Association, Tucson, AZ.

Drewes, H. 1977. Geologic map and sections ofRincon Valley Quadrangle, Pima County, Arizona.U.S. Geological Survey, Miscellaneous Investi-gations Series Map I–997.

Duncan, D. K. 1990. Nocturnal rodent populationsand associated vegetation with implications of


human use at Saguaro National Monument,Arizona. Technical Report 35. CooperativeNational Park Service Resources Study Unit,School of Natural Resources, University ofArizona, Tucson, AZ.

Felley, D. L., and T. E. Corman. 1993. Spring 1993cactus ferruginous pygmy-owl surveys in Arizona.Unpublished report. Nongame and endangeredwildlife program, Arizona Game and Fish Depart-ment, Phoenix, AZ.

Fishbein, M. 1995. Flora of Madrona Canyon.Unpublished report to Saguaro National Park,Tucson, AZ.

Fishbein, M., and J. E. Bowers. 1996. Additions andcorrections to the Flora of the Rincon Mountains,Pima County, Arizona. Unpublished report toSaguaro National Park, Tucson, AZ.

Fishbein, M., S. McMahon, G. Ferguson, V. Stein-mann, and A. Johnson. 1994a. Flora of ChimeneaCanyon, Saguaro National Monument, East Unit.Unpublished report to Saguaro National Park,Tucson, AZ.

Fishbein, M., V. Steinmann and A. Johnson. 1994b.Floristic Survey of the proposed Box CanyonProtected Natural Area: Final report. Unpublishedreport to Saguaro National Park, Tucson, AZ.

Fitzgerald, C. S. 1996. Large mammal inventory. InWildlife inventory of the Saguaro National Parkexpansion area, edited by L. K. Harris, pp. 20–23.Unpublished report to Saguaro National Park,Tucson, AZ.

Funicelli, C. S., P. J. Anning, and D. S. Turner. 2001.Long-term vegetation monitoring at SaguaroNational Park: A decade of change. TechnicalReport No. 70. USGS Sonoran Desert ResearchStation, University of Arizona, Tucson, AZ.

Goode, M. J., D. Turner, and S. Breeden. 1998.Herpetofauna of Saguaro National Park expansionarea: Habitat assessment, inventory, threats andmanagement recommendations. Unpublishedreport to Saguaro National Park, Tucson, AZ.

Griscom, H. 2000. Saguaro National Park, Arizona,sensitive raptor survey, 2000. Unpublished reportto Saguaro National Park, Tucson, AZ.

Halvorson, W. L., and B. S. Gebow, editors. 2000.Floristic surveys of Saguaro National Parkprotected natural areas. Technical Report No. 68.USGS Sonoran Desert Research Station, Univer-sity of Arizona, Tucson, AZ.

Halvorson, W. L., and P. P. Guertin. 2003. USGSWeeds in the West project: Status of introducedplants in southern Arizona Parks. USGS, SonoranDesert Research Station, University of Arizona,Tucson, AZ.

Hoffmeister, D. F. 1986. Mammals of Arizona.University of Arizona Press and Arizona Gameand Fish Department, Tucson, AZ.

Integrated Taxonomic Information System (ITIS).2005. Accessed on 20 March 2005 from:http://www.itis.usda.gov/index.html.

Lowe, C. H., and P. A. Holm. 1991. The amphib-

ians and reptiles at Saguaro National Monument,Arizona. Technical Report No. 37. CooperativeNational Park Resources Studies Unit, Universityof Arizona, Tucson, AZ.

M’Closkey, R. T. 1980. Spatial patterns in sizes ofseeds collected by four species of heteromyidrodents. Ecology 61: 486–489.

McColly, R. A. 1961. Geology of the SaguaroNational Monument Area, Pima County, Arizona.Masters Thesis, Department of Geology, Univer-sity of Arizona, Tucson, AZ.

McPherson, G. R. 1993. Effects of herbivory andherb interference on oak establishment in a semi-arid temperate savanna. Journal of VegetationScience 4: 687–692.

Mott, D. N. 1997. Saguaro National Park, ArizonaWater Resources Report. Unpublished report bythe National Park Service, Fort Collins, CO.

Murray, R. C. 1996. Amphibian and reptile inven-tory. In Wildlife inventory of the Saguaro NationalPark expansion area, edited by L. K. Harris, pp.31–46. Unpublished report to Saguaro NationalPark, Tucson, AZ.

National Park Service (NPS). 1992. Update to thenatural/cultural resources management plan.Saguaro National Monument, Tucson, AZ.

National Park Service (NPS). 2005. NPS visitationdatabase reports. Accessed on 09 January 2005from: http://www2.nature.nps.gov/NPstats/npstats.cfm.

North American Ornithological Atlas Committee(NAOAC). 1990. Handbook for atlasing NorthAmerican breeding birds, edited by C. Smith.Accessed 13 July 2001 from: http://american-birding.org/norac/atlascont.htm.

Pima Association of Governments (PAG). 2005.Population estimates 1980–2004. Accessed onAug 22, 2005 from: http://www.pagnet.org/popu-lation/data/Est1980-2004.htm.

Pima County Flood Control District (PCFCD).2005. Unpublished data from Manning CampALERT rain gauge. Accessed 23 June 2005 from:http://rfcd.pima.gov/alertsys/.

Powell, B. F., W. L. Halvorson, and C. A. Schmidt.2006. Vascular plant and vertebrate inventory ofSaguaro National Park, Rincon Mountain District.USGS Open File Report 2006-1075. U.S.Geolog-ical Survey, Southwest Biological SciencesCenter, Sonoran Desert Research Station, Univer-sity of Arizona, Tucson, AZ.

Rondeau, R., and R. Van Devender. 1992. Floristicsurvey of the proposed Wildhorse CanyonProtected Natural Area. Unpublished report toSaguaro National Park, Tucson, AZ.

Schwalbe, C. R., T. C. Esque, M. J. Nijhuis, D. F.Haines, J. W. Clark, and P. J. Swantek. 1999.Effects of fire on Arizona Upland Desertscrub atSaguaro National Park. In A century of parks insouthern Arizona: Second conference on researchand resource management in southern Arizonanational parks, extended abstracts, edited by L.


Benson and B. Gebow, editors, pp. 107–109.National Park Service, Southern Arizona Officeand USGS Sonoran Desert Research Station,University of Arizona, Tucson, AZ.

Short, K. C. 1996. Results of the Arizona BreedingBird Atlas effort in Saguaro National Park.Unpublished report to Saguaro National Park,Tucson, AZ.

Sidner, R. 1991. Survey for the endangered lesserlong-nosed bat (Leptonycteris curasoae) onSaguaro National Monument: May–September1991. Unpublished report to Saguaro NationalPark, Tucson, AZ.

Sidner, R. 2003. Bat diversity at the Madrona Poolsarea. In Madrona pools PULSE study, edited byT. Edwards and D. E. Swann, pp. 74–83. Unpub-lished report, Saguaro National Park, Tucson.

Sidner, R., and R. Davis. 1994. Bats of SaguaroNational Monument, 1992–1994. Unpublishedreport to Saguaro National Park, Tucson, AZ.

Sprouse, T., R. Emanuel, and B. Tellman. 2002.Final report: Surface water quality monitoringoverview and assessment for the Sonoran DesertNetwork, National Park Service. Unpublishedreport. Water Resources Research Center, Univer-sity of Arizona, Tucson, AZ.


Steenbergh, W. F., and C. H. Lowe. 1977. Ecologyof the saguaro: II. Reproduction, germination,establishment, growth, and survival of the youngplant. National Park Service Monograph SeriesNo. 8, Washington, D.C.

Sumner, L. 1951. A biological survey of SaguaroNational Monument. Unpublished report toSaguaro National Park, Tucson, AZ.

Swann, D. E. 1997. Status of leopard frogs (GenusRana) in the Rincon Mountain Unit of SaguaroNational Park: Results of surveys, summer 1996.Draft final report to Saguaro National Park,Tucson, AZ.

Swann, D. E. 1999. Surveys of reptiles and amphib-ians, Rocking K and Saguaro National ParkExpansion Area. Unpublished report to SaguaroNational Park, Tucson, AZ.

Swann, D. E. 2003. Mammal inventory of theMadrona Pools area. In Madrona pools PULSEstudy, edited by T. Edwards and D. E. Swann, pp.88–102. Unpublished report, Saguaro National

Park, Tucson, AZ.Swann. D. E. 2004. A checklist of herpetofauna and

mammals of Saguaro National Park, Arizona.Western National Parks Association, Tucson, AZ.

Swann, D. E., V. Gempko, and B. F. Powell. 2003a.Special surveys for lowland leopard frogs,mammals, and plants on the remote east slope ofthe Rincon Mountains. Unpublished report to theSonoran Desert Inventory and MonitoringProgram, Saguaro National Park, and DesertSouthwest Cooperative Ecosystems Study Unit,Tucson, AZ.

Swantek, P. J., W. L. Halvorson, and C. R.Schwalbe. 1999. GIS database development toanalyze fire history in southern Arizona andbeyond, an example from Saguaro National Park.Technical Report 61. U.S. Geological Survey,Arizona National Park Resources Unit, Univer-sity of Arizona, Tucson, AZ.

Tucson Bird Count (TBC). 2005. Accessed fromweb on August 4, 2005: http://www.tucson-birds.org/index.html.

United States Department of Agriculture (USDA).2005. The PLANTS Database, Version 3.5.National Plant Data Center, Natural ResourcesConservation Service, Baton Rouge, LA.Accessed on 10 March 2003 from:http://plants.usda.gov.

Warshall, P. 1994. The biopolitics of the Mt.Graham red squirrel (Tamiasciuris hudsonicusgrahamensis). Conservation Biology 8: 977–988.

Western Regional Climate Center (WRCC). 2005.Arizona climate summaries for the University ofArizona. Accessed on 15 February 2005 from:http://www.wrcc.dri.edu/summary/climsmaz.html.

White, G. C., K. P. Anderson, K. P. Burnham, andD. L. Otis. 1983. Capture-recapture and removalmethods for sampling closed populations. LosAlamos National Laboratory, Los Alamos, NM.

Whittaker, R. H., and W. A. Niering. 1965. Vegeta-tion of the Santa Catalina Mountains, Arizona: Agradient analysis of the south slope. Ecology 46:429–452.

Wolf, S. A., and D. E. Swann. 2002. Infrared-trig-gered photography of mammals in the RinconMountain District of Saguaro National Park,Arizona. Final report to Western National ParksAssociation and Saguaro National Park, Tucson,AZ.

VEGETATIVE CHARACTERISTICS OF OAK SAVANNAS IN THE SOUTHWESTERN BORDERLANDS REGIONPeter F. Ffolliott and Gerald J. Gottfried

Much has been learned about the oak (encinal)woodlands of the Southwestern Borderlandsregion in recent years. Ecological, hydrologic,and environmental characterizations have beenobtained through collaborative effortsinvolving a large number of people(McPherson 1992, 1997; DeBano et al. 1995;Ffolliott 1999, 2002; McClaran andMcPherson 1999; Gottfried et al. 2005; andothers). However, comparable characteriza-tions of the lower-elevation oak savannas arealso necessary to enhance the knowledge of allof the oak ecosystems in the SouthwesternBorderlands region. Oak savannas differ fromthe more extensive oak woodlands in that theyare more open in stand structure, and, there-fore, a higher level of herbaceous productionmight be expected.

Species compositions, density patterns, andannual growth rates of the tree overstories andspecies compositions and seasonal and annualproduction (standing biomass) of the herba-ceous plants in the understories of“representative” oak savannas of the South-western Borderlands regions are presented inthis paper. Comparisons of these vegetativecharacteristics with tree overstories and herba-ceous understories in the more densely stockedoak woodlands are also made. This collectiveinformation should contribute to the “knowl-edge-base” on the ecological relationships inthe oak savannas of the region.

STUDY PROTOCOLStudy Areas

The study areas consist of twelve watersheds,ranging from 8 to 35.5 ha in size, on theeastern side of the Peloncillo Mountains insouthwestern New Mexico. These watershedswere established by the U.S. Forest Service toevaluate the impacts of cool- and warm-seasonburning treatments on the ecological andhydrologic characteristics of the oak savannas(Gottfried et al. 2000, 2005; Neary andGottfried 2004). The areal aggregation of thesewatersheds, called the Cascabel Watersheds, is182.7 ha. The Cascabel Watersheds are situ-ated within the Malpai Borderlands in theeastern part of the Coronado National Forest,on the western edge of the Animas Valley(Figure 1). The watersheds are 1,640 to 1,705m in elevation. The nearest long-term precip-itation station indicates that annual precipitationaverages 597 mm, with nearly one-half occur-ring in the summer monsoonal season.Hendricks (1985), Vincent (1998), Osterkamp(1999), and Gottfried et al. (2000, 2005) havedescribed the geologic, physiographic, andhydrologic characteristics of the general areaof the Cascabel Watersheds. Bedrock geologyis Tertiary rhyolite overlain by Oligocene-Miocene conglomerates and sandstone. Soilsare classified as Lithic Argustolls, LithicHaplustrolls, or Lithic Ustorthents. These soilsare generally less than 50 cm to bedrock.Streamflow originating in the oak savannas ismostly intermittent in nature, and large flowsfollow high-intensity rainfall events.

The study areas that formed the basis tocompare tree overstories in the oak savannaswith those in the oak woodlands are locatedalong the southern slopes of the HuachucaMountains within the Coronado NationalForest. Touchan (1988), Gottfried and Ffolliott(2002), and Ffolliott and Gottfried (2005) hadselected these areas for earlier studies of treeoverstories in the oak woodlands. Averageelevation of these areas is 1,750 m. Annualprecipitation averages about 655 mm, nearlyequally split between the summer and winterseasons. The topography is gently sloping to amostly level terrain. Hendricks (1985) classi-fied the soils in the Casto-Martinez-CaneloAssociation. These soils are moderately fine tofine-textured and relatively deep.

The area in the oak woodlands that wasused to estimate herbage production forcomparison with the herbage production in theoak savannas was located in the CoronadoNational Memorial also on the southern slopesof the Huachuca Mountains. The area is situ-ated at 1,756 to 1,785 m in elevation. Annualprecipitation amount and its seasonal distribu-tion are similar to that of the study areasselected to compare the tree overstories. Steepslopes up to 35 percent characterize the topog-raphy. Hendricks (1985) classified the soils inthe Casto-Martinez-Canelo Association.

Study MethodsOn each of the Cascabel Watersheds, between35 and 45 sample points were establishedalong transects that were perpendicular to thestream system and situated from ridge to ridge.The intervals between the sample points variedamong the watersheds depending on the sizeand configuration (shape) of the watershedsampled. A total of 421 sample points werelocated on the watersheds. Measurements oftree overstory and herbaceous understory char-acteristics were obtained on varying-sizedplots centered over these sample points.Sampling designs and intensities on the otherstudy areas are discussed below.

Tree Overstory MeasurementsSpecies compositions and the densities of thetree overstories in the oak savannas on theCascabel Watersheds were measured on 0.1-ha circular plots. Single-stemmed trees weremeasured in terms of their diameter root collar(drc) and multiple-stemmed trees in equivalentdiameter root collar (edrc) following the proce-dures outlined by Chojnacky (1988). Adifferent study protocol was followed tomeasure the tree overstories in the oak wood-lands on the southern slopes of the HuachucaMountains. Species compositions and thedensities of tree overstories were tallied on atotal of 80 0.04-ha and 25 0.1-ha randomlylocated circular plots established as a samplingbasis for published (Touchan 1988; Gottfriedand Ffolliott 2002; Ffolliott and Gottfried2005) and unpublished tree overstory invento-ries. The procedures outlined by Chojnackywere again the basis for the individual treemeasurements in these inventories. Estimatesof the tree densities on both study areas areexpressed in numbers of trees per hectare andcorresponding cubic-meter volumes perhectare.

Annual growth rates of the tree overstoriesin both of the oak ecosystems were estimatedwith applications of a growth and yield model(Fowler and Ffolliott 1995) that is based on thevariable-density yield-table method of calcu-lating growth rates (Avery and Burkhart 2002;Husch et al. 2003). Conversions of the solu-tions of the basic growth equation in thismethod from English to metric units providedestimates of the growth rates in terms of cubic-meter volume, with tree age classes, sitequality values (Callison 1988), and current treeoverstory densities of the oak stands sampledforming the input variables. Current and futurevolumes of the stands were estimated fromthese input variables. Differences betweenthese two volume estimates represented aprediction of (net) growth rates for the periodof consideration, which is one year (annual) inthis study.

72 Ffolliott and Gottfried

Ffolliott and Gottfried 73

Figure 1. The Cascabel Watersheds (arrow) are located within the Malpai Borderlands, a325,000-ha area managed by a landowner organization implementing ecosystemmanagement on a virtually unfragmented open-space landscape.

Herbaceous Understory MeasurementsSpecies compositions and seasonal (spring andautumn) estimates of the production of thegrass, forb, and shrub plants comprising theherbaceous understories of the oak savannason the Cascabel Watersheds were obtained in2003, 2004, and 2005. The spring estimate wastaken to represent the production of early-growing plants, with the autumn estimatereflecting the production of the late-growingplants. Temperatures and antecedent soil waterderived from winter precipitation are thefactors favorable to the early growers, whileplant species that are late growers respond tothe summer monsoonal rains. Summations ofthe production of grass, forb, and shrubcomponents represented estimates of totalherbage production for the estimation periods.Total herbage production of late-growingplants in the oak woodlands was estimated inthe Coronado National Memorial at the sametime as that in the oak savannas on theCascabel Watersheds in the fall of 2005 toprovide a basis for comparison. This estimatewas obtained on six randomly located transectsof 10 plots each.

All of the estimates of herbage productionobtained in this study were expressed in kilo-grams per hectare. The weight-estimateprocedures originally outlined by Pechanecand Pickford (1937) were followed. Estimatesof on-site green weight of the herbaceousplants sampled were obtained on (appro-ximately) 0.90 m2 circular plots. Appropriatecorrections were applied in converting theseestimates of on-site green weight to actualoven-dry weights. Utilization of herbage byherbivores was minimal on the study areas.Less than 5 percent of the forage plants wasutilized annually.

RESULTS AND DISCUSSIONTree Overstories

Species CompositionsTree species comprising the oak savannas onthe Cascabel Watersheds and the oak wood-lands on the southern slopes of the HuachucaMountains were largely the same. The domi-

nant species tallied on the Cascabel watershedsincluded Emory (Quercus emoryi) (60.1% ofall trees tallied), Arizona white (Q. arizonica)(11.9%), and Toumey oak (Q. toumeyi) (4.4%),and alligator juniper (Juniperus deppeana)(15.3%). Minor components were redberryjuniper (J. coahuilensis) (2.0%), Mexicanpinyon (Pinus cembroides) (5.6%), andmesquite (Prosopis velutina) (0.7%). Speciesof trees observed in the forest inventories ofthe oak woodlands on the southern slopes ofthe Huachuca Mountains were dominantlyEmory oak (89.3%), with inter- minglingArizona white oak (8.7%), and a few scatteredalligator juniper (1.3%) and Mexican pinyon(0.7%).

Tree DensitiesAverage numbers of trees per acre and 90%confidence intervals of the tree overstories inthe oak ecosystems on the two study areas arepresented by size-class categories in Figure 2.These size-classes coincide with those listedby O’Brien (2002) in describing the resourcecharacteristics of the woodland types in theregion; that is, saplings (2.5 to 12.5 cm drc),medium trees (12.6 to 22.5 cm drc), and largetrees (22.6 cm drc and larger). The overall testof significance in analyzing the differences innumbers of trees per hectare for all of the size-classes (all trees) combined was evaluated at a0.10 level. However, because the three size-classes shown were nested within the overalltest, a Bonferroni adjustment was applied tomaintain the Type I error in a test of signifi-cance for the size-classes at a 0.30 level.

Average numbers of medium and largetrees and all trees combined per hectare in theoak savannas on the Cascabel Watersheds weresignificantly less than the correspondingnumbers of trees in the oak woodlands on thesouthern slopes of the Huachuca Mountains.However, the numbers of saplings weresimilar. The reason for the similar densities ofsaplings might be a result of the episodiccycles of obtaining (sexual) reproduction oftrees in the oak ecosystems of the South-western Borderlands region (Borelli et al.



1994). Numerous plantlets blanket the land-scape when the reproduction events occur,largely obscuring the effects of site onsurviving tree numbers. Such a regenerationevent could have coincided with the approxi-mate age of the saplings on the two studyareas.

A local volume table based on the treevolumes calculated by Chojnacky (1988) wasthe basis for converting the average numbersof trees per hectare to corresponding estimatesof average cubic- meter volume per hectare. Itwas estimated by these conversions that thevolume of the trees in the oak savannas on theCascabel Watersheds was 11.2 ± 3.46 (mean ±[t0.10 x standard error]) m3/ha, while the esti-mated volume of the trees in the more denseoak woodlands on the southern slope of theHuachuca Mountains was 18.5 ± 2.43 m3/ha.Statistically significant differences in thecubic-meter volumes of the trees grouped bysize-class categories (O’Brien 2002) and all

trees combined were similar to the patternobserved in the estimated numbers of trees perhectare. Over 90 percent of the volume of thetrees in the two oak ecosystems was containedin large trees.

Annual Growth RatesTrees in the oak ecosystems in the South-western Borderlands region grow slowly,rarely exceeding a fraction of a cubic meter perhectare each year (McPherson 1992, 1997;Ffolliott 1999, 2002; McClaran andMcPherson 1999), a value equivalent to agrowth rate that is less than one percent of thevolume of the trees. More specifically in thisstudy, the estimated annual growth rate of thetree overstory in the oak savannas on theCascabel Watersheds was nearly 0.049 ± 0.016m3/ha, while that in the oak woodlands on thesouthern slope of the Huachuca Mountainswas 0.078 ± 0.011 m3/ha. While a differencein annual growth rates is suggested, confidence

Figure 2. Average numbers of trees per hectare, and the 90% confidence intervals of thetree overstories, in the oak savannas on the Cascabel Watersheds and the oakwoodlands on the southern slopes of the Huachuca Mountains.

intervals for the respective estimates are onlyapproximate because the errors in the growthand yield model forming the basis for theseestimates are unknown.

Increment-core analyses indicated that theannual growth rates of trees in the oak standson both study areas were “comparatively fast”in their early and middle stages of their devel-opment; that is, for the saplings and mediumtrees. However, growth rates slowed as thetrees got older to the point where the annualgrowth rates become negligible.

Spatial Distributions of TreesThe relative variability in the spatial distribu-tions of tree overstories in the oak savannas onthe Cascabel Watersheds was greater (coeffi-cient of variation = 7.98) than that observedfor tree overstories in the oak woodlands onthe southern slopes of the Huachuca Moun-tains (coefficient of variation = 1.08). Thisdifference in spatial distributions suggests amore heterogeneous stocking condition in oaksavannas than in oak woodlands (Ffolliott andGottfried 2005). However, this possibility mustbe “conditioned” by the fact that differentsampling designs and intensities wereemployed on the two study areas. Neverthe-less, the openings of varying sizes and shapesthat are interspersed among the tree oversto-ries in the oak savannas were less commonlyencountered in the oak woodlands.

Herbaceous UnderstoriesSpecies CompositionsPlant species observed in the herbaceous under-stories of the oak savannas on the CascabelWatersheds and the oak woodlands on theCoronado National Memorial were similar.Included among the perennial grasses wereblue (Bouteloua gracilis), sideoats (B. curtipen-dula), slender (B. repens), and hairy (B. hirsuta)grama, bullgrass (Muhlenbergia emersleyi),common wolfstail (Lycurus phleoides), andTexas bluestem (Schizachyrium cirratum).Forbs were a minor herbaceous component onboth sites. There were also only minimal occur-rences of half-shrubs such as beargrass (Nolina

microcarpa), fairyduster (Calliandra erio-phylla), and common sotol (Dasylirionwheeleri). Scattered shrubs included Fendler’sceanothus (Ceanothus fendleri), Mexicancliffrose (Purshia mexicana), and pointleafmanzanita (Arctostaphylos pungens). Occur-rences of annual plants in the herbaceousunderstories of the two oak ecosystems wereminimal.

Annual ProductionAverages and 90% confidence intervals for theestimated seasonal production of early- andlate-growing grasses, forbs, and shrubs and theannual production of all early- and late-growing herbaceous plants in the oak savannason the Cascabel Watersheds for the three yearsof the study are summarized in Table 1.Because the values for the production ofgrasses, forbs, and shrubs were nested withinthe overall estimate of total herbage produc-tion, a Bonferroni adjustment was also appliedin determining the statistically significantdifferences in the components of herbageproduction that are shown in this table. Thelevels of significance in these analyses weresimilar to those specified in analyzing thedifferences in numbers of trees per acre.

There were no significant differencesbetween early- and late-growing forbs or early-and late-growing shrubs in the three years ofherbage estimation. However, this was not thecase with grasses, with the production of early-growing grasses significantly less than theproduction of late-growing plants. The occur-rence of higher than the normally expected latesummer rains in 2005 was the likely cause ofa significantly higher estimated production oflate-growing grasses (mainly grama) thanearly-growing grasses in that year. This differ-ence in the production of grasses was alsoreflected by the significant differences in theproduction of total herbage in the oaksavannas. With the exception of the estimatesof the production of late-growers in 2005, allof the herbage estimates on the CascabelWatersheds were obtained at a time ofprolonged drought.



Estimated total herbage production of late-growing plants in the understories of the oakwoodlands on the Coronado National Memo-rial in the autumn of 2005 was 174.8 ± 26.5kg/ha. This value was significantly less thanthe 297.1 ± 13.6 kg/ha of late-growing plantsin the oak savannas on the Cascabel Water-sheds at the same time. It appears from thisinitial analysis of the comparison, therefore,that a higher level of production of late-growing herbage plants might (in fact) beexpected in the more open oak savannas thanthe oak woodlands as suggested above.However, the estimates of herbage productionshould continue for a sufficient number ofyears to adequately represent the variabilitythat precipitation in the region might have onherbage production before drawing definitiveconclusions on the reported comparison.

Overstory-Understory Relationships. Comparisons of the frequently observed rela-tionship of increasing herbage production withdecreasing tree-overstory densities (Ffolliottand Clary 1982) indicated that there were nostatistically significant correlations betweeneither the production of early- or late-growingherbaceous plants and tree overstory densityin the oak savannas on the Cascabel Water-sheds or the production of late-growingherbaceous plants and tree overstory densityin the oak woodlands on the CoronadoNational Memorial. That is, though theaverage tree overstory density in the oaksavannas was significantly less than that in the

oak woodlands, the relationships between thecorresponding estimates of herbage productionand tree overstory density on the sample plotson the respective study areas were not statisti-cally significant for the range of tree overstorydensities sampled on the areas.

Similar results showing little statisticalcorrelation between herbage production andtree overstory density have also been reportedin earlier studies in the oak woodlands of theregion (Gottfried and Ffolliott 2002; Ffolliottand Gottfried 2005). It is possible, therefore,that tree overstory density in itself might notbe a significant factor in “controlling” theproduction of herbaceous plants in the oakecosystems of the southwestern United States.This possibility is further strengthened by thefinding that there was no statistical correlationbetween annual herbage production and treedensities in the Gambel oak (Q. gambelii)stands that intermingle with the ponderosa pine(Pinus ponderosa) forests on the Beaver CreekWatersheds in northern Arizona (Reynolds etal. 1970).

While the densities of the oak overstorieson the savannas and the oak woodlands mightnot affect the level of herbage production inthese two oak ecosystems, the reverse situa-tion might not always be true. McClaran andMcPherson (1999) reported that a dense coverof perennial grasses can limit successfulEmory oak regeneration on ungrazed sites inthe Southwestern Borderland region. However,further study is necessary to verify the extentof this situation.

Table 1. Averages and 90% confidence intervals of the seasonal production of herbaceousplants in the oak savannas on the Cascabel Watersheds for 2003, 2004, and 2005.

Plant group Spring Autumn

————————-kg/ha—————————Grasses 105.6 ± 13.3 176.6 ± 23.5Forbs 26.0 ± 12.7 36.2 ± 12.1Shrubs 22.8 ± 11.9 21.5 ± 11.9 Total 154.4 ± 13.5 234.3 ± 17.6

SUMMARYIn addition to presenting baseline informationon the vegetative characteristics of the oaksavannas in the Southwestern Borderlandsregion, this chapter also reports on the differ-ences in the respective tree overstories andherbaceous understories characterizing the oaksavannas on the Cascabel Watersheds and theoak woodlands on the southern slopes of theHuachuca Mountains, respectively. Tree over-stories in the oak savannas were less dense(more open) and stocking conditions appearedto be more heterogeneous than in the oakwoodlands. As expected, therefore, a higherestimate of the production of late-growingherbaceous plants was observed in the oaksavannas than in oak woodlands in the oneyear of estimation.

Assuming that the findings reported in thischapter represent the more general case, itmight be that the two oak ecosystems in theSouthwestern Borderlands region — which areboth in the Upper Encinal Type (Turner et al.2003) and classified in the Quercus emoryi/Bouteloua curtipendula (Emory oak / sideoatsgrama) habitat type — should not be consid-ered “homogeneous management units” interms of their tree overstory and herbaceousunderstory characteristics or other naturalresources values linked to these characteristicssuch as livestock production potentials or thequality of wildlife habitat. While such a differ-entiation is often made in the management ofoak savannas and more extensive oak wood-lands in California (Standiford 1991, 2002;Pillsbury et al. 1997), more extensive study isnecessary to verify this possibility in theSouthwestern Borderlands region.

ACKNOWLEDGMENTSThis research and the preparation of this paperwas supported by the Southwestern Border-lands Ecosystem Management Project of theRocky Mountain Research Station, U.S. ForestService, and the Arizona Agricultural Experi-ment Station, University of Arizona, Tucson.

LITERATURE CITED Avery, T. E., and H. E. Burkhart. 2002. Forest meas-

urements. McGraw-Hill Book Company, NewYork, NY.

Borelli, S., P. F. Ffolliott, and G. J. Gottfried. 1994.Natural regeneration in encinal woodlands ofsoutheastern Arizona. The Southwestern Naturalist39: 179–183.

Callison, J. C. 1988. Site quality indices in oakwoodlands. Ph.D. Dissertation, University ofArizona, Tucson, AZ.

Chojnacky, D. C. 1988. Juniper, pinyon, oak, andmesquite volume equations for Arizona. USDAForest Service, Research Paper INT-391, Logan,UT.

DeBano, L. F., P. F. Ffolliott, G. J. Gottfried, R. H.Hamre, and C. B. Edminster, technical coordina-tors. 1995. Biodiversity and management of theMadrean Archipelago: The sky islands of south-western United States and northern Mexico.USDA Forest Service, General Technical Report.RM- 264, Fort Collins, CO.

Ffolliott, P. F. 1999. Encinal woodlands in the south-western United States. In Ecology andmanagement of forests, woodlands, and shrub-lands in the dryland regions of the United Statesand Mexico: Perspectives for the 21st century,edited by P. F. Ffolliott and A. Ortega-Rubio, pp.69–81. Centro de Investigaciones Biologicas delNoroeste, La Paz, Baja California Sur, Mexico.

Ffolliott, P. F. 2002. Ecology and management ofevergreen oak woodlands in Arizona and NewMexico. In Oak forest ecosystems: Ecology andmanagement for wildlife, edited by W. J. McSheaand W. M. Healy, pp. 304–316. The JohnsHopkins University Press, Baltimore, MD.

Ffolliott, P. F., and W. P. Clary. 1982. Understory-overstory vegetation relationships: An annotatedbibliography. USDA Forest Service, GeneralTechnical Report INT-136, Logan, UT.

Ffolliott, P. F., and G. J. Gottfried. 2005. Vegetativecharacteristics of oak savannas in the southwesternUnited States: A comparative analysis with the oakwoodlands in the region. In Connecting mountainland and desert seas: Biodiversity and manage-ment of the Madrean Archipelago II, complied byG. J. Gottfried, B. S. Gebow, L. G. Eskew, and C.B. Edminster, pp. 399–402. USDA Forest Service,Proceedings, RMRS-P-36, Fort Collins, CO.

Fowler, W. P., and P. F. Ffolliott. 1995. A growth andyield model of Emory oak: Applications on water-shed lands in the southwestern United States. InBiodiversity and management of the MadreanArchipelago: The sky islands of southwesternUnited States and northern Mexico, technicalcoordination by L. F. DeBano, P. F. Ffolliott, G. J.Gottfried, R. H. Hamre, and C. B. Edminster, pp.347–350. USDA Forest Service, General Tech-nical Report RM-264, Fort Collins, CO.

Gottfried, G. J., and C. B. Edminster. 2005. The



Forest Service, Rocky Mountain ResearchStation’s Southwestern Borderlands EcosystemManagement Project: Building on 10 years ofsuccess. In Connecting mountain islands anddesert seas: Biodiversity and management of theMadrean Archipelago II, compiled by G. J.Gottfried, B. S. Gebow, L. G. Eskew, and C. B.Edminister, pp. 237–240. USDA Forest Service,Proceedings RMRS-P-36, Fort Collins, CO.

Gottfried, G. J., and P. F. Ffolliott. 2002. Notes onherbage resources in encinal woodlands. InMeeting resource management information needs:Fourth conference on research and resourcemanagement in the southwestern deserts, editedby W. L. Halvorson and B. S. Gebow, pp. 53–55.U.S. Geological Survey, Sonoran Desert FieldStation, University of Arizona, Tucson, AZ.

Gottfried, G. J., D. G. Neary, and R. J. Bemis, 2000.Watershed characteristics of oak savannas in thesouthwestern borderlands. Hydrology and WaterResources in Arizona and the Southwest 30:21–28.

Gottfried, G. J., B. S. Gebow, L. G. Eskew, and C.B. Edminister, compilers. 2005. Connectingmountain islands and desert seas: Biodiversity andmanagement of the Madrean Archipelago II.USDA Forest Service, Proceedings RMRS-P-38,Fort Collins, CO.

Hendricks, D. M. 1985. Arizona soils. College ofAgriculture, University of Arizona, Tucson,AZ.

Husch, B., T. W. Beers, and J. A. Kershaw, Jr. 2003.Forest mensuration. John Wiley & Sons, Inc.,New York, NY.

McClaran, M. P., and G. R. McPherson. 1999. Oaksavannas in the American Southwest. In Savannas,barrens, and rock outcrop plant communities ofNorth America, edited by R. C. Anderson, J. S.Fralish, and J. M. Baskin, pp. 275–287.Cambridge Press, New York, NY.

McPherson, G. R. 1992. Ecology of oak woodlandsin Arizona. In Ecology and management of oakand associated woodlands, technical coordinationby P. F. Ffolliott, G. J. Gottfried, D. A. Bennett, V.M. Hernandez C., A. Ortega-Rubio, and R. H.Hamre, pp. 24–33. USDA Forest Service, GeneralTechnical Report RM-218, Fort Collins, CO.

McPherson, G. R. 1997. Ecology and managementof North American savannas. University ofArizona Press, Tucson, AZ.

Neary, D. G., and G. J. Gottfried. 2004. Geomor-phology of small watersheds in an oak encinal inthe Peloncillo Mountains. Hydrology and WaterResources in Arizona and the Southwest 34:65–71.

O’Brien, R. A. 2002. Arizona’s forest resources,1999. USDA Forest Service, Resource BulletinRMRS-RB-2, Logan, UT.

Osterkamp, W. R. 1999. Runoff and sediment yieldfrom proxy records: Upper Animas Valley, NewMexico. USDA Forest Service, Research PaperRMRS-18, Fort Collins, CO.

Pechanec, J. F., and G. D. Pickford. 1937. A weightestimate method for determination of range orpasture production. Journal of the AmericanSociety of Agronomy 29: 894–904.

Pillsbury, N. H., J. Verner, and W. D. Tietje, tech-nical coordinators. 1997. Oak woodlands:Ecology, management, and urban interface issues.USDA Forest Service, General Technical ReportPSW-160, Berkeley, CA.

Reynolds, H. G., W. P. Clary, and P. F. Ffolliott.1970. Gambel oak for southwestern wildlife.Journal of Forestry 68: 45–547.

Standiford, R B., technical coordinator. 1991.Proceedings of the symposium on oak woodlandsand hardwood rangeland management. USDAForest Service, General Technical Report PSW-126, Berkeley, CA.

Standiford, R. B. 2002. California’s oak woodlands.In Oak forest ecosystems: Ecology and manage-ment for wildlife, edited by W. J. McShea and W.M.. Healy, pp. 280–303. The Johns HopkinsUniversity Press, Baltimore, MD.

Touchan, R. 1988. Growth and yield of Emory oak.In Oak woodland management: Proceedings ofthe workshop, edited by P. F. Ffolliott, and J. D.Hasbrouck, pp. 11–18. School of RenewableNatural Resources, University of Arizona, Tucson,AZ.

Turner, R. M., J. W. Webb, J. E. Bowers, and J. R.Hastings. 2003. The changing mile revisited.University of Arizona Press, Tucson, AZ.

Vincent, K. R. 1998. Tectonics and earthquakehazards of the southern Animas Valley, HidalgoCounty, New Mexico. Bureau of Mines andMineral Resources, State of New Mexico, SantaFe, NM.

DISTRIBUTION AND CONSERVATION PROTECTION OF NATURAL LAND COVERIN THE SONORAN DESERT ECOREGION IN ARIZONA, AS DESCRIBED BY THESOUTHWEST REGIONAL GAP ANALYSIS PROJECTKathryn A. Thomas, Keith A. Schulz, and Ken Boykin

The goal of the U.S. Geological Survey’s(USGS) National Gap Analysis Program(GAP) is to keep common species common byidentifying vertebrate species and vegetation(plant communities) not adequately repre-sented on lands managed for conservation(Gap Analysis Program 2006). The conserva-tion assessment provided by gap analysis isproactive in the sense that it provides infor-mation on the conservation management ofvertebrates and vegetation not listed as threat-ened and endangered, as well as those that are.Gap analysis products can be used for better-informed policy and land managementdecisions, which can potentially avert theecological, economic, and legal dilemmasposed when a species is classified as legallyendangered.

Gap analysis is conducted using threeprimary types of digital map products: 1) aland cover map, 2) a conservation managementmap consisting of stewardship (i.e., ownership)land boundaries and the conservation statuslevels assigned to those lands, and 3) vertebratepredicted-distribution maps. While we refer tothese digital products as “maps” in this text,each is a digital geodatabase developed withina geographic information system (GIS). Theconservation management map indicates thelegal mandates for conservation managementfor each ownership parcel, coded by four statuslevels (1–4). The fundamental assumption isthat management status 1 and 2 provideadequate protection for the long-term viability

of biota on those lands (Gap Analysis Program2000). The gap analysis is conducted as anoverlay of the conservation management mapwith the land cover map and with each of thevertebrate species maps to determine the repre-sentation of vegetation and vertebrate speciesin each of the status levels (Scott et al. 1993).

The Southwest Regional Gap AnalysisProject (SWReGAP) has mapped land cover,stewardship, conservation management, andpredicted vertebrate distribution in the five-state Southwest region. This multi-institutional effort included Arizona, Colorado,Nevada, New Mexico, and Utah and representsthe first formal regional Gap Analysis Project(Prior-Magee et al. 2007). Utah State Univer-sity coordinated land-cover mapping efforts,and New Mexico State University coordinatedvertebrate, stewardship, and conservationmanagement mapping. Other collaboratinginstitutions included the U.S. GeologicalSurvey Southwest Biological Science Centerin Arizona, the U.S. Environmental ProtectionAgency in Nevada, the Colorado Division ofWildlife, and Colorado State University. Theproject was assisted by NatureServe ecologistswho coordinated development and applicationof the ecological system legend for land cover.Final products for the entire region can beobtained from the National Gap AnalysisProgram website (Gap Analysis Program2006).

The land cover map describes natural andsemi-natural vegetation as identified by

82 Thomas, Schulz, and Boykin

ecological systems developed from the Inter-national Terrestrial Ecological SystemsClassification (Comer et al. 2003). Ecologicalsystems “represent recurring groups of biolog-ical communities that are found in similarphysical environments and are influenced bysimilar dynamic ecological processes, such asfire or flooding” (NatureServe Explorer 2006).They relate to the National Vegetation Classi-fication System (Federal Geographic DataCommittee 1997) in that both describe char-acteristic vegetation associations and alliances.The online NatureServe Explorer (NatureServeExplorer 2006) provides a description of the636 ecological systems currently identified.The 109 ecological systems identified for the5-state SWReGAP region are also describedonline (SWReGAP Landcover LegendDescriptions 2006).

While the final products for SWReGAP arenow available for the five-state region and foreach of the five participating states, additionalgap analyses can be applied to other bound-aries within the region, such as ecoregions.Fifteen ecoregions are contained in part orentirely within the SWReGAP study area. Weare here reporting gap analysis findings withinthe Arizona portion of the Sonoran ecoregion(hereafter referred to as the Arizona Sonoran),which includes much of the Arizona Uplandand Lower Colorado Valley portions of theSonoran Desert as described by Shreve (1951,1964). The entire Sonoran ecoregion (Figure1) is 223,425 km2 and is comprised of parts ofArizona and California in the United States,and Baja del Norte and Sonora in Mexico. Weconducted the ecoregional gap assessment forthe Arizona Sonoran only, since comparabledistribution data for land cover, vertebratespecies, and conservation management do notexist for the California or Mexican portion ofthe Sonoran ecoregion. To define the ArizonaSonoran we used ecoregion boundariesdefined by The Nature Conservancy’s Ecore-gions and Divisions Map and modified fromthe U.S. Forest Service Bailey’s (1995) ecore-gion boundaries (The Nature ConservancyEcoregions and Divisions Map 2000). The

Arizona Sonoran (Figure 1) is 90,190 km2 orapproximately 40 percent of the entire Sonoranecoregion.

METHODSWe used the regional products of SWReGAPto conduct this gap analysis of the ArizonaSonoran ecological systems. We also reportdistribution statistics for mapped ecologicalsystems, stewardship, and conservationmanagement within the Arizona Sonoran.

The land cover and conservation manage-ment maps were constrained by the ArizonaSonoran boundary and summary statistics werederived for the distribution of the map cate-gories. Gap analysis was conducted by creatinggeographic intersection of the land cover withthe conservation management map, and calcu-lating summary statistics of the intersectedmaps. A geographic information system(GIS)—with ESRI ® ArcGIS Desktop 9 soft-ware and Spatial Analyst extension—was usedfor map intersection and derivation of summarystatistics. Microsoft Excel was used to furthergroup and analyze summary statistics.

DatasetsThe SWReGAP land cover map was publishedas provisional in 2004 and contains 125 legendclasses, of which 109 are ecological systemsand 16 land-use (Prior-Mcgee et al. 2007). Theminimum mapping unit is one acre (.4 ha). Themap was developed using decision-tree (alsoknown as classification and regression tree, orCART) modeling (Breiman et al. 1984,Lawrence and Wright 2001) applied to LandsatEnhanced Thematic Mapper Plus (ETM+)remotely sensed images and derivates, a digitalelevation model (DEM) and derivatives, andextensive ground observations. Full methodsfor the land cover map are reported in Lowryet al. (2007)

The SWReGAP conservation managementmap was published as provisional in 2005 andconsists of boundary information for public andprivate lands (where provided) at variable reso-lution and with conservation management statusassigned to each delineated land area (Ernst et

Thomas, Schulz, and Boykin 83

Figure 1. The extent of the Arizona portion of the Sonoran Ecoregion.

al. 2007). Four status levels are used to describethe legal mandate for conservation management(Crist 2000). A key to conservation managementstatus was used to assign the appropriate levelfor the SWReGAP project (USGS SouthwestRegional Gap Analysis Project 2006).

Status 1: An area having permanent protec-

tion from conversion of natural land cover anda mandated management plan in operation tomaintain a natural state within which distur-bance events (of natural type, frequency,intensity, and legacy) are allowed to proceedwithout interference or are mimicked throughmanagement.


Status 2: An area having permanent protec-tion from conversion of natural land cover anda mandated management plan in operation tomaintain a primarily natural state, but whichmay receive uses or management practices thatdegrade the quality of existing natural commu-nities, including suppression of naturaldisturbance.

Status 3: An area having permanent protec-tion from conversion of natural land cover forthe majority of the area, but subject to extrac-tive uses of either a broad, low-intensity type(e.g., logging) or localized intense type (e.g.,mining). It also confers protection to federallylisted endangered and threatened speciesthroughout the area.

Status 4: There are no known public orprivate institutional mandates or legally recog-nized easements or deed restrictions held bythe managing entity to prevent conversion ofnatural habitat types to anthropogenic habitattypes. The area generally allows conversion tounnatural land cover throughout.

RESULTSLand cover in the Arizona Sonoran

Thirty-eight land cover classes are mappedwithin the Arizona Sonoran; eight of theseclasses represent development, agriculture,water, and/or altered vegetation. High- andlow-intensity developed lands constitute 4.5percent of the ecoregion, and agriculture 5.5percent of the area.

Approximately 88 percent of the ecoregionwas mapped as natural land cover andconsisted of 30 ecological systems (Table 1).Example descriptions of the meso-scale terres-trial ecological systems characteristic of theArizona Sonoran are shown in Table 2.

Sonoran Paloverde-Mixed Cacti DesertScrub (Figure 2) and Sonora-MojaveCreosotebush-White Bursage Desert Scrubdominate the landscape and together accountfor nearly 77 percent of all Arizona Sonoranlandcover. The next most abundant ecologicalsystem, occurring as a distant third andcovering 3.5 percent of the land area, isApacherian Chihuahuan Mesquite Upland

Scrub, considered a range extension ofmesquite. Twenty-three ecological systemsoccur with less than 1 percent cover. Of these,nine are peripheral to the Arizona Sonoranwith less than 1 percent of their statewideoccurrence in the Arizona Sonoran (Table 1).

Six ecological systems are representedlargely in the Arizona Sonoran when consid-ered on a statewide basis (Table 1); more than70 percent of the statewide occurrence for eachis within the Arizona Sonoran ecoregion.These systems are: Sonoran Paloverde-MixedCacti Desert Scrub, Sonora-Mojave Creosote-bush-White Bursage, North American WarmDesert Active and Stabilized Dune, Sonora-Mojave Mixed Salt Desert Scrub, NorthAmerican Warm Desert Riparian MesquiteBosque, and North American Warm DesertRiparian Woodland and Shrubland. TheSonoran Paloverde-Mixed Cacti Desert Scrubis unique in that nearly its entire regional rangeis endemic to the Arizona Sonoran.

Some ecological systems known to occurin the Arizona Sonoran occur in relativelysmall patches naturally (< 1 ha) and were notmapped as separate units. These include:Chihuahuan-Sonoran Desert Bottomland andSwale Grassland, North American WarmDesert Badland, North American Warm DesertCienega, North American Warm Desert Pave-ment, North American Warm Desert Playa,Sonora-Mojave Semi-Desert Chaparral,Sonoran Fan Palm Oasis, and Sonoran GraniteOutcrop Desert Scrub.

Conservation management in the Arizona Sonoran

The major land stewards (Table 3) within theArizona Sonoran are federal agencies (51% ofthe land area), tribes (18%), private lands(17%), and state lands (13%). The Bureau ofLand Management (BLM) manages nearly26.5 percent of the land with the next largestfederal land manager being the Department ofDefense (12%). Nearly eighteen percent(17.7%) of the land area has status 1 or 2conservation management (Table 3), thegreatest legally mandated protection. Forty-


Table 1. Ecological systems within the Arizona Sonoran ecoregion and the proportion of their state andregional range that occurs within the ecoregion.

Proportion Proportion Arizona Arizona state cover regional cover Sonoran Sonoran in Arizona in Arizona

Ecological System (km2) cover (%)1 Sonoran (%) Sonoran (%)

Sonoran Paloverde-Mixed Cacti Desert Scrub 36,988.5 41.0% 93.0% 92.3%Sonora-Mojave Creosotebush-White

Bursage Desert Scrub 32,075.7 35.6% 93.0% 92.3%Apacherian-Chihuahuan Mesquite

Upland Scrub 3,185.5 3.5% 19.3% 9.9%Sonoran Mid-Elevation Desert Scrub 1,607.4 1.8% 29.8% 29.8%Chihuahuan Creosotebush, Mixed Desert

and Thorn Scrub 1,109.4 1.2% 17.6% 4.0%North American Warm Desert Active

and Stabilized Dune 1,016.1 1.1% 100.0% 35.7%Chihuahuan Mixed Salt Desert Scrub 865.1 1.0% 30.7% 19.5%Sonora-Mojave Mixed Salt Desert Scrub 817.3 0.9% 80.9% 31.8%North American Warm Desert Riparian

Mesquite Bosque 588.7 0.7% 74.1% 69.5%Apacherian-Chihuahuan Piedmont

Semi-Desert Grassland and Steppe 395.9 0.4% 3.5% 0.9%Madrean Pinyon-Juniper Woodland 265.0 0.3% 2.0% 1.2%Mogollon Chaparral 216.8 0.2% 2.3% 1.9%North American Warm Desert Riparian

Woodland and Shrubland 216.8 0.2% 80.5% 47.0%North American Warm Desert Bedrock

Cliff and Outcrop 138.9 0.2% 18.3% 3.8%Colorado Plateau Pinyon-Juniper Woodland 67.9 0.1% 0.2% 0.1%North American Warm Desert Wash 45.6 0.1% 30.0% 6.9%Chihuahuan Succulent Desert Scrub 42.8 0.1% 39.3% 22.7%Chihuahuan Stabilized Coppice Dune

and Sand Flat Scrub 41.1 0.1% 21.9% 0.7%North American Warm Desert Volcanic Rockland 22.2 < 0.1% 10.8% 2.2%Madrean Encinal 14.3 < 0.1% 0.5% 0.3%North American Warm Desert Lower

Montane Riparian Woodland and Shrubland 11.9 < 0.1% 6.6% 2.8%Madrean Pine-Oak Forest and Woodland 11.5 < 0.1% 0.3% 0.2%Madrean Juniper Savanna 11.5 < 0.1% 3.4% 1.2%Inter-Mountain Basins Juniper Savanna 2.8 < 0.1% 0.1% 0.1%Great Basin Pinyon-Juniper Woodland 1.5 < 0.1% < 0.1% < 0.1%Colorado Plateau Mixed Bedrock Canyon

and Tableland 1.2 < 0.1% < 0.1% < 0.1%Rocky Mountain Ponderosa Pine Woodland 1.0 < 0.1% < 0.1% < 0.1%Mojave Mid-Elevation Mixed Desert Scrub 0.6 < 0.1% < 0.1% < 0.1%Chihuahuan Sandy Plains Semi-Desert Grassland 0.3 < 0.1% 1.9% < 0.1%North American Arid West Emergent Marsh 0.1 < 0.1% 0.2% < 0.1%1 Arranged in descending order


Table 2. Descriptions of two extensive ecological systems in the Arizona Sonoran.

Sonoran Paloverde-Mixed Cacti Desert ScrubPrimary division: North American Warm Desert (302)Land cover class: ShrublandSpatial scale and pattern: MatrixConcept Summary: This ecological system occurs on hillsides, mesas, and upper bajadas in southernArizona and extreme southeastern California. The vegetation is characterized by a diagnostic sparse,emergent tree layer of Carnegia gigantea (3–16 m tall) and/or a sparse to moderately dense canopycodominated by xeromorphic deciduous and evergreen tall shrubs Parkinsonia microphylla and Larreatridentata, with Prosopis sp., Olneya tesota, and Fouquieria splendens less prominent. Other commonshrubs and dwarf-shrubs include Acacia greggii, Ambrosia deltoidea, Ambrosia dumosa (in driersites), Calliandra eriophylla, Jatropha cardiophylla, Krameria erecta, Lycium spp., Menodora scabra,Simmondsia chinensis, and many cacti including Ferocactus spp., Echinocereus spp., and Opuntiaspp. (both cholla and prickly pear). The sparse herbaceous layer is composed of perennial grassesand forbs with annuals seasonally present and occasionally abundant. On slopes, plants are oftendistributed in patches around rock outcrops where suitable habitat is present.

Sonora-Mojave Creosotebush-White Bursage Desert ScrubPrimary division: North American Warm Desert (302)Land cover class: ShrublandSpatial scale and pattern: MatrixConcept Summary: This ecological system forms the vegetation matrix in broad valleys, lower bajadas,plains, and low hills in the Mojave and lower Sonoran Deserts. This desert scrub is characterized bya sparse to moderately dense layer (2–50% cover) of xeromorphic microphyllous and broad-leavedshrubs. Larrea tridentata and Ambrosia dumosa are typically dominants, but many different shrubs,dwarf-shrubs, and cacti may codominateor form typically sparse understories. Associated species may include Atriplex canescens, Atriplexhymenelytra, Encelia farinosa, Ephedra nevadensis, Fouquieria splendens, Lycium andersonii, andOpuntia basilaris. The herbaceous layer is typically sparse, but may be seasonally abundant withephemerals. Herbaceous species such as Chamaesyce spp., Eriogonum inflatum, Dasyochloapulchella, Aristida spp., Cryptantha spp., Nama spp., and Phacelia spp. are common.

eight percent has status 3 management and 34percent status 4. The U.S. Fish and WildlifeService (FWS) manages the largest area ofstatus 1 and 2 lands (6,282 km2) with the BLMsecond (5,901 km2). While all of the FWSlands are status 1 and 2, 24 percent of the BLMlands are status 1 and 2, and the remainderstatus 3. The lands with the least legallymandated conservation protection are the statetrust lands (11,191 km2) and private lands, bothwith no known conservation restrictions(15,515 km2). State trust and private landrepresent nearly 87 percent of all status 4 landin the Arizona Sonoran, those with no legalmandate for conservation management.

Conservation status of ecological systems

Although we conducted the gap analysis forall 30 ecological systems of the ArizonaSonoran, we report here the results for elevenecological systems that are characteristic of theecoregion (Table 4). These eleven ecologicalsystems represent the three most abundantecological systems and additional ecologicalsystems with 30 percent or more of their stateor regional range within the Arizona Sonoran,as shown in Table 1. Collectively these 11ecological systems comprise 86 percent of theland cover of the Arizona Sonoran.

Status 1 and 2 lands provide the most


legally mandated protection to biota; so, often,the combined area of status 1 and 2 is consid-ered in assessing the level of conservationprotection. There is no set threshold of landswith protection to evaluate the adequacy ofconservation protection for a plant communityor vertebrate species. Various authors (Souléand Sanjayan 1998; Odum and Odum 1972;Specht et al. 1974; Miller 1984; Noss andCooperrider 1994) have recommended thresh-olds of 10, 20, or 50 percent of the distributionof a vegetation type or vertebrate species tohave protected management.

The percentages of status 1 and 2 protectionin the Arizona Sonoran for the 11 ecologicalsystems ranged from 0.7 percent to 40.1percent, all under the broadest threshold of 50percent (Table 4). Nine of the ecologicalsystems have 20 percent or less conservationprotection and five with less than 10 percentconservation protection. Those five systemsare: Apacherian-Chihuahuan Mesquite UplandScrub (7.9%), Chihuahuan Mixed Salt DesertScrub (1.1%), Chihuahuan Succulent Desert

Scrub (0.7%), North American Warm DesertRiparian Mesquite Bosque (5.1%), and Sonora-Mojave Mixed Salt Desert Scrub (3.0%).

Conversely lands with status protectionrepresent areas with the highest risk of loss ofbiotic populations due to habitat conversion.Sonora-Mojave Mixed Salt Desert Scrub hasthe highest proportion of area in status 4, 78percent. All other ecological systems, exceptfor North American Warm Desert Active andStabilized Dune, have between 30 and 50percent in status 4 (Table 4).

An additional consideration is the propor-tion of the state and regional status 1 and 2protection for an ecological system that occurswithin the Arizona Sonoran (Table 4). For fiveecological systems, the status 1 and 2 conser-vation management provided in the ArizonaSonoran desert provides all or a high propor-tion of the systems status 1 and 2 protectionwithin Arizona: Sonoran Paloverde-MixedCacti Desert Scrub (96%), Sonora-MojaveCreosotebush-White Bursage Desert Scrub(76%), North American Warm Desert Active

Figure 2. A typical aspect of Sonoran Paloverde-mixed Cacti Desert Scrub


Table 3. Land stewardship and conservation management in the Arizona Sonoran ecoregion.

Land area Status 1 Status 2 Status 3 Status 4km2 % km % km2 % km2 % km2 %

Federal LandsBureau of LandManagement 23,912 26.5% 63 0.3% 5,838 24.4% 18,011 75.3% 0 0.0%

Bureau of Reclamation 382 0.4% 0 0.0% 0 0.0% 0 0.0% 382 100.0%Fish and WildlifeService 6,282 7.0% 5,984 95.2% 298 4.8% 0 0.0% 0 0.0%

Forest Service 3,429 3.8% 10 0.3% 736 21.5% 2,683 78.3% 0 0.0%Defense or Energy 10,611 11.8% 0 0.0% 1,243 11.7% 9,368 88.3% 0 0.0%National Park Service 1,455 1.6% 1,453 99.9% 2 0.1% 0 0.0% 0 0.0%Tribal 15,838 17.6% 0 0.0% 0 0.0% 12,952 81.8% 2,886 18.2%

StateState Parks andRecreation 181 0.2% 0 0.0% 0 0.1% 1,819 99.9% 0 0.0%

State Land Board 11,465 12.7% 0 0.0% 237 2.1% 37 0.3% 11,191 97.6%State Wildlife Reserves 111 0.1% 0 0.0% 99 89.0% 121 0.7% 0 0.3%

Local GovernmentsRegional Government 486 0.5% 0 0.0% 0 0.0% 0 0.0% 486 100.0%City Land 165 0.2% 0 0.0% 0 0.0% 0 0.0% 165 100.0%County Land 156 0.2% 0 0.0% 0 0.0% 0 0.1% 156 99.9%Local Land Trust 2 0.0% 0 0.0% 1 65.7% 1 34.3% 0 0.0%

PrivateTNC 29 0.0% 3 10.9% 15 53.4% 10 35.7% 0 0.0%Private/Unrestricted 15,515 17.2% 0 0.0% 0 0.0% 0 0.0% 15,515 100.0%

Water1 169 0.2% 0 0.0% 0 0.0% 0 0.0% 0 0.0%

Total2 90,190 100.0% 7,514 8.3% 8,469 9.4% 43,255 48.0% 30,782 34.1%

1 Water was not assigned a status category2 Column totals indicate the total land and porportion of land in each status category; rows indicate the proportion foreach category of land owner

and Stabilized Dune (100%), Sonora-MojaveMixed Salt Desert Scrub (79%), and NorthAmerican Warm Desert Riparian Woodlandand Shrubland (71%). For the region, theproportion of status 1 and 2 protection withinthe Arizona Sonoran is considerably less forall except for the nearly endemic SonoranPaloverde-Mixed Cacti Desert Scrub (96%,Table 4) and North American Warm DesertActive and Stabilized Dune (77%).

DISCUSSIONGap analysis provides a “coarse filter” assess-ment of the conservation protection of biota(The Nature Conservancy 1982). There is nosingle statistic that fully describes the adequacyof the representation on lands with highconservation protection of an ecologicalsystem. Ecological condition of the land is notpart of the gap analysis; and past managementand other stresses (such as invasion of intro-


duced species, altered ecological processes,and habitat fragmentation) may greatly influ-ence the adequacy of land to support targetedbiodiversity.

This gap analysis provides information thatcan be used in several ways:

1) The analysis identifies ecologicalsystems or vertebrate animal habitats that lacklegal protection or have minimal legal protec-tion. Additional legal protection may berequired for adequate conservation manage-ment;

2) The analysis indicates ecological systemswhose representation within status 1 and 2

lands may be adequate, but the adequacy ofmanagement at status 1 or 2 level is notknown. Inadequate status 1 and 2 managementmay be due to lack of appropriate managementor may occur because impacts to the landpreclude the ability to manage at the status 1or 2 level. The gap analysis statistics canprovide an indication of the importance ofproviding management or mitigating impactsso that the status 1 or 2 level conservationmanagement is achieved where mandated.

3) The analysis identifies the proportion ofan ecological system within status 4 lands.This provides an indication of the impact on

Table 4. Representation of ecological systems in status 1 and 2 for the Arizona Sonoran ecoregion,Arizona state and the region and in status 4 for the ecoregion.

Proportion state Proportion regionalStatus 1 & 2 Status 4 status 1 & 2 status 1 & 2 in Arizona in Arizona in Arizona in Arizona

Ecological System Sonoran (%) Sonoran (%) Sonoran (%) Sonoran (%)

Sonoran Paloverde-MixedCacti Desert Scrub 22.9% 24.3% 96% 96%

Sonora-Mojave Creosotebush-White Bursage Desert Scrub 18.9% 27.5% 76% 37%

Apacherian-ChihuahuanMesquite Upland Scrub 7.9% 33.4% 42% 15%

Sonoran Mid-ElevationDesert Scrub 17.6% 32.7% 23% 23%

North American Warm DesertActive and Stabilized Dune 40.1% 2.9% 100% 77%

Chihuahuan Mixed SaltDesert Scrub 1.1% 23.9% 6% 6%

Sonora-Mojave Mixed SaltDesert Scrub 3.0% 78.3% 79% 5%

North American Warm DesertRiparian Mesquite Bosque 5.1% 29.2% 44% 38%

North American Warm DesertRiparian Woodland andShrubland 20.3% 48.4% 71% 52%

North American Warm DesertLower Montane RiparianWoodland and Shrubland 30.4% 40.6% 11% 5%

North American WarmDesert Wash 15.0% 47.1% 22% 7%

Chihuahuan SucculentDesert Scrub 0.7% 34.9% 5% 2%


the ecological system if its occurrence onstatus 4 lands were lost.

While status 1 and 2 lands indicate that landhas a legal mandate to be managed so as toconserve biota, legal mandate does not insurethat such management is actually imple-mented. For example, illegal immigration andvarious smuggling activities across theArizona-Mexico border have resulted in aproliferation of informal foot trails, informalroads, and deposition of large amounts of trashand human waste on status 1 and 2 lands in theborder states (Segee and Neeley 2006). Anextensive barrier is being constructed at theinternational border that has the potential tofragment movement of vertebrates across theborder (Vacariu 2004-2005). Uncontrolledinfestations of invasive biota have degradedhabitat in the Arizona Sonoran and are linkedwith changes in fire magnitude and frequency(Brooks and Pyke 2001) that can convert avegetation type.

Conversely, the assignment of land to status4 does not necessarily mean that land is notbeing managed for conservation of naturalbiota. Private land owners may enter intoconservation easements that have not beenincluded in the stewardship mapping (sincethese data are included only if voluntarilyprovided to SWReGAP). Private land ownersmay choose to manage for biodiversity withoutany such legal mandates on the land.

Our assessment of ecological systems in theArizona Sonoran identified 11 systems that areeither spatially dominant or have over a thirdor more of their range within the ArizonaSonoran. Five ecological systems have verylow status 1 and 2 protection (< 10%).However, four of those ecological systemsappear to have half or more of their state andregional status 1 and 2 protection in otherecoregions. The fifth, Sonora-Mojave MixedSalt Desert Scrub is an exception in that it haslow status 1 and 2 protection within theArizona Sonoran and this protection is nearly80 percent of all status 1 and 2 protection forthe ecological system within the state. In addi-tion, within the Arizona Sonoran 78 percent of

Sonora-Mojave Mixed Salt Desert Scrub is instatus 4 lands, which have no conservationmandate. Sonoran Paloverde-Mixed CactiDesert Scrub, Sonora-Mojave Creosotebush-White Bursage Desert Scrub, and NorthAmerican Warm Desert Riparian Woodlandand Shrubland are also notable in that theyhave between 18 and 23 percent status 1 and 2protection within the Arizona Sonoran, whichrepresents over 70 percent of the state status 1and 2 protection for these ecological systems.North American Warm Desert Active andStabilized Dune is the best protected within theArizona Sonoran (40.1% status 1 and 2) yetthis protection represents all of the state protec-tion and over three-fourths of the regionalprotection.

These results of the ecoregional gapanalysis suggest that conservation manage-ment should be maintained or increased for:Sonoran Paloverde-Mixed Cacti Desert Scrub,Sonora-Mojave Creosotebush-White BursageDesert Scrub, North American Warm DesertActive and Stabilized Dune, Sonora-MojaveMixed Salt Desert Scrub, and North AmericanWarm Desert Riparian Woodland and Shrub-land using the criteria of the high proportionof statewide status 1 and 2 protection occur-ring in the Arizona Sonoran. The SonoranPaloverde-Mixed Cacti Desert Scrub andSonora-Mojave Mixed Salt Desert Scrub areadditionally highlighted in that the former isnearly endemic to the ecoregion and the smallarea of the latter with status 1 and 2 protectionin the Arizona Sonoran represents most of thestatus 1 and 2 protection in the state.

This gap assessment for the ArizonaSonoran provides a first-level, coarse-filteranalysis of conservation protection. Our resultsshould be considered with respect to conser-vation protection provided on adjacent landsand to “fine filter” information potentiallyavailable on the biota for the areas of interest.Other criteria, such as the actual total landcover of an ecological system, especially incombination with a view of the historicalextent of the ecological system or the numberof vertebrate species the ecological system


supports, may highlight the need for mainte-nance and/or increase in the conservationprotection of other Arizona Sonoran ecolog-ical systems. The results of our ecoregionalgap analysis can be used to derive location datafor examining potential additional legal protec-tion or changes in conservation managementplanning.

ACKNOWLEDGMENTSThe SWReGAP was conducted with fundingprovided by the U.S. Geological Survey’sNational Gap Analysis Program. The ArizonaSonoran analysis was made possible by thesupport of the University of Arizona School ofNatural Resources, the New Mexico Cooper-ative Fish and Wildlife Research Unit, and theU.S. Geological Survey’s Southwest Biolog-ical Science Center Sonoran Desert ResearchStation. We thank the organizers of theBorders, Boundaries, and Time Scales confer-ence for the opportunity to present thisinvestigation. Julie Prior-Magee and DaleTurner provided invaluable comments on themanuscript. Use of product names does notimply endorsement by the U.S. GeologicalSurvey.

LITERATURE CITEDBailey, R. G. 1995. Descriptions of the Ecoregions

of the United States. U.S. Forest Service Miscel-laneous Publication #1391 (Revised), withSeparate Map at a Scale of 1:7,500,000. U.S.Department of the Interior, Forest Service, Wash-ington D.C.

Breiman, L., J. H. Friedman, R. A. Olshen, and R.and C. J. Stone. 1984. Classification and Regres-sion Trees. Wadsworth, Inc., Belmont, CA.

Brooks, M. L., and D. A. Pyke. 2001. Invasiveplants and fire in the deserts of North America. InProceedings of the Invasive Species Workshop:The Role of Fire in the Control and Spread ofInvasive Species. Fire Conference 2000: The FirstNational Congress on Fire Ecology, Prevention,and Management, edited by K. E. M. Galley andT. P. Wilson, pp. 1–14. Miscellaneous PublicationNo. 11, Tall Timbers Research Station, Talla-hassee, FL.

Comer, P., D. Faber-Langendoen, R. Evans, S.Gawler, C. Josse, G. Kittel, S. Menard, S. Pyne,M. Reid, K. Schulz, K. Snow, and J. Teague. 2003.Ecological systems of the United States: Aworking classification of U.S. terrestrial systems.

NatureServe, Arlington, VA.Crist, P. 2000. Mapping and Categorizing Land

Stewardship. Gap Analysis Program Handbook.http://gapanalysis.nbii.gov/ Accessed 11/20/06.

Ernst, A. E., S. S. Schrader, V. Lopez, J. Prior-Magee, K. Boykin, D. Schrupp, L. O’Brien, W.Kepner, K. Thomas, J. Lowry, and B. Thompson.2007. Land stewardship. In Southwest RegionalGap Analysis Final Report, edited by J. S. Prior-Magee, K. G. Boykin, D. F. Bradford, W. G.,Kepner, J. H. Lowry, D. L. Schrupp, K. A.Thomas, and B. C. Thompson, Chapter 4. U.S.Geological Survey, Gap Analysis Program,Moscow, ID.

Federal Geographic Data Committee, VegetationSubcommittee. 1997. FGDC Vegetation Classifi-cation and Information Standards – June 3, 1996Draft. FGDC Secretariat, Reston, VA.

Gap Analysis Program. 2000. A handbook forconducting Gap Analysis. U.S. Geological SurveyGap Analysis Program, Moscow, Idaho. Availableonline at: ftp://ftp.gap.uidaho.edu/products/hand-bookpdf/CompleteHandbook.pdf

Gap Analysis Program 2006. http://gapanalysis.nbii.gov/portal/server.pt?. Accessed 11/20/06.

Lawrence, R. L., and A. Wright. 2001. A rule-basedclassification system using classification andregression trees (CART) analysis. Photogram-metric Engineering and Remote Sensing 67(10):1137–1142.

Lowry, J., R. Ramsey, K. Thomas, D. Schrupp, T.Sajwaj, J. Kirby, E. Waller, S. Schrader, S.Falzarano, L. Langs, G. Manis, C. Wallace, K.Schulz, P. Comer, K. Pohs, W. Rieth, C.Velasquez, B. Wolk, W. Kepner, K. Boykin, L.O’Brien, D. Bradford, B. Thompson, and J. Prior-Magee. 2007. Mapping moderate-scale land coverover very large geographic areas within a collab-orative framework: A case study of the SouthwestRegional Gap Analysis Project (SWReGAP).Remote Sensing of Environment 108: 59–73.

Miller, K. R. 1984. The Bali action plan: A frame-work for the future of protected areas. In NationalParks, Conservation, and Development, edited byJ. A. McNeely and K. R. Miller, pp. 756–764.Smithsonian Institution Press, Washington D.C.

NatureServe Explorer. October 6, 2006.http://www.natureserve.org/explorer/#. Accessed11/20/06.

NatureServe Publications. 2006. http://www.natureserve.org/publications/usEcologicalsystems.jsp.Accessed 5/11/06.

Noss, R. F., and A. Y. Cooperrider. 1994. SavingNature’s Legacy. Island Press. Washington, D.C.

Odum, E. D., and H. T. Odum. 1972. Natural areasas necessary components of man’s total environ-ment. Proceedings of the North American Wildlifeand Natural Resources Conference. 37: 178–189.

Prior-Magee, J. S., K. G. Boykin, D. F. Bradford,W. G. Kepner, J. H. Lowry, D. L. Schrupp, K. A.Thomas, and B. C. Thompson (eds.). 2007. South-


west Regional Gap Analysis Project Final Report.U.S. Geological Survey, Gap Analysis Program,Moscow, ID.

Scott, J. M., F. Davis, B. Csuti, R. Noss, B. Butter-field, C. Groves, H. Anderson, S. Caicco, F.D’Erchia, T. C. Edwards, Jr., J. Ulliman, and G.Wright. 1993. Gap analysis: A geographicapproach to protection of biological diversity.Wildlife Monographs 123: 3–41.

Segee, B. P., and J. L. Neeley. 2006. On the line:The impacts of immigration policy on wildlife andhabitat in the Arizona borderlands. Defenders ofWildlife. Washington, D.C.

Shreve, F. 1951. Vegetation and Flora of theSonoran Desert. Volume I. Vegetation. CarnegieInstitution of Washington Publication 591.

Shreve, F. 1964. Vegetation and Flora of theSonoran Desert. Stanford University Press, Stan-ford, CA.

Soulé, M. E., and M. A. Sanjayan. 1998. Conser-vation targets: Do they help? Science 297(5359):2060–2061.

Specht, R. L., E. M. Roe, and V. H. Boughlon. 1974.Supplement No. 7. Australian Journal of Botany

Supplement Series.SWReGAP Landcover Legend Descriptions. 2006.

http://earth.gis.usu.edu/swgap/legend_desc.html(Last accessed 11/18/08).

The Nature Conservancy Ecoregions and DivisionsMap. 2000. http://gis.tnc.org/data/MapbookWeb-site/map_page.php?map_id=9. Accessed 5/15/06.

The Nature Conservancy. 1982. Natural HeritageProgram Operations Manual. The Nature Conser-vancy, Arlington, VA (unpublished).

USGS Southwest Regional Gap Analysis Project.Land Stewardship. 2006. http://fws-nmcfwru.nmsu.edu/swregap/Stewardship/Default.htm Accessed 11/20/06.

USGS Southwest Regional Gap Analysis Project.Stewardship Categorization. 2006. http://fws-nmcfwru.nmsu.edu/swregap/Stewardship/Categorization.htm Accessed 11/20/06.

Vacariu, K. 2004–2005. Ecological security on theborder: A day of reckoning for wildlife linkagesbetween the United States and Mexico. WildEarth. http://twp.org/cms/page1130.cfm. Lastaccessed 4/27/09.

VERTEBRATE SPECIES IN DESERT CAVES AND MINES — A COMPARISON BETWEEN THE CHIHUAHUAN AND SONORAN DESERTS

Thomas R. Strong

Caves are widely known to provide habitat fora variety of vertebrate species that spend all orsignificant portions of their life cycles insidethe totally dark areas. It is less well known thatcaves, and particularly cave entrance areas, canprovide an important resource for a widevariety of other species. Especially in aridregions, caves may provide temporary relieffrom extreme temperature or low humidity(Kingsley et al. 2001), they may provide densites, nest substrates, hunting locations forpredators, or hiding places to escape predators.

While many individual observations ofvertebrate species in caves have been reported,there have been few attempts to compile thisinformation in any systematic way. Kingsleyet al. (2001) compiled a list of species knownto use caves or abandoned mines in theSonoran Desert. This list was based primarilyon personal observations or knowledge of thefour authors and on records listed inHoffmeister (1986), but it did not include athorough literature review. Still, this prelimi-nary list included 67 vertebrate species (4amphibians, 14 reptiles, 11 birds, and 38mammals). The primary conclusion of thepaper was that caves and mines are significantwildlife habitat resources and worthy ofprotection through the Sonoran Desert Conser-vation Plan proposed by Pima County.

The Chihuahuan Desert covers a large areaof southern New Mexico, western Texas, andthe extreme southeastern corner of Arizona.This desert includes an even larger area ofMexico, extending far south in the central

plateau. Average annual rainfall in the Carlsbad,New Mexico, vicinity is about 35.9 cm (14.1in), with most precipitation coming during asummer rainy season. Summer daytime hightemperatures are around 35° C (95° F), andwinter lows are around –2° C (28° F).

Many caves are located in extensivedeposits of limestone and gypsum in theChihuahuan Desert. These caves provide moremoderate conditions of temperature andhumidity that may be a critical resource formany species. As in other desert regions ofNorth America, there have been no systematicstudies of vertebrate species using the cavehabitats, although Bailey (1928) mentioned theuse of caves by many species in his account ofthe vertebrate biology of the Carlsbad Cavernsregion.

The Sonoran Desert covers most of southernand western Arizona, southeastern Californiain the U.S.; and Sonora and the Baja Peninsulain Mexico. Within Arizona, typical annualprecipitation values are 6.9 cm (2.7 in) at Yumaand 27.9 cm (11.0 in) at Tucson. Summermonthly average daytime high temperaturesrange from about 38° to 42° C (100° to 107°F), and monthly average winter lows rangefrom about 4° to 8° C (39° to 46° F).

This part of Arizona is in the Basin andRange physiographic province, with broad,low valleys separated by mountain ranges.Some of these mountain ranges may be severalthousand feet above the valleys, such that themountain tops are above the Sonoran Desertbiome. Geology of this region is complex, with

94 Strong

bedrock exposures in the mountains separatedby wide valleys of alluvial deposits. Sedimen-tary rocks, including limestone and gypsum,are present in isolated patches, and igneous andmetamorphic rocks are common in the moun-tain ranges.

METHODSData in this analysis were compiled from avariety of sources. The primary sources ofinformation for the Chihuahuan Desert werein the unpublished records in the files ofCarlsbad Caverns National Park and theCarlsbad Field Office of the Bureau of LandManagement. Several internet sites have exten-sive information on vertebrate species and theirhabitat usage and requirements. In particular,Biotic Information System of New Mexico(BISON-M; BISON 2004), supported by theNew Mexico Department of Game and Fish,and NatureServe Explorer (NatureServe 2004),supported by natural heritage programs. Otherstandard literature sources were also searchedfor relevant information. Underground featuresin the Chihuahuan Desert are limited to cavesin soluble rocks. Extensive limestone andgypsum karst areas with many caves arereadily accessible in the vicinity of Carlsbad,New Mexico. Direct observations were madein many limestone and gypsum caves of thisregion.

For the Sonoran Desert, there are no denseconcentrations of caves on public lands, andthere are no file records comparable to thosein New Mexico. Kingsley et al. (2001) providea reliable list of species known to utilize cavesand mines in southern Arizona, but they didnot document species in relation to specificlocations. Other sources, in particularHoffmeister (1986) and Cockrum andPetryszyn (1991), provided valuable informa-tion, including site records. While caves arerelatively rare in the Sonoran Desert, aban-doned mines are relatively common andprovide underground resources similar tonatural caves. Direct observations were madein numerous caves and abandoned mines insouthern Arizona.

Observational evidence of vertebratespecies can take several forms. Visual obser-vations of living animals are the most reliableevidence of species using undergroundresources. These animals got into the cave ormine under their own power, which suggeststhat they made a conscious choice to use theunderground site for some purpose. The pres-ence of nests or den sites is also positiveevidence of the mammal or bird thatconstructed the nest, and it also providesevidence of the reason for the use of the under-ground feature. Similarly, the presence offeathers and egg shells indicates that birdswere nesting in the cave or mine, and thecondition of the egg shells can indicatesuccessful hatching. The presence of scat in acave or mine demonstrates that the animal wasalive while in the site, and it also indicates thatthe animal was in the site for a significantperiod of time. Tracks indicate the presence ofa live animal in the site, but they provide noevidence for the reason for or the duration ofthe visit. Skeletal material in a cave or mine iseven more ambiguous. While it can generallybe identified to species, it does not necessarilyprovide confirmation that the animal enteredthe site by its own choice, or even whether itwas alive when it entered the site. Skeletalmaterial at the bottom of a pit generally indi-cates that the animal died as a result of the fall.Although it may have entered the cave bychoice, it probably did not intend to fall into apit.

Data from all of these sources were enteredinto spreadsheets, including species, caves,dates of observation (if known), and originalobservers (if known). Each report of a speciesfrom a cave or mine was entered as a separaterecord. For common species, there were oftenmultiple reports from a single site.

RESULTSChihuahuan Desert

The literature and file searches and personalobservations provided over 730 reports of 81species of vertebrates in the caves of theChihuahuan Desert in the vicinity of Carlsbad,

Strong 95

New Mexico. These records are a little vague,because a primary source of information(Bailey 1928) reports many species as beingfound in numerous other caves near CarlsbadCavern, without any specific details. Anotherdrawback in the data obtained from the filesearches is that most reports are not fromtrained biologists, and many sightings are notidentified to species. In addition, large, easilyrecognizable, or dangerous animals, such asmountain lions, porcupines, or rattlesnakes, aremore likely to be reported than wood rats thatare so common no one feels a need to reportseeing them. However, reports for a singlecave as “unidentified bat,” “unidentifiedrodent,” and “unidentified snake” would stillcount as at least three different vertebratespecies for that cave. Vertebrate records areavailable for at least 160 caves in the Carlsbadvicinity, including 80 caves in CarlsbadCaverns National Park, 79 caves on BLMlands, and 1 cave on State of New Mexicoland. Additional details of the ChihuahuanDesert species diversity and distribution willbe provided in another publication (Strong, inpreparation).

Sonoran DesertData for the Sonoran Desert are relativelysparse compared with the Chihuahuan Desert,for the reasons given above. However, becauseof the sources of data, there are very few obser-vations that are not identified to species. I havecompiled records with at least 210 reports ofvertebrates, including 31 species documentedwith specific locations. Another 43 species areconsidered confirmed, but without specificlocations, based on Kingsley et al. (2001).Confirmed records are known for 13 caves,including 11 limestone caves and two basaltcaves, and for 52 abandoned mines.

Species DistributionThe data from these two desert regions may beanalyzed is several ways. The simplest andmost direct approach is to compile a list ofspecies for each region, as shown on Table 1.This table is limited to those reports that iden-

tified an organism to species. In this table, thedata are the number of different sites fromwhich a species was reported at least once. Forthe species in the Sonoran Desert that wereconsidered confirmed but not documented inspecific sites, a minimum number of one wasassigned. It is apparent from this list that a fewspecies are reported from many sites and manyspecies are reported from only one or two sites.

The data from Table 1 are summarized inTable 2 to show the numbers of vertebratespecies of each class found in each desertregion, and the numbers shared by the twodeserts. Mammals dominate the vertebratespecies sets in each desert region, and notsurprisingly, bats provide the most species ineach area. Birds and reptiles are reasonablywell-represented in the underground featuresof both desert regions. Amphibians and fish arepoorly represented, but they are relativelyscarce in most other habitats of these deserts.

These data on species distribution in thecaves and mines are presented graphically inFigures 1 and 2 for the Chihuahuan andSonoran Deserts, respectively. The horizontalaxis of this figure represents the number ofvertebrate species that are found in the numberof underground features shown on the verticalaxis. These graphs clearly illustrate that manyspecies are found in very few sites, while onlya few species are widely distributed in manylocations. The graph for the Sonoran Desert isskewed because of the large number of speciesthat are not documented for specific sites andare assigned the minimum value of one site.Additional information would likely changethe shape of the graph and it would retain thesame general pattern as in the ChihuahuanDesert.

Site DiversityAn alternative way to analyze these data is toexamine the number of vertebrate speciesusing each cave or mine. These results areillustrated in Figures 3 and 4 for theChihuahuan and Sonoran Deserts, respectively.Even though data are available for many moresites in the Chihuahuan Desert, the diversity

96 Strong

Table 1. Vertebrate species in underground sites. Data indicate the number of sites in which eachspecies has been documented. Nomenclature follows ITIS 2007.

Chihuahuan SonoranScientific Name Common Name Desert Desert

MammalsNotiosorex crawfordi Desert Shrew 2 1Macrotis californicus California Leaf-nosed Bat 21Choeronycteris mexicana Mexican Long-tongued Bat 6Leptonycteris yerbabuena Lesser Long-nosed Bat 15Myotis velifer Cave Myotis 14 16Myotis thysanodes Fringed Myotis 4Myotis volans interior Long-legged Myotis 2Myotis californicus California Myotis 1 2Myotis leibii Small-footed Myotis 2 1Myotis yumanensis Yuma Myotis 2 1Lasiurus cinereus Hoary Bat 2Lasiurus borealis Eastern Red Bat 2Lasionycteris noctivagans Silver-haired Bat 1Pipistrellus hesperus Western Pipistrelle 3 1Eptesicus fuscus Big Brown Bat 2 2Euderma maculatum Spotted Bat 1Idioinycteris phyllotis Allen’s Big-eared Bat 1Plecotus townsendii Townsend’s Big-eared Bat 16 5Antrozous pallidus Pallid Bat 2 8Tadarida brasiliensis Brazilian Free-tailed Bat 5 4Nyctinimops macrotus Big Free-tailed Bat 1 1Nyctinimops femorosacca Pocketed Free-tailed Bat 1 1Eumops perotis Western Mastiff Bat 1Eumops underwoodi Underwood’s Mastiff Bat 1Lepus californicus Black-tailed Jack Rabbit 4 2Sylvilagus auduboni Desert Cottontail 5 3Ammospermophilus harrisii Harris’ Antelope Squirrel 2Eutamias canipes Gray-footed Chipmunk 1Spermophilus variegatus Rock Squirrel 3 1Thomomys bottae Valley Pocket Gopher 2 1Pappogeomys mexicana Mexican Pocket Gopher 1Pappogeomys castanops Yellow-faced Pocket Gopher 2Chaetodipus intermedius phasma Rock Pocket Mouse 1Chaetodipus baileyi baileyi Bailey’s Pocket Mouse 1Peromyscus eremicus Cactus Mouse 1Peromyscus crinitus Canyon Mouse 1Peromyscus boylii Brush Mouse 1 1Peromyscus leucopus White-footed Mouse 1Peromyscus pectoralis White-ankled Mouse 1

Strong 97

Dipodomys spectabilis Banner-tailed Kangaroo Rat 1Dipodomys merriami Merriam’s Kangaroo Rat 3 3Dipodomys deserti Desert Kangaroo Rat 1Neotoma albigula White-throated Wood Rat 1 3Neotoma lepida Desert Wood Rat 1Neotoma mexicana Mexican Wood Rat 2Sigmodon hispudus Hispid Cotton Rat 1Erethizon dorsatum Porcupine 37 1Canis latrans Coyote 4Urocyon cinereoargenteus Gray Fox 2 1Vulpes velox Swift Fox 1Vulpes microtus Kit Fox 2Ursus americanus Black Bear 2Procyon lotor Raccoon 6 1Nasua nasua Coati 1Bassariscus astutus Ringtail 39 4Mustela frenata Long-tailed Weasel 1Taxidea taxus Badger 1Conepatus leuconotus Common Hognosed Skunk 2 1Spilogale gracilis Western Spotted Skunk 1 1Mephitis mephitis Striped Skunk 3 1Puma concolor Mountain Lion 18 1Lynx rufus Bobcat 2 1Peccari tajacu Collared Peccary 1 3Odocoileus hemionus Mule Deer 15 1Ovis canadensis Desert Bighorn Sheep 2 1Ammotragus lervia Barbary Sheep 12Capra hircus Domestic Goat 15BirdsCoragyps atratus Black Vulture 1Cathartus aura Turkey Vulture 5 3Buteo regalis Ferruginous Hawk 1Tyto alba Barn Owl 5 2Megascops kennicottii Western Screech-Owl 1Bubo virginianus Great Horned Owl 20 2Phalaenoptilus nuttallii Common Poorwill 1Zenaida macroura Mourning Dove 1Aeronautes saxatalis White-throated Swift 3 1

Hummingbird sp. 1Colaptes auratus Northern Flicker 5Sayornis saya Say’s Phoebe 5 2

Table 1. continuedChihuahuan Sonoran

Scientific Name Common Name Desert Desert

98 Strong

Myiarchus cinerascens Ash-throated Flycatcher 1Petrochelidon fulva Cave Swallow 19Petrochelidon pyrrhonota Cliff Swallow 1Tachycineta thalassina Violet-green Swallow 1Salpinctes obsoletus Rock Wren 10 1Catherpes mexicanus Canyon Wren 8 1Junco hyemalis Dark-eyed Junco 1Spizella atrogularis Black-throated Sparrow 1ReptilesTerrapene ornata Ornate Box Turtle 1Chrysemys picta Painted Turtle 1Gopherus agassizii Desert Tortoise 1Crotophytus collaris Collared Lizard 2Sceloporus magister Desert Spiny Lizard 1Sceloporus clarkii Clark’s Spiny Lizard 1Sceloporus undulatus Eastern Fence Lizard 1 1Uta stansburiana Common Side-blotched Lizard 1Urosaurus ornatus Ornate Tree Lizard 1Heloderma suspectum Gila Monster 1Pituophis catenifer Gopher Snake 2Salvadora grahamiae Mountain Patch-nosed Snake 1Thamnophis sp. Garter Snake 1Tantilla sp. Black-headed Snake 1Crotalus atrox Western Diamondback Rattlesnake 8 1Crotalus lepidus Mottled Rock Rattlesnake 7Crotalus lepidus Banded Rock Rattlesnake 1Crotalus mitchelii Speckled Rattlesnake 1Crotalus molossus Black-tailed Rattlesnake 3 2Crotalus tigris Tiger Rattlesnake 1Crotalus scutulatus Mojave Rattlesnake 1AmphibiansAmbystoma mavortium Tiger Salamander 6 1Anaxyrus punctatus Red-spotted Toad 1 1Lithobates yavapaiensis Lowland Leopard Frog 1FishFundulus zebrinus Plains Killifish 1

Table 1. continuedChihuahuan Sonoran

Scientific Name Common Name Desert Desert

Strong 99

within sites has a similar pattern for bothdeserts, with many sites having only a fewspecies and a few sites having many species.The site with the greatest number of species inthe Chihuahuan Desert is Carlsbad Cavern, alarge cave which has been studied intensivelyfor several decades by many researchers. Incontrast, the site with the greatest number ofspecies in the Sonoran Desert is a small basaltcave with all species recorded in a single visit.

DISCUSSIONConfirmed Resource Uses

While the diversity and distribution of verte-brates using underground features is of interest,from an ecological point of view the reasonswhy vertebrates seek out these resources iseven more interesting. Data on the uses ofcaves or mines are generally more availablefor the Chihuahuan Desert, but some data areavailable for the Sonoran Desert. The simplestuse of a cave or mine would be as a temporaryroost site. Several species of birds have beenobserved roosting in caves, particularly owlsusing caves as daytime roost sites (T. Strong,pers. obs.). Similarly, many species of bats usecaves or mines as temporary nighttime roostsbetween foraging bouts.

Caves and abandoned mines provide goodsites for nests or dens for many species of birdsand mammals. Wood rat (Neotoma sp.) nestsare found in many sites in both the Chihuahuanand Sonoran deserts, but the species building

the nests cannot be identified without visualconfirmation (Novack 2004; Allison 2004).Bailey (1928) reported that mountain lionswere using caves as den sites in the CarlsbadCaverns vicinity, and lions have been encoun-tered in other caves in this region (Parent 1998;Allison and Roemer 1998; J. Goodbar, pers.comm.). Piles of small mammal bones mayindicate the presence of carnivore den sites (T.Strong, pers. obs.). Porcupine den sites havebeen noted in numerous caves in theChihuahuan Desert (Fleming and Hummel1977b; Hummel 1977; Belski 1979), and liveporcupines have been encountered (Pate 1992;Fleming 1977; T. Strong, pers. obs.). Birds alsouse caves and mines as nest sites, withconfirmed nesting for at least nine species incaves of the Chihuahuan Desert and at leastfour species in the Sonoran Desert (Bailey1928; Belski 1989; Fleming and Hummel1977a; Lindsley 1967; Pate et al. 1995;Spangle and Thompson 1958; Corman andWise-Gervais 2005; T. Strong, pers. obs.). Thecave swallow (Petrochelidon fulva) is the mostcommon nesting bird in Chihuahuan Desertcaves, with nesting confirmed in at least 15caves.

Caves and mines are also important asmaternity sites for many bat species. TheBrazilian free-tailed bat (Tadarida brasiliensis)is well-known for its maternity colony inCarlsbad Caverns. Maternity colonies for cavemyotis (Myotis velifer) and fringed myotis

Table 2. Summary of species numbers in each desert by class.

Chihuahuan Sonoran Species in Vertebrate Class Desert Desert Both Deserts

Mammals 51 47 32Birds 16 11 7Reptiles 11 13 4Amphibians 2 3 2Fish 1 0 0Total 81 74 45

100 Strong

Figure 1. Distribution of vertebrate species found in caves of the Chihuahuan Desert nearCarlsbad, New Mexico.

(M. thysanodes) (Ek 1990; Pate 1994) havealso been reported for the Carlsbad Cavernvicinity. The only reported maternity sites forthe endangered lesser long-nosed bat(Leptonycteris yerbabuenae) are in caves andmines of the Sonoran Desert (Cockrum andPetryszyn 1991).

Many vertebrate species are known to usecaves or abandoned mines as hibernation sites.In particular, bats have been observed hiber-nating in several caves of the ChihuahuanDesert (Bailey 1928; Belski 1988; Baker et al.,no date; Kerbo 1978; Ek 1991). It is likely thatseveral bats hibernate in Sonoran Desert cavesand mines, but this behavior has not beenspecifically reported. The poorwill is the only

known bird confirmed to hibernate, and itcould use caves or mines in these deserts ashibernation sites. It has been reported hiber-nating (not in a cave) at Carlsbad CavernsNational Park (S. West, pers. comm.), and ithas been observed in a crevice in a pit entranceto a cave in the Park (P. Seiser, pers. comm.).It is likely that several reptiles and amphibiansuse caves as hibernation sites, but there are nodocumented observations in the caves or minesof either desert.

Probable Resource UsesUnderground features in arid regions are likelyto provide water sources for a variety ofanimals. Caves and mines may provide points

Strong 101

of access to groundwater or small perchedaquifers in areas where surface water is rare.Mule deer and bighorn sheep have beenreported to obtain water from pools in caveswith large entrances in Slaughter Canyon in233 National Park (Bailey 1928; Welbourn1978). It seems almost certain that other verte-brate species are using these water sources inboth desert regions, but if it has been observed,it has not been reported.

Some species are apparently using cavesas foraging sites. Bailey (1928) reported thatwhite-footed mice were common throughoutCarlsbad Cavern and were feeding on cricketsand food dropped by tourists. Ringtails arelikewise found in deep areas of CarlsbadCavern (Bailey 1928; D. Pate, pers. comm.),and it seems likely that they are feeding onmice. Bailey (1928) also suggested thatmountain lions were using a cave with a largeentrance as a hunting site. Snakes are also

likely to use caves as foraging sites. A para-lyzed mouse seen in the entrance of a cave on233 National Park had probably been bittenby a rattlesnake that was seen nearby(Reames and Barber 2003). A live mountainpatch-nosed snake (Salvadora grahamiae) atthe bottom of a 13 meter pit was probablyforaging on mice that have been seen at thebottom of the pit (T. Strong, pers. obs.). Inanother cave, the presence of a pile of caveswallow feathers on a ledge with ringtail scatsuggests the possibility of predation (T.Strong, pers. obs.).

Another probable use of caves or mines isto obtain relief from extreme environmentalconditions. In desert regions with very highsummer temperatures and very low humidity,caves and mines could provide sites with lowertemperatures and much higher relative humidi-ties. It seems likely that many vertebratespecies would deliberately select these favor-

Figure 2. Distribution of vertebrate species found in caves and mines of the SonoranDesert of Arizona.

able microclimates, although choice might bevery difficult to demonstrate.

Incidental or Unintentional UsesNumerous species of mammals have beenidentified through the presence of tracks orscat, indicating use of the cave. Birds andreptiles may also leave evidence of this type incaves. These observations could fall into thecategory of incidental use. While evidence ofthis sort confirms that an animal was alivewhile in the cave or mine, it cannot be inter-preted to explain why the animal was using thesite.

The presence of skeletal material in cavesand mines may mean that the animals enteredthe cave deliberately, but they may have beenunable to find their way back out. Animalswhose remains are found at the bottom of a pitmay have entered the site deliberately, butprobably did not intend to fall into a pit. Someof these animals may have been killed by the

fall, or they died because they were unable toclimb out. Some vertebrates, particularlyreptiles, appear to be able to survive relativelylong drops. As noted above, a mountain patch-nosed snake apparently survived a 13 m fallinto an overhanging pit in a Chihuahuan Desertcave (T. Strong, pers. obs.).

Skeletal material found in some caves mayalso suggest that these animals (or partsthereof) had been carried into the caves as preyitems. For example, jackrabbit and cottontailbones were found in a cave with a short climbdown into the entrance of this cave whichwould make it very difficult for these animalsto get out of the cave. However, carnivorescould easily climb in and out, and ringtail andbobcat tracks were common in the cave,suggesting that these carnivores could havecarried the rabbits into the cave (T. Strong,pers. obs.). Deer legs found in a cave with alarge entrance were probably brought into thecave by a large predator (Carrington 1999),

102 Strong

Figure 3. Diversity of vertebrate species in caves of the Chihuahuan Desert nearCarlsbad, New Mexico.

Strong 103

and mountain lions are known to use this cave(Roemer 2000). In some caves in both deserts,quantities of rodent bones are probably relatedto the use of these sites by great horned owlsand barn owls, which eject pellets containingmany small bones (T. Strong, pers. obs.).

ComparisonsSome of the observed differences in speciesdistributions may be explained by a variety ofdifferences between the underground featuresin the Sonoran and Chihuahuan Deserts. Thefollowing comments are generalizations, andsome exceptions are likely for each of thesecomparisons. Geologically, the Carlsbadregion of the Chihuahuan Desert is underlainby soluble rocks, primarily limestone andgypsum. The Carlsbad Caverns National Parkis dominated by limestone formations in theGuadalupe Mountains, including Permian reef,fore-reef, and back-reef units. BLM landsnorth, east, and south of Carlsbad include

extensive areas of gypsum karst. These solublerocks allow the formation of large and smallcaves. Other rock types are rare in this vicinity,and none of the vertebrate records in this partof the Chihuahuan Desert are from mines.Topographically, the gypsum areas are on rela-tively flat-lying plains east and southeast of theGuadalupe Mountains. The mountain rangerises gradually from the eastern plains and thendrops abruptly to the south and west, and it iscut by numerous deep canyons.

In the Sonoran Desert of Arizona, limestoneand gypsum are relatively rare, and only a fewof the recorded underground sites are lime-stone or gypsum caves. This region is in theBasin and Range topographic province, withnumerous mountain ranges separated by broadalluvial valleys. Some extrusive volcanics arepresent with vertebrates in two small caves ofa basalt flow. Metamorphic rocks are commonin the mountain ranges, and hydrothermalintrusions have created many deposits of

Figure 4. Diversity of vertebrate species in caves and mines of the Sonoran Desert ofArizona.

104 Strong

copper and other metals. Early mining activityin these mineral deposits created large numbersof now-abandoned mine adits and shafts, manyof which are now used by vertebrates. Most ofthe underground records of vertebrates in theSonoran Desert are from abandoned mines.Topographically, most sites in the SonoranDesert are located in the lower parts of moun-tain ranges where bedrock is exposed. Thesesites also tend to be at slightly lower elevationsthan the sites in the Chihuahuan Desert,leading to slightly higher average undergroundtemperatures.

The general morphology of the under-ground features also varies between desertregions. Many of the limestone caves in theGuadalupe Mountains have large entrancesthat are visible from long distances. Theselarge entrances provide easy access for smalland large vertebrates. Large entrances alsoallow light to penetrate far into a cave, makinga larger area suitable for animals that arelimited to twilight zones of caves. Most of thegypsum caves in the Carlsbad vicinity havesmaller entrances in sinkholes in the plains.These smaller entrances would limit the extentof light penetration and could preclude theiruse by larger animals.

The limestone and basalt caves in theSonoran Desert have relatively smallentrances, with limited light penetration andlimited accessibility for larger animals. Theabandoned mine entrances tend to be roughlythe same size, being high enough and wideenough to allow relatively easy human entry,which also allows easy access for mostanimals. Light penetration in these featureswill be controlled to some extent by orienta-tion.

The availability of water within the under-ground features also differs between deserts,and even among caves or mines within eitherthe Sonoran or Chihuahuan Desert. In theChihuahuan Desert, the gypsum caves on theplains east of the mountains are generallyconduits for ephemeral floodwaters. Anysurface water from rainfall runoff is capturedin sinkholes and transported through the caves

toward the Pecos River. In contrast, the lime-stone caves often have dripping water or poolsof water, but flowing streams are rare. In theSonoran Desert, the basalt caves are very dry,and the limestone caves are generally dry,although some may have some dripping wateror small pools. The abandoned mines rangefrom very dry to moderately wet, but there areno underground flow channels in the caves andmines of the Sonoran Desert. However, evenfeatures that have no free water are likely tohave relative humidity conditions that arehigher than the surrounding ambient condi-tions. The differing availability of water willinfluence the variety of vertebrate species thatare able to utilize specific sites.

Another obvious difference between theSonoran and Chihuahuan Deserts is the poolof species present in these two regions. Thespecies found in underground features will berestricted to the available set of species in thevicinity of the underground sites. As can beseen in Table 1, some species are common tocaves or mines in both deserts, but many otherspecies are found in only one region or theother.

In summary, there are wide contrastsbetween the physical conditions of the under-ground features in the Sonoran andChihuahuan Deserts. In addition, the data setsfor each region are based on diverse sources,with non-uniform collection or observationprocedures. In spite of these drawbacks, thepatterns of species distribution among cavesand mines and the patterns of species diversitywithin these features are similar in the twodeserts. It is anticipated that further, system-atic collection of data on vertebrates in thesefeatures will increase our list of species usingthese features as well as the numbers of sitesfor each species and will increase the record ofthe number of species within each site.

CONCLUSIONSThe results of this study demonstrate thatunderground features of both theChihuahuan Desert and the Sonoran Desert arebeing used regularly by a wide variety of verte-

Strong 105

brates. This level of usage and the documentedtypes of usage by species demonstrates thatthese underground sites provide a habitatfeature that is an important resource. In an aridenvironment with extremes of high tempera-tures and low relative humidity, caves could becritical to the survival of many vertebrates.Although few of the species observed in thesesites are listed as threatened or endangered,their continued presence in the Chihuahuan orSonoran Desert may depend on these under-ground resources.

With so many species depending on theunderground features of these deserts, it isimperative that federal and local land manage-ment agencies maintain policies that provideprotection for cave and mine resources. Mostcaves on Carlsbad Caverns National Park areadministratively closed, although three cavesare open for commercial tours (including toursin undeveloped areas) and eight others areopen for recreational caving. As noted above,Carlsbad Cavern has a high diversity of speciesin spite of the heavy annual visitation. TheBureau of Land Management maintains apermit system for several of its caves, andthere are some seasonal restrictions on visita-tion because of bats. However, many BLMcaves are open for recreational caving with norestrictions. Many of the abandoned mines inthe Sonoran Desert are controlled by largemining corporations, although they may belocated on federal land. However, because ofrelatively easy public access, many of thesesites are in danger of being closed for publicsafety and liability reasons.

Under the National Environmental PolicyAct (NEPA), federal agencies are required toanalyze potential environmental impacts priorto taking any action. Based on the evidence ofvertebrate use of these caves and mines, poten-tial impacts on these wildlife species and theirhabitat requirements must be considered. Inaddition, whether these agencies are givingpermits for recreational caving or for scientificresearch, they should be encouraged to providethe permitees with information about wildlifespecies using the caves and any precautions

they should take when visiting the caves.When mines are to be closed for public safety,surveys should determine the use or potentialfor use by vertebrates, primarily bats, andspecially-designed gates could provide accessfor bats and other vertebrates while preventinghuman access.

LITERATURE CITEDAllison, S. 2004. Cave C-108 trip report, 4/18/04.

Unpublished data in Carlsbad Caverns NationalPark files.

Allison, S. and G. Roemer. 1998. Cave C-25 tripreport, 9/7/98. Unpublished data in CarlsbadCaverns National Park files.

Bailey, V. 1928. Animal life of the Carlsbad Cavern.Monographs of the American Society ofMammalogists, No. 3. Williams and Wilkins Co.,Baltimore, MD.

Baker, J. K., B. Crisman, and D. Geise. No date.Report on Cave C-18, early 1950s. Unpublisheddata in Carlsbad Caverns National Park files.

Belski, D. 1979. Trip report on Porcupine Shit Pit.Unpublished data in BLM files, Carlsbad FieldOffice.

Belski, D. 1988. Trip report on Rocking Chair Cave,GYPKAP report. Unpublished data in BLM files,Carlsbad Field Office.

Belski, D. 1989. Trip report on Chalk and SkyliteCaves. Unpublished data in BLM files, CarlsbadField Office.

Biotic Information System of New Mexico(BISON). 2004. Species Account. Internet site:http://fwie.fw.vt.edu/states/nm.htm. Accessedmultiple dates, 2004.

Carrington, P. 1999. Goat Cave trip report. Unpub-lished data in Carlsbad Caverns National Parkfiles.

Cockrum, E. L. and Y. Petryszyn. 1991. The long-nosed bat, Leptonycteris: An endangered speciesin the Southwest? Occasional Papers. MuseumTexas Tech University. 142: 1–32.

Corman, T. E. and C. Wise-Gervais. 2005. Arizonabreeding bird atlas. University of New MexicoPress, Albuquerque, NM.

Ek, D. 1990. Memo on Lake Cave, 28 July 1990.Unpublished data in Carlsbad Caverns NationalPark files.

Ek, D. 1991. Memo on Goat Cave. Unpublisheddata in Carlsbad Caverns National Park files.

Fleming, S. 1977. Trip Report on Root Cave.Unpublished data in BLM files, Carlsbad FieldOffice.

Fleming, S. and B. Hummel. 1977a. Cave descrip-tion of Boyd’s Cave. Unpublished data in BLMfiles, Carlsbad Field Office.

Fleming, S. and B. Hummel. 1977b. Cave descrip-tion of Honest Injun Cave. Unpublished data in

106 Strong

BLM files, Carlsbad Field Office.Hoffmeister, D. F. 1986. Mammals of Arizona.

University of Arizona Press, Tucson, AZ.Hummel, B. 1977. Cave description of Klutz

Caverns. Unpublished data in BLM files, CarlsbadField Office.

Integrated Taxonomic Information System (ITIS).2007. Taxonomic information database, availableat http://www.itis.gov. Accessed multiple dates.

Kerbo, R. 1978. Deep Cave trip report, 27 January1978. Unpublished data in Carlsbad CavernsNational Park files.

Kingsley, K. J., T. R. Strong, E. L. Smith, and T. K.Snow. 2002. Caves and mine adits as wildliferesources in the Sonoran Desert region. InProceedings of the 2001 national cave and karstmanagement symposium, 16–19 October 2001,Tucson, AZ, edited by G. T. Rea, pp. 138–140.U.S. Forest Service, Coronado National Forest.

Lindsley, P. 1967. Lake Cave trip report, 5 July1967. Unpublished data in Carlsbad CavernsNational Park files.

NatureServe. 2004. NatureServe Explorer, an onlineencyclopedia of life. Internet site: http://www.natureserve.org/explorer/. Accessed multipledates, 2004.

Novack, A. 2004. Cave C-72 trip report, 13 April2004. Unpublished data in Carlsbad CavernsNational Park files.

Parent, L. 1998. Cornering a lion! Canyons andCaves No. 11, Winter 1998.

Pate, D. 1992. Trip report on Adobe Cave, 31 May1992. Unpublished data in BLM files, CarlsbadField Office.

Pate, D. 1994. Lake Cave trip report, 12 August1994. Unpublished data in Carlsbad CavernsNational Park files.

Pate, D., H. Burgess, and J. Richards. 1995. CaveC-60 trip report, 7 June 1995. Unpublished datain Carlsbad Caverns National Park files.

Reames, S. and R. Barber. 2003. Lechuguilla Cavetrip report, 13 October 2003. Unpublished data inCarlsbad Caverns National Park files.

Roemer, D. 2000. Mountain lion observationsreport. Unpublished data in Carlsbad CavernsNational Park files.

Spangle, P. and M. Thompson. 1958. Report onCave C-20, 18 May 1959. Unpublished data inCarlsbad Caverns National Park files.

Welbourn, W. C. 1978. Biology of Ogle Cave witha list of the cave fauna of Slaughter Canyon. NSSBulletin 40(1): 27–34.

Status of Resources

IMPLICATIONS OF ILLEGAL BORDER CROSSING AND DRUG TRAFFICKING ONTHE MANAGEMENT OF PUBLIC LANDS

Craig C. Billington, Randy Gimblett, and Paul R. Krausman

Increasing numbers of illegal immigrants arecrossing rugged parts of the 3,122 km U.S.-Mexico border that traverses the SonoranDesert. With increases in border-crossing comelarger numbers and groups captured by BorderPatrol and other agencies on state, federal andtribal lands. Increases in border security inurban portions of the border over the pastseveral years have caused entrants and smug-glers to attempt more difficult crossings,traversing public and tribal lands in SouthernArizona.

Between 1977 and 1987 the number ofundocumented aliens (UDA) apprehended bythe U.S. Border Patrol in the U.S. doubledfrom 1 to 2 million (Weber 1988). The reportedincreases in apprehensions have continued andrepresent increased enforcement, and increasedUDA activity (Andreas 2000; Nevins 2002).The U.S. Immigration and NaturalizationService (INS) estimated that for every UDAapprehended two to three more cross theborder undetected (Weber 1988). This does notinclude human traffic related to narcoticssmuggling. Since immigration reform in 1996and the increasing enforcement under Opera-tion Gatekeeper (i.e., an increased effort andconcentration of federal agents to tightenborder security along U.S.-Mexico bordercities, e.g., El Paso, Texas; San Diego, Cali-fornia), UDA and narcotics traffic shifted toremote areas (Cohen 1997; Andreas 2000;Nevins 2002; McIntyre and Weeks 2002;Jacoby 2004). Border Patrol activity has alsoincreased in wildlands, which has caused more

vehicle traffic, new roads, off-road vehicle(ORV) traffic, and aerial surveys associatedwith patrolling. The cumulative effect of thisincreased activity and the natural resourcedegradation is unknown.

Illegal activities in Arizona are indicativeof the severity of the problem throughout thesouthwestern U.S. In 1999, the INS appre-hended 470,499 UDA in Arizona, double thenumber apprehended in 1995 (McIntyre andWeeks 2002). In the Tucson sector, apprehen-sions by the U.S. Border Patrol related to ille-gal immigration increased from 333,648 in2002 to 491,771 in 2004. Marijuana seized bythe U.S. Border Patrol in the Tucson sector in-creased 31.6% between 2002 and 2004. Thisillegal activity contributes to the overall in-crease in human activity leading to natural re-source degradation in remote areas of thedesert near the Mexican border (McIntyre andWeeks 2002; Jacoby 2004; Tobin 2004). Areasthat have been set aside as reserves — i.e.,Organ Pipe Cactus National Monument(OPCNM), Cabeza Prieta National WildlifeRefuge (CPNWR), Buenos Aires NationalWildlife Refuge (BANWR), and IronwoodForest National Monument (IFNM) — haveexperienced increases in human activity (T.Tibbitts, biologist, OPCNM; C. McCasland,biologist, CPNWR, pers. comm.). Natural re-source degradation due to humans occurs inall these areas, but remains largely undocu-mented (T. Tibbitts; C. McCasland, pers.comm.). Furthermore, the numbers of migrantdeaths has climbed steadily since 1994 when

Operation Gatekeeper imposed strict enforce-ment on crossings in the urban areas sur-rounding Tijuana, San Diego, Ciudad Juarezand El Paso. In 1998, 28 people died duringborder crossings. In 2005, 7,200 were docu-mented as deceased. This dramatic increase re-flects the growing problem along theU.S.-Mexico border.

There have been corresponding increasesin Border Patrol and other law enforcementactivities in the Tucson sector. In fact, an initia-tive called the Arizona Border Control (ABC)has increased the number of Border Patrolagents in the Tucson sector (from 200 togreater than 2,000 agents), the number ofaircraft, and unmanned aerial vehicles, and thedetention center mentioned above (Departmentof Homeland Security 2004). Law enforce-ment agencies have been trying to decrease thenumber of migrant deaths by imposing strictenforcement on smugglers putting their humancargo in danger, using a search and rescuegroup, and avoiding apprehending immigrantsnear water stations. Humane Borders, aTucson-based non-profit, has established waterstations along the U.S.-Mexico border toimprove migrant safety. Some view thesewater stations as encouraging illegal immigra-tion, which may lead to vandalism. Stations inIFNM have been destroyed (Vanderpool 2003;Marizco 2004).

The vast stretches of unpopulated border-lands in Southern Arizona are a major corridorof human and drug smuggling. The TohonoO’odham Nation law enforcement officersroutinely apprehend drug smugglers, nettingbetween 2,000 and 6,000 pounds of cocaineand marijuana each month that have crossedtheir 75-mile stretch of border with Mexico(Hinkle 2001). These figures do not accountfor drugs confiscated by the Border Patrol orother law enforcement agencies operating inthe area. More recently, the Border Patrol hasbeen confiscating at least 5,000 pounds ofdrugs per month in the Tucson sector.

IMPACTS IN THE IFNMPublic lands in Southern Arizona, including theIFNM, have become a crossroads of illegal

immigrant and drug smuggling, law enforce-ment activities, ranchers and recreational users.The IFNM, which was established by a Presi-dential Proclamation in 2000 (Clinton 2000),lies between the Tohono O’odham Nation tothe south and west, and Interstate 10 to thenortheast (Figure 1). The area contains manyroutes between the international border sharedby the Tohono O’odham Nation, Mexico andInterstate Highway 10. Illegal immigrants anddrug smugglers impact the natural environ-ment of IFNM, leaving trash and creating newdouble-track, or “wildcat,” roads. These illegalactivities also impact the broader landscape ofIFNM and its surroundings. The smugglingtraffic in IFNM is a threat to many of thenatural features unique to the area. Biologistsfrom the Arizona Sonoran Desert Museum(ASDM) documented the impacts of smug-gling traffic in the area. They noted impactsconsisting of trash at camps, pickup sites, andtravel routes within the monument (ASDMwebsite). The ranchers living on the privateproperty in and around IFNM are concernedabout the dangers posed by drug smugglersand UDAs (Turf 2002).

Trash and illegal activity affects the wholecommunity of IFNM users. Law enforcementand undocumented immigrants have been citedas major issues that have received considerablepublic comment during the scoping process(Bureau of Land Management 2004). In fact,38.3% of the letters received in this processspecifically cited these concerns. Theseconcerns have also been voiced in the tripreports filled out by recreation users and recip-ients of hunting permits as part of therecreation study conducted by the Universityof Arizona (Gimblett 2004). In this study, over50% of respondents ranked garbage dumpingand illegal immigrant activity as the mostserious problems facing the management of themonument.

The conflicts between drug smuggling,illegal immigration, illegal dumping, recre-ation use and ranching all pose challenges toIFNM’s long-term planning and managementstrategies. The purpose of this paper is todescribe a study that was intent on under-

110 Billington, Gimblett, and Krausman

Billington, Gimblett, and Krausman 111

Figure 1. Location and land ownership for Ironwood Forest National Monument, Arizona,2004.

standing the spatial distribution of human useacross the monument and documenting theamount of nighttime use that was occurring.

STUDY AREAThe IFNM was established on 9 June 2000 bypresidential order and is managed by theBureau of Land Management (BLM). The

IFNM encompasses 76,781 ha in the SonoranDesert and includes 52,232 ha (68%) offederal, 22,135 ha (29%) State of Arizona, and2,433 ha (3%) of private land (Figure 1). It islocated within Pima and Pinal counties, and iseasily accessible from the Tucson (50 km) andPhoenix (100 km) metropolitan areas frompaved and dirt roads connecting with Interstate

Highway 10 (Figure 1). The monument wasestablished to protect natural and culturalresources and is one of the most biologicallydiverse areas within the Sonoran Desert with> 670 species of flora and fauna (Tersey et al.2001). The area contains federally listed threat-ened and endangered species — the lesserlong-nosed bat (Leptonycteris curasoaeyerbabuenae) and cactus ferruginous pygmy-owl (Glaucidium brasilianum). Other wildlifein the area include: cougars (Puma concolor),mule deer (Odocoileus hemionus), mountainsheep (Ovis canadensis), collared peccary(Pecari tajacu), coyotes (Canis latrans),bobcats (Lynx rufus), desert tortoise (Gopherusagassizii), and a variety of small mammals,migratory and non-migratory birds, reptiles,and amphibians.

The IFNM is made up of a series of inter-connected mountain ranges separated andsurrounded by valleys of creosote bush (Larreatridentata). These mountain ranges extendingsouth to north include the Roskruge,Waterman, Silver Bell, West Silver Bell,Samaniego Hills, and Sawtooth Mountains(Figure 1). The IFNM is bordered by theTohono O’odham Nation to the south andwest, cotton fields on the Santa Cruz floodplain on the north, and private lands to the east.It extends 48 km from south to north and 50km from west to east.

Four main vegetation associations occurwithin IFNM. Lowland plains are dominatedby creosote bush and bursage (Ambrosiadeltoidea). Hillsides are made up of palo verde(Parkinsonia microphylum), ocotillo(Fouquieria splendens), and mixed cacti —e.g., cholla (Opuntia spp.), saguaro (Carnegieagigantea), and prickly pear (Opuntia). Desertriparian communities exist along washes thatinclude palo verde, mesquite (Prosopisvelutina), ironwood (Olneya tesota), andacacia (Acacia spp.). A chaparral communityof jojoba (Simmondsia chinensis) and mixedscrub occurs at higher elevations (Bristow etal. 1996).

Elevation ranges from 580 m on the valleyfloor to 1,290 m at Silver Bell Peak. Rainfallis bimodal with rainfall primarily in winter

(December–February) and late summer (July–September). Since 1956 the average annualprecipitation has been 31.8 cm. The daytimehigh temperatures range from 15 to 20° C inthe winter to > 40° C during summer (NationalOceanic and Atmospheric Administration offi-cial website; http://www.noaa. com/). The yearwas divided into 5 seasons: winter(December–January), spring (February–April),dry summer (May–June), wet summer(July–September), and autumn (October–November) based on temperature and precip-itation (Sellers and Hill 1974). Humanactivities that take place in and around IFNMinclude mining, ranching, livestock grazing,hunting, camping, hiking, off-highway vehicleuse, target shooting, research and conservation,and illegal immigrant and drug traffic (Bristowet al. 1996; Tersey et al. 2001).

METHODSWe evaluated human activities in IFNM from1 January 2002 to 13 July 2004 generallyfollowing the methods of Gimblett (2004) andTitre et al. (2004). We used sensors and surveydata to determine human activity patterns andprovide estimates of visitor use. Sensorsprovide data on level of use in specific areas,entry and exit points, time of use, and routestraveled. Surveys provided data on group size,entry and exit points, destination, time of use,and activity type. We downloaded data fromdata loggers connected to the sensors. Wedesigned specific Microsoft Access queries toassist in data analysis. We constructed queriesfor each individual sensor and all sensorscombined to evaluate human traffic on anhourly, weekly, monthly and annual basis.

We used a global positioning system (GPS)receiver (Trimble Navigation, Lafayette, LA)to map all of the roads in IFNM. Twenty-sixtraffic/trail sensor pads (Scott Technical Instru-ments Ltd., Hamilton, New Zealand) wereplaced on routes to monitor vehicle traffic for8–30 months (Figure 2). Sensors were placedat a sampling of access points throughoutIFNM. Sensor pads were concealed 0.3 munder the ground. A HOBO® data logger(Onset Computer Corporation, Pocasset, MD)



was attached by a cable buried 2–3 m from theroute that recorded date and time of event thateach vehicle or human that moved over them.Four vibration sensors (Scott Technical Instru-ments Ltd.; Hamilton, New Zealand) were

concealed inside cattle guards or gates onroads.

In addition to the counter pads, we placednine survey stations throughout IFNMbetween September 2002 and March 2004, in

Figure 2. Trail sensor and survey station locations in Ironwood Forest National Monument,Arizona, 1 January 2002–13 July 2004.


areas where people were observed to stop (e.g.,recreation sites, gates, intersections) (Figure2). Survey stations provided a questionnairewith a map and instructions for visitors to fillout describing their use of IFNM (Billington2005). Respondents recorded their trips on themaps and answered questions that related toissues of concern and encountered in IFNMand assigned ratings 1 (not serious) to 4 (mostserious) to indicate the degree of the problemencountered. Other questions included dateand time of visit. The survey was self-admin-istered with forced choice questions/answersrelated to management issues and a map itin-erary. The survey was intended to inform thevisitor about the IFNM and capture moredetailed information on the distribution of visi-tation in the monument.

Trail sensors and survey stations werechecked twice a month and data were down-loaded into a laptop computer. Survey stationswere stocked each time completed surveyswere collected. We spent 2 nights (1800–0600hr) during 4 seasons (i.e., spring, dry summer,wet summer, and autumn) to observe who wasdriving at night by sitting near the side of a dirtroad that had high night use (based on previoustraffic pad data and personal communicationwith Border Patrol and BLM). All humanactivity was recorded that passed by the obser-vation point. The research team was positionedwithin a safe distance of the observation point,but had a clear vision of visitors. It wasassumed that a vehicle was involved in UDAtransportation or illegal activity if it wasdriving with its lights off, had > 5 occupants,was an older vehicle in relatively poor condi-tion, and did not have any government or lawenforcement markings and license plates.Based on these criteria vehicles were classifiedeither as illegal, government, or other.

We compared night (1800–0600) and day(0600–1800) data using a simple 2-tailed t-test.We modeled seasonal data using analysis ofvariance (ANOVA) to compare levels ofhuman activity among seasons. We modeledannual, seasonal, and night and day eventstogether using MANOVA with all possibleinteractions considered. We used linear regres-

sion to model and compare the overall trendsin nighttime use and daytime use over time.Statistical analysis was all performed usingJMP IN version 5 (Sall et al. 2005). Confidenceintervals and correction factors for sensorcounts were calculated and performed usingthe procedures established by Titre et al.(2004).

RESULTSOur data were normally distributed. Twelve tofifteen thousand visitors (i.e., recreation visi-tors, local residents, law enforcement, UDA)travel through IFNM annually based on sensorand survey data (Gimblett 2004). Because wedid not monitor every entrance our estimatesrepresent a minimum. The estimated numberof annual visitors includes residents who liveand work within the IFNM boundary.

The number of recorded events (i.e., vehi-cles recorded by a sensor, 1 event = 1 vehicledriving over a sensor) varied among sensors(Figure 3). Manville Gate had the greatestnumber of recorded events (n = 3,075). Sasco2 had the fewest recorded events (n = 128).When factoring in the number of days in oper-ation, Pipeline was the most active (n = 15.53events/day, 95% CI = 11.54–19.53) and RedHill 3 was the least active (n = 0.45 events/day,95% CI = 0.33–0.57). The number of recordedevents on the 23 sensors from 1 January 2002to 13 July 2004 was 19,369 during 9,893sensor days. Events/day on all 23 sensorscombined was 76.92 (95% CI = 57.27– 96.87).Visitor use followed a general loop followingmain roads through the central section ofIFNM. Primary destinations for recreationusers included trails and campsites surroundingRagged Top Mountain. Peak hours of usecompared between sensor data and survey data(Billington 2005) differed. Peak hours of recre-ation use were between 0700 and 1500 withlittle nighttime use. Sensor data showed morenighttime use (31.2%) than the survey respon-dents (10.9%).

Survey respondents rated trash dumpingand illegal immigration as the two greatestproblems in IFNM (Table 1). About 36% ofrespondents who reported these as problem


situations consider them to be most serious(scored 4). Garbage dumping (x = 2.56, SD =1.24, n = 88) and illegal immigrant activity (x= 2.34, SD = 1.38, n = 89) were consideredproblematic by 83 and 84% of the respondents,respectively.

Nighttime activity was common (31.2%) inIFNM (Figure 4). Sensors in the central sectionof IFNM near Ragged Top Mountain had the

greatest proportion of daytime events. Sensorslocated in the southern section of IFNM nearthe border of the Tohono O’odham Nation andextending northeast towards Interstate 10 hadthe greatest proportion of nighttime events(Figure 4). Eight sensors that formed a looparound Silver Bell and Ragged Top Mountains(i.e., Silver Peak, Ragged South, Pad 4, RedHill 3, Ragged Top, Ragged Rear, Silver 2, and

Figure 3. Events (i.e., vehicular traffic) recorded on sensors in Ironwood Forest NationalMonument, Arizona, 1 January 2002–13 July 2004.

Cemetery) were compared to the 8 sensors thatextended from the Reservation border towardsmain roads (i.e., Agua Blanca, Pad 5, Pad 6,Pad 2, Pad 3, Red Hill 2, Red Hill 1, andSamm Hill) (Figure 4). The Ragged Top loopsensors averaged 0.39 (95% CI = 0.30–0.50)events/night and overall 31.8% of the eventsrecorded by these sensors were at night. Thesensors extending from the reservation aver-aged 1.56 (95% CI = 1.10–1.85) events/nightand overall 67.8% of the events recorded bythese sensors were at night.

Events per 24 hour period by year, timeperiod (night or day), season, and all possibleinteractions using MANOVA (F19,6 = 4.074,P = 0.045). Time period influenced the model(F1,6 = 11.747, P = 0.014). Means were testedfor all sensor events per 24 hour periodcombined between night and day. Weconcluded that there was a strong indication ofa difference between means (t12 = 2.01, P =0.057) of night (x = 0.86, 95% CI = 0.39–1.32)and day (x = 1.38, 95% CI = 1.04–1.72). Theyear did not influence the MANOVA model(F1,6 = 0.980, P = 0.359). Season warrantedfurther investigation (F4,6 = 3.809, P = 0.071)using ANOVA.

The number of events recorded duringwinter was 4,070 (2.31 events/day), spring n =4,555 (2.22 events/day); dry summer, n = 4,757(1.78 events/day); wet summer, n = 1,983 (1.76events/day); and autumn, n = 4,004 (1.85events/day). The means of events per 24 hourperiod for the 5 seasons were compared usingANOVA. None of the means were significantlydifferent (F4,12 = 0.646, P = 0.645); events per24 hour period ranged from a mean of 1.68(95% CI = 0–3.65) in wet summer to a meanof 2.86 (95% CI = 1.33–4.40) in spring. Meansof events/day compared by season were notsignificantly different (F4,8 = 0.545, P = 0.708);events per 24 hour period ranged from a meanof 1.03 (95% CI = 0.13–1.95) in dry summerto a mean of 1.83 (95% CI = 0.70–2.96) inautumn. Means of events/night by season werenot significantly different (F4,8 = 0.619, P =0.662); events per 24 hour period ranged froma mean of 0.40 (95% CI = 0–1.73) in autumnto a mean of 1.25 (95% CI = 0.16–2.34) inspring.

Linear regressions indicated no change innighttime and daytime events over the studyperiod, when analyzed separately. The slope ofthe line was not significantly different from 0


Table 1. Problems reported by visitors (n = 103) to Ironwood Forest National Monument, Arizona, 1 January 2002 through 13 July 2004.

Situation n x SD

Garbage dumping 88 2.56 1.24Illegal immigrant activity 89 2.34 1.38Lack of visitor information 76 1.96 1.14Lack of law enforcement 79 1.75 1.11Unsafe target shooting 73 1.66 1.03Impacted visual resources 73 1.60 1.01Damage/collection of vegetation 72 1.56 1.01Damage/collection of petroglyphs 72 1.56 1.07Lack of trails for non-motorized activities 86 1.55 1.00Reckless drivers on or off trails 76 1.51 0.95Feeling safe 81 1.51 0.90Lack of camping facilities 74 1.32 0.74Damage to livestock 72 1.25 0.78Conflicts with other users 71 1.21 0.56Noise 72 1.18 0.61


Figure 4. Ratios of nighttime and daytime vehicular traffic, determined from sensors inIronwood Forest National Monument, Arizona, 1 January 2002–13 July 2004.

(F1,53 = 0.048, P = 0.827) for events/night orevents/day (F1,53 = 0.127, P = 0.723). Therewas no overall increase in nighttime ordaytime events during the study. However,when night and day data were combined andmodeled by year and season there was a

general overall decline in activity (F1,24 =4.407, P = 0.047) (Figure 5).

When examined within season, nighttimeevents (Figure 6) and daytime events (Figure7) revealed no consistent pattern of increase ordecrease. For example, in 2002 there were 0.87


Figure 5. Linear regression model of events/day by year/month, Ironwood Forest NationalMonument, Arizona, 1 January 2002–13 July 2004.

Figure 6. Events/night recorded by season and year in Ironwood Forest NationalMonument, Arizona, 1 January 2002 – 13 July 2004.


events/night (95% CI = 0.64–1.09), in 2003n = 1.16 (95% CI = 0.86–1.45), and in 2004n = 0.57 (95% CI = 0.42–0.72). In 2002 therewere 1.65 events/day (95% CI = 1.23–2.08),in 2003 n = 2.15 (95% CI = 1.59–2.70), and in2004 n = 1.57 (95% CI = 1.16–1.97).

During the 8 nights (1800–0600 hr) thattraffic was observed (Table 2), 6 vehicles werecounted and 5 met the criteria for being consid-ered illegal. The level of nighttime traffic alongthis road averaged 3.58 (95% CI = 2.66–4.50)vehicles/night according sensor data.

DISCUSSION AND CONCLUSIONSThis study started out with the assumption thathuman use in IFNM was increasing, particu-larly nighttime use. Results of this studyindicate that both day and nighttime use actu-ally have slightly declined over the years forwhich we collected data, but remain signifi-cant. Perhaps shifting drug operations,

increased media attention and law enforcementactivities along the border have lead to thisreduction in activity. It is possible that thisslight decline might be a result of OperationGatekeeper launched in 1994 or at least sinceIFNM was established in 2000. It is alsopossible that illegal activity has decreased orshifted to other areas (e.g., near Yuma,Arizona) as the result of the Arizona BorderControl Initiative, which increased enforce-ment efforts in the Tucson sector especially inand around the Tohono O’odham Nationstarting in Fiscal Year 2004 (C. Griffin, pers.comm.; LoMonaco 2005).

What we did find in this study is that humanuse by season in IFNM is fairly consistentthroughout day, night, season, and year. Otherstudies in arid regions that examined humanrecreation patterns found heavier use in coolerseasons than warmer seasons (Purdy and Shaw1981; Gimblett 2002; Gimblett and Sharp

Figure 7. Events/day recorded by season and year in Ironwood Forest National Monument,Arizona, 1 January 2002 – 13 July 2004.

2005). This pattern was not observed for night-time activity during this study and overallseasonal differences were not significant.Illegal traffickers traveling through the desertat night are likely responding to factors otherthan temperature (e.g., law enforcementactivity, economic influences).

Based on the survey responses related to thetimes of visitor use and sensor data on timesof recorded events it is reasonable to assumethat the majority of nighttime use is illegal.This was also supported by our night observa-tions of traffic where 5 of 6 vehicles fit ourcriteria for illegal activity. Illegal activityoccurring in IFNM is also evidenced by theamount of refuse (e.g., water bottles, foodcontainers, and clothing) left behind andcollected by BLM (F. Mendoza, pers. comm.).

The value of incorporating visitors into thedecision making and the management ofIFNM was revealed in their responses to oursurveys. This important information fromperceptions of visitors did indicate that trashdumping and illegal immigration were theirgreatest concerns. While a detailed study of the

spatial pattern of trash dumping has not specif-ically been undertaken, it is impossible todirectly link such activity to illegal immigra-tion. Circumstantial evidence does indicatehowever, that in specific areas on the monu-ment (pick up or stop over areas) there is anaccumulation of specific types of trash that areassociated with illegal immigrants (whiteplastic water containers, clothing and otheridentifiable supplies). Certainly, more workneeds to be done in this area.

One of the interesting findings of this studyis that the pattern of illegal activity is notevenly distributed throughout IFNM but inten-tionally spatially focused and concentrated.The main corridor for illegal activity extendsfrom the southwestern section along the IFNMborder with the Tohono O’odham Nationnortheast towards the main roads (e.g., AvraValley Road and Interstate 10) (Figure 4). Thisis one of the critical areas where IFNM inter-sects with a main corridor for illegal traffickingcoming across the United States-Mexicoborder and traveling north through the TohonoO’odham Nation (F. Mendoza, pers. comm.).


Table 2 - Night (1800-0600 hr) observations of vehicles in Ironwood Forest National Monument, Arizona, 2004 for 8 nights (96 hr total).

Date Direction Vehicle type Number of Lights Activity Time of travel Age occupants on/off category

5/12/04 North Mid size pickup truck 18 Off Illegal immigration1920 Early 1980s

5/13/04 North Pickup truck w/camper > 5 Off Illegal immigration0440 1990s

8/22/04 North Full size pickup truck > 5 Off Illegal immigration0515 Early 1990s

11/05/04 South Small pickup truck 2 Off Illegal immigration1159 Late 1990s

11/06/04 North Mid size pickup truck > 5 Off Illegal immigration0320 Late 1990s

11/07/04 South Sport utility 2 On Recreation0545 Early 1990s


Trafficking activity probably leaves the Nationand spreads through IFNM as it connects withmain roads. Coincidentally, the patterns ofillegal or nighttime use in the southwesternsection of the IFNM coincided with the loca-tion of the highest concentration of recreationsites (campsites). These recreation sites repre-sent a majority of the class 4 (high impact)recreation sites in the monument. A majorityof these recreation sites are commonly usedtemporarily by illegal visitors as rest spots andby legal visitors undertaking recreationalcamping in the area. Evidence of trash associ-ated with illegal entrants is commonly foundin some of these and other sites in the area.While there is no observed use data on thesesites by illegal activities, just knowing thatnighttime use frequents these sites and is rela-tively high at times when legal visitors arepotentially camped at these locations repre-sents a serious conflict and ultimately a safetyconcern for the management staff of IFNM. Inaddition, anecdotal evidence suggests thistraffic pattern of illegal entrants and traffickingis not only going northbound but humans alsotravel south to bring money across the borderand pick up people and goods to smuggle.Other areas in IFNM have a much smallerproportion of nighttime activity (e.g., thecentral section of IFNM around the Silver Belland Ragged Top Mountains). This area is apopular destination for recreation activity andother daytime use (Figure 4).

Illegal activity is a significant portion of theactivity occurring in IFNM, however, it is hasremained consistent from 2002–2004. Thetraffic, illegal roads and trails, trash, andnatural resource degradation associated withUDA trafficking are a significant concern inIFNM and a major reason that neighboringOPCNM has been placed on the list of theUnited States’ 10 most endangered nationalparks (National Parks Conservation Associa-tion Official website; http://www.npac.org).

In conclusion, studying spatial/temporalmovement patterns of humans for public landsmanagement is imperative. Resource managersneed to understand how the landscape theymanage is being used by legal and/or illegal

activities. Studying the spatial/temporalpatterns of human use reveals that areas withinpublic lands are not being equally visited ratherthat some areas are more heavily visited andimpacted than others and by certain types ofgroups. Resource managers need to understandsuch patterns of use and the seasonal distribu-tions in order to develop effective managementplans and recommendations that respond toand address such disparities. Too often visitorsurvey tools employed to gather informationabout visitors for human management havedone little to provide useful information tomanagers who need to manage for dispersedand concentrated use in large scale settings.Too often this information is used to charac-terize the visitor but says nothing about howthey interact with other visitors and impact thesetting they frequent. This non-traditionalmethod for monitoring, evaluating andmanaging human use described in this chapterprovided information on the nature of the visit,and captured the spatial/temporal patterns ofdaily and seasonal use. This information isvaluable for resource managers attempting tobalance recreation use with resource protec-tion.

While this study was undertaken to eval-uate the spatial/temporal distribution of visitorswho frequent IFNM, through a sampling andmonitoring plan, it was never intended tocapture illegal or nighttime activity. Fewstudies, if any have been specifically designedto evaluate the impacts of illegal entrants andtrafficking on public lands. By using non-inva-sive pressure sensitive pads and othertechnologies such as Global PositioningSystems (GPS), radio frequency technologyand others in combination with surveys andobservation, visitor use patterns can be discov-ered and documented. Future work needs toaddress the short and long-term impacts ofillegal activities before our public land is soseverely impacted for future generations.

LITERATURE CITEDAndreas, P. 2000. Border games: policing the U.S.-

Mexico divide. Cornell University Press, Ithaca,New York, NY.

Billington, C. C. 2005. Monitoring illegal activityon wildlands. Matsters Thesis, School of NaturalResources. University of Arizona, Tucson.

Bureau of Land Management. 2004. Scoping report:Ironwood Forest National Monument resourcemanagement plans and environmental impactstatement. February 12, 2004. http://www.az.blm.gov/lup/ironwood/scoping_repoort.pdf.

Bristow, K. D., J. A. Wennerlund, R. E.Schweinsburg, R. J. Olding, and R. E. Lee. 1996.Habitat use and movements of desert bighornsheep near the Silver Bell Mine, Arizona. ArizonaGame and Fish Technical Report 25.

Clinton, W. J. 2000. Ironwood Forest NationalMonument Proclamation. June 9, 2000.http://www.az.blm.gov/lup/ironwood/ironproc.htm.

Cohen, P. 1997. Rare plants trampled in immigra-tion crackdown. New Scientist, 153: 66.

Department of Homeland Security. 2004. Depart-ment of Homeland Security Announces ArizonaBorder Control Initiative. March 16, 2004.http://www.dlhs.gov.

Gimblett, H.R., editor. 2002. Integrating geographicinformation systems and agent-based modelingtechniques for simulating social and ecologicalprocesses. Oxford University Press, London,England.

Gimblett, H. R. 2004. Ironwood Forest NationalMonument access, travel route inventory andvisitor use study. Final Report. Prepared for theBureau of Land Management.

Gimblett H. R. and C. Sharp. 2005. Monitoring andestimating visitor use levels at the MadronaRanger Station. Prepared for the Rincon Institute.May 2005.

Hinkle, Jeff. 2001. Heavy traffic: The TohonoO’odham border crossing remains a hot spot in thedrug was. Tucson Weekly. 14 June 2001.

Jacoby, T. 2004. Borderline: Why we can’t stopillegal immigration. The New Republic, 26January 2004.

LoMonaco, C. 2005. No drop in migrant flow ordeaths: Multimillion-dollar border effort had littleimpact, analysis shows. Tucson Citizen, 16 March2005.

Marizco, M. 2003. Desert water stop quicklyvandalized. The Arizona Daily Star. 5 May 2003.

McIntyre, D. L., and J. R. Weeks. 2002. Environ-mental impacts of illegal immigration on theCleveland National Forest in California. TheProfessional Geographer 54: 392–405.

Nevins. J. 2002. Operation Gatekeeper: The rise ofthe ”illegal alien” and the making of the U.S.Mexico boundary. Routledge, New York, NY.

Newell, L. Anne. 2004. Drop house not likely atrend, officials say. The Arizona Daily Star. 3 May2004.

Purdy, K., and W. W. Shaw. 1981. An analysis ofrecreational use patterns in desert bighorn habitat.Desert Bighorn Council Transactions 25: 1–5.

Sall, J., L. Creighton, and A. Lehman. 2005. JMPIN the statistical discovery software version 5.1.Brooks/Cole-Thomson Learning, Belmont, CA.

Sellers, W. D., and R. H. Hill. 1974. Arizonaclimate: 1931–1972. University of Arizona Press,Tucson, AZ.

Teibel, D. L. 2003. Man shot by ranger ID’d. TucsonCitizen. 14 May 2003.

Tersey, D., B. Drennen, and D. Moore. 2001. Pre-plan analysis for the Ironwood Forest NationalMonument. BLM Field Office, Tucson, AZ.

Titre, J., M. Bates, R. Gumina, and M. Jones. 2004.Boulder open space and mountain parks (OSMP)use estimation and visitor survey study for theChautauqua study area. Fall 2003 Project FinalReport. Park Studies, Boulder, CO.

Tobin, M. 2004. A run along the abyss. The ArizonaDaily Star, 7 March, 2004.

Turf, L. 2002. When smugglers roam. The ArizonaDaily Star. 27 December 2002.

Vanderpool, Tim. In Desert heat, furor over migra-tion simmers on. Christian Science Monitor. 14June 2003.

Weber, S. 1988. USA by numbers: A statisticalportrait of the United States. Zero PopulationGrowth, Washington, D.C.

Whitford, W. G., Y. Steinberger, W. MacKay, L. W.Parker, D. Freckman, J. A. Wallwork, and D.Weems. 1986. Rainfall and decomposition in theChihuahuan Desert. Oecologia 68: 512–515.


THE U.S.-MEXICO BORDER AND ENDANGERED SPECIESDouglas K. Duncan, Erin Fernandez, and Curtis McCasland

THE SPECIESThe Arizona-Sonora Border area has a greatdiversity of ecosystems and species. TheEndangered Species Act (ESA) protects 26listed species of plants and animals in the area,18 of which are endangered (10 of those havedesignated critical habitat) and 8 of which arelisted as threatened. Species that occupyriparian and aquatic communities make up halfthe group. At least 25 of these are certain tooccur in the border region (Table 1). Addi-tionally, species that are candidates for listing,managed under conservation agreements orconsidered sensitive, also occur in this area.

Listed species occur along much of the inter-national border in Arizona. From the ColoradoRiver in the west to the Peloncillo Mountains inthe east, species occurring along the borderlandsof Arizona include the Yuma clapper rail (Ralluslongirostris yumanensis), Sonoran pronghorn(Antilocapra americana sonoriensis), Pimapineapple cactus (Coryphantha scheeri var.robustispina), Chiricahua leopard frog (Ranachiricahuensis), and the New Mexico ridge-nosed rattlesnake (Crotalus willardi obscurus).Many of the species, such as the Chiricahualeopard frog and Huachuca water umbel(Lilaeopsis schaffneriana ssp. recurva) occupyriparian and aquatic communities, so their distri-bution is discrete. Other species that are moremobile, such as the lesser long-nosed bat(Leptonycteris curasoae yerbabuenae), couldoccur almost anywhere in suitable habitat. Thespecies most affected by illegal border crossingand law enforcement is the endangered Sonoranpronghorn.

Sonoran PronghornSonoran pronghorn inhabit the hottest anddriest portions of the Sonoran Desert. Histor-ically this subspecies ranged in the U.S. fromthe Imperial Valley, California to near theSanta Cruz River in the east, from Gila Bendand the Kofa Mountains south to the interna-tional border. Currently the subspecies rangesfrom the Copper and Cabeza Prieta Mountainseast to State Route 85, and from Interstate 8south to the international border. Thesubspecies’ range and population size declineddramatically in the early 20th century, howeverthere were no definitive studies or informationavailable to document population size before1925. Within the last 12 years, the U.S. popu-lation of pronghorn has ranged from 142individuals in 1998 to 21 animals in 2002. Wecurrently (2009) estimate the wild populationat about 85 individuals, which includesanimals released from the semi-captivebreeding pen.

Fawn survival is considered to be one of themost important factors affecting populationsize, and this is our most significant manage-ment issue. Pronghorn fawn survival isstrongly correlated with the length of timebetween winter and summer rain events. Thenumber of fawns surviving until the firstsummer rains is also correlated with theamount of winter rainfall (Hervert et al. 2000).Obviously, severe drought affects not onlyfawns but the whole population; the severedrought of 2001 and 2002 nearly extirpated theentire U.S. population. We are currently imple-menting numerous recovery actions to increase

124 Duncan, Fernandez, and McCasland

Table 1. Listed species status and presence1 along the Arizona-Mexico border.

Species group Listed Endangered Critical habitat Present

Plants 5 4 1 5Amphibians 2 1 0 2Reptiles 2 0 0 2Fish 8 5 6 8Birds 7 5 3 5Mammals 6 6 0 3Total 30 21 11 25

1 Certain to be found within 100 kilometers of the border at least part of the year.

fawn survival both through direct intervention(semi-captive rearing), irrigation of nativeforage plots, and providing numerous waterssources around heavily used portions of theirrange.

EFFECTSEffects to listed species and the ecosystems onwhich they depend occur both from illegalcross border traffic and law enforcement activ-ities aimed at interdicting it. Illegal traffic canbe categorized in several ways. The illegalborder crossers themselves are generally eitherimmigrants seeking work or smugglers ofcontraband, largely drugs. The effects ofcoyotes, or people smugglers, will be consid-ered the same as illegal immigrants. Typically,illegal immigrants cross the border via foot,rarely using a vehicle, while smugglers mostoften use vehicles to cross even in roadlessareas. Bicycles, especially in flat terrain, andstock animals are also used.

Law enforcement interdiction activities,chiefly by U.S. Border Patrol (USBP), arelargely conducted by vehicle. Most vehiclesbeing used have four-wheel drive, and range insize from quads to heavy specialty vehicles. Agreater proportion of law enforcement vehicleuse is on designated roads and trails, whencompared to vehicle use by illegal entrants. Law

enforcement personnel also use motorcycles andhorses, and they sometimes patrol on foot.Unmanned aerial vehicles are also deployedalong the Arizona border.

Direct, indirect, and cumulative effectsfrom the above border activities are analyzedunder section 7 consultations on border activ-ities. All the classes of effects occur from bothlaw enforcement and illegal entrants. Many ofthe specific effects are common to both groups.For example both illegal and law enforcementactivities can create new roads and trails;disturb vegetation and soils; disturb wildlife;impact wildlife movement corridors; and movenonindigenous species. Some specific effects,however, such as entrant trash, are only asso-ciated with illegal entrants.

Actions by illegal entrants are basically allassociated with the act of crossing the interna-tional border and journeying towards adestination. Smugglers usually travel bothdirections. Some smugglers turn themselves into the USBP, posing as illegal immigrants, andget a free trip back to the border. Illegalentrants travel on all roads and trails, fromInterstate Highways to designated trails infederally designated wilderness areas. Vehic-ular travel, both illegal and law enforcement,on designated roads has little to minimaleffects, except for back country roads that

Duncan, Fernandez, and McCasland 125

receive a lot more use now than they didhistorically. Significant effects to resourcesoccur when the illegal entrants and pursuantlaw enforcement travel off of roads. Off-roadtravel may adversely affect listed species fromthe creation of roads and trails; increasedhuman disturbance, especially in wildernessareas or caves and mines; use of water sources;soil compaction and resultant changes inhydrology; introduction of nonindigenousspecies; spread of disease; destruction orvandalism of wildlife waters or other wildlifeprojects such as fences (e.g., Sonoran prong-horn pen fence); trash; pollution; destructionof vegetation; and fires.

Direct effects occur when an individuallisted species or occupied habitat are impactedby an activity. Both illegal and law enforce-ment activities have direct effects on listedspecies. Direct effects could occur to listedspecies when vehicles cross a stream and runover an individual listed plant or animal, byaircraft disturbing nesting listed birds, or evenhumans capturing listed animals for food.Direct effects are usually more obvious thanthe impacts discussed below, however, effectsthat are not direct often cause the greatest harmor concern because they can be more subtleand difficult to analyze, can occur outside ofthe foot print of an action, can occur later, orcan be difficult to measure.

Recovery actions designed to help recoverthe U.S. population are significantly impactedby illegal cross-border traffic. Illegal bordercrossers have exploited water sources we knoware important for pronghorn, smashed irrigationlines used to promote quality native forage forpronghorn, and cut and dug under fences aroundthe pronghorn captive rearing pen. Vehiclessmuggling drugs and people routinely drive off-road through areas used by pronghorn duringthe hot dry summer months. These actionslikely result in the most significant effects topronghorn from illegal traffic. Any activityforcing pronghorn to flee from habitat used tostay cool when free water and water found inforage is scarce to absent has the potential todevastate the pronghorn population. These

activities may force females to stop nursingfawns, leading to greater fawn mortality.Recently weaned fawns may also be affectedfrom these disturbances, and adult mortalitymay even occur. Furthermore, law enforcementpersonnel responding to sensors, vehicle tracks,or foot sign also travel through these areas.Unfortunately, it is not possible to specificallyquantify the impacts associated with these activ-ities; however, we believe the impacts aresignificant. The drought of 2001-2002 and thedramatic pronghorn population decline occurredat a time of rapid increase in illegal smugglingof people and contraband into the U.S. and asubsequent increase in interdiction efforts byFederal law enforcement agencies.

Indirect effects of an action are those thatoccur later in time but are reasonably certainto occur. An example would be a roadconstructed through a Sonoran pronghornfawning area, and the later use of that roadcausing fawning to cease or occur in less suit-able habitat, resulting in decreased fawnrecruitment.

An interdependent action is one that has noindependent utility apart from the action beingconsidered. An example would be if the USBPproposes building a barrier fence at the border.To construct the barrier, vehicles must useexisting or new roads to access the project site.The use of these roads is interdependent toconstruction of the barrier, and would notoccur “but for” the barrier. The road itself isnot interdependent.

Interrelated actions rely on the larger actionfor their justification. Using the previousexample, the construction of the barrierrequires water, but there is no well nearby. Tofacilitate construction of the barrier, USBPproposes to drill a well that will be used fornothing but construction. Because the wellmust be constructed to construct the barrier,drilling the well is an action that is interrelatedto the proposed action of barrier construction.

Under section 7 regulations [50 CFR§402.02], cumulative effects are those effectsof future State or private activities, notinvolving Federal activities, that are reasonably


certain to occur within the action area of theFederal action subject to consultation. Thisdefinition applies only to section 7 and shouldnot be confused with the broader use of thisterm in the National Environmental Policy Actor other environmental laws. Let’s suppose thebarrier example above is proposed to be builtnear Lukeville. About that time, the town ofLukeville announces plans to expand (and wasable to do so), and had the appropriate approvalto do so. As long as there is no Federal connec-tion or tie to the Lukeville expansion (permit,funding, or approval), then the effects of thatexpansion to the Sonoran pronghorn would becumulative to the construction of the barrier.Cumulative effects only come into play insection 7 consultation during formal consulta-tion, to determine if the proposed action mayjeopardize the continued existence of thespecies. No incidental take is anticipated forcumulative effects.

Incidental take is take of listed fish orwildlife species that results from, but is not thepurpose of, carrying out an otherwise lawfulactivity conducted by a Federal agency orapplicant (50 CFR §402.02). Take is furtherdefined as to harass, harm, pursue, hunt, shoot,wound, kill, trap, capture, or collect or attemptto engage in any such conduct (ESA §3[19]).“Harm” is further defined by U.S. Fish andWildlife Service (USFWS) to include signifi-cant habitat modification or degradation thatresults in death or injury to listed species bysignificantly impairing behavioral patternssuch as breeding, feeding, or sheltering.“Harass” is defined by USFWS as actions thatcreate the likelihood of injury to listed speciesto such an extent as to significantly disruptnormal behavior patterns which include, butare not limited to, breeding, feeding or shel-tering (50 CFR §17.3).

ESA COMPLIANCEEndangered Species Act compliance along theborder is extremely complicated because of thenumber of land owners and managers in theregion, the number of Federal agenciesinvolved with border issues, the geographic

and technical scope of border problems, andthe ever-changing nature of illegal cross-bordertraffic and law enforcement. Additionalcomplication stems from the creation of theDepartment of Homeland Security (DHS) andU.S. Customs and Border Protection (CBP)and its resultant complex internal structure.

Every Federal agency must determine ifany of their actions may affect a listed speciesor critical habitat, and if so, consult with theUSFWS. All Federal agencies are required bysection 7 (Interagency Cooperation) of theESA to ensure that their actions do not jeop-ardize a listed species or result in thedestruction or adverse modification of criticalhabitat. A Federal action or Federal nexus isdefined as an action that is authorized, funded,or carried out by the agency. Consultation canbe either informal or formal. If an actionagency determines that there is no effect tolisted species from the Federal action, thennothing further is required by the Act. If theagency determines the action may affect, butis not likely to adversely affect listed species,they then seek the written concurrence of theUSFWS (informal consultation). When theaction agency determines that their action mayaffect and is likely to adversely affect listedspecies, then formal consultation with theUSFWS begins and a biological opinion iswritten by the USFWS. The biological opinionexamines the effects of the proposed action,analyzes the proposed effects against thespecies’ status and environmental baseline, anddetermines whether the proposed effects willjeopardize the continued existence of the feder-ally listed species.

The ESA allows for actions to occur beforesection 7 consultation if the action agencydetermines there is an emergency. An emer-gency is a situation involving natural disasters,casualties, national defense or security emer-gencies, and includes response activities thatmust be taken to prevent imminent loss ofhuman life or property. Broad discretion isgiven to the action agency to determine whenan emergency has occurred. Many BorderPatrol actions have been deemed emergencies;


creation and operation of camp details, variousborder initiatives, and helicopter landings inwilderness were all considered emergencyactions by the Border Patrol.

In an emergency, an action agency must firstdetermine if their response to the emergencymay affect listed species and, if so, notify theUSFWS. Upon notification, the USFWS iden-tifies conservation measures that may reducethe adverse effects of the emergency responseto listed species. Conservation measures arediscretionary and may not interfere with theemergency response. The USFWS follows upthe notification from the agency noting anyconservation measures and emergency consul-tation procedures. The agency can respond tothe emergency before notifying the USFWS.When an action agency determines an emer-gency has occurred and their response to it mayaffect listed species or critical habitat, they mustconsult under section 7.

Section 7 consultation on an emergencyaction analyzes only the response to the emer-gency, and not the emergency itself. Forexample, when the USBP established campdetails (a remote base) in southwesternArizona in response to illegal entrant healthand safety issues, the section 7 consultationonly analyzed the effects to federally listedspecies from establishment and operation ofthe camps. The action causing the emergencydeclaration, illegal border traffic, is analyzedas part of the environmental baseline. Theenvironmental baseline section in the biolog-ical opinion includes past and present effectsof all Federal, State, or private actions in theaction area, the anticipated effects of allproposed Federal actions in the action area thathave undergone formal or early section 7consultation, and the impact of State andprivate actions which are contemporaneouswith the consultation process. The environ-mental baseline defines the current status ofthe species and its habitat in the action area toprovide a platform to assess the effects of theaction now under consultation (USFWS 1998).

The above described section 7 consultationprocess for some CBP projects was signifi-

cantly affected by the enactment of the Real IDAct of 2005, which allows the Secretary of theDHS to waive many laws, including the ESA,to expeditiously construct border fences androads. The April 1, 2008, waiver covers envi-ronmental laws, including the ESA. Thiswaiver covers the construction, operation, andmaintenance of tactical infrastructure to includefixed and mobile barriers (such as fencing,vehicle barriers, towers, sensors, cameras, andother surveillance, communication, and detec-tion equipment) and roads near the internationalborder (Department of Homeland Security2008; website accessed on 12 January 2010;http://cbp.gov/xp/cgov/border_security/ti/ti_docs/esp_information.xml). About 470 milesof the U.S./Mexico international border arecovered by the waiver.

As of December 25, 2009, CBP completedroughly 643 miles of fencing (344 miles ofprimary pedestrian fence and 299 miles ofvehicle fence) and accompanying roads alongthe U.S./Mexico border (about half were inArizona) without undergoing section 7 consul-tation (Department of Homeland Security2009; website accessed on 12 January 2010;http://cbp.gov/xp/cgov/border_security/ti/ti_news). In lieu of consultation, CBP addressedits environmental impacts through a voluntaryprocess that included the preparation of Envi-ronmental Stewardship Plans and BiologicalResource Plans to substitute for EnvironmentalAssessments and Biological Assessmentsrespectively. However, many of these planswere completed without USFWS input and didnot thoroughly address effects on listedspecies.

The U.S. Army Corps of Engineers (COE)manages environmental compliance for borderfences and related roads for the USBP inArizona. The COE often contracts with outsideparties to assist with environmental compli-ance. Private consultants awarded thesecontracts often utilize subject matter experts assubcontractors. In addition, about 65 percentof the Arizona-Mexico border is managed byFederal agencies; USBP must work with landmanagement agencies to ensure the USBP can


effectively complete their mission while notcompromising the missions of the various landmanagement agencies. Thus, in addition toUSBP, CBP, and COE, various consultants,and the land manager(s) are involved in section7 consultations.

The USBP conducts a variety of actions thataffect listed species in the border region. USBPactivities fall into two broad categories: patroland infrastructure. Patrol activities include allnormal patrol mechanisms, including all typesof vehicles on and off road, horse patrol, footpatrol, and the use of drag roads. Discussionon patrol activities has been batched into oneconsultation for each sector. Consultation oninfrastructure tends to be project specific.However, we have consulted on projects thatinclude both patrol and infrastructure, such asdetection towers. Infrastructure can includenew roads, camps, fences, detection equip-ment, and offices.

One of the major problems regardingconsultation on border law enforcement effortsis that law enforcement responses must changein response to constantly changing bordercrossing patterns. Consultations examining theeffects of patrol activities on federally listedspecies are extremely difficult; project descrip-tions and impacts to listed species candramatically change within extremely shorttime frames, rendering finalized biologicalopinions outdated.

Infrastructure projects that are small or nextto urbanized areas tend not to need consulta-tion because they normally do not affect listedspecies. Listed species rarely occur in or nextto developed areas, and small projects aremuch less likely to intersect with listed speciesor their habitat. Many of the infrastructureprojects completed by the USBP or in variousplanning stages are small, or in developed orotherwise already disturbed areas.

However, the USBP has also completed andproposed the development of larger infrastruc-ture projects that are within listed specieshabitat. A few completed projects have beenthrough section 7 consultation, but there aremany more potential infrastructure projects

likely to need section 7 consultation. There aretwo basic kinds of infrastructure projects: linearand areal. Linear projects can include develop-ment of roads, fences, drag roads, and lights.Areal projects which cover a discrete area caninclude the development of offices, camps,helipads, or surveillance systems. Thoughlinear projects usually do not have a wide foot-print, their extensive length is likely to affectlisted species. Furthermore, there is concernthat linear infrastructure projects may act asbarriers, further fragment habitat, or act as path-ways for increased predation or disease,affecting the potential survival and recovery offederally listed species in the area. Much of theArizona-Sonora border has some sort of linearproject associated with border enforcement.Also, since many of the projects need to be asnear the border as possible, there are feweroptions in situating those projects to reducepotential effects to listed species. Complicatingthis matter is the issue of timing; the USFWSoften negotiates for timing restrictions to mini-mize or avoid impacts to federally listedspecies. Unfortunately, timing restrictions makescheduling work even more problematic.

The complex operational structure of theDHS/CBP/USBP makes environmentalcompliance difficult. Since DHS is a relativelynew Department, program responsibility andproject roles are sometimes unclear betweenconstituent bureaus. USBP or CBP actionsoften require review and approval of multipleagencies in DHS. As an example of thecomplexity, a meeting regarding endangeredspecies issues on a larger project may includeCBP, USBP, COE, USFWS, and consultingfirms, as well as multiple levels (e.g., local andnational) and offices within one agency orconsulting firm. Federal land manager(s) andthe Arizona Game and Fish Department(AGFD) may also be involved. Coordinatingmeetings, following approval processes, andcompleting project and document review iscomplicated with so many entities involved.

An additional complicating factor incompliance is the direct and indirect involve-ment of other Federal agencies, mainly land


management agencies. Though not as often asUSBP, most Federal land managers are directlyinvolved with border enforcement issues,mainly through their law enforcement officersand coordinating with USBP on law enforce-ment activities within each land managementunit. Initially USBP assumed they hadcompleted all environmental regulatoryrequirements upon discussing needs with themanagers of the various land managementunits. Federal land managers are indirectlyinvolved with USBP ESA compliance throughoperational coordination between their lawenforcement staff and USBP, and with pre-planning infrastructure projects. An interestingexample of another complicated and complexissue is that other agencies sometimes requestconsultation basically on behalf of USBP.

A relatively recent improvement for envi-ronmental compliance was the creation ofPublic Lands Liaison Agents in the Tucson andYuma Sector USBP offices. Though the mainduties of the agents are to coordinate accessand other issues with public land managers,they have also been involved with project plan-ning and environmental compliance. Theliaisons have helped with several endangeredspecies issues, including the reduction ofUSBP helicopter over flights of the Sonoranpronghorn semi-captive breeding pen.

Avenues other than the more formalizedstructure provided by section 7 consultationhave been developed which allow for input onUSBP actions. The Borderlands ManagementTask Force (BMTF) includes mostly Federalagencies, but also includes Tribal governments,state agencies, and Congressional staff.Membership is limited because sensitive lawenforcement information is discussed. Themission of the BMTF is:

“to facilitate an intergovernmental forumfor cooperative problem-solving oncommon issues related to the Arizona-Mexico border. The primary mission isto address border security, human safety,and natural and cultural resourceprotection through shared resources,

information, communication, problem-solving, standardization and training.”

Border infrastructure project planningoccurs through monthly Project Delivery Team(PDT) meetings. The main purpose of the PDTis to construct new and enhance existingborder infrastructure to improve the effective-ness of USBP operations and activities (CBP2005). Most staff involved with the PDT areproject managers, engineers, and USBP agents,though certain state and Federal agencies mayattend. The PDT develops and tracks projectsfrom concept through construction.

The BMTF and PDT meetings facilitatecommunication between the USBP and others,though specific endangered species issues orcompliance are rarely discussed.

The AGFD may also be involved with ESAcompliance and border issues. There is anexisting Memorandum of Agreement betweenthe USFWS and AGFD that allows the AGFDto participate in Section 7 consultation, subjectto action agency approval (CBP or USBP inthis case). AGFD staff periodically attendsBMTF and PDT meetings. AGFD’s concernsinclude effects to all fish and wildlife and theirhabitats.

SOLUTIONSThere are procedures and measures thatFederal agencies can take to make ESAcompliance more efficient from a proceduralperspective and more protective of listedspecies and their habitats. The PDT, BMTF,and Public Lands Liaison agents have allhelped improve communication and processesregarding natural resource protection. Theircontinued use and staffing, especially by FWSand USBP will continue to improve ESAcompliance and resource protection. Addi-tional measures that can also improve process,compliance, and resource conservationinclude: batch or programmatic section 7consultations, environmental staff at USBPSectors, streamlined and simplified processes,and environmental training for CBP, USBPagents, consultants and other staff.


A batch consultation is one where similaractions are grouped together. Examples mightbe all USBP road projects in the Tucson Sector,or all infrastructure in the San Rafael Valley.A programmatic section 7 consultation is onethat covers a program. An example is a USBPoperations consultation on border activities thatcovers an entire sector of USBP. This consul-tation will cover all ongoing patrol actions, andwill also address the potential for an increasein USBP staffing levels and patrol activities.Batch and programmatic consultations takemore effort initially, but they can precludework later, or at least make it simpler.

The addition of environmental staff at theUSBP Sector level would address severalissues. A main complicating factor with CBPand USBP ESA compliance is the difficulty incommunicating with all the entities involved,and within the CBP/USBP hierarchy. Havinga local CBP/USBP environmental staff tocoordinate compliance and local CBP/USBPcommunications would greatly enhance ESAcompliance and resource protection. A relatedaction that would also streamline and enhancecompliance would be to have CBP and USBPdelegate ESA compliance to the lowest levelspossible. To be effective that would requiresufficient staffing, funding, training, andresources.

Finally, the real solutions to illegal cross-border traffic and its impacts to listed specieswill not be achieved by the CBP/USBP orUSFWS, or even any of the Federal agenciesor consultants involved. The root causes ofcontraband smuggling and illegal immigrationneed to be dealt with. The incentives for crossborder traffic must be removed and will beaddressed where policy, legislation, andeconomics intersect. The problems have beendecades in the making, and will likely takedecades to truly fix.

The findings and conclusions in this article arethose of the authors and do not necessarily repre-sent the views of the U.S. Fish and WildlifeService.

LITERATURE CITEDHervert, J. J., J. L. Bright, M. T. Brown, L. A. Priest,

and R. S. Henry. 2000. Sonoran pronghorn popu-lation monitoring: 1984-1998. Nongame andEndangered Wildlife Program, Technical ReportNo, 162. Arizona Game and Fish Department,Phoenix, AZ

U.S. Customs and Border Protection. 2005. TucsonSector: Project Delivery Team (PDT) Work Plan.107pp.

U.S. Fish and Wildlife Service and National MarineFisheries Service. 1998. Endangered speciesconsultation handbook: Procedures for conductingconsultation and conference activities undersection 7 of the Endangered Species Act. Wash-ington, D.C. 290pp.

ADAPTIVE MANAGEMENT OF THE GRASSLAND-WATERSHED AT LAS CIENEGASNATIONAL CONSERVATION AREA: THE ROLE OF MONITORING, RANCHER ENGAGEMENT, AND MULTI-STAKEHOLDER ADVISORY TEAMSDavid Gori, Karen Simms, Mac Donaldson, Gitanjali Bodner, and Heather Schussman

Land managers are frequently forced to makeland management decisions with little or noinformation regarding the outcomes of thosedecisions. Science-based adaptive manage-ment is a process designed to change thisthrough the collection of information thatassists in evaluating the effects of managementactions and in identifying knowledge gaps thatcan be addressed by research or additionalmonitoring. The result is a decision-makingprocess based on learning and information.

The adaptive management process requiresseven primary steps:1) Identification of management goals or

objectives and measurable thresholds inresource condition.

2) Development of a monitoring protocol withadequate sampling effort to detect biologi-cally meaningful change in resourcecondition over a specified time period andan optimal frequency and timeframe duringwhich monitoring should be conducted.

3) Consistent implementation of the moni-toring protocol.

4) Analysis of data.5) Review of data against established goals

and thresholds to determine the need forchanges in management.

6) Implementation of needed managementchanges (and continued monitoring).

7) Implementation of follow-up scientificstudies to fill identified information gaps.

In this paper, we describe an ongoing adap-tive management process at Las CienegasNational Conservation Area that assists theBureau of Land Management (BLM) and therancher who holds a grazing permit there(hereafter, referred to as the permittee) inmaking grazing management decisions. Aspart of this adaptive management process, weevaluated the upland monitoring protocol andproposed revisions that increased theprotocol’s statistical power to detect change ingrassland-watershed condition. Data collectedin 2004 and 2005 using the revised protocolwere analyzed and used by the BLM, thepermittee and multi-stakeholder teams tomodify the permittee’s proposed annualgrazing plan, including stocking numbers andplanned pasture rotation. Follow-up moni-toring documented the effect of these grazingmanagement decisions.

THE SITELas Cienegas National Conservation Area(NCA) and the surrounding Sonoita ValleyAcquisition Planning District (SVAPD) are amix of BLM, Arizona State Trust, and privatelands, encompassing nearly 39,000 ha (96,371ac). Both were designated by Congress andsigned into law by the President in December,2000, in order to conserve, protect, and

132 Gori, Simms, Donaldson, Bodner, and Schussman

enhance the unique and nationally importantresources there. Among these resources, theNCA and SVAPD support five of the raresthabitat types in the Southwest—cienegawetland, cottonwood-willow riparian forest,sacaton riparian grassland, mesquite bosque,and desert grassland—as well as six federally-listed species. The NCA’s enabling legislationalso allows for the continuation of livestockgrazing and recreation. Las Cienegas NCAforms the northern anchor of a 323,887 ha(800,342 ac) conservation area identified inThe Nature Conservancy’s (TNC) ApacheHighlands ecoregional assessment of conser-vation priorities (Marshall et al. 2004). In ananalysis of 600 TNC conservation areas iden-tified in the five ecoregions overlappingArizona, the conservation area that includesLas Cienegas NCA ranked highest in terms ofbiological uniqueness and irreplaceability.

The NCA contains a portion of one of thebest remaining open native grasslands in theborderland region; however, shrub encroach-ment is becoming an increasing problem in themiddle and northern parts of the NCA andSVAPD (Enquist and Gori 2005; Gori andSchussman 2005). This grassland forms thewatershed for upper Cienega Creek and theriparian and aquatic habitats there, making itan important focus for BLM management.

The Las Cienegas Resource ManagementPlan (RMP) was approved in July 2003,completing a collaborative planning processthat occurred over an 8-year period with theSonoita Valley Planning Partnership (SVPP).The SVPP is a voluntary association of federal,state, and local agencies, organizations, andprivate citizens who share a common interestin the resources and management of the publiclands within the Sonoita Valley, an area thatincludes the NCA and the entire upper water-shed of Cienega Creek. The BLM incorporatedthe vision, goals, and objectives of the SVPPas the foundation for the Las Cienegas RMP(BLM 2003).

BIOLOGICAL PLANNINGThe Las Cienegas RMP states that the BLMwill adopt an adaptive management strategy

(BLM 2003). As part of this strategy, the RMPprescribes a flexible grazing program wherebyauthorized use (stocking numbers and pasturerotation) is varied annually based on an assess-ment of range conditions including monitoringdata. The RMP formalized an ongoing Biolog-ical Planning process that assists the BLM byproviding input and information that theagency can use in making livestock grazingand other management decisions. Through theprocess, the Biological Planning Team assistsBLM with review of monitoring data andprovides input into proposed actions. Compo-sition of the team is a balance betweenresource managers and users. The BiologicalPlanning Team includes the RangelandResource Team (RRT), the Technical ReviewTeam (TRT) and other interested agencies andpublic. The RRT is a committee of ninemembers representing commercial and recre-ational users, environmental organizations,academia, and elected officials, and the TRT iscomposed of state and federal agency resourcespecialists with expertise in range, riparian,watershed, and wildlife management. In addi-tion, public participants, including otherinterested agencies, provide input on TRT andRRT recommendations during BiologicalPlanning meetings (see below).

The Biological Planning process generallyconsists of the following steps that incorporatecomponents of the adaptive managementprocess described above:1) Proposed annual grazing plan is developed

by the permittee.2) Monitoring data are collected and analyzed

by members of the BLM, TRT, and othercollaborators; based on this analysis, modi-fications to the initial grazing plan areproposed.

3) Monitoring data are reviewed by the RRTin context of other issues that may havearisen for the NCA. The RRT reviews theTRT’s recommended modifications to thegrazing plan or other proposed actionsbased on the monitoring data, and makesadditional recommendations, as needed.

Gori, Simms, Donaldson, Bodner, and Schussman 133

4) All recommendations are discussed by theBiological Planning Team as a wholeduring twice-yearly Biological Planningmeetings and receive additional publicinput at each meeting.

5) After review of existing data and recom-mendations from the Biological PlanningTeam, the BLM Field Manager will thenapprove or make any necessary changes tothe annual grazing plan.

RESOURCE OBJECTIVES AND AN EVALUATION OF EXISTING

MONITORING: AN ILLUSTRATIONThe Las Cienegas RMP identifies two generalupland objectives for the grassland-watershed(BLM 2003): 1) Maintain or achieve properly functioning

upland condition and a high similarity index(> 50% by weight) to the historical climaxplant community; and

2) Maintain or achieve ground cover in grass-land communities in excess of 70% (< 30%exposed bare soil) on 80% or more of theecological sites on the NCA and SVAPD by2015.Ecological sites are land units classified by

the USDA Natural Resources ConservationService (NRCS) to assist management; theyare defined by soils, terrain, climate, andpotential and current vegetation (USDA 2009).

Because this paper describes a process forevaluating monitoring protocols used in adap-tive management, and because the similarityindex is a derivative parameter that does notlend itself to an analysis of statistical power,we focus on the second general objective forthe grassland watershed and its associatedmonitoring protocol for illustrative purposes.

To determine if this objective was beingmet, BLM established 31 key areas between1995 and 1998 on the Empire-Cienega Allot-ment (30,000 ha / 74,132 ac), the largest of fourgrazing allotments in the NCA and SVAPD.The key areas, scattered in different pasturesand ecological sites within the allotment, wereperiodically monitored between 1995 and

2003 using a pace-frequency method for esti-mating plant species frequency and apoint-intercept method for estimating substratecover (Herrick et al. 2005; Ruyle, no date).One hundred quadrats, each 40 cm x 40 cm(pace frequency) and 100 points (point inter-cept) were sampled at one-pace intervals (~ 1.5m) along two parallel, 76 m (833 yd) transects.

For frequency sampling, the occurrence ofall plant species, herbaceous and woody,within quadrats was recorded. From this, thefrequency of occurrence for each species wascalculated by dividing the number of quadratsthe species occurred in by the total number ofquadrats sampled. For dry weight ranksampling, the three most abundant species ona dry weight basis were identified in thequadrat and ranked. A pointer attached to thequadrat frame was used to collect point-inter-cept data. The following categories of substratecover were distinguished: bare ground, gravel,rock, litter and live basal vegetation. Descrip-tions of these categories can be found in BLM(2003). Percent cover by substrate categorywas calculated by dividing the number of“hits” in each category by the total number ofpoints sampled (n = 100). All measurementswere conducted in the fall (September-October) after the summer rains whenflowering and annual production of warm-season grasses and herbs, which dominatethese grasslands, had been completed.

We evaluated the upland monitoring proto-cols described above based on three criteria:• How well do the parameters derived from

the existing monitoring protocols addressmanagement objectives;

• What is the statistical power of these proto-cols to detect change; and

• How much time do they take to implement?In addition, we identified several corollary

considerations to facilitate our evaluation,including:• Estimated parameters should explicitly

address upland objectives or critical stresses


that, if unchecked, would prevent key areasfrom meeting these objectives (our logichere is that since resources for monitoringare limited, parameters that directlymeasure progress toward objectives orchanges in critical stresses are more impor-tant than those that do not);

• Sampling effort should be adequate todetect biologically-meaningful change inparameter values over appropriate timeframes;

• Existing agency protocols should be usedor expanded upon when possible; and

• Modifications to protocols should notsignificantly increase the total time alreadyspent monitoring; placing greater emphasison one set of measurements may be offsetby de-emphasizing others.

A comparison of the upland resource objec-tive with information obtained frompace-frequency and point-intercept methodsindicated that only the point intercept methodyielded data pertinent to the bare-groundobjective (Table 1). In addition, neither methoddirectly estimated the extent of shrubencroachment, a critical threat in grasslands.Shrubs may compete with perennial grasses

for soil moisture, reducing perennial grasscover and increasing the amount of exposedsoil and erosion rates (Hennessey et al. 1983;McPherson 1995, 1997; McPherson andWeltzin 2000; Schlesinger et al. 1990). Thesefactors may prevent sites from achievingupland objectives.

We also evaluated the stratification of keyareas to determine if their distribution wasadequate to meet the RMP’s upland objectiveof “80% or more of the ecological sites on theNCA.” Twelve of 13 (92%) of the possibleecological sites had key areas located in them,while nine of the 13 (69%) of the ecologicalsites had key areas that were appropriate forevaluating livestock grazing effects, i.e., keyarea located 0.8–1.6 km (0.5–1 mi) fromwater. Additional key areas that can be usedto evaluate grazing effects are needed toobtain adequate representation across ecolog-ical sites on the allotment.

We performed a power analysis for thepace frequency and point intercept methodsto determine the sampling effort needed todetect a biologically-meaningful change infrequency or cover between two points intime with a specified level of false-change (α)and missed-change (β) errors. Since bothparameters are expressed as a proportion orpercent, the analysis is similar for both. For

Table 1. A comparison of the parameters obtained from specific monitoring methods with informationneeded to address management objectives. Prior to 2004 there was no monitoring protocol in place tomeasure shrub cover, a critical threat in grasslands, in key areas.

Monitoring Protocol Grassland Variable Estimate Addresses Management Objective?

Point Cover Substrate cover including bare ground Yesand litter cover

Change in perennial grass cover/composition

Increase in exotic grass coverFrequency Combination of density and dispersion No

of plant species Line-intercept Cover Shrub cover by species Critical threat, may prevent key

area/pasture from meeting substrate cover objectives


simplicity, we begin our discussion with thesubstrate cover/point-intercept method. Thekey step in the analysis is to identify what ameaningful level of change is from a biolog-ical or management standpoint. The objectivefor bare ground cover requires that theprotocol is sensitive to changes that approachand pass the 30% value so that pro-activechanges in grazing management can be madebefore the system is severely damaged.Furthermore, in desert grasslands, perennialgrass basal cover normally ranges between10% and 30% and reaches a critical thresholdnear 5% cover; when grass cover is less thanor equal to 5%, erosion potential on manyecological sites increases dramatically andgrazing rest may not result in vegetationrecovery if soil movement and loss are severe(Hennessey et. al. 1983; P. Warren, unpubl.data). Therefore, we determined that a biolog-ically meaningful level of change was 2-5%when perennial grass basal cover was low(less than 10%), and a 7-10% change wasmeaningful when bare ground cover was 20-40% (near the BLM threshold value of 30%for bare groundcover).

Once identified, the biologically mean-ingful level of change can be substituted intothe following equation, as p1 and p2, and thenecessary sample size for detecting differencesbetween two proportions calculated (Elzingaet al. 2001):

n = (ZA + ZB)2 (p1q1 + p2q2) / (p2-p1)2

where n is the estimated necessary sample size;ZA is the Z coefficient for the false-change errorrate; ZB is the Z coefficient for the missed-change error rate; p1 is the proportion value forthe first sample, expressed as a decimal; q1 is(1 - p1); p2is the proportion value for the secondsample, expressed as a decimal; and q2 is(1 – p2).We set the acceptable level for false-change(Type I) and missed-change (Type II) errors at10%. A 20% probability for Type I and IIerrors is usually recommended for monitoringprograms; however, we decreased the proba-bility because of the regional and nationalimportance of the NCA and BLM’s manage-

ment prescription of a flexible livestockgrazing strategy.

Sample numbers necessary to detect 5%,10% and 20% changes in cover with a 10%probability of false- and missed-change errorsvary with the percent change in basal cover aswell as the original percent cover (Figure 1).Our results showed that a 20% change can bedetected with 102 points for all cover values(0% to 100%), a 10% change can be detectedwith 422 points for all cover values, and a 5%change can be detected with 981 points forvalues from 0% to 20% and 85% to 100%.Comparison of the level of change detectablewith 102, 422, and 981 sample points with thedesired level of detectable change (e.g., 2-5%for perennial grass basal cover and 10% forbare ground cover) indicates that a minimumof 981 sampling points are needed. With thissampling effort, at least a 6-7% change incover (bare ground, perennial grass or other)can be detected across the full range of initialparameters values (0%-100%). With only 100points sampled in the original protocol,sampling effort was insufficient to detect thedesired level of change; however, that effortcould detect a 20% change in cover for covervalues from 0% to 30% and 50% to 100% anda 10% change in cover for cover valuesbetween 0% and 10%.

The preceding analysis assumes that pointsalong transects are independent and that repli-cation (i.e., sample size) is at the level of thepoint. If the transect is the unit of replication,then the analysis differs (Sundt 2002) and anestimate of the pooled sample variancebetween transects (s2) is required to solve thefollowing equation:

MDC = [√(s2/n)] (ZA + ZB)where MDC is the size of the minimumdetectable change, expressed in absolute termsrather than as a percentage; and n, ZA, and ZBare as above (Elzinga et al. 2001). Our resultsindicated that 10 transects (with 100 points pertransect) were sufficient to detect the targetedchanges in parameter values identified abovebut 2 transects, as currently implemented, werenot (Gori and Schussman 2005).


Figure 1. Comparison of the number of sampling points needed to detect a 5%, 10%, and20% change in basal cover or canopy cover, assuming a 10% probability of false ormissed change errors.


As indicated, the first analysis is also appli-cable to the frequency protocol. However,without specific management objectives forplant species’ frequency, it is difficult to eval-uate if a sample size of 100 quadrats will beable to detect changes that are meaningful froma management perspective because the contextfor what management is trying to achieve withrespect to the density or dispersion of individualspecies or identification of a threshold valuethat could trigger a change in management islacking. Although an advantage of frequencysampling is its low level of observer bias, thereare several disadvantages. Since frequency is afunction of both density and dispersion,frequency data can show significant changes inpercent values where no real changes in abun-dance (or cover) exist. Therefore, frequencysampling is of limited use as an “early warning”system to detect changes in watershed condi-tion or the abundance and cover of perennialgrasses. In addition, frequency data are timeconsuming to collect because monitors musttypically recognize 30 to 40 plant species perkey area at Las Cienegas and, with a 40 cm x40 cm quadrat frame, most species are too rare(< 5% frequency) to detect a statistically signif-icant change in frequency over time (Ruyle, nodate).

UPLAND MONITORING PROTOCOL REVISIONS

Based on the above analyses, the originalprotocol for estimating substrate cover wasrevised, building on an existing NRCS-BLMprotocol (Herrick et al. 2005), and field testedin the fall of 2004. Key areas were enlarged to50 m x 100 m to accommodate the greaternumber of transects (10) and sampling points(1,000). Point-intercept measurements weremade at 0.5 m intervals along 10 transects,each 50 m in length, for a total of 1,000sampling points. Plot size was increased to 100m x 100 m and sequential point measurementswere made at 2 m intervals in loamy bottomecological sites dominated by giant sacaton(Sporobolus wrightii) to maintain independ-ence of sequential sampling points (Gori andSchussman 2005). Substrate categories were

expanded from the original protocol to include,as components of live basal vegetation, peren-nial grasses by species, annual grasses, andforbs. No species identifications were made forannual grasses and forbs. This modificationallowed us to track changes in perennial grasscover and composition over time, informationthat is important in evaluating grassland condi-tion and that can be inferred only with someuncertainty from pace-frequency measure-ments (Thurow et al. 1986, 1988a, b). Canopy(1st hit) and basal (2nd hit) cover hits wererecorded separately. Canopy cover estimatesprovide information on watershed conditionbeyond what basal cover estimates of bareground and live (basal) vegetation can providebecause grass, forb, or litter canopies canprotect bare ground (soil) beneath them fromthe erosive impacts of raindrops (Thurow et al.1988a; Pellant et al. 2000).

The revised protocol for substrate cover isequivalent (i.e., 10 transects, 1,000 points) tosubstrate cover protocols being applied at theSan Rafael Ranch and San Rafael RanchNatural Area (Arizona State Parks) and on theDiamond A Ranch in the Peloncillo Mountainsand in the San Bernardino Valley (MalpaiBorderlands Group). A similar protocol (15transects, 750 points) is also being applied atTNC’s Aravaipa Canyon Preserve and atMuleshoe Ranch Cooperative ManagementArea to measure prescribed burn effects onshrub-invaded grassland watersheds.

Because it doesn’t directly address anyupland objective, TNC recommended that thepace-frequency sampling be discontinued.However, a couple of TRT members stressedthe importance of maintaining this long-termdata set. BLM decided that the pace-frequencymeasurements will be continued at a reducedfrequency (every 3-5 years) assuming there areadequate resources to complete the otherupland monitoring.

No estimates of shrub cover were made inthe original protocol. Because this informationrelates to an important stress on grasslandsystems and will be useful in planning, prior-itizing and evaluating the success of shrubcontrol treatments, the revised protocol calls


for measuring shrub cover by species using aline-intercept method. Measurements occurredalong the same 10 transects (per key area) usedfor substrate cover monitoring but will berepeated at a frequency of every 5-6 years.

The upland monitoring plan for LasCienegas consists of the revised protocol anda protocol for estimating the similarity indexvalue of key areas (via the dry weight-rankmethod), which addresses the first generalupland objective (see above). The monitoringplan increases the total monitoring effort forkey areas annually only because BLM decidedto continue implementation of the original pacefrequency sampling (Gori and Schussman2005). That is, the time it takes to implementthe revised substrate cover protocols annuallyis essentially the same as the estimated time toimplement the original BLM annual moni-toring protocols for substrate cover, dry-weightrank, and pace frequency: approximately 2hours per key area for a 4-person crew.Sampling for shrub cover and the similarityindex both add an additional 1 hour per keyarea, but the recommended samplingfrequency is every 5-6 years and 10 years,respectively. An increase in the total numberof key area plots to better represent ecologicalsites and pastures in the allotment will alsoincrease the total monitoring effort. However,we are now considering a monitoring schedulewhere individual key areas that have metresource objectives will be sampled every 3years. This schedule should partially offset theincreased time costs of continuing the pace-frequency sampling and sampling additionalkey areas. Thus, implementation of the revisedprotocol can be accomplished with existingresources.

Additional details on the revised substratecover protocol, evaluation of the similarityindex protocol, and their time costs can befound in Gori and Schussman (2005).

THE ADAPTIVE MANAGEMENTPROCESS

The revised upland monitoring plan wasreviewed by the TRT and RRT as well asoutside experts in 2004. In addition, a work-

shop for these teams was held in fall 2005 tocompare the results of the original and revisedprotocols for two adjacent key-area sites. Theparticipants confirmed that the protocols gavecomparable results for substrate cover andplant species’ composition.

Following completion of the protocolreview in 2004, the monitoring plan was field-tested by TNC and BLM staff in Septemberand October, 2004. Monitoring was repeatedat the same time of year in 2005. In both 2004and 2005, key areas were prioritized andselected for measurement based on thepermittee’s proposed annual grazing plan toensure that information collected would begermane to decision-making. Twenty-four keyareas were measured in 2004 and 22 in 2005.The monitoring data was analyzed andpresented to the TRT, RRT, and interestedpublic during Biological Planning in fall 2004and 2005. Monitoring information included:(1) bare ground cover, basal and canopy coverof perennial grasses, shrub canopy cover, anddominant perennial grass and shrub species forall key areas; (2) a proposed livestock grazingplan for fall, winter, spring and summer, 2004-2005 and 2005-2006, including stocking rateand pasture rotation schedule; and (3) acomparison of bare ground and perennial grasscover with summer precipitation records bykey area from 2001-2005.

In 2004, the TRT’s review of the proposedgrazing plan and monitoring data revealedconcern over two northern pastures withproposed winter and spring use in 2004-2005.The concern was due to low basal cover ofperennial grasses in 2004 (< 5% basal cover),a significant downward trend in live basalvegetation cover between 1995 and 2004(R2s > 0.48, p < 0.025), and relatively highshrub cover in two key areas in the twopastures (shrub cover > 27.4%, n = 2 keyareas). Based on these concerns, the permitteerevised his grazing plan and decreased by 2months (50 %) the total amount of time thatlivestock would be grazing in these pastures.Specifically, the 300 head of livestockproposed to enter the two pastures onDecember 1, 2004, were held on sacaton


pastures until February 1, 2005, in an effort todecrease the overall time livestock weregrazing these areas. In addition, the permitteeproposed using water sources strategically,keeping livestock on the sacaton portions ofthese northern pastures and away from moresensitive upland areas with lower perennialgrass cover.

These two northern pastures received poorsummer rainfall again in 2005, as did an adja-cent pasture that also showed low perennialgrass basal cover in 2005 (basal cover = 4%)and a declining trend in live basal vegetationcover since 1995 (R2 = 0.94, 4df, p < 0.01).The latter pasture was proposed for winter-spring use in 2005-2006, but all three pastureswere excluded from grazing in 2005-2006.Instead, the permittee kept livestock on nearbysacaton pastures with suitable forage duringthis period.

Another pasture, West Pasture, failed tomeet bare ground objectives in fall 2004. TheTRT and RRT recommended that livestock beremoved immediately and the pasture rested.Summer rainfall for that pasture in 2005 waswell above average and by fall 2005, perennialgrass basal cover had increased from 12 to19% (a 58% increase) and bare ground coverhad decreased from 33 to 24% (27%); bothchanges were statistically significant (Fisherexact tests: perennial grass, p < 0.001; bareground, p < 0.001). As a result of this recovery,the pasture was scheduled for 3 months ofwinter use in 2006. Although changes weremade in pasture use, proposed stocking rateswere not adjusted in 2004 or 2005.

The RRT concurred with the TRT’s recom-mendation following careful review of themonitoring data analysis in both years. TheBLM also concurred with these recommenda-tions and the proposed grazing plans weremodified accordingly.

Overall, there was no relationship betweenincreases or decreases in basal grass cover orbare ground cover on key areas between fall2004 and fall 2005 and whether a pasture wasgrazed in the intervening months, though thisanalysis did not account for the intensity of use(Fisher exact tests: perennial grass, p = 0.61;

bare ground, p = 0.56). The change in peren-nial grass basal cover between 2004 and 2005was positively related to the amount ofsummer rainfall key areas received in 2005,however, there was no relationship betweenbare ground cover and rainfall (bare ground:R2 = 0.01, 16 df, p > 0.65; perennial grass: R2

= 0.22, 16 df, p = 0.05). That is, key areas thatshowed greater increases in perennial grasscover between the 2 years received moresummer rainfall in 2005 than did plots showingsmaller increases or decreases in perennialgrass basal cover between 2004 and 2005.Similarly, the change in perennial grass basalcover on key areas between 2004 and 2005was positively related to the summer rainfalldeviation in 2005 but the change in bareground cover was not (perennial grass: R2 =0.32, 16 df, p = 0.014; bare ground: R2 = 0.03,16 df, p > 0.5); rainfall deviations were calcu-lated as the difference between the amount ofsummer rainfall a key area received in 2005and its mean summer rainfall calculated fromgauge records starting in 2001 or earlier.Another precipitation variable, the differencein summer rainfall that key areas received in2004 vs. 2005 was unrelated to changes in bareground and perennial grass basal cover. Addi-tional factors besides summer rainfall (e.g.,differential effects of the ongoing drought onperennial grass species, degree of shrubencroachment, and intensity of livestock use)may also be contributing to changes in grass-land condition on key areas in 2005.

CONCLUSIONSThe BLM, permittee, TRT, and RRT success-fully completed steps 1 through 6 of theadaptive management process. In addition, theTRT and RRT, with BLM concurrence, iden-tified the following information gaps or needsthat would assist in future decision-making(i.e., Step 7): • Identify biologically important thresholds

for basal cover of perennial grasses byecological site, including thresholds thatmay trigger a change in grazing manage-ment; and


• Establish additional key areas in northernpastures and on ecological sites that areunder-sampled on the allotment to assist inthe evaluation of livestock grazing andclimatic factors in these pastures. This willbe addressed through the addition of pairedkey-area plots and livestock exclosures onsouth-facing slopes. The TRT and RRT also identified some

refinements needed in the RMP’s resourceobjectives, as well as additional planning toimprove BLM’s ability to manage the diverseexpression of desert grassland on different soiltypes. These refinements include:• Define desired objectives for basal cover of

perennial grasses by ecological site; and• Develop a shrub control plan for the

Empire-Cienega Allotment, includingfuture prescribed burns to reduce shrubcover in affected areas.In February, 2004, the TRT selected five

sites for additional key areas and paired adja-cent livestock exclosures in northern pasturesand fencing was completed in spring of 2006.Two of these pairs were on ecological sites thatpreviously lacked key areas appropriate forevaluating livestock effects, increasing the totalto 11 of 13 ecological sites (85%) on theEmpire Cienega Allotment with key areas. Inaddition, the TRT will identify desired objec-tives and biologically important thresholds forbasal cover of perennial grasses by ecologicalsite by fall 2006.

The collaborative adaptive managementprocess at Las Cienegas NCA has had anumber of expected and unexpected benefitsthat ultimately contribute to BLM’s and thepermittee’s success in managing the grassland-watershed. Specifically, Biological Planningwith its reliance on interactions between BLM,the permittee, multi-stakeholder teams, and theinterested public has:• Built trust among participants and devel-

oped a solution-oriented approach toaddress potential conflicts over grazing andother resource concerns;

• Fostered an environment where decision-making is based on information instead ofemotions;

• Increased knowledge of grazing effects toimprove resource management;

• Provided access to different perspectivesand expertise that has given participants theability to go beyond measurements of vege-tation change to consider wildlife use andneeds; and

• Demonstrated that rigorous monitoring canbe implemented without increasing theoverall cost of monitoring.Additional details on the science-based

adaptive management process at Las CienegasNCA can be found in Gori and Schussman(2005), BLM (2006), and Bodner et al. (2007).

ACKNOWLEDGMENTSThe authors would like to thank R. Marshall,G. McPherson, D. Robinett, E. Carrillo, K.Hughes, G. Drennan, and P. Heilman forsharing their knowledge and insights with usand for providing comments on earlier reportsand drafts of this manuscript.

LITERATURE CITEDBodner, G., J. Simms, and D. Gori. 2007. State of

the Las Cienegas National Conservation Area:Gila topminnow population status and trends1989-2005. Report by The Nature Conservancy.50 pp. (can be downloaded from www.azconser-vation.org)

Bureau of Land Management (BLM). 2003.Approved Las Cienegas Resource ManagementPlan and Record of Decision. U.S. Department ofInterior, Bureau of Land Management, ArizonaField Office.

Bureau of Land Management (BLM). 2006. Reporton the Biological Planning process for livestockmanagement at Las Cienegas NCA: Fall 2005upland monitoring results and adaptive manage-ment by the Bureau of Land Management,Technical Resource Team, and RangelandResource Team. 29 pp.(can be downloaded fromwww.azconservation.org)

Elzinga, C. L., D. W. Salzer, J. W. Willoughby, andJ. P. Gibbs. 2001. Monitoring plant and animalpopulations. Blackwell Science, Malden, MA.

Enquist, C. A. F., and D. F. Gori. 2005. An assess-ment of the spatial extent and condition of


grasslands in the Apache Highlands Ecoregion. InConnecting mountain islands and desert seas:Biodiversity and management of the MadreanArchipelago II. 2004 May 11-15; Tucson, AZ,compiled by G. J. Gottfried, B. S. Gebow, L. G.Eskew and C. B. Edminster. Proceedings RMRS-P-36. Department of Agriculture, Forest Service,Rocky Mountain Research Station, Fort Collins,CO.

Gori, D. F., and H. Schussman. 2005. State of theLas Cienegas National Conservation Area. Part I.Condition and trend of the desert grassland andwatershed. Report by The Nature Conservancy.63 pp. (can be downloaded from www.azconser-vation.org).

Hennessey, J. T., R. P. Gibbens, J. M. Tromble, andM. Cardenas. 1983. Vegetation changes from 1935to 1980 in mesquite dunelands and former grass-lands of southern New Mexico. Journal of RangeManagement 36: 370-374.

Herrick, J. E., J. W. Van Zee, K. M. Havstad, L. M.Burkett, and W. G. Whitford, with contributionsfrom B. T. Bestelmeyer, A. Melgoza, M. Pellant,D. A. Pyke, M. D. Remmenga, P. L. Shaver, A. G.de Soyza, A. J. Tugel, and R. S. Unnasch. 2005.Monitoring manual for grassland, shrubland andsavanna ecosystems. Volume II: Design, supple-mentary methods and interpretation. USDA-ARSJornada Experimental Range, Las Cruces, NM.

McPherson, G. R. 1995. The role of fire in desertgrasslands. In The desert grassland, edited by M.P. McLaren and T. R. Van Devender, pp. 130-151.University of Arizona Press, Tucson, AZ.

McPherson, G. R. 1997. Ecology and managementof North American savannas. University ofArizona Press, Tucson, AZ.

McPherson, G. R., and J. F. Weltzin. 2000. Distur-bance and climate change in the UnitedStates/Mexico borderland plant communities: Astate-of-the-knowledge review. U.S.D.A. ForestService General Technical Report RMRS-GTR-50. Rocky Mountain Research and ExperimentStation, Fort Collins, CO.

Marshall, R. M., D. Turner, A. Gondor, D. Gori, C.Enquist, G. Luna, R. Paredes Aguilar, S.Anderson, S. Schwartz, C. Watts, E. Lopez, and

P. Comer. 2004. An ecological analysis of conser-vation priorities in the Apache Highlandsecoregion. Report by The Nature Conservancy inArizona, Instituto del Medio Ambiente y Desar-rollo Sustentable del Estado de Sonora, agencyand institutional partners. 152 pp. (can be down-loaded from www.azconservation.org).

Pellant, M., P. Shaver, D. A. Pyke, and J. E. Herrick.2000. Interpreting indicators of rangeland health,Version 3. Technical Reference 1734-6. U.S.Department of Interior, Bureau of Land Manage-ment, Denver, CO.

Ruyle, G. B. No date. Some methods for monitoringrangelands and other natural area vegetation.Extension Report 9043. Division of RangeManagement, University of Arizona, College ofAgriculture, Tucson, AZ.

Schlesinger, W. H., J. F. Reynolds, G. L.Cunningham, L. F. Huenneke, W. M. Jarrell, R.A. Virginia, and W. G. Whitford. 1990. Biologicalfeedbacks in global desertification. Science 247:1043–1048.

Sundt, P. 2002. The statistical power of rangelandmonitoring data. Rangelands 24: 16–20.

Thurow, T. L., W. H. Blackburn, and C. A. Taylor,Jr. 1986. Hydrologic characteristics of vegetationtypes as affected by livestock grazing systems,Edwards Plateau, Texas. Journal of RangeManagement 39: 505–509.

Thurow, T. L., W. H. Blackburn, and C. A. Taylor,Jr. 1988a. Infiltration and interril erosion responsesto selected livestock grazing strategies, EdwardsPlateau, Texas. Journal of Range Management 41:296–302.

Thurow, T. L., W. H. Blackburn, and C. A. Taylor,Jr. 1988b. Some vegetation responses to selectedlivestock grazing strategies, Edwards Plateau,Texas. Journal of Range Management 41:108–114.

U.S. Department of Agriculture, Natural ResourcesConservation Service (USDA). 2009. EcologicalSite Information System. http://esis.sc.egov.usda.gov/.

Zar, J. H. 1984. Biostatistical analysis. Prentice-Hall. Englewood Cliffs, NJ.

ECOLOGY AND CONSERVATION IN THE SONOYTA VALLEY, ARIZONA AND SONORAPhillip C. Rosen, Cristina Melendez, J. Daren Riedle, Ami C. Pate, and Erin Fernandez

Lowland desert rivers, streams, and ciénegasare the most severely impacted environmentaltype in the southwestern United States (Bryan1928; Hendrickson and Minckley 1985;Minckley and Deacon 1991; Logan 2002;Turner et al. 2003) and northwestern Mexico(Hendrickson et al. 1980; Hendrickson 1983;Miller et al. 2005). Huge areas of riparianforest have been lost in the low deserts(Ohmart 1982); most southwestern fishes inthe U.S. have long been recognized to beendangered, threatened, or sharply declining(Miller 1961; Minckley 1973; Minckley andDeacon 1991); and many of the aquaticamphibians and reptiles are following suit(e.g., Rosen and Schwalbe 2002).

Río Sonoyta, Sonora, Mexico (Figure 1),which adjoins Organ Pipe Cactus NationalMonument and passes through the town ofSonoyta, Sonora, is a rare lowland desert streamthat has not been completely degraded anddesiccated, although the regionwide down-cutting of ciénegas (Hendrickson and Minckley1985) that began about 1891 has severelyimpacted the system at its upstream end. Itretains perennial surface flow in a local area,which is protected within the Reserva de laBiosfera El Pinacate y Gran Desierto de Altar,and still supports its originally known native fishfauna (Snyder 1915; Miller et al. 2005). Nativeturtles and diverse, productive riparian forestsand woodlands remain widespread despiteserious environmental degradation during thepast approximately 118 years.

Discussions regarding the status of theendemic Sonoyta mud turtle (Kinosternonsonoriense longifemorale) were convened by

U.S. Fish and Wildlife Service (USFWS) in1997. In 2001 a multi-agency conservationteam consisting of representatives fromUSFWS, Instituto del Medio Ambiente y elDesarrollo Sustentable del Estado de Sonora(IMADES, now CEDES), Pinacate BiosphereReserve, Organ Pipe Cactus National Monu-ment (ORPI), Arizona Game and FishDepartment (AGFD) and University ofArizona (UA) began conservation planningand initiated ecological monitoring at Quito-baquito, Quitovac, and the Río Sonoyta proper.As a result of small population sizes and otherpotential problems facing all populations ofthis turtle, which are detailed here, USFWSidentified the Sonoyta mud turtle as a candi-date for listing as threatened under theEndangered Species Act of the United States(Knowles et al. 2002a & b).

Sampling to determine the distribution,abundance, and status of the Sonoyta mudturtle, which has been conducted on an annualor bi-annual basis from 2001 through 2006,quickly led to the realization that a broaderresource was at risk and to an expansion oftaxonomic focus and conservation objectives.In this chapter we offer details on the status ofand threats to the natural biodiversity resourcesof Río Sonoyta based primarily on publishedliterature and our work there during 2001-2006. We also outline ideas on what might bedone to reconcile the conservation of theseresources with present and future human uses.

METHODS AND MATERIALSSurveys were conducted once to twice per yearduring 2001 through 2006, generally during

144 Rosen, Melendez, Riedle, Pate, and Fernandez

March-November, and mostly in April orOctober. We selected sites to visit based onmaps, aerial images, museum records, andinterviews. We established two intensive studyareas, one at and adjoining Presa Xochimilcoon the east margin of Sonoyta and one at alower section of perennial stream at theRancho Agua Dulce pers. comm. or“Papalote”) reach, in the Pinacate Reservesouth of Quitobaquito. We recorded riparianand aquatic conditions photographically andaccording to annotated lists of dominant plantsand notable bird species, and we sampled forfishes using minnow traps, dipnets, hoop nets,

and seines. Trap locations and other places ofinterest were recorded with GPS units and withdescriptive notes. Fishes and invertebratescaptured in the traps were recorded. Detailednotes on ecological observations were main-tained in an itinerary-based notebook, whileturtle data were recorded on data sheets.

Potential habitat for the Sonoyta mud turtlein Sonora was surveyed using visual surveysand trapping. Visual surveys involvedapproaching water sources quietly andsearching with binoculars for basking andsurfacing turtles, looking for tracks in moistsand or mud, and muddling (probing by hand

Figure 1. Location map showing Río Sonoyta, Quitovac, and low-elevation cities in theSonoran Desert. All of the cities except Sonoyta and Yuma have dried up their lowlandstreams or rivers, and native fishes are absent or nearly so, except in Río Sonoyta.

Rosen, Melendez, Riedle, Pate, and Fernandez 145

in burrows and undercut banks). Trappingemployed 46 and 76 cm (18 and 30 in) diam-eter, 2.54 cm (1 in) mesh hoop nets baited withsardines and hot dogs for juvenile and adultturtles, and galvanized steel 6 mm (1/4 in)mesh minnow traps with openings enlarged to5-8 cm (2-3 in) and baited with a piece ofsardine for hatchling and juvenile turtles. Trapswere set for 2-6 hours in the immediate areaadjoining Sonoyta and for approximately 24hours per checking cycle elsewhere.

Turtle study in the region by one of us(Rosen) began at Quitobaquito, in Organ PipeCactus National Monument in 1983 (Rosenand Lowe 1996), with occasional visits to RíoSonoyta starting the same year. Capturedturtles were marked for individual recognitionby notching marginal scutes on the carapace,using a nick on a bridge marginal to signify thegeneral region — 5R for the study area at PresaXochimilco and adjoining river reaches inSonoyta, 5L for the Agua Dulce region, and nobridge mark for Quitobaquito. No movementamong sites has ever been detected. Althoughwe observed turtles at Quitovac on three trips,no turtles were marked there. Various shellmeasurements were taken on captured turtles,and growth rings were either measured toquantify yearly growth or counted backwardto the year of hatching to determine age.

A principal threat identified during thisstudy was groundwater depletion and loss ofaquatic habitat. Sonoyta is presumed to havehad a ciénega, and Río Sonoyta is thought tohave originally had long-flowing reaches, andwe therefore reviewed available historicdescriptions of water in the Sonoyta Valley tomore carefully evaluate former conditions. Wecontrasted our evaluation of former conditionswith our field observations to assess hydro-logical changes and consequent threat levelsof aquatic vertebrate species. We utilized thesethreat evaluations and discussions with peoplein the Sonoyta region as a basis for conserva-tion recommendations.

STUDY REGION DESCRIPTIONBackground details on current environmentalconditions along Río Sonoyta, current place

names, and conditions are provided in thissection. Historical literature was reviewed fordescriptions of original aquatic ecosystemconditions which allowed an assessment ofchanges in the Sonoyta Valley during the pastcentury and a half. The reader should refer toBennett and Kunzmann (1989), Fisher (1989),Felger et al. (1992), Nabhan et al. (1982,2000), and Weir and Azary (2001) for descrip-tions of the oases at Quitobaquito andQuitovac, and for historical references forthese largely intact spring systems.

Riparian and Aquatic Environments in 2001-2006

Hydrology and Extent of Water Although the hydrology of Río Sonoyta basinis not fully understood, it appears that presentlyexisting surface waters (Figure 2A) are closelyassociated with shallow rock which forces theaquifer to the surface, rather than imperviousfine sediment layers — although MacDougal(1908) and others have suggested the latter. In2001 and 2002 we found two areas, each withabout 6-10 semi-perennial river pools withcatfish and turtles, in reaches without stream-flow adjoining hills just east of Sonoyta(“Vidrios” and “San Raphael” on topographicmaps). Some of these pools were apparentlyperched on lenses of clay, as they were in areaswithout evidence of submerged bedrock andsat above 2 m (6.5 ft) drop-offs in thestreambed. Below the drop-offs there was nosurface water. We do not know if these poolscontinued to support aquatic vertebrates duringand after the severe droughts of 2002-2004.Such perched aquifers were not evident indownstream reaches that currently supportperennial water.

At Sonoyta, the only apparently naturalperennial water in 2001-6 adjoined the southtip of the Sonoyta Hills, where bedrock isexposed in the river channel. At the head ofthis spring is Presa Xochimilco (Figure 3);which is apparently the same as “PresaDerivadora” of Broyles and Felger, in Felger(2000). The Presa occupies the site of thehistoric laguna (Hoy 1990), which wasretained by a large stone dam, possibly the one


Figure 2. Hydrological conditions at Río Sonoyta. Perennial sections are indicated byblack line outlined in white, with ciénega near Sonoyta indicated by marsh symbols.Figure 2A (top) shows conditions observed during current study, with lower end of flowshown as variable depending on year and season. Figure 2B (bottom) shows inferredoriginal conditions (see text) with uncertainty suggested by narrowing white outlining.


still extant, placed on this bedrock. The damserved as a headgate for acequia water(Hornaday 1908) prior to the large-scale devel-opment of electric and gasoline-powered wellsand the desiccation of the local aquifer. Earlyin our study, runoff maintained a pond wellover 0.5 km long, which occupied the mappedlaguna area. During most of our study peren-nial water occurred only as sewage effluentfrom an army base at the site of old Sonoytareleased in the upper 0.3 km (0.2 mi) of themapped laguna. Except in the kilometer (0.6mi) above the dam, the river bottom nearSonoyta is entrenched within vertical walls upto 8 m (26.2 ft) deep.

The perennial spring below the dam wasflowing for several hundred meters in 2001,but was reduced in 2005-6 to approximately

10 pools in the kilometer below the dam. Thepools ranged from 0.3 to > 1.5 m (1-5 ft) deep,and some received small inputs of sewageeffluent from individual homesteads atop thesouth arroyo wall. Lacking garbage collectionservice, many of these homesteads alsodisposed of copious domestic trash in thearroyo.

The wettest part of the riverbed we wereable to locate between Sonoyta and AguaDulce was at Ejido Josefa Ortiz, just upstreamfrom the historic site of Santo Domingo. Thisnon-perennial spring in the riverbed occurredat the river’s narrow point between tworhyolitic hills. It was reported to flow for about1 km (0.6 mi) during winter, when saltcedar(Tamarix ramosissima) was leafless and there-fore not transpiring, but we found it no more

Figure 3. Presa Xochimilco at the east edge of present-day Sonoyta. The photo wastaken in 2001 from the dam, looking east. Photo by Rafaela Paredes.


than 300 m (328 yd) in length on 8-9 March2002. A field of salt grass (Distichlis spicata)adjoins the spring source outside the ripariansaltcedar thicket. This spring stands at the headof the irrigation developments of SantoDomingo, most of which was not in use duringour surveys.

Downstream from Sonoyta, to SantoDomingo, we found decreasingly active use ofthe irrigation systems. From Santo Domingoto downstream reaches associated with AguaDulce and Agua Salada, most irrigationsystems were in disuse. Further, a number ofthose closest to the river lacked evidence ofmodern infrastructure and were overgrownwith salt-tolerant desertscrub, especially desertsaltbush (Atriplex polycarpa) and broom seep-weed (Sueda moquinii), and thus appeared tohave been long-abandoned.

The Agua Dulce reach (Figure 4) as definedhere is centered about 1.6 km (1 mi) south of

Quitobaquito and Aguajita springs, near currentRancho Agua Dulce, rather than at a site 6.5 km(4 mi) southwest of Quitobaquito as stated insome earlier literature (e.g., Schott, in Emory1857; Hoy 1990), which is close to the site weknow as Agua Salada or La Salada. This peren-nial reach occurs at the next major rockychoke-point after Santo Domingo. There is iden-tical-looking decomposing granite to the north(Quitobaquito Hills) and south (Cerro de TresVerredos, an outlier of the Sierra los Tanques).It appears plausible that the river course ispushed to the south against the flank of Sierralos Tanques by the large bajada of the PuertoBlanco Mountains and water is forced to thesurface by subterranean granitic rock.

Goodman (1992) reported deep alluviabetween Sonoyta and Santo Domingo, atwhich point the valley is thought to be shal-lowly underlain by the Aguajita Springs BiotiteGranite, of late Cretaceous age. Further details

Figure 4. The Agua Dulce, or “Papalote” reach of Río Sonoyta in April 2003. Photo by P.C. Rosen.


are poorly known or unknown, althoughCaruth (1996) reported that QuitobaquitoSprings depend on water shed by the PuertoBlanco Mountains, which requires 500 toseveral thousand years to pass through bajadasediments until encountering subsurface rockwhich forces it up to emerge as springs.Goodman also found that the water of thesprings along Quitobaquito Hills are all iden-tical in chemical composition, but differ fromthat at Agua Dulce.

Until further information becomes avail-able, the source and fate of the Agua Dulcereach are unknown and not preciselypredictable. However, this last remaining free-flowing reach of the river is likely the productof the aquifer between Sonoyta and SantoDomingo and its attenuating continuation overthe subsurface Aguajita Springs granite. Eventhough water may be in the ground for manycenturies prior to emerging as spring flow, adepleted aquifer may reverse the upwardmovement of the groundwater, and weaken orkill the surface flow. Continued pumping forirrigation could desiccate Agua Dulce, and thecloser to the perennial reach it occurs, the morelikely that strong or immediate impacts will beseen.

The Agua Dulce reach fluctuates markedlyin extent. During our current period of study,it usually varied from 1.6-2.7 linear km (1-1.7mi) of surface water from a consistent sourceto a variable site of disappearance beneath thesandy riverbed. During the wet period of 1977-1984, and shortly thereafter, it was reported tobe “at least 5-6 km (3-3.7 mi)” in extent byMiller and Fuiman (1987, p. 606), whosediscussion implies that fish were found over alength of 10-13 km (6.2-8.1 mi) of river during5-7 May 1986 (following a wet winter).

We found the Agua Dulce reach character-ized by shallow runs of 3-10 cm (1.2-3.9 in)deep water over sand and fine gravel, withdeeper — 10-30-cm (3.9-11.8 in) — sandyriffles and glides. There were local sandy tomud-bottomed pools that were mostly < 0.6 m(2 ft) deep prior to 2005. During 2005-6,several pools had maximum depths of 0.8-1.1m (2.6 -3.6 ft). During the driest parts of our

study, surface flow disappeared from most ofthe Agua Dulce reach, which was reduced topools supported by subsurface flow, particu-larly in June and July from 2002-2005.

At Agua Salada in 2001, we successfullydug to water at a depth of about 0.6 m (2 ft).

Riparian Forest, Woodland, and Bank Vegetation We observed riparian forest and woodlandalong Río Sonoyta over all of the region weinvestigated, from Colonia La Nariz down-stream to the northeast side of the Pinacate lavaregion at Los Vidrios Viejos. Upper reacheswere often characterized by a dominance ofvelvet mesquite (Prosopis velutina) bosque.We surveyed this region only briefly in 2001.

From Presa Xochimilco to at least 4.5 km(2.8 mi) upstream, we found a structurallydiverse gallery forest composed primarily ofvelvet mesquite, Goodding willow (Salixgooddingii), and saltcedar mixed in relativelyeven abundances. This forest also contained avariety of other small desert trees and shrubs.Downstream of the presa there was a denseforest of saltcedar mixed with lesser represen-tations of mesquite, willow, and local patchesof giant reed (Arundo donax), bulrush (Scirpesamericanus), and southern cattail (Typhadomingensis). Throughout these reachesBermuda grass (Cynodon dactylon) was thedominant groundcover.

Drier reaches below this area — i.e., belowthe Mexican Highway 2 bridge in Sonoyta —usually supported more open stands of velvetmesquite and saltcedar. Saltcedar formed densethickets at Ejido Josefa Ortiz and a substantialpart of Santo Domingo, at Agua Dulce, and atAgua Salada. Except for the Agua Dulceregion, each of these areas was only examinedbriefly 1-3 times, and our survey of the riverdid not include 100% coverage of all reaches.

The Agua Dulce reach supported small,local stands of willow within the saltcedarthicket in at least two places. It had locallyextensive stands of arrowweed (Plucheasericea) and marginal bosque outside thesaltcedar thickets, comprising primarily velvetmesquite and saltcedar. Despite heavy grazing


by cattle, stream margins were stabilized bysalt grass, Bermuda grass, spike rush(Eleocharis sp.), bulrush, and sedges (Cyperussp.) in numerous areas free of saltcedar shade,and there were small patches of cattails.However, saltcedar provided the main stabi-lization of the flow channel and much of themost resilient stabilization of streambanksproper. Floristic information for parts of RíoSonoyta is available in Felger (2000).Aquatic and Riparian Fauna Dense accumulations of filamentous algae(e.g., Cladophora sp.) were found in thestream when flood scour was absent. A richinvertebrate fauna and numerous tiny juvenilefishes were found in these conditions despiteextremely high densities of adult fishes, espe-cially in extensive areas of shallow water,particularly in the stream margins.

One endemic invertebrate, the Quitobaquitospring snail, Tryonia quitobaquitae, has alsobeen described from the Río Sonoyta basin. Itbelongs to the family Hydrobiidae, which hasnumerous endemic populations found inisolated springs, seeps, and drainagesthroughout the southwestern United States.The snail was first collected at QuitobaquitoSpring in 1963 and was assigned to T. imitator,a more widespread species. Hershler andLandye (1988) described it as a separatespecies based on morphometric analysis ofshell characteristics and soft body tissue.Tryonia quitobaquitae has been found atQuitobaquito Spring, Burro Spring, andWilliams Spring approximately 2-3 km (1.2 –1.9 mi) away, all along the south and westflanks of Quitobaquito Hills.

The aquatic vertebrates in the SonoytaValley are the Sonoyta mud turtle, the Arizonamud turtle (K. arizonense, which was collectedonce at Quitobaquito (Smith and Hensley1957) and for which we have recently receiveda photographic record from Presa Xochimilco(A. Pate, pers. comm. 2007), anuran amphib-ians of at least six species (Rosen 2007), alltoads, and five species of fishes, two of themnative. The status of the amphibians is poorlyknown, but two species of interest deserve

mention, the Great Plains narrow-mouthedtoad (Gastrophryne olivacea) and the Sonorangreen toad (Bufo retiformis), both of which,like the Arizona mud turtle, have regionaldistributions centered on the Tohono O’odhamlands in south-central Arizona (Sullivan et al.1996; Rosen and Funicelli 2008). We have notseen tadpoles or amphibians breeding in anyof the perennial waters of the region.

At least 157 species of birds are knownfrom the Sonoyta region, including 47breeding species and 70 migratory species thatwinter in the region (Russell and Monson1998). We found riparian birds to be abundantand diverse especially at and above PresaXochimilco. Migratory and riparian birds werealso prominent at the Agua Dulce reach, aswere birds characteristic of aquatic habitats,including belted kingfisher, green heron, andblack-crowned night heron.

Río Sonoyta appears to be preferentiallyutilized by some bats, such as Underwood’smastiff bat (Eumops underwoodi) (Barns andPate 2004), that would likely be regionallyabsent without the river’s resources. Similarly,a number of medium-sized mammal specieshave been recorded at ORPI that likely dependon the river for regional persistence, such as theraccoon, striped skunk, and hognose skunk(Cockrum and Petryszyn 1986). Pronghorn andmule deer are occasionally observed utilizingmarginal mesquite and saltcedar woodland inthe lowermost parts of the area we examined(Israel Barba, pers. comm., 2001-2).

Original ConditionsHistorical Review Hoy (1990) provides a summary statementabout ciénega and stream conditions inSonoyta Valley based on reports from Amer-ican and some earlier Mexican sources. Herewe examine this record in detail to determinewhat can definitely be said about originalconditions — those that existed at and shortlyafter the arrival of Europeans in the region —and habitat losses in the river environment.

Padre Eusebio Kino, who first arrived atSonoyta in 1698, is vague about ecologicalconditions at Sonoyta, but noted, upon his


arrival, a Tohono O’odham community withextensive irrigation agriculture already fullydeveloped. He described about 1,000 peopleat Sonoyta, and a considerable number morein surrounding communities (Bolton 1936). Asthe only perennial stream between Arivaca andthe Colorado River, Río Sonoyta was a hub ofthe Tohono O’odham culture.

The boundary survey party headed byEmory (1857) and the railroad route survey ofAndrew B. Gray (1856) and Peter R. Bradyreported briefly on the environment atSonoyta. Gray was there in May 1854, andreported (p. 87):The valley is broad, with springs and a small stream(the Sonoita) which flows for a few miles in the drymonths, when it sinks…

This suggests that in times of average orlow rainfall the river was not as large as mightbe surmised from other descriptions, such asin Bryan (1925, map based on a 1917 survey).

Arthur Schott, who was apparently in theSonoyta region during 1853-5, reported hydro-logical details in Part II, Chapter IV of Emory(1857, pp. 73-75). He describes a considerablyless extensive river scene, after first confirmingthat a ciénega-like environment was found atSonoyta:Besides numerous deep charcos and even smalllagoons in its lower part, this cienega is blessed witha small stream fed at its outset by a number of smallsprings [which] afford a constant flow of water …[which] is clear, of a bluish hue, but warm andslightly brackish. Notwithstanding this perennialsupply, the little river of Sonoyta continues but abouta mile as a running stream.

This probably cannot be entirely attributed tointensive diversions for irrigation at the time,since Schott says the inhabitants of Sonoytaonly “irrigate a small patch of ground.”However, at this time Sonoyta may havereferred to the Mexican town, rather than theTohono O’odham community near the westend of present day Sonoyta. Downstream,Schott reported similar conditions:Following the bed of the Sonoyta river, a narrow butsmooth pass leads to another cienega … but the

water, except in two or three places does not come tothe surface, and it is necessary to dig for it every-where during the dry season. … Upon some risingground in the west end of the last-mentioned cienegathere is a settlement, or, more properly, cattle rancho,the inhabitants of which are favored with spring waterflowing out in abundance from a dozen little springs.… This water resembles … that of Sonoyta …

Since Schott then immediately goes on tomention “Quitobaquita” separately as heproceeds westward in his description, the “last-mentioned cienega” would almost certainlyrefer to Santo Domingo. He continues, clearlyreferring next to the Agua Dulce (Papalote)reach, and lastly Agua Salada, in order:The water of the Rio Sonoyta appears above groundfor the last time near Quitobaquita. On the southeastside of the Cerros de la Salada fresh palatable watercan be got in its bed by digging to a depth of aboutthree feet. Just below it, it becomes so salty that evenfamishing mules will not touch it.

Edgar Mearns (1907) reported detailedbiological observations at Sonoyta and SantoDomingo from 9-25 January 1894, and then atQuitobaquito from 25 January to 8 Februarybefore proceeding west by way of old AguaDulce. He camped at Sonoyta, not long after aperiod of heavy rains which had recentlywashed out the marshy ciénega, and describeda “pretty creek containing several species offishes.” (McMahon and Miller [1985] reportedthat pupfish and dace were in the Río Sonoytaat Sonoyta at least until 1950.) He reported thatthe river rose south of monument 164, whichis near present-day Ejido Desierto de Sonora,10 km (6.2 mi) ESE of present-day Sonoyta,and flowed about 40 km (24.8 mi). In 2001,we found a couple of small pools of water inthe river bed near heavily agricultural EjidoDesierto de Sonora.

Mearns found that recent floods had largelywrecked irrigation agriculture at Sonoyta;fields that remained were a few that “receivedtheir water from springs at some distance[north] from the river” — evidence of multiplespring sources characteristic of ciénega condi-tions. Although old Sonoyta was small, theTohono O’odham village at or just west of thesite of present day Sonoyta was not:


Below the village of Sonoyta is a village of PapagoIndians, who successfully irrigate large fields fromthe Sonoyta River. … A few miles further downstreamis the Mexican town of Santo Domingo, distant about14.5 km (9 miles) from Sonoyta.

Ives (1936) describes a ciénega-like forma-tion at the east edge of Sonoyta that lasted untilflood scour caused an incision, reportedly(Lumholtz 1912 [1990]) on 6 August 1891: Just east of the town is a meadow, which is beingchanneled at the present time. Until about 1890 thismeadow was a swamp, but in that year, a barrier ofcalcareous mineral-spring deposit was washed out,draining the swamp and lowering the localwatertable.

Lumholtz was told that the ciénega hadextended 4.8 km (3 mi) east from Sonoyta, andthat “The swamps dried up in 3 years … .Where there had been before only a llano [atreeless plain, likely a wet meadow], a forestof mezquites sprang up.” Lumholtz reportedthat abundant perennial surface water couldalways be found to Agua Dulce, which is atvariance with Schott’s and our information,and apparently reflects misinterpretationsstemming from the large changes in flow ofRío Sonoyta that track rainfall trends. Ives alsonotes (1936) that only on rare occasions isthere any continuous surface flow betweenSonoyta and Agua Dulce.

InterpretationThere is consistent evidence that the site ofSonoyta had multiple springs that would haveproduced an extensive marshy meadow withstream flow, and this would certainly have alsoincluded scour pools associated with the mainpaths of storm runoff (Figure 2B). Thus, refer-ences to a sizable ciénega are justified. Alreadyby the time of the first written descriptionsdiscussed here, the scene had long since beentransformed by irrigation agriculture, andprimeval habitat may have differed in extentfrom the original conditions inferred here. Theevidence presented here indicates that RíoSonoyta existed in three distinct reaches thatwere perennial even during dry periods –Sonoyta, Santo Domingo, and Agua Dulce.

These reaches occupied a minority portion(perhaps a fourth to a third) of a 25 – 35 km(15.5 – 21.7 mi) length of the Sonoyta Valley,rising first about 4 km (2.5 mi) above Sonoytaand ending due west or southwest of Quito-baquito. Original conditions probably involvedalternative states of ciénega and scouredstream associated with each perennial reach,as suggested by Hendrickson and Minckley’s(1985) model.

Prior to the flood damage and entrenchmentof the river’s arroyo beginning about 1890, thelivestock and irrigation agriculture practicedin the valley probably altered the ecosystem inrelatively benign ways. However, the extent offlowing stream and wet meadow must havebeen significantly reduced by diversion ofsprings and the river onto irrigated fields.

Now the large perennial springs and flowingstreams at Sonoyta and Santo Domingo areessentially gone, but the location and probablythe length of the Agua Dulce reach remainssimilar to conditions in 1855. We cannot inferwhether the latter reach has already contracteddue to upstream activities, but the location ofits emergence is similar to Schott’s description.Descriptions of the need to dig in the sand bedfor water at the old site (La Salada) alsosuggests that conditions in this part of the riverwere much like those existing now, and theAgua Dulce perennial flow reach may not havebeen very much longer than it is now.

RESULTSStatus of the Sonoyta Mud Turtle

The Sonoyta mud turtle(for explanation of thisuse of the English name, see Rosen andMelendez [2010, this volume]), a locallyendemic subspecies of the Sonoran mud turtle,remains common at Quitovac, Presa Xochim-ilco and the spring pools below it, the Sonoytasewage lagoon, the Agua Dulce reach of RíoSonoyta, and Quitobaquito. It was rare andlocalized in the 4 km (2.5 mi) of river formerlysupporting the main ciénega above the presa,but we trapped or hand-captured several indi-viduals in each intermittent reach of poolsthere. In the Santo Domingo area, we found asingle turtle at a soup-bowl-sized remnant of


water at the semi-perennial spring at EjidoJosefa Ortiz, in October 2001. We did not findturtles upstream or downstream of these areas,but our surveys further upstream were notexhaustive, and turtles might exist there inriver pools and artificial ponds.

Originally, the principal turtle populationcan be presumed to have been in the Ciénegade Sonoyta, and had already been greatlyreduced during the period from the 1891 inci-sion to 1970, when modern groundwaterpumping began. Relatively wet conditionsprevailed from about 1977 through 1995, andwe found the remnant population at Xochim-ilco to consist mostly of turtles recruited in1997 or earlier. In contrast, there was abundantevidence of consistent recruitment at Quito-baquito (Rosen and Lowe 1996) and at AguaDulce. The turtle’s persistence at the town ofSonoyta is largely dependant on sewageeffluent from the army base, and at the Sonoytasewage lagoon, which is about 150 m (164 yd)south of the river and 1.5 km (0.9 mi) down-stream of the highway bridge. Thus, the entireSonoyta population may become threatened bythe imminent relocation and modernization ofthe town’s wastewater system.

As described below, aquifer withdrawalswest of Sonoyta may also threaten the turtlepopulation at Agua Dulce. Mining interests inthe springs at Quitovac (http://www.copper-ridge.com/s/Quitovac.asp) create a potentiallyimmediate threat to that population, and thepopulation at Quitobaquito (estimated at 90 –150 by Rosen and Lowe 1996) is too small toprovide a reliably viable long-term populationto preserve the taxon. As a result of thesethreats, the USFWS recognizes the Sonoytamud turtle as a candidate for threatened status,and AGFD is preparing a candidate conserva-tion agreement.

Native and Non-native Fishes in Río Sonoyta

The Quitobaquito pupfish (Cyprinodoneremus) and longfin dace (Agosia chryso-gaster) were confined to the Agua Dulce reachduring our survey period. Non-native fisheswere more widespread. Mosquitofish

(Gambusia affinis) were found at all non-effluent perennial and semi-perennial waterswe sampled, including at Quitovac, ColoniaLa Nariz, Santo Domingo, Ejido Josefa Ortiz,and southeast of Ejido Desierto de Sonora. AnAfrican cichlid population, probably Tilapiazillii, was found at Quitovac, and an individualwas reported at Agua Dulce reach in 2003 butnot seen subsequently. Black bullhead catfish(Ameiurus melas) were found at Agua Dulce,throughout the Xochimilco area, and in pools4 km (2.5 mi) east of the presa; although wesaw no evidence of them at Quitovac, the localpeople reported “bagre” (catfish) there. Asingle individual black bullhead was removedfrom Quitobaquito in 1995, and this speciesmay occur elsewhere in the Sonoyta Valley.We did not observe fishes at the Sonoytasewage lagoon during observations there in2001-2.

Our results are similar to findings reportedand summarized by Miller and Fuiman (1987)and Hendrickson and Varela (1989) with twoimportant exceptions: (1) native fishes werefound in Sonoyta in 1950 and in the reachbetween Sonoyta and Santo Domingo in 1987,and (2) our turtle sampling allowed us to detectnon-native fishes, particularly catfish, at morelocalities.

The native fishes were numericallypredominant at the Agua Dulce reach during2001-6 (Figure 5), and both species were abun-dant until dry conditions produced die-offs ofthe longfin dace during foresummer droughtsof 2002 (I. Barba, pers. comm.) and 2004 (ourobservations). During 2004 our trappingoccurred in July at low water and over-repre-sented dace, which were confined to poolswhere we could trap, and under-representedpupfish and mosquitofish, which were abun-dant in areas too shallow for our traps. In 2005and 2006 we found the dace confined to thelower third of the reach and we estimated it at< 1% of the total fish population. Meanwhile,the pupfish population had continued to thriveand the mosquitofish population had increasedmarkedly by 2006. A monitoring and exoticspecies removal program for fishes was insti-tuted by the Pinacate Biosphere Reserve at


Agua Dulce beginning in 2005 (E. Larios,pers. comm.).

Black bullheads are able to survive oxygendepletion by gulping air at the water surface,and thus they can remain in some of the mostinhospitable waters, even in effluents at PresaXochimilco. Mosquitofish may survivehypoxic water by respiring close to the watersurface; in addition, they are widely distributedin the Sonoyta Valley upstream of Agua Dulcereach. We suspect they are planted formosquito control and dispersed during floods,rapidly but temporarily occupying reaches atEjido Josefa Ortiz and Santo Domingo duringhigh flow in October 2003. Pupfish are highlytolerant of low oxygen, whereas longfin daceare hypoxia-sensitive like other Southwestern

stream fishes (Lowe et al. 1967). Thus, thedrought-associated decline of the longfin dacemay be explained primarily by respiratoryphysiology, exacerbated by groundwaterpumping; the longfin dace may also be threat-ened by the non-native fishes.

DISCUSSIONThe Hydrologic Threat

Brown (1991) provides details showing thatgroundwater pumping for agriculture wasvirtually non-existent prior to the 1970s. Itrapidly increased to a peak of 102.6 millionm3/yr (83,000 ac-ft/yr) in the 1980s, which heestimated to exceed natural recharge of theaquifer by 68 million m3/yr (55,000 ac-ft/yr).At that time, approximately 8000 ha (20,000

Figure 5. Relative abundance of fishes in the Agua Dulce reach of Río Sonoyta based onminnow trapping and direct counts during nighttime observations. Dace abundance issignificantly over-represented in results for 2004 (see text). The catfish abundance indexis an average of capture rates in hoop nets and minnow traps, standardize to unit mean.


ac) were being farmed, with 13,000 ha (32,000ac) developed for irrigation. The total farmedin the most recent two decades is probablylower, judging from records through 1997 (C.Conner, pers. comm.) and the large number ofunused wells, irrigation works, and fields seenduring our surveys. Local farmers ascribed thisto the high price of electricity for the pumps.Although a moratorium on new pumps andfurther irrigation developments went into effectin the 1980s, groundwater withdrawals vastlyexceed recharge, and existing infrastructurehas the capacity to more than double themaximal pumping of 1987 (Brown 1991).

Hendrickson and Varela (1989) reportednative fishes and perennial water during 1-7September 1987 near Sonoyta to a point 11 km(6.8 mi) below the highway bridge, where flowinfiltrated into the streambed. Above the presadam and town they found only non-nativefishes. They reported only about 6 km (3.7 mi)of dry riverbed separating the Sonoyta reachfrom the Agua Dulce reach, which was appar-ently about 13 km (8.1 mi) long and supportednative fishes. Although 1987 was a moderatelydry year, local rains were good during August1987 and the period 1977-1984 was the wettestin the instrumental record for the regionreaching back to 1896.

In 1983-1987 we saw more water in RíoSonoyta than thereafter, with a large perennialpool at the highway bridge in Sonoytabecoming predominantly dry by 1989. During2001-2006, we only found flow conditionsequaling those reported for the early 1980sshortly after major cyclonic storm events inOctober 2003. During our study period, wenever found evidence of perennial flow, oreven clearly perennial pools, anyplace exceptbelow the bedrock outcrops near PresaXochimilco and at the Agua Dulce reach. Wesuspect that normal drought conditionscombined with aquifer depletion accounts fordifferences between our observations and theprevious reports we have reviewed.

Our observations indicate that the springs(or ciénegas) at Sonoyta and Santo Domingohave almost completely disappeared and arenot currently supporting native fish habitat. We

found pools in the riverbed with catfish andturtles as far as 4.5 km (2.8 mi) east ofSonoyta, and with seepwillow (Baccharis sali-cifolia; indicating at least perched shallowgroundwater) north of Ejido Desierto deSonora in 2001. The river and its riparian forestare not fully degraded at Sonoyta despite thetrend of rapid decline.

The Agua Dulce reach retains generally thecharacter it had in our earliest availabledescriptions. However, we observed thatalmost all the formerly extensive irrigationagriculture developments below SantoDomingo were inactive during 2001-6. TheAgua Dulce reach could face rapid, completedesiccation if large-scale agricultural pumpingresumes in the lower Sonoyta Valley, espe-cially from Santo Domingo to Agua Salada.Murguía (1998) provides an example of howsensitive this reach and its native fishes noware to human-caused desiccation. Whether thecontinuing aquifer overdraft closer to Sonoytawill impact streamflow in this reach cannot bedetermined at present.

Non-native Species ThreatsThe nature of exotic species threats to nativefishes in Río Sonoyta is not entirely clear.Which species may have the greatest potentialimpacts, and which are likely to suffer fromsuch impacts are not definitely known. AtAgua Dulce, the four established species havepersisted together since at least 1983 (Millerand Fuiman 1987; Rosen, pers. obs.), but theyhave not done so through protracted droughtlike the current one.

As detailed above, our observations indi-cate that the proportional decline of the longfindace during 2001-6 was primarily caused bysuffocation during foresummer droughts.However, in April 2006 mosquitofish reachedtheir highest abundance during our observa-tions, and were occupying the deeper currents(10-30 cm / 4-12 in) in the stream that previ-ously comprised the spatial niche occupied bylongfin dace. Abundant mosquitofish maydepress dissolved oxygen levels sufficiently toincrease hypoxia mortality in the dace. Further,mosquitofish, which are highly predatory and


markedly piscivorous, might significantlyimpact recruitment in longfin dace andpupfish. Both introduced species — themosquitofish and black bullhead — were mostabundant in the deeper pools The catfish preyactively on mosquitofish (Rosen, pers. obs.2001-2), and might possibly exercise controlover mosquitofish density. The currentprogram of attempting to eliminate catfish atAgua Dulce, which is designed to protect thepupfish from predation (K. Larios, pers.comm. 2006), could benefit the mosquitofishby removing its principal predator, andmosquitofish could then disproportionatelyimpact the two native fishes. Thus, the fishcommunity structure may be dynamic in unex-pected ways, and we lack a firm scientific basisfor deciding management strategies.

No nonnative species of fish are currentlyestablished in Quitobaquito. An individualblack bullhead was removed from the pond(Rosen and Lowe 1996); golden shiner(Notemigonus crysoleucas) that were intro-duced in the 1960s were removed by drainingthe pond in 1969 (Minckley 1973; Bennett andKunzmann 1989).

Individuals of several species of non-nativeturtles have been documented in QuitobaquitoSprings. Smith and Hensley (1957) collecteda mating pair of Arizona mud turtles in 1955,which may have been the last of a populationor releases from east or north of Organ PipeCactus NM. A painted turtle, Chrysemys pictadorsalis was collected by Hulse (1974), acooter (Pseudemys sp.) was reported by P.Bennett and M. Kunzmann (pers. comm.1989), and a red-eared slider was removed byRosen and Lowe (1996).

As with introduced fishes, the threat fromnon-native saltcedar is not entirely clear.Saltcedar uses water as liberally as otherriparian forest and woodland trees, but growsmore densely, and thus may have a negativeimpact on streamflow in the Agua Dulce reach.Obviously, it is also supplanting native riparianplant biodiversity, and thus likely reducing birdspecies diversity. Saltcedar thickets usually donot support large lizard populations (Jakle andGatz 1985; Griffen et al. 1989), a finding

consistent with what we have seen during oursurveys. Nonetheless, attempts to eradicate itall at once could have dramatic impacts onthreatened aquatic animals. Under presentconditions, saltcedar is stabilizing the riverchannel at Agua Dulce. It provides bank struc-ture utilized by the mud turtles for burrowingto retreat from predators and presumably tosurvive flood scour that would be severewithout established woody, firmly rootedstreamedge plants. Without saltcedar or someother plant with sufficient root structure, floodscour might also become severe enough todramatically impact fish populations.

Grazing by introduced livestock may simi-larly have non-obvious effects. While grazingmay be responsible for saltcedar recruitmentdominance, this is presently moot becausesaltcedar is already thoroughly established.Removal of grazing from Agua Dulce wouldlikely lead to further stabilization of the low-flow channel, producing more and deeperpools, and favoring the pool-dwelling non-native fishes (as well as the turtles) at theexpense of the native fishes.

Conservation MeasuresThe most serious threats to the Río Sonoytaaquatic ecosystem come from water demandson the aquifer for agriculture and a growingtown population. Virtually all other desertcities in the Sonoran and Mojave Desert regionhave desiccated the rivers upon whose banksthey were situated. To prevent this pattern fromrepeating at Sonoyta, three general approachesmay be considered. First, we suggest thatfuture quality of life for human beings may bebeing sacrificed for present gains by poor plan-ning or resource valuation, as may besuggested in the case of riparian and riverinedamage that occurred at Tucson, Caborca, andother desert communities. Second, the directeconomic value of the river resource may beincreased by promoting ecotourism. Third,infrastructure improvements in the regionmight be promoted to enhance rather thandegrade the river environment.

The riparian gallery forest adjoiningSonoyta on the east exists as a significant


amenity and may be developed, along withother river reaches, for townspeople, travelers,and nature enthusiasts. Building on the richbird diversity and pleasant shade, a park andbirdwatching tour stop might be successfullyestablished. The Agua Dulce reach in thePinacate Reserve could also support naturetourism, although the sensitivity of theresource there currently suggests that lessintensive use should be planned there. Threegeneral steps can be recommended to facilitatethis process:1) Involve local community members and

leaders in park development.2) Conduct research with public participation

in identifying bird and other natureresources for ecotourism development.

3) Establish an urban and agricultural wildlifeprogram to preserve and enhance birdhabitat quality within human-occupiedlandscapes.Modernization of the sewage treatment and

sanitation system of Sonoyta was under activeconsideration during our surveys. It is currentlyin progress, and some of the following recom-mendations may be implemented as a result ofcooperation among several agencies andgovernmental entities from local to bi-nationallevels (Rosen 2009). Prospects for findingsufficient outside funds to support a highquality infrastructure upgrade may be signifi-cantly improved if the design will benefithabitat and rare and interesting animals as wellas improve health and quality of life. Sewageeffluents provide the only major aquatic andriparian environments near Nogales, Sonora,Tucson, Arizona, and parts of Phoenix,Arizona. These systems fail as biodiversityconservation for two principal reasons.

First, water quality is often very poor, withhigh nitrogen content which prevents fish fromthriving. Lack of fish then contributes tohuman health concerns involving mosquito-borne pathogens. Improvements in techniqueand infrastructure design may readily resolvethis problem in the future. For example, therelatively modern Pima County wastewater

treatment facility in southern Avra Valley nearTucson, supports reproducing salamander andtoad populations (Rosen, pers. obs. 2004).

Second, most wastewater treatment facili-ties and urban ecological “restoration” projectsin the U.S. Southwest have emphasized the useof ponds and lakes. Deep, standing waterponds provide habitat suitable for establish-ment and population expansion of exoticaquatic species that can have negative impactson native aquatic biodiversity. Such harmfulspecies could include introduced crayfishes,American bullfrogs (Rana catesbeiana), andvarious fishes, including those already in RíoSonoyta and others, such as Tilapia, that arelikely to be introduced (Rosen and Schwalbe2002; Minckley and Deacon 1991). Nativefishes are generally more flood resistant thannon-natives (Meffe 1984; Minckley and Meffe1987). These considerations suggest thefollowing recommendations:4) A modernized sewage treatment facility can

be designed to produce high quality watersuitable for native fishes, which can beintroduced for mosquito control.

5) The infrastructure design can includecreation of habitat preferentially suitable fornative fishes by releasing suitably treatedwastewater into the river channel, where itmay also create riparian development withother benefits for people and wildlife.

6) This design may be presented to fundingagencies likely to assist with capital andoperations.Control of non-native species is a less

costly and socially complicated, yet still essen-tial, issue for conservation in the SonoytaValley. A number of issues remain, requiring acombination of action and research:7) Administrative and educational efforts

should be directed toward preventing thespread of non-native cichlid fishes or othernew aquatic species in the Sonoyta Valley.

8) Quitobaquito pupfish and longfin dace,rather than or in addition to mosquitofish,


could be used as vector control or estab-lished in other parts of the Sonoyta Valleywhere suitable waters exist or may becreated. (Pupfish are likely to be superior,if not merely complementary additions tomosquitofish as agents of mosquito control[reviewed by Schoenherr 1988].) Arefugium for the longfin dace seems mostimmediately pressing at the present time.

9) Research is needed to understand the inter-actions of the four existing fishes in RíoSonoyta prior to large-scale efforts topartially remove the non-native species.

10) Saltcedar control and eradication measuresmay be combined with novel, simultaneousrevegetation programs to avoid inadver-tently destroying some of the fewremaining populations of threatened aquaticvertebrates.

Finally, the community’s interest inprotecting the aquifer and river are likely to bethe crucial determinants of the success ofbiodiversity conservation in the SonoytaValley.

SUMMARYRío Sonoyta, which adjoins Organ Pipe CactusNational Monument and the town of Sonoyta,Sonora, is a rare lowland desert stream that hasnot been completely degraded. Little perennialsurface water remains at Sonoyta, although theriver and aquifers near the town still support awoodland with many riparian birds, along withpopulations of the endemic Sonoyta mud turtle(a candidate for listing under the EndangeredSpecies Act). Degradation has been less severedownstream, within the perennial Agua Dulce(or “Papalote”) reach in the Pinacate BiosphereReserve. This reach supports a bird-rich,saltcedar-dominated woodland, the mud turtleand two native fishes (longfin dace and theendangered Quitobaquito pupfish). Severalspecies of bats are dependent on the river andoasis springs for their regional existence, andthe endangered Sonoran pronghorn utilizes thedownstream river corridor. Springs atQuitovac, Sonora, and Quitobaquito, Arizona,

also support the mud turtle and, at Quito-baquito the pupfish, as well as riparian biotas.Much of the Sonoyta River valley facesthreats, especially groundwater withdrawalsthat could lead to losses of species and riparianvalues. Additional threats are exotic saltcedarand introduced cichlid fish. There is interestfrom Mexican and U.S. agencies and NGOs ininitiating conservation efforts in the region.Opportunities for conservation includeecotourism, urban park development, modern-ized sewage treatment with effluent availablefor use in river restoration, and landscape-levelconservation planning.

ACKNOWLEDGMENTSWe thank Mike Baltzy, Israel Barba, LisaBucci, Charles Conner, Pepe Davila, AnaLuisa Gallardo, Rocio Gama, David Hall, IzarIzaguirre, Glen Knowles, Keno Larios, KellyLyons, Bryan Milstead, Juan Miranda, MatiMiranda, Amanda Moors, Lorena Morales,Gigi Owen, Rafaela Paredes, Jim Rorabaugh,and Tim Tibbitts for assisting us with fieldwork and information, and Arizona Game andFish Department and U.S. Fish and WildlifeService for grants that funded the field work.

LITERATURE CITEDBarns, B. and A. Pate. 2004. Determining foraging

and roosting areas for mastiff bats (Eumops spp.)using radio telemetry. http://www.nps.gov/gis/mapbook/tech/31.html

Bennett, P. S. and M. R. Kunzmann. 1989. A historyof the Quitobaquito Resource Management Area,Organ Pipe Cactus National Monument, Arizona.Technical Report No. 26, Cooperative NationalPark Resource Studies Unit (USGS SonoranDesert Research Station), University of Arizona,Tucson, AZ. 80pp.

Bolton, H. E. 1936. Rim of Christendom; a biog-raphy of Eusebio Francisco Kino, Pacific coastpioneer. Macmillan, New York, NY.

Brown, B. 1991. Land use trends surroundingOrgan Pipe Cactus National Monument. TechnicalReport 39, Cooperative National Park ResourcesStudies Unit (USGS Sonoran Desert FieldStation), University of Arizona, Tucson, AZ. 65pp.

Bryan, K. 1925. The Papago Country, Arizona. U.S.Geological Survey, Water-Supply Paper 499,Washington, D.C.

Bryan, K. 1928. Change in plant associations bychange in groundwater level. Ecology 9: 474–478.

Caruth, R.L. 1996. Hydrogeology of the Quito-


baquito Springs, and La Abra Plain, Arizona,Organ Pipe Cactus National Monument, andSonora, Mexico. U.S. Geological Society WaterResources Investigations Report 95–4295.

Cockrum, E. L. and Y. Petryszyn. 1986. Mammalsof the Organ Pipe Cactus National Monument.Special Report No. 5. Cooperative National ParkStudies Unit (USGS Sonoran Desert ResearchStation), University of Arizona, Tucson, AZ.

Emory, W. H. 1857. Report on the United States andMexico Boundary Survey, Made under the Direc-tion of the Secretary of the Interior. Volume I.Washington, D.C.

Felger, R. S. 2000. Flora of the Gran Desierto andRío Colorado of Northwestern Mexico. Univer-sity of Arizona Press, Tucson, AZ.

Felger, R. S., P. L. Warren, S. Anderson, and G. P.Nabhan. 1992. Vascular plants of a desert oasis:Flora and ethnobotany of Quitobaquito, OrganPipe Cactus National Monument, Arizona.Proceedings of the San Diego Society of NaturalHistory 8: 1–39.

Fisher, S. 1989. Hydrologic and limnologic featuresof Quitobaquito Pond and springs, Organ PipeCactus National Monument. Technical Report No.22, Cooperative National Park Resource StudiesUnit (USGS Sonoran Desert Research Station),University of Arizona, Tucson, AZ.

Goodman, B. S. 1992. Hydrogeology of the Quito-baquito Springs area, La Abra Plain, and the RíoSonoyta Valley, Organ Pipe Cactus NationalMonument, Arizona and Sonora, Mexico. MastersThesis, University of Arizona, Tucson, AZ.

Gray, A.B. 1856. Survey of a route for the SouthernPacific R.R., on the 32nd parallel. Wrightson &Co., Cincinnati, OH.

Griffen, G. F., D. M. Stafford Smith, S. R. Morton,G. E. Allan, and K. A. Masters. 1989. Status andimplications of the invasion of tamarisk (Tamarixaphylla) on the Finke River, Australia. Journal ofEnvironmental Management 29: 297–315.

Hendrickson, D.A. 1983. Distribution records ofnative and exotic fishes in Pacific drainages ofnorthern Mexico. J. Ariz.-Nev. Acad. Sci. 18: 33–38.

Hendrickson, D.A., and W.L. Minckley. 1985.Ciénegas—vanishing climax communities of theAmerican Southwest. Desert Plants 6: 131–175.

Hendrickson, D. A. and A. Varela Romero. 1989.Conservation status of the desert pupfish, Cyprin-odon macularius, in México and Arizona. Copeia1989: 478–483.

Hendrickson, D.A, W.L. Minckley, R.R. Miller, D.J.Siebert, and P.H. Minckley. 1980. Fishes of theRio Yaqui Basin, Mexico and United States. J.Ariz.-Nev. Acad. Sci. 15: 65–106.

Hershler, R., and J. J. Landye. 1988. Arizona Hydro-biidae (Prosobranchia: Rissoacea). SmithsonianContributions to Zoology, Number 459: 1–63.

Hornaday, W. T. 1908. Campfires on Desert andLava. Charles Scribner and Sons, NY.

Hoy, B. 1990. Sonoyta and Santo Domingo: A storyof two Sonoran towns and the river that ran by.Journal of Arizona History 31: 117–140.

Hulse, A. C. 1974. An autecological study of Kinos-ternon sonoriense LeConte (Chelonia;Kinosternidae). Ph.D. Dissertation. Arizona StateUniversity, Tempe, AZ.

Ives, R. L. 1936. Desert floods in the SonoytaValley. American Journal of Science 32: 349–360.

Jakle, M. D. and T. A. Gatz. 1985. Herpetofaunaluse of four habitats in the middle Gila Riverdrainage, Arizona. USDA Forest Service GeneralTechnical Report RM-120: 355–358.

Knowles, G. W., J. Rorabaugh, R. Paredes Aguilar,P. C. Rosen, D. H. Hall, and D. Riedle. 2002a.Distribution, status and conservation of theSonoyta mud turtle (Kinosternon sonorienselongifemorale). Presentation to the Desert FishesCouncil, November 2002.

Knowles, G. W., R. Paredes Aguilar, D. H. Hall, J.C. Rorabaugh, and P. C. Rosen. 2002b. Status,distribution, and recommendations concerning theSonoyta mud turtle (Kinosternon sonorienselongifemorale) in Sonora and Arizona. Posterpresentation, and abstract for conference. InMeeting resource management information needs,Tucson, Arizona, May 15-17, 2002, pp. 27–28.U.S. Geological Survey, Sonoran Desert FieldStation, University of Arizona, Tucson, AZ.

Logan, M.F. 2002. The lessening stream: An envi-ronmental history of the Santa Cruz river.University of Arizona Press, Tucson, AZ.

Lowe, C. H., D. S. Hinds, and E. A. Halpern. 1967.Experimental catastrophic selection and tolerancesto low oxygen concentration in native Arizonafreshwater fishes. Ecology 48: 1013–1017.

Lumholtz, C. 1912. New Trails in Mexico.Reprinted in 1990 by The Southwest Center,University of Arizona Press, Tucson, AZ.

MacDougal, D. T. 1908. Across Papagueria.Bulletin of the American Geographical Society 40:1–21.

McMahon, T. E., and R. R. Miller. 1985. Status ofthe fishes of the Rio Sonoyta Basin, Arizona andSonora, Mexico. Proceedings Desert FishesCouncil 13-15: 237–245.

Mearns, E. A. 1907. Mammals of the MexicanBoundary of the United States. U.S. NationalMuseum Bulletin No. 56.

Meffe, G. K. 1984. Effects of abiotic disturbance onco-existence of predator-prey fish species.Ecology 65: 1525–1534.

Miller, R. R. 1961. Man and the changing fish faunaof the American Southwest. Papers of theMichigan Academy of Sciences, Arts, and Letters46: 365–404.

Miller, R. R., and L. A. Fuiman. 1987. Descriptionand conservation status of Cyprinodon maculariuseremus, a new subspecies of pupfish from OrganPipe Cactus National Monument, Arizona. Copeia1987: 593–609.


Miller, R. R., W. L. Minckley, and S. M. Norris.2005. Freshwater Fishes of Mexico. The Univer-sity of Chicago Press, Chicago, IL.

Minckley, W. L. 1973. Fishes of Arizona. ArizonaGame and Fish Department, Phoenix, AZ.

Minckley, W. L. & J. E. Deacon (eds.). 1991. Battleagainst extinction: Native fish management in theAmerican west. University of Arizona Press,Tucson, AZ.

Minckley, W.L., and G.K. Meffe. 1987. Differentialselection by flooding in stream-fish communitiesof the arid American Southwest. In Communityand evolutionary ecology of North Americanstream fishes, edited by W. J. Matthews and D. C.Heins, pp. 93–104. University of Oklahoma Press,Norman, OK.

Murguía, L. 1998. El pez pupo del desierto(Cyprinodon macularius) en la Reserva de laBiosfera El Pinacate y Gran Desierto de Altar,Sonora, México. Cross Border Waters: Fragiletreasures for the 21st century. Ninth U.S./MexicoBorder States Conference on Recreating, Parks,and Wildlife, Proceedings, edited by Gottfried etal., pp. 200–208. RMRS-P-5, USDA ForestService, Fort Collins, CO.

Nabhan, G. P., A. M. Rea, K. L. Reichardt, E.Melink, and C. F. Hutchinson. 1982. Papago influ-ence on habitat and biotic diversity: Quitovacoasis ethnoecology. Journal of Ethnoecology 2:124–143.

Nabhan, G. P., A. M. Rea, K. L. Reinhardt, E.Melink and C. F. Hutchinson 2000 Papago(O’odham) influences on habitat and biotic diver-sity: Quitovac oasis ethnoecology. InEthnobotany: A reader, edited by P. E. Minnis, pp.4–62. University of Oklahoma Press, Norman,OK.

Ohmart, R. D. 1982. Past and Present Bioticcommunities of the Lower Colorado River Main-stem and Selected Tributaries. Vol. V. The GilaRiver, San Pedro River, Santa Cruz River. Bureauof Reclamation, Boulder City, NV.

Rosen, P. C. 2007. The amphibians and reptiles ofthe dry borderlands of northwestern Sonora andsouthwestern Arizona. In Dry borderlands, editedby R. S. Felger and B. Broyles, pp. 310–337.University of Utah Press, Salt Lake City, UT.

Rosen, P. C. 2009. Integration of the decliningSonoyta mud turtle into modernization of munic-ipal Sonoyta, Sonoran, Mexico. Abstract of

presentation at IUCN Turtle Survival Alliance andIUCN Tortoise and Freshwater Turtle SurvivalGroup 2009 Annual Meeting, St. Louis, Missouri.August 5-8, 2009. http://www.turtlesurvival.org/resources/document-library/2009annualmeeting/

Rosen, P.C. and C. Funicelli. 2008. Conservation ofurban amphibians in Tucson: Salvage projectreport and discussion of habitat issues. Report toPima County Flood Control District and Board ofSupervisors, Tucson, AZ. http://rfcd.pima.gov/reports/pdfs/conservation_urban_amphibs_2008.pdf

Rosen, P. C. and C. H. Lowe. 1996. Populationecology of the Sonoran Mud Turtle (Kinosternonsonoriense) at Quitobaquito Springs, Organ PipeCactus National Monument, Arizona. Final Reportto Arizona Game & Fish Dept. Heritage Program,Phoenix, AZ.

Rosen, P. C. and C. R. Schwalbe. 2002. Widespreadeffects of introduced species on aquatic reptilesand amphibians in the Sonoran Desert region. InExotic species in the Sonoran Desert, edited by B.A. Tellman, pp. 220–240. University of ArizonaPress, Tucson, AZ.

Russell, S. M. and G. Monson. 1998. The Birds ofSonora. University of Arizona Press, Tucson, AZ.

Schoenherr, A. A. 1988. A review of the life historyand status of the desert pupfish, Cyprinodonmacularius. Bulletin of the Southern CaliforniaAcademy of Sciences 87: 104–134.

Smith, P. W. and M. M. Hensley. 1957. The mudturtle Kinosternon flavescens stejnegeri Hartweg,in the United States. Proceedings of the Biolog-ical Society of Washington 70: 201–204.

Snyder, J. O. 1915. Notes on a collection of fishesmade by Dr. Edgar A. Mearns from rivers tribu-tary to the Gulf of California. Proc. U. S. Natl.Mus. 49: 573–586

Sullivan, B.K., R.W. Bowker, K.B. Malmos andE.W.A. Gergus. 1996. Arizona distribution ofthree Sonoran Desert anurans: Bufo retiformis,Gastrophryne olivacea, and Pternohyla fodiens.Great Basin Naturalist. 56 (1): 38–47.

Turner, R. M., R. H. Webb, J. E. Bowers, and J. R.Hastings. 2003. The Changing Mile Revisited.University of Arizona Press, Tucson, AZ.

Weir, D. R. and I. Azary. 2001. Quitovac oasis: Asense of home place and the development of waterresources. The Professional Geographer, 53: 45–55.

OVERVIEW OF THE LOWER COLORADO RIVER MULTI-SPECIESCONSERVATION PROGRAMWilliam E. Werner

Development of the lower Colorado River,including the construction of four large damsand five smaller weirs that provide waterstorage, pumping forebays, and diversion facil-ities (Table 1), has resulted in physical changesthat, in combination with the introduction ofnon-native species, have affected the status ofnative fish and wildlife species. Species nowlisted under the Endangered Species Act (ESA)include the Yuma clapper rail (Ralluslongirostris yumanensis), Southwest willowflycatcher (Empidonax traillii extimus), deserttortoise (Gopherus agassizii), bonytail (Gilaelegans), Colorado pikeminnow (Ptycho-cheilus lucius), humpback chub (Gila cypha),and razorback sucker (Xyrauchen texanus).Agencies responsible for water and power-related operation, maintenance, and distri-bution, developed, in collaboration with otherentities, the Lower Colorado River Multi-Species Conservation Program (LCR MSCP)to address and offset effects of ongoing andanticipated river operations on the listedspecies presently found in the area.

The origin of the LCR MSCP lies in thelisting of the four “Big River” fish (Table 2)by the U.S. Fish and Wildlife Service(Service). In general terms, the listing rationalefor the bonytail and razorback sucker werehabitat fragmentation, alteration of the pre-development hydrograph, and introduction ofand interaction with non-native aquaticspecies.

Following listing of the razorback suckerand designation of critical habitat for all four

Big River fish in 1994, the Service advised theBureau of Reclamation (Reclamation) thatthey should consult under Section 7(a)(2) ofthe ESA on their operation and maintenance ofthe Lower Colorado River (LCR) system.

The role of the Secretary of the Interior,acting through Reclamation, as watermasteron the LCR is defined in the decree of theU.S. Supreme Court in Arizona v. California(1964). Also, Reclamation has had much ofthe responsibility in the development of theriver, and title to many of the river facilitiesremains with the federal government eventhough some of those facilities are operatedand maintained by non-federal water users.An example is Imperial Dam, which is ownedby the federal government but operated by theImperial Irrigation District to divert water tothat district in California as well as theWellton-Mohawk Irrigation and DrainageDistrict and Yuma area districts in Arizona.Because of these multiple roles, many actionshave both federal and non-federal compo-nents. As Reclamation began to address theissue of consultation on operation and main-tenance activities they chose to include theirstakeholders in the process. Further, stake-holders were concerned that their interestswould not be given sufficient considerationin a traditional ESA consultation processbetween Reclamation and the Service.

DEVELOPMENT OF THE LCR MSCPThe LCR MSCP planning area comprisesareas up to and including full-pool elevations

162 Werner

of Lakes Mead, Mohave, and Havasu and thehistorical floodplain of the Colorado Riverfrom Lake Mead to the international boundarywith Mexico at San Luis, Arizona. Participantsincluded agencies, water providers, powerdistributors, water and power contractors, andenvironmental organizations.

Federal participants were the Departmentof the Interior (including Bureau of Reclama-tion, Bureau of Land Management, Bureau ofIndian Affairs, Fish and Wildlife Service, andNational Park Service) and the Department ofEnergy’s Western Area Power Administration.

Arizona participants included Arizona Elec-tric Power Cooperative, Inc., Arizona PowerAuthority, City of Bullhead City, CentralArizona Water Conservation District, CibolaValley Irrigation and Drainage District, Elec-trical District No. 3 – Pinal County, Town ofFredonia, Golden Shores Water ConservationDistrict, City of Lake Havasu City, City ofMesa, Mohave County Water Authority,Mohave Valley Irrigation and Drainage

District, Mohave Water Conservation District,North Gila Valley Irrigation and DrainageDistrict, Salt River Project AgriculturalImprovement and Power District, City ofSomerton, Town of Thatcher, Unit “B” Irriga-tion and Drainage District, Wellton-MohawkIrrigation and Drainage District, Town ofWickenburg, City of Yuma, Yuma CountyWater Users’ Association, Yuma IrrigationDistrict, Yuma Mesa Irrigation and DrainageDistrict, Arizona Game and Fish Commissionand Department, and Arizona Department ofWater Resources.

California participants included Bard WaterDistrict, City of Needles, Coachella ValleyWater District, Colorado River Board of Cali-fornia, Imperial Irrigation District, Los AngelesDepartment of Water and Power, Palo VerdeIrrigation District, San Diego County WaterAuthority, Southern California EdisonCompany, Southern California Public PowerAuthority, and The Metropolitan Water Districtof Southern California.

Table 1. Lower Colorado River dams.

Name Year Completed TypeLaguna Dam 1924 DiversionHoover Dam 1936 StorageParker Dam 1938 StorageImperial Dam 1938 DiversionMorelos Dam 1950 DiversionDavis Dam 1950 StorageHeadgate Rock Dam 1950 DiversionPalo Verde Weir 1957 DiversionGlen Canyon Dam 1963 Storage

Table 2. “Big river” fish of the Colorado River.

Species Status under ESA, Year Critical Habitat Designated

Colorado pikeminnow (Ptychocheilus lucius) Endangered, 1967 1994Humpback chub (Gila cypha) Endangered, 1967 1994Bonytail (Gila elegans) Endangered, 1980 1994Razorback sucker (Xyrauchen texanus) Endangered, 1991 1994

Werner 163

Nevada participants included the ColoradoRiver Commission of Nevada, SouthernNevada Water Authority, Nevada Departmentof Wildlife, and Basic Water Company.

Participation by environmental organiza-tions varied through time with NationalWildlife Federation participating at the time ofcompletion.

Analysis of EffectsConstruction of dams on the Colorado Riverhas resulted in alteration of the hydrograph.Prior to the construction of Hoover Dam theriver experienced large spring floods on anearly annual basis. The baseline condition atthe initiation of planning for the LCR MSCPwas an altered riverine system and decoupledfloodplain. The hydrograph of the river isaltered by floodwaters being stored. Largeintra-annual variability was dampened by theconstruction of Hoover Dam (Figure 1). Thesediment regime has been altered in that sedi-ment drops in calm waters of reservoirs andthe water released from dams is clear. Clear

water picks up new sediment and channel inci-sion occurs. Consequently, a much greaterdischarge is needed to overtop the riverchannel and floodplain flooding no longeroccurs. These changes are significant to thenative biota in that the processes that naturallymaintained riparian forest are altered, newlyhatched and young fish don’t have access toextra cover in flooded bottomlands, in clearwater young fish are more susceptible topredation by sight-feeding predators, andsystem “reset” no longer occurs.

As Anglo settlement occurred in theColorado River basin in the late 1800s and theearly 1900s, fish and other aquatic organismsfrom the Mississippi Basin were introducedinto the Colorado River system. The aquaticfauna of the Colorado River was distinct withrelatively few species prior to settlement. Fishand other organisms were introduced fromelsewhere as a food source and for sportfishing. Many of these introduced speciesquickly occupied newly changed habitats asriver development occurred. Construction of

Figure 1. Colorado River flows below Hoover Dam (Courtesy U.S. Bureau of Reclamation)

164 Werner

dams limited movement of fish, such asColorado pikeminnow, that once apparentlymoved great distances. The water of reservoirsbecame clear to the greater advantage to sight-feeding predators than species that evolved inthe more turbid water of the Colorado.

The species addressed by the LCR MSCPare largely associated with riparian woodlandhabitat, marsh, and backwaters as those habitatshave been most affected by river development.Species listed under the ESA include Yumaclapper rail (Rallus longirostris yumanensis),Southwest willow flycatcher (Empidonax trailliiextimus), desert tortoise (Gopherus agassizii),bonytail (Gila elegans), humpback chub (G.cypha), and razorback sucker (Xyrauchentexanus). Other covered species include westernred bat (Lasiurus blossevillii), western yellowbat (L. xanthinus), Colorado River cotton rat(Sigmodon arizonae plenus), Yuma hispidcotton rat (S. hispidus eremicus), western leastbittern (Ixobrychus exilis hesperis), Californiablack rail (Laterallus jamaicensis coturniculus),yellow-billed cuckoo (Coccyzus americanusoccidentalis), elf owl (Micrathene whitneyi),gilded flicker (Colaptes chrysoides), Gila wood-pecker (Melanerpes uropygialis), vermilionflycatcher (Pyrocephalus rubinus), ArizonaBell’s virio (Vireo bellii arizonae), Sonoranyellow warbler (Dendroica petechia sonorana),summer tanager (Piranga rubra), flat-tailedhorned lizard (Phrynosoma mcalli), relictleopard frog (Rana onca), flannelmouth sucker(Catostomus latipinnis), MacNeill’s sootywingskipper (Pholisora gracielae), sticky buckwheat(Eriogonum viscidulum), and threecornermilkvetch (Astragalus geyeri var. triquetrus).Additional species are included for which suffi-cient information was not available for permitprocessing but will receive research effort andmay be added as covered species if warranted,including California leaf nosed bat (Macrotuscalifornicus), pale Townsend’s big eared bat(Corynorhinus townsendii pal-lescens), lowlandleopard frog (Rana yavapaiensis), and desertpocket mouse (Chaetodipus penicillatussobrinus).

Four categories of covered activities areaddressed by the LCR MSCP including

ongoing flow-related activities, future flow-related activities, ongoing non-flow-relatedactivities, and future non-flow-related activi-ties. Included are operation and maintenanceactivities related to responsibilities of Recla-mation for: flood control, improvement ofnavigation, and river regulation; storage anddelivery of Colorado River water for agricul-tural, municipal, industrial, and other beneficialpurposes; and for generation of electricalpower, all as described in the LCR MSCPBiological Assessment (LCR MSCP 2004c).The LCR MSCP Habitat Conservation Plandescribes non-federal activities (LCR MSCP2004b). Ongoing non-federal flow-relatedactivities include water diversions and returnsof up to 9.251 billion cubic meters per year(gcmy) (7.5 million acre feet per year (mafy))from existing facilities and diversions andreturns of any surplus waters for water contrac-tors in Arizona, California, and Nevada. Futureflow-related activities include power produc-tion and changes in points of diversion up to1.941 gcmy (1.574 mafy) by water contractorsin Arizona, California, and Nevada ofColorado River water and consequent reduc-tion in releases from Hoover, Davis, andParker Dams. Changes in point of diversionare anticipated in response to changes in waterdemand within the states. Ongoing non-flow-related activities include operation, main-tenance, and replacement of existing waterdiversion and distribution works and electricalgeneration and distribution facilities within theplanning area and certain activities conductedby Arizona Game and Fish Department andNevada Department of Wildlife along the river.

Hydrologic modeling and subsequentanalysis were completed for flow-relatedcovered activities to describe changes tohydrologic conditions and associated changesto river surface elevations, reservoir elevations,and groundwater levels (LCR MSCP 2004b,c).The analysis addressed impacts to ground-water, marsh, and backwater habitats from thelower river surface elevations caused bychanges in point of diversion. Changes togroundwater may affect surface vegetation andsurface elevations of backwaters not directly

Werner 165

connected to the river. Changes in daily lowriver surface elevations may affect connectedbackwaters, marsh, and riverine habitats.Changes to Lake Mead elevations may affectaquatic habitats, such as spawning areas, andterrestrial vegetation within the pool. Reduc-tions in flows past Morelos Dam, at the lowerend of the system, may affect terrestrial habitatin the Limitrophe Division. Non-flow-relatedactivities were typically analyzed as “foot-print” impacts, i.e., quantification of the areaunder the action.

Conservation PlanThe basic conservation strategy for the LCRMSCP includes 3 core elements, a habitat-based approach for most species (Table 3),population augmentation for listed fish (Table4), and a program to maintain habitat baselinevalues. During development of the proposedconservation strategy we used native habitaton the lower Bill Williams River as a concep-tual model of pattern, juxtaposition, and

interspersion for an integrated mosaic ofhabitat to be created through the LCR MSCP.On the Bill Williams River we see yellow-billed cuckoo in the cottonwood overstory,Southwest willow flycatcher in the mid-storyand understory levels, Yuma clapper rail in theTypha marsh, and razorback sucker in inter-spersed open water. In the process of offsettingeffects of covered activities it is not necessaryto create separate acreage to mitigate effects toeach species, provided that created acreageprovides the habitat needs of target species.Also, by providing habitat for species that useearly seral stages (e.g. sapling willow trees)such as the Southwest willow flycatcher, andlate seral stages (e.g. mature cottonwoodcanopy forest) such as the yellow-billedcuckoo, we have addressed or have largelyaddressed the needs of many other species aswell. Addition of missing components to basicearly or late seral cottonwood/willow commu-nities can ensure that the needs of additionalspecies are met.

Table 3. Habitat creation.

Area Affected by Area to be CreatedLand Cover Type Covered Activities as Offset to Effects

Cottonwood-Willow 866 ha 2404 haMesquite 239 ha 534 haMarsh 104 ha 207 haBackwater 161 ha 146 ha*TOTALS 1,370 ha 3,291 ha

* Habitat created to be higher quality than that impacted.

Table 4. “Big river” fish of the Colorado River.

Species Activity

Razorback Sucker Raise and release 660,000 fish over 50-year periodBonytail Raise and release 620,000 fish over 50-year periodHumpback Chub $10,000/year to GCDAMP* for 50 yearsFlannelmouth Sucker $80,000/5 years + 85 acres (34 ha) of backwaters

* Glen Canyon Dam Adaptive Management Program.

166 Werner

Major conservation themes for listed fishinclude maintenance of existing genetic diver-sity, noting that Lakes Mohave and Havasucurrently serve as genetic refuges. To maintaingenetic diversity the Program will raise nativefish to a size less vulnerable to predation priorto release. This will be accomplished throughcapture of wild larvae and through hatcheryproduction of fish from broodstock (withcareful genetic monitoring and management).Habitat creation and management for fishincludes providing predator-free environments(refuges) for native fish. These areas will eitherserve as destinations for released fish or maybe periodically connected back to the main-stream to allow adult fish to utilize naturalhabitat in the mainstream Colorado.

The LCR MSCP is continuing efforts of anad hoc Lake Mohave Native Fish Workgroupto capture wild-hatched larvae of razorbacksucker in Lake Mohave, and is now imple-menting similar efforts in Lake Mead. Theselarvae are raised in aquaria and then grow outponds until large enough for release. This ispossible because newly hatched larvae areattracted to light so biologists are able tocapture the larvae with a dip net after hanginga light over the side of a boat at night.

Humpback chub are found in the ColoradoRiver in Grand Canyon with potential overlapwith the LCR MSCP planning area at theupper limits of Lake Mead. Conservation forthis chub is to support efforts of the adjacentGlen Canyon Dam Adaptive ManagementProgram. The flannelmouth sucker is foundas an introduced population below DavisDam that is naturally reproducing. Theemphasis for this species is to better under-stand how it is successfully using habitat inthat area, and to ensure that the amount ofexisting habitat remains, as that under-standing may improve conservation measuresfor other species.

To maintain habitat baseline values, thoseexisting at the time of plan development, a$25,000,000 Habitat Maintenance Fund wasestablished. This fund, established up-front inthe process in interest-bearing accounts, canbe used to fund actions to maintain existing

habitat values within the planning area. Therationale is that absent the processes that main-tain habitat value through time, such as springflood flows, existing habitat will degrade orchange through natural aging and successionalprocesses and will be of lesser value to coveredspecies in the future, unless maintained. Thisfund is available to land managers with theconsent of Reclamation, the Service, and Stateparticipants.

Because of uncertainty about the status andhabitat requirements of some species, andregarding effective approaches for creation offunctional habitat, research and adaptivemanagement are important elements of theLCR MSCP. The program includes funding forresearch, with emphasis early in the program,to gain knowledge to design conservationmeasures to address species for which suffi-cient knowledge was lacking, to refineknowledge of specific habitat requirements orthresholds in variables, and to develop tech-niques for cost and water efficient creation andmaintenance of habitat.

The LCR MSCP includes extensive moni-toring of covered species, both at newlycreated habitat areas and throughout the lowerColorado River corridor, to document status,trends, and use of habitat.

During development of the conservationplan proposal a conservation opportunity area(COA) analysis was conducted. The purposeof this effort was to determine feasibility ofimplementation of conservation measures forcovered species in the planning area. Thisanalysis included a review of the geomorphicsetting, a review of LCR hydrology, a surveyof existing features such as relic sloughs andbackwaters, an investigation of opportunitiesto establish an integrated mosaic of aquatic,marsh, and riparian habitats, and an assessmentof availability of lands and water that could beused for conservation purposes. Generalizedconclusions from the COA analysis are: • The morphology of the lower basin is a

constraint as it is a series of canyons andvalleys with the canyons not providingsufficient space for large stands of trees,

Werner 167

• Because of physical changes to the river theacreage achieved through modification offlows is not great,

• Entrenchment and lack of sediment supplyare system changing conditions,

• Soil salinity is increasing throughout theLower Basin as “flushing” of salts has beenreduced with changes to the hydrograph,and

• Large patch size is important for somespecies and to develop large acreages inblocks, conversion of some agriculturallands will be necessary. The overall goal of the LCR MSCP is

implementation of a program that will: • Conserve habitat and work toward recovery

of threatened and endangered species, aswell as reduce the likelihood of additionalspecies being listed;

• Accommodate present water diversions andpower production and optimize opportuni-ties for future water and powerdevelopment, to the extent consistent withthe law, and

• Provide the basis for incidental take author-izations.

IMPLEMENTATIONImplementation of the LCR MSCP is speci-fied in a Funding and ManagementAgreement (FMA) among the parties (LCRMSCP, 2005a) and is carried out by Recla-mation with oversight from a SteeringCommittee that includes permittees and otherstakeholders, including a Federal ParticipantGroup, an Arizona Participant Group, a Cali-fornia Participant Group, a NevadaParticipant Group, a Native American Partic-ipant Group, a Conservation ParticipantGroup, and an Other Interested Parties partic-ipant Group. The Steering Committeeoperates by consensus, with a dispute resolu-tion process. Through the FMA funding is50% from the federal government and 50%from the states of Arizona, California, and

Nevada. The state share is split 50% Cali-fornia, 25% Arizona, and 25% Nevada.

As described in the FMA, the ProgramManager (a Reclamation employee) shall annu-ally develop and present to the SteeringCommittee an Implementation Report, WorkPlan, and Budget. Those documents shallinclude: a current financial report; a descriptionof all conservation measures initiated, continuedor completed during the previous year; adescription, purpose, and cost of all conserva-tion measures to be initiated during the nextthree years; a running tabulation and descrip-tion of all conservation measures completed todate; a description of any take of coveredspecies during the previous budget period; arunning tabulation of habitat created or restoredby the MSCP; a description of all findings,conclusions, and results of monitoring, research,or conservation measures undertaken; anyrecommendations by the Service or statewildlife agency regarding the LCR MSCP; andapproval or rejection of any minor modification.The annual workplan process is the means bywhich conservation measures are modifiedthrough adaptive management based on knowl-edge gained through research. A ScienceStrategy, which will be revised on a five-yearcycle, complements the process to ensure thatpriority research needs are met to support adap-tive management. Once reviewed by theSteering Committee, annual work plans areforwarded to the Service for their review priorto implementation. In addition to the AnnualImplementation Report, Work Plan, Budget, andContribution Payment Schedule, other mattersthat must be presented to the SteeringCommittee for its consideration include: addi-tional or modified conservation measuresproposed pursuant to adaptive management;land and water acquisitions; reports andresponses to Congress and Federal and Stateregulatory agencies; and financial reports andaccountings.

Two large blocks of agricultural land wereavailable to the LCR MSCP at initiation ofimplementation, the Palo Verde EcologicalReserve (PVER) in the Palo Verde Valley inCalifornia and the Cibola Valley Conservation

168 Werner

Area (CVCA) in Arizona. Through agreementwith Reclamation, the PVER was made avail-able to the LCR MSCP without cost to offsetcosts to implement requirements of the Cali-fornia Endangered Species Act (CESA)Section 2081 permit. Non-California partieswere concerned that CESA requirementswould increase overall program costs and werenot interested in sharing those additional costs.The CVCA is agricultural land in the CibolaValley Irrigation and Drainage Districtpurchased by Mohave County Water Authorityfor the associated water, part of which theyintend to transfer upstream in the future.Access to these two large blocks of agriculturalland has enabled Reclamation to move forwardwith pilot scale efforts to establish native vege-tation as habitat. Initial efforts, plantinggallon-size rooted cuttings by hand in auguredholes, have evolved to mechanized massplanting of greenhouse-raised cuttings utilizinga planter designed for cauliflower. As adaptivemanagement is a fundamental tenet of the LCRMSCP knowledge gained from research andpilot projects, such as the mass-plantingproject, will be fed back into the program witha goal of efficient and effective implementa-tion. Examples of the more than twenty-fiveresearch projects in the first five years of theprogram include: a study to determine salinity,temperature and oxygen limits for bonytail andrazorback sucker to facilitate design andmanagement of off-channel habitats; a studyof effectiveness of quagga mussel protocols forremoving mussels from transport water duringmovement of native fish by truck; a study ofnest predation on avian-covered species thatmay improve the design of habitat patches; astudy of western red bat and western yellowbat roosting characteristics; and a study ofhydrologic conditions such as soil moisture,depth to ground water, and amount of standingwater underneath occupied habitat for thewillow flycatcher and yellow-billed cuckoosin order to duplicate such conditions at habitatcreation sites.

ENVIRONMENTAL COMPLIANCEEnvironmental compliance documents for theLCR MSCP include the Final ProgrammaticEnvironmental Impact Statement/Environ-mental Impact Report (LCR MSCP, 2004a)that analyses, as required by the National Envi-ronmental Policy Act (NEPA) and CaliforniaEnvironmental Quality Act (CEQA), effects ofissuance of the ESA Section 10(a)(1)(B)permit by the Fish and Wildlife Service, imple-mentation of conservation actions by theBureau of Reclamation, implementation of theconservation plan and funding by Californiaentities, and issuance of a CESA Section 2081permit by California Department of Fish andGame. The Environmental Impact Statementdoes not provide NEPA compliance forongoing operations and maintenance or futureactions other than conservation measures. TheFinal Habitat Conservation Plan (LCR MSCP,2004b) was submitted by the non-federalparticipants to the Fish and Wildlife Servicewith an application for the Section 10(a)(1)(B)permit. A Final Biological Assessment (LCRMSCP, 2004c) was submitted by Reclamationto the Fish and Wildlife Service during formalconsultation under ESA on ongoing operationand maintenance activities and implementa-tion of the conservation measures. Because ofthe size of the documents that shared muchinformation, a volume of appendices (LCRMSCP, 2004d) and a separate volume thatincludes responses to comments on the entirepackage (LCR MSCP, 2004e) were produced.An Implementation Agreement (LCR MSCP2005b) associated with the Section 10(a)(1)(B)permit defines terms between permittees andthe Fish and Wildlife Service. The Funding andManagement Agreement (LCR MSCP 2005a)defines funding, management, and termsbetween the permittees and Reclamation.

LITERATURE CITEDLower Colorado River Multi-Species Conservation

Program. 2004a. Lower Colorado River Multi-Species Conservation Program, Volume I: FinalProgrammatic Environmental Impact State-ment/Environmental Impact Report. December17, 2004. Jones and Stokes, Inc. Sacramento, CA.

Werner 169

Lower Colorado River Multi-Species ConservationProgram. 2004b. Lower Colorado River Multi-Species Conservation Program, Volume II: FinalHabitat Conservation Plan. December 17, 2004.Jones and Stokes, Inc. Sacramento, CA.

Lower Colorado River Multi-Species ConservationProgram. 2004c. Lower Colorado River Multi-Species Conservation Program, Volume III: FinalBiological Assessment. December 17, 2004. Jonesand Stokes, Inc. Sacramento, CA.

Lower Colorado River Multi-Species ConservationProgram. 2004d. Lower Colorado River Multi-Species Conservation Program, Volume IV: FinalAppendices for Volumes I-III and V. December17, 2004. Jones and Stokes, Inc. Sacramento, CA.

Lower Colorado River Multi-Species ConservationProgram. 2004e. Lower Colorado River Multi-Species Conservation Program, Volume V:Responses to Comments on LCR MSCP VolumesI-IV. December 17, 2004. Jones and Stokes, Inc.Sacramento, CA.

Lower Colorado River Multi-Species ConservationProgram. 2005a. Lower Colorado River Multi-Species Conservation Program, Funding andManagement Agreement. April 4, 2005.

Lower Colorado River Multi-Species ConservationProgram. 2005b. Lower Colorado River Multi-Species Conservation Program, ImplementationAgreement. April 4, 2005.

The giant saguaro (Carnegiea gigantea), anicon of the Sonoran Desert due to its unique-ness in the landscape, has had a role as adelimiter of the Desert’s boundaries and plantcommunities (Shreve and Wiggins 1964). Thesaguaro is a long-lived perennial cactus inwhich many different biotic interactions havebeen characterized, such as its dependence onnurse plants during establishment, bird nestingand insect presence in its trunk, and bat polli-nation of flowers and bird dispersion of seeds(Steenbergh and Lowe 1969, 1977; Turner etal. 1969; Turner et al. 1966). Even diseased,necrotic and dead saguaro stems create habi-tats for insects. Given its importance in saguaroperformance, biotic interactions may be theexplanation for saguaro population decline andrecovery at different times and places (Alcornand May 1962; Steenbergh and Lowe 1969,1977; Turner 1990).

Declines in saguaro forests have beennoticed in different parts of the Sonoran Desertand at least one study has proposed a correla-tion between increased saguaro mortality andepidermal browning (Turner and Funicelli2000). Although its causes have not beenclearly identified, epidermal browning refersto a series of symptoms in saguaros, the mostconspicuous being changes in color of above-ground tissues over time from green to yellowand brown (Duriscoe and Graban 1992; Evanset al. 1995; Turner and Funicelli 2000),decreasing UV-B protection, photosynthesisand gas exchange (Evans et al. 1994b; Lajthaet al. 1997). As the saguaro is a dominant and

keystone species of the Sonoran Desert,changes in its population dynamics will likelyhave an effect on the functioning of this desertecosystem.

Termites are important components indesert ecosystem function. Their role in cellu-lose decomposition and nutrient recycling hasbeen well studied in North American deserts(Schaefer and Whitford 1981; Whitford2002). In most cases, it has been shown thatdesert termites are restricted to undergroundnesting and feeding on woody and dead plantmaterials. Most of the forty species in theSonoran Desert are specialists to particularplant species (Smith 2000), and those limitedto saguaro, were found only on its lignifiedbase (Olson 2000).

In the past, termite activity on live tissueshas been reported anecdotally, and mostly inurban settings (Jones and Nutting 1989). Thereare no reports or studies of termite activity onliving tissues of saguaro and most other desertplants. Population density and/or controlprocedures of Heterotermes aureus,Gnathamitermes tubiformans (subterraneantermites), and other termite species have beenreported in live tissue of several grass anddesert plants, but without any quantification ofdamage (Allen et al. 1980; Bodine and Ueckert1975; Spears et al. 1975; Ueckert et al. 1976).

In this paper, we report for the first time thepresence of termites on aboveground livinggreen tissues of saguaro. We studied termitepresence on a number of sites and over severalyears in order to increase our understanding

TERMITE ACTIVITY ON GREEN TISSUES OF SAGUARO (CARNEGIEA GIGANTEA) IN THE SONORAN DESERT

Alejandro E. Castellanos, Reyna A. Castillo, Adrian Quijada

172 Castellanos, Castillo, and Quijada

about the role and causes of this previouslyundocumented phenomena as well as itscontribution to the ecology of saguaro popu-lations throughout the Sonoran Desert.

METHODSSaguaro populations were sampled fromMesquite–Palo Verde plant communitieswithin the Central Gulf Coast subdivision ofthe Sonoran Desert (Brown 1982; Shreve andWiggins 1964) in Sonora, Mexico. The siteswere located at 29°22'15" lat N and 112°00'29"long W (Figure 1). The associated naturalvegetation showed little evidence of distur-bance, and consisted of a mixture of trees(Prosopis glandulosa, Cercidium micro-phyllum, Bursera microphylla, Olneya tesota,Pachycereus pringlei) and shrubs (Fouquieriasplendens, Lophocereus schotti, Stenocereusthurberi, Larrea tridentata, Ambrosia dumosa,Encelia farinosa (species name follows Shreveand Wiggins 1964). The nearest climatologicalrecords available were obtained from PuertoLibertad, Sonora, a coastal town, approxi-mately 60 km from the farthest site. PuertoLibertad has a mean annual temperature of19° C and average annual precipitation of 83.8mm (14-year average). Winter rainfall is animportant and significant percentage (40.9 %)of total annual rainfall.

Fieldwork was performed in December1989, March 1990, January 2001, and Mayand August 2003. Saguaro populations weresampled using point-centered quarter transectsof 700-1500 m length. Sampled points weretaken every 50 m, and at each point, the fournearest saguaros were measured for height,diameter, vigor, presence or absence of termiteactivity (nesting tubes) along their ribs, numberof ribs with such activity, and height of termitetubes on saguaro living tissues. We reportedsaguaro vigor based on tissue color; greentissue was considered healthy, yellow wasintermediate, and brownish color unhealthy.For most analyses, saguaros were grouped inthree size classes: saplings (up to 1 m), juve-niles (1 to 2.5 m) and adults (above 2.5 m). InJanuary 2001, as it was evident that termiteswere notoriously present on aboveground

tissues of a diverse number of desert species,we sampled saguaros and other plant speciesin the same community for termite activity onliving tissues.

At each date, we sampled abovegroundnesting tubes for species identification.Although a number of other insect andarthropod species were found along the exfo-liating bark tissues at the base of saguaros, wesampled only those found inhabiting the aerialnest remains on saguaro green tissues. Speci-mens identified as Heterotermes aureus (PaulBaker, pers. com.) were collected in August2003 from aerial nesting tubes on affectedgreen epidermal tissues of saguaro.

We used χ2 statistical tests to identify therelationship between termite presence andvigor (or browning) in the sampled saguaropopulations. Here we report only thoseanalyses performed on data from 1989, 1990and 2001. We used statistical software toperform our analyses (SPSS, SAS Institute)with a significance level of p < = 0.05 (Sokaland Rohlf 1995).

RESULTSPresence of Termites on Aboveground

Green Tissues of SaguaroOur analysis at the study site focused onsaguaros because at our first sampling dates(1989 and 1990) this was the species thatconspicuously showed a presence of termiteson aboveground living tissues (Figure 2). Pres-ence of termite nesting — tubes were foundmost frequently in adult and juvenile saguarosand less in saplings, although the height ofaffected individuals went from a few centime-ters to 8 to 10 m (Figure 3). Presence oftermites on juvenile saguaros was mostfrequent in 1989, but not in 1990 whenfrequency was higher on mature individuals.A comparison between year and age group(Figure 2) resulted in highly significant statis-tical differences for presence of termites onadult saguaros in each year (1989, χ2 = 15.29;1990, χ2 = 27.55; 2001, χ2 = 25.12; all signif-icantly different χ2 < =0.001).

Termite nesting tubes started at the bottomof saguaros and went, in most cases, up to the

Castellanos, Castillo, and Quijada 173

green epidermal tissues. Most adult individ-uals with termite presence had either somesort of epidermal damage or different degreesof epidermal browning from previous years.In saguaros, height of termite tubes wasstatistically and significantly correlated withheight of browning surface area (χ = 0.8041,p < = 0.0001; Table 1), although not allsaguaros with browning in their tissues hadtermite nesting tubes on them, and somehealthy saguaros had termites present only intheir green living tissues. Healthy saguaros,apparently colonized for the first time bytermites, had the shortest height of termite

tubes along their ribs. Height of termite tubeswas also positively correlated with numberof affected ribs (χ = 0.7957, p < = 0.0001),and height of saguaro (χ = 0.6668, p < =0.0001).

We did not notice a strong directionalcomponent on saguaro termite nesting tubes,although a formal geostatistical analysis wasnot performed. At our sites about 51% of allsampled individuals were affected by termiteactivity in 75% or more of their ribs, and asmaller percentage of saguaros had every oneof their ribs affected to different degrees (datanot shown).

Figure 1. Location of sampled saguaro sites in Sonora (dots). Other locations in BajaCalifornia (near La Paz), Mexico and Arizona (east of Casas Grandes), USA, are notshown.


Presence of Termites on AbovegroundTissues of Sonoran Desert Plant Species

We found widespread termite presence onsaguaros over a large portion of their distri-butions in the Sonoran Desert (Figures 1 and3a, b, c). as well as on Pachycereus pringlei(Figure 3d). During the 2001 field season, wealso found termites in aboveground tissues ofcacti such as cholla (Opuntia fulgida; Figure3e), senita cactus (Lophocereus schotti) andpitahaya (Stenocereus thurberi; Figure 3f) aswell as herbaceous and woody species suchas Fouquieria splendens, Ambrosia deltoidea,Ambrosia dumosa and Encelia farinosa toname the most conspicuous in our sample(Table 2). In 2001, termites were found ashigh as 7–8 m in saguaro and cardon (Figures3c and d).

DISCUSSIONOur study is the first to describe termite pres-ence on aboveground green epidermal tissuesof live saguaros. Previous descriptions oftermite ecology in North American desertshave been restricted to underground nestingand feeding habits with decayed wood anddead plant materials (Schaefer and Whitford1981). Termite presence on aboveground livetissue in Sonoran Desert plants has only inci-dentally been mentioned for species in urbanand agricultural environments (Jones andNutting 1989). There is only anecdotal refer-ences for native desert plants like Opuntia,Acacia gregii, Dalea and Atriplex (Jones andNutting, 1989) from other North Americandeserts. Jones and Nutting (1989) mention thatParaneotermes simplicicornis was found on

Figure 2. Percentage of saguaros with (black columns) and without (white) termite nestingtubes on aboveground green tissues. Data grouped for three age categories (see text)were significantly different χ2 < 0.001 in all years (1989, χ2 = 15.29; 1990, χ2 = 27.55;2001, χ2 = 25.12).


Figure 3. a) Termites along ribs in a young saguaro individual (study site); b) Adultsaguaro with termites only on ribs (Organ Pipe Cactus National Monument); Termitescan reach considerable heights in saguaro (c) and cardon (d), and have been found incholla (e) and pitahaya (f) at different localities within the Sonoran Desert.

Table 1. Spearman non-parametric test for Height of termite tubes in live tissues as theycorrelated with Saguaro height, diameter, number of damaged ribs and height of browningin tissues. Highly statistical significance for p < = 0.0001.

Saguaro ρ p

Height of termite tubes Height 0.6668 < .0001Diameter 0.2958 0.0206Damaged ribs 0.7957 < .0001Browning height 0.8041 < .0001


cactus skeletons of chollas and Carnegieagigantea (saguaro), but not on their live tissues.

Termites and DisturbanceTermite presence on living saguaro and above-ground tissues of desert species may beincreasing within the Sonoran Desert. At our sitein 1989, termites were noticed only on saguaro,but by 2003, they were present on shrubs andother cactus species (Table 2). Termites may nothave been previously noticed on saguarobecause termite tubes persist for only a shortperiod of time during fall and early winter, theyare washed away with light winter rains.Although termite activity typically ceases whentemperatures are cooler than 9° C (Ueckert et

al. 1976), we have no data on the physiologicalconstraints of the species we found in our study.Given the large variability in the amount anddistribution of termite presence in abovegroundtissues in different years, it is possible that soilmoisture and air temperature are factors thatinfluence termite nesting tube presence.

There are a number of possible causes thatmay be influencing the increase of termitepresence in aboveground tissues of plantspecies from Sonoran Desert ecosystems.Change in temperature and rainfall patternseither human (land-use change) or climatically(global change) induced may be importantfactors for increased termite activity above-ground. Termite activity increases during wet

Table 2. Number and presence percent of termite tubes on live tissue of trees, shrub andsub-shrub species at the study site. Data are from three transects (900 * 2 m each) made in2001.

Termite presence SampleSpecies % n

TreesCercidium microphyllum 25.0 4Olneya tesota 0 6ShrubsFouquieria splendens 54.4 54Ambrosia dumosa 33.3 57Lippia palmeri 28.6 14Ambrosia deltoidea 23.3 129Condalia globosa 16.7 12Encelia farinosa 15.4 78Hyptis emoryi 9.5 21Krameria parvifolia 5.1 39Larrea tridentata 0 33Lycium spp 0 5Jatropha cinerea 0 8CactiFerocactus wislizenii 90.0 10Pachycereus pringlei 68.5 89Lophocereus schotti 49.1 57Stenocereus thurberi 42.4 59Carnegiea gigantea 33.5 243Opuntia fulgida 25.9 27Opuntia leptocaulis 46.7 15


years (Bodine and Ueckert 1975) and withanthropogenic disturbance. Increased abun-dance of certain termite functional types hasbeen noticed in deserts and the tropics aftervegetation disturbance (Black and Okwakol1997; Davies et al. 2003). In the tropics, distur-bance has lead to termites nesting in the stemsof live plants (Hegh 1922), often associatedwith sites with poor soil nutrient conditions(Black and Okwakol 1997).

Human disturbance is increasingly wide-spread in the Sonoran Desert. The plantcommunities at the sites we studied have beenincreasingly decimated in their populations ofkeystone species such as Prosopis spp(mesquite) and Olneya tesota (ironwood),which have been selectively and heavilyharvested for fencing, firewood, charcoal, andcrafts. Over the last 15-20 years, charcoal hasdramatically increased as a profitable domesticand export business. The statistics on woodand charcoal production from our study regionare unreliable, however there is evidence thatextraction levels are considerable (Taylor2006). Given the low rate of establishment andslow growth of desert trees, overexploitationof these key species is to be expected after afew years (Suzán et al. 1997), as shown in theabsence or almost complete absence of thesespecies from our vegetation sampling (Table2). We think that as wood removal increases tomeet demand, physical and biotic processesaffected by diminished litter and dead mate-rial, such as termite dynamics, may becomedisrupted. Wood extraction may affect primaryproductivity, availability and seasonality ofwood litter presence, alter decompositiondynamics (Nash et al. 1999), and indirectlyalter termite behavior, perhaps triggeringincreased aboveground presence in saguaroand other desert plants. If this is the case, weshould be seeing increased termite activity inareas where land-use patterns cause impover-ished litter and dead root biomass input to thesoil. We know very little about the causes andecological and functional implications ofaboveground termite activities.

Other natural or human-induced changes inrainfall and land cover may be responsible for

increasing aboveground termite presence onother desert species (Table 2). Increased vari-ability in temporal rainfall patterns andincreasing temperatures will occur as globalchange progresses. Climate models forSonoran Desert predict an increase in rainfalland temperatures during the autumn-winterseason (SRAG 2000). Because termite activityincreases with late summer and fall precipita-tion at our sites, a change in rainfall patternsand increased temperatures during fall andwinter seasons could be already happening andhelp explain the increased presence of termiteson live plant tissues over the years, and onsaguaro and cardon from many distant loca-tions from Baja California, Arizona(Castellanos, pers. obs.), and islands of theGulf of California (Castillo, pers. obs.) in theSonoran Desert.

Termites and Saguaro BrowningPresence of termites on living saguaroepidermal tissues has not been described andthe consequences of their presence on saguarophysiology and population dynamics is yetunknown. We found however, important simi-larities between our observations of termitevisible effects to living epidermal tissues ofsaguaro with those described as first occurringfor epidermal browning (Turner and Funicelli2000). We propose that termite effects onsaguaro green epidermal tissues and epidermalbrowning may be related phenomena. At ourstudy site, termite presence was correlated withsome of the characteristics of tissue damagethat had been ascribed to epidermal browning(Table 1). Compared to healthy saguaros withno spine damage (Figure 4a), spines withtermite nesting tubes (Figure 4b) showed signsof loosening first the middle spine and damageat the base of the spine areole and along the rib(Figures 4b and c), with total spine loss appar-ently after several events. Spine loosening andfall is a first sign of saguaro browning (Turnerand Funicelli 2000).

We observed other similarities betweentermite effects on saguaro epidermal tissuesand saguaro browning. Browning spreads fromribs to adjacent epidermal tissues, a pattern


similar to that found with termites initiallyclimbing only along ribs and progressivelyspreading to adjacent green tissues (Figure 4d).Damage induced each year, either physicallyor possibly as a wounding response, will allowtermites to spread further from lignified spinesover the next years (Figure 4d). Progressiveeffects spreading from spine ribs to epidermaltissues (Figures 4c, d, and e) were associatedwith termites, with symptoms very similar towhat has been reported for epidermal browning(Figure 1 in Turner and Funicelli 2000). Duringyears with high termite aboveground activity,browning and termites were relatedphenomena as found in browned saguaros withtermites spread all over them (Figure 4f). Sinceaboveground termite presence on epidermaltissues of saguaros differs from year to year(Figure 2), spread and damage may takedecades.

Epidermal browning in saguaro does not yethave a known definitive causal agent. Severalphysical variables such as freezing (Steenberghand Lowe 1977), air pollution (Stolte 1992),heat and UV-B load (Evans et al. 1994b; Evanset al. 1992; Turner and Funicelli 2000) havebeen proposed as causes of saguaro epidermalbrowning. Increased heat and UV-load onsouthern exposure saguaro stems seem to berelated to the directional effects of browningfound in some saguaro populations of thenorthern Sonoran Desert region (Evans et al.1994a; Evans et al. 1992; Turner and Funicelli2000). Termite effects on saguaros were visu-ally well correlated with most previouslydescribed symptoms of saguaro browning. Thesymptoms we uncovered under termite nestingtubes on spines (Figure 4b), and scars left onepidermal tissues (Figure 4e), were similar tothose previously associated with browning

Figure 4. Saguaro browning correlates with termite activity. Healthy saguaro ribs andspines (a); Termite tubes along spines (b); Spine damage after termite presence (c);Termite damage spread to green tissue (d) (tubes were removed from region in the centerof the photography to expose the kind of termite damage); Spread of termite tubes to allsaguaro ridges and aerial tissues (e); Saguaro epidermal browning (f). Photographs weretaken in 2001


(Turner and Funicelli 2000) and tissue hard-ening and lignifications of saguaro epidermaltissue (Evans et al. 1994b; Olson 2000).Although we did not specifically measuredirectional effects, at our sites there were noobvious directional patterns in the epidermalbrowning on saguaro tissues, since more than75% of ribs were affected in at least 51% ofaffected adult saguaros. Some of the studiedsites are at more southern, non-freezing lati-tudes and in regions with warmer temperaturesthan the locations where directional effectshave been reported (McAuliffe 1993). If direc-tional effects are present at more northernlatitudes of saguaro distribution, that wouldn’tcancel the possibility that increasing heat loadsmay benefit termite activity and abovegroundgrowing conditions or that both effects couldbe related. Spine loss may diminish the coolingeffects on epidermal tissues, and increase heatload, but this should be tested. As an end result,saguaro browning decreases photosyntheticcarbon gain, and induces a positive feedbackloop of diminishing physiological performance(Lajtha et al. 1997), inducing an early senes-cence process.

Termites and Sonoran DesertEcosystem

In this paper we have described a previouslyundocumented and seemingly increasingphenomena of termite presence on above-ground plant species in the Sonoran Desert. Wethink that presence of termites abovegroundmay be an early warning signal that importantchanges are happening in Sonoran Desertecosystems, and here we bring some evidencethat such phenomena will be playing anincreasingly important role in shaping thestructure and function of species and ecosys-tems in this North American desert.

It is of outmost importance to determine theecological and environmental factors that haveenabled termites to spread aboveground, and tounderstand the ecological consequences of thisincreased termite activity. As a small glimpseof what could happen, we are amazed that sucha sturdy plant as saguaro, a plant that can with-stand years and possibly centuries of stressful

physical conditions, may be so greatly affectedby a small, almost inconspicuous insect.

ACKNOWLEDGMENTSAECV acknowledges support fromCONACYT in the form of research andsabbatical grants and David Williams andSAHRA at the University of Arizona forproviding a stimulating atmosphere while onsabbatical. Iván Parra and José Llano helpedwith fieldwork. Paul Baker, University ofArizona, helped with taxonomic identificationof termites. Alejandro Nava helped with Figure1. Francisco Molina-Freaner, Richard Felgerand other anonymous reviewers providedhelpful comments on earlier drafts. We thankLaura Hoffman for greatly improving ourEnglish.

LITERATURE CITED Alcorn, S. M., C. May, 1962. Attrition of a saguaro

forest. Plant Disease Reporter 46: 156–158.Allen, C. T., D. E. Foster, and D. N. Ueckert, 1980.

Seasonal food habits of a desert termite,Gnathamitermes tubiformans in West Texas. Envi-ronmental Entomology 9: 461–466.

Black, H. I. J., and M. J. N. Okwakol, 1997. Agri-cultural intensification, soil biodiversity andagroecosystem function in the tropics: The role oftermites. Applied Soil Ecology 6: 37–53.

Bodine, M. C., and D. N. Ueckert, 1975. Effect ofdesert termites on herbage and litter in a shortgrassecosystem in West Texas. Journal of RangeManagement 28: 353–358.

Brown, D. E., 1982. Biotic communities of theAmerican Southwest—United States and Mexico.Desert Plants 4: 1–342.

Davies, R. G., L. M. Hernández, P. Eggleton, R. K.Didham, L. L. Fagan, and N. N. Winchester, 2003.Environmental and spatial influences upon speciescomposition of a termite assemblage acrossneotropical forest islands. Journal of TropicalEcology 19: 509–524.

Duriscoe, D. M. and S. L. Graban. 1992. Epidermalbrowning and population dynamics of giantsaguaros in long-term monitoring plots. InProceedings of the symposium on research inSaguaro National Monument, edited by C. P.Stone and E. S. Bellantoni, pp. 237–257. NationalPark Service Cooperative National ParkResources Studies Unit, School of RenewableNatural Resources, University of Arizona, Tucson,AZ.

Evans, L. S., V. A. Cantarella, L. Kaszczak, S. M.Krempasky, and K. H. Thompson, 1994a.Epidermal browning of saguaro cacti (Carnegiea


gigantea): Physiological effects, rates of browningand relation to sun/shade conditions. Environ-mental and Experimental Botany 34: 107–115.

Evans, L. S., V. A. Cantarella, K. W. Stolte, and K.H. Thompson, 1994b. Epidermal browning ofsaguaro cacti (Carnegiea gigantea): Surface andinternal characteristics associated with browning.Environmental and Experimental Botany 34:9–17.

Evans, L. S., K. A. Howard, and E. J. Stolze, 1992.Epidermal browning of saguaro cacti (Carnegieagigantea): Is it new or related to direction? Envi-ronmental and Experimental Botany 32: 357–363.

Evans, L.S., V. Sahi, and S. Ghersini, 1995.Epidermal browning of saguaro cacti (Carnegieagigantea): Relative health and rates of surficialinjuries of a population. Environmental and Exper-imental Botany 35: 557–562.

Hegh, E., 1922. Les Termites, partie générale.Description.—Distribution géographique.—Classi-fication.—Biologie.—Vie sociale.—Alimenta-tion.—Constructions.—Rapports avec le mondeextérieur. Impr. industrielle & financiére, Bruxelles.

Jones, S. C., and W. L. Nutting, 1989. Foragingecology of subterranean termites in the SonoranDesert. In Special biotic relationships in the aridsouthwest, edited by J. O. Schmidt, pp. 79–106.University of New Mexico Press, Albuquerque,NM.

Lajtha, K., K. Kolberg, and J. Getz, 1997. Ecophys-iology of the saguaro cactus (Carnegiea gigantea)in the Saguaro National Monument: Relationshipto symptoms of decline. Journal of Arid Environ-ments 36: 579–590.

McAuliffe, J. R., 1993. Case study of research,monitoring, and management programs associatedwith saguaro cactus (Carnegiea gigantea) atSaguaro National Monument. Technical ReportNPS/ WRUA/NRTR-93-01. National Park Ser-vice Cooperative National Park Resources StudiesUnit, School of Renewable Natural Resources,University of Arizona, Tucson, AZ.

Nash, M. S., J. P. Anderson, and W. G. Whitford,1999. Spatial and temporal variability in relativeabundance and foraging behavior of subterraneantermites in desertified and relatively intactChihuahuan Desert ecosystems. Applied SoilEcology 12: 149–157.

Olson, C. A., 2000. Insects and the saguaro. In Anatural history of the Sonoran Desert, edited byArizona-Sonora Desert Museum, pp. 353–356.University of California Press.

Schaefer, D., and W. G. Whitford, 1981. Nutrientcycling by the subterranean termiteGnathamitermes tubiformis in a Chihuahuandesert ecosystem. Oecologia 48: 277–283.

Shreve, F., and I. L. Wiggins, 1964. Vegetation andFlora of the Sonoran Desert. Stanford UniversityPress.

Smith, R. L., 2000. Termites. In A natural history ofthe Sonoran Desert, edited by Arizona-Sonora

Desert Museum, pp. 316–319. University of Cali-fornia Press.

Sokal, R. R., F. J. Rohlf, 1995. Biometry. W. H.Freeman, New York.

Spears, B. M., D. N. Ueckert, T. L. Whigham, 1975.Desert termite control in a shortgrass prairie:Effect on soil physical properties. EnvironmentalEntomology 4: 899–904.

SRAG (Space Radiation Analysis Group), 2000.Preparing for a changing climate. The potentialconsequences of climate variability and change.In Preparing for a changing climate: The potentialconsequences of climate variability and change,edited by NOAA, ISPE, and USGS, p. 60. Tucson,AZ.

Steenbergh, W. F., and C. H. Lowe, 1969. Criticalfactors during the first years of life of the saguaro(Carnegiea gigantea) at Saguaro National Monu-ment, Arizona. Ecology 50: 825–834.

Steenbergh, W. F., and C. H. Lowe, 1977. Ecologyof the saguaro. II. Reproduction, germination,establishment, growth and survival of young plant,p. 242. U.S. Government Printing Office, Wash-ington, D.C.

Stolte, K.W., 1992. Air pollution threats to biolog-ical resources in Saguaro National Monument. InAir pollution threats to biological resources inSaguaro National Monument, edited by C. P. Stone,and E. S. Bellantoni, pp. 217–222. Southwest Parksand Monuments Association, Tucson, AZ.

Suzán, H., D. T. Patten, and G. P. Nabhan, 1997.Exploitation and conservation of ironwood(Olneya tesota) in the Sonoran Desert. EcologicalApplications 7: 948–957.

Taylor, M. J., 2006. Biomass in the borderlands:Charcoal and firewood production in Sonoranejidos. Journal of the Southwest 48: 63–90.

Turner, D. S., and C. S. Funicelli, 2000. Ten-yearresurvey of epidermal browning and populationstructure of saguaro cactus (Carnegiea gigantea)in Saguaro National Park. Technical Report 69,U.S. Geological Survey, Sonoran Desert FieldStation, University of Arizona, Tucson,AZ.

Turner, R. M., 1990. Long-term vegetation changeat a fully protected Sonoran Desert site. Ecology71: 464–477.

Turner, R. M., S. M. Alcorn, and G. Olin, 1969.Mortality of transplanted saguaro seedlings.Ecology 50: 834–844.

Turner, R. M., S. M. Alcorn, G. Olin, and J. A.Booth, 1966. The influence of shade, soil andwater on saguaro seedlings establishment. Botan-ical Gazette 127: 95-102.

Ueckert, D. N., M. C. Bodine, and B. M. Spears,1976. Population density and bio-mass of thedesert termite Gnathamitermes tubiformans(Isoptera: Termitidae) in a shortgrass prairie: Rela-tionship to temperature and moisture. Ecology 57:1273–1280.

Whitford, W. G., 2002. Ecology of Desert Systems.Academic Press, San Diego, CA.

2002a; Segee and Neeley 2006). With theexception of mountainous regions, pedestrianfences or vehicle barriers now span the borderin the study area from San Luis east onto theTohono O’odham Nation. These structureshave created barriers not only to people andtheir vehicles, but to wildlife and cultures aswell (Cordova and de la Parra 2007). Also ofgreat concern in the desert lands is the spreadof flammable non-native winter annual plantsthat are capable of fueling large fires.

This description of the herpetofaunaemphasizes species or groups of species thatare imperiled within the region or throughouttheir ranges as well as the conservation effortsthat would facilitate the maintenance of thesethreatened, but important populations. Someare conserved by species-specific plans andactions, while others are afforded protectionvia land management practices, includingprotective land use designations. Conservationefforts are occurring within biological, cultural,political, and fiscal environments that in somecases constrain their effectiveness, but in othersprovide significant conservation opportunities.

STUDY AREAThis area of northwestern Sonora, Mexico, andsouthwestern Arizona, U.S., is bounded by theColorado River on the west, the Gila River onthe north, a line from Gila Bend, Arizona, toBahia San Jorge, Sonora, on the east, and theUpper Gulf of California on the south; alsoincluded is a small portion of Baja Californiaeast of the Colorado River (Figure 1).

The 4.26 million hectare (10.5 mil ac) regionof Sonoran Desert in southwestern Arizona andnorthwestern Sonora is surprisingly diverse inregard to its herpetofauna, owing to its variedlandscapes and habitats: from the marinewaters of the Gulf, the remnant riparian wood-lands and marshlands along the Colorado,Sonoyta, and Gila Rivers, the volcanic slopesof the Sierra Pinacate, the arid and relativelybarren dunes of the Gran Desierto (the largestdune system in the western hemisphere,Bowers 1998), to the broad desert valleys andmountains.

The region’s rivers and wetlands have beendramatically altered by water diversions,upstream dams, groundwater pumping, leveesand bankline structures, introduction of sportfishes and other non-native plants and animals,and development of the floodplains for agri-culture and urban uses (Ohmart et al. 1988;Fradkin 1996; Felger 2000; Glenn et al. 2001).However, the desert lands are relatively intact,being too dry to support much livestockgrazing and generally lacking minerals ofsufficient quality for large-scale mining.Current uses of the desert lands, mainly aerialmilitary training, border security, dispersedrecreation, and wildlife management, have leftrelatively few scars on the landscape, althoughroads, trails, camps, and other disturbance nearthe border in Arizona — resulting from illegalborder crossings and response by U.S. BorderPatrol and other law enforcement agencies —have increased greatly in recent years(Milstead and Barns 2002; Rorabaugh et al.

CONSERVATION OF AMPHIBIANS AND REPTILES IN NORTHWESTERN SONORAAND SOUTHWESTERN ARIZONAJames C. Rorabaugh

Turner and Brown 1982). Washes and slopessupport a more diverse flora, both structurallyand in species richness; diversity also increaseswith rainfall from west to east (Felger 2000;Turner and Brown 1982). Mountains andbajadas in the region’s eastern portion are char-acterized by the more diverse flora of theArizona Upland subdivision of Sonorandesertscrub (Brown and Lowe 1994).

Variability in precipitation brings aboutdramatic floristic changes, both seasonally andin the longer term. This is most apparent in wetwinters that produce luxuriant ephemeral,herbaceous floras, but can also be seen in wetand dry cycles during which perennial cacti,shrubs, and trees may die back in largenumbers or much regeneration occurs. Adrought from 1931-34 in the Sierra Pinacatekilled large numbers of trees and shrubs andthey have yet to return to their prior abundance(Hayden 1998).

This region of the Sonoran Desert is partic-ularly hot and dry. High temperaturescombined with low relative humidity insummer result in high evaporation and tran-spiration rates. Yuma has a potentialevapotranspiration-to-precipitation ratio ofabout 30, which is particularly challenging forplant survival and growth (Dimmitt 2002).Freezing temperatures are relatively rare (15-25 days a year at Yuma, and about one day atPuerto Peñasco) and becoming rarer (Weissand Overpeck 2005), but occasional damagingfreezes occur. Freezing temperatures in 1978resulted in die back of elephant trees (Burseramicrophylla) and organ pipe cactus (Steno-cereus thurberi) at OPCNM (S. Rutman, pers.comm. 2005). Lows of -5.5° C (22.1° F) wererecorded at Bahia Adair and Sierra del Rosarioin December 1972 (May 1973).

Rainfall is bimodal (winter and summer)and highly variable; mean annual precipitationranges from 40 mm (1.6 in) at San Luis RíoColorado to 267 mm (10.5 in) at Aguajita /Quitobaquito (Felger 2000; Comrie andBroyles 2002). Periodic severe droughtcontrasting with brief wet periods stronglyinfluence plant and animal communities(Felger 2000; Rosen 2000). For instance, no

Common biological, climatic, geological,and cultural threads run through the study areathat bind it into a discrete unit, despite a myriadof land management jurisdictions and an inter-national border (Cornelius 1998). Elevationsvary from sea level to 1,466 m (4,810 ft) at AjoMountain and 1,290 m (4,232 ft) at PinacatePeak. Dunes and sandy flats of silica sandsbrought to the region by the Colorado Riverstretch from Yuma south and east through theGran Desierto to Bahia San Jorge (Kresan2007; Bowers 1998). The Pinacate Volcanicfield covers about 2,000 km2 (722 mi2), whileelsewhere northwest to southeast trendinggranitic and basaltic mountain ranges occur.Large management jurisdictions cover muchof the landscape; they include the Barry M.Goldwater Range (BMGR, U.S. Departmentof Defense), Cabeza Prieta National WildlifeRefuge (CPNWR, U.S. Fish and WildlifeService), and Organ Pipe Cactus NationalMonument (OPCNM, U.S. National ParkService); and in Mexico the Reserva de laBiosfera Pinacate y Gran Desierto de Altar(Pinacate Reserve, La Comisión Nacional deÁreas Naturales Protegidas (CONANP)) , andthe Reserva de la Biosfera Alto Golfo de Cali-fornia y Delta del Río Colorado (Alto GolfoReserve, CONANP).

The study area is relatively sparsely popu-lated; the largest city is San Luis Río Colorado,with roughly 160,000 inhabitants (Schmidt2005), followed by Yuma (77,515) and PuertoPeñasco (roughly 35,000). Smaller communi-ties include Sonoyta (10,817), Somerton(7,266), Ajo (3,705), Gila Bend (1,980), andWellton (1,829) (U.S. Census Bureau 2000,Border Environment Cooperation Commission2005). Arizona communities grow in thewinter months, often substantially, due to aninflux of winter visitors, and the population ofPuerto Peñasco can reportedly double onholiday weekends.

Floristically the area is largely representedby simple, sparse shrub communities, oftendominated by creosote (Larrea tridentata),white bursage (Ambrosia dumosa), and otherspecies of the Colorado Desert subdivision ofSonoran desertscrub (Brown and Lowe 1994;

182 Rorabaugh

Figure 1. Study area and protection/management areas.

Rorabaugh 183

184 Rorabaugh

measurable precipitation fell in the Sierra delRosario from September 1969 to August 1972,a period of 34 months (May 1973). Yet,winters can be relatively wet, and occasionalmonsoons or late summer-early fall rains maybring more than the annual average rainfall ina short period. From 24-26 September 1997,the remnants of Hurricane Nora dropped 86.6mm (3.4 in) of rain on Yuma. Monsoon stormsin the afternoons of July 27 and August 9, 1989dropped a total of 151.6 mm (6.9 in) of rain onYuma (National Oceanic and AtmosphericAdministration 1989a,b), which is 217% of themean annual rainfall (Holcombe et al. 1997).

Water is a “master limiting factor” indeserts (Pianka 1986). Net primary produc-tivity, insect abundance, and lizard clutch sizesand frequency, body size, and density havebeen found in various studies to all be posi-tively correlated with amounts of recentprecipitation (Rosenzweig 1968; Turner et al.1969, 1982; Pianka 1970, 1986; Parker andPianka 1975; Anderson 1994; Dickman et al.1999; Rosen 2000; Young and Young 2000).That said, timing of rainfall (summer versuswinter), density dependent factors, predation,and potentially inter-specific competition canact to obscure these climatic-driven relation-ships (Turner 1977; Turner et al. 1982; Rosen2000; Young and Young 2000).

Populations of longer-lived species, such asmany larger snakes, desert tortoise (Gopherusagassizi), Gila monster (Helodermasuspectum), desert iguana (Dipsosaurusdorsalis), desert night lizard (Xantusia vigilis),and common chuckwalla (Sauromalus ater),vary less, as most adults of these species tendto survive drought, but reproductive output,growth, diet, and behavior often vary betweenwet and dry periods (Nagy 1973; Krekorian1984; Smits 1985; Rosen 2000; Averill-Murrayet al. 2002; Beck 2005). At OPCNM, Rosen(pers. comm. 2006) found that the distributionof snakes varied between dry and wet periods— in drought, snakes were typically found indrainages, where vegetation was more abun-dant and occasional runoff occurred, but inwetter periods snakes ventured more into thedrier uplands and flats between the drainages.

In contrast to the desert areas, populationdynamics of riparian, marine, agricultural, andurban herpetofauna are probably not limitedby precipitation. Historically, amphibians andreptile populations in the riparian corridors ofthe Gila and Colorado rivers probably variedwith periodic flood or low water events, theformer of which were extreme in some years(Mueller and Marsh 2002). This is much lessthe case now, as these rivers are regulated byupstream dams and at least some flow in thestudy area is maintained by irrigation returnflow, both of which tend to create much morestable conditions. Disturbance today in thesetwo rivers is more likely to occur from peri-odic fire, and in the Gila River from aboutHighway 95 to Texas Hill, by channel-clearingactivities conducted by the Wellton-MohawkIrrigation and Drainage District.

AMPHIBIAN AND REPTILEINHABITANTS

The herpetofauna of the study area (Appendix1) includes 12 amphibians (one salamander,six toads, one spadefoot, and four frogs) and58 reptiles (nine turtles and tortoises, 20lizards, and 29 snakes). One lizard, two turtles,two frogs, and one salamander are non-nativeintroductions. Two introduced species, the tigersalamander (Ambystoma mavortium) and pondslider (Trachemys scripta), are probably main-tained in the study area by continuedintroductions and may not maintain breedingpopulations.

Some species reported from the study areaare not included in the Appendix. A pair ofbreeding mud turtles (probably Arizona mudturtle, Kinosternon arizonense) was collectedat Quitobaquito, OPCNM in 1955 (Rosen andLowe 1996), and the yellow mud turtle(Kinosternon flavescens) was reported fromYuma (Stebbins 1966) and collected on theGila River near Fort Yuma (Van Denburgh1922). The Arizona mud turtle occurs onTohono O’odham Nation lands to the east ofthe study area as well as south in Sonora, andcould have conceivably existed historically inthe valley of the Río Sonoyta, Sonora (Iverson1985; Jones et al. 2007). Today, yellow mud

Rorabaugh 185

turtles occur from Cochise and Graham coun-ties in southeastern Arizona east tonorthwestern Illinois (Ernst et al. 1994;Brennan and Holycross 2006). The disjunctnature of the Yuma records suggests misiden-tification, incorrect locality data, or anintroduction (Ohmart et al. 1988; Mellink andFerreira-Bartrina 2000). In any case thisspecies and the Arizona mud turtle haveapparently not persisted in the study area andtherefore are not considered further. Americanalligators (Alligator mississipiensis) wereintroduced to the Colorado River probablybefore 1938 and persisted at least into the1950s, although no evidence of reproductionwas ever found (Glaser 1970; Vitt and Ohmart1978). I also observed an alligator at FortunaPond, Gila River near Yuma (pers. obs., 1988),but the species has not persisted and thus wasalso not included. Baegert (1952) reported“alligators” at the mouth of the ColoradoRiver in 1751, which may have been Amer-ican crocodiles (Crocodylus acutus); however,Felger (2007) believed that report was notcredible. In any case, the species is now extir-pated from the coastal waters of Sonora(Rorabaugh 2008), and is also not consideredfurther.

The information in the Appendix is basedon a variety of sources, including literature,museum collections, my own personal experi-ences, and communications with others whohave worked in the study area. The followingliterature was particularly valuable: Vitt andOhmart (1978), Lowe et al. (1986), Ohmart etal. (1988), Gonzalez-Romero and Alvarez-Cardenas (1989), Rosen and Lowe (1996),Turner et al. (1997), Rosen (2000, 2007), Steb-bins (2003), Brennan and Holycross (2005,2006), Rorabaugh (2008), and speciesaccounts in the Catalogue of AmericanAmphibians and Reptiles.

Seventy species represents substantialdiversity for an arid region of the Southwest.Arizona supports 128 species (Brennan andHolycross 2006), while Sonora (not includingislands in the Gulf of California) is home to atleast 178 species (Rorabaugh 2008). Hence,the study area includes 55% and 39% of the

number of species found in Arizona and main-land Sonora, respectively.

CONSERVATION CONTEXTAbout 50% of the 4.26 million ha (10.5 mil ac)study area is designated as biosphere reserves,national monument, or wildlife refuge,including the 0.94 million ha (2.3 mil ac) AltoGolfo Reserve, the 0.71 million ha (1.7 mil ac)Pinacate Reserve, the 0.35 million ha (0.86 milac) CPNWR, the 0.13 million ha (0.32 mil ac)OPCNM, and the 0.13 million ha (0.32 mil ac)Refugio Vaquito (part of which also lies withinthe Alto Golfo Reserve, Figure 1). The bios-phere reserves include core protection zones(Zona Nucleo in the Alto Golfo Reserve, andZonas Sierra el Pinacate and Sierra del Rosarioin the Pinacate Reserve) where protective regu-lations are very strict; the remaining areaswithin the reserves are buffers with less strin-gent regulations. These buffer zones oftenoverlay ejido (private cooperative) lands. Thestudy area also includes the 0.75 million ha(1.85 mil ac) BMGR, which is primarily amilitary aerial training range. Much of theBMGR is off limits to the public, and only2.5% of lands in the BMGR have beensubjected to moderate to high severity militaryactivities; another 7.5% have been affected toa lesser degree (U.S. Departments of the Navy,Air Force, and Interior 2006). These land usedesignations and accompanying restrictionsand regulations provide substantial protectionfor the region’s herpetofauna and its habitats.

The study area is rich biologically, andsupports many imperiled faunal and floralspecies. Marshall et al. (2000) evaluatedconservation priorities in the Sonoran Desertecoregion and found the Pinacate/OPCNM/BMGR complex to contain moreconservation targets (sensitive faunal and floralspecies) than any of the other 99 conservationsites they evaluated. The Colorado River deltaranked eighth on that list.

A major challenge to amphibian and reptileconservation has been a negative publicperception. Characteristics often associatedwith these animals, such as “venomous, slimy,cold, lowly, and creepy”, have evolved into

Grande leopard frog (Rana berlandieri), spinysoftshell (Apalone spinifera), and pond slider,are only represented as introduced species, andin the study area do not warrant protection.

The only species that may warrant specialstatus, but currently is not afforded one, is theSonoran Desert toad. Although probablysecure in the eastern portion of the study area(Rosen and Lowe 1996) and along the GilaRiver east of Dome Valley (pers. obs.), I amnot aware of any documented observations orcollections of the species along the ColoradoRiver Arizona/Mexico in the study area, Yumaarea, or Gila River west of Dome Valley sincethe 1940s (Vitt and Ohmart 1978; Jennings andHayes 1994, pers. obs.). Yet Sonoran Deserttoads may still persist in the Río Coloradovalley of Sonora (Grismer 2002; pers. comm.with resident of Ejido Johnson, Sonora, 2007)and potentially at the Foothills or FortunaWash east of Yuma based on unconfirmedreports in the 1980s.

If we remove from special status designa-tion the 15 species on the Mexican list thatappear not to be significantly threatened in thestudy area, and the four introduced species, butadd in Sonoran Desert toad, then a short list of19 (27%) species would be considered imper-iled. Grouping of these species by communitytypes — marine, riparian/wetland, desertvalleys, and desert bajadas and mountains,reveals an interesting pattern (Table 1): marineand riparian/wetland species are particularlyimperiled when compared to desert species.Five of six marine species (all sea turtles) andsix of 13 (46%) riparian/wetland species areimperiled, compared to 13% (4 of 31) of desertvalley and 13% (5 of 39) of desertbajada/montane species. Another indication ofthreats is the presence of introduced species, ofwhich five of six are riparian/wetland species.

Marine SpeciesFive species of sea turtles (Green, Hawksbill,Olive Ridley, Leatherback, and Loggerhead)are known from the Gulf of California (Table1). All but the Hawksbill have been recordedin the Upper Gulf (Grismer 2002), but it likelyvisits the study area. Only the Olive Ridley has

quite a mythology of half truths and miscon-ceptions. The end result is that many snakesare killed on sight, frogs and salamanders areavoided, and in general, amphibians andreptiles have not received the same affectionor appreciation that colorful birds, big gamespecies, sport fishes, or other charismatic orcommercially significant species are afforded.Even in the scientific and conservationcommunities, amphibians and reptiles are rela-tively unknown compared to other vertebrategroups in the study area. Negative publicperception is perhaps abating (e.g., frogs arenow well represented in popular culture), butstill remains a problem for conservation.

IMPERILED SPECIESThirty-seven of 70, or 53% of amphibian andreptiles species in the study area, are consid-ered special conservation status species underthe U.S. Endangered Species Act (8 species),Mexican regulations (36 species: ‘Lista deEspecies en Riesgo”, SEMARNAT 2008), orare on the IUCN’s Redlist (8 species,Appendix). Some species on the Mexican listare either introductions or their populationsappear fairly secure in the study area. Most ofthese are “Pr” species, which is a categoryindicating possible threat, but not enoughinformation is available to list the species asthreatened or endangered. For instance, thereis no indication that the western banded gecko(Coleonyx variegatus, “Pr”), zebra-tailed lizard(Callisaurus draconoides, “A” or threatened),long-nosed leopard lizard (Gambeliawislizenii, “Pr”), common kingsnake (Lampro-peltis getula, “A”), coachwhip (Masticophisflagellum, “A”), nightsnake (Hypsiglenachlorophaea, “Pr”), and all six of therattlesnakes (all “Pr”) are threatened or endan-gered in any substantial way in the study area.The variable sandsnake (Chilomeniscusstramineus, “Pr”), saddled leaf-nosed snake(Phyllorynchus browni, “Pr”) and Sonorancoralsnake (Micruroides euryxanthus, “A”),although of limited distribution in the studyarea, are additional species that are probablynot threatened in the study area. Four “Pr”species, including the tiger salamander, Rio

186 Rorabaugh

Table 1. “Short list” of imperiled species by community type.

Community1 Imperiled Species Total No. Species % Imperiled % IntroducedMarine Green Sea Turtle, Hawksbill Sea Turtle,

Loggerhead Sea Turtle, Olive Ridley Sea Turtle, Leatherback Sea Turtle 6 83 0

Riparian/Wetland Sonoran Desert Toad, Lowland Leopard

Frog, Sonoran Mud Turtle, Black-necked Gartersnake, Mexican Gartersnake, Checkered Gartersnake 13 46 42

Desert Valleys Sonoran Green Toad, Western Narrow-

mouthed Toad, Flat-tailed Horned Lizard, Yuman Fringe-toed Lizard 31 13 3

Desert Bajada/Mountain Desert Tortoise, Gila Monster, Common

Chuckwalla, Rosy Boa, Sonoran Desert Toad 39 13 01 Communities where species primarily occur or are most commonly encountered.

Since 1991 all sea turtles in Mexico and itswaters have been protected by presidentialdecree; however, the law is difficult to enforceand some poaching still occurs. For such prohi-bitions to be effective, an understanding of theneed for conservation must be promoted at thecommunity level. Several organizations aredoing just that. Drs. Gary Nabhan and JeffreySeminoff and others have worked closely withthe Seri (Coomcáac) people on the north-central Sonoran coast to recruit the Seri andlocal fisherman into sea turtle conservation. Atraining session co-taught by biologists andSeri elders for Seri youth “para-ecologists”introduced conservation techniques and simplemonitoring protocols to promote conservationof the Gulf’s sea turtles (Nabhan et al. 1999).Similarly, on the Baja peninsula, GrupoTortuguero de las Californias — an alliance ofcommunities, fishing cooperatives, NGOs,tourism outfitters, scientists, government agen-cies, and others — is working to promote theecological, economic, and cultural role of seaturtles in the Gulf of California. CEDO (ElCentro Intercultural de Estudios de Desiertosy Océanos) located in Puerto Peñasco, hasdeveloped educational and outreach programs

been known to nest in the Upper Gulf; Nabhan(2003) and Peggy Turk-Boyer (pers. comm.2006) report nesting on beaches near PuertoPeñasco in the 1990s, and Seminoff andNichols (2007) report evidence of nesting nearEl Golfo de Santa Clara in 2000. All fivespecies are listed as threatened, endangered, orcritically endangered on the U.S., Mexican,and IUCN lists, and all have declined in theGulf of California. A long history of harvestingturtles, legally and illegally, and unintentionalbycatch in nets and on longlines in the Gulf ofCalifornia have contributed to that decline. Butthe threats to sea turtles are neither unique tothe study area nor to the Gulf of California, asall five species are distributed throughout thewarmer waters of the world, and turtles thatvisit the Gulf come from as far away assouthern Mexico and Japan. Throughout theirdistributions, the Gulf’s sea turtles are alsodying from pollution-related disease; ingestionof debris and trash; intentional harvest ofturtles for food, aphrodisiacs, and the jewelryand souvenir industry; incidental capture innets; boat strikes; destruction or disturbance ofbeaches where sea turtles nest; and predationof nests by people and dogs.

Rorabaugh 187

1991; Mellink and Ferreira-Bartrina 2000;Mueller and Marsh 2002).

Although it is impossible to determinecauses after the fact, predation by non-nativeAmerican bullfrogs, potentially spiny softshellturtles, and a diversity of non-native fishes, isa likely cause of the demise of the lowlandleopard frog, Mexican gartersnake, andSonoran mud turtle on the lower Colorado andGila Rivers. American bullfrogs and a varietyof non-native fishes eat gartersnakes, frogs,and turtles, and each one alone may be capableof eliminating populations of lowland leopardfrogs, Sonoran mud turtles, and Mexicangartersnakes (Hayes and Jennings 1986; Rosenand Schwalbe 1988, 2002). Red swamp cray-fish (Procambrus clarkia) and Rio Grandeleopard frogs were likely introduced after theextirpation of these native species (Inman etal. 1998; Platz et al. 1990), and thus did notcontribute to their demise.

In addition to the many introductions ofnon-native predators, wetland habitats on theColorado and Gila Rivers have been lost anddegraded due to upstream dams and diver-sions, levees and bankline structures to controlflows and flooding, introductions of non-nativeplants, groundwater pumping, and return agri-cultural flow, which is typically saline andcarries with it pesticides and fertilizers(Ohmart et al. 1988; Garcia-Hernandez et al.2001; Glenn et al. 2001).

Restoration of flows and some aspects ofnative wetland and riparian communities arepossible along the big rivers (Ohmart et al.1988; Pitt 2001). When completed, the YumaEast Wetlands Park, which straddles theColorado River just east of Yuma, will encom-pass about 565 ha (1,400 ac) of wetlands andnative riparian communities. Another 55 ha(135 ac) park and restoration area — “YumaWest Wetlands” — is close to completionnearby. These projects are excellent examplesof restoration potential. Similarly, wetlands andriparian restoration has been accomplished atseveral sites along the Gila River in the studyarea with varying success (Rorabaugh 1995),and wetland and riparian restoration projectsare occurring as well in the delta and on the

and promotes conservation of Gulf of Cali-fornia resources and species, including seaturtles. In addition, staff at the Alto GolfoReserve work with fisherman and localcommunities to develop an understanding ofthe need for conservation of sea turtles andother species of the upper Gulf.

Riparian and Wetland SpeciesThirteen species are obligate or primarilywetland or riparian species (Appendix, Table1), including tiger salamander, Woodhouse’stoad, Sonoran Desert toad, Great Plains toad,three ranid frogs, three turtles, and all threegartersnakes. Note that in the study area theblack-necked gartersnake is only known frommesic canyons of the Ajo Mountains. All of theextant wetland species, except the black-necked gartersnake, can be found inagricultural ditches, canals, and fields, as wellas along rivers or backwaters. The SonoranDesert toad and occasionally Great Plains toadare also found at stock tanks, tinajas, and rainpools in the desert in the eastern portion of thestudy area.

Five of the 13 species are introduced, andof the other eight, 6 are considered imperiled(I included the Sonoran Desert toad). Two ofthe six imperiled species are almost certainlyextirpated — Mexican gartersnake andlowland leopard frog — which have not beenrecorded in the study area since 1890 and1942, respectively. These species were notfound in recent investigations: Vitt and Ohmart(1978), Rosen and Schwalbe (1988), Clarksonand Rorabaugh (1989), Rorabaugh et al.(2002b). Nor have Sonoran mud turtles beenfound in the region since 1962, except on theRío Sonoyta, where they still persist. TheSonoran Desert toad is extirpated from theU.S. portion of the Colorado River, but maystill persist along that river in Mexico (Grismer2002), and still occurs in ephemeral waters andthe Gila River corridor in the eastern portionof the study area. The high percentage of intro-duced species and imperilment of nativewetland and riparian species is consistent withthe status of other wetland and riparian speciesgroups (Ohmart et al. 1988; Rosenberg et al.

188 Rorabaugh

the ejido will use all or portions of the outflowfor agriculture, with resulting reductions inriverine aquatic habitat. See Rosen et al. (thisvolume) for further information.

Desert Valley SpeciesThirty-one species are considered inhabitantsof desert valleys and flats. Of those, four areon the list of imperiled species (Table 2),including Sonoran green toad, western narrow-mouthed toad, Yuman fringe-toed lizard, andflat-tailed horned lizard. The Sonoran greentoad is known only from a few localities fromWhy to Lukeville and breeding sites have notyet been documented (Rosen and Lowe 1996).The western narrow-mouthed toad is knownonly from an individual that was calling atLukeville (Rosen and Lowe 1996). Both areexpected in the Río Sonoyta Valley. In thestudy area, these species are at the westernextremes of their ranges. Breeding sites mayinclude cattle tanks or other artificial impound-ments, as well as playas and intermittentreaches of the Río Sonoyta. A key conserva-tion need for these species is locating andprotecting any breeding sites in the study area.

The range of the Yuman fringe-toed lizardlies mostly within the study area, although italso occurs south along the coast from BahiaSan Jorge to Bahia Tepoca. It is associated withand highly adapted to living in dunes, but Ihave also observed it in windblown sandy flatsin the Yuma Desert. This species and its habitatare vulnerable to habitat degradation. In theAlgodones Dunes, Imperial County, Cali-fornia, Luckenbach and Bury (1983)documented declines in plants, arthropods,mammals, and lizards (including the closely-related Colorado Desert fringe-toed lizard,Uma notata) in areas with relatively low levelsof OHV use. In heavily used areas, virtuallyno plants or animals existed. However, thelargest tracts of habitat for the Yuman fringe-toed lizard in the study area are either closedto public use, or vehicles are limited to existingroutes. In Arizona, no off-highway vehicle(OHV) activity is allowed on the MohawkDunes, and the Yuma Dunes (on the BMGR)are in an area where public use is prohibited.

Río Hardy (J. Campoy, pers. comm. 2006).However, these projects probably do notbenefit imperiled amphibians and reptilesbecause wetlands are rapidly colonized bynon-native predators. Loss of native herpeto-faunal diversity on the Colorado and GilaRivers may be irreversible, although there issome potential for establishing highlymanaged and isolated refugia populations oflowland leopard frogs, Sonoran mud turtles, orother species. The recently completed LowerColorado River Multi-Species ConservationPlan may provide opportunities for estab-lishing such refugia — the Sonoran Deserttoad and lowland leopard frog are consideredevaluation species under the Plan (LowerColorado River Multi-Species Program 2004).

In contrast to the situation on the big rivers,the Río Sonoyta and associated Quitobaquito,Quitovac, and sewage treatment ponds atSonoyta present an opportunity to conserveand potentially restore a desert streamcommunity. The Quitobaquito/Río SonoytaWorking Group, a bi-national coalition ofagencies and biologists, has developed aproposal for comprehensive conservation forwhich they are seeking funding. The proposalincludes community-based, ecologicallysound planning to develop: (1) ecotourism andan urban park along the stream corridor, (2)aquifer conservation or restoration, (3)modernized infrastructure planning regardingsewage treatment, (4) landscape-level conser-vation planning and urban design, and (5)monitoring and educational programs for keyresources. This proposal would help conservethe Sonoran (Sonoyta) mud turtle (Kinos-ternon sonoriense longifemorale) and othernative species. A recent threat to the RíoSonoyta is a planned modernized sewagetreatment plant to be located on ejido landswest of the town of Sonoyta. The plant wouldreplace the existing sewage lagoons, whichprovide habitat for mud turtles. As of thiswriting, the disposition of the treated waterthat would exit the plant is uncertain. Ifreturned to the river, it could bolster flows andbenefit mud turtles and other native aquaticspecies; however, there is some potential that

Rorabaugh 189

stated by the courts on three occasions. Theproposed rule was withdrawn, in part, due tothe development of a multi-party conservationagreement and strategy, in which 196,425 ha(485,000 ac) were designated in five Manage-ment Areas, one in Arizona and four inCalifornia, to be managed for the long-termviability of the flat-tailed horned lizard(Rorabaugh et al. 2000; Flat-tailed HornedLizard Interagency Coordinating Committee2003). Implementation of the conservationstrategy has been good, as documented byannual progress reports.

Despite recent reductions in illegal borderactivities, habitats along the entire U.S.-Mexico border in the Arizona portion of thestudy area are threatened by these activitiesand law enforcement response (Milstead andBarnes 2002; Kralovec 2006; Segee andNeeley 2006; Cordova and de la Parra 2007).The pedestrian fence that runs through flat-tailed horned lizard habitat east of San Luisprobably is, to some extent, a barrier to move-ment of lizards across the border, but it isapparently not an absolute barrier. In May2008, I observed several miles of this fencethrough the Yuma Dunes that had been under-mined by wind erosion. Animals, and in somecases people, could pass under the fence,which posed no barrier to a flat-tailed hornedlizard. Even in places where the fence was notundermined, I observed western whiptailsrunning into burrows at the base of the fencethat appeared to provide access to the otherside. The fence has apparently reduced illegalcross-border traffic and associated OHVactivity in the Yuma Desert Management area(Allen and Rorabaugh 2007).

Construction of the Area Service Highwaywas completed in 2009 along the western andnorthern boundaries of the Yuma DesertManagement Area from a new port of entry atAvenue E on the international border to ArabyRoad at Interstate 8. It eliminated about 283 ha(699 ac) of flat-tailed horned lizard habitatalong the road corridor, and isolated another1,480 ha (3,657 ac) (U.S. Fish and WildlifeService files). Although the project impactedhabitat, it only represented a loss of about 2.7%

Military activities have had little or no impacton these areas, and Border Patrol or smugglervehicle tracks are uncommon (pers. obs.). AtPinta Sands on CPNWR, a road traverses thenorthern end of the dunes and lava flow, butno vehicles are allowed off-road (althoughBorder Patrol and smugglers drive off-road inthis area). A dune south of County 19th Streetand west of Avenue 4E near Yuma, where Ihave observed fringe-toed lizards, has beendeveloped into a sand and gravel pit and is alsoimpacted by OHVs and trash dumping;however, in regard to impacts to dune habitats,this is the exception to the rule. In Sonora, thelargest and most extensive dunes lie within thePinacate and Alto Golfo reserves. IncreasingOHV activity in the Puerto Peñasco/ChollaBay area impacts some fringe-toed lizardhabitat, and a recently completed highwayfrom Peñasco to El Golfo de Santa Clara isbringing additional visitors, their OHVs, andassociated recreational impacts to remoteportions of the Gran Desierto.

The flat-tailed horned lizard is known frommostly sandy flats and low dunes. In Arizonait occurs in the Yuma Desert west of the ButlerHills and south of the Gila River (Rorabaughet al. 1987), and in Sonora from San Luis RíoColorado south and east to the Gulf of Cali-fornia and Bahia San Jorge (Rodriguez 2002;Rorabaugh 2008). Piest and Knowles (2002)estimated 89,522 ha (221,214 ac) of habitatexisted historically in Arizona, of which25,190 ha (62,246 ac, 28%) have been lostprimarily to agricultural and urban develop-ment. No recent estimates of habitat loss areavailable for Sonora; however, Johnson andSpicer (1985) estimated 14% of the habitat inSonora was threatened by human activities.Their estimate of occupied habitat in Sonorawas conservative, hence the percentage of thelizard’s habitat that was threatened at that timewas probably an overestimate.

In the U.S., the status of the flat-tailedhorned lizard has been the subject of muchrule-making and litigation. It was firstproposed for threatened status under theEndangered Species Act in 1993. Theproposed rule has been withdrawn and rein-

190 Rorabaugh

Additional non-native plants, such as Russianthistle (Salsola tragus), the perennial buffel-grass (Pennisetum ciliare), and many othersare common locally, particularly in disturbedareas with relatively fertile soils (Brooks 1999;Wilson et al. 2002).

Probably the greatest effect non-nativeplants can have on the herpetofauna isincreasing the fire frequency with subsequentmajor effects to vegetation communities.Historically, fire occurred only very rarely inSonoran desertscrub (Schmid and Rogers1988). Many Sonoran Desert plants are poorlyadapted to fire and are readily killed.Desertscrub communities that burn, and partic-ularly if they burn repeatedly, may takedecades or centuries to recover (Schmid andRogers 1988; Schwalbe et al. 2000; Narog andWilson 2003), if such recovery is possible. Inthe spring of 2005, after the previous winter’sluxurious growth of annual plants dried out,fires scorched at least 25,500 ha (63,010 ac)on the BMGR (Luke Air Force Base 2005) and2,025 ha (5,000 ac) on CPNWR (C.McCasland, pers. comm. 2006). Fires burnedespecially hot in desert washes, killing paloverdes (Parkinsonia floridum and P. micro-phylla) and other trees, as well as saguaros(Carnegiea gigantea). Removing trees fromthis community would likely reduce or elimi-nate tree-dwelling species, such as ornate treelizards (Urosaurus ornatus) and desert spinylizards (Sceloporus magister), but would alsoreduce the availability of relatively mesicmicrosites that may be important for lizard andsnake survival during drought (see discussionabove). Desert Bajada and Mountain Species

Five imperiled species occur primarily indesert mountain or bajada habitats, includingdesert tortoise, Gila monster, common chuck-walla, rosy boa, and Sonoran Desert toad(Appendix, Table 1). The latter species wasdiscussed above under Riparian and WetlandSpecies, and is likely fairly secure where itoccurs as a montane or bajada species in theeastern study area (Rosen and Lowe 1996).The common chuckwalla is widespread in

of current habitat in Arizona, and the highwayand its right-of-way fence could serve as adefensible barrier against OHV and other intru-sions into the Management Area.

In Sonora, much of the habitat of the flat-tailed horned lizard lies within the Pinacateand Alto Golfo reserves, although there issignificant habitat north of the latter reserveand west of the former reserve, as well ashabitat in the southeastern portion of the studyarea that is unprotected. As discussed for theYuman fringe-toed lizard, OHV activity isoccurring in the Puerto Peñasco/Cholla Bayarea, and such activity is likely to increase inthe Gran Desierto with the recent completionof the Peñasco to El Golfo de Santa Clarahighway. Another recently completed highway,from San Luis Río Colorado to roughly ElDoctor, also eliminated habitat and providesnew access to the Gran Desierto. These typesof activities degrade habitat and result inmortality of lizards (Flat-tailed Horned LizardInteragency Coordinating Committee 2003).Due to road mortality, there will likely be a“dead zone” along the new highways, in whichdensities of reptiles are much reduced andsome species may be absent (Boarman et al.1992; Rosen and Lowe 1994). Grant et al.(2001) found 87 percent fewer flat-tailedhorned lizards within 0.72 km (0.45 mi) ofHighway 98 in California than in areas fartherfrom the road.

A growing threat to valley species is inva-sion of non-native plants, particularly winterannuals. Mediterranean grass (Schismusarabicus and S. barbatus) is widespread andoften abundant, and was first recorded inArizona in the 1920s and ‘30s (Felger 1990).Sahara mustard (Brassica tournefourtii), amore recent introduction (Felger 2000), isoften common in sandy flats and dunes. Thisspecies covered the Mohawk Dunes andsurrounding areas during the wet winter of2004-2005, and appears to be on the increasein the Yuma area. Also during the winter of2004-2005, salad rocket (Eruca vesicaria) wasthe dominant vegetation in a swath along Inter-state 8 from Gila Bend to west of Sentinel (~53km), and was also prominent near Sonoyta.

Rorabaugh 191

plans, but its management guidance is notbinding. A conservation agreement andstrategy is in development for the Sonoranpopulation in Arizona. If adopted, signators tothe agreement would commit to implementingthe guidance in the conservation strategy.

The common chuckwalla, the Gila monster,and rosy boa are desirable to herpetoculturists,and are often collected or poached. Gilamonsters can bring $1,200 to $1,500 apiece onthe black market. In Arizona, it is illegal underthe Arizona Game and Fish Commission Order43 to collect or possess a Gila monster;however, under the same order, limitednumbers of common chuckwallas and rosyboas may be collected with a valid huntinglicense. No collection of these species isallowed in Sonora without a special permit.Goode et al. (2005) documented collection-related habitat destruction and declines incommon chuckwalla populations in areasfrequented by collectors on South Mountainnear Phoenix. However, common chuckwallapopulations persist and have been stable in fivePhoenix mountain parks despite four of thembeing relatively small (less than 1,500 ha, or3,706 ac), surrounded by urban development,and subjected to substantial recreational useand access (Sullivan and Flowers 1998;Sullivan and Sullivan 2008). Rosy boas maybe impacted by collecting (Fisher 2003), whichoften occurs along paved roads through theirhabitat. However, even in heavily collectedareas, such as Whitewater Canyon in RiversideCounty, California, rosy boas still persist(Spillman 2006). Hence, it may be difficult toextirpate a population of common chuckwallasor rosy boas via collecting or poaching alone,and in the remote regions of the study area,populations are probably relatively safe fornow. Although some poaching occurs, collec-tion is not known to be a significant threat tothe Gila monster in the study area. Further-more, the species is likely most abundant inprotected areas, such as OPCNM.

Over a four-year period, Rosen and Lowe(1994) recorded snake mortality along a 44.1km section of Route 85, which runs along themiddle bajada through OPCNM from Why to

montane areas, whereas the rosy boa is limitedto mountains in the northeastern portion of thestudy area, the Gila Mountains, and perhapselsewhere. The Gila monster and deserttortoise are more common in the study area’seastern portion, but occur as far west as theGila Mountains (Appendix).

In general, montane habitats are much lessaffected by anthropogenic activities than desertflats and valleys. Urban and agricultural devel-opment, roads, OHV and most otherrecreational pursuits, military training, utilitycorridors, and other activities typically occurin the flats, rather than on hillsides or bajadasheavily dissected by drainages. Fire also doesnot burn as readily in montane habitats, wherefuels are often discontinuous because of barrenrock outcrops. The 2005 fires on the BMGRand CPNWR generally burned around moun-tain ranges and only rarely burned onto slopes.

The desert tortoise has been the subject ofmuch conservation planning and implementa-tion for more than two decades in Arizona(Howland and Rorabaugh 2002). The deserttortoise, including the Sonoran population(south and east of the Colorado River) and theMojave population (north and west of theColorado River), was petitioned for listingunder the U.S. Endangered Species Act in1985 and 1989, but only the Mojave popula-tion was listed as a threatened species (1990).The Sonoran population was petitioned againin 2008; and in 2009, the U.S. Fish andWildlife Service found that it may warrantlisting as a threatened or endangered species.The Mojave population occurs primarily inflats and valleys, while the Sonoran populationoccurs on bajadas and montane slopes.

The Arizona Interagency Desert TortoiseTeam, which has included representatives fromseveral land management agencies in the studyarea, developed a “Management Plan for theSonoran Population of the Desert Tortoise inArizona, 1996,” which provides a number ofmanagement options for minimizing effects ofhuman activities on the desert tortoise and itshabitat, as well as research and monitoringrecommendations. The plan is used by landmanagers in the development of land use

192 Rorabaugh

Tortuguero de las Californias, and CEDO, aswell as outreach carried out by staff of the AltoGolfo Reserve, are models for this kind ofwork and should be supported and expanded.

2) Conserve the wetlands of the RíoSonoyta Valley. Unlike the Colorado and GilaRivers, threats to the wetlands of the RíoSonoyta are manageable. Proposals by the RíoSonoyta/Quitobaquito Working Group shouldbe funded and implemented. Furthermore,U.S. interests should continue to work with thePinacate Reserve and others to find designs forthe new wastewater treatment plant that willmaintain flows in the Río Sonoyta and facili-tate conservation of the sensitive species thatdepend upon it.

3) Breeding habitats of the Sonoran greentoad and western narrow-mouthed toad shouldbe located and protected.

4) Thorough surveys for special statusriparian and wetland species should beconducted in the Río Colorado Valley. Thestatus of the Sonoran Desert toad in theFoothills and Fortuna Wash should also beevaluated. Any breeding sites of this speciesshould be protected.

5) Opportunities should be sought with theAlto Golfo Reserve in Sonora and through theLower Colorado River Multi-Species Conser-vation Plan in Arizona to develop refugiapopulations of special status amphibians andreptiles of the Colorado and Gila Rivers,including those that have been extirpated.These managed wetland sites could potentiallyserve as refugia for a variety of native fishes,as well as native turtles, snakes, and amphib-ians.

6) Fires in Sonoran desertscrub should beaggressively suppressed, and feasible controlsfor non-native invasive plants should be devel-oped and implemented. Aggressivesuppression of fires can be an expensivemanagement option, and may not be feasiblein Sonora. Some plants with localized distri-butions in the study area, such as buffelgrass,can be mechanically controlled (Rutman andDickson 2002); however, others will require

Lukeville. The number of snakes found deadwas equivalent to the estimated snake popula-tion in a 5.0 km2 (1.9 mi2) area, or toeliminating all snakes within 65 m (213 ft) ofthe road. No rosy boas were found on the roadduring the study, but the authors presentedevidence of it being collected on the road inprevious decades. They suggested that itspopulations adjacent to the road had declinedsubstantially due to road mortality.

CONCLUSIONSThe following conclusions target conservationpriorities for imperiled amphibians andreptiles, but also address threats to the fourmajor herpetofaunal communities in the studyarea. To maintain biodiversity, conservationoften must focus on specific at-risk species andthe sometimes very specialized habitats inwhich they occur. But taking a broader conser-vation perspective, an entire suite of speciescan be protected by targeting ecologicalcommunities. The most imperiled herpeto-fauna in the study area occur in communitiesthat also support a variety of other sensitive orendangered plants and animals. If thesecommunities can be protected, many of theirimperiled species (herpetofauna and others)will be protected as well.

Recommended Actions1) Conserve sea turtles in the Upper Gulf.

Ultimately, if worldwide efforts to protect seaturtles are successful, all five species will prob-ably continue to visit the Upper Gulf. Ourfocus locally should be to minimize or elimi-nate poaching and incidental bycatch, and tomonitor sea turtle populations. Althoughsignificant threats remain, conservation of seaturtles in the Upper Gulf is achievable andmuch work has been accomplished towardsthat goal (Seminoff and Nichols 2007). Fishingregulations protective of sea turtles should beaggressively enforced in the Gulf of California,but there is also a critical need to work withlocal fishing communities and ejidos to buildan understanding of the need for sea turtleconservation at the community level. The workof Drs. Nabhan and Seminoff, Grupo

Rorabaugh 193

(including the implementation of plans andagreements), extent and trends in humanimpacts, climate change, and changes inbiotic communities across the study areawould be desirable. Feasible, consistent, andrepeatable monitoring strategies that meet theneeds of differing jurisdictions and planscould potentially be developed by the bi-national working group proposed inRecommended Action 9).

12) Conduct research to identify new toolsfor conservation or to improve the effective-ness of existing tools. In particular, newmethods are needed to control non-native plantinvasions in desertscrub. If effective means canbe developed to control crayfish or bullfrogsin complex aquatic systems, some wetlandhabitats could potentially be restored for nativeamphibians and reptiles.

ACKNOWLEDGMENTSA thorough understanding of species’ occur-rences, habitat use, distributions, andconservation needs and threats would not havebeen possible without the help of numerousbiologists and museums who readily shareddata and their knowledge of the study area.Specimens and accompanying data in thefollowing museums were crucial in clarifyingspecies presence and distribution in Sonora:University of Arizona, Arizona State Univer-sity herpetology collection, CaliforniaAcademy of Science (including StanfordUniversity collections), Los Angeles CountyMuseum, Museum of Vertebrate Zoology,University of Kansas Natural HistoryMuseum, and the Field Museum of NaturalHistory. The following individuals providedinformation about one or more species: GeorgeBradley, Jose Campoy, Charles Conner, MarkFisher, Martha Gomez, Andy Holycross, MattJohnson, Ami Pate, Lin Piest, RamsesRodriguez, Phil Rosen, Jeffrey Seminoff, TimTibbitts, and Peggy Turk-Boyer. MaryRichardson prepared Figure 1. Sherry Barrett,Cecil Schwalbe, Bill Halvorson, and twoanonymous reviewers reviewed a draft of thispaper and provided valuable comments.

biological or other controls. Fire enhanced byinvasive plants may be the most serious threatto biotic communities in desert valleys, and yetcurrent control techniques are expensive andrelatively ineffective or temporary.

7) Continue implementation of the Flat-tailed Horned Lizard Rangewide ManagementStrategy and Agreement, with an emphasis onbi-national coordination and cooperation.Protection of areas for the flat-tailed hornedlizard also benefits other valley species, suchas the Yuman fringe-toed lizard. An attemptshould be made to bring the Border Patrol intothe agreement as a signatory.

8) Finalize and implement the conserva-tion agreement and strategy for the Sonoranpopulation of the desert tortoise. This agree-ment and strategy should be a bi-nationaldocument.

9) Develop a bi-national working group tocoordinate herpetofaunal conservation, moni-toring, and research across jurisdictionalboundaries north and south of the border. Thisgroup could perhaps examine herpetofaunalconservation needs and targets more thor-oughly than I did here, and develop furtherrecommendations for conserving the biodi-versity of the area. This group could be asubcommittee of Partners in Amphibian andReptile Conservation (PARC), and manage-ment recommendations could tier from andelaborate on PARC’s “Habitat ManagementGuidelines for Amphibians and Reptiles of theArid Southwest” (Woods et al. 2004). Thisgroup could also seek funding for key projectsand conduct outreach to communities andstakeholders to build support for conservationactions.

10) Continue and strengthen managementof protected areas. Additional staff, funding,and resources are needed in all protected areasto fully implement management plans, conductpublic outreach, and enforce regulations.

11) Presence and distribution of the imper-iled amphibians and reptiles listed in thischapter should be monitored. Additionalmonitoring of conservation management

194 Rorabaugh

Environment and Natural Resources, NationalInstitute of Ecology, El Colegio de la FronteraNorte, and Southwest Consortium for Environ-mental Research and Policy. Tlalpan, Mexico,D.F. 206pp.Cornelius, S. 1998. Fragmentation of naturalresource management in the Sonoran Desert:Issues and resolution. The Sonoran Institute,Tucson, Arizona. 57pp.Crother, B. I. 2008. Scientific and standard Englishnames of amphibians and reptiles of NorthAmerica north of Mexico, with commentsregarding confidence of our understanding. Sixthedition. Society for the Study of Amphibians andReptiles, Herpetological Circular No. 37.Dickman, C. R., M. Letnic, and P. S. Mahon. 1999.Population dynamics of two species of dragonlizards in arid Australia: The effects of rainfall.Oecologia 119: 357–366.Dimmitt, M. A. 2002. Biomes and communities ofthe Sonoran Desert region. In A natural history ofthe Sonoran Desert, edited by S. J. Phillips, P. W.Comus, pp. 3–18. Arizona-Sonora DesertMuseum Press, Tucson, AZ; and University ofCalifornia Press, Berkeley, CA.Ernst, C. H., R. W. Barbour, and J. E. Lovich. 1994.Turtles of the United States and Canada. Smith-sonian Institution Press, Washington D.C. 578pp.Felger, R. S. 1990. Non-native plants of Organ PipeCactus National Monument. University ofArizona, Cooperative National Park ResourceStudies Unit, Technical Report 31: 1–93.Felger, R. S. 2000. Flora of the Gran Desierto andRío Colorado of northwestern Mexico. Universityof Arizona Press, Tucson, AZ. 673pp.Felger, R. S. 2007. Living resources at the center ofthe Sonoran Desert. In Dry borders, great naturalreserves of the Sonoran Desert, edited by R.Felger and B. Broyles, pp. 147–192. University ofUtah Press, Salt Lake City, UT.Fisher, R. N. 2003. Population status and conserva-tion of the rosy boa (Lichanura trivirgata). U.S.Geological Survey, Western Ecology ResearchCenter, San Diego Field Station.http://www.werc.usgs.gov/sandiego/ boas.html.Flat-tailed Horned Lizard Interagency CoordinatingCommittee. 2003. Flat-tailed horned lizardrangewide management strategy. 78pp. + appen-dices.Fisher, R. N. 2003. Population status and conserva-tion of the rosy boa (Lichanura trivirgata). U.S.Geological Survey, Western Ecology ResearchCenter, San Diego Field Station.http://www.werc.usgs.gov/sandiego/boas. html.Fradkin, P. L. 1996. A river no more, the ColoradoRiver and the West. University of California Press,Berkeley, CA. 360pp.Garcia-Hernandez, J., K. King, A.Velasco, E.Shumilin, M. Mora, and E. Glenn. 2001. Sele-nium, selected inorganic elements, andorganochlorine pesticides in bottom material and

LITERATURE CITEDAllen, T., and J. Rorabaugh. 2007. Human distur-bance in the flat-tailed horned lizard Yuma DesertManagement Area. Marine Corps Air Station,Yuma, AZ.Anderson, R. A. 1994. Functional and populationresponses of the lizard Cnemidophoris tigris toenvironmental fluctuations. American Zoologist34: 409–421.Averill-Murray, R. C., P. Woodman, and J. M.Howland. 2002. Population ecology of theSonoran Desert tortoise in Arizona. In TheSonoran desert tortoise, natural history, biology,and conservation, edited by T. R. Van Devender,pp. 109–134. Arizona-Sonora Desert MuseumPress, Tucson; and University of California Press,Berkeley, CA.Baegert, J. J. 1952. Observations in Lower Cali-fornia: Translated from the original German withan introduction and notes by M. M. Brandenburgand C. L. Baumann. University of California,Berkeley, CA.Beck, D. D. 2005. Biology of Gila monsters andbeaded lizards. University of California Press,Berkeley, CA.Boarman, W. I, M. Sazaki, K. H. Berry, G. Goodlet,B. Jennings, A. P. Woodman. 1992. Measuringeffectiveness of a tortoise-proof fence andculverts. Status report from the first field season.In Proceedings of the 1992 Desert TortoiseCouncil Symposium, edited by K. R. Beaman, pp.126–142. Border Environment Cooperation Commission.2005. Air quality and street paving project inSonoyta, Sonora. http://www.cocef.org/aproyectos/469_PCD_Sonoyta_ Eng.pdf.Bowers, J. E. 1998. Dune country. University ofArizona Press, Tucson, AZ. 156pp.Brennan, T. C., and A. T. Holycross. 2005. A fieldguide to amphibians and reptiles of MaricopaCounty. Arizona Game and Fish Department,Phoenix, AZ. 68pp. Brennan, T. C., and A. T. Holycross. 2006. Amphib-ians and reptiles in Arizona. Arizona Game andFish Department, Phoenix, AZ. 151pp.Brooks, M. L. 1999. Habitat invasibility and domi-nance by alien annual plants in the westernMojave Desert. Biological Invasions 1: 325–337.Brown, D. E., and C. H. Lowe. 1994. Bioticcommunities of the Southwest. University of UtahPress (map). Clarkson, R. W., and J. C. Rorabaugh. 1989. Statusof leopard frogs (Rana pipiens Complex) inArizona and southeastern California. The South-western Naturalist 34(4): 531–538.Comrie, A. C., and B. Broyles. 2002. Variability andspatial modeling of fine-scale precipitation datafor the Sonoran Desert of southwest Arizona.Journal of Arid Environments 50: 573–592.Cordova, A., and C. A. de la Parra (eds.). 2007. Abarrier to our shared environment. Secretariat of

Rorabaugh 195

biota from the Colorado River delta. Journal ofArid Environments 49: 65–89. Glaser, H. S. R. 1970. The distribution of amphib-ians and reptiles in Riverside County, California.Riverside Museum Publications, Natural HistorySeries 1: 1–140.Glenn, E. P., F. Zamora-Arroyo, P. L. Nagler, M.Briggs, W. Shaw, and K. Flessa. 2001. Ecologyand conservation biology of the Colorado RiverDelta, Mexico. Journal of Arid Environments 49:5–15.González-Romero, A., and S. Alvarez-Cárdenas.1989. Herpetofauna de la region del Pinacate,Sonora, Mexico: Un inventario. The SouthwesternNaturalist 34(4): 519–526.Goode, M. J., W. C. Horrace, M. J. Sredl, and J. M.Howland. 2005. Habitat destruction by collectorsassociated with decreased abundance of rock-dwelling lizards. Biological Conservation 125:47–54.Grant, T., M. McGrann, J. Neilans, and G. Wright.2001. A mark-recapture pilot study of the flat-tailed horned lizard, Phrynosoma mcallii, in theYuha Basin of southeastern California. Bureau ofLand Management, El Centro, CA.Grismer, L. L. 2002. Amphibians and reptiles ofBaja California, including its Pacific Islands andthe Islands of the Sea of Cortéz. University ofCalifornia Press, Berkeley, CA. Hayden, J. D. 1998. The Sierra Pinacate. Universityof Arizona Press, Tucson, AZ. 87pp.Hayes, M. P., and M. R. Jennings. 1986. Decline ofranid frog species in Western North America: Arebullfrogs (Rana catesbeiana) responsible? Journalof Herpetology 20(4): 490–509.Holcombe, T. L., T. Ley, and D. A. Gillette. 1997.Effects of prior precipitation and source area char-acteristics on threshold wind velocities forblowing dust episodes, Sonoran Desert 1948–78.Journal of Applied Meteorology 36: 1160–1175. Howland, J. M., and J. C. Rorabaugh. 2002. Conser-vation and protection of the desert tortoise inArizona. In The Sonoran desert tortoise, naturalhistory, biology, and conservation, edited by T. R.Van Devender, pp. 334–354. Arizona-SonoraDesert Museum Press, Tucson, AZ; and Univer-sity of California Press, Berkeley, CA.Inman, T. C., P. C. Marsh, B. E. Bagley, and C. A.Pacey. 1998. Survey of crayfishes of the GilaRiver basin, Arizona and New Mexico, with noteson occurrences in other drainages and adjoiningstates. Report to U.S. Department of Interior,Bureau of Reclamation, Phoenix, AZ.Iverson, J. B. 1985. Status of the Arizona mud turtleKinosternon flavescens arizonense in Arizona andSonora. Report to the Office of EndangeredSpecies, U.S. Fish and Wildlife Service,Albuqurque, NM. Jennings, M. R., and M. P. Hayes. 1994. Amphibianand reptile species of special concern in California.Report to the Inland Fisheries Division, California

Department of Fish and Game, Rancho Cordova,CA.Johnson, T. B., and R. B. Spicer. 1985. Phrynosomamcallii (Hallowell 1852) Flat-tailed horned lizard.Contract Report No. 14-16-002-81-224 to U.S.Fish and Wildlife Service, Albuquerque, NM.Jones, T. R., B. D. Taubert, M. L. Zornes, and R. D.Babb. 2007. Kinosternon arizonense (Arizonamud turtle). USA: Arizona: Maricopa County.Herpetological Review 38(4): 479–480.Kralovec, M. L. 2006. Past, present, and futuremanagement of border related impacts on OrganPipe Cactus National Monument resources. InPrograms and Abstracts: Borders, Boundaries, andTime Scales, Sixth Conference on Research andResource Management in the SouthwesternDeserts, Tucson, AZ. 41pp.Krekorian, C. O. 1984. Life history of the desertiguana, Dipsosaurus dorsalis. Herpetologica40(4): 415–424.Kresan, P. 2007. A geologic tour of the dry bordersregion. In Dry borders, great natural reserves ofthe Sonoran Desert, edited by R. Felger and B.Broyles, pp. 31–45. University of Utah Press, SaltLake City, UT.Liner, E. A., and G. Casas-Andreu. 2008. Nombresestándar en Español en Inglés y nombrescientíficos de los anfibios y reptiles de México.Segundia Edición. Society for the Study ofAmphibians and Reptiles, Herpetological CircularNo. 38.Lowe, C. H., C. R. Schwalbe, and T. B. Johnson.1986. The venomous reptiles of Arizona. ArizonaGame and Fish Department, Phoenix, AZ. 115pp.Lower Colorado River Multi-Species ConservationProgram. 2004. Lower Colorado River Multi-Species Program, Volume II: Habitat ConservationPlan. Final. December 17. (J&S 00450.00). Sacra-mento, CA. Luckenbach, R. A., and R. B. Bury. 1983. Effectsof off-road vehicles on the biota of the AlgodonesDunes, Imperial County, California. Journal ofApplied Ecology 1983(20): 265–286.Luke Air Force Base. 2005. Biological assessmentfor the emergency fire suppression activities onthe Barry M. Goldwater Range East duringsummer 2005. Range Management Office, LukeAir Force Base, AZ. 14pp.Marshall, R. M., S. Anderson, M. Batcher, P. Comer,S. Cornelius, R. Cox, A. Gondor, D. Gori, J.Humke, R. Parades Aguilar, I. E. Parra, and S.Schwartz. 2000. An ecological analysis of conser-vation priorities in the Sonoran Desert ecoregion.The Nature Conservancy Arizona Chapter andInstituto del Medio Ambiente y el DesarrolloSustentable del Estado de Sonora. 51pp. + appen-dices.May, L. A. 1973. Resource reconnaissance of theGran Desierto region, northwestern Sonora,Mexico. Masters thesis, University of Arizona,Tucson, AZ. 173pp.

196 Rorabaugh

Mellink, E., and V. Ferreira-Bartrina. 2000. Onwildlife of wetlands of the Mexican portion of theRio Colorado delta. Bulletin of Southern Cali-fornia Academy of Science 99(3): 115–127.Milstead, B, and B. Barns. 2002. Life on the border:Monitoring the effects of border-crossing and lawenforcement on natural resources. In Meetingresource management information needs: Fourthconference on research and resource managementin the southwestern deserts, extended abstracts,edited by W. L. Halvorson and B. S. Gebow, pp.87–88. USGS Sonoran Desert Field Station,University of Arizona, Tucson, AZ.Mueller, G. A., and P. C. Marsh. 2002. Lost, a desertriver and its native fishes: A historical perspectiveof the lower Colorado River. Information Tech-nology Report, USGS/BRD/ITR-2002-0010.69pp.Nabhan, G. P., H. Govan, S. A. Eckert, and J. A.Seminoff. 1999. Sea turtle workshop for theindigenous Seri Tribe. Marine Turtle Newsletter86: 14.Nabhan, G. P. 2003. Singing the turtles to sea: TheComcáac (Seri) art and science of reptiles. Univer-sity of California Press, Berkeley, CA. 317pp.Nagy, K. A. 1973. Behavior, diet, and reproductionin the desert lizard, Sauromalus obesus. Copeia1973(1): 93–102.Narog, M. G., and R. C. Wilson. 2003. Delayedmortality: Saguaro cacti are still dying 10 yearsafter wildfire!, Extended Abstracts, Second Inter-national Wildland Fire Ecology and FireManagement Congress and Fifth Symposium OnFire And Forest Meteorology, American Meteor-ological Society, Boston, MA.National Oceanic and Atmospheric Administration.1989a. July 1989. Hourly Precipitation Data,Arizona 39(7): 1–11.National Oceanic and Atmospheric Administration.1989b. August 1989. Hourly Precipitation Data,Arizona 39(8).Ohmart, R. D., B. W. Anderson, and W. C. Hunter.1988. The ecology of the lower Colorado Riverfrom Davis Dam to the Mexico-United Statesinternational boundary: A community profile.Biological Report 85(7.19), U.S. Fish and WildlifeService, Research and Development, Washington,D.C. 296pp.Parker, W. S., and E. R. Pianka. 1975. Comparativeecology of populations of the lizard Uta stans-buriana. Copeia 1975: 615–632.Pianka, E. R. 1970. Comparative autecology of thelizard Cnemidophorus tigris in different parts ofits geographic range. Ecology 51: 703–720.Pianka, E. R. 1986. Ecology and natural history ofdesert lizards. Princeton University Press,Princeton, NJ. 208pp.Piest, L., and B. Knowles. 2002. An analysis of thestatus and land ownership of flat-tailed hornedlizard habitat in Arizona. Arizona Game and FishDepartment, Yuma, AZ.

Pitt, J. 2001. Can we restore the Colorado Riverdelta? Journal of Arid Environments 49: 211–220.Platz, J. E., R. W. Clarkson, J. C. Rorabaugh, andD. M. Hillis. 1990. Rana berlandieri: Recentlyintroduced populations in Arizona and south-eastern California. Copeia 1990(2): 324–333.Rodriguez, R. 2002. Evaluation of the status, distri-bution, and development of education/interpretation materials of the flat-tailed hornedlizard, Phrynosoma mcallii (Hallowell) in Mexico.Contract - 00PG340230. Centro Intercultural deEstudios de Desiertos y Océano, A.C. PuertoPeñasco, Sonora, Mexico. 34pp. + appendices.Rorabaugh, J. C. 1995. A superior accession ofwestern honey mesquite (Prosopis glandulosa var.torreyana) for riparian restoration projects. DesertPlants 11(4): 30–38.Rorabaugh, J. C. 2008. An introduction to theherpetofauna of mainland Sonora, México, withcomments on conservation and management.Journal of the Arizona-Nevada Academy ofScience 40(1): 20–65.Rorabaugh, J. C., M. Coffeen, and L. Piest. 2002a.Human disturbance in the flat-tailed horned lizardYuma Desert Management Area. U.S. Fish andWildlife Service, Phoenix, AZ. 14pp.Rorabaugh, J. C., C. L. Palermo, and S. C. Dunn.1987. Distribution and abundance of the flat-tailedhorned lizard (Phrynosoma mcallii) in Arizona.The Southwestern Naturalist 32(1): 103–109.Rorabaugh, J. C., M. J. Sredl, V. Miera, and C. A.Drost. 2002b. Continued invasion by an intro-duced frog (Rana berlandieri): SouthwesternArizona, southeastern California, and RioColorado, Mexico. The Southwestern Naturalist47(1): 12–20. Rorabaugh, J. C., S. Vissman, and B. L. Morrill.2000. A multi-agency conservation agreement forthe flat-tailed horned lizard, Phrynosoma mcallii,in the Sonoran Desert of southwestern Arizonaand southeastern California. In Creative coopera-tion in resource management: Third conferenceon research and management in the southwesterndeserts, extended abstracts, edited by W. L.Halvorson and B. S. Gebow, pp. 75–76. USGSSonoran Desert Field Station, University ofArizona, Tucson, AZ. Rosen, P. C. 2000. A monitoring study of vertebratecommunity ecology in the northern SonoranDesert, Arizona. Dissertation, University ofArizona, Tucson, AZ. 307pp.Rosen, P. C. 2007. Amphibians and reptiles in aridsouthwestern Arizona and northwestern Sonora.In Dry borders, great natural reserves of theSonoran Desert, edited by R. Felger and B.Broyles, pp. 310–337. University of Utah Press,Salt Lake City, UT.Rosen, P. C., and C. H. Lowe. 1994. Highwaymortality of snakes in the Sonoran Desert ofsouthern Arizona. Biological Conservation68(1994): 143–148.

Rorabaugh 197

198 Rorabaugh

Rosen, P. C., and C. H. Lowe. 1996. Ecology of theamphibians and reptiles at Organ Pipe CactusNational Monument, Arizona. Technical ReportNo. 53. Tucson, AZ: Cooperative Park StudiesUnit, National Biological Survey, University ofArizona. 136pp.Rosen, P. C., and C. R. Schwalbe. 1988. Status ofthe Mexican and narrow-headed garter snakes(Thamnophis eques megalops and Thamnophisrufipunctatus rufipunctatus) in Arizona. Report tothe Office of Endangered Species, U.S. Fish andWildlife Service, Albuquerque, NM. 50pp. +appendices.Rosen, P. C., and C. R. Schwalbe. 2002. Effects ofexotics on reptiles and amphibians. In Invasiveexotic species in the Sonoran region, edited by B.Tellman, pp. 220–240. University of Arizona Pressand the Arizona-Sonora Desert Museum, Tucson,AZ.Rosenberg, K. V., R. D. Ohmart, W. C. Hunter, andB. W. Anderson. 1991. Birds of the lowerColorado River Valley. University of ArizonaPress, Tucson, AZ. 416pp.Rosenzweig, M. L. 1968. Net primary productivityof terrestrial communities: Prediction from clima-tological data. American Naturalist 102: 683–718.Rutman, S., and L. Dickson. 2002. Management ofbuffelgrass on Organ Pipe Cactus National Monu-ment, Arizona. In Invasive exotic species in theSonoran region, edited by B. Tellman, pp. 311–318. University of Arizona Press and theArizona-Sonora Desert Museum, Tucson, AZ.424pp.Schmid, M. K., and G. F. Rogers. 1988. Trends infire occurrence in the Arizona upland subdivisionof Sonoran Desert, 1955–1983. The SouthwesternNaturalist 33: 437–444.Schmidt, B. 2005. Lack of water squeezes San LuisRio Colorado industry to a trickle. July 19, 2005Yuma Daily Sun.Schwalbe, C. R., T. C. Esque, P. J. Anning, and W.L. Halvorson. 2000. Exotic grasses, long-livedspecies, and managing desert landscapes: A casehistory at Saguaro National Park. In Creativecooperation in resource management: Thirdconference on research and management in theSouthwestern deserts, extended abstracts, editedby W. L. Halvorson and B.S. Gebow, p. 87. USGSSonoran Desert Field Station, University ofArizona, Tucson, AZ.Segee, B. P., and J. L. Neeley. 2006. On the line, theimpacts of immigration policy on wildlife andhabitat in the Arizona borderlands. Defenders ofWildlife, Washington, D.C. 40pp.SEMARNAT (Secretaría de Medio Ambiente yRecursos Naturales). 2008. Proyecto demodificación a la Norma Oficial Mexicana NOM-059-SEMARNAT-2001, Protección ambiental –Especies naturales de México de flora y faunasilvestres – categorías de riesgo y especificacionespara su inclusión o cambio – Lista de especies en

riesgo. Diaro Oficial, viernes 5 de diciembre de2008 (Segunda Sección).Seminoff, J. A., and W. J. Nichols. 2007. Sea turtlesof the Alto Golfo: A struggle for survival. In Dryborders, great natural reserves of the SonoranDesert, edited by R. Felger and B. Broyles, pp.505–518. University of Utah Press, Salt Lake City,UT.Smits, A. W. 1985. Behavioral and dietary responsesto aridity in the chuckwalla, Sauromalus hispidus.Journal of Herpetology 19(4): 441–449.Spillman, B. 2006. Poachers squeeze the fun out ofsnake collecting. USA Today, August 28, 2006.Stebbins, R. C. 1966. A field guide to westernreptiles and amphibians. Houghton MifflinCompany, NY. 279pp.Stebbins, R. C. 2003. A field guide to westernreptiles and amphibians. Houghton MifflinCompany, NY. 533pp.Sullivan, B. K., and M. A. Flowers. 1998. Largeiguanid lizards of urban mountain preserves innorthern Phoenix, Arizona. Herpetological NaturalHistory 6(1): 13–22.Sullivan, B. K., and K. O. Sullivan. 2008. Commonchuckwalla (Sauromalus ater) populations in thePhoenix metropolitan area: Stability in urbanpreserves. Herpetological Conservation andBiology 3(2): 149–154.Turner, D. S., C. R. Schwalbe, and P. L. Warren.1997. Reptile and plant inventory of the MohawkDunes, inventory and monitoring of unique sanddune ecosystems, Barry M. Goldwater Air ForceBase, Arizona. Legacy Resource ManagementProgram Project #95-1009. University of Arizona,Tucson, and The Nature Conservancy, ArizonaChapter. 46pp.Turner, F. B. 1977. The dynamics of populations ofsquamates, crocodilians and rhynocephalians. InBiology of the reptilia, volume 7, ecology andbehavior A, edited by C. Gans and D. W. Tinkle,pp. 157–261. Academic Press, London, England.Turner, F. B., P. A. Medica, K. W. Bridges, and R.I. Jennrich. 1982. A population model of the lizardUta stansburiana in southern Nevada. EcologicalMonographs 52(3): 243–259.Turner, F. B., P. A. Medica, J. R. Lannon, Jr., and G.A. Hoddenbach. 1969. A demographic analysis offenced populations of the whiptail lizard, Cnemi-dophorus tigris, in southern Nevada. TheSouthwestern Naturalist 14: 189–202.Turner, R. M., and D. E. Brown. 1982. Sonorandesertscrub. In Biotic communities of the Amer-ican Southwest—United States and Mexico,edited by D. E. Brown. Desert Plants 4(1-4): 181–222.U.S. Census Bureau. U.S. Census 2000.http://www.census.gov/main/www/cen2000.htmlU.S. Departments of the Navy, Air Force, and Inte-rior. 2006. Final environmental impact statementfor the Barry M. Goldwater Range IntegratedNatural Resources Management Plan, Vol. 1.

Rorabaugh 199

Marine Corps Air Station-Yuma and Luke AirForce Base, AZ. 616pp.Van Denburgh, J. 1922. The reptiles of westernNorth America. Occasional Papers, CaliforniaAcademy of Sciences 10: 1–1028.Vitt, L. J., and R. D. Ohmart. 1978. Herpetofaunaof the lower Colorado River: Davis Dam to theMexican border. Western Foundation of VertebrateZoology 2(2): 35–72.Weiss, J. L., and J. T. Overpeck. 2005. Is theSonoran Desert losing its cool? Global ChangeBiology 11: 2065–2077.Wilson, M. F., L. Leigh, and R. S. Felger. 2002.Invasive exotic plants in the Sonoran Desert. InInvasive exotic species in the Sonoran region,edited by B. Tellman, pp. 81–90. University of

Arizona Press and the Arizona-Sonora DesertMuseum, Tucson, AZ. 424pp.Woods, L. J., J. MacAller, J. L. Smeltzer, R. A.Murray, P. Rosen, C. Schwalbe, T. Hare, andcontributors. 2004. Habitat management guide-lines for amphibians and reptiles of the aridSouthwest (draft). Partners in Amphibian andReptile Conservation, Technical PublicationHMG-4, Tucson, AZ. 100pp.Young, K. V., and A. T. Young. 2000. Final report:Scientific study of the flat-tailed horned lizard,Phrynosoma mcallii. Marine Corps Air Station,Yuma, Arizona. U.S. Dept. of Navy ContractsN68711-95-LT-C0032, N68711-95-LT-C0035.72pp.

AMPHIBIANSSalamandersAmbystoma mavortium, TigerSalamander, Ajolote, SalamandraTigre

Frogs and ToadsAnaxyrus cognatus, Great PlainsToad, Sapo de las Grandes Planicias,SapoAnaxyrus punctatus, Red-spottedToad, Sapo Manchas Rojas, SapoPintoAnaxyrus retiformis, Sonoran GreenToad, Sapo Sonorense, Sapo

Anaxyrus woodhousii, Woodhouse’sToad, Sapo de Woodhouse, SapoGastrophryne olivacea, WesternNarrow-mouthed Toad, RanitaOlivo, SapitoOllotis alvaria, Sonoran DesertToad, Sapo Verde, Sapo Grande

Collected at the Pozo Municipal, Puerto Peñasco, occasionallyfound in Colorado River region, probably due to escaped/re-leased fish bait. Introduced. Pr

Gila River and associated croplands, Río Sonoyta and croplands,and occasional at impoundments in the eastern portion of studyarea. Rare on Colorado River.Found widely at tanks, tinajas, and ponds particularly in moun-tains and in the eastern portions of the study area. Occasional inagriculture.Breeds in ephemeral rain pools and cattle tanks in summer. Val-ley bottoms on eastern edge of study area from Lukeville toWhy. PrColorado and Gila River valleys. Common in agriculture andalong rivers.Recorded near Lukeville, may occur elsewhere in eastern edgeof study area. Breeds in ephemeral pools in summer. Pr

Eastern portions of study area in Arizona and Sonora, as well asalong the Gila River. May still occur at the Foothills and FortunaWash east of Yuma. Extirpated from the Colorado River, Ari-zona, but may still persist in the Río Colorado valley, Sonora,and Baja California.

Appendix. Checklist of the amphibians and reptiles of northwestern Sonora and southwestern Arizona.

Scientific and Common Names1 Distribution, Status, and Notes2

continued

200 Rorabaugh

Rana berlandieri, Río GrandeLeopard Frog, Rana Leopardo

Rana catesbeianus, AmericanBullfrog, Rana de Toro, RanaMugidoraRana yavapaiensis, LowlandLeopard Frog, Rana de Yavapai,Rana LeopardoScaphiopus couchii, Couch’sSpadefoot, Sapo de Espuela, Sapo

Smilisca fodiens, Lowland Burrow-ing Treefrog, Ranita Minera, Rana

REPTILESTurtlesApalone spinifera, Spiny Softshell,Tortuga VerdeCaretta caretta, Loggerhead SeaTurtle, Tortuga Caguama, TortugaPerica Chelonia mydas, Green Sea Turtle,Parlama, Tortuga Negra,

Dermochelys coriacea, LeatherbackSea Turtle, LaúdEretmochelys imbricate, HawksbillSea Turtle, CareyGopherus agassizii, Desert Tortoise,Tortuga del Monte, Galápago deDesierto Kinosternon sonoriense, SonoranMud Turtle, Tortuga de Agua,Casquito de SonoraLepidochelys olivacea, Olive RidleySea Turtle, Tortuga Golfina

Trachemys scripta, Pond Slider,Tortuga Pinta

Colorado River valley, Sonora and Arizona, and Gila River val-ley – introduced from New Mexico or Texas in the 1960s or1970s. Found in riverine (including Cienega de Santa Clara) andagricultural habitats. PrColorado and Gila River valleys, in riverine and agriculturalhabitats. Introduced.

Historically found along Colorado and presumably Gila Rivers.Extirpated. Pr

Breeds in ephemeral rain pools and cattle tanks in summer.Widespread, but absent from valleys of the Gran Desierto andthe Yuma Desert.Breeds in ephemeral rain pools and cattle tanks during summer.Recorded near Why, may occur elsewhere in eastern edge ofstudy area.

An introduced aquatic turtle of the Río Colorado and Gila River,as well as adjacent agricultural ditches and canals. PrLarge (to 1.1 m) and rare turtle of the Gulf of California. Notknown to nest in Sonora. P, EN, T

Large (to 1.1 m) marine turtle of the Gulf of California. Mostcommonly encountered marine turtle on the coast of Sonora. Notknown to nest in study area. P, EN, ELarge (to 2.4 m, but typically < 1.0 m in the Gulf of California)and rare marine turtle. Not known to nest in study area. P, CR, ELarge (to 0.9 m, but most are 0.3-0.5 in the Gulf) marine turtle,which is more common in the southern portion of the Gulf. Notknown to nest in Sonora. P, CR, EA large (to 0.4 m) terrestrial tortoise found in mountains and ba-jadas. Apparently absent from some ranges in western studyarea. A, VUCurrently an aquatic stream and pond turtle of the Río Sonoytaand adjacent wetlands, including Quitobaquito. Likely extirpatedfrom Colorado and Gila Rivers. VU, CLarge (to 0.7 m) marine turtle of the Gulf of California. Nestsrarely on beaches as far north as Puerto Peñasco and El Golfo deSanta Clara. P, EN, TOccasional in aquatic habitats near Yuma. Introduced. No evi-dence of breeding. Pr

Appendix. Checklist of the amphibians and reptiles of northwestern Sonora and southwestern Arizona.continued


continued



Rorabaugh 201

LizardsAspidoscelis tigris, Tiger Whiptail,Huico OccidentalAspidoscelis xanthonota, Red-backed Whiptail, Huico, Huico deDorso RojoCallisaurus draconoides, Zebra-tailed Lizard, CachoraColeonyx variegatus, WesternBanded Gecko, Cuija Manchada Oc-cidentalCrotaphytus nebrius, SonoranCollared Lizard, Lagartija de Collar,CachorónDipsosaurus dorsalis, DesertIguana, Iguana, Iguana del DesiertoGambelia wislizenii, Long-nosedLeopard Lizard, Cachorón, Ca-chorón Mata CaballoHeloderma suspectum, GilaMonster, Escorpíon, EscorpíonPintadoHemidactylus turcicus,Mediterranean House Gecko, GecoPinto

Phrynosoma goodei, Goode’sHorned Lizard, Camaleón de SonoraPhrynosoma mcallii, Flat-tailedHorned Lizard, Camaleón de ColaPlana, Camaleón del Gran DesiertoPhrynosoma solare, Regal HornedLizard, Camaleón RealSauromalus ater, CommonChuckwalla, Iguana, Sceloporus clarkii, Clark’s SpinyLizard, Lagartija Espinosa de Clark,CachoraSceloporus magister, Desert SpinyLizard, Lagartija Espinosa delDesierto, Cachorón

Throughout study area in valleys, mountains, and riparianwoodlands. Known from mountains in northeastern portion of study area.Not documented in Sonora, but likely occurs in one or more ofthe Sierras Cubabi, San Francisco, and Pinacate. Flats and valleys. Particularly common in sandy or gravelly flatsand washes. AThroughout valleys and bajadas in the study area, occasional inlower montane canyons. Nocturnal. Pr

Rocky mountains and bajadas, throughout the study area.

Throughout flats and bajadas. Particularly abundant in the val-leys of the arid Gran Desierto and Yuma Desert.Flats and valleys throughout the study area. Pr

Primarily a montane and bajada species, most common in theeastern half of the study area. Venomous. A, VU

This introduced nocturnal gecko is found in cities and towns,often in and around buildings at night. Occurs at Yuma, GilaBend, and El Golfo de Santa Clara. Reported from San Luis RíoColorado. Likely occurs locally at Puerto Peñasco. Widespread, but within the range of P. mcallii, typically found incoarser sands closer to mountains.Arid, sandy flats and low dunes of the Gran Desierto and Yumadesert from Yuma southeast to Bahia San Jorge. PT, A

Flats in eastern portion of study area. Absent from west of CraterMacDougal and Agua Dulce Mountains.Large-bodied lizard of rocky slopes and outcrops throughout thestudy area. ARelatively mesic canyons in the Ajo and other mountains in theeastern edge of the study area. Often found in trees, sometimeson the ground or in rocks.Throughout the study area, except for treeless valleys in the GranDesierto and Yuma Desert. Found on trees or rocks.

continued

Uma rufopunctata, Yuman Fringe-toed Lizard, Cachora, Lagartija deManchas LateralesUrosaurus graciosus, Long-tailedBrush Lizard, Cachorrita, Lagartijade Matorral.Urosaurus ornatus, Ornate TreeLizard, Lagartija de Árbol,CachoritaUta stansburiana, Common Side-blotched Lizard, Cachora GrisXantusia vigilis, Desert NightLizard, Cuija, SalamanquesaSnakesArizona elegans, Glossy Snake,Culebra BrillanteChilomeniscus stramineus, VariableSandsnake, Culebra de los Médanos,CoralilloChionactis occipitalis, WesternShovel-nosed Snake, Culebra Pala-naria Occidental Chionactis palarostris, SonoranShovel-nosed Snake, Coralillo Falso

Coluber bilineatus, Sonoran Whip-snake, Chirrionera, AlicanteColuber flagellum, Coachwhip,Chirrionera, AlicanteCrotalus atrox, Western Diamond-backed Rattlesnake, Víbora Serrana,CascabelCrotalus cerastes, Sidewinder,CuernitosCrotalus mitchelli, Speckled Rat-tlesnake, Víbora Blanca

Crotalus molossus, Black-tailed Rat-tlesnake, Cola Prieta, Cascabel

Patchily distributed in dunes and flats with fine, windblown sandfrom near Dateland to the Yuma Desert and south to the Gulf ofCalifornia. ATypically found on the branches of creosote, mesquite and othershrubs or trees in the valleys in the study area. Rare in easternstudy area and distribution poorly known in Sonora.Typically found on trees or rocks. Throughout study area, butpatchy distribution in the western deserts. Common in riparianwoodlands and urban landscaping. Absent from treeless valleys.Widespread throughout the study area in flats and mountains.

Isolated montane localities. Typically found under dead agaves,Yuccas, and Nolinas. Primarily nocturnal.

Primarily a species of flats and valleys. Throughout the studyarea.Primarily a burrowing species of sandy or gravelly arroyos withleaf litter. Probably absent from western valleys. PrFlats and valleys throughout. Abundant in sandy flats and lowdunes.

Eastern edge of study area from near Why south into Sonora ingravelly bajadas and sandy to rocky flats with relatively openvegetation. Mountains and bajadas in northeastern portion of study area; alsoin Sierra Pinacate. Often arboreal.Throughout, primarily in valleys and bajadas. Found primarilyon the ground, but occasionally in trees and shrubs. ASpecies of bajadas, washes, and riparian areas. Common in thelatter habitat. Absent from valleys of the western Gran Desiertoand Yuma Desert. Highly venomous. PrSmall (< 0.8 m) horned rattlesnake of sandy flats and bajadasthroughout. Abundant in sandy western valleys. Highly ven-omous. PrRocky mountains and bajadas in the Pinacate region of Sonoraand throughout Arizona portion, except for most ranges at OrganPipe Cactus NM. Highly venomous. PrRocky slopes and canyons likely throughout the study area.Highly venomous. Rarely encountered in western mountainranges. Pr

continued



202 Rorabaugh

continued

Crotalus scutulatus, Mojave Rat-tlesnake, CascabelCrotalus tigris, Tiger Rattlesnake,Cascabel, Cascabel del TigreHypsiglena chlorophaea,Nightsnake, Culebra NocturnaLampropeltis getula, CommonKingsnake, Culebra Real ComúnLeptotyphlops humilis, WesternThreadsnake, Culebrilla Ciega deOccidenteLichanura trivirgata, Rosy Boa, BoaRosadaMicruroides euryxanthus, SonoranCoralsnake, Coralillo, Coralillo Oc-cidentalPelamis platurus, Yellow-belliedSea Snake, Alicante del MarPhyllorhynchus browni, SaddledLeaf-nosed Snake, CulebritaPhyllorhynchus decurtatus, SpottedLeaf-nosed Snake, CulebritaPituophis catenifer, Gophersnake,Cincuate, Víbora Sorda

Rhinocheilus lecontei, Long-nosedSnake; Coralillo FalsoSalvadora hexalepis, Western Patch-nosed Snake, Culebra ChataSonora semiannulata, WesternGroundsnake, Culebra de ArenaTantilla hobartsmithi, Smith’sBlack-headed Snake, CulebraCabeza Negra del SuroesteThamnophis cyrtopsis, Black-neckedGartersnake, Culebra de Agua, Cule-bra Lineada de Bosque

Valleys and bajadas, particularly in the east. Absent from westernGran Desierto (west of Sierra del Rosario) and Yuma Desert.Highly venomous. PrSmall (< 0.9 m) rattlesnake of rocky slopes and canyons in east-ern portions of study, especially at Organ Pipe Cactus NM.Highly venomous. PrMountains and bajadas, absent from valleys of the Gran Desiertoand the Yuma Desert. PrMost common in riparian corridors and agriculture, likely absentfrom drier western ranges and valleys. AProbably more common in eastern portion of study area. Likelyabsent from arid valleys of the Gran Desierto and the YumaDesert. Small, secretive, burrowing species.Montane species of the northeastern study area, but also the Gilaand Tinajas Altas mountains and Sierra Pinacate. Likely occurselsewhere. AEastern edge of study area. Organ Pipe Cactus NM, Agua DulceMountains, presumably mountains and bajadas in eastern studyarea in Sonora. Highly venomous. ARanges throughout the Gulf of California, but does not breed inour area. Highly venomous, although it rarely bites. Northeastern portion of study area in foothills and bajadas. Oftenin gravelly or sandy soils. PrTypically found in sandy or gravelly soils in valleys and bajadas.Absent from western valleys. Widespread, but likely absent from arid valleys in the westernGran Desierto and Yuma Desert. Common in riparian and agri-culture.Likely absent from arid valleys in the Gran Desierto and YumaDesert. Elsewhere fairly common. Occasional in agriculture. Throughout, but most common in valleys. Diurnal.In urban, agricultural, and riparian habitats along the Río Col-orado and presumably Gila River. Distribution elsewhere poorlyunderstood. Mesic canyons in the Ajo Mountains, but also near Sonoyta. Reported from the Gila River at Wellton.

Mesic canyons in the Ajo Mountains. Diurnal. A



Rorabaugh 203

204 Rorabaugh

Thamnophis eques, MexicanGartersnake, Culebra de Agua,Culebra de Agua Nómado MexicanoThamnophis marcianus, CheckeredGartersnake, Culebra de Agua,SochuateTrimorphodon biscutatus, WesternLyresnake, Víbora Sorda

A highly aquatic species, formerly occurred near Yuma and pre-sumably elsewhere on the Colorado and Gila Rivers. Extirpated.A, CMarshy wetlands and agriculture along or near the Colorado andGila Rivers near Yuma. Ciénega de Santa Clara and likely else-where in the Río Colorado Valley of Sonora.ARocky hillsides and bajadas throughout the study area.



1 Nomenclature based on Crother (2008), except that Spanish common names are taken from Liner and Casas-Andreu(2008) or are names used locally. 2 Status corresponds to Mexico’s ‘Lista de Especies en Riesgo” (SEMARNAT 2008),including P = in danger of extinction, A = threatened, and Pr = species of special protection; IUCN’s Redlist, includ-ing CR = critically endangered, EN = endangered, VU = vulnerable, and NT = near threatened; and U.S. EndangeredSpecies Act, including C = candidate for listing and PT = proposed threatened, T = threatened, and E = endangered.

OBSERVATIONS ON THE STATUS OF AQUATIC TURTLES AND THEOCCURRENCE OF RANID FROGS AND OTHER AQUATIC VERTEBRATESIN NORTHWESTERN MEXICOPhilip C. Rosen and Cristina Melendez

The aquatic vertebrate fauna in the arid south-western United States is highly imperiled bythe combined action of habitat destruction,habitat modification, introduction of exoticspecies, and introduced diseases and parasites(Minckley and Deacon 1991; Rinne andMinckley 1991; Rosen and Schwalbe 2002a).Although much less completely documented,it appears that similar processes are morerecently (Unmack and Fagan 2004) becomingmanifest in arid northwestern Mexico(Hendrickson et al. 1980; Hendrickson 1983;Contreras-Balderas and Escalante Cavasos1984; Miller et al. 2005). Although severe inMexico, the problem appears to be developinglater than in the U.S., with similar ecologicalsymptoms apparently time-lagged by about 4decades. (Unmack and Fagan 2004). Armedwith this knowledge, perhaps Mexico mightabate and mitigate the catastrophic ecologicalsituation that is occurring in the U.S.The status of aquatic, perennial-water

herpetofauna in Mexico is poorly known anddocumented, and no published details areavailable in arid northwestern Mexico for thisassemblage, which shares many species,habitat characteristics, and conservation prob-lems with the aquatic herpetofauna of the U.S.Southwest. In connection with a survey of mudturtles (genus Kinosternon) to obtain additionalmaterial for an ongoing genetic study at local-ities in Sonora and northwestern Chihuahua,we had the opportunity to conduct a rapidassessment of occurrence and status of aquatic

herpetofauna. We focused on the turtles and,especially, ranid frogs (family Ranidae), whichare impacted by exotic species and diseasethroughout the American West (Rosen 2008;Lannoo 2005). Based on long-term familiaritywith this ecological system and its componentsin Arizona and New Mexico, our surveysallowed us to assess habitat conditions andintroduced species such as bullfrogs, crayfish,and non-native fishes known to be affectingnative frogs in adjoining parts of the U.S.(Hayes and Jennings 1986; Rosen et al. 1995;Rosen and Schwalbe 2002b). Here we presentthe results of this autumn 2005 survey, andcompare them to species status and habitatconditions in the Sonoran Desert and Madreanregions of the United States.

METHODSSampling and Identification

During the period 22 September through 11November 2005 we sampled 39 localities inSonora and northwestern Chihuahua (Figure 1,Table 1). The targeted localities were chosenbased on museum records for mud turtles(genus Kinosternon), with additional sitessampled whenever time and access permittedalong the extensive survey routes. Sites weresampled using a combination of baited hooptraps, dipnetting, seining, and visual encountersurvey (Crump and Scott 1994) for periods of0.5-13.25 hours. Although the objective was toobtain blood samples from turtles for a contin-uing phylogeographic study (Rosen 2003), and

206 Rosen and Melendez

much time was spent traveling between thewidely scattered localities, sites reported herewere sampled as thoroughly as possible forfishes, amphibians, reptiles, and crayfish.However, since only single site visits weremade, our data do not represent an exhaustivesampling of the aquatic fauna.We made efforts to capture, identify, and

obtain photographic vouchers for all aquaticspecies observed. This was not alwayspossible, and some taxa could not be reliably

identified from our photographs. For leopardfrogs, we assigned species based ongeographic range indicated by Frost andBagnara (1976), which largely correspondedto our estimates based on our gestalt (field-based) identifications. Reptiles that were notcaptured but were tentatively identified withbinoculars are indicated with a “?” in thetables. Although specimens could not becollected, all crayfish we observed appeared tobe the exotic northern crayfish (Orconectes

Figure 1. Localities sampled for turtles and other aquatic vertebrates during fall 2005.

Rosen and Melendez 207

virilis), as in most of Arizona and previouslyreported for Chihuahua (Hobbs 1989). Fishes were identified in hand and from

voucher photographs, and confirmed withreference to Miller et al. (2005). Although wesuspect that most of the cichlid fishes weobserved were redbelly tilapia (Tilapia zillii),except at Alamos and in Ojo de Agua on RíoMátape, and are reasonably certain all werenon-native, we could not confirm cichlid iden-tifications and therefore report all cichlids as“Tilapia sp.” Similarly, topminnows wereidentifiable as Poeciliopsis occidentalis (or P.sonoriensis) based on the presence of darkmales; we made no attempt to treat theproblem of accompanying all-female species,and list all our records under the genus name.Names of animals discussed in text and

tables are collated in Table 2. We have utilizedcommon names as in Crother (2001), Liner(1994), Schwalbe and Lowe (2000), and Milleret al. (2005) with slight modifications forclarity and conciseness, as follows. We retainthe generic names Bufo and Rana, followingPauly et al. (2009). For Kinosternonsonoriense, we use the name, "Sonoran MudTurtle", rather than "Sonora Mud Turtle"because the latter is a recent misnomer thatmisleads regarding the distribution of thespecies, which is clearly associated with theSonoran biogeographic region but notrestricted to or centered on the state of Sonora.

Relative Abundance and RankingFor numerical analysis of abundance patterns,we converted categorical abundance values toordinal ranks (Tables 3 and 4). The method-ological design and ordinated categories usedin this rapid assessment follow from experi-ence surveying for and categorizing abundanceof aquatic herpetofaunal taxa in Arizona(Rosen and Schwalbe 1988, 2002b; Rosen1987). This ensures that within-study compar-isons among sites are un-biased and reflectsimilar observations reported for Arizona bythe senior author. Here, details are provided to make clear the

observational bases underlying the ranked valuesfor abundance to facilitate comparison to other

and future studies. Survey times (Table 1) werematched, as nearly as possible, to the needspresented by habitat complexity and extent. Inall cases, environments were searched for visible(resting, sun-basking, under-cover hiding, andsurfacing) animals; clear water was searchedvisually for visible fish and tadpoles. If appro-priate, each site was sampled using along-handled dipnet (2 m handle, 4 mm mesh)including random sweeps and directed captureto identify animals. Trapping (Table 1) wascarried out for turtles and large fishes for the timeperiods indicated in the table. Deeper waterswere seined (25 ft x 6 ft deep, 4 mm mesh bagseine) when feasible to capture large fish foridentification and when low water clarityprevented observations of fish relative abun-dance. However, most species identifications oflarger fish taxa were confirmed based oncaptured small conspecifics. For springs and small ponds, the entire

perimeter was walked and searched, anddipnetting was carried out systematically insuitable habitat cover. Similar methods wereused in larger systems, except that (a) loticsystems or large reservoirs were not sampledin their entirety, but only within 1-3 km of thecoordinates given in Table 1, (b) searches weremade widely to encompass maximal habitatdiversity within available time, (c) auspicioushabitat, based on prior experience, wasaccorded a majority of search time, while (d)apparently suboptimal habitat was spot-sampled to avoid allowing pre-conceived ideasabout habitat to go unchecked under the poten-tially different circumstances of Sonoracompared to Arizona.In a rapid assessment survey, direct, numer-

ical data on abundance and population densitycannot be collected. Experience has shown thatthese assessments, with ordinal ranking of abun-dance, have great value for resource managersand subsequent surveyors (e.g., Holycross et al.2006 use of Rosen and Schwalbe 1988). Theranked categories are defined as follows: 0(absent) = none seen; 1 (rare) = one or a veryfew (generally less than four) specimens seencompared to taxon-specific expectation (seebelow); 2 (uncommon) = few seen (e.g.,


generally 8 or less), and generally less thanexpected for abundant populations observed inArizona; 3 (common) = individuals seen inmoderate to low numbers over extensiveportions of the sample area, and total numbersgenerally 9 or greater for abundant species likefrogs or involving schools of fish; 4 (abundant)= consistently observed individuals throughoutextensive portions of sample area, and gener-

ally always present in suitable habitat; 5 (super-abundant) = species visible in high numbersthroughout sampling area, at densitiesexceeding those seen under usual circumstancesfor the taxon, and in numbers totaling in theteens (turtles), hundreds (ranid frogs or largerfish species), or thousands in extensive schools(tadpoles, smaller fish species, and juvenilefish). Naturally uncommon or rarely observed

Table 1. Localities surveyed, survey time, and sampling methods (V = visual encounter survey; T = baitedhoop net; D = dipnetting) for turtles and other aquatic animals during Fall 2005 in Sonora and Chihuahua..

Location UTM (NAD-27) Elev. Sampling No. Locality Description Zone Easting Northing (ft.) Date Time Method

1 Son., Río Magdalena - Teranate 12R 507832 3398832 NA 09/22/05 2:00 V, D2 Son., Río Magdalena - bridge to

Teranate 12R 513072 3404692 NA 09/22/05 0:45 V, D3 Son., Rió San Miguel de Horcasitas

near Cucurpe 12R 528002 3354356 NA 09/22/05 2:00 V, D4 Son., Rancho Agua Fria wash,

Saracachi Ciénega 12R 541549 3357976 NA 09/23/05 1:00 V5 Son., Saracachi Ciénega 12R 539026 3358469 NA 09/23/05 8:00 V, T, D6 Son., Río Sonora just N of Arizpe 12R 583972 3358173 NA 09/24/05 1:00 V, D7 Son., Sierra Aconchi at Agua Caliente

Spring Canyon 12R 569694 3301863 NA 09/24/05 2:45 V, D8 Son., Presa Molinito, Rió Sonora E

of Hermosillo 12R 526392 3231339 NA 09/24/05 1:15 V, D9 Chih., Río Casas Grandes Hwy bridge

N Casas Grandes Viejo 13R 217822 3364072 4813 10/11/05 2:25 V, T, D10 Chih., Río Casas Grandes, tajo W

Nuevo Casas Grandes 13R 217321 3368690 NA 10/12/05 2:40 V, D11 Chih., Casas Grandes, Rancho La

Princesa house pond 13R 217532 3368366 NA 10/12/05 1:45 V, D12 Chih., Río Piedras Verdes ca. 1.7 mi

SE of Colonia Juarez 12R 783651 3354182 NA 10/12/05 3:49 V, D13 Chih., Galeana, Angostura pool/spg

near Ojo de Arreys 13R 250247 3327924 4900 10/13/05 6:00 V14 Chih., Galeana, Angostura town,

bulldozed pond & vicinity 13R 249745 3326672 4900 10/13/05 0:30 V, D15 Chih., Río Santa Maria, Buenaventura,

S of Chih Hwy 28 13R 261048 3303204 5131 10/14/05 2:00 V, T, D16 Chih., Río Santa Maria, Buenaventura, rastro N Chih Hwy 28

13R 260339 3304324 5110 10/14/05 1:15 V, T, D17 Chih., floodplain acequia pond ca.

4 mi S Buenaventura 13R 262484 3300382 NA 10/14/05 0:30 V, D18 Chih., Río Santa Maria irr. headgate,

ca. 5 mi S Buenaventura 13R 263772 3296396 NA 10/14/05 0:30 V, Dcontinued


species, or those for which our samplingmethods or seasonality were insufficient topermit estimates of relative abundance, werecategorized as absent (0) or present (+); theseobservations are included for future referenceand were not used in the analyses presented inthis paper. Although it is admittedly impossibleto precisely specify the way ordinal ranks wereassigned, we have little doubt that experiencedobservers, accustomed to assigning such values

in the field, would produce roughly similar andconsistent relative abundance values.In order to summarize the ordinal abun-

dance data into broader measures reflective ofoverall fish abundance patterns (native fishabundance, exotic fish abundance, exoticpredatory fish abundance) we created indicesby summing the ordinal ranks for these cate-gories. From these, standardized indices of therelative dominance of native versus exotic fish

Location UTM (NAD-27) Elev. Sampling No. Locality Description Zone Easting Northing (ft.) Date Time Method19 Chih., Río Papagochic at Yepomera,

at Chih Hwy 16 13R 221193 3218796 6349 10/14/05 2:30 V, T, D20 Son., Yecora, stream in meadow

0.4 mi S of Mex Hwy 16 12R 703736 3138788 5075 10/15/05 4:45 V, T, D21 Son., Río Altar, Presa Cuatemoc 12R 450785 3416147 NA 10/22/05 0:30 V22 Son., Río Altar at Tubutama bridge 12R 455537 3417098 2050 10/22/05 1:49 V, D23 Son., tank along Mex Hwy 14 W of

Moctezuma 12R 617942 3293168 3545 10/27/05 1:00 V, D24 Son., Río Moctezuma S of Jecori 12R 621269 3310352 NA 10/27/05 3:30 V, D25 Son., Río Moctezuma at presa near

Mex Hwy 14 12R 625987 3296664 2048 10/28/05 0:30 V, D26 Son., Río Moctezuma N of Jecori

(Jamaica) 12R 619951 3315235 NA 10/28/05 4:45 V, T, D27 Son., Río Bavispe, N of Huasabas 12R 665423 3314947 1809 10/28/05 12:15 V, T, D28 Son., Río Bavispe, S of Bavispe 12R 699744 3367785 3284 10/30/05 13:15 V, T, D29 Son., La Colorada (presa just N of

town) 12R 540865 3186789 1267 11/05/05 1:08 V, D30 Son., Río Mátape at Hwy 16 near

San Jose de Pima 12R 563683 3176913 1180 11/05/05 2:14 V, D31 Son., Río Mátape ca. 4 mi S of

San Jose de Pima 12R 562640 3173996 NA 11/05/05 0:45 V, D32 Son., Rancho Ojo de Agua, NE of

La Misa 12R 567439 3149138 826 11/05/05 12:00 V, T, D33 Son., isolated spring just NW Punta

de Agua, vic. La Misa 12R 558735 3145084 773 11/06/05 0:50 V, D34 Son., laguna below dam at Punta de

Agua (nr La Misa) 12R 558739 3144330 NA 11/06/05 0:30 V, D35 Son., Sierra de Alamos, aguaje SE of

Rancho La Sierrita 12R 704379 2985841 1520 11/09/05 0:45 V, D36 Son., Alamos, Rancho Las Cabras 12R 708143 2987805 NA 11/10/05 0:45 V, D37 Son., Alamos, tank nr Río Cuchujaqui,

SSE Las Uvalamas 12R 707428 2977582 NA 11/11/05 1:25 V, D38 Son., Alamos, pool along El Chinal

Rd 4.2 mi S Las Uvalamas 12R 706211 2977489 NA 11/11/05 0:30 V, D39 Son., Alamos, concrete well tank at

(S of) Rancho Las Sierrita 12R 703997 2984964 1770 11/11/05 0:30 V

Table 1. continued

at individual sites were constructed by dividingtheir abundance indices by total abundanceindex for all fish.

Statistical Tests and AssumptionsTwo-sample tests and non-parametric analysisof covariance (ANCOVA on the ordinal rank

values; Quade 1967, Shirley 1981, Marascuiloand McSweeney 1977) were performed usingSAS-JMP version 5.1 on a personal computer.Sample sizes were not large, and we thereforeperformed visual checks of graphical materialfor approximate normality and homogeneityof slopes for parametric two-sample and


Table 2. Scientific and common names of animals referred to in text and tables. Asterisk indicates thecommon name given differs from cited sources.

Scientific NameEnglish Name Spanish Name Family

Bufo alvarius Sonoran Desert Toad sapo grande sonoriense * BufonidaeHyla arenicolor Canyon Treefrog ranita de canon HylidaePachymedusa dacnicolor Mexican Leaf-frog rana verduzca HylidaeSmilisca fodiens Lowland Burrowing Treefrog ranita excavador * HylidaeSmilisca baudini Mexican Treefrog rana trepadora HylidaeLeptodactylus melanonotus Sabinal Frog ranita sabinal LeptodactylidaeGastrophryne olivacea Great Plains Narrow-mouthed

Toad ranita boca cónica * MicrohylidaeRana catesbeiana American Bullfrog rana de toro * RanidaeRana magnaocularis Big-eyed Leopard Frog rana leopardo ojos grandes * RanidaeRana tarahumarae Tarahumara Frog rana de Tarahumara RanidaeRana yavapaiensis Lowland Leopard Frog rana leopardo de Yavapai * RanidaeOrconectes virilis Northern Crayfish cangrejo norteño CambaridaeCatostomus bernardini Yaqui Sucker matalote Yaqui CatastomidaeCatostomus wiggensi Opata Sucker matalote opata CatastomidaeLepomis cyanellus Green Sunfish pez sol CentrarchidaeMicropterus salmoides Largemouth Bass lobina negra Centrarchidaecichlid sp. unknown cichlid mojarra no conocido CichlidaeTilapia zillii Redbelly Tilapia tilapia CichlidaeAgosia chrysogaster Longfin Dace pupo panzaverde CyprindaeAgosia sp. Mexican Longfin Dace pupo mexicano CyprindaeCampostoma ornatus Mexican Stoneroller rodapiedras mexicano CyprindaeCyprinella formosa Beautiful Shiner carpita Yaqui CyprindaeCyprinus carpio Common Carp carpa comun CyprindaeGila ditaenia Sonora Chub carpa sonoriense CyprindaeGila eremica Desert Chub carpa del desierto CyprindaePimephales promelas Fathead Minnow carpita cabezona CyprindaeRhinichthys osculus Speckled Dace carpita pinta CyprindaeCyprinodon eremus Quitobaquito Pupfish cachorrito del Sonoyta CyprindontidaeAmeiurus melas Black Bullhead bagre torito negro IctaluridaeGambusia affinis Western Mosquitofish guayacon mosquito PoeciliidaePoeciliopsis occidentalis Gila Topminnow guatapote de Sonora PoeciliidaeTrachemys yaquia Yaqui Slider jicotea Yaqui EmydidaeKinosternon hirtipes Rough-footed Mud Turtle casquito de pata rugosa KinosternidaeKinosternon integrum Mexican Mud Turtle casquito de burro KinosternidaeKinosternon sonoriense Sonoran Mud Turtle casquito de Sonora KinosternidaeThamnophis cyrtopsis Black-necked Gartersnake culebra lineada de bosque NatricidaeThamnophis eques Mexican Gartersnake culebra de agua Mexicana * Natricidae


ANCOVA tests, respectively. The ANCOVA ispresented as a means to summarize the dataset:this study was conducted under uncontrolled,natural conditions with many operative factorsand possible, unknown covariates. TheANCOVA has heuristic value for under-standing the dataset numerically, but its resultis interpreted as a preliminary evaluation andhypothesis regarding the relative impact ofexotic fishes versus bullfrogs on native ranidfrog abundance.In this analysis we have assumed that native

leopard frogs originally occurred in all regionsand major waters sampled, an assumptionjustified by examination of museum records inthe University of Arizona herpetology collec-tion, as well as in the literature (Frost andBagnara 1976; Platz and Frost 1984). Thisdoes not imply that these frogs are or wereequally abundant within each region, but onlythat it was potentially possible to find nativeranid frogs in the each of the regions, had theseregions not been differentially occupied byexotic species. Our experience and the moreextensive museum voucher record in Arizonaand New Mexico support this assumption. Onearea we visited during 2005, Río Sonoyta, hasnever yielded leopard frog specimens, and webelieve they were not present in historic times,and have therefore excluded it from the ranidfrog analysis. This site will be reported on else-where (Rosen et al., this volume).

RESULTSWe observed 16 species of aquatic herpeto-fauna (Table 3), defined to include thosespecies requiring perennial water or at leastusing it occasionally for breeding. We alsoobserved crayfish, in Chihuahua only (Table3), and at least 14 species of fishes (Table 4).

Aquatic ReptilesAlthough all historic museum localities that werevisited were considered suitable for mudturtles, we found mud turtles at only 8 of 15(53 %) of them. Not all of the other visitedsites were considered suitable for mud turtles:some were isolated and semi-perennial,concreted springs, or very large reservoirs, and

these were considered unsuitable (i.e., non-habitat) and, in the case of large reservoirs,were impractical to survey adequately in anycase. Of the non-historic, but apparently suit-able areas sampled, 12 of 17 (71 %) wereoccupied, not significantly different from theproportion of historic localities where wefound turtles (χ2 = 2.0, p > 0.1), indicating, notunexpectedly, a wider distribution of mudturtles than defined by the museum localities.Major impacts on turtle populations were

apparent in the Guzman Basin (Ríos delCarmen, Santa Maria, and Casas Grandes) ofnorthwestern Chihuahua. Río Casas Grandessupported many exotic species reported orsuspected to impact the Sonoran mud turtle(Rosen and Schwalbe 2002a; Rosen 2008),including bullfrogs, crayfish, and commoncarp. Habitat desiccation and modification waswidespread, and we could not capture any mudturtles in this region. Similarly, near Galeana,Chihuahua, drought and agricultural water usewere reported to have completely desiccatedall major streams, springs, and cienegas. Wefound only two terrestrially active rough-footed mud turtles at a recently desiccatedhistoric locality at which turtles had previously(J. Iverson, pers. comm. 2009) and recently (S.McGaugh, pers. comm. 2005) been very abun-dant. Although drought conditions werewidespread in the region, local people weinterviewed indicated the widespread failureof springs was a recent phenomenon, and theyattributed it to groundwater pumping for agri-culture. Thus, although aquatic turtles arepersisting in the Guzman Basin (Table 3), wewere able to locate little suitable habitat forsampling during our 1.5-day survey effort.In Sonora, most of the flow of major

streams including the Río San Miguel de Horc-asitas, Río Sonora, and Río Moctezuma wasbeing diverted by small, established or tempo-rary, diversion dams and levees, and almost allwater deep enough for the Sonoran mud turtlewas localized as a result. Small habitat areasremained at the diversion structures, or inwidely scattered scour pools adjoining or nearthe main channel. Although we captured turtlesin some of these streams, it was apparent that


Table 3. Herpetofauna, and crayfish, observed at localities surveyed during Fall 2005 in Sonora andChihuahua. Symbols are as follows: r = rare, u = uncommon, c = common, a = abundant, s = super-abundant. A + indicates “present”, without estimate of relative abundance. Question mark indicates thespecies was observed but identification was not confirmed by capture.

LocalityNumber

1 - - - u - - - - - - - - - - - - - 12 - - - u - - - - - - - - - - - - - 13 u - - - - - - - - - - - - - - - - 14 - - - - - - - - - - - - - - - - - 05 s - - - - + - - - a - - c - - - - 46 u - - - - - - - - u - - - - - - - 27 u - u - c - - - + u - - - - + - - 68 - - - - - - - - - - - - - - - - - 09 - - - c - - - - - - - - - - - - c 210 - - - a - - - - - u? - - u? + - - - 411 - - - a - - - - - - - - - - - - - 112 - - - u - - - - - a - - - - - - u 313 - - - - - - - - - - u - - - - - - 114 - - - c - - - - - - - - - - - - - 115 - - - u - - - - - - c - - - - - c 316 - - - c - - - - - - a - - - - - c 317 - - - c - - - - - - - - - - - - u 218 - - - u - - - - - - - - - - - - c 219 - - - - - - - - - - a - - - - - u 220 - a - - - - - - - s - - - - - - - 221 - - - - - - - - - - - - - - - - - 022 - - - a - - - - - s - - - - - + - 323 u - - - - - - - - - - - - - - - - 124 c - - - - - - - - - - - - - - - - 125 a - - - - - - - - - - - - - - - - 126 u - - - - - - - - r - - - - - - - 227 c - - - - - - - - a - - a - - - - 328 r - - - - - - - - c - - - - - - - 229 - r - - - - - - - - - - - - + - - 230 - u - - - - - - - - - a - - + - - 331 - u - - - - - - - - - a - - - - - 232 - a - - - - - - - - - a - - - - - 233 - a - - - - - - - - - - - - - - - 134 - - - - - - - - - - - u - - - - - 135 - c - - - - - - - - - c - - - - - 236 - - - - - - - + - - - a - - - - - 237 - a - - - - - + - - - - - - - - 238 - - - - - - - - - - - a - - - - - 139 - u - - - - + - - - - - - - - - - 2

Total Localities10 9 1 12 1 1 1 2 1 10 4 7 3 1 3 1 7 74

Rana

yavapa

iensis

R.ma

gnaocu

laris

R.tarahu

marae

R.cat

esbeia

naHyla a

renico

lorSm

ilisca

fodien

sS. b

audini

Leptod

actylu

smelan

onotus

Gastro

phryne

olivac

eaKin

ostern

onson

oriens

eK.hirtipes

K.integr

umTra

chemys sc

ripta y

aquia

Chrys

emyspic

taTham

nophis

cyrtop

sisT. e

ques

Orcon

ectescf.

virilis

TotalSpecies


Table 4. Fishes observed at turtle survey localities surveyed during Fall 2005 in Sonora and Chihuahua.Symbols are as in Table 3. Details for determination of abundance categories and species identity are intext.

Native Fish Species Exotic Fish Species

LocalityNumber

1 a - - - c - a c - r - - - - - 52 - - - - - - - - - - - - - - a 13 a c - u - u a - - - - - - - - 54 u - - - - - - - - - - - - - - 15 - c - c - c a - - - - - - - - 46 - - - - - - - - - - - - - - u 17 - - - - - - - - - - - - - - - 08 - - - - - - - s u - a - - a - 49 - - - - - - - - c c - - c? - - 310 - - - - - - - - - - - - - - c 111 - - - - - - - - - - - - - - c 112 - - - - - - - - - - - - - - u 113 - - - - - - - - - - - - - - - 014 - - - - - - - - - - - - - - - 015 - - - - - - - c - - - - - - c 216 - - - - - - - - - - c - - - c 217 - - - - - - - - - - - - - - - 018 - - - - - - - - - - c - - - - 119 - - - - - - - - - - - - - - u 120 - - - - - - - - - u - - - - a 221 - - - - - - - - - - - - - a a 222 - - - - - - - - - - - - - a a 223 - - - - - - - - - - - - - - - 024 a - - - - - c - - - - - - - u 325 - - - - - - - - - - - - - - ? 026 a - - - - - c - - - - - - - - 227 - - - - - - - - - - - - - - s 128 - - c - - - - - c u - a - - a 529 - - - - - - - - - - - - - - - 030 - - - - - - a - - - - - - c u 331 - - - - - - - - - - - - - r - 132 - - - - - - - - - - - - - c u 233 - - - - - - - - - - - - - - - 034 - - - - - - - - - - - - - a - 135 - - - - - - - - - - - - - - - 036 - - - - - - - - - - - - - c - 137 - - - - - - - - - - - - - - 038 - - - - - - - - - - - - - - - 039 - - - - - - - - - - - - - - - 0

# Occurrences5 2 1 2 1 2 6 3 3 4 3 1 1 8 16

Agosia c

hrysog

aster

Camp

ostom

a orna

tumCatostom

usber

nardin

iC.wig

ginsi

Gila d

itaeni

aG.

eremic

aPoeciliopsis sp.

Ameiu

rusmelas

Cyprinus

carpio

Lepom

is cyan

ellus

Gamb

usia a

ffinis

Micro

pterus

salmo

ides

Pimeph

ales prom

elas

Tilapi

a sp.

Unidentified Fish

TotalSpecies


diversions simplified the channels, leavingthem almost uniformly very shallow. Streamdepletion by agricultural diversion at small ormedium-sized dams was observed at localities3, 6, 8, 18, 21, 24-27, and 30, and probably atlocality 31. We never observed the large Yaqui slider,

which is found in deep water environments(unpubl. obs. 2006-2009, Figure 2), in thesesituations. This form of habitat modificationaffects it severely, even in larger streams suchas Río Bavispe, although habitat suitable for ithas been created behind some of the medium-sized diversion dams. In areas with such diversions, most habitat

suitable for Sonoran mud turtles had been lost.Sonoran mud turtles were rarely observed in

the main stream channel, except in pools at thediversion itself. They occurred primarily inmarshy backwaters produced by inflow scourand off-channel scour pools. By contrast, wefound the Mexican mud turtle abundant undera wide variety of conditions, including streamsimpacted by water diversion, such as RíoMátape (Figure 3). Black-necked gartersnakes (Thamnophis

cyrtopsis) were observed at three localities, butMexican gartersnakes (T. eques) were found atonly one locality — the ciénega at the bridgecrossing at Tubutama on Río Altar — wherethree were found in under two hours ofsampling despite the presence of exotic cich-lids and bullfrogs. Most of our survey efforttook place under conditions too cool for high

Figure 2. A large pool behind a diversion dam on Río Bavispe north of Huasabas, in whichYaqui Sliders (upper right) were abundant. Photographs by P. C. Rosen, 29 October2005.


activity of gartersnakes, and therefore we havenot evaluated gartersnake status except toprovide presence data (Table 3).

Aquatic FrogsWe found the introduced American bullfrog at3 of 4 sites in the Concepción basin (Río Altarand Río Magdalena) in northwestern Sonoranear Arizona, and 9 of 10 sites in the endorheicGuzman Basin in northwestern Chihuahua,both near the United States. In contrast, nobullfrogs were seen in the interior of Sonoraand Chihuahua (0 of 25 sites). Leopard frogsshowed the opposite pattern, being seen only,and regularly, in the interior (19 of 25 sites,with a mean relative abundance of 1.9 ± 0.31SE). This pattern is highly significant (χ2 =38.3, p < 0.001).

We detected similar numbers of fish speciesat bullfrog and non-bullfrog sites (1.58 versus1.44 species / site), but there were more exoticfish species detected at bullfrog versus non-bullfrog sites (1.33 ± 0.26 versus 0.85 ± 0.22;t = 1.41, P < 0.09) and fewer native fish species(0.25 ± 0.25 SE versus 0.59 ± 0.25 SE; P >0.2). Similarly, using the index derived fromrank abundance values, exotic fish were moreprominent, and again non-significantly, at bull-frog versus non-bullfrog sites (3.08 ± 0.77versus 2.13 ± 0.70; t = 0.92, P < 0.19).However, considering the index of exoticspecies dominance (the summed abundanceranks for exotic fishes / summed abundanceranks for all fishes) exotic fish were signifi-cantly more dominant at bullfrog versusnon-bullfrog sites (0.54 ± 0.10 SE versus 0.33

Figure 3. Shallow stream micro-habitat in Río Mátape Basin occupied by the MexicanMud Turtle. Photographs at Rancho Ojo de Agua, by P. C. Rosen, 5 November 2005.


± 0.08 SE; t = 1.72, P = 0.049). Crayfish abun-dance was more strongly and positivelyassociated with bullfrog occurrence (1.33 ±0.41 SE versus 0.07 ± 0.07 SE; t = 2.99, P <0.01); as such, the bullfrog and crayfish effectscannot be statistically separated in this dataset.To further explore the relative weights of

bullfrog presence-absence and measures ofexotic fish predominance, we regressed theexotic fish measures (species richness, indexof abundance, and index of dominance) againstthe rank-abundance values for leopard frogs inthe upland regions from which bullfrogs andcrayfish were absent. For fish species richness,these data yielded no significant or nearlysignificant correlations, whereas there was asignificant correlation between the index ofexotic predatory fish abundance and rankedleopard frog abundance (r = 0.406, n = 24, P =0.049; for the index of exotic predatory fishdominance this statistic was marginally signif-icant: r = 0.366, n = 24, P = 0.078).To attempt to express, in a single analysis,

the relative importance of the bullfrog-crayfishand exotic fish associations with leopard frogabundance, we computed the analysis ofcovariance of the leopard frog abundanceusing bullfrog plus crayfish presence-absenceas a category and the abundance index forexotic predatory fishes as the covariate. Thisanalysis also suggested that bullfrog-crayfishpresence was a stronger effect (P < 0.0001)than exotic predatory fish (P = 0.03; Table 5),although crayfish and bullfrog effects couldnot be statistically separated. In the interior of Sonora, native leopard

frogs were often seen in moderate to highabundance, especially where native fishespredominated. The most spectacular examplewas at Saracachi Ciénega, in the Río Sonorabasin. This largely intact, extensive headwatermarsh of Río San Miguel de Horcasitassupported 2 turtle species, 5 fishes, and at least4 anurans (Tables 3 and 4; an additional twoanurans, the Couch’s spadefoot [Scaphiopuscouchii] and red-spotted toad [Bufo punctatus]were also observed), including thousands ofadult and subadult lowland leopard frogs. Thisspecies count is especially notable because all

aquatic vertebrate taxa observed at Saracachiwere native, with several of them occurring inremarkable abundances. The widespread pres-ence of native leopard frogs in the bullfrog-freeinterior of Sonora (Table 3) was notablyobserved even in places (e.g., upper RíoBavispe) where non-native fishes were diverseand numerically predominant (Table 4).

DISCUSSIONStatus of Aquatic Turtles

Observations of habitat conditions madeduring our survey indicated that aquatic turtlespecies in northwestern Mexico were beingimpacted by water diversions. Simply put, thewater was so shallow in reaches downstreamfrom these diversions that certain specieswould not be expected, and were rarely ornever detected, in the diversion-modifiedhabitat. Diversions of this kind were seenwherever canyon bottoms were broad enoughfor small-scale farming; this included most ofthe stream areas accessible in our survey.Although this impact is undoubtedly smallerin remote, rugged canyons, such canyon-bound, scour-prone environments are also lesssuitable for turtles than the slower, deeper,more productive waters of fertile lowlands. Webelieve that none of the aquatic turtles westudied are rare or threatened in Sonora andnorthwestern Chihuahua, but that all havelikely been affected to some degree by habitatloss. Water diversion impacts on individual turtle

species can be judged by interspecific varia-tion in habitat needs and predation resistanceknown or inferred for members of the turtleassemblage. For example, the large Yaquislider would be highly vulnerable to predationand collection by humans without deep waterfor escape; we have seen it at many localitiesacross Sonora, but never in shallow-waterenvironments. Similarly, the Sonoran mudturtle is generally uncommon or rare in theSonoran Desert region where it lacks deepwater, available shelter in tree roots or undercutbanks, and often with muddy substrata; preda-tion has been noted when shallow waterexposes this species to medium-sized stream-


side predators (Rosen, unpublished observa-tions at Tule Creek, Yavapai County, Arizona).We expect that the rough-footed mud turtle,which is morphologically similar and hassimilar habitat utilization, will be affected likethe Sonoran mud turtle. The Mexican mudturtle provided a telling contrast. This rela-tively large mud turtle is able to closeremarkably tightly within its shell, and thisapparently permits it to remain active, and topersist as populations, in shallow waters rarelyutilized by other regional Kinosternon. Thesight of these animals clambering over sand inscant centimeters of water depth can be strik-ingly reminiscent of the terrestrial capabilityafforded terrestrial box turtles (Terrapene) bytheir tight shell closure. Based on these consid-

erations, and the occurrence and abundance weobserved during the survey, we rate turtlesusceptibility to water diversions as follows:Yaqui slider > Sonoran mud turtle ≈ rough-footed mud turtle > Mexican mud turtle. Our survey covered little of the level terrain

of the Altiplano or lowland plains, whereimpacts of habitat modification and introducedanimals are most potent (Hendrickson et al.1980; Juárez-Romero et al. 1991; Miller et al.2005). Our observations in the Guzman Basinand in the lowland reservoirs we examined inSonora (Presa Molinito and Presa Cuatemoc)suggest that turtles and other aquatic herpeto-fauna are indeed suffering severe impacts fromhabitat desiccation and modification in broadvalleys and on the coastal plain. However, we

Table 5. Analysis of covariance of leopard frog (Rana yavapaiensis, R. magnaocularis) abundance in north-western Mexico in relation to American bullfrog (R. catesbeiana) and northern crayfish (Orconectes virilis)presence/absence and predatory exotic fish abundance.

Summary of FitR2 0.40R2 – adjusted 0.35Root Mean Square Error 1.28Mean of Response 1.33

Analysis of VarianceSource DF Sum of Squares Mean Square F Ratio

Model 3 39.0 13.0 7.9Error 35 57.6 1.6 Prob > FC. Total 38 96.7 0.0004

Parameter EstimatesTerm Estimate Std Error t Ratio Prob > |t|Intercept 2.25 0.285 7.87 < .0001Bullfrog-Crayfish Presence / Absence -1.94 0.446 -4.34 0.0001Predatory Exotic Fish Abundance -0.25 0.110 -2.24 0.0319(Bullfrog P/A) x (Predatory Exotic Fish - 1.26) 0.25 0.247 1 0.3256

Effect TestsSource DF Sum of Squares F Ratio Prob > FBullfrog-Crayfish Presence / Absence 1 31.07 18.87 0.0001Predatory Exotic Fish Abundance 1 8.23 5.00 0.0319(Bullfrog-Crayfish) x (Predatory Exotic Fish) interaction 1 1.64 0.99 0.3256


subsequently (2006) observed Yaqui sliders incanals on the heavily agricultural Yaqui Riverfloodplain west of Ciudad Obregón.

Native and Non-native FishesEven without intensive fish sampling, patternsimportant for conservation biology wereevident that are consistent with broaderpatterns for Mexico described in Miller et al.(2005). Non-native fishes are now dominantin a number of major waters: the currentterminus of Río Sonora at Presa Molinito nearHermosillo, the two areas we sampled in RíoAltar, the upper Río Bavispe near Bavispe, andthe Guzman Basin in Río Casas Grandes andRío Santa Maria. Native fishes predominatedin many other waters, including RíoMagdalena and throughout most of the interiorhighlands of Sonora (Table 4). Most of our records of non-native fishes in

the Río Yaqui Basin are similar to thosereported by Hendrickson et al. (1980) based onsurveys nearly three decades earlier, but ourrecords add the green sunfish (Lepomiscyanellus) at Yecora. Rinne and Minckley(1991) suggested that non-native fish invasionin Mexico may be 2-3 decades lagged behindthat in the U.S., whereas Unmack and Fagan(2004) estimated a 4-5 decade lag comparingthe Río Yaqui Basin to the Gila River Basin ofthe United States. Our analysis does not helpdiscriminate between these estimates, but doesdemonstrate that the highlands of the Yaquiand other basins are considerably less impactedby the invasion of exotics than surroundinglowland aquatic environments. A review of the published literature

(Hendrickson 1983; Campoy-Favela et al.1988; Hendrickson and Juarez-Romero 1990;Abarca et al. 1995; Miller et al. 2005) suggeststhat the following exotic fish locality records,in addition to green sunfish at Yecora, are alsonew: non-native African cichlids (Tilapia sp.)abundant in mainstem Río Altar at Tubutama(two localities), at Quitovac (Rosen et al., thisvolume), and in Río Mátape (Juarez-Romeroet al. 1988); and black bullhead catfish(Ameiurus melas) in Río Magdalena belowImuris, Sonora.

Introduced Species and Native Ranid Frogs

Some other records for exotic species in thestudy region are also new. Most of theobserved bullfrog localities are new records,and the apparent absence of this frogthroughout most of Sonora was unexpected,based on experience in Arizona. Exotic cray-fish (Orconectes cf. virilis), previouslyreported in Chihuahua (Hobbs 1989), wereonly observed in Guzman Basin streams andat Yepomera, Chihuahua. All of the exoticspecies recorded on this survey may be contin-uing to spread. Tracking, understanding, andworking toward a solution to this problem is acritical element for native biodiversity conser-vation in the region.The frequency of leopard frog observa-

tions during our surveys, and the observedabundances, suggested better populationstatus in Sonora than in Arizona. The remark-able superabundance of leopard frogs atSaracachi Ciénega is described below; else-where, leopard frog abundances wereconsistently moderate to high in all areaswhere habitat appeared suitable, except inlowlands where multiple exotic species werepredominant. Even in marginal habitat condi-tions or isolated waters, we often foundleopard frogs. These results resemble 1971-1976 observations by John Frost in Arizona(Frost, unpublished field notes, P. Fernandez,pers. comm. 1996), but since then equivalentabundances of native ranid frogs have notbeen reported in Arizona. We did not findleopard frogs in major reservoirs, as expectedbased on previously known effects of exoticfish species that were typically dominant inthe reservoirs.It was also surprising to find native leopard

frogs persisting even at the upper Río Bavispewhere non-native fishes have dominated thesystem for several decades (Hendrickson et al.1980; Leibfried 1991; Abarca et al. 1995).Mechanisms permitting native leopard frogsto persist as they have in Sonora deserveinvestigation, and the rate at which these frogpopulations may be declining should be eval-uated.


The lack of observed co-occurrence ofleopard frogs with introduced bullfrogs andcrayfish in northwestern Sonora was not unex-pected, although it was more similar to resultsfor the Chiricahua leopard frog in southeasternArizona (Rosen et al. 1995; Rosen andSchwalbe 1995, 2001, 2002b) than for thelowland leopard frog in Arizona (Sartorius andRosen 2000). Although many authors haverecognized bullfrogs as a primary cause ofranid frog declines in the American West,others have suggested that non-native fishesmay have stronger negative impacts (Hayesand Jennings 1986; Adams 1999, 2000). The results of our analysis of covariance

(Table 5) suggest that bullfrogs or bullfrogstogether with crayfish may have greater impactthan exotic fishes on leopard frogs. Perhapsexotic fish impacts are intensifying (e.g.,Unmack and Fagan 2004) and will becomeprogressively more important. Predatory nativefishes that may be replaced by exotics mighthave already limited leopard frog populationsunder original conditions, so exotic fishes maynot alter the predation and competitive regimesfor leopard frogs in the ways bullfrogs might.The bullfrog may also be vectoring theemerging amphibian disease chytridiomycosis(Garner et al. 2006), but this pathogen is wide-spread in Sonora even in large regions wherebullfrogs are absent (Hale et al. 2006), sodisease transmission by bullfrogs seems anunlikely explanation for bullfrog impacts inSonora. Two alternative ecological mechanisms

strengthening bullfrog impacts may include:(1) not only do adult and large juvenile bull-frogs prey on native ranids, but bullfrogtadpoles can be strong competitors with otherranids (Kupferberg 1997); and (2) bullfrogsand leopard frogs are phylogenetically relatedand hence ecologically similar in myriad ways,including subtle aspects of habitat and micro-habitat utilization, competition, and predationsusceptibility. As an example, leopard frogsutilize peripheral or semi-perennial pools inriver bottoms where interactions with fish areminimal, but bullfrogs are also able, and likely,to invade such sites.

Perhaps, however, the apparent priority ofbullfrog impacts reflects the current restrictionof bullfrogs in our survey areas to lowlandregions that are heavily impacted by anthro-pogenic habitat modification and have highdiversities of other exotics, particularly cray-fish, which have known impacts onSouthwestern ranid frogs (Fernandez andRosen 1996). We suspect, based on theirobserved abundances, and based on theiroccurrence in assemblages of exotics withabundant species known to impact leopard frogpopulations, that crayfish may not have beenprimary impacts causing leopard frog popula-tion losses in northwestern Mexico. However,further investigation involving fieldwork andexperimental conservation will be required toclarify this issue.

An Exemplary Site: Saracachi CiénegaWe have visited most, if not all, major ciénegasin the U.S. Southwest, as well as several innorthernmost Sonora: most of these sites havebeen ecologically dominated by exotic fishesand bullfrogs for two or more decades, andmany have been severely degraded by anthro-pogenic erosion (Hendrickson and Minckley1985). We therefore found our observations atSaracachi Ciénega, at Rancho Agua Fria, to beremarkable in several ways. The site is largelyintact, with little evidence of erosion. Struc-turally, it may be the most natural, leastimpacted major ciénega remaining in the South-west; we believe there is no equivalent ciénegaexample remaining in the U.S., either from thestandpoint of aquatic vertebrate species diver-sity or the absence of exotic aquatic vertebrates.No ciénega in the U.S. Southwest supports suchhigh diversity of native fishes (maximum 3species, at Empire Ciénega) and aquatic turtles(1 species, Sonoran mud turtle, being the norm).Other ciénegas in northern Sonora, such as thoseat Cocospera and Los Fresnos, are alreadyinvaded by bullfrogs (J. Rorabaugh, T. Jones,pers. comm. 2006). The abundance of leopard frogs at Sara-

cachi equaled or exceeded the superabundanceof bullfrogs found in comparable Arizonaciénegas (Rosen and Schwalbe 1995),


including extreme examples observed atArivaca and San Bernardino National WildlifeRefuge (Schwalbe and Rosen 1988). Thissupports inferences that valley bottom ciénegawetlands originally were critical parts of theoriginal leopard frog population of southernArizona, and may have supported core abun-dances that drove regional metapopulationsource-sink and related dynamics. Weobserved 24 egg masses during our singlereconnaissance at Saracachi, during what isusually the secondary breeding season for thelowland leopard frog (Collins and Lewis 1979;Frost and Platz 1983; Sartorius and Rosen2000). Leopard frog reproductive success wasapparently supported by extensive shallow,productive waters with open aquatic vegeta-tion maintained by moderate livestock grazingon the un-incised ciénega bottomland (Figure4). These observations of leopard frogs there-fore carry implications for recovery prospects

for threatened species (USFWS 2002). Sara-cachi Ciénega offers outstanding opportunitiesfor basic and conservation-oriented ecologicalresearch, and deserves high conservationpriority.

Modes of Persistence of Native AquaticVertebrates in Arid Northwestern MexicoThe persistence and abundance of leopardfrogs as well as native fishes in northwesternMexico appears to be associated with thedelayed and — thus far, compared to impactsin the United States — relatively limitedimpact of the biological invasion by non-native predatory fishes (Unmack and Fagan2004) and crayfish and bullfrogs. However,the regional persistence of native taxa in ourdataset is markedly uneven: in the lowlandsnear Hermosillo, Sonora, and Casas Grandes,Chihuahua, we observed aquatic vertebratefaunas in much the same degraded condition

Figure 4. Lowland Leopard Frog egg-laying site with extensive, moderately vegetatedshallows at Saracachi Ciénega, Sonora, 23 September 2005. Photograph by P. C. Rosen.


found over most lowland areas in the U.S.Southwest (Campoy-Favela et al. 1989;Juárez-Romero et al. 1991; Varela-Romero etal. 1992a & b for related examples). NearAltar, Sonora, we believe degradation likethat seen in the U.S. was currently underwayat the time of our survey, and unpublishedreports regarding headwater ciénegas in theSan Pedro and Santa Cruz river basin innorthern Sonora (J. Rorabaugh, D. Duncan,pers. comm. 2006-2009) point toward thissame conclusion.The parallel, but time-lagged, buildup of

exotics in the Gila River Basin and in interiorhighlands of Sonora (Unmack and Fagan2004) is not a simple lock-step with time-lag,but reflects two pairs of divergent forces: (1)the smaller number of major dams in interiorSonora in tandem with habitat modification bysmall-scale water diversions, and (2) the lessintensive and extensive government involve-ment in propagation and dispersal of exoticaquatic vertebrates in northwestern Mexicocompared to the U.S. Southwest. The exampleof exotic species problems in the U.S., and thedifficulties they create for biodiversity conser-vation, may motivate proactive conservationprograms in Mexico at an earlier stage ofdevelopment of the exotic species advance(e.g., Varela-Romero et al. 2004). Whereas dams create habitat, refugia, and

source populations favoring exotic species,shallow, heavily flood-affected waters favornative aquatic species (Meffe 1984; Minckleyand Meffe 1987; Sartorius and Rosen 2000).Although the widespread diversion of watersfrom streams onto small arable floodplain landsin Sonora has negative impacts on turtles, asdescribed above, it also keeps many streamreaches shallow, flood-prone, and thus unsuit-able for a number of exotic species that wouldotherwise likely be having greater impacts onnative aquatic species. It seems plausible that thelarge- and small-scale patterns of water diver-sion described above might be combined withenvironmentally foresightful governmentalpolicy to prevent the U.S. exotic species biodi-versity catastrophe from being replayed in thehighlands of northwestern Mexico.

ACKNOWLEDGMENTSWe thank Jesús Alaya, Luis Castillo, Ana LuisaGallardo, Yadid León, Suzanne McGaugh,Juan Miranda, José Peraza, Daren Riedle,Martín Villa, and the staff of Sierra de AlamosBiosphere Reserve for assistance with accessto and sampling at localities, as well as manyother courtesies. The project was carried outduring sampling funded by Arizona Game andFish Department. The work was conducted inaccordance with permit number 11543 fromthe Secretaría de Medio Ambiente y RecursosNaturales (SEMARNAT), Mexico.

SUMMARYA total of 39 localities were sampled in Sonoraand northwestern Chihuahua during a 24-dayrapid assessment survey conducted in tandemwith mud turtle genetics sampling during fall2005. Mud turtles were generally abundant,and were captured at 8 of 15 historic localitiesand 12 of 17 newly sampled localities withpotentially suitable habitat (Sonoran mudturtle, 4 of 9 historic, plus 6 of 9 newlysampled sites; rough-footed mud turtle, 4 of 5historic; and Mexican mud turtle, 1 of 2historic plus 6 of 8 new localities). Althoughthe Sonoran mud turtle probably had not beenextirpated from any major locality, river diver-sions for agriculture restricted it to localizeddeeper-water micro-sites in some localities.The rough-footed mud turtle was impacted bydesiccation of springs near Galeana,Chihuahua, whereas the Mexican mud turtleappeared tolerant of shallow water. Theendemic Yaqui Slider was recorded at only 2localities, and in many areas water diversionhas probably eliminated much of its habitat.Exotic crayfish were found only in Chihuahua.Exotic fishes were dominant in certain majorwaters, whereas native fishes predominated inmany others. Although native leopard frogswidely co-occurred with exotic fishes, theirabundance was negatively correlated withmeasures of exotic fish abundance, moststrongly involving the predominance of exoticpredatory fishes. Exotic American bullfrogsoccurred at most localities in northwesternChihuahua, where crayfish also occurred


widely, and northwestern Sonora, and neitherbullfrogs nor crayfish were detected in interiorSonora. Native Lowland and Big-eyedLeopard Frogs had a pattern opposite that ofbullfrogs and crayfish, and were found at 19of 25 localities sampled in interior Sonora,suggesting a particularly strong negativeimpact of bullfrogs, and possibly crayfish, onleopard frogs. As in native fishes, impacts ofexotic species on aquatic herpetofauna innorthwestern Mexico appear to have laggedbehind the impacts of exotics and habitatmodification in adjacent areas of the UnitedStates, but our observations indicate substan-tial impacts that are very likely expanding. Wenote sites with outstanding habitat character-istics, some, particularly Saracachi Cienega,deserving special protection.

LITERATURE CITEDAbarca, F. J., K. L. Young, B. L. Jensen, I. Parra, R.H. Bettaso, and K. Cobble. 1995. Yaqui Riverfishes relevant to the Madrean Province: U.S.-Mexico Collaborations. In Biodiversity andmanagement of the Madrean archipelago: The skyislands of southwestern United States and north-western Mexico, techincal coordinators L. F.DeBano, P. F. Ffolliott, A. Ortega-Rubio, G. J.Gottfried, R. H. Hamre, and C. B. Edminster, pp.370–378. USDA Forest Service, General Tech-nical Report RM-GTR-264.Adams, M. J. 1999. Correlated factors in amphibiandecline: Exotic species and habitat change inwestern Washington. Journal of Wildlife Manage-ment 63: 1162–1171.Adams, M. J. 2000. Pond permanence and theeffects of exotic vertebrates on anurans. Ecolog-ical Applications 10: 559–568.Branson, B. A., C. J. McCoy, Jr., and M. E. Sisk.1960. Notes on the freshwater fishes of Sonorawith an addition to the known fauna. Copeia 1960:217–220. Campoy-Favela. J., A. Varela-Romero, and L.Juárez-Romero. 1989. Observaciones sobre laictiofauna nativa de la cuenca del Río Yaqui,Sonora, México. Ecologica 1(1): 1–29.Campoy-Favela. J., A. Varela-Romero, L. Juárez-Romero, and D. A. Hendrickson. 1988. Apreliminary report on the ichthyofauna of the RíoAros, sub-basin of the Río Yaqui basin, Sonora,Mexico. Proc. Desert Fishes Council 19(1987):28.Collins, J. P. and M. A. Lewis. 1979. Overwinteringtadpoles and breeding season variation in the Ranapipiens complex in Arizona. Southwestern Natu-ralist 24: 371–373.

Contreras-Balderas, S. and M. A. EscalanteCavasos. 1984. Distribution and known impactsof exotic fishes in Mexico. In Distribution,biology, and management of exotic fishes, editedby W. R. Courtenay, Jr., and J. R. Stauffer, Jr., pp.102–130. Johns Hopkins University Press, Balti-more, MD.Crother B. I. (ed.) 2001. Scientific and StandardEnglish Names of Amphibians and Reptiles ofNorth America North of Mexico, with CommentsRegarding Confidence in Our Understanding.SSAR Herpetological Circular 29.Crump, M. L., and N. J. Scott, Jr. 1994. Visualencounter surveys. In Measuring and monitoringbiological diversity: Standard methods foramphibians edited by W. R. Heyer, M. A.Donnelly, R. W. McDiarmid, L. C. Hayek, and M.S. Foster, pp. 84–92. Smithsonian InstitutionPress, Washington, D.C.Fernandez, P. J. and P. C. Rosen. 1996. Effects ofthe introduced crayfish Orconectes virilis onnative aquatic herpetofauna in Arizona. FinalReport to Arizona Game & Fish Dept. HeritageProgram, Phoenix, AZ.Frost, J. S. and J. T. Bagnara. 1976. A new speciesof leopard frog (Rana pipiens Complex) fromNorthwestern Mexico. Copeia 1976: 332–338. Frost, J. S. and J. E. Platz. 1983. Comparativeassessment of modes of reproductive isolationamong four species of leopard frogs (Rana pipienscomplex). Evolution 37: 66–78.Garner, T. W. J., M. W. Perkins, P. Govindarajulu,D. Seglie, S. Walker, and A. A. Cunningham.2006. The emerging amphibian pathogen Batra-chochytrium dendrobatidis globally infectsintroduced populations of the North Americanbullfrog, Rana catesbeiana. Biology Letters 2:455–459.Hale, S. F., P. C. Rosen, J. L. Jarchow, and G. A.Bradley. 2006. Effects of the chytrid fungus on theTarahumara frog (Rana tarahumarae) in Arizonaand Sonora, Mexico. In Connecting mountainislands and desert seas: Biodiversity and manage-ment of the Madrean Archipelago II Proceedings,complied by G. J. Gottfried, B. S. Gebow, L. G.Eskew, and C. B. Edminster, pp. 407–411. Tech-nical Report RMRS-P-36. U.S. Department ofAgriculture, Forest Service, Rocky MountainResearch Station. Fort Collins, CO. Hayes, M. P., and M. R. Jennings. 1986. Decline ofranid frog species in western North America: Arebullfrogs (Rana catesbeiana) responsible? Journalof Herpetology 20: 490–509.Hendrickson, D. A. 1983. Distribution records ofnative and exotic fishes in Pacific drainages ofnorthern Mexico. J. Ariz.-Nev. Acad. Sci. 18: 33–38.Hendrickson, D. A. and L. Juárez-Romero. 1990.Los peces de la cuenca del Río de la Concepción,Sonora, México, y el estatus del charalitosonorense, Gila ditaenia, una especie en amenaza


de extincion. Southwestern Naturalist 35: 177–187.Hendrickson, D. A., and W. L. Minckley. 1985.Ciénegas—vanishing climax communities of theAmerican Southwest. Desert Plants 6: 131–175.Hendrickson, D. A., W. L. Minckley, R. R. Miller,D. J. Siebert, and P. H. Minckley. 1980. Fishes ofthe Río Yaqui Basin, Mexico and United States.J. Ariz.-Nev. Acad. Sci. 15: 65–106.Hobbs, H. H. 1989. An illustrated checklist of theAmerican crayfishes (Decapoda: Astacidae,Cambaridae, and Parastacidae). SmithsonianContributions to Zoology 480.Holycross, A., W. P. Burger, E. J. Nigro, and T. C.Brennan. 2006. Surveys for Thamnophis equesand Thamnophis rufipunctatus in the Gila RiverWatershed of Arizona and New Mexico. Finalreport to Arizona Game and Fish Department,Phoenix, AZ.Juárez-Romero, L., A. Varela-Romero, and J. R.Campoy-Favela. 1988. Observacionespreliminares sobre la ictiofauna del Río Mátape,Sonora, México. Memorias IX del CongresoNacional de Zoologia. Villahermosa, Tabasco.October 13-16, 1987. 9(II): 27–33.Juárez-Romero, L., A. Varela-Romero, and J.Campoy-Favela. 1991. Ecological observations ofnative fishes from the lower Río Yaqui, Sonora,México. Proc. Desert Fishes Council 20(1988):79–80.Kupferberg, S. J. 1997. Bullfrog (Rana catesbeiana)invasion of a California river: The role of larvalcompetition. Ecology 78: 1736–1751.Lannoo, M. 2005. Amphibian Declines: TheConservation Status of United States Species.University of California Press, Berkeley, CA.Leibfried, W. C. 1991. A recent survey of fishesfrom the interior Río Yaqui Drainage, with arecord of the flathead catfish Pylodictis olivarisfrom the Río Aros. Proc. Desert Fishes Council20(1988): 79–80.Liner, E. A. 1994. Scientific and Common Namesfor the Amphibians and Reptiles of Mexico inEnglish and Spanish / Nombres Científicos yComunes en Ingles y Español de los Anfibios ylos Reptiles de México. Society for the Study ofAmphibians and Reptiles, St. Louis, MO.Marascuilo, L. A. and M. McSweeney. 1977.Nonparametric and Distribution-Free Methods forthe Social Sciences. Brooks/ Cole, Monterey, CA.Meffe, G. K. 1984. Effects of abiotic disturbance onco-existence of predator-prey fish species.Ecology 65: 1525–1534.Miller, R. R., W. L. Minckley, and S. M. Norris.2005. Freshwater Fishes of Mexico. University ofChicago Press, Chicago, IL.Minckley, W. L., and J. E. Deacon (editors). 1991.Battle Against Extinction: Native Fish Manage-ment in the American West. University of ArizonaPress, Tucson, AZ.Minckley, W. L., and G. K. Meffe. 1987. Differen-

tial selection by flooding in stream-fish commu-nities of the arid American Southwest. InCommunity and Evolutionary Ecology of NorthAmerican Stream Fishes, edited by W. J.Matthews and D.C. Heins, pp. 93–104. Universityof Oklahoma Press, Norman, OK.Pauly, G. B., D. M. Hillis, and D. C. Cannatella.2009. Taxonomic freedom and the role of officiallists of species names. Herpetologica 65: 2:115–128. Platz, J. E. and J. S. Frost. 1984. Rana yavapaiensis,a new species of leopard frog (Rana pipienscomplex). Copeia, 1984: 940–948.Quade, D. 1967. Rank analysis of covariance.Journal of the American Statistical Association 62:1187–1200.Rinne, J. N. and W. L. Minckley. 1991. Nativefishes in arid lands: A dwindling resource of theDesert Southwest. General Technical Report, RM-206, USDA Forest Service, Rocky MountainForest and Range Experiment Station, FortCollins, CO.Rosen, P. C. 1987. Female reproductive variationamong populations of Sonoran mud turtles(Kinosternon sonoriense). Master’s Thesis,Arizona State University, Tempe, AZ. Rosen, P. C. 2003. Taxonomic status of the Sonoytamud turtle (Kinosternon sonoriense longifemoraleIverson) based on mitochondrial d-loop sequence,with a discussion of phylogeography. Final reportto Arizona Game and Fish Department, Phoenixand U.S. Fish and Wildlife Service, Albuquerque,NM.Rosen, P. C. 2008. Aridlands turtles in SouthwesternNorth America. II. A review of conservation statusand problems. Sonoran Herpetologist 21(12):130–135.Rosen, P. C. and C. R. Schwalbe. 1988. Status andecology of the Mexican and narrow-headedgartersnakes (Thamnophis eques megalops andThamnophis rufipunctatus rufipunctatus) inArizona. Unpublished Final Report from ArizonaGame and Fish Dept. (Phoenix, Arizona) to U.S.Fish and Wildlife Service, Albuquerque, NM.Rosen, P. C. and C. R. Schwalbe. 1995. Bullfrogs:Introduced predators in southwestern wetlands. InOur living resources: A report to the nation on thedistribution, abundance, and health of U.S. plants,animals, and ecosystems, edited by E. T. LaRoe,G. S. Farris, C. E. Puckett, P. D. Doran and M. J.Mac, pp. 452–454. U.S. Dept. Interior, National.Biological Service, Washington, D.C.Rosen, P. C. and C. R. Schwalbe. 2001. Bullfrogimpacts on native wetland herpetofauna insouthern Arizona. Final report to Arizona Game& Fish Dept. Heritage Program (IIPAM I97041),Phoenix, AZ and the U.S. Fish and WildlifeService, Albuquerque, NM.Rosen, P. C. and C. R. Schwalbe. 2002a. Wide-spread effects of introduced species on aquaticreptiles and amphibians in the Sonoran Desert


region. In Invasive exotic species in the SonoranDesert region, edited by B. A. Tellman, pp. 220–240. University of Arizona Press, Tucson, AZ.Rosen, P. C. and C. R. Schwalbe. 2002b. Conser-vation of wetland herpetofauna in southeasternArizona. Final report to Arizona Game & FishDept. Heritage Program (IIPAM I99016),Phoenix, AZ and U.S. Fish and Wildlife Service,Albuquerque, NM. Rosen, P. C., C. R. Schwalbe, D. A. Parizek, P. A.Holm, and C. H. Lowe. 1995. Introduced aquaticvertebrates in the Chiricahua region: Effects ondeclining native ranid frogs. In Biodiversity andmanagement of the Madrean archipelago: The skyislands of southwestern United States and north-western Mexico, technical coordination by L. F.DeBano, P. F. Ffolliott, A. Ortega-Rubio, G. J.Gottfried, R. H. Hamre, and C. B. Edminster, pp.251–261. Gen. Tech. Rep. RM-GTR-264.U.S.D.A. Forest Service, Fort Collins, CO.Sartorius, S. S., and P. C. Rosen. 2000. Breedingphenology of the lowland leopard frog (Ranayavapaiensis): Implications for conservation andecology. Southwestern Naturalist 45(3): 267–273.Schwalbe, C. R. and P. C. Rosen. 1988. Preliminaryreport on effects of bullfrogs on wetland herpeto-fauna in southeastern Arizona. In Management ofamphibians, reptiles, and small mammals in NorthAmerica, edited by R. C. Szaro, K. E. Severson,and D. R. Patton, pp. 166–173. General TechnicalReport RM-166, USDA Forest Service, FortCollins, CO.Schwalbe, C. R., and C. H. Lowe. 2000. Amphib-ians and reptiles of the Sierra de Alamos. In The

tropical deciduous forest of Alamos, edited by R.H. Robichaux and D. A. Yetman, pp. 172–199.University of Arizona Press, Tucson, AZ.Shirley, E.A. 1981. A distribution-free method foranalysis of covariance based on ranked data.Journal of Applied Statistics 30: 158–162.Unmack, P. and W. F. Fagan. 2004. Convergence ofdifferentially invaded systems toward invader-dominance: Time-lagged invasions as a predictorin desert fish communities. Biological Invasions6: 233–243.U.S. Fish & Wildlife Service. 2002. Endangered andthreatened wildlife and plants; proposal to list theChiricahua leopard frog as threatened with aspecial rule. Federal Register 65: 115: 37343–37357.Varela-Romero, A., J. Campoy-Favela, and L.Juárez-Romero. 1992a. Fishes of the rios Mayoand Fuerte Basins, Sonora and Sinaloa, México.Proc. Desert. Fishes Council. 22 (1990) (21[1989]): 70–71.Varela-Romero, A., C. Galindo-Duarte, E. Saucedo-Monarque, L. S. Anderson, P. Warren, S.Stefferud, J. Stefferud, S. Rutman, T. Tibbits, andJ. Malusa. 1992b. Re-discovery of Gila inter-media and G. purpurea in northern Sonora,Mexico. Proc. Desert. Fishes Council.22 (1990):33.Varela-Romero, A, G. Ruíz-Campos, and H.Blasius. 2004. Advances during 2003 in the studyand conservation of native fishes in northwesternMexico (Area report). Proc. Desert. FishesCouncil. 35 (2003): 87.

CHALLENGES TO NATURAL RESOURCE MONITORING IN A SMALL BORDERPARK: TERRESTRIAL MAMMALS AT CORONADO NATIONAL MEMORIAL,COCHISE COUNTY, ARIZONA

Don E. Swann, Melanie Bucci, Amy J. Kuenzi, Barbara N. Alberti, and Cecil Schwalbe

Since 1916, the responsibility of the U.S.National Park Service (NPS) has been toprovide for public enjoyment of America'sgreatest natural and cultural resources whilesimultaneously conserving these resources forfuture generations. Although NPS has longrecognized the value of long-term monitoring(e.g., Robbins et al. 1963), only a few long-term data sets exist for any national park. Inthe early 1990s, in response to criticism thatmost parks do not even know what speciesthey have, let alone how park ecosystemsmay be changing over time (NationalResearch Council 1992), NPS initiated aservice-wide Inventory and Monitoring(I&M) Program that was greatly expandedwith the Natural Resource Challenge begin-ning in 1999.

NPS management policy for naturalresources ultimately directs that managementof parks be based on knowledge of resourcesand their conditions (NPS 2006). A majorpurpose of this policy is to monitor (andmanage) biological diversity — in otherwords, to document trends in the inherentintegrity of the entire natural community, notjust one or more individual species (NPS 2006,Chapter 4.1). Although the level of preserva-tion required in parks — explicit protection ofeven the smallest scrap of petrified wood orliving flower — may impact field work (forexample, in the collection of voucher speci-mens), it provides the conceptual frameworkfor monitoring.

Unfortunately, failure to monitor biologicaldiversity is also an important backdrop formonitoring in the Park Service. NPS and otherland management agencies have been criti-cized for their failure to prevent, and evenrecognize, the loss of species from protectedareas. These failures are explicit in studies thatreport high extinction rates of native mammals(Newmark 1995), amphibians (Drost andFellers 1996) and plants (Drayton and Primack1996) in parks, and implicit in inventories andstudies of park biodiversity. The need forinventory and monitoring programs is thusgrounded in public expectations that NPS willbe proactive in recognizing and preventingdamage to all natural resources under theagency’s care, not simply high profile, threat-ened, or endangered species.

Nevertheless, financial realities, includingthe ability to retain and rehire specialized biol-ogists, mean that it is impossible to monitor allpark resources all of the time. Due to fluctua-tions in NPS funding, many monitoringprograms designed for parks in the past havebeen discontinued or were never implemented.Thus there is a great need to develop moni-toring that can be maintained when fundingdeclines, yet adequate to detect importantecological changes.

Coronado National Memorial is a NationalPark Service (NPS) interpretive site of approx-imately 1900 ha that was established because itoverlooks the probable route of FranciscoVasquez de Coronado, an early Spanish explorer

of North America. Located on the Mexico-United States boundary at the junction of severalmajor biogeographic provinces, the Memorialsupports a rich flora and fauna, but has also beengreatly impacted in recent years by human trafficalong the international border, as well as wild-fire, cattle grazing, and development outsidepark boundaries. Monitoring to provide infor-mation for management is considered essentialat the Memorial, as in many small parks, but isone of many management priorities.

The goal of our study was to conduct a thor-ough and repeatable inventory of terrestrialmammals and to make recommendations forlong-term monitoring that would be relevantfor small national park units. During an inten-sive, one-year inventory phase (1996–1997),we used multiple techniques to confirm asmany of the memorial's terrestrial mammalspecies as possible. To develop data to informmonitoring decisions, we then continued tomonitor mammals for seven additional years(1997–2003) using infrared-triggered camerasand, in particular, annual small mammal trap-ping on two large permanent plots.

STUDY SITECoronado National Memorial is located at thesouth end of the Huachuca Mountains, on theMexico-United States border (Figure 1). Thememorial contains 1900 ha in the upper water-shed of Montezuma Canyon and ranges inelevation from approximately 2,386 m to 1,471m. The steep northern and western portions ofthe memorial are predominantly oak woodland(Ruffner and Johnson 1991), while the lesssteep eastern portion is primarily semi-desertgrassland. Little surface water is present. Thecomplex geology includes metamorphosedsedimentary deposits, volcanic deposits, andmore recent sedimentary deposits (Doe 1986),including limestone outcrops that form under-ground chambers. A small portion of thememorial was grazed by cattle during ourstudy. The Huachuca Mountains are one of anumber of "sky island" ranges in southeasternArizona, southwestern New Mexico, andMexico that include biota from several biogeo-

graphic provinces, including the Sierra MadreOccidental to the south, the Rocky Mountainsto the northeast, the Sonoran Desert to thewest, and the Chihuahuan Desert to the east(Van Devender and Reina 2005).

METHODSDetecting presence and absence of terrestrialmammals can be difficult due to their diverselifestyles, including nocturnal and undergroundhabits. In addition, small mammals may bevery specific in their microhabitat require-ments, while larger species, especiallycarnivores, may occur naturally at very lowpopulation densities. To detect as many speciesthat occur in the memorial as possible, we useda wide variety of techniques as outlined below.

Small mammal trappingWe trapped small mammals using extra largeSherman and Tomahawk brand live traps forsmall mammals using standard methods (fordetails see Swann et al. 2007). During the1996–1997 inventory, we established 65 plotsof 25 traps each in areas that represented thegeographic, topographic, and vegetative diver-sity of the memorial (Figure 2), includingburned and unburned oak woodland areas, wetseeps, cattle tanks, riparian corridors, high andlow elevation grasslands, grazed and ungrazedsemi-desert grassland areas, and areas alteredby human activity. At each grid we placedtraps 10 m apart in a square or rectangularpattern and trapped for one to four nights,marking mammals captured on each night.

To evaluate the sample size and cost ofannual monitoring of small mammals weestablished two grids of 100 traps each (Figure2). One grid was randomly located in semi-desert grassland below 1524 m; this grid wasdominated by non-native grasses, particularlyLehmann’s lovegrass (Eragrostis lehman-niana). The second was randomly located ona south-facing slope in oak savanna above1921 m and was dominated by native grassessuch as bullgrass (Muhlenbergia emersleyi)and small trees including Emory oaks(Quercus emoryi). Individual animals were

226 Swann, Bucci, Kuenzi, Alberti, and Schwalbe

Swann, Bucci, Kuenzi, Alberti, and Schwalbe 227

Figure 1. Location map of Coronado National Memorial, Cochise County, Arizona. FromSchmidt et al. 2007.

uniquely marked using permanent color pensin order to develop a capture history for eachand each grid was annually trapped for 3–6nights during late October – early Decemberfrom 1997 through 2003. Abundance on eachgrid was estimated using the ProgramCAPTURE for closed populations (Otis et al.1978). CAPTURE chooses a model based onthe data given; we generally chose this modelto estimate abundance unless the null modelwas chosen and a model based on behavior,heterogeneity or time was available with onlya slightly lower rating.

Infrared-triggered photography Infrared-triggered photographs of large andmedium-sized mammals were obtained using

the model 1500 Trailmaster camera system(Goodson and Associates, Inc., Lenaxa, KS),where a single infrared beam is emitted by atransmitter and detected by a receiver; a photo-graph is taken when this beam is broken by ananimal. We placed camera units in vegetatedareas that were protected from visitors andillegal border crossers, and represented thegeographic diversity of the memorial (Figure 3).Cameras were set for intervals of two weeks ormore at a natural water source, or baited withsardines, cat or other carnivore lure, a visuallure, or some combination of these. We recordedall changes of film and bait used, and map coor-dinates were obtained using GPS. Animals ineach photograph were identified to species ifpossible, and times and dates recorded.


Figure 2. Map of Coronado National Memorial, showing locations of mammal trappinggrids used during this study. Long-term monitoring grids (trapped 1997-2003) were 2a6(Joe’s Canyon grid, oak woodland, southwest section) and 9a18 (Grassland grid, semi-desert grassland, southeastern section). All others were inventory grids.

As with small mammal trapping, weconducted an extensive inventory at 25 loca-tions for one year (1996–1997), then continuedphotography at a selected smaller number ofsites through 2003.

ObservationsTo supplement records of mammal distributionand relative abundance, we recorded allmammals observed during field work on thisand a related study of herpetofauna (Swannand Schwalbe 2007) during 1996–1997. Parkstaff continued to look for and document newspecies during 1998–2003. We looked formammals while driving on a road transect,conducting trapping and habitat analysis, andunder boards and other materials. We recordeddate, time, species, and location associatedwith each animal observed.

Species identification and permitsSpecimens or photographs of difficult-to-identify mammals were brought to theUniversity of Arizona Mammal Collection andconfirmed by Yar Petryszyn, Assistant Curator.Research was conducted under permits to CRSfrom Coronado National Memorial and theArizona Game and Fish Department andfollowed University of Arizona animalhandling protocols (Institutional Animal Careand Use Committee, Control #94-067).

RESULTSWe confirmed 45 terrestrial mammals at Coro-nado National Memorial during 1996-2003,including three non-native species — feral dog(Canis familiaris), feral cat (Felis catus) andhouse mouse (Mus musculus). Species wereconfirmed by voucher photos, specimens, or


Figure 3. Locations of infrared-triggered (Trailmaster) cameras used during this study.

observations where a specimen previouslyexisted, or (in three cases) by reliable obser-vations only. We believe that two other speciesoccur at the memorial, but did not documentthem during 1996–2003: mule deer(Odocoileus hemionus), which were reliablyobserved by park staff prior to our study andare probably occasional visitors; and easterncottontails (Sylvilagus floridanus), which webelieve we observed and photographed, but forwhich a voucher specimen is required forunambiguous identification.

We confirmed a number of species not previ-ously confirmed for the memorial and, exceptfor one species of pocket gopher — Southernpocket gopher (Thomomys umbrinus, for whichwe observed sign in appropriate habitat — weobserved all species confirmed in a previousinventory (Petryszyn and Cockrum 1979).

Small mammal trapping. We trapped small mammals during the inven-tory phase, September 1996 – December 1997for a total of 4628 trap nights. We made 540captures for a mean trap success of 11.7%. Wecaptured 416 individuals of 17 species (Table1). A few species were captured throughout thememorial, but many were confined to eitherhigh or low elevations; a few were captured inspecial habitats such as wet seeps or opendesert habitat.

On the 100-trap oak savanna grid nearCoronado Peak, we captured 412 individualsof 11 species during 3,700 trap nights during1997–2003. On the lower grassland grid justnorth of Montezuma Canyon near the eastboundary, 478 individuals of 14 species werecaptured during 3,800 trap nights. Over sevenyears, on both grids combined, we trapped 890


Table 1. Species of small mammals and numbers captured (captures – minus recaptures) at CoronadoNational Memorial during the intensive inventory (1996-1997) and long-term monitoring (1997-2003). *Note that the desert shrew was not trapped during the inventory phase, but one individualwas hand-captured and photo-vouchered.

1996-1997Species Inventory 1997 1998 1999 2000 2001 2002 2003

Desert shrew (Notiosorex crawfordii) 0* 0 0 1 0 0 0 0

Hispid pocket mouse (Chaetodipus hispidus) 23 1 0 4 1 5 2 1

Rock pocket mouse (C. intermedius) 2 0 4 3 0 8 5 8

Silky pocket mouse(Perognathus flavus) 0 0 0 0 1 2 1 0

Sonoran Desert pocket mouse (C. penicillatus) 51 3 0 0 0 0 4 1

Unknown pocket mouse 0 0 0 0 0 0 0 1Ord’s kangaroo rat (Dipodomys ordii) 2 0 0 0 0 0 0 9

Banner-tailed kangaroo rat(D. spectabalis) 1 0 0 0 0 0 0 0

Merriam’s kangaroo rat (D. merriami) 5 0 0 0 0 0 0 0

House mouse (Mus musculus) 0 0 0 0 0 1 0 0

individuals of 19 species (Table 1). Twospecies trapped during the intensive inventory— Merriam’s kangaroo rat (Dipodomysmerriami) and banner-tailed kangaroo rat (D.spectabalis) — were not trapped on the moni-toring grids during 1997–2003, and fourspecies trapped during the monitoring effort— spotted ground squirrel (Spermophilusspilosoma), silky pocket mouse (Perognatusflavus), house mouse, and desert shrew(Notiosorex crawfordi) — were not trappedduring the inventory in 1996–1997. In addi-tion, we captured a number of species in areaswhere they had not been previously trapped,including several low elevation “grassland”species, such as Northern pygmy mice(Baiomys taylori), on the high elevation oaksavanna grid.

Abundance and species richness on the twogrids varied greatly among years (Table 2,Figure 4). For example, on the grassland grid

estimated abundance ranged from a low of 46in 1997 to a high of 192 in 2000. Abundance ofcotton rats (Sigmodon spp.) was higher during2000 and 2001 than during 1997–1999 or2002–2003 (Figure 5). Increases in rodent abun-dance, and especially in abundance of Arizonacotton rats (S. arizonae) the dominant specieson the grassland grid) appeared to be correlatedwith higher than normal summer rains in 1999and very high rainfall in 2000 (Figure 6).

Infrared-triggered photographyDuring the inventory of September 1996 –December 1997, three Trailmaster cameraswere used for approximately 1142 nights at 25locations (Figure 3). During this period, 379photographs of 17 mammal species wereobtained (Figure 7). Cameras were often notoperational because a roll of film had beencompletely exposed or an equipment malfunc-tion had occurred; we estimate the cameras


Northern pygmy mouse (Baiomys taylori) 34 7 5 15 26 30 18 23

White-throated woodrat (Neotoma albigula) 50 8 9 13 10 12 8 20

Southern grasshopper mouse (Onychomys torridus) 15 4 10 10 22 14 6 9

Deer mouse (Peromyscus maniculatus) 5 2 0 0 3 3 3 13

White-footed mouse (P. leucopus) 26 9 4 26 6 3 2 3

Brush mouse (P. boylii) 80 3 7 3 15 23 4 28Unknown white-footed

mouse 3 0 0 1 0 0 0 1Fulvous harvest mouse (Reithrodontomys fluvescens) 16 14 4 12 13 15 10 14

Western harvest mouse (R. megalotis) 8 7 8 1 4 10 8 11

Unknown harvest mouse 0 0 1 0 2 0 0 0Arizona cotton rat (Sigmodon arizonae) 27 4 3 10 39 69 11 8

Table 1. continued

1996-1997Species Inventory 1997 1998 1999 2000 2001 2002 2003


Table 2. Abundance estimates for small mammals on two long-term monitioring plots at CoronadoNational Memorial, 1997-2003. JC = Joe’s Canyon grid (oak woodlands). GR = semi-desert grass-land. Estimates were derived using mark-recapture based on CAPTURE; we used model selected byCAPTURE unless null model was selected and another model scored within 10 percentage points, inwhich case we chose the other model. M(o) is null model, M(h) is heterogeneity model, M(b) isbehavior model, M(t) is time model, and M(bh) is combined behavior and heterogeneity. Plot Year All Species Cotton rats (Sigmodon spp.). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

#Cap Model Est SE CI #Cap Model Est SE CIJC 1997 56 M(o) 66 4.62 61-79 18 M(o) 21 2.59 19-31

1998 38 M(h) 67 3.91 52-97 11 M(o) 11 0.85 11-141999 60 M(b) 89 21.14 68-164 5 M(th) 5 0.42 5-72000 53 M(h) 102 11.54 82-129 5 M(o) 5 0.39 5-52001 88 M(bh) 114 15.02 98-162 17 M(h) 18 1.31 17-242002 30 M(t) 39 4.78 34-53 8 M(o) 11 3.19 9-242003 87 M(h) 137 15.08 116-176 15 M(h) 19 3.63 16-33

GR 1997 27 M(h) 46 7.48 36-66 7 M(h) 15 4.5 10-291998 35 M(h) 68 11.8 52-100 7 M(o) 8 1.73 9-271999 45 M(b) 65 17.48 50-132 11 M(th) 14 3.60 12-292000 100 M(h) 192 18.45 163-235 44 M(b) 63 16.00 49-1232001 119 M(bh) 174 18.17 149-222 64 M(b) 102 28.72 75-2052002 74 M(bh) 82 8.64 75-117 24 M(bh) 34 7.75 27-622003 79 M(bh) 118 15.39 98-161 9 M(h) 24 6.12 16-44

were operational for a total of approximately640 nights.

During the monitoring period of January1998 – January 2004 (excluding 2001 and2002, when we did not operate cameras), Trail-master cameras were used at 9 locations(Figure 3), and we estimate that cameras wereoperational for approximately 972 nights.During this period, 1028 photographs of 18native species (plus feral dogs) were obtained(Figure 7). One species photographed in1996–1997, the common raccoon (Procyonlotor), was not photographed during1998–2004; two species not photographedduring 1996–1997 — Virginia opossum(Didelphis virginiana) and Arizona graysquirrel (Sciurus arizonensis) — werephotographed during 1998–2004.

ObservationsThree species — desert shrew, feral cat, andBotta’s pocket gopher (Thomomys bottae) —

were documented during the inventory byhand-capture or collection of a dead individual.Three species — Arizona gray squirrel,southern pocket gopher, and black-tailedjackrabbit (Lepus californicus) — were nottrapped or photographed but were observed(or, in the case of the pocket gopher, their signin appropriate habitat was observed) byresearchers. Two additional species — Amer-ican badger (Taxidea taxus) and mule deer —had been recently observed by expert staff butwere not observed, captured, or photographedduring the period.

During the monitoring phases, no newspecies were observed. Of the five speciesobserved but not otherwise documented duringor just prior to the 1996–1997 inventory, onespecies (Arizona gray squirrel) was confirmedby infrared-triggered photography and one(American badger) was confirmed by roadkillin 2004. Three species (southern pocketgopher, black-tailed jackrabbit and mule deer)


Figure 4. Estimated abundance and 95% confidence interval of small mammals on Joe’sCanyon (oak woodland) and Grassland (semi-desert grassland) monitoring grids atCoronado National Memorial, 1997–2003.

known to occur at the memorial remain undoc-umented.

DISCUSSIONSpecies diversity andinventory completeness

Coronado National Memorial has a greatdiversity of mammals compared to other parksin southern Arizona, especially considering itssmall size. For example, the Rincon MountainDistrict of Saguaro National Park is more than14 times larger than Coronado, but has thesame number (45) of confirmed species(Swann and Powell 2006). In addition to theterrestrial mammals we studied, Coronado alsohas many bats (6 confirmed and 17 possiblespecies; Swann et al. 2007). This diversity isprobably due to a number of factors, includingthe memorial's location in the HuachucaMountains at the northern end of the Madreanmountains; connectivity with other natural

areas, including Coronado National Forest, theSan Pedro River, and undeveloped areas inMexico; and the presence of areas of ungrazedsemi-desert grasslands with high grass cover.In addition, small parks such as Coronado mayhave higher diversity per unit area becausetheir mammalian fauna include most of thehabitat generalists — e.g., species like desertcottontail and mountain lion (Puma concolor)— that occur in larger parks. Mammal diver-sity is a significant natural resource value atthe memorial that is worthy of preservation.

We believe that we detected most terrestrialmammals at the memorial, given our long-termeffort using multiple methods, although wefailed to confirm two other species (mule deerand eastern cottontail) that probably also occur.We did not detect seventeen native species thatoccur or have occurred historically within theHuachuca Mountains (Hoffmeister 1986) andnearby valleys. These include species which


Figure 5. Estimated abundance and 95% confidence interval of cotton rats (Sigmodonspp.) on Joe’s Canyon (oak woodland) and Grassland (semi-desert grassland) monitoringgrids at Coronado National Memorial, 1997–2003. Most cotton rats on the Joe’s Canyongrid were yellow-nosed cotton rats, and most on Grassland grid were Arizona cotton rats.

have been found very close to the memorialand are likely to occur there (e.g., commonporcupine [Erithizon dorsatum]) from time totime; species which occur nearby but areunlikely at our site due to lack of suitablehabitat, e.g., round-tailed ground squirrel(Spermophilus tereticaudus); species which arecertainly not resident but range widely andmay pass through the memorial from time totime, e.g., jaguar (Panthera onca); and a fewspecies which are now certainly extirpated inthe area, e.g., gray wolf (Canis lupus). Swannet al. (2000) provides detailed species accountsfor all known and potential species, includingsummaries of historic and museum records.

Comparison with past studiesOur study confirmed a number of rodentspecies not confirmed by an earlier inventory(Petryszyn and Cockrum 1979). Although their

study was shorter and involved fewer trapnights than ours, Petrysyn and Cockrum(1979) trapped in similar areas and did notdetect several grassland species, such aspygmy mice, three species of kangaroo rats,and three species of cotton rats. Indeed, theyellow-nosed cotton rat (Sigmodon ochrog-nathus), one of the most abundant rodentspecies at both higher and lower elevations,was never trapped by Petryszyn and Cockrum(1979).

The most plausible explanation for thesedifferences, in addition to reduced effort, is thatgrasses in the low elevation grasslands southof East Montezuma Canyon Road, which werevery robust during our study, were sparse in1977–78 due to heavy cattle grazing and lowrainfall. Although grazing occurred in thememorial during our study, it had been absentfrom this area for at least 8 years when our


Figure 6. Annual and summer rainfall totals, Coronado National Memorial, 1996-2003.MM = Mean monsoon, MA = Mean annual. “Monsoon” is rainfall total during June, July,August, and September. Source: National Weather Service.

inventory began (E. Lopez, Coronado NMsuperintendent, pers. comm.). Our long-termsemi-desert grassland grid was dominated bythe non-native Lehmann’s lovegrass, butnonetheless had a very high abundance andrichness of native rodents.

Another major change that has occurred atCoronado during recent decades has been theloss of trees (particularly conifers) and theresultant growth of high elevation grasses sincethe severe Peak Fire of June 1988 (Ruffner andJohnson 1991). Our oak woodland monitoringgrid was located in an area of thick nativegrasses that had completely burned during thePeak Fire. This vegetative change may also berelated to the cessation of grazing at higherelevations. The resulting increase in grass seedcrop has clearly been favorable to smallrodents. It will be interesting to track changesin the species diversity of these oak savannas

if they become revegetated with oaks andpiñon pine.

Monitoring implicationsAlthough monitoring is essential if the NPS isto fulfill the important mission of preservingbiodiversity on its lands, monitoring of verte-brates can be time-consuming and expensive.Many mammal species are difficult to observeand count, and their populations often fluctuategreatly due to natural causes. Monitoring isparticularly difficult in small park areas wherehuman and financial resources may be morevariable and limited than in larger parks(Swann 1999). Coronado does not have anythreatened and endangered terrestrialmammals, so our study examined the costs andbenefits of two other methods proposed formonitoring terrestrial mammals, annual trap-ping of small mammals and repeated


Figure 7. Number of photographs of 19 species of mammals detected by infrared-triggered (Trailmaster) cameras during inventory phase (1996-1997) and long-termmonitoring (1998-2003).

inventories of the terrestrial mammal commu-nity as a whole.

Annual trapping of small mammals is oftenrecommended for monitoring programs forvarious reasons, including documentation ofchanges in prey base and overall environmentalconditions. In our study, annual monitoringclearly provided many economic and non-economic benefits, including a demonstrationof the dynamic nature of mammal communi-ties at the memorial. As seen in other studies insemi-desert grasslands (e.g., Heske and Brown1990), abundance of grassland rodents variedamong years and appeared to respond tochanges in precipitation. We observed a largeincrease in Arizona cotton rats, normally asso-ciated with riparian grasslands, during the verywet period during 1999–2000. The 1997–2003trapping effort also gradually increased ourdetection of very rare species in the memorialand added new species.

Non-economic benefits of our effortincluded opportunities to work with volunteers

in the field and a higher profile for research inthe park. Like many parks, Coronado relies onvolunteers for many daily tasks, and manyvolunteers in visitor contact positions increasedtheir knowledge of natural resources byassisting with trapping efforts.

Our economic costs during the period wereminimal, but this was because of continuity inpersonnel and a high reliance on volunteers.One of us (BA) was employed by the memo-rial during most of this study and was able toassist with mammal trapping, recruit parkvolunteers, secure park funds to pay two of us(DES and MB) for short periods to processmammals, coordinate infrared-triggeredcamera efforts, and coordinate securityarrangements directly with law enforcementstaff. However, staff changes at the memorialwere one factor in our decision to discontinuemonitoring in 2003, as by that time there werehigher priority issues at the memorial thanmonitoring non-threatened or endangeredmammals.


Border security, not an issue when westarted the inventory in 1996, became the over-riding issue at the memorial during 1998 (andremains so as of this writing). During1998–2003 we were unable to set or checktraps before or after sunset, which greatlyreduced the number of traps we could setwithin our budget, which in turn reduced powerto detect significant trends. In addition, our datasuggest that small mammals are probably nota good indicator for impacts associated with theinternational border; mammals were abundantduring periods where border crossings andenforcement activities were high.

Efficiency of intensive inventories indescribing the mammal community

Another goal of this study was to assess howcompletely the mammal community could bedescribed with an intensive, relatively short-term effort (approximately one year). For smallparks such as Coronado, monitoring mammaldiversity through periodic (repeat) inventorieshas been proposed as a cost-effective approachto long-term monitoring (Swann 1999). Theadvantage of repeat inventories is that they donot require permanent staff (indeed, our inten-sive inventory was conducted through acooperative agreement with the University ofArizona). Repeat inventories may provideinsight into important changes that are occur-ring, such as the loss of native species. But dueto the difficulty of detecting rare species, thereis a danger that short inventories may fail toadequately describe the community in a waythat is adequate for long-term monitoring.

During the inventory phase of our study weconfirmed or reliably observed 42 species;during the monitoring phase we trapped,photographed, or observed all but three ofthese species; collected or photographed twoof the five observed species; and confirmed anadditional three species. Thus, if mule deer andeastern cottontail are excluded, and reliableobservations included, the 1996–1997 inven-tory documented 42 of the 45 species (93.3%)known to occur at Coronado during the entire1996–2003 period. This high percentage

suggests that for monitoring long-term changesin diversity of mammals, at least, inventoriesof short duration can be reasonably completeif they are of sufficiently high intensity and aredesigned to be repeatable.

However, the species we detected only inthe monitoring phase are obviously rare at thememorial and therefore may be of greatermanagement interest. Long-term studies ofmany taxonomic groups (see summary inRozensweig 1995) indicate that known speciesrichness in an area increases over time, thoughat a declining rate. Species detected only aftersignificant effort probably exist in small popu-lations or expand their ranges to include thememorial during years with favorable envi-ronmental conditions. For this reason,estimation of species richness, or at least useof species accumulation curves, should be partof the inventory process.

Monitoring recommendationsNational Park Service programs to monitormammals often emphasize estimating popula-tion size of threatened or endangered speciessuch as Florida panthers, or common speciessuch as deer mice, but are not concerned withmonitoring other species. Threatened or endan-gered species often must be studied for legalreasons, and long-term studies of singlespecies (Gibbons 1990; Pierson and Turner1998) have given tremendous insight into theirnatural history and endangerment factors.Nevertheless, understanding changes in abun-dance of selected species is no substitute fordata on trends in the presence and distributionof all species in the park. Loss of species fromnational parks and other natural areas due tohuman impacts is a major concern (Newmark1995) yet one that few monitoring programsare designed to measure. In addition totracking potential changes in abundance ofcommon species, monitoring must provideinformation on species that may be in dangerof extirpation because of their rarity or loss ofspecialized habitat.

Our study suggests that intensive invento-ries may be a useful and more sustainable tool

for monitoring mammals in small nationalparks and natural areas than annual moni-toring. Inventories designed to be repeatedevery 10-20 years can allow assessment ofoverall changes in species richness as well asin the individual status of different species. Tobe effective, repeat inventories should take asystematic approach to sampling all species inthe area of interest and include trapping anduse of infrared-triggered cameras or similartechnology in all vegetation communities.

Like all monitoring approaches, repeatedinventories may be helpful in detecting manytypes of ecological changes, both positive andnegative. Comparison of our results withCockrum and Petryzyn (1979) suggests thatnative species diversity may be increasing atCoronado and that the memorial may be arefugium for many mammals, including notonly grassland rodents, but also hunted animalssuch as deer and predators such as mountainlion. Because of this, the memorial may alsoplay a future role in the return of species whichare presently extirpated from the area. Black-tailed prairie dogs (Cyonomys ludovicianus),which occurred near the memorial earlier inthis century, occur in Mexico less than 4.8 km(three miles) south of the Huachucas (Ecolog-ical Center of Sonora 1994) and couldnaturally recolonize the memorial in the future.Coronado, as part of the Greater HuachucaMountains–San Pedro River area, similarlyprovides habitat for jaguars, ocelots (Leop-ardus pardalis), and gray wolves movingnorthward, should these species increase innumber. On the other hand, future develop-ment of a border fence could restrictmovements of animals. The memorial by itselfis not large enough to sustain populations oflarge species such as mountain lions, blackbears (Ursus americanus), and coatis (Nasuanarica), so habitat loss in the San Pedro Valleycould lead to local extirpation of these species.

While monitoring of mammal diversitythrough repeated inventories should be apriority, it goes without saying that otherresearch and monitoring should occur when-ever possible. In addition to small mammals,

efforts should be made to study selectedmammal species intensively, measuring abun-dance as well as parameters such asreproduction and survival. Long-term studiesof species that are of management interest atCoronado, such as ringtails (Bassariscusastutus), skunks, or pygmy mice, areextremely rare, and would benefit not onlyNPS but other land management and wildlifeconservation agencies.

ACKNOWLEDGMENTSWe thank Coronado National Memorial,particularly superintendents Ed Lopez, JimBellemy, and Kym Hall, and many other staffincluding Henry Ruiz, David Chavez, CyndyDonaldson, Michael Hardin, Fred Moosman,Stock Sticha, William Smith, KennethThompson and Nancy Willcox, as well asvolunteers Paul Funk, Alicia Gutzman,Heather Guest, Michael Guest, Mary Megill,Phi Pham, Vicki Powers, Bonnie Ranier, DavidRanier, Mary Helen Surovik, Will Sergent, andPete Van Cleve. We are very grateful to SandyWolf and Sherry Mann for assistance in thefield, and for Zoe Hackl, Melanie Culver, andLisa Haynes for sharing data from their studyof native cats. We thank Yar Petryzyn formammal identifications, and Matt Daniels,Kristen Beupre, and Kara O’Brien for datamanagement and mapping. Thanks to SueBenson, Joan Ford, Brenda Carbajal, MattGoode, Mary Greene, Kathy Hiett, MikeKunzmann, Dale Linder, Brian F. Powell,Cecilia A. Schmidt, Sue Benson, and CarolWakely for additional support. We greatlyappreciate the long-term support of WilliamHalvorson for this and many other long-termmonitoring projects in southern Arizona parks.Funding for this project was provided by Coro-nado National Memorial and the U.S.Geological Survey, Sonoran Desert ResearchStation.

LITERATURE CITEDDoe, M. F. 1986. Geology in the vicinity of Mon-

tezuma's Cave, southern Huachuca Mountain,Cochise County, Arizona. Unpublished report



submitted to NPS-SOAR, Phoenix, AZ. Ecological Center of Sonora. 1994. Management

plan for the protected area of wild and aquaticflora and fauna in the Sierras of la Mariquita-laElenita-Rio San Pedro, in the municipalities ofCananea, Naco and Santa Cruz, Sonora, Mexico.Unpublished report by the ECS and System ofProtected Natural Areas in the State of Sonora,Hermosillo, Sonora.

Brown, J. H., and E. J. Heske. 1990. Temporalchanges in a Chihuahuan Desert rodent commu-nity. Oikos 59: 290–302.

Drayton, B., and R.B. Primack. 1996. Plant specieslost in an isolated conservation area in m e t r o -politan Boston from 1894 to 1993. ConservationBiology 10: 30–39.

Drost, C. A., and G. M. Fellers. 1996. Collapse ofa regional frog fauna in the Yosemite area of theCalifornia Sierra Nevada, USA. Conservation Bi-ology 10: 414–425.

Gibbons, J. W. 1990. Turtle studies at SREL: A re-search perspective. In The natural history andecology of the slider turtle, edited by J.W. Gib-bons, pp. 19–44. Smithsonian Institution Press,Washington D.C.

Hoffmeister, D. F. 1986. Mammals of Arizona. Uni-versity of Arizona Press. Tucson, Arizona.

National Park Service. 2006. National Park Servicemanagement policies. Washington, D.C.

National Research Council. 1992. Science and thenational parks.National Academy Press, Wash-ington, D.C.

Newmark, W. D. 1995. Extinction of mammal pop-ulations in western North American nationalparks. Conservation Biology 9: 512–526.

Otis, D. L., K. P. Burnham, G. C. White, and D. R.Anderson. 1978. Statistical inference from cap-ture data on closed animal populations. WildlifeMonographs 62: 1–135.

Petryszyn, Y., and E. L. Cockrum. 1979. A surveyof the mammals of Coronado National Memorial,Arizona. In Survey of the vertebrate fauna ofCoronado National Memorial, edited by E. L.Cockrum, C. H. Lowe, S. M. Russell, D. Dan-forth, T. B. Johnson, and Y. Petryszyn, pp.96–124. Technical Report 5, NPS CooperativePark Studies Unit, University of Arizona.

Pierson, E. A. and R. M. Turner. 1998. An 85-yearstudy of saguaro (Carnegiea gigantea) demogra-phy. Ecology 79: 2676–2693.

Robbins, W. E., A. Ackerman, M. Bates, S. A. Cain,F. F. Darling, J. M. Fogg, Jr., T. Gill, J. M. Gilli-son, E. R. Hall, and C. L. Hubbs. 1963. A reportby the advisory committee to the National ParkService on research. National Academy ScienceResearch Council.

Rosenzweig, M. L. 1995. Species diversity in spaceand time. Cambridge University Press, Cam-bridge, England.

Ruffner, G. A. and R. A. Johnson. 1991. Plant ecol-

ogy and vegetation mapping at Coronado Na-tional Memorial, Cochise County, Arizona. Tech-nical Report 41, NPS Cooperative Park StudiesUnit, University of Arizona.

Schmidt, C. A., B. F. Powell, D. E. Swann, and W.L. Halvorson, eds. 2007. Vascular plant and ver-tebrate inventory of Coronado National Memo-rial. Open File Report 2007-1393, U. S.Geological Survey, Southwest Biological ScienceCenter, Sonoran Desert Research Station, Uni-versity of Arizona, Tucson, AZ.

Swann, D. E. 1999. Evaluating approaches formonitoring terrestrial vertebrates in U.S. nationalparks: An example from Tonto National Monu-ment, Arizona. Master’s thesis, School of Re-newable Natural Resources, University ofArizona, Tucson, AZ.

Swann, D. E., C. R. Schwalbe, A. J. Kuenzi, and B.N. Alberti. 2000. Inventory of the terrestrial mam-mals of Coronado National Memorial, CochiseCounty, Arizona. Final report to National ParkService, Southern Arizona Parks, Phoenix, AZ.

Swann D. E, and B. F. Powell. 2006. Mammal in-ventory of Saguaro National Park, Rincon Moun-tain District. In Vascular plant and vertebrateinventory of Saguaro National Park, RinconMountain District, edited by B. F. Powell, W. L.Halvorson, and C. A. Schmidt, pp. 69–97. OpenFile Report 2006-1075, U. S. Geological Survey,Southwest Biological Science Center, SonoranDesert Research Station, University of Arizona,Tucson, AZ.

Swann, D. E., C. R. Schwalbe, A. J. Kuenzi, M.Bucci, and B. N. Alberti. 2007. Mammal inven-tory. In Vascular plant and vertebrate inventory ofCoronado National Memorial, edited by C. A.Schmidt, B. F. Powell, D. E. Swann, and W. L.Halvorson, pp. 47–54. Open File Report 2007-1393, U. S. Geological Survey, Southwest Bio-logical Science Center, Sonoran Desert ResearchStation, University of Arizona, Tucson, AZ.

Swann, D. E. and C. R. Schwalbe. 2007. Reptilesand amphibians inventory. In Vascular plant andvertebrate inventory of Coronado National Me-morial, edited by C. A. Schmidt, B. F. Powell, D.E. Swann, and W. L. Halvorson, pp. 11–26. OpenFile Report 2007-1393, U. S. Geological Survey,Southwest Biological Science Center, SonoranDesert Research Station, University of Arizona,Tucson, AZ.

Van Devender, T. R., and A. L. Reina G. 2005.Floristic comparison of an Arizona sky island andthe Sierra Madre Occidental in eastern Sonora:The Huachuca Mountains and the Yecora area. InConnecting mountain islands and desert seas:Biodiversity and management of the Madrean ar-chipelago II, compiled by G. J. Gottfried, B. S.Gebow, L. G. Eskew, and C. B. Edminster, pp.154–157. Proceedings RMRS-P-36. Fort Collins,CO. U. S. Department of Agriculture, Forest Serv-ice, Rocky Mountain Research Station.

One of the most studied assemblies (Fauth etal. 1996) are the small mammals in the NorthAmerican deserts. Through these studies,mechanisms have begun to be proposed tounderstand the structure of a community; thisis in part because of the relative simplicity ofthe microenvironments present in arid zones(Shepherd and Kelt 1999; Kelt et al. 1999;Giannoni et al. 2001). In arid environments theassociation of small mammals with vegetationis important because they modify it directly(Kelt et al. 1999); one of the main functions ofthe community of small mammals is togenerate changes in the distribution and struc-ture of the vegetation, mainly of herbs andshrubs because these are their main sources offood (Curtin et al. 1999; Price and Correll2001; Sassi et al. 2004).

Small mammals are one of most abundantgroups in North American deserts and theircoexistence has been attributed to processesthat imply differences in the use of microhab-itat and variation in food availability (Kelt etal. 1999). In relation with microhabitat use, ithas been established that rodents with bipedallocomotion (e. g. Dipodomys) are species thatprefer to forage in areas with low cover andsparse vegetation (Thompson 1982; Hallet1982). Regarding the species with quadrupedlocomotion (e.g. Chaetodipus, Perognathusand Peromyscus), they concentrate theirforaging in areas with dense vegetation andcover which protect them from predators(Hallet 1982; Kotler 1984).

Rodents in arid zones act as seed dispersersand they feed mainly on plants (Brown et al.1979; Price and Correll 2001; Folgarait andSala 2002). It is important therefore, to eval-uate the degree of association between rodentsand plants on the community level. Thelegume ironwood (Olneya tesota) has beenbroadly studied in both the United States andthe state of Sonora, México. It is the mostabundant species of the coast of the Gulf ofCalifornia and the plains of Sonora (Nabhanand Behan 2000). The ironwood, an extremelyhard wood (hence the name), has remarkablyslow growth (it has been estimated that sometrees are 800 years old) and can grow toheights exceeding 15 meters. (Tewksbury andPetrovich 1994; Nabhan and Behan 2000).

Ironwood is considered one of the mainSonoran Desert nurse species, because underits canopy microenvironments are createdwhich increase the germination and establish-ment of different seedlings (Búrquez andQuintana 1994). Nurse trees protect variousspecies of plants like saguaros and othercolumnar cacti from sun radiation, hightemperatures, and water stress (Búrquez andQuintana 1994; Tewksbury and Petrovich1994).

The majority of ecological studies have beenfocused on the association of ironwood with theseedlings that use it as a nurse species, leavingaside the relationship that it has with the fauna.Of the few existing studies of the relationshipwith fauna, it has been determined that the pres-

THE SMALL MAMMAL COMMUNITY ASSOCIATED WITH IRONWOOD (OLNEYA TESOTA)

Helí Coronel-Arellano and Carlos A. López-González

242 Coronel-Arellano and López-González

ence of ironwood, palo verde and acaciasincreases the richness of bird species up to63.7%, and in absence of such vegetal associ-ation, owls and woodpeckers disappear fromthe community (Nabhan and Behan 2000).

Tewksbury and Petrovich (1994) carriedout counts of bird nesting sites in the canopyof the ironwood and they compared it withother types of bushy vegetation (Acacia,Bursera, Colubrina, Condaliopsis, Jatropha,Prosopis and Cercidium, Zizyphus andJaquinia). The authors studied the numbers ofanimals and their activity in sites with iron-wood (e.g., bedding places and directobservations). These authors found a largernumber of nesting sites in ironwood (27) thanin the rest of trees and bushes (19). In regardto the activity of other animals, they found arelationship between the canopy cover and theincrease of usage from mammals and reptiles.Observations show that the birds, mammalsand reptiles use with major frequency habitatassociated with ironwood trees (Tewksburyand Petrovich 1994).

There is no reliable inventory of the faunaassociated with ironwood, because the infor-mation available is the product of question-naires sent to different researchers. From thisquestionnaire, the probable fauna associatedwith O. tesota are 25 species of ants, 25 or-thopterans, 188 bees, 12 anura, 19 lizards, 24serpents, 57 birds, and 64 mammals (Nabhanand Behan 2000).

Regarding mammals, the species that mayhave a major use of ironwood sites are the onesbelonging to the order Rodentia; one of themain groups that feeds on and harvests seed inNorth American deserts. McAuliffe (1984)registered consumption of ironwood seedlingsby pocket mice (Perognathus sp.), packrats(Neotoma sp.) and ground squirrels (Sper-mophilus sp.). Nevertheless, the conditions thatfavor the frequency and quantity of seed depre-dation is unknown.

Due to the fact that ironwood in theSonoran Desert is one of the main nursespecies and is under anthropogenic pressure, itis necessary to evaluate the relationship that

this tree has with rodents. Therefore, the objec-tives of this work were: (1) to determine if thecommunity of small mammals selects siteswith ironwood, (2) to describe and comparethe richness, structure and composition of thecommunity of small mammals, and (3) todetermine if the abundance of rodents has arelationship with the cover and structure of thevegetation.

STUDY AREA The study area was located within the SonoranDesert in the state of Sonora, México. Weconducted sampling at four localities (Figure1). The first study area was San Judas Ranch,located at coordinates 29˚22'43" north latitudeand 111˚06'37" west longitude. The secondwas the Pozo Hondo Ranch, located atHermosillo municipality at 29°43'05" northlatitude and 111°25'11" west longitude and atan altitude of 638 meters above sea level. Thethird study area was Las Glorias Ranch, withlocation at 29°38'05" north latitude and111°29'12" west longitude.

These three study areas are located in the“plains of Sonora” of the Sonoran Desert(Turner and Brown 1994). The climate of theselocations is classified as “very dry semi-hotwith rains in summer.” The average annualtemperature is 21.8° C, the highest tempera-tures are in the months of July (31.6° C) andAugust (30.6° C) and the lowest temperaturesare in the months of December and January(13.5° C and 12.7° C respectively); the ave-rage annual precipitation is 278.4 mm, themajor quantity of rain (66.3 and 67.4 mm)occurring in the months of July and August(Instituto Nacional de Estadística, Geografía eInformática 2000).

The higher canopy of the vegetation isdominated by Olneya tesota, it is also charac-terized by Fouquieria macdougalii, Burseraconfusa, Cercidium praecox, C. microphyllumand Prosopis; the most common cactus speciesare Lophocereus schottii, Stenocereus thurberiand S. alamosensis. Regarding the shrubs,Encelia farinosa is the most abundant. Thereare also large size bushes such as Mimosa laxi-

Coronel-Arellano and López-González 243

flora, Guaiacum coulteri, Phaulothamnusspinescens and Randia thurberi (Turner andBrown 1994).

The last place that we studied is LobosRanch, located at 50 kilometers to the south-east of Puerto Libertad in the municipality ofPitiquito Sonora, at 29°34'36" north latitudeand 112°10'10" west longitude, elevation 198meters above sea level. This place is located inthe “central gulf coast” (Turner and Brown1994). This private property has 22,800hectares and is managed for hunting and forestexploitation. Forest management includes theextraction of rubber, charcoal, mesquite wood,and jojoba seeds. The climate of this locationis semi-hot with cool winter. The averageannual temperature is 20° C and the averageannual precipitation is 250 mm. This property

consists of 95% plains with small hills in thecenter (Information Systems and Automation2005). The main vegetation elements arebushes such as Jatropha, Euphorbia,Fouquieria splendens, Larrea tridentata andsmall trees like Cercidium microphyllum,Olneya tesota and Bursera; other bushes witha minor representation are Encelia farinosa,Solanum and Ambrosia, the representativecactus in the area is Lophocereus schottii(Turner and Brown 1994).

METHODSCapture of Organisms

At each study area, we selected 20 sites withironwood and 20 sites without ironwood.Within each site we captured small mammalsby placing Sherman traps of three different

Figure 1. Localities sampling in the Sonoran Desert.


sizes (small, medium and large) in a circularpattern, taking as center one ironwood or theabsence of it (Figure 2; Coronel et al. 2005).We baited the traps with oats mixed with peanutbutter and vanilla. The captured organismswere identified to species, sexed, weighed andreleased. The identification of the capturedorganisms was made using species keys presentin MacMahon (1994), Álvarez-Castañeda andPatton (2000) and Whitaker (2002).

The relative abundance of each species wascalculated dividing the number of individuals bythe sampling effort (Nichols and Conroy 1996).

Association with Ironwood To establish if there were any differences inselection by the small mammal community atsites with ironwood, we used a two-wayANOVA. Also, the species diversity wascompared in sites with and without ironwoodby ways of the equity index of Shannon-Weaver using a paired t-test (Zar 1999).

Community AnalysisWe described species richness at each site, bycounting the number of species of smallmammals at the four study areas. The Margalefindex (Dmg) was also calculated for each site.To evaluate the efficiency of the sampling, wecalculated a species accumulation curve, usingas effort the 511 locations where the traps were

placed. This analysis was elaborated with dataof two places (Pozo Hondo and Lobosranches). To construct the curve three non-parametric estimates of alpha diversity, wereused Chao2, Jackknife of first order and Boot-strap; the estimate Chao2 is recommendedwhen dealing with small sample size and it iscalculated as Chao2 = S + (L2 / 2 M) where, S= the total number of species, L = number ofspecies that occur only in one sample (uniquespecies) and M = is the number of species thatoccur in two samples.

The Jackknife estimate of first order isbased on the number of species that occur onlyin one sample (L); this estimate reduces thesubestimation of the real number of species inthe area of study and is calculated as follows:Jackknife 1 = S + L (m -1/m), where S = is thetotal number of species, L = number of speciesthat occur only in one sample (unique species),m = number of samples. The third estimateused was Bootstrap which is calculated asfollows: Bootstrap = S + Σ (1 – pj)n. This esti-mate is based on pj, the proportion of samplingunits that contain each species j, but it appearsthat this estimate is less precise than the priortwo (Magurran 2004). To calculate the alphadiversity values we used the algorithms of theprogram Estimates 7.5.0 (Colwell 2004).

To analyze the structure of the community,abundance rank curves were created to deter-

Figure 2. Spatial diagram of Sherman traps in vegetation sites, where I = Ironwood, NI = No Ironwood.


mine abundant and rare species in the sampledsites. These curves were constructed calcu-lating the abundance logarithm of each speciesand are graphed (on the x axis) ordering fromthe most abundant species to the least abun-dant. This type of curve allows one to observethe dominance or evenness of the differentspecies within the community. When acommunity is highly equitable the curve of thetable is horizontal. On the other hand whenthere is a high dominance species the curve hasa vertical form. To analyze the heterogeneitywe calculated the index of equity of Shannon-Weaver and Pielou (J´). This last one measuresthe proportion of the observed diversity in rela-tion to the maximum expected diversity. Itsvalues go from 0 to 1, this last value isachieved when all the species abundance isequally distributed in the community(Magurran 2004).

Structure of the Microhabitat Additionally, in the sites where traps werelocated (with or without ironwood), we quan-tified canopy cover by measuring twodistances: from the tree or shrub to the north(D1) and another to the east direction (D2); thecanopy cover was calculated with thefollowing formula of area: A =0.25 * 3.1416 *D1* D2. To determine if there are differencesin cover between sites with and without iron-wood a t-test was applied. To establish ifcanopy cover has an effect on the number ofcaptured rodents, a simple linear regressionwas made. The analysis of cover was madewith 256 sites per treatment.

In the sites where the traps were located, thedistance to the nearest tree and shrub wasmeasured, using the point-quarter samplingmethod (Brower et al. 1998); to establish ifthere were significant differences between thedistances to the nearest tree and bush of thesampling sites, a t-test (Zar 1999) was appliedto each case. To determine if the structure ofthe vegetation (distance to the nearest tree andbush) of the sampled sites had an effect on thenumber of captured organisms, a Pearsoncorrelation was made for each case (Zar 1999).

The previous analyses were made with 200sites per treatment.

RESULTSThe sampling effort was 6,732 trap nights. Atotal of 599 rodents were captured, concen-trated in three families, seven genera and 10species. The relative abundance of eachspecies of rodent was different. The mostabundant rodent was the spiny pocket mouse(Chaetodipus baileyi), followed by thekangaroo rat (Dipodomys merriami), desertpocket mouse (C. penicillatus) and the white-throated woodrat (Neotoma albigula). Theleast abundant rodents were two species ofPeromyscus (Table 1).

Association with Ironwood We found no preference for sites with iron-wood, because there were no significantdifferences in the treatments with or withoutironwood (F = 0.012, df = 1, P = 0.9130),nevertheless, a significant difference betweenthe abundance of the species (F = 8.823, df =9, P = 0.0001; Figure 3) was found.

Species diversity in the sites with ironwoodwas higher (H´= 1.50) in comparison with thesites without ironwood (H´= 1.26), neverthe-less, there was not a significant difference inthe diversity between both treatments (t =0.276, df = 598, P = 0.05).

When diversity was analyzed between sites,Las Glorias and Lobos, the diversity of rodentswas higher in areas with ironwood (LasGlorias H´= 0.8019; Lobos H´=1.624) than inareas without ironwood (Las Glorias H´=0.7785; Lobos H´= 1.573). For the San Judasand Pozo Hondo ranches, diversity was higherin the areas without ironwood (San Judas H´=0.7646; Pozo Hondo H´= 1.297), than in areaswith ironwood (San Judas H´= 0.6588; PozoHondo H´= 1.072).

Community AnalysisThe site with higher specific richness (Dmg)was Lobos Ranch, followed by San Judas,Pozo Hondo, and Las Glorias ranches. In addi-tion, sites that presented a higher diversity and


dominance were Lobos and Pozo Hondoranches (Table 2).

The estimates of alpha diversity Chao2 andBootstrap predicts that there are seven speciesof rodents in the Pozo Hondo and Lobosranches, while the estimate of first order Jack-knife predicts eight species (Figure 4).

The rank-abundance curves (Figure 5)showed in the four ranches a high dominanceof the family Heteromvidae, representedmainly by the species Chaetodipus baileyiand Dipodomys merriami. The speciesChaetodipus baileyi was the dominant speciesin three sites (Lobos, San Judas and Las

Table 1. Species richness and relative abundance of small mammals captured.Family Species AIW* PIW* Total Relative AbundanceHeteromyidae Perognathus flavus 18 12 30 0.0045

Chaetodipus baileyi 148 158 306 0.0455Chaetodipus penicillatus 31 35 66 0.0098Dipodomys merriami 66 49 115 0.0171

Muridae Neotoma albigula 18 37 55 0.0082Peromyscus maniculatus 1 0 1 0.0001Peromyscus eremicus 3 9 12 0.0018Peromyscus boylii 0 1 1 0.0001Onychomys torridus 5 3 8 0.0012

Sciuridae Spermophilus tereticaudus 3 2 5 0.0007*AIW = Ironwood absence, *PIW = Ironwood presence.

Figure 3. Number of individuals by species between treatments (with or without ironwood).


Table 2. Species, number of individuals and diversity indexes for the study areas.Species Lobos San Judas Pozo Hondo Las GloriasSpermophilus tereticaudus 4 1 0 0Dipodomys merriami 65 13 19 18Chaetodipus baileyi 129 90 2 85Chaetodipus penicillatus 47 2 5 10Perognathus flavus 23 5 0 2Neotoma albigula 53 0 2 0Onychomys torridus 5 0 2 0Peromyscus eremicus 3 1 6 0Peromyscus boylii 1 0 0 0Peromyscus maniculatus 1 0 0 0Number of Individuals 331 112 36 115Total Number of Species 10 6 6 4Margalef Index 1.551 1.060 1.395 0.632Shannon-Wiener Index 1.637 0.721 1.392 0.797Pielou Index 0.711 0.313 0.605 0.346

Figure 4. Species accumulation curve of rodents using three non-parametric estimates ofalpha diversity.


Glorias ranches), while in Pozo Hondo Ranchthe dominant species was Dipodomysmerriami.

Structure of the Microhabitat Average cover was higher on sites with iron-wood (x =14.36, SD = 21.76, n = 256) than onsites without ironwood (x = 6.16, SD = 10.23,n =256; t = 5.46; df = 510; P = 0.0001). Inaddition, we found a negative associationbetween cover and number of animalscaptured (r = 0.651, df = 26, P = 0.0002;Figure 6). Dipodomys merriami presents ahigher number of captures in areas withreduced cover (r = 0.669, df = 9, P = 0.0342;Figure 7).

Average distance between trees on siteswith ironwood (x = 12.94; DE = 8.84; n = 200)and without ironwood (x = 13.12; SD = 9.11;n =200) was not significantly different (t = -

0.2, df = 398, P = 0.8414). Nevertheless, wefound that shrub distance was higher (t = 4.14,df = 398, P = 0.0001) on sites with ironwood(x = 3.76; SD = 1.61; n = 200) than on siteswithout ironwood (x = 3.10; SD = 1.58; n =200). Regarding the correlations, a relationshipbetween the distances to shrubs and captureswas not found (r2 = -0.0732; P = 0.059), butwe did find a relationship between distance totrees and small mammal captures (r2= -0.732;P = 0.010).

DISCUSSIONThe species with higher number of captureswere the ones belonging to the familyHeteromyidae and the species Neotomaalbigula. These results coincide with thosereported in other studies of small mammals inthe Sonoran Desert (Duncan 1990; Petryszynand Russ 1996).

Figure 5. Rank-abundance curve for the species of small mammals in the four study sites.Where CB corresponds to Chaetodipus baileyi, DM to Dipodomys merriami, NA toNeotoma albigula, CP to Chaetodipus penicillatus, PF to Perognathus flavus, OT toOnychomys torridus, ST to Spermophilus teriticaudus, PE to Peromyscus eremicus, PMto Peromyscus maniculatus and PB to Peromyscus boylii.


In this study we found that ironwood wasnot selected as a nurse species by the commu-nity of small mammals, because there were nodifferences in the abundance between the treat-ments (with and without ironwood), also, wefound that the diversity in both treatments wassimilar, so these results differ with thosereported for perennial seedlings and birds.Other authors have found that in sites withironwood the diversity and species abundanceis higher when compared with sites without O.tesota (Tewksbury and Petrovich 1994;Nabhan and Behan 2000).

Regarding the evaluation of the samplingby the curves of accumulation of species, thedifferent estimates of alpha diversity Chao2,Jackknife of first order and Bootstrap predictbetween seven and eight species per site. Fromthe previous supposition in three sites (PozoHondo, San Judas and Las Glorias) we docu-mented 100% of the diversity of small

mammals, while for Lobos we documented80%, nevertheless, in this site we counted onlyone record of Peromyscus boylii and P. manic-ulatus, which may mean that these individualshave an erroneous identification and belong tothe species Peromyscus eremicus (Table 1).

In this study, we found a high dominanceby the species Chaetodipus baileyi which canbe explained by this species preference forareas with trees and bushes of large size(Paulson 1988), characteristics that werepresented at all study sites.

The rank-abundance curves (Figure 5)show that C. baileyi was the dominant speciesin three sites, followed by D. merriami, whichcan mean that these two species are in directcompetition; as has been observed in otherstudies, where C. baileyi excludes D. merriami(Paulson 1988).

Despite the fact that the ironwood did notexplain the presence of small mammals, it was

Figure 6. Simple lineal regression of the ranks of cover and number of capturedindividuals.


found that the structure of the habitat (cover ofthe canopy and distance to trees) has an influ-ence on the abundance of rodents, as we founda higher usage of sites with small cover (0.05to 11.9 m2) and small distances to trees (6 to15 m), which may indicate that the docu-mented species in this study concentrate theirforaging activity on sites with closed vegeta-tion structure, which reduces the detectionfrom predators (i.e., owls, coyotes, kit fox, orsnakes), because closed areas or sites are suit-able to reduce the risk of predation (Thompson1982; Kotler 1984; Brown et al. 1988).

The relationship that we found of a highernumber of captures of Dipodomys merriami inopen sites has been reported in other studiesand it has been attributed to the biped loco-motion of this heteromyid, which favors itsability to escape from predators (Brown andLieberman 1973; Price 1978; Hallet 1982;Thompson, 1982).

LITERATURE CITED Alvarez-Castañeda., S. T., and J. L. Patton. 2000.Mamíferos del noroeste de México. Vols. I and II.CIBNOR, La Paz, Baja Cailifornia Sur.Brower, E. S., H. J. Zar, and C. N. von Ende, editors.1998. Field and laboratory methods for generalecology. 4th edition. McGraw-Hill. Brown, H. J., and G. A. Lieberman. 1973. Resourceutilization and coexistence of seed-eating desertrodents in sand dune habitats. Ecology 54:788–797.Brown, H. J., O. J. Reichman, and D. W. Davidson.1979. Granivory in desert ecosystems. AnnualReview Ecology System 10: 201–217.Brown, S. J., B. Kotler, J. R. Smith, and W. O.Wirtz. 1988. The effects of owl predation on theforaging behavior of heteromyd rodents.Oecologia. 76: 408–415.Búrquez, A. and M. Quintana. 1994. Islands ofdiversity: Ironwood ecology and the richness ofperennials in a Sonoran Desert biological reserve.In Ironwood: An ecological and cultural keystoneof the Sonoran Desert, edited by G. P. Nabhan andJ. L. Carr, pp. 9–27. Occasional Papers in Conser-vation Biology, Occasional Paper No.1.Washington, D.C.

Figure 7. Simple lineal regression of cover and captures of Dipodomys merriami.


Campos, C., R. Ojeda, S. Monge, and M. Dacar.2001. Utilization of food resources by small andmedium-sized mammals in the Monte Desertbiome, Argentina. Austral Ecology. 26: 142–149.Colwell, R. K. 2004. Estimates: Statistical estima-tion of species richness and shared species fromsamples. Version 7. User´s Guide and Applicationpublished at: http://purl.oclc.org/estimates.Coronel-Arellano, H., I. Sáyago-Vázquez, and C.A. López-González. 2005. Pequeños mamíferosasociados al Palo Fierro (Olneya tesota) en eldesierto sonorense. In XXll simposio sobre faunasilvestre. Gral. MV. Manuel Cabrera Valtierra, pp.94–102. Universidad Nacional Autónoma deMéxico.Curtin, G. C., A. D. Kelt, C. T. Frey, and J. H.Brown. 1999. On the role of small mammals inmediating climatically driven vegetation change.Ecology Letters 2: 309–317.Duncan, K. D. 1990. Small mammal inventory ofChiricahua National Monument Cochise County,Arizona. Technical Report No. 30. National ParkService Cooperative Studies Unit, University ofArizona.Fauth, J. E., J. Bernardo, M. Camara, W. J. Rese-tarits, J. Van Buskirk, and S. A. McCollum. 1996.Simplifying the jargon of community ecology: Aconceptual approach. The American Naturalist147: 282–286.Folgarait, J. P., and E. O. Sala. 2002. Granivory ratesby rodents, insects, and birds at differentmicrosites in the Patagonian steppe. Ecography25: 417–427.Giannoni, M. S., M. Dacar, P. Taraborelli, and E. C.Borghi. 2001. Seed hoarding by rodents of theMonte Desert, Argentina. Austral Ecology 26:259–263.Hallet, G. J. 1982. Habitat selection and the commu-nity matrix of a desert small-mammal fauna.Ecology 63: 1400–1410.Information System and Automation, Corp. 2005.Rancho Lobos. http://www.rancholobos.com/rancho-lobos/goma.shtml. Consulted on 15 May2006.Instituto Nacional de Estadística, Geografía eInformática. 2000. Síntesis de informacióngeográfica del Estado de Sonora. Aguascalientes,México.Kays, W. R., and D. E. Wilson. 2002. Mammals ofNorth America. Princeton Field Guides. Kelt, A. D., K. Rogovin, G. Shenbrot, and J. H.Brown. 1999. Patterns in the structure of Asianand North American desert small mammalcommunities. Journal of Biogeography 26:825–841.Kotler, P. B. 1984. Risk of predation and the struc-ture of desert rodent communities. Ecology 65:689–701.MacMahon, A. J. 1994. Desert. National AudubonSociety Nature Guides. Magurran, A. E. 2004. Measuring biological diver-

sity. Blackwell Publishing. McAuliffe, J. R. 1984. Prey refugia and distribu-tions of two Sonoran Desert cacti. Oecologia 65:82–85.McShea, J. W., J. Pagels, J. Orrock, E. Harper, E.,and K. Koy. 2003. Mesic deciduous forest aspatches of small-mammal richness within anAppalachian Mountain forest. Journal ofMammalogy. 84: 627–643.Nabhan, G. P., and M. Behan. 2000. Desert iron-wood primer, biodiversity and uses associatedwith ancient legume and cactus forest in theSonoran Desert. Arizona-Sonora Desert Museum.Nichols, J. D., and M. J. Conroy. 1996. Techniquesfor estimating abundance and species richness. InMeasuring and monitoring biological diversity:Standard methods for mammals, edited by D. E.Wilson, F. R. Cole, J. D. Nichols, R. Rudran, andM S. Foster, pp. 177–234. Smithsonian InstitutionPress, Washington, D.C.Paulson, D. D. 1988. Chaetodipus baileyi.Mammalian Species 297: 1–5.Petryszyn, Y., and S. Russ. 1996. Nocturnal rodentpopulation densities and distribution at Organ PipeCactus National Monument, Arizona. TechnicalReport No. 52. National Biological Survey, Coop-erative Park Studies Unit, University of Arizona.Price, V. M. 1978. The role of microhabitat in struc-turing desert rodent communities. Ecology 59:910-921. Price, V. M. and R. A. Correll. 2001. Depletion ofseed patches by Merriam’s kangaroo rats: Are GUDassumptions met? Ecology Letters 4: 334-343.Sassi, L. P., J. Tort, and E. C. Borghi. 2004. Effectof spatial and temporal heterogeneity on granivoryin the Monte Desert, Argentina. Austral Ecology29: 661–666.Shepherd, L. U., and D. A. Kelt. 1999. Mammalianspecies richness and morphological complexityalong an elevational gradient in the arid southwest.Journal of Biogeography 26: 843–855.Tewksbury, J. J., and C. Petrovich. 1994. The influ-ences of ironwood as a habitat modifier species:A case study on the Sonoran Desert coast of theSea of Cortez. In Ironwood: An ecological andcultural keystone of the Sonoran Desert, edited byG. P. Nabhan and J. L. Carr, pp. 29–54. OccasionalPapers in Conservation Biology, Occasional PaperNo.1. Washington, D.C. Thompson, S. D. 1982. Microhabitat utilization andforaging behavior of bipedal and quadrupedalheteromyid rodents. Ecology 63: 1303–1312.Turner, M. R., and D. E. Brown. 1994. Tropical-subtropical desertlands. In Biotic communitiessouthwestern United States and northwesternMéxico, edited by D. E. Brown, pp. 180–221.University of Utah Press, Salt Lake City, UT.Whitaker, O. J. 2002. Field guide to mammals.National Audubon Society. Zar, J. H. 1999. Bioestatistical analysis. PrenticeHall.

FINDING THAT 4-STAR DINER OR HOW BATS MIGHT “ANTICIPATE” PRODUCTIVEFORAGING AREASDebbie C. Buecher, Ronnie Sidner, and John L. Koprowski

Riparian corridors, particularly in the Amer-ican Southwest, are critical habitat for manyplant and animal species (Ohmart andAnderson 1982). The structural complexity ofstreamside landscapes often contains tremen-dous biodiversity — often contributing muchgreater diversity than their proportion of landarea (Neary et al. 2005). Unfortunately,riparian environments are often at risk due toanthropogenic pressures such as water extrac-tion, cattle grazing, urbanization, mining, andtimber harvest (Steiner et al. 2000). Theseimpacts have caused riparian landscapes todwindle to less than 1% of the area known inthe late 19th century (Ffolliott and Thorud1974). As well as a reduction and degradationof historic riparian environments, the intro-duction of exotic species has added additionalpressure to the quality of these landscapes.Because of concern for native speciesimpacted by these changes, studies have beenconducted to evaluate resource use across taxa(Greier and Best 1980; Hunter et al. 1988;Anderson 1994; Ellis et al. 2000). Riparianlandscapes are also important foraging habitatfor bats and may provide their only access todrinking water (Grindal et al. 1999; Hollowayand Barclay 2000). We used a Sonoran Desertriparian corridor to define and quantify theassociated bat community. Our study is aneffort to provide land management agencieswith critical information necessary to makeinformed decisions regarding wildlife in anarid region.

Historically it has been difficult to studyand monitor habitat use by nocturnalmammals, due to the inability to observebehavior in low light levels. An assessment ofpeer-reviewed literature indicated a paucity ofresearch publications, particularly on bats androdents in Arizona (Koprowski et al. 2005). Inaddition, because body mass of most temperatebats is ≤ 15 grams, Aldridge and Brigham(1988) found advances in radio-transmittertechnology (i.e., miniaturization) necessary toallow for the effective monitoring of batsduring their study of night-time foraging. Dueto the labor-intensive nature of these studies,there is still only a limited understanding ofhow individuals use their environment. Studiesexamining broad landscape questions amongmultiple species are still prohibitively expen-sive. On the other hand, due to use of biosonarby bats for spatial orientation and to forage forfood (Griffin 1958), recent advances inacoustic technology have given us another toolto monitor bats across landscapes. This studybegan as an investigation of bat use along aSonoran Desert riparian corridor but serendip-itously gave us some greater insight into howbats perceive and perhaps even anticipate foodresources in heterogeneous environments.

STUDY SITESabino Canyon is a Sonoran Desert ripariancorridor administered as a recreational area byCoronado National Forest, U.S. Departmentof Agriculture. This rugged canyon system,

254 Buecher, Sidner, and Koprowski

just north of Tucson, Arizona, is in the frontrange of the Santa Catalina Mountains. Thesemountains lie on the northeastern edge of theSonoran Desert and rise to 2791 m above the728 m desert floor. The Santa Catalinas arecomposed of hard metamorphic granites andthe mid to lower reaches of Sabino Canyon arecharacterized by dramatic vertical cliffs (Bezy2004). This popular oasis is situated in theArizona Upland subdivision of the SonoranDesertscrub Biome (Turner and Brown 1994).Although the Forest Service has closed thearea to private vehicles, they provide areas tohike, picnic, and bicycle, plus a tram systemtransports visitors 5.6 km along the bottom ofthe canyon. The proximity of Sabino Canyonto a large metropolitan area (~ 800,000 resi-dents) ensures heavy visitation (> 1.4 millionvisitors/year) to this scenic attraction. In addi-tion, two major forest fires (Bullock Fire 2002and Aspen Fire 2003) in the upper watershedof this canyon have applied further pressure ona resource already impacted by recent years ofdrought (Swetnam and Betancourt 1998).Annual precipitation for Tucson is approxi-mately 300 mm/year, with the surroundingmountain ranges getting up to 750 mm/year.The Sonoran Desert has a bimodal rain pattern,with about 50% of the annual precipitationoccurring during December–February,producing snow above 1500 m. The rest of theprecipitation generally falls during July–September as intense summer monsoonstorms. Sabino Creek begins high in the SantaCatalina Mountains (1800 m elevation) inmixed conifer forest and has one of the largerwatersheds in the Santa Catalina Mountains(~ 92 km2), resulting in Sabino Creek flowingfor longer periods than other canyons. Duringsnow melt or summer monsoons the canyoncan flood, but the normal pattern of precipita-tion produces moderate to low flows forapproximately 8 months of the year. Earlysummer and late fall are the creek’s driestperiods. During these months the water oftensinks into the channel’s sandy floor, flowsunderground, and is inaccessible to wildlife.The banks of Sabino Creek support deciduous

riparian vegetation dominated by cottonwood(Populus fremontii), willow (Salix gooddingii),ash (Fraxinus pennsylvanica), walnut (Juglansmajor), and sycamore (Platanus wrightii)(Minckley and Brown 1994). The steep slopesof Sabino Canyon are dominated by saguaro(Carnegiea gigantea), prickly pear and chollacacti (Opuntia spp.), palo verde (Cercidiumspp.), and brittle bush (Encelia farinosa)(Turner and Brown 1994).

METHODSPassive Acoustic Sampling

The recent availability of portable, field- robustyet affordable ultrasonic bat detectors hasmade evaluation of activity patterns mucheasier to conduct than previously (Murray etal. 1999; Johnson et al. 2002). In addition batdetectors have greatly increased our knowl-edge of how bats use a resource (Hayes 1997;Vaughn et al. 1997; Humes et al. 1999;Kalcounis et al. 1999). During 2005 weconducted monthly passive (i.e., researcherabsent) acoustic sampling in an effort tocompare bat use between two environments.Our sampling protocol involved deploying 4Anabat II frequency division bat detectors(Titley Electronics, Ballina, N.S.W. Australia)simultaneously each month, January throughDecember, in two different habitats alongSabino Creek. We chose two sites from thelower canyon to represent riparian environ-ments with deciduous streamside vegetation(Minckley and Brown 1994). We anticipatedthat this habitat, characterized by dense vege-tation, could support greater insect biomass forforaging bats. We selected two additional sitesin a more arid, open habitat with less vegeta-tion. These upper sites were chosen whereSonoran desertscrub extends from the hillsidesto the canyon floor. The null hypothesis for thisstudy was that bats would use both habitatsequally, and the alternative hypothesis was thatbats would prefer the lower riparian area dueto its additional foraging resources. Locationsof the four sites were: Site 1, ~ 300' aboveSabino Canyon Dam; Site 2, a pool area justupstream from Shuttle Stop #1; Site 3, just

Buecher, Sidner, and Koprowski 255

upstream from Tram Bridge #7; and Site 4, justupstream from the concrete foundation forAnderson Dam (Figure 1). Sites 1 and 2 werecategorized as “Riparian” sites and Sites 3 and4 as “Desertscrub” sites (Figure 2). Wesampled on a monthly basis (except October),plus we sampled twice in January, July andSeptember.

Each month we synchronized internalclocks in each detector unit and programmedinternal timers to turn the sampling equipmenton 30 minutes prior to sunset, running contin-uously all night and turning off 30 minutes pastsunrise. The afternoon prior to sampling, weset equipment at each site in a locked, water-

proof housing designed by T. Messina(http://home.earthlink.net/~nevadabat/). Be-cause the microphone system for the Anabat IIis sensitive to moisture, we placed the micro-phone at a 45° angle downward to a horizontal11 x 11 cm plexiglass reflector placed on topof a 2 m high PVC pole (Livengood et al.2003). The next morning we downloaded data(i.e., sound files of bat calls labeled withdate/time stamps) onto a laptop computer. Weused the number of call files as a measure ofbat activity and evaluated echolocation callsusing Anabat6, Analook and Anamusic soft-ware (O’Farrell et al. 1999). With this systemwe sampled simultaneously at 4 sites per night

Figure 1. Map of Lower Sabino Canyon showing the sites used for the comparativeacoustic sampling study. Sites 1 and 2 are in riparian deciduous vegetation and Sites 3and 4 are in Sonoran desertscrub.


per month, allowing us to evaluate andcompare resource use and determine wherebats are most active along the canyon bottom.

RESULTSDuring 14 nights of acoustic sampling in 2005,we passively recorded approximately 35,000files of bat calls. Often Anabat detectors willrecord more than an individual bat’s call on thesame file, however we used the number of filesas a general measure of overall activity at eachsite. Figure 3 shows the calls by samplingperiod distributed throughout the year. The batdetector placed at Site 1 recorded a total of7,234 files, the detector at Site 2 recorded5,436 files, the detector at Site 3 recorded12,354 files and the detector at Site 4 recorded9,823 files. We expected that the greatest levelof foraging activity, reflected by the number ofcall files, would be along the lower portion ofSabino Creek (Sites 1 and 2) where densedeciduous riparian vegetation exists to supportgreater insect biomass.

However, we found that the upper reachesof the canyon (Sites 3 and 4) often had greaterbat activity. In addition, the Anabat detectorjust upstream from Bridge #7 (Site 3) had thegreatest number of calls of all sites alongSabino Creek. These results support neitherour null hypothesis nor our alternative hypoth-

esis and required a new evaluation of thehabitat parameters along Sabino Creek.

DISCUSSIONThe outcome of this study was unexpected andrequired further analysis to explain the results.We expected greater foraging by bats, reflectedin the number of calls, along riparian stretchesof Sabino Creek with characteristically densevegetation. However, we believe that thedifferences that occurred, particularly the highlevels of foraging activity above Bridge #7, arereal and explainable when we use additionalscientific disciplines during analysis.

Physics and MeteorologyWhen we review basic laws of physics, weknow that warm air is less dense than cold air,causing warm air to rise; whereas cold air sinks(Geiger 1957; Halliday and Resnick 1966).Additionally, topography has a huge influenceon the microclimate in valleys or canyons(Geiger 1957), particularly in the mountainous,semi-arid American Southwest. During the daythe basin floor heats up, causing hot air to riseand move up-canyon. On a warm summer day,warm up-slope currents influence the micro-climate of the valley and will move small bitsof dust and detritus up-canyon. As the warmair rises to higher elevations, the air loses heat

Figure 2a. Riparian sampling site aboveSabino Canyon Dam.

Figure 2b. Desertscrub sampling siteabove Bridge #7.


Figure 3. Plot of monthly totals of the acoustic calls of bats by sampling site.

due to adiabatic cooling, and this cooler airmass sits on the “cushion” of rising warmer air.However, once the sun sets, the “engine” thatdrove this phenomenon reverses and the coldair sitting in the upper watershed on the moun-tain, being heavier, settles and flowsdown-canyon much like a river of cold water(Geiger 1957). This can be observed at nightwhen crossing a dry arroyo or streambedwhere the air temperature at the flow-line ofthe canyon floor is often colder than thesurrounding terrain.

HydrologyCombining physics and hydrology, the flow ofwater down valleys can be modeled usingcomputer software packages (HEC 2 Manual1990) to predict where water will be slowed orimpounded by topographic irregularities.These variable landscapes can affect flowvelocities and create currents that produce sideeddies in the channel. These circular currentscan capture leaves and detritus in small

vortices along the side of the main flow, andactually move this debris upstream along thebackside of a circular eddy until fluctuation inthe current occurs that changes the flow regime(Rouse 1978). A practical application of thisphenomenon is seen in guidebooks availableto kayakers and river-runners that show thesafest way to maneuver rapids in rivers. Thesehydrologic models might also allow us to eval-uate other “fluids,” such as cold-air drainage,under similar topographic constraints. Wesuggest that after sunset, cold air moves alongcanyon floors much like water flows downhilland will create eddies in the air column just aseddies occur in the water flow (Geiger 1957).The cool air will be more dense than an equalamount of warm air flowing up- canyon duringthe day and will capture dust and detritus ineddies.

EntomologyFlight for insects is an effort to counteractforces acting in opposition (i.e., gravity and


drag) to desired motion. A method to evaluateinsect flight through air uses a proportionalrelationship of inertial to frictional forces(Chapman 1998). Forces acting on a flyinginsect include “lift” created by wing move-ments, “drag” or friction of the mediumbecause of the insect’s mass and shape, plus“thrust,” a function of the wings lifting theinsect up and forward (Brodsky 1994). “Drag”is the force acting in opposition to forwardmovement (Drag = 0.5 CD ρ S V2), where ρ =the density of the medium, V = the velocity ofthe medium, S = the area of the insect perpen-dicular to the direction of movement and CD= the Drag Coefficient (Chapman 1998). TheDrag Coefficient depends on the shape of theinsect (i.e., how streamlined its presence is toair resistance), the insect’s surface texture(smooth creating less drag than rough) and itsReynolds Number. The Reynolds Number(Re) is a dimensionless value that measures therelative importance of the speed and size of aninsect to the density and viscosity of themedium through which the insect is flying (Re= ρVl / µ; Chapman 1998), where ρ = thedensity of the medium, µ = the viscosity of themedium, V = the speed of the insect, and l =the size of the insect. Generally, the larger theinsect, the larger the Reynolds number becausethe viscosity becomes less critical as the sizeof the insect increases (Brodsky 1994;Lehmann 2002). However, for prey-sizedinsects (≤ 5 mm), flight may be influenced bythe air’s density and — particularly fornocturnal insects flying along a cooler canyonbottom — could be equated to swimmingthrough molasses. It is in this way that insectsmay be at the mercy of airflow patterns and thefluid nature of air (Lewis 1965), allowing batsto predict where to forage efficiently on patchyfood resources.

Extensive research in Europe to modelinsect infestations between crop fields providesan example of how insects are distributed bywinds (Lewis 1970; Landon et al. 1997).Lewis showed that smaller insects (< 3 mm2)drift along with the breeze — like inert parti-cles and are often retained in eddy zones.

However, larger insects (> 4-5 mm2) will actu-ally use eddies as a refuge and will alsoaccumulate in shelter zones or eddies in theflow. In England this often occurred alonghedgerows; however any break or protectionfrom the main flow of the wind will providean opportunity for insects to accumulate(Lewis 1965). Epila (1988) also showed thatinsects are poor “aeronauts” and, onceairborne, are at the mercy of the wind untilthey encounter a wind shelter or shadow thataffords protection. These studies may provehelpful when determining why bats arecongregating in larger numbers at a particularsite along Sabino Creek.

MammalogyInsectivorous bats often cope with heteroge-neous patches of prey, both temporally andspatially (Barclay 1991; Brigham 1991).Research has shown that bats often anticipateproductive feeding areas, given their spatialmemory and knowledge of the landscape (Bell1980). It is not unexpected that bats foragingalong Sabino Creek would anticipate pocketsof prey that occur dependably due to topo-graphic irregularities along the canyon bottom.These would actually be more predictable thanthe heterogeneous emergence patterns ofinsects during the summer months.

Disciplines CombinedWhen we investigate the topography alongSabino Canyon at Site 3, where we saw thegreatest foraging activity, we observe that aprominent rocky ridge projects into thecanyon cross-section from the west (Figure 4).As Sabino Creek flows under the bridge justupstream from Bridge #7, it turns slightly andflows parallel to the paved road. However,when the main flow of the creek hits the ridgeof rock at Bridge #7, it forces the creek tomake an abrupt 90° bend to the east. Thislikely produces a large eddy in the flowregime and essentially traps insects in aprotected column of air upstream of the ridge.This could explain the high number offoraging calls and suggests that bats are able


to take advantage of rich food resources,particularly if they are not as patchy asephemeral insect swarms.

MANAGEMENT IMPLICATIONSThree primary factors characterize the distri-bution of bats across the landscape: availabilityof appropriate roosts, accessibility to drinkingwater and adequate food resources (Racey andEntwistle 2003). Resource managers oftenfocus on preservation of bat populationsthrough appropriate management of dayroosts, especially maternity colonies and hiber-nacula (Brady 1982; Brigham 1993; Thomaset al. 1990; Tuttle 1977). Although this hasproven critical in maintaining healthy batpopulations, proper land management policiesmust also consider availability of other limitingresources. When protecting bat populations,habitat associated with roosts must guaranteecontinued accessibility to food and drinkingwater.

Our study used a multi-discipline approachto evaluate how bats locate food at a landscapelevel, given the heterogeneous nature of theseresources (Buecher 2007). This approachshows promise in predicting the presence ofbats at the scale at which these animals must

evaluate their habitat. Hopefully, these datawill provide resource managers with additionalinsight and information critical to determiningresource use by bats along an invaluableriparian corridor. Given the many pressures onthis unique environment, this study willprovide Forest Service personnel with infor-mation necessary when making decisionsregarding the protection and conservation ofthis resource.

ACKNOWLEDGMENTSThis study was made possible with logisticalsupport of Coronado National Forest, SantaCatalina Ranger District. They were mostgenerous in allowing us access to SabinoCanyon Recreational Area at all hours of theday and night. Also, the Nongame Branch ofthe Arizona Game and Fish Department helpedus acquire the necessary Scientific CollectingPermits to conduct this project. This projectwas funded by an Arizona Game and FishUrban Heritage Grant, a grant from T & E,Inc., and by a financial donation from TheFriends of Sabino Canyon. This financial aidprovided additional Anabat II ultrasonic batdetectors, allowing us to conduct simultaneoussampling along Sabino Creek.

Figure 4. Note granite ridge extending into Sabino Canyon flow-line.


LITERATURE CITEDAldridge, H. D. J. N., and R. M. Brigham 1988.

Load carrying and maneuverability in aninsectivorous bat: A test of the 5% “rule” of radio-telemetry. Journal of Mammalogy 69(2):379–382.

Anderson, D. C. 1994. Demographics of smallmammals using anthropogenic desert riparianhabitat in Arizona. Journal of WildlifeManagement 58(3): 445–453.

Barclay, R. M. R. 1991. Population structure oftemperate zone insectivorous bats in relation toforaging behavior and energy demand. Journal ofAnimal Ecology 60: 165–178.

Bell, G. P. 1980. Habitat use and response topatches of prey by desert insectivorous bats.Canadian Journal of Zoology. 58: 1876–1883.

Bezy, J. V. 2004. A guide to the geology of SabinoCanyon and the Catalina Highway. Down-to-Earth #17. Arizona Geological Survey, Tucson,AZ.

Brady, J. T. 1982. Cave management for theendangered Indiana bat (Myotis sodalis) and graybat (Myotis grisescens). In National cavemanagement proceedings, edited by H. Thornton& J. Thornton, pp. 86–95. The American CaveConservation Association, Richmond, VA.

Brigham, R. M. 1991. Fidelity in foraging androosting behaviour by the big brown bat(Eptesicus fuscus). Canadian Journal of Zoology69: 117–121.

Brigham, R. M. 1993. The implications of roostsites for the conservation of bats. In Proceedingsof the third prairie conservation and endangeredspecies workshop. Provincial Museum of AlbertaNatural History Occasional Paper #19: 361–365.

Brodsky, Andrei K. 1994. The evolution of insectflight. Oxford University Press, Oxford, England.

Buecher, D. C. 2007. Bat diversity, resource use andactivity patterns along a Sonoran Desert ripariancorridor. M.S. Thesis, University of Arizona,Tucson, AZ.

Chapman, R. F. 1998. The insects: Structure andfunction. 4th ed. Cambridge University Press,Cambridge, England.

Ellis, L. M., M. C. Molles, Jr., C. S. Crawford, andF. Heinzelmann. 2000. Surface-active arthropodcommunities in native and exotic riparianvegetation in the Middle Rio Grande Valley, NewMexico. The Southwestern Naturalist 45(4): 456–471.

Epila, J. S. O. 1988. Wind, crop pests and agroforestdesign. Agricultural Systems 26: 99–110.

Ffolliott, P. F. and D. B. Thorud. 1974. Vegetationmanagement for increased water yield in Arizona.Agricultural Experiment Station TechnicalBulletin 215. University of Arizona, Tucson, AZ.

Geiger, R. 1957. The climate near the ground.Harvard University Press, Cambridge, MA.

Greier, A. R. and L. B. Best. 1980. Habitat selection

by small mammals of riparian communities:Evaluating effects of habitat alteration. Journal ofWildlife Management 44(1): 16–24.

Griffin, D. R. 1958. Listening in the dark, theacoustic orientation of bats and men. YaleUniversity Press, New Haven, CT.

Grindal, S. D., J. L. Morissette, and R. M. Brigham.1999. Concentrations of bat activity in riparianhabitats over an elevational gradient. CanadianJournal of Zoology 77: 972–977.

Halliday, D. and R. Resnick. 1966. Physics, Part Iand II. John Wiley and Sons, New York, NY.

Hayes, J. P. 1997. Temporal variation in activity ofbats and the design of echolocation-monitoringstudies. Journal of Mammalogy 78(2): 514–524.

Holloway, G. L. and R. M. R. Barclay. 2000.Importance of prairie riparian zones to bats insoutheastern Alberta. Ecoscience 7(2): 115–122.

Humes, M. L., J. P. Hayes, and M. W. Collopy.1999. Bat activity in thinned, unthinned, and old-growth forests in western Oregon. Journal ofWildlife Management 63(2): 553–561.

Hunter, W. C., R. D. Ohmart, and B. W. Anderson.1988. Use of exotic saltcedar (Tamarix chinensis)by birds in arid riparian systems. The Condor 90:113–123.

Johnson, J. B., M. A. Menzel, J. W. Edwards, andW. M. Ford. 2002. A comparison of two acousticalbat survey techniques. Wildlife Society Bulleting30(3): 931–936.

Kalcounis, M. C., K. A. Hobson, R. M. Brigham,and K. R. Hecker. Bat activity in the borealforest: Importance of stand type and verticalstrata. Journal of Mammalogy 80(2): 673–682.

Koprowski, J. L., A. J. Edelman, B. S. Pasch, and D.C. Buecher. 2005. A dearth of data on mammals ofthe Madrean Archipelago: What do we think weknow and what we actually do know. InConnecting mountain islands and desert seas:Biodiversity and management of the MadreanArchipelago II, compiled by G. J. Gottfried, B. S.Gebow, L. G. Eskew, and C. B. Edminster, pp.412–415. General Technical Report RMRS-P-36,U.S. Department of Agriculture, Forest Service,Rocky Mountain Research Station, Fort Collins,CO.

Landon, F., J. Levieux, D. Rougon, and P. Taupin.1997. Influence of the immediate environment onthe dispersal of Sitona lineatus L. (Col. Curc.)during crop colonization. Agriculture, Ecosystemsand Environment 64: 261–266.

Lehmann, F. O. 2002. The constraints of body sizeon aerodynamics and energetics in flying fruitflies: An integrative view. Zoology 105: 287–295.

Lewis, T. 1965. The effects of an artificialwindbreak on the aerial distribution of flyinginsects. Annals of Applied Biology 55: 503–512.

Lewis, T. 1970. Patterns of distribution of insectsnear a windbreak of tall trees. Annals of AppliedBiology 65: 213–220.

Livengood, K., R. Drobney, and C. Corben. 2003.


What good is a raincheck? Weather proofing forbat detectors and its effect on call reception(Poster). 33rd Annual North AmericanSymposium on Bat Research, Lincoln, NE.

Minckley, W. L. and D. E. Brown. 1994. Sonoranriparian deciduous forest and wetlands. In Bioticcommunities, southwestern United States andnorthwestern Mexico, edited by D. E. Brown, pp.269–273. University of Utah Press, Salt LakeCity, UT.

Murray, K. L., E. R. Britzke, B. M. Hadley, and L.W. Robbins. 1999. Surveying bat communities: Acomparison between mist nets and the Anabat IIbat detector system. Acta Chiropterologica 1(1):105–112.

Neary, D. G., P. F. Ffolliott and L. F. DeBano. 2005.Hydrology, ecology, and management of riparianareas in the Madrean Archipelago. In Connectingmountain islands and desert seas: Biodiversity andmanagement of the Madrean Archipelago II,compiled by G. J. Gottfried, B. S. Gebow, L. G.Eskew, and C. B. Edminster, pp. 316–319.General Technical Report RMRS-P-36, U.S.Department of Agriculture, Forest Service, RockyMountain Research Station, Fort Collins, CO.

O’Farrell, M. J., B. W. Miller, and W. L. Gannon.1999. Qualitative identification of free-flying batsusing the Anabat detector. Journal of Mammalogy80(1): 11–23.

Ohmart, R. D. and B. W. Anderson. 1982. NorthAmerican desert riparian ecosystems. InReference Handbook on the Deserts of NorthAmerica, edited by G. L. Bender, pp. 433–479.Greenwood Press, Westport, CT.

Racey, P. A. and A. C. Entwistle. 2003.Conservation ecology of bats. In Bat Ecology,edited by T. H. Kunz and M. B. Fenton, pp. 680–743. University of Chicago Press, Chicago, IL.

Rouse, H. 1978. Elementary mechanics of fluids.Dover Publications, Inc. New York, NY.

Steiner, R., J. Blair, L. McSherry, S. Guhathakurta,J. Marruffo, and M. Holm. 2000. A watershed ata watershed: The potential for environmentallysensitive area protection in the upper San PedroDrainage Basin (Mexico and USA). Landscapeand Urban Planning 49: 129–148.

Swetnam, T. W. and J. L. Betancourt. 1998.Mesoscale disturbance and ecological response todecadal climatic variability in the AmericanSouthwest. Journal of Climate 11: 3128–3147.

Thomas, D. W., M. Dorais and J. M. Bergeron.1990. Winter energy budgets and cost of arousalsfor hibernating little brown bats, Myotis lucifugus.Journal of Mammalogy 71(3): 475–479.

Turner, R. M. and D. E. Brown. 1994. Tropical-subtropical desertlands. In Biotic communities,southwestern United States and northwesternMexico, edited by D. E. Brown, pp. 180–221.University of Utah Press, Salt Lake City, UT.

Tuttle, M. D. 1977. Gating as a means of protectingcave dwelling bats. In National cave managementsymposium proceedings, 1976, edited by T. Aleyand D. Rhodes, pp. 77–82. Speleobooks,Albuquerque, NM.

Vaughn, N., G. Jones, and S. Harris. 1997. Habitatuse by bats (Chiroptera) assessed by means of abroadband acoustic method. Journal of AppliedEcology 34: 716–730.

ADVANCING LARGE CARNIVORE RECOVERY IN THE AMERICAN SOUTHWESTTony Povilitis and C. Dustin Becker

By passing the Endangered Species Act (ESA)of 1973, the United States Congress committedthe nation to restoring endangered species byprotecting their ecosystems from “economicgrowth and development untempered byadequate concern and conservation” (USC1973). Since then, the science of conservationbiology has confirmed the act’s core impera-tive by demonstrating the tight connectionbetween ecosystem conservation and therecovery of endangered species populations.However, economic interests have increasedtheir political influence on government to apoint where ecosystem conservation is diffi-cult even when it is essential for speciesrecovery. Povilitis et al. (2006) cite examplesof species that are “bureaucratically imperiled”because public agencies have failed to upholdthe ESA by circumventing ecosystem conser-vation in deference to economic interests. TheSan Diego ambrosia (Ambrosia pumila),Pacific fisher (Martes pennanti pacifica),humpback chub (Gila cypha), Florida panther(Puma concolor coryi), leatherback turtle(Dermochelys coriacea), Mexican wolf (Canislupus baileyi), and jaguar (Panthera onca) area few of many bureaucratically imperiledspecies. Bureaucratic practices that stymierecovery of endangered species include with-holding protective listing under the ESA,failure to develop adequate recovery goals andplans, and failure to implement agreed-uponconservation plans.

We discuss how wildlife agencies respon-sible for the recovery of the Mexican wolf andthe jaguar in the southwestern United States

have faltered in their efforts to recover thesespecies. In both cases, deference to stake-holders with economic interests or ideologicalopposition has hampered efforts of the U.S.Fish and Wildlife Service (USFWS) and coop-erating state wildlife agencies to fullyimplement the ESA. After reviewing historicaland contemporary reasons for why thesespecies are endangered, we discuss a numberof actions that could make it easier for agencyprofessionals, conservation organizations, andcitizen wildlife advocates to achieve largecarnivore recovery.

MEXICAN WOLFMexican wolves once ranged from the border-lands of the American Southwest to centralMexico numbering in the thousands, but by theearly 1930s they had been extirpated in theUnited States primarily as a result of acampaign led by the U.S. Bureau of BiologicalSurvey (Brown 1983; Robinson 2005) andpartner agencies. On behalf of the livestockindustry, the Biological Survey’s successoragency, the USFWS, extended the extermina-tion program south of the border and all buteliminated Mexico’s wolves by the mid-1970s.In 1976, the Mexican wolf joined four otherwolf subspecies on the U.S. EndangeredSpecies list (USFWS 2003). Only the timberwolf’s existence in the Great Lakes region wascertain at that time, whereas any wild Mexicanwolves in the Southwest were presumedwanderers from Mexico.

After the ESA listings, concern arose thatindividual wolves that could not be identified

264 Povilitis and Becker

to subspecies in the lower 48 states would lacklegal protection under the ESA. In response,the USFWS ruled that the problem could be“handled most conveniently by listing only thespecies name” (USFWS 1978); consequently,Mexican wolf and other subspecies wereconsolidated into a generalized gray wolflisting. USFWS offered its “firmest assurance”that it would continue to recognize andconserve valid subspecies of gray wolf(USFWS 1978) because the ESA calls forrecovering imperiled subspecies and popula-tions, as well as full species (USC 1973). Thelikelihood that subspecies, like the Mexicanwolf, carry ecologically relevant adaptationsand the potential to become a unique newspecies are compelling biological reasonsunderlying that mandate (O’Brien and Mayr1991).

Edward Nelson and Edward Goldman ofthe Biological Survey first identified theMexican wolf as a unique subspecies andnamed it for their fellow employee, VernonBailey (Nelson and Goldman 1929). Goldmandelineated the northern range of C. l. baileyi,approximately at today’s U.S. Interstate 10 inArizona and New Mexico (Young andGoldman 1944). Other mammalogists reaf-firmed baileyi as a distinct subspecies andconfirmed Goldman’s boundary line (Hall andKelson 1959; Nowak 1995). The uniquenessof the Mexican wolf has also been corrobo-rated by genetic studies (Wayne et al. 1992;Hedrick et al. 1997), with Garica-Moreno etal. (1996) concluding that Mexican wolves arethe “most distinct grouping of North Americanwolves supporting their distinction as anendangered subspecies.”

In 1980, biologists suggested that the twoextinct southwestern subspecies that occurredjust to the north and east of the Mexican wolf’srange (the Mogollon mountain wolf and theTexas gray wolf) could be consolidated taxo-nomically with C. l. baileyi (Bogan andMehlop 1980). They found the two mostdistinct southwestern subspecies to be baileyiand youngi (the southern Rocky Mountainwolf), and believed that the Mogollon wolf

represented an intermediate form between thetwo (Bogan and Mehlop 1983). The USFWSadopted their view (USFWS 1982) and recog-nized an area extending approximately 200miles to the north of Goldman’s boundary lineas a “zone of subspecies intergradation”(USFWS 1996). This interpretation ofMexican wolf taxonomy and geographic rangeallowed the choice of what was apparently theleast politically sensitive area in the Southwestfor a reintroduction attempt — a 17,752 squarekilometer portion of the Gila and ApacheNational Forests in east-central Arizona andwest-central New Mexico, now known as theBlue Range Wolf Recovery Area (USFWS1996).

Twenty years after the Mexican wolf wasplaced on the U.S. Endangered Species list alawsuit filed by wildlife advocates promptedUSFWS to begin releasing Mexican wolvesback into the wild from the only existingsource, a captive population. The goal for theBlue Range Wolf Recovery Area (BRWRA)was to establish a single population of at least100 wild wolves in 9 years (by 2006) (USFWS1996). This was an important initiative butwell short of the widely accepted recovery goalin conservation biology of inter-connected,multiple populations within a species’ naturalgeographic range.

To meet the requirement for multiple popu-lations, the USFWS Recovery Team for graywolves in the Southwest has suggested intro-ducing Mexican wolves well beyond theirhistoric range to the north, in northern NewMexico, northern Arizona, and Colorado(Draper 2004). Their apparent rationale is thatthese areas have larger blocks of habitat withless potential for livestock and developmentconflicts than the “Sky Island” mountainranges of southern Arizona and New Mexico.Unfortunately, this approach jeopardizes thefuture of the Mexican wolf as a subspecies inviolation of the ESA requirement to conservevalid subspecies. If introduction of theMexican wolf proceeds north, hybridizationand genetic swamping by the northern graywolf are likely. Wolves transplanted to the

Povilitis and Becker 265

Yellowstone region from western Canada haveflourished and are dispersing south, with aconfirmed report of a wolf reaching centralColorado (Gebhart 2004). Mexican wolveswould also probably lose territories whenconfronted by northern wolves, as they are onaverage 44% smaller (J. Oakleaf, USFWS,pers. comm.; USFWS 2003).

Even in the absence of northern wolves,Mexican wolves placed to the north of theBRWRA would face a different evolutionaryenvironment from their original range. Forexample, the historic northern limit of theMexican wolf’s range corresponds with amajor shift in prey. To the south, Coues white-tailed deer (Odocoileus virginianus couesi) andcollared peccary (Tayassu tajacu), both amongthe smallest of North American ungulates, areprevalent (Hall 1981). On the other hand,large-bodied elk (Cervus elaphus), abundantto the north, are absent over most of theMexican wolf’s original range (Leopold 1959;Hoffmeister 1986). Policy regarding regionalboundaries for recovery of subspecies shouldnot ignore evolutionary biology—the processby which genetic distinctiveness develops.Isolation by distance, climate, and habitat on acontinental scale, along with prey specializa-tion by wolves in different regions explainsvariability in genetic structure of wolf popula-tions in North America (Carmichael et al.2001; Geffen et al. 2004).

By 1998, when the USFWS beganreleasing Mexican wolves in the BRWRA,policy had firmly pitted wolf restorationagainst perceived limits of social and politicaltolerance (USFWS 1998; Robbins 2005). Thepolicy was and continues to be heavily influ-enced by opponents of large predator recovery.Mexican wolves that depredate livestock aretranslocated or killed. On the other hand, noreductions in livestock numbers or distribution,or changes in livestock husbandry practices arerequired to better accommodate wolves.Unlike in Yellowstone National Park andcentral Idaho, where northern gray wolveswere successfully reintroduced in the mid-1990s, the BRWRA lacks a large, core area of

livestock-free habitat where Mexican wolvesare lightly managed or left alone (USFWS1996; Bangs et al. 1998). Moreover, unlikewolf recovery programs elsewhere in thewestern U.S. (USFWS 1994), Mexican wolvesare not allowed to colonize public landsbeyond BRWRA area boundaries (USFWS1998). Finally, initial releases of captive-bornwolves were limited to a “primary recoveryzone” comprising part of the smaller Arizonaportion of the BRWRA (USFWS 1996). Thesepolicies hampered the program’s ability torelease wolves in suitable areas lackingwolves, for replacement of lost mates, or forgenetic enhancement (Bergman et al. 2004).

The USFWS (1996) anticipated that theBRWRA wolf population would grow toaround 83 wolves and 15 breeding packs bythe end of 2005 (year 8 of the program) andthat the need to release wolves from captivitywould end after 2002. In contrast, at the end of2005, the USFWS estimated that there were 35to 59 wolves in 5 breeding packs in theBRWRA (AZGFD 2006). This shortfalloccurred despite the release of 90 captivewolves, including 18 released in 2004 and2005. Removals for management purposes(58) and human-caused mortality (31) largelyexplained the high mortality of collaredwolves, averaging 64% annually from 1998 to2003 (AMOC 2005). Most removals were ofwolves moving outside the BRWRA boundary(36%) or depredating livestock (24%), andmost mortality involved illegal gunshot (61%),showing that policy constraints on the Mexicanwolf program had not deterred poaching.

Empirical data collected by USFWS biolo-gists from 1998–2003 (USFWS 2004) provideestimates of annual survival (S=0.36) andbirths per individual (B=0.44). When used ina simple linear population viability analysis(PVA): Nt + 1 = (Nt * S) + (Nt * B*S), whereNt is the initial population and Nt+1 is thepopulation a year later, it is obvious why wolfrecovery is failing. Even with an optimisticestimate of a starting population of 50 Mexicanwolves of reproductive age, the populationdeclines to near zero in 7 years without new


releases (Figure 1). While it is clear thatmortality rates are excessive because ofremovals, translocations, and illegal shootingof wolves, the cause of the low birthrate isuncertain but may be related to biologicalconditions on the recovery area (USFWS2004). Wolf pack productivity and pupmortality in the BRWRA are importantresearch areas (USFWS 2004). Still, ifmanagement policies are not changed to allowfor improved survival, a self-sustaining popu-lation of Mexican wolves will not beachievable.

In 2005, USFWS imposed additionalrestrictions on wolf recovery through the inter-agency Adaptive Management OversightCommittee (AMOC 2005), chaired by theArizona Game and Fish Department (AGFD).Additional restrictions were advanced at closedmeetings arranged by a U.S. Congressmanbetween USFWS officials and opponents ofthe Mexican wolf reintroduction program(Soussan 2005). These added restrictionsincluded a moratorium on new releases of

captive-bred wolves in 2006 (if the number ofbreeding pairs in the wild were to reach six ormore on December 31, 2005), and automaticremoval or killing of wolves known or likelyto have been involved in three livestock depre-dation incidents in a single year, regardless oftheir genetic significance to the population orany subsequent cessation of depredations.AMOC also sought authority for Arizona andNew Mexico to permanently remove wolvesthat prey on livestock or create locally unac-ceptable impacts on native ungulatepopulations once the Mexican wolf populationreached 125 individuals (AMOC 2005). Undercurrent policies, it is unlikely that the popula-tion will attain that size.

While policy makers have the technicaldata for making adaptive managementchanges, they fail to fully address what isbiologically required for Mexican wolfrecovery, apparently to minimize politicalconflict. A successful recovery program willrequire policy changes to enhance survival ofBRWRA wolves by reducing control activities

Figure 1. Population viability of 50 Mexican wolves based on empirical values for annualsurvival (S) and births per individual (B) (S = 0.36; B = 0.44) (USFWS 2004). Extinction ofthe wild population results in the absence of continued release of wolves from captivestock.


involving their capture and relocation orremoval. Recovery of the Mexican wolf willultimately depend on policy changes in favorof reestablishing the subspecies within its orig-inal range to the south.

JAGUARThe northern end of the jaguar’s geographicrange historically included much of Arizona,New Mexico, Texas, and perhaps parts ofsouthern California and Louisiana (Hall 1981;Valdez 2000). The jaguar was known by thePapago, Pima, Yavapai, Hopi, Apache, andother Native Americans of the Southwest(Daggett and Henning 1974; Brown and López2001). Bounties offered by Spanish authoritiesmay have significantly diminished jaguar pres-ence in the Southwest prior to Americansettlement (Matthiessen 1987). Later, hunting,trapping, and poisoning by hunters, ranchers,and U.S. government agents depleted jaguarnumbers, with over 60 animals reportedlytaken in Arizona and New Mexico aloneduring the last century (Brown and López2001). While jaguars were apparently extir-pated from Texas by 1948 (Nowak 1975), theanimals continue to be recorded in southernArizona and New Mexico, with two or threeidentified individuals photographed in recentyears.

As with the Mexican wolf, habitat loss andfragmentation are a threat to the jaguar(Johnson and Van Pelt 1997; USFWS 1997).In Arizona and New Mexico, the most favor-able jaguar habitat includes mountain ranges,canyons, and washes, interconnected by unde-veloped riparian areas, wash complexes, andmesquite grasslands (Sierra Institute 2000).Jaguar habitat is jeopardized by land develop-ment, road construction, and depletion ofsprings and surface waters, especially alongthe base of mountain ranges, major washcomplexes, and riparian areas.

In response to the killing of a jaguar by anArizona rancher, a petition was filed in 1992to list the jaguar as endangered in the UnitedStates (Povilitis 1992). The USFWS had listedthe jaguar 20 years earlier, but protection

applied only to animals south of the U.S.border (USFWS 1972). In 1979 the USFWSacknowledged the oversight and its intentionto rectify the matter “as quickly as possible”(USFWS 1979). Two years after receiving the1992 petition, the agency issued a proposedrule for U.S. listing of the jaguar, and, uponlegal action taken by the Center for BiologicalDiversity and a subsequent U.S. District Courtorder, completed the listing process in 1997(USFWS 1997). However, the USFWS did notdesignate critical habitat, arguing that identifi-cation of critical habitat would likely makejaguars more vulnerable to illegal killing, andsuggesting that critical habitat for the speciesis not determinable. In response, the petitionersasked USFWS to reconsider its decision sincedesignation of large blocks of habitat (asopposed to localized areas) required for thiswide ranging carnivore would not aid would-be poachers. They further explained thatessential features of jaguar habitat were in factknown (i.e., concealing vegetation and topog-raphy, adequate prey base, perennial orseasonal water sources, limited risk frompoaching and human disturbance, and spatialrequirements based on body size and diet), anddemonstrated how these features could be usedto identify and map core and connectinghabitat for jaguar in southern Arizona and NewMexico (Sierra Institute 1999). In rejectingcritical habitat, the USFWS also claimed thatthe U.S. cannot be considered “essential to theconservation of the species” as “the key to thespecies’ conservation in the northern part of its[global] range lies closer to the core of thespecies range in Mexico” (USFWS 1999). Itassured that Section 7 of the ESA prohibitsfederal agencies from taking actions likely tojeopardize the continued existence of a listedspecies, including activities which impacthabitat. In contrast, the petitioners reasonedthat the “jeopardy” issue would be ignored byagencies given the jaguar’s sparse presence,that conservation of the species in the U.S. wasthe issue at hand, and that responsible federaland state agencies need to understand whichareas are of salient importance for habitat


monitoring, protection, and restoration.The USFWS decision on critical habitat

was legally challenged, and a new ruling wasrequired by a court-approved settlement agree-ment. In 2006, the USFWS again opposed theprotection of critical habitat but dropped theargument that designation was not deter-minable or prudent because it would makejaguars more vulnerable to poaching (USFWS2006). Instead, it presented a three-part argu-ment as to why critical habitat designationwould not be beneficial because “jaguarconservation does not require habitat withinthe United States”: first, they reasoned, theSouthwest currently does not have the essen-tial physical and biological features for jaguarsince there are few recent reports of theanimals and no confirmed females or cubs (anon sequitur); second, “loss of or threats to[habitat] features in the United States...is notlimiting the recovery of the species” (contraryto previous USFWS assessments of this threat[e.g., USFWS 1997]); and third, “the range ofthe jaguar in the United States is not enougharea to provide for the conservation (i.e.,recovery) of the jaguar or even make a signif-icant contribution to the conservation of thejaguar” (a position that ignores the ESA’s focuson the U.S. and the urgent need to conserve theintegrity of the greater ecosystem for the north-ernmost jaguar population that extends into theU.S. from Sonora, Mexico). With this secondcritical habitat ruling, the USFWS jacketed thejaguar in a classic Catch-22: no need forhabitat conservation because there are too fewjaguars, but without such conservation thefewer jaguars there will be. Immediately afterthe critical habitat decision, the Center forBiological Diversity filed a notice of its intentto sue the USFWS for ESA violations, namelyfailure to designate critical habitat for thejaguar and failure to develop a recovery planfor the species (Augustine 2006).

In 1997, a Jaguar Conservation Team(JAGCT) led by the AGFD was created in lieuof a federal recovery process. The JAGCTstemmed from a formal agreement by federaland state agencies and a number of counties in

southern Arizona and New Mexico to adoptthe AGFD’s jaguar conservation strategy(Johnson and Van Pelt 1997). The initiativewas conceived with the intent to convince theUSFWS that the ESA listing of the jaguar inthe United States was not needed (Shroufe1997; Brown and López 2001). The JAGCTpledged to address threats to the jaguar byproviding for conservation “consistent with theintent of the Act” (Johnson and Van Pelt 1997).JAGCT’s accomplishments include a jaguarbibliography, collaboration on research, acapture risk assessment, educational materialsand school curriculum, and reports on poten-tial habitat (Van Pelt 2006). On the other hand,JAGCT’s promise to undertake habitat conser-vation has been entirely unfulfilled. Under theconservation strategy, the team was to provide“land management cooperators with guidelinesfor assessing impacts of current and plannedactions on the jaguar and its currently knownor suspected habitat” (Johnson and Van Pelt1997). Cooperators, in turn, were to “evaluatethe potential impact on jaguars and jaguarhabitat of each new project” while JAGCTwould recommend how to address impacts andconcerns. The JAGCT would also “encouragepublic and private land managers to conserveor enhance suitable or potentially suitablejaguar habitat, including corridors connectingthose habitat blocks, to ensure that the jaguar’scurrent and future habitat needs (includingnatural dispersal and habitat expansion) areappropriately addressed.” State wildlife agen-cies would pursue formal agreements withpublic land agencies and private landownersto get the job done. These and other habitatobjectives, most slated for completion by1999, were not met.

JAGCT’s neglect to advance habitat conser-vation appears to be in deference to membersand work group participants concerned thatconcrete measures might ultimately conflictwith ranching and property rights interests. Atthe same time JAGCT existence has succeededin keeping the politically reluctant USFWSfrom developing a recovery plan (ESA Section4) for the jaguar. The prognosis for genuine


conservation through policy change is notpromising: In 2006, the AGFD and the NewMexico Department of Game and Fish(NMDGF) revised the original 1997 “Conser-vation Assessment and Strategy for the Jaguarin Arizona and New Mexico” under which theJAGCT was to operate (AGFD and NMDGF2006a). The new draft document lacked meas-urable recovery goals, failed to recognize lossor degradation of habitat as a clear threat to thejaguar in the United States, was less specificabout conservation measures, lacked timeschedules for implementation, and would onlybe implemented depending on “funding,personnel availability, and other responsibilitiesof the individual signatories.” It ignored impor-tant previous work of JAGCT’s HabitatSubcommittee (for example, on habitat conser-vation guidelines) or simply rehashed the 1997document’s call for tasks (for example, assess-ment of impacts to jaguar habitat) that were tohave been completed long ago. In short,JAGCT’s new 2006 framework was weaker andonly “strategic in nature, not a detailed imple-mentation plan” (AGFD and NMDGF 2006b)

Ironically, while touting the “metapopula-tion concept for species persistence and anecosystem approach for habitat conservation”the new JAGCT document failed to applythese key concepts despite numerous effortsby participant conservationists to encourageand assist the team in doing so. When queriedon the matter by conservationists, the respon-sible agencies, instead of outlining actions,simply removed these principles from a laterdraft of the plan (AGFD and NMDGF 2006c).

The AGFD and the NMDGF appear to holdambiguous concepts of habitat conservationfor the jaguar. For example, in responding to apublic comment in favor of keeping ranchersand farmers on the land to prevent subdivision,the agencies acknowledged “the importance ofhabitat loss (including fragmentation) throughsubdivision or other causes relative to jaguarconservation,” elsewhere noting that “habitatsmust remain sufficiently intact (and barrier-free) and otherwise suitable for jaguars...toallow them to travel and sustain themselves in

Arizona and New Mexico” (AGFD andNMDGF 2006b). In contrast, in replying tocriticism that JAGCT had not delivered aspromised on habitat conservation, the agenciesasserted that “habitat protection or enhance-ment [in Arizona and New Mexico] is notneeded for jaguars,” and that as “evidenced bypresence of mountain lions and bears...habitatand prey base exist in sufficient quantity andquality to sustain carnivores, including thejaguar” (AGFD and NMDGF 2006b). Webelieve these contradictory statements by statewildlife agencies on jaguar habitat and itsconservation reflect pressures to present polit-ically “balanced,” as opposed to scientificallycredible, assessments.

By April 2006, livestock and property rightsinterests had taken even greater control of theJAGCT when the team allowed a large numberof sympathetic state soil and water conserva-tion districts to vote on important habitatdecisions along with the smaller pre-existingslate of members, mostly representing higherlevels of government (e.g., state wildlife andagricultural departments, and regional or stateoffices of federal natural resource agencies).An outcome was de facto exclusion of poten-tial habitat areas in New Mexico fromJAGCT’s program. In response to a conserva-tionist’s concern about such “block voting,”the state wildlife agencies conceded the issue“might warrant further discussion and consid-eration” (AGFD and NMDGF 2006b).

BUREAUCRATIC REFORMHistorically, the Mexican wolf and the jaguarwere persecuted under institutionalized anti-predator campaigns sponsored by the U.S.Government for the benefit of the livestockindustry. Both species ultimately required legalaction from private entities to prompt agencymeasures under the ESA. Requiring largeblocks of habitat configured for movement anddispersal, large carnivores face a regional land-scape increasingly rendered unfavorable tothem by economic activities and development.

Public wildlife agency staff understandfully that current human population growth


and development patterns in the states in ques-tion, and that a host of related stressors onwildlife will lead to further habitat loss anddegradation, declines in water availability, andincreased disturbance and displacement (e.g.,AGFD 2005). Mexican wolf and jaguarrecovery cannot be achieved until federal andstate wildlife agencies enact policy andprogrammatic changes (Povilitis et al. 2006)that engage them in regional-scale habitatconservation for these species. The changeswould involve collaborative planning for landdevelopment, highway design and modifica-tion (for safe wildlife passage), andinternational border security in order to protectkey habitat areas and connecting corridors. Aslead government agencies responsible forwildlife recovery, the USFWS, the AGFD, andthe NMDGF would promote primacy or co-equal status for endangered carnivores onpublic lands (e.g., reduction of livestock androad densities where needed), and providecandid assessments of on-the-ground recoveryprogress, or lack thereof. Politically-basedsubstitutes for habitat conservation measureswould be avoided. For example, repeated callsfor more research on jaguars and their habitat(Johnson and Van Pelt 1997; AGFD andNMDGF 2006a) in lieu of on-the-groundhabitat conservation measures do not advancespecies recovery. Lastly, we believe thatagency officials would accomplish betterrecovery by avoiding concessions to economicinterests that undermine the effectiveness ofhabitat conservation programs.

How can agency professionals end bureau-cratic imperilment of endangered carnivoreswhile at the same time protecting themselvesfrom political retaliation? First, they can callupon professional organizations for support,favorable publicity, and, if need be, whistle-blowing protection. Examples of theseorganizations include the Society for Conser-vation Biology, a group concerned with publicpolicy as well as conservation science (Meffeeet al. 2006), and the Public Employees forEnvironmental Responsibility, a nationalalliance of scientists, law enforcement officers,

land managers, and other professionals dedi-cated to upholding environmental laws andvalues. Second, they can work closely withfederal, state, and local political leaders whosupport the ESA and other wildlife laws andwho would exert their influence, when needed,on behalf of those who steadfastly work toimplement those laws. Third, wildlife officialscan employ professional facilitators to leadstakeholder meetings to avoid direct politicalattacks, improve ad hoc decision making, andensure that meetings stay focused on speciesrecovery and the means to achieve it.

Many individuals and private conservationorganizations have fought arduously forMexican wolf and jaguar recovery. Neverthe-less, they have had little success. A stronger,more unified public voice to overcome bureau-cratic obstacles to wildlife recovery is neededto accomplish real recovery. We suggest thecreation of carnivore-recovery oversightgroups (CROGS) whose mission would be topromote ecosystem-based recovery planning,to publicly expose any agency misdoings, andto garner the political support needed toadvance species recovery. CROGS wouldcommission scientifically-based (and peer-reviewed) recovery plans that would setprogram standards by way of example and beready in anticipation of an end to (or at least alessening of) bureaucratic obstructionism andan improved political climate for speciesconservation. To stimulate public interest andconcern, CROG leaders would discussMexican wolf and jaguar conservation throughpopular articles, editorials, talk radio, newsreleases, and other media and would engageindividuals in state and federal wildlife agen-cies and elected officials in respectful dialogueor joint discussions.

As federal and state agencies have increas-ingly embraced community-based initiatives(Brunner et al. 2002), the ability of localeconomic and political interests to overrideconservation science has grown proportionally,undermining in many instances the greaterpublic interest in recovering endangeredwildlife. Endangered species management can


be co-opted by stakeholders for private inter-ests (Becker 2002). Being “caught in themiddle” of what appears to be a democraticprocess of opposing values and goals canprevent wildlife agencies from attaining theirstated conservation goals. Agency officialshave an ethical and professional obligation notto abandon species recovery objectives, and toengage stakeholders in genuine community-based initiatives to fulfill the nation’scommitment to endangered species recovery.

LITERATURE CITEDAdaptive Management Oversight Committee

(AMOC). 2005. Mexican Wolf Blue Range Rein-troduction Project. Arizona Game and FishDepartment, Phoenix, AZ.

Arizona Game and Fish Department. 2005.Arizona’s comprehensive wildlife conservationstrategy: State of the State (Companion DocumentB), Phoenix, AZ.

Arizona Game and Fish Department. 2006.Mexican wolf project holds public meetings todiscuss 2006 and beyond. Endangered SpeciesUpdates. January 23, Phoenix, AZ.

Arizona Game and Fish Department and NewMexico Department of Game and Fish. 2006a.Conservation assessment and framework for thejaguar in Arizona, New Mexico, and northernMexico. April 13 draft. Arizona Game and FishDepartment, Phoenix, AZ.

Arizona Game and Fish Department and NewMexico Department of Game and Fish. 2006b.Responses to public comment on the draft Conser-vation Assessment and Framework for the jaguarin Arizona, New Mexico, and northern Mexico.Arizona Game and Fish Department, Phoenix,AZ.

Arizona Game and Fish Department and NewMexico Department of Game and Fish. 2006c.Jaguar conservation assessment and frameworkfor Arizona, New Mexico, and northern Mexico.May 31 draft. Arizona Game and Fish Depart-ment, Phoenix, AZ.

Augustine, J. 2006. Notice of intent to sue for viola-tions of the Endangered Species Act; failure todesignate critical habitat for the jaguar; failure tocomplete a recovery plan for the jaguar. Letter tothe U.S. Fish and Wildlife Service. July 13. South-west Center for Biological Diversity, SanFrancisco, CA.

Bangs, E. E., S. H. Fritts, J. A. Fontaine, D. W.Smith, K. M. Murphy, C. M. Mack, and C. C.Niemeyer. 1998. Status of gray wolf restoration inMontana, Idaho, and Wyoming. Wildlife SocietyBulletin 26: 785–798.

Becker, C. D. 2002. Finding common ground:

Governance and natural resources in the AmericanWest (book review). Wildlife Society Bulletin 31:909–913.

Bergman, D., C. Buchanan, C. Dale, C. Hayes, T.Johnson, and W. Murphy. 2004. Mexican GrayWolf Reintroduction Project. Five-year reviewdraft. Administrative Component. Mexican WolfAdaptive Management Oversight Committee.U.S. Fish and Wildlife Service, Albuquerque, NM.

Bogan, M. A., and P. Mehlhop. 1980. Systematicrelationships of gray wolves in southwesternNorth America, National Fish and Wildlife Labo-ratory, Washington, and University of NewMexico, Albuquerque, NM.

Bogan, M. A., and P. Mehlhop. 1983. Systematicrelationships of gray wolves (Canis lupus) insouthwestern North America. Occasional PaperNo. 1. Museum of Southwestern Biology, Univer-sity of New Mexico, Albuquerque, NM.

Brown, D. (ed.). 1983. The wolf in the Southwest.The University of Arizona Press, Tucson, AZ.

Brown, D. E., and C. A. López G. 2001. BorderlandJaguars. The University of Utah Press, Salt LakeCity, UT.

Brunner, R., C. Colburn, C. Cromley, R. Klein, andE. Olson. 2002. Finding Common Ground: Gover-nance and Natural Resources in the AmericanWest. Yale University Press. New Haven, CT.

Carmichael, L. E., J. A. Nagy, N. C. Larter, and C.Strobeck. 2001. Prey specialization may influencepatterns of gene flow of wolves of the CanadianNorthwest. Molecular Ecology 10: 2787–2798.

Dagget, P. M., and D. R. Henning. 1974. The jaguarin North America. American Antiquity 39(3):465–469.

Draper, E. 2004. Gray wolves lope toward Coloradorelease. The Denver Post. 13 December 2004.

Garcia-Moreno, J., M. D. Matocq, M. S. Roy, E.Geffen, and R. K. Wayne. 1996. Relationships andgenetic purity of the endangered Mexican wolfbased on analysis of microsatellite loci. Conser-vation Biology 10: 376-389.

Gebhart, R. 2004. When wolves arrive: Manage-ment team ponders plan for established packs.Craig Daily Press (Craig, CO). 29 October 2004.

Geffen, E., M. J. Anderson, and R. K. Wayne. 2004.Climate and habitat barriers to dispersal in thehighly mobile gray wolf. Molecular Ecology 13:2481–2490.

Hall, E. R. 1981. The mammals of North America,2nd edition. John Wiley & Sons, New York, NY.

Hall, E. R., and K. R. Kelson. 1959. The Mammalsof North America. The Ronald Press. New York,NY.

Hedrick, P. W., P. S. Miller, E. Geffen, and R. K.Wayne. 1997. Genetic evaluation of three captiveMexican wolf lineages. Zoo Biology 16: 47–69.

Hoffmeister, D. F. 1986. Mammals of Arizona. TheUniversity of Arizona Press, Tucson, and ArizonaGame and Fish Department, Phoenix, AZ.

Johnson, T. B., and W. E. Van Pelt. 1997. Conser-


vation assessment and strategy for the jaguar inArizona and New Mexico. Nongame and Endan-gered Wildlife Program Technical Report 105.Arizona Game and Fish Department, Phoenix,AZ.

Leopold, A. S. 1959. Wildlife of Mexico: The GameBirds and Mammals. University of CaliforniaPress, Berkeley, CA.

Matthiessen, P. 1987. Wildlife in America. Viking.New York, NY.

Meffe, G. K., D. Ehrenfeld, and R. F. Noss. 2006.Conservation biology at twenty. ConservationBiology 20: 595–596.

Nelson, E. W., and E. A. Goldman. 1929. A newwolf from Mexico. Journal of Mammalogy 10:165–166.

Nowak, R. M. 1975. Retreat of the jaguar. NationalParks and Conservation Magazine 17(4):430–432.

Nowak, R. M. 1995. Another look at wolftaxonomy. In Ecology and conservation of wolvesin a changing world, edited by L. N. Carbyn, S.H. Fritts, and D. R. Seip, pp. 375–398. CanadianCircumpolar Institute, Edmonton, Canada.

O’Brien, S. J. and E. Mayr. 1991. Bureaucraticmischief: Recognizing endangered species andsubspecies. Science 251: 1187–1188.

Povilitis, T. 1992. Petition to the U.S. Fish andWildlife Service to list the jaguar as endangeredin the United States. Submitted to the USFWS,Washington, D.C., by the instructor and studentsof the Sierra Institute, Southwest Program,University of California Extension, Santa Cruz,and Life Net, Montezuma, NM. July 26.

Povilitis, T., D. R. Parsons, M. Robinson, and D.Becker. 2006. The bureaucratically imperiledMexican Wolf. Conservation Biology 20(4):942–945.

Robbins, J. 2005. The look of success. Conserva-tion in Practice 6(4): 28–34.

Robinson, M. J. 2005. Predatory bureaucracy: TheExtermination of Wolves and the Transformationof the West. University Press of Colorado,Boulder, CO.

Shroufe, D. 1997. Arizona Game and Fish Depart-ment director’s letter to Arizona Game and FishCommissioners. January 15.

Sierra Institute 1999. Designation of critical habitatfor jaguar (Panthera onca) in the United States:A request to the U.S. Fish and Wildlife Service.Sierra Institute Field Studies Program in Arizona.University of California Extension, Santa Cruz,CA.

Sierra Institute. 2000. Jaguar habitat in southernArizona and New Mexico: A report to the habitatsubcommittee of the jaguar conservation team.Sierra Institute Field Studies Program in Arizona.University of California Extension, Santa Cruz,CA.

Soussan, T. 2005. U.S. proposes new restrictions onwolf releases. Albuquerque Journal. 27 April

2005.U.S. Congress. 1973. Endangered Species Act

(ESA). Washington, D.C.U.S. Fish and Wildlife Service. 1972. List of endan-

gered foreign fish and wildlife. Federal Register37(62): 6476.

U.S. Fish and Wildlife Service. 1978. Reclassifica-tion of the gray wolf in the United States andMexico, with determination of critical habitat inMichigan and Minnesota. Federal Register 43(47):9607–9615.

U.S. Fish and Wildlife Service. 1979. Endangeredand threatened wildlife and plants: U.S. popula-tions of seven endangered species. FederalRegister 44(144): 43705.

U.S. Fish and Wildlife Service. 1982. Mexican wolfrecovery plan. Albuquerque, NM.

U.S. Fish and Wildlife Service. 1994. The reintro-duction of gray wolves to Yellowstone NationalPark and central Idaho. Final EnvironmentalImpact Statement. U.S. Fish and Wildlife Service,Helena, MT.

U.S. Fish and Wildlife Service. 1996. Reintroduc-tion of the Mexican wolf within its historic rangein the southwestern United States. Final Environ-mental Impact Statement. Albuquerque, NM.

U.S. Fish and Wildlife Service. 1997. Endangeredand threatened wildlife and plants: Final rule toextend endangered status to the jaguar in theUnited States. Federal Register 62: 39147–39157.

U.S. Fish and Wildlife Service. 1998. Endangeredand threatened wildlife and plants; establishmentof a nonessential experimental population of theMexican gray wolf in Arizona and New Mexico.Final rule. Federal Register 63(7): 1752–1772.

U.S. Fish and Wildlife Service. 1999. Letter fromActing Regional Director to T. Povilitis. October1. Albuquerque, NM.

U.S. Fish and Wildlife Service. 2003. Final rule toreclassify and remove the gray wolf from the listof endangered and threatened wildlife in portionsof the conterminous United States. FederalRegister 68(62): 15804–15875.

U.S. Fish and Wildlife Service. 2004. Mexican wolfrecovery program: Progress report 7.Albuquerque, NM.

U.S. Fish and Wildlife Service. 2006. Determina-tion that designation of critical habitat is notprudent for the jaguar. Federal Register 71(133):39335–39337.

Valdez, R. 2000. Jaguar, In Ecology and Manage-ment of Large Mammals in North America, editedby S. Demarais and P. R. Krausman. Prentice-Hall, Inc., Upper Saddle River, NJ.

Van Pelt, W. E. 2006. Potential jaguar habitat inArizona and New Mexico: Summary of work andrecommendations of the Jaguar Habitat Subcom-mittee for the Jaguar Conservation Team. ArizonaGame and Fish Department, Phoenix, AZ.

Wayne, R. K., N. Lehman, M. W. Allard, and R. L.Honeycutt. 1992. Mitochondrial DNA variability


of the gray wolf: Genetic consequences of popu-lation decline and habitat fragmentation.Conservation Biology 6: 559–569.

Young, S. P., and E. A. Goldman. 1944. The Wolvesof North America. American Wildlands Institute,Washington, D.C.

CONSERVATION GENETICS OF BLACK BEARS IN THE SKY ISLANDS OF ARIZONA AND NORTHERN MEXICO

Cora Varas, Carlos A. López-González, Paul R. Krausman,and Melanie Culver

Black bears (Ursus americanus) were firstdescribed in Arizona in the early 1800s. Fromthe 1930s to the 1950s black bears were clas-sified as predatory species (Hoffmeister 1986)and some populations were nearly extirpated.From 1958 to 1968 bears were classified assmall game and received some protection. In1968 their status changed to big game(Hoffmeister 1986). From the 1950s until2001, the black bear population in Arizonawas stable (n = 2,500-3,500) (Cunningham etal. 2001; McCracken et al. 1995).

Reduction in bear numbers is of increasingconcern to their long-term survival in Arizona(LeCount and Yarchin 1990). Bear populationsthat are unintentionally reduced significantlyoften take years to recover (Miller 1990) dueto delayed reproductive maturity (firstbreeding at 3-7 years) and low reproductiverate (2 cubs every 2 to 6 years) (LeCount1982a, 1983; Kolinosky 1990).

Concerns for black bears in Arizona includeharvest numbers, anthropogenic use of landthat could threaten population connectivity,and inaccurate population estimates. From1964 to 1989 a mean of 239 bears wereharvested annually (6.8% of the maximumpopulation estimate), which increased to 368bears in 2001 (10.5% of the maximum popu-lation estimate).

Since the 1980s, the Arizona black bearseason length has been based on the numberof females harvested within a given gamemanagement unit. When the harvest objectiveof 5% of total females is reached, the bear

season is closed in that area. However, in the1970s greater than 5% females in some popu-lations were harvested (R. Olding and T.Waddell, Arizona Game and Fish Department,unpubl. data). Also, the black bear populationin east-central Arizona was over harvested inthe 1980s with 15% adult annual mortalityaffecting recruitment by reducing the breedingage of females and, therefore, reducing thenumber of cubs available for replacement(LeCount 1982b). Additionally, liberal huntingseasons in the sky islands (i.e., isolated moun-tains surrounded by desert and grasslandecosystems) combined with limited habitatavailable, have produced low populationnumbers in Coronado National Forest insouthern Arizona (R. Olding and T. Waddell,unpubl. data).

In Mexico, black bears are listed as endan-gered (Servheen et al. 1999). There are recordsof black bears in Sonora, Chihuahua, Coahuila,Nuevo León, Zacatecas, and Durango (Sierra-Corona et al. 2005). However there is notmuch information on current black bear popu-lations in Mexico. The published informationis mainly from populations in the northernSonora and Coahuila (Doan-Crider and Hell-gren 1996; Sierra-Corona et al. 2005; Onoratoet al. 2004), and bear occurrence is not welldocumented in other parts of Mexico.Although little scientific information is avail-able for Mexico, it is known that black bearshave lost at least 30% of their historical rangein Mexico (Pelton et al. 1997). The mainfactors threatening black bear survival in

276 Varas, López-González, Krausman, and Culver

northern Mexico are habitat loss and poaching(MacCracken et al. 1995); in addition the pooreconomy prevents enforcement of poachingand habitat destruction regulations. The lackof information about migration patterns andconnectivity among populations withinMexico and neighboring Arizona furtherhampers potential conservation and manage-ment efforts. To enhance conservation,biologists and managers should concentratetheir efforts on sky islands.

SKY ISLANDSPrimary habitats for black bears are coniferousand broadleaf deciduous woodlands inArizona and northern Mexico. These habitatsoccur on mountains that rise from the desertand are isolated from each other. Of necessityblack bears periodically move through thedesert lowlands to other sky islands (LeCountand Yarching 1990).

Sky islands of the desert Southwest haveproduced population isolation in many speciesoccupying the region resulting in morpholog-ical and genetic differentiation of flora andfauna. Morphological diversity has beendemonstrated in lemon lily (Linhart andPremoli 1993), snails (Bequaert and Miller1973), beetles (Ball 1966), the jumping spider(Maddison and McMahon 2000), mountainspiny lizards (Stebbins 1985), canyon treefrog(Barber 1999), and the Mt. Graham redsquirrel (Riddle et al. 1992).

Molecular genetic studies have been usedto investigate the mechanisms of sky islandisolation and how they affect population struc-ture of the species that inhabit them. Geneticdifferentiation has been studied in terms ofisolation due to biogeography barriers, distanceto the source of migrants, and sky island sizewith respect to population structure. Also,genetic analyses have been useful to estimatewhether time of speciation is concordant withisland formation. Molecular studies have notbeen reported in the literature for largemammals in the sky islands. And there is noknowledge of how sky island size, configura-tion, distance from other sky islands, and

proximity to barriers affect connectivity (geneflow) of large mammals such as black bears.

Factors, such as distance, that influence thedispersal of plants, insects, reptiles, and smallmammals could have little or no effect onblack bears due to their capacity for longdistance dispersal up to 230 km. Also, barrierssuch as rivers or patches of desert that affectsmaller species could have little effect on bearmovement. However, a combination ofdistance and unsuitable habitat (e.g., humanuse of desert lowlands including housingdevelopments in the valleys between mountainranges, recreational use of the land, agricul-tural land use, summer home developments,and highways) may cause significant barriersfor black bears (Schenk 1996) and disrupt theconnectivity among bear populations.

North of the Mogollon Rim in Arizona is anarea of continuous bear habitat that extendsfrom west of Williams southeastward to theArizona-New Mexico border. This area has thegreatest number of black bears in Arizona andthe bears appear to move large distances(Cunningham et al. 2001). This northern popu-lation could potentially be the source of bearsfor at least some of the sky islands in southernArizona and northern Mexico. We examinedthis possibility by using molecular DNA tech-niques to study the population structure ofblack bears on sky islands in Arizona (i.e.,Mogollon Rim, Four Peaks-Mt. Ord, NutriosoMountains) and Mexico (Sierra San Luis,Sierra Madre Occidental).

STUDY TECHNIQUES. PREVIOUS STUDIES

Bear populations are difficult to inventory andmonitor because the animals occur in lowdensities and are secretive by nature. A varietyof techniques have been used to obtain popu-lation numbers, density, and movementestimates for bears. Direct observation can beused to estimate small population sizes andtrends as with the brown bears in Glacier, andYellowstone National Parks (Hayward 1989).Capture-mark-recapture (Kolenosky 1986) andradio telemetry (Vashon et al. 2003) have been

Varas, López-González, Krausman, and Culver 277

the most commonly used techniques.Recently, molecular markers in combinationwith non-invasive sampling techniques haveprovided an inexpensive and efficient alternatemethod to obtain information about speciesand populations.

Where paleontological and morphologicaldata have revealed inconclusive results,molecular DNA techniques have been usefulin delineating evolutionary relationships of theeight species of bears. The giant panda is themost ancestral, followed by the spectacledbear. This was determined using six genesegments of mitochondrial DNA (Waits et al.1999). The American and Asiatic black bearare closely related, and the youngest groupincludes brown and polar bears.

During the Pleistocene, the most recentglaciation, deciduous forests occurred mainlyin eastern and western refugia in NorthAmerica. Mitochondrial DNA studies ofblack bears have confirmed the existence ofthese two refugia by identifying two majorgroups (i.e., clades): one east of the RockyMountains (including the southern RockyMountains), and another west of the RockyMountains (California and southern BritishColumbia), with an area of contact where bothclades are present in northern BritishColumbia and Alberta (Wooding and Ward1997). This suggests that, at least in part, theextant patterns of diversity in black bears isdue to post Pleistocene colonization followedby woodlands retreating to higher elevationsin the southwestern United States.

Analysis of population genetic structure inblack bears has identified evolutionary historyand level of genetic differentiation among popu-lations (Peacock et al. 2007; Robinson 2007).Genetic structure of black bear populations hasbeen examined in several studies using nuclearDNA microsatellite loci (highly variable regionsof nuclear DNA that are not usually containedwithin genes) fragment analysis and mitochon-drial DNA (maternally inherited extra-nuclearDNA) sequence analysis.

Microsatellite DNA variation has been usedin black bears to understand how fragmenta-

tion affects population structure and manage-ment decisions. For example, black bears onNewfoundland Island, Canada, had lowerlevels of genetic variation than mainland popu-lations (Paetkau and Strobeck 1994). In Florida,black bears have currently at least eight genet-ically distinct subpopulations from what oncewas a large single population (Dixon et al.2007). Louisiana black bears showed a signif-icant population differentiation between thecoastal and inland populations (Triant et al.2004). Microsatellite loci were useful to detectthe origins of black bear populations after rein-troduction programs from Minnesota andManitoba to Arkansas and Louisiana. Bearsfrom Ozark and Ouachita in Arkansas andinland Louisiana descended from reintroducedbears; whereas, bears from southeasternArkansas and coastal Louisiana were geneti-cally unique and isolated populations (Csiki etal. 2003). Mitochondrial DNA has also beenuseful in detecting population isolation, forexample, black bears in the Kenai Peninsulaand adjacent coastal populations are not closelyrelated, showing the lack of connectivitybetween the peninsula and coastal populations(Robinson et al. 2007). Finally, a lack ofconnectivity has been confirmed between theAlexander Archipelago black bears and themainland bears of southeast Alaska (Peacocket al. 2007; Stone and Cook 2000). In contrast,a lack of differentiation has been observed inone bear study. Black bears from northernSierra Madre Oriental in Mexico and westernTexas show connectivity between them viadesert corridors (Onorato et al. 2004).

Genetic analyses have been useful in exam-ining phylogenetic relationships and level ofconnectivity among bear populations. Theearliest studies using allozymes were mostlyuninformative due the little genetic variabilitydetected. Microsatellite loci and mtDNAcontrol region sequences, used more recentlyin black bear population studies, have revealedsubstantial genetic variation. Genetic data hasbeen used to develop augmentation plans inconservation planning and bear management(Waits et al. 2001).


METHODSWe collected hair, scat and tissue samples frombears in Arizona (Mogollon Rim, Four Peaks-Mt. Ord, Nutrioso Mountains) and Mexico(Sierra San Luis, Sierra Madre Occidental). Thesamples from the Mexican populations werecollected from field studies, using non-invasivetechniques. In Arizona samples collectedincluded scats, hair and tissues from huntedanimals and from other field studies thatinvolved handling of bears. We extracted DNAusing QIAGEN’s DNeasy tissue or scats kits(for instructions visit http://qiagen.com/litera-ture/genomlit.asp) and amplified mitochondrialDNA and six microsatellites using polymerasechain reaction (PCR) (Paetkau et al. 1995). Adetail of the laboratory methodology we used inthis study is described in Paetkau et al. (1998).We checked for errors and quality controlaccording to Paetkau (2003).

Mitochondrial DNAMitochondrial DNA is maternally inherited,and is very useful to detect historical evolu-tionary lineages. We amplified a 310 base pairs(bp) region of the mitochondrial DNA ControlRegion (CR). For details see Varas et al. (2006).

Nuclear MicrosatellitesMicrosatellite loci are useful to resolve recentpopulation structure (i.e., determination of thetwo alleles present for an individual at onegenetic locus). The number of alleles andheterozygosity (i.e., the state where the twoalleles at a genetic locus are different) are indi-cators of how variable the populations are. Weamplified and genotyped six microsatellites:G1A, G10B, G1D, G10C, G10M, G10X(Paetkau and Strobeck 1994, 1998) for allArizona and Mexico populations in this study.

AnalysisMitochondrial DNAWe edited and aligned sequences usingSequencher 4.6 (Gene Codes Corporation,2006) and PAUP 4.0 (Phylogenetic AnalysisUsing Parsimony; Swofford 2002), to resolvephylogenetic relationships among the

sequences. In our analyses, we includedsequences from Onorato et al. (2004), Woodingand Ward (1997) and one sequence of Ursusamericanus Kermodei (Paetkau and Strobeck1996). We performed parsimony and distanceanalysis using PAUP 4.0 (Swofford 2002).

MicrosatellitesWe tested departures from Hardy-Weinbergequilibrium and genetic differentiation inGenepop 3.4 (Raymont and Rousset 1995).The Fixation Index (FST) is an indicator ofconnectivity among populations with FST = 0indicating the populations are fully connected,and FST = 1 indicating there is no connectionbetween populations.

RESULTSMitochondrial DNA

Preliminary results show five mitochondrialDNA haplotypes (i.e., unique set of mutationswithin a region of DNA). Two haplotypes areshared with bears from western New Mexicoin roughly the same frequencies. These sametwo haplotypes were also shared with bearsfrom the Rocky Mountains (Wooding andWard 1997). The other three haplotypes wererare and found in only one population from thisstudy. No sharing was observed between haplo-types from this study and the single haplotypefound in south central New Mexico, or with thetwo haplotypes found in southern Texas and theSierra Madre Oriental in Mexico.

Nuclear MicrosatellitesThe number of alleles ranged from two to nineper locus and the heterozygosity ranged from0.17 to 0.42. Randomly mating closed popu-lations should conform to Hardy-Weinbergequilibrium, whereby observed and expectedheterozygosities do not differ significantly. Theobserved versus expected heterozygosity indi-cated our populations conformed toHardy-Weinberg expectation at four micro-satellite loci but not at two loci. Consideringall loci together, only the Four Peaks-Mt. Ordpopulation deviated from Hardy Weinbergequilibrium. The FST values in this study were


highest (0.1131) between the Mogollon Rimin Arizona and the Sierra Madre Occidental inMexico. The range in FST values was 0.0002to 0.1131 and the average FST was 0.0477.

DISCUSSIONOur mitochondrial DNA data shows that blackbears from Arizona are more closely related toblack bears in western New Mexico and alongthe Rocky Mountains (east of the RockyMountain clade) than to bears in the westerngroup (California) presented by Wooding andWard (1996). This is reasonable because thereis a geographical connection between northernArizona (Mogollon Rim) and the southernRocky Mountains, and because black bears inArizona are separated from California by along stretch of desert. Our results indicate theArizona/Mexico populations are not closelyrelated to populations further east than westernNew Mexico. This pattern of diversity likelyrepresents historical dispersal since the lastglaciation.

Black bears in the sky islands in Arizonaand in the Sierra Madre Occidental in northernMexico are closely related. MitochondrialDNA and microsatellites show the Arizonasky islands and Sierra Madre Occidental popu-lations share mtDNA haplotypes and manymicrosatellites alleles. Therefore, we couldconsider the sky island region in Arizona andnorthern Mexico one connected population interms of management and conservation.

Preliminary mitochondrial data shows thatthere is a moderate level of gene flow amongthe sky islands in Arizona and Sierra MadreOccidental in Mexico, with the highest FSTvalue being 0.1131 between the populationsin the Sierra Madre Occidental and the popu-lation on the Mogollon Rim in Arizona (thepopulations separated by the longest distance).These results suggest that the primary factorinfluencing gene flow among bear populationsis the distance between populations. There-fore, neighboring populations are lessdifferentiated than distant ones.

Black bears in the two studied Mexicanpopulations, Sierra Madre Oriental (Coahuila/

Sierra el Burro) (Onorato et al. 2007), and theSierra Madre Occidental (Sierra San Luis)(Varas et al. 2006), are not closely related. Itseems likely that these populations were histor-ically separate and have not experiencedsignificant gene flow since the last glaciation.This means that there are at least two differentblack bear lineages occurring in Mexico.

Our results indicate a connected populationof black bears in the sky islands in Arizona andthe Sierra Madre Occidental in northernMexico. Black bears are moving among thesesky islands within and between the U.S. andMexico. Management options need to considergenetic differentiation and levels of gene flowamong populations, and to strive to maintaingenetic variability of population to promotelong-term survival of wildlife. The increasedmilitarization on the border may be disruptingand reducing the movement of bears across theborder. Also, the addition of an impermeablefence across the border would stop bear migra-tion between the United States and Mexico.Arizona populations may be the only source ofmigrants to the endangered black bear popu-lation in Sonora, Mexico. Habitat connectivitybetween Texas and Mexico has alloweddispersal between populations in Coahuila,Mexico (source population) and Texas(subpopulations) (Doan-Crider and Hellgren1996, Onorato at al. 2007). As a result thisenhances the long-term viability of themetapopulation in Big Bend National Park,Texas provided that the border remains opento bear migration. Two-way movementbetween source populations and subpopula-tions is vital to the survival of the black bearin the desert Southwest.

Challenges are huge in terms of preservingthe connectivity among populations in the skyislands. This connectivity is vital so that largemammals in general have the genetic vari-ability they need to adapt to the fast changingenvironment. International cooperation, bina-tional agreements, and education of the publicare the keys to maintaining the rich biodiver-sity we have in this unique sky islandecosystem.


LITERATURE CITEDBall, G. E. 1966. The taxonomy of the subgenusScaphinotus dejean with particular reference tothe subspecies of Scaphinotus petersi Roeschke(Coleoptera: Carabidae Cychrini). Transactions ofthe American Entomological Society 92: 687–722.

Barber, P. H. 1999. Patterns of gene flow and popu-lation genetic structure in the canyon treefrog,Hyla arenicolor (cope). Molecular Ecology 8:563–576.

Csiki, I., C. Lam. A. Key, E. Coulter, J. D. Clark, R.M. Pace II, K. G. Smith, and D. D. Rhoads. 2003.Genetic variation in black bears in Arkansas andLouisiana using microsatellite DNA markers.Journal of Mammalogy 84: 691–701.

Bequaert, J. C. and W. B, Miller. 1973. Themollusks of the arid southwest. University ofArizona Press. Tucson, AZ.

Cunningham, S. C., L. M. Monroe, A. Kirkendall,C. L. Ticer. 2001.Arizona Game and Fish Depart-ment. Technical Guidance Bulletin 5.

Dixon, J. D., M. K. Oli, M. C. Wooten, T. H. Eason,J. W. McCown, M. W. Cunningham. 2007.Genetic consequences of habitat fragmentationand loss: The case of the Florida black bear (Ursusamericanus floridanus). Conservation Genetics 8:455–464.

Doan-Crider, D. L., E. C. Hellgren. 1996. Popula-tion characteristics and winter ecology of blackbears in Coahuila, Mexico. Journal of WildlifeManagement 60: 398–407.

Gene Codes Corporation. 2006. Sequencher 4.6.Ann Arbor, MI.

Hayward, G. D. 1989. Historical grizzly bear trendsin Glacier National Park: A critique. WildlifeSociety Bulletin 17: 195–197.

Hoffmeister, D. F. 1986. Mammals of Arizona.University of Arizona and the Arizona Game andFish Department, Tucson, AZ.

Kolinosky, G. B. 1986. Reproductive biology ofblack bears in east-central Ontario. InternationalConference on Bear Research and Management8: 385–392.

LeCount, A. L.1982a An Analysis of the black bearharvest in Arizona (1968–1978) Arizona Gameand Fish Department Wildlife Bulletin. SpecialReport 12.

LeCount, A. L. 1982b. Characteristics of a centralArizona black bear population. Journal of WildlifeManagement 46: 861–868.

LeCount, A. L. 1983. Denning ecology of blackbears in central Arizona. International Conferenceof Bear Research Management 5: 71–78.

LeCount, A. L., and J. C. Yarchin. 1990. Black bearhabitat use in east central Arizona. Arizona Gameand Fish Department Technical Report 4.

Linhart, Y. B., and A. C. Premoli. 1993. Geneticvariation in central and disjunct populations ofLilium parryi. Canadian Journal of Botany 72: 79–85.

Maddison, W. P. and M. McMahon. 2000. Diver-gence and reticulation among montanepopulations of a jumping spider (Habronattuspugillis Griswolds). Systematic Biology 49: 400–421.

McCracken, C., K. A. Johnson, and D. Rose. 1995.Status, management, and commercialization of theAmerican black bear (Ursus americanus). TrafficUSA. Washington, D.C.

Miller, S. D. 1990. Population management of bearsin North America. International Conference ofBear Research and Management 8: 357–373.

Onorato, D. P., E. C. Hellgren, R. A. Van DenBussche, D. L. Doan-Crider, J. R. Skyles, Jr. 2007.Genetic structure of American black bears in thedesert southwest of North America: Conservationimplications for recolonization. ConservationGenetics 8: 565–576.

Paetkau, D., and C. Strobeck. 1994. Microsatelliteanalysis of genetic variation in black bear popu-lations. Molecular Ecology 3: 489–495.

Paetkau, D., G. F. Shields, and C. Strobeck. 1998.Gene flow between insular, coastal and interior

populations of brown bears in Alaska. MolecularEcology 7: 1283–1292.

Paetkau, D., and C. Strobeck. 1998. Ecologicalgenetic studies of bears using microsatelliteanalysis. Ursus 10: 299–306.

Paetkau D. 2003. An empirical of data quality inDNA-based population inventories. MolecularEcology 6: 1375–1387.

Peacock, E., M. M. Peacock, K. Titus. 2007. Blackbears in Southeast Alaska: The fate of two ancientlineages in the face of contemporary movement.Journal of Zoology 271: 445–454.

Pelton, M. R., F. VanManen, A. Coley, K. Weaver,J. Pederson, T. Eason. 1994. Black bear conser-vation action plan North America. The WorldConservation Unit / Species Survival CommissionBear Specialist Group.

Raymond, M. and F. Rousset. 1995. GENEPOP1.02: Population genetics software for exact testand ecumenicism. Journal of Heredity 86: 248–249.

Riddle, B. R., T. L. Yates, T. E. Lee. 1992. Molec-ular divergence and variation in the endangeredMount Graham red squirrel.

Robinson, S. J., L. P. Waits, I. D. Martin. 2007.Evaluating population structure of black bears onthe Kenai Peninsula using mitochondrial andnuclear DNA analyses. Journal of Mammology88: 1288–1299.

Schenk, M. 1996. North American bear society—caring for Arizona’s black bears. Arizona WildlifeViews 39: 16–17.

Servheen, C., S. Herrero, B. Peyton. 1999. StatusSurvey and Conservation Action Plan. IUCN/SSCBear and Polar Bear Specialist Groups. The WorldConservation Unit, Gland, Switzerland andCambridge, England.

Sierra-Corona, R., I. A. Sáyago Vázquez, M. del C.


Silva Hurtado, C. A. López-González. 2005.Black bear abundance, habitat use, and foodhabits in the Sierra San Luis, Sonora, México. InConnecting mountain islands and desert seas:Biodiversity and management of the MadreanArchipelago II Proceedings. Tucson, AZ.

Stebbins, R. C. 1985. Western reptiles and amphib-ians. Houghton Miffin, Boston, MA.

Stone, K. D., and J. A. Cook. 2000. Phylogeog-raphy of black bears (Ursus americanus) of thePacific Northwest. Canadian Journal of Zoology78: 1218–1223.

Swofford, D. 2002. PAUP 4.0 (PhylogeneticAnalysis Using Parsimony and other methods),Sinauer Associates, Inc. Sunderland, MA.

Triant, D. A., R. M. Pace III, M. Stine. 2004. Abun-dance, genetic diversity and conservation ofLouisiana black bears (Ursus americanus lute-olus) as detected through noninvasive sampling.Conservation Genetics 5: 647–659.

Varas, C., C. A. López-González, J. Ramirez. P.Krausman, M. Culver. 2006. Population geneticsof black bears in northern Mexico. Borders,boundaries, and time scales. Procedings of theSixth Conference on Research and ResourceManagement in the Southwest Deserts edited byW. L. Halvorson, pp. 80–82.

Vashon, J. H., M. R. Vaughan, A. D. Vashon, D. D.Martin, and K. N. Nichols. 2003. An expandableradio collar for black bear cubs. Wildlife SocietyBulletin 31: 380–386.

Waits L. P., S. L. Talbot, R. H. Ward. G. F. Shields.2001. Mitochondrial DNA phylogeography of theNorth American brown bear and implications forconservation. Conservation Biology 12: 408–417.

Wooding, S., R. Ward. 1997. Phylogeography andpleistocene evolution in the North American blackbear. Molecular Biology and Evolution 14: 1096–1105.

RECOVERY EFFORTS FOR THE SONORAN PRONGHORN IN THE UNITED STATESRyan R. Wilson, Paul R. Krausman, and John R. Morgart

The goal of listing a species as endangeredunder the Endangered Species Act (ESA) is torecover that species from the threat of extinc-tion (Yoakum 2004a). An amendment to theESA in 1978 requires a recovery plan be devel-oped for all endangered species (Clark 1994),outlining the steps that are required for therecovery of the species and designating criteriafor delisting (Scott et al. 1996). Recovery plansand efforts made to recover endangeredspecies alone do not always make a difference(Scott et al. 1996), as many species withrevised recovery plans are more imperiled thanthey were when their original recovery planwas written (Tear et al. 1995). Reviewingrecovery efforts for a species is important todetermine what has worked and what has not,and to provide insight into endangered speciesrecovery that may improve the recoveryprocess for other species (Clark et al. 1994).

The Sonoran pronghorn was on the first listof endangered species in 1967 (U.S. Fish andWildlife Service 1967), six years before theenactment of the ESA. The subspecies is stilllisted as endangered and until 2003 was prob-ably more imperiled than when originallylisted. Reviewing recovery efforts for Sonoranpronghorn is now appropriate because theFinal Revised Recovery Plan for SonoranPronghorn (U.S. Fish and Wildlife Service1998) stated that if actions in the plan werecompleted successfully, then downlisting tothreatened was anticipated for Sonoran prong-horn by 2005; an action that did not happen.

In this chapter we review the history of theSonoran pronghorn as an endangered species

and outline the conservation and recoveryefforts initiated for the subspecies before andafter its listing in 1967. We limited our reviewof Sonoran pronghorn recovery efforts to thoseinitiated for the U.S. subpopulation; two othersubpopulations occur in Sonora, Mexico, andare functionally separated by a highway andagricultural developments (Arizona Game andFish Department 1981).

REASONS FOR ENDANGERMENTSonoran pronghorn were historically distrib-uted in the U.S. from the Imperial Valley,California, east to the Altar Valley, Arizona,and from near the Gila River in the north to theinternational boundary with Mexico in thesouth (Wright and deVos 1986). The currentdistribution of Sonoran pronghorn is almostentirely limited to Cabeza Prieta NationalWildlife Refuge (CPNWR), Organ PipeCactus National Monument (OPCNM), andthe Barry M. Goldwater Range (BMGR)(Hervert et al. 2000).

The Sonoran pronghorn numbers, whichlikely numbered in the thousands in the mid-19th century, declined rapidly by the end ofthat century due to over-harvest and loss ofhabitat (Yoakum 2004b). By 1907 they werealready rare along the U.S.-Mexico border(Mearns 1907). This led to the appointment ofa special game warden to patrol the interna-tional border with Mexico to protect pronghornand bighorn sheep (Ovis canadensis) frompoaching (Leopold 1959).This appointment,however, only lasted a few years and likely hadlittle impact on the population. Unlike other

284 Wilson, Krausman, and Morgart

populations of pronghorn in North America,Sonoran pronghorn did not increase innumbers during the 1900s (Yoakum 1968).Estimates (better classified as “guesstimates”because the populations were not systemati-cally sampled) of Sonoran pronghorn numbersin the U.S. during the 1900s varied between100 and 200 and never exceeded 250 (U.S.Fish and Wildlife Service 1998).

Even with hunting of Sonoran pronghornbanned for nearly 60 years (U.S. Fish andWildlife Service 1998) and the preservation oflarge areas of Sonoran pronghorn habitat(610,000 ha; Hervert et al. 2000) in the late1930s and early 1940s (Phelps 1978), thepopulation never increased, indicating otherfactors limited the population. The mostcommonly cited suggestion for the ultimatecause of the population’s endangerment is lossof habitat due to the creation of roads and otherbarriers to movement, in addition to over-grazing by livestock (U.S. Fish and WildlifeService 1982; Wright and deVos 1986; U.S.Fish and Wildlife Service 1998). The drying ofthe Gila and Sonoyta rivers in Arizona andSonora, respectively, may also havecontributed to the decline in numbers ofSonoran pronghorn (Carr 1972). Sonoranpronghorn may have used these areas duringdry periods as sources of succulent and nutri-tious forage and drinking water (ArizonaGame and Fish Department 1981), althoughmultiple waters were developed for Sonoranpronghorn in the 1950s and 1960s (Morgart etal. 2005).

RECOVERY AND CONSERVATION EFFORTS

Following listing of Sonoran pronghorn underthe Endangered Species Preservation Act in1967, the Arizona Game and Fish Departmentinitiated a study to collect biological informa-tion on the subspecies (Arizona Game and FishDepartment 1981).

The Sonoran Pronghorn Recovery Teamfirst met in 1975 (U.S. Fish and WildlifeService 1998) producing the first recovery planfor Sonoran pronghorn in 1982 (U.S. Fish and

Wildlife Service 1982). The recovery team seta recovery goal of maintaining an averagepopulation of 300 Sonoran pronghorn in theU.S. over a 5-year period. If this goal could bemet and the recovery team believed that majorthreats to the subspecies were eliminated, theU.S. Fish and Wildlife Service would considerdelisting the Sonoran pronghorn (U.S. Fish andWildlife Service 1982). Little was known,however, about the subspecies’ basic lifehistory characteristics (e.g., survival andmortality rates, home range size, seasonalmovements, habitat selection, productivity, andrecruitment estimates).

The recovery team recognized that inade-quate knowledge of methods to increase thenumbers or range of Sonoran pronghorn wasa serious problem inhibiting its recovery. Therecovery team also stated that while it could bepossible to transplant Sonoran pronghorn toother areas as a means of increasing the overallpopulation, at that time there was inadequateknowledge of suitable transplant sites, capturemethods, or numbers of animals required tosuccessfully establish a new population (U.S.Fish and Wildlife Service 1982). The 1982recovery plan did not outline a proposedmethod for reaching the recovery goal. There-fore, the objective set forth in the plan was tomaintain Sonoran pronghorn numbers untiltechniques were developed to reach therecovery goal.

Actions proposed in the 1982 recovery planto maintain Sonoran pronghorn numbersincluded: population surveys, maximize publicownership of habitat, preserve existing habitat(i.e., minimizing human disturbance and cattletrespass), determine life history, modifylimiting factors (e.g., predation, forage quan-tity and quality, and water) when they aredetermined, establish a captive breeding popu-lation for transplant stock, and reestablishSonoran pronghorn in historic habitat.

The first conservation action with thepotential to increase Sonoran pronghornnumbers was the removal of cattle on most ofthe current Sonoran pronghorn range in the late1970s and early 1980s (1978 on OPCNM,

Wilson, Krausman, and Morgart 285

1983 on CPNWR, and 1986 on BMGR;O’Gara and McCabe 2004). On ranges in goodecological condition, cattle and pronghorn donot normally compete for forage (Yoakum etal. 1996), however, on marginal pronghornhabitat (Yoakum 2004a) cattle may competewith pronghorn (Ellis 1970). Cattle may alsochange the vegetation associations so the land-scape supports fewer pronghorn (Wagner1978). Removing livestock from the currentrange of Sonoran pronghorn may have bene-fited pronghorn, however, reverting the areasto better habitat for native ungulates may takedecades (Valone et al. 2002) or may even beimpossible (Van Auken 2000).

Between the mid-1980s and 1990s, threestudies on life history characteristics ofSonoran pronghorn were conducted (Wrightand deVos 1986; Hughes 1991; Hervert et al.2000). In addition, three water developmentswere created for Sonoran pronghorn (Morgartet al. 2005), all fences were removed fromguzzlers and drinkers on CPNWR to facilitatetheir use by pronghorn, OPCNM modifiedtheir boundary fences with CPNWR to facili-tate pronghorn movements, and the firstfull-time ecologist was employed at CPNWR(U.S. Fish and Wildlife Service 1998). Variousstudies were also conducted to determine whateffects military operations on BMGR mighthave on pronghorn behavior and survival (seeKrausman et al. 2005 for a review).

In 1992, a systematic population monitoringprogram was initiated to conduct biennialpopulation surveys (Snow 1994). At the time,Sonoran pronghorn were the only endangeredmammal in Arizona that had not been inten-sively surveyed, and prior to 1992, there hadnot been a range-wide population survey(Snow 1994). Therefore, as late as 1992 thepopulation status of Sonoran pronghorn in theU.S. was not known. Prior to 1992, there hadbeen periodic attempts to estimate pronghornnumbers in the U.S., but they were not true esti-mates and therefore their reliability is unknown.Since 1992, the entire range of Sonoran prong-horn in the U.S. has been surveyed bienniallyto obtain population estimates.

In 1996, a population viability analysis(PVA) was used to model the probability ofSonoran pronghorn becoming extinct givenpopulation status and conditions present in1996 (Hosack et al. 2002). The PVA alsoexamined the sensitivity of the remainingSonoran pronghorn population to varying esti-mates of population parameters and frequencyof severe droughts. Using an estimate of 100animals in the population at the start of themodeling exercise, the probability of extinc-tion in the next 50 years was 12%. Results ofthe PVA also revealed that populations withnumbers < 100 have a 10-65% increased riskof extinction (Hosack et al. 2002). An increasein the frequency of catastrophic droughts (i.e.,severe enough to cause > 50 % mortality of thepopulation) caused greater population fluctu-ations, an increase in loss of genetic variation,and a decreased population growth rate. Moreimportantly, the PVA revealed that reducedfawn survival (i.e., < 25%) might affect thepopulation more than reduced adult survival(i.e, < 78% for males and < 90% for females)(Hosack et al. 2002).

The second Sonoran Pronghorn RecoveryPlan was written in 1998 (U.S. Fish andWildlife Service 1998) and updated therecovery criteria based on the results of thePVA (Hosack et al. 2002) and the three studieson Sonoran pronghorn life history (Wright anddeVos 1986; Hughes 1991; Hervert et al.2000). The new recovery criteria stated thatSonoran pronghorn would be considered fordownlisting when there are 300 Sonoranpronghorn in one U.S. population, and asecond population is established in the U.S.that remains stable over five years, or whennumbers are determined to be adequate tosustain a viable population (U.S. Fish andWildlife Service 1998). The 1998 recoveryplan also stated that if actions presented in theplan were successfully completed, Sonoranpronghorn were anticipated to be downlistedto threatened by 2005. The plan also acknowl-edged that significant aspects of Sonoranpronghorn life history were not known and thatthis hampered the ability to estimate a delisting


date and possibly to develop effective recoveryactions.

The 1998 recovery plan, like the one in1982, mentioned that captive breeding and thepossibility of reintroductions to areas ofhistoric range should be further investigated.The 1998 recovery plan also called for theinvestigation of habitat modification (i.e.,forage plots, water catchments, chain fruitcholla [Opuntia fuligida] establishment), land-use restrictions in areas of high pronghorn use,and further research on limiting factors.

By the beginning of 2002, none of theactions that were to be investigated in the 1998recovery plan (i.e., forage plots, captivebreeding, reintroductions, land-use restrictions)had been implemented. However, by the endof the year, many of those proposed recoveryactions were implemented or were beingimplemented because nearly 80% of theSonoran pronghorn population in the U.S.perished after a severe drought in 2002 (Brightand Hervert 2003).

Hervert et al. (2001) suggested the creationof forage enhancement plots in key areas ofSonoran pronghorn habitat to increase fawnsurvival by providing lactating females andforaging fawns access to more succulent andnutritious forage during times of the year withlimited rainfall. Since 2002, four forageenhancement plots have been established (onein 2002, three in 2005). Each of the forageenhancement plots also provides a source offree-standing water for Sonoran pronghorn(Figure 1). There are plans to expand three ofthe existing forage enhancement plots prior tosummer 2006 (Arizona Game and Fish Depart-ment, unpublished report). Additionally, the2002 drought spurred the creation of 6 emer-gency water catchments for Sonoran pronghornbetween 2003 and 2004 (Morgart et al. 2005).Two additional emergency waters were createdin February 2006 within CPNWR in the SierraPintas and Pinta sands, (Arizona Game andFish Department, unpublished report).

Following the 2002 drought, plans weremade to implement a captive-breedingprogram for Sonoran pronghorn (Arizona

Game and Fish Department 2003). The plansfor a captive-breeding facility for Sonoranpronghorn were modeled after a facility devel-oped for captive-breeding of peninsularpronghorn in Mexico (Cancino et al. 2005).The Sonoran pronghorn captive-breedingfacility (enclosure) was built in 2003 and islocated on CPNWR (Figure 1). The enclosureencompasses 260 ha and is sectioned into twohalves to manage the genetic diversity of thecaptive population. Forage enhancement plots(Hervert et al. 2001) and drinkers were createdin the enclosure (Figure 1) to enhance thenatural forage available to captive pronghornand provide water throughout the year.

Captive breeding began in early 2004when two females from Sonora (Januarycapture) and 1 male from the U.S. subpopula-tion (April capture) were captured andtransported to the enclosure. Four additionalfemales from the U.S. subpopulation werecaptured and released into the enclosure inDecember 2004. At the time of capture, ultra-sound revealed that all four females werepregnant; most with twins. By mid-March2005, all six females gave birth, increasing thetotal captive population to 17 animals.However, in July 2005, four fawns (threefemale, one male) died from unknown causes,and in November 2005, 1 adult female died ofunknown causes. On 2 December 2005,supplemental feeding of ad libitum alfalfa haywas initiated within the enclosure after captiveindividuals began to look emaciated due topoor forage conditions within the enclosure.

In December 2005, three adult femaleswere captured within the U.S. and transportedto the enclosure and placed in the then unoc-cupied section (i.e., southern half). Also,during the December 2005 captures, two adultfemales were captured and fitted with radio-collars and re-released, to assist in populationmonitoring efforts and other research. Fouradditional Sonoran pronghorn (one male, threefemale) were captured in Sonora in January2006 and transported to the southern half ofthe enclosure. Ultrasound revealed that allthree of these females were pregnant upon


their release. Approximately one month aftertheir release, one of the Mexican females in thesouthern half of the enclosure died of unknowncauses. Between February and March 2006,females in the enclosure gave birth to eightfawns; three sets of twins and two singletons(Arizona Game and Fish Department, unpub-lished report).

One of the goals of the Sonoran pronghorncaptive breeding program is to produce healthy

individuals so a second population of Sonoranpronghorn can be established in the U.S. (U.S.Fish and Wildlife Service 1998). To determinewhere a future reintroduction might occur, ahabitat evaluation study was conducted(O’Brien et al. 2005). Six areas outside of thecurrent distribution of Sonoran pronghornwere identified as potential habitat for a rein-troduced population (O’Brien et al. 2005).However, the models used in the study only

Figure 1. Location of forage enhancement plots and waters in the currently occupiedportion of the Sonoran pronghorn captive breeding enclosure on Cabeza Prieta NationalWildlife Refuge, Arizona, 2005.


contained coarse vegetation and landscapefeatures (i.e., slope, aspect, biome, distance towash, and soil type) so future ground-basedstudies should be conducted to further evaluatethe identified areas (O’Brien et al. 2005).

Another conservation effort, enacted in 2002in response to the catastrophic drought andmentioned in the 1998 recovery plan, weretemporary land-use closures on CPNWR,portions of OPCNM, and surrounding Bureauof Land Management (BLM) lands from 15March until 15 July each year to limit distur-bance from recreationists to Sonoran pronghornduring fawning. While disturbance of Sonoranpronghorn during fawning could be detrimentalto individual productivity (Phillips andAlldredge 2000), the effectiveness of thisconservation measure is likely reduced becauseof the increase in numbers of illegal immigrantsand the subsequent increase in border lawenforcement activity (Goodwin 2000). This isfurther supported by studies that documentpronghorn being more easily disturbed byrecreationists than other species of largemammals (Taylor and Knight 2003) andshowing no signs of habituation to continueddisturbance (Fairbanks and Tullous 2002).

Another action enacted to benefit Sonoranpronghorn was the retirement of the CameronGrazing Allotment on BLM land south of Ajo,Arizona, in September 2004 (T. Hughes, BLM,pers. comm.). This allotment is knownSonoran pronghorn habitat — the removal ofcattle and the subsequent removal of fencesmay allow more pronghorn to use the area.This action might also increase the number ofSonoran pronghorn that can be supported ontheir current range by increasing access toavailable habitat and allowing more flexibilityin responding to seasonal rainfall events.

DISCUSSIONRecovery efforts for the Sonoran pronghornover the last three decades have focused onstudying the subspecies’ natural history andpotential impacts of military operations(Krausman et al. 2004, 2005); little habitatmanipulation to benefit Sonoran pronghorn

occurred until recently (U.S. Fish and WildlifeService 1998). O’Gara and McCabe (2004)suggested that listing as endangered under theESA has not hastened the recovery of theSonoran pronghorn. To effectively conserve anendangered species, reasons why the speciesis imperiled and what factors contribute to thisimperilment must first be determined (Scott etal. 1996), which requires knowledge of aspecies’ natural history.

Biologists and managers lacked informa-tion on basic life history characteristics ofSonoran pronghorn until the studies of Wrightand deVos (1986), Hughes (1991), and Hervertet al. (2000) were completed. Estimates ofsurvival and mortality rates and of productivityand recruitment are important for endangeredspecies management because they allow biol-ogists to determine potential factors limitingpopulation growth. Biologists can then developstrategies to increase survival and recruitmentto stimulate population growth even if thelimiting factors are only proximate causes ofthe species’ endangered status (Mills et al.2005). Knowledge of home range size,seasonal movements, and habitat use are alsoneeded for effective management of endan-gered species because they identify theminimum area needed to maintain an indi-vidual, habitat requirements, and importantareas that need to be protected for survival ofthe species (Hervert et al. 2005). This infor-mation can then be used to more effectivelyimplement habitat management by consideringhabitat preferences of the species (Hervert etal. 2005), and to find potential habitat forfuture reintroductions.

Knowledge of a species’ natural history,however, will not facilitate recovery unlessconcomitant recovery actions can minimize oreliminate limiting factors. Implementation ofrecovery actions is probably the most chal-lenging part of the recovery process (Culbertand Blaire 1989). Until these basic life historydata were known, efforts to manage the prox-imate factors of Sonoran pronghornendangerment were not suggested and imple-mented. Both the 1982 and 1998 Sonoran


pronghorn recovery plans discussed furtherresearch into implementing habitat manage-ment actions and captive breeding (U.S. Fishand Wildlife Service 1982, 1998), howeverthese actions were not initiated until the middleof a severe drought, during which there was an80% reduction of an already small population(Bright and Hervert 2003).

Prior to 2002, much had been said aboutpotential negative impacts of a severe droughton the remaining Sonoran pronghorn in theU.S.. The results of the 1996 PVA (Hosack etal. 2002) suggested that an increased frequencyof catastrophic droughts increased the proba-bility of extinction over the next 100 years by46%. It was, therefore, recommended thatmanagement actions that reduce the impactsof drought on a population be implemented(provisioning of food and water) to reduce thechances of the population going extinct(Hosack et al. 2002).

Hosack et al. (2002) noted that it may alsobe beneficial to establish a captive populationto guard against the extinction of the remainingU.S. subpopulation of Sonoran pronghorn. The1998 recovery plan stated that “actions thatresult in a decrease in mortality rates for adultsand juveniles would be expected to provide themost drastic benefits for Sonoran pronghorn.”(U.S. Fish and Wildlife Service 1998:26). Anextreme drought provided the impetus for theinitiation of recovery efforts mentioned in1982, 1998, 2001, and 2002 (U.S. Fish andWildlife Service 1982, 1998; Hervert et al.2001; Hosack et al. 2002).

Forage enhancement plots and captivebreeding may provide the best tools forprotecting the remaining Sonoran pronghornin the U.S. from extinction. Vegetation manip-ulation is a common management techniquefor increasing the number of pronghorn thatcan be supported on an area (Yoakum et al.1996), but this is the first time it has beenimplemented to help increase Sonoran prong-horn numbers. Because one goal of Sonoranpronghorn recovery is to increase the popula-tion size, it is important to initiate managementactions that assure adequate forage is available

(Yoakum 2004a). One of two situationsrequiring the manipulation of habitat toincrease pronghorn numbers occurs whenfood, water, or cover are limiting factors (i.e.,forage and water in the case of Sonoran prong-horn; Fox et al. 2000) and the possibility existsfor improvement of those factors (Yoakum andO’Gara 2000). Forage enhancement plots(Hervert et al. 2001) will hopefully increasesurvival and recruitment by allowing individ-uals to meet their nutritional demands,especially during periods of drought, preg-nancy, and lactation (Fox et al. 2000; Koerthet al. 1984; Hervert et al. 2001).

Forage enhancement plots are still anexperimental management tool as there havebeen no studies that show the plots areincreasing survival and recruitment of Sonoranpronghorn. However, some information maybe gained on Sonoran pronghorn use of forageenhancement plots by the two animals radio-collared in December 2005. The size andnumber of forage enhancement plots that willbe adequate to enhance forage for the popula-tion is unknown and a study to quantify theincrease in forage quality and quantity has notbeen conducted.

Studies have indicated that supplementalfeeding of wild ungulates is either ineffectiveor detrimental to the management of thosepopulations. In a supplementally fed popula-tion of white-tailed deer (Odocoileusvirginianus), as density increased so didneonatal mortality of fawns born to two- andthree-year-old females (Ozaga and Verme1982). Also, in other studies of deer, whenlimited food is provided to starving individualsin a patchy environment, adult males usuallydominate other deer in obtaining forage(Ozaga 1972; Grenier et al. 1999). Similarobservations have been made of adult andyearling male Sonoran pronghorn in the enclo-sure chasing away females from feedingstations (Arizona Game and Fish Department,unpublished report). Supplemental feeding ofelk (Cervus elaphus) did not increase fecun-dity, but may have influenced sex ratios at birthin favor of males (Smith 2001). These studies


present possible implications of forageenhancement plots for Sonoran pronghorn.While the potential exists for forage enhance-ment plots and supplemental feeding to bepositive, they may not be effective. Therefore,a study should be conducted to determine theeffects of forage enhancement plots on foragequantity, quality, and water content. Until sucha study is conducted, forage enhancement plotsshould continue to be operated.

The Sonoran pronghorn captive breedingprogram has potential to aid in the conserva-tion and recovery of the subspecies in the U.S.In addition to serving as a source of stock forsupplementing the existing wild subpopula-tion, the program will be able to provide asource of animals for translocations to portionsof historic range. The 1998 recovery plan (U.S.Fish and Wildlife Service 1998) suggested thatthe most effective recovery effort for Sonoranpronghorn may be expanding the current rangeof Sonoran pronghorn. The 1982 recovery plan(U.S. Fish and Wildlife Service 1982) alsodiscussed translocation of Sonoran pronghornas a way to increase their numbers.

Other endangered species recoveryprograms have been successful at rearing indi-viduals in captivity for translocations toreestablish populations in historic habitat(Stüwe and Nievergelt 1991). The captivebreeding facility for peninsular pronghorn(Cancino et al. 2005) has been successful inrearing large numbers of individuals for even-tual release into historic habitat, but the releaseof animals into habitat took longer thaninitially expected. Raising Sonoran pronghornin a large enclosure in their habitat likelyincreases the chances that they will exhibitnatural behaviors once released and, therefore,will increase the chance of successful futurereintroductions, as has been demonstrated withblack-footed ferrets (Mustela nigripes) rearedin a naturalistic captive environment (Vargaset al. 1999). The Sonoran pronghorn captivebreeding facility could also be a useful tool forincreasing genetic diversity, especially afternearly 80% of the U.S. population perished in2002. This will be accomplished by capturing

and transferring Sonoran pronghorn fromMexico into the enclosure (Arizona Game andFish Department 2003).

The ability to save an endangered speciesbecomes more limited when fewer animalsexist. Tear et al. (1995) recommended thataggressive and proactive efforts need to beinitiated sooner rather than later for the conser-vation of endangered species. In the case ofSonoran pronghorn, funding for their conser-vation has recently increased, likely due to thenear extinction of the U.S. subpopulation.Additionally, the amount of research onSonoran pronghorn has increased as there weremore peer-reviewed publications on Sonoranpronghorn from 1996 to 2005 (n = 17) thanfrom 1926 to 1995 (n = 10) (Krausman et al.2005). It is important to review past recoveryefforts for Sonoran pronghorn to determinepast successes and shortcomings of therecovery program. Managers should then focuson maximizing the effectiveness of the currentrecovery efforts by investigating their efficacy(Jarman and Brock 1996) and by imple-menting future recovery efforts experimentally(Sinclair 1991). More effective recoveryefforts will aid in reaching the eventual goal ofrecovery and serve as a model for the recoveryof other threatened and endangered species.

ACKNOWLEDGMENTSWe would like to thank J. J. Hervert, J. D.Yoakum, and two anonymous reviewers forvaluable comments on an earlier version of thismanuscript. This report was funded by the U.S.Fish and Wildlife Service.

LITERATURE CITEDArizona Game and Fish Department. 1981. The

Sonoran pronghorn. Special Report Number 10.Arizona Game and Fish Department, Phoenix,AZ.

Arizona Game and Fish Department. 2003. FinalDraft: Sonoran pronghorn semi-captive breedingfacility management plan. Arizona Game and FishDepartment, Yuma, AZ.

Bright, J. L., and J. J. Hervert. 2003. Sonoran prong-horn 2002 aerial survey summary. Arizona Gameand Fish Department Nongame and EndangeredWildlife Program Technical Report 236. ArizonaGame and Fish Department, Phoenix, AZ.


Cancino, J., V. Sanchez-Sotomayor, and R. Castel-lonos. 2005. From the field: Capture, hand raising,and captive management of peninsular pronghorn.Wildlife Society Bulletin 33: 61–65.

Carr, J. N. 1972. Endangered species investigation:Sonoran pronghorn. Arizona Game and FishDepartment, Federal Aid in Wildlife RestorationProject W-53-22, Work Plan 7, Job 1.

Clark, J. A. 1994. The Endangered Species Act: Itshistory, provisions and effectiveness. In Endan-gered species recovery: Finding the lessons,improving the process, edited by T. W. Clark, R.P. Reading, and A. L. Clarke, pp. 19–43. IslandPress, Washington, D C.

Clark, J. A., R. P. Reading, and A. L. Clarke. 1994.Introduction. In Endangered species recovery:Finding the lessons, improving the process, editedby T. W. Clark, R. P. Reading, and A. L. Clarke,pp. 3–17. Island Press, Washington, D.C.

Culbert, R., and R. B. Blair. 1989. Recovery plan-ning and endangered species. Endangered SpeciesUpdate 6: 2–8.

Ellis, J. E. 1970. A computer analysis of fawnsurvival in the pronghorn antelope. Ph.D. Disser-tation, University of California, Davis, CA.

Fairbanks, W. S., and R. Tullous. 2002. Distributionof pronghorn (Antilocapra Americana Ord) onAntelope Island State Park, Utah, USA, beforeand after establishment of recreation trails. NaturalAreas Journal 22: 277–282.

Fox, L. M., P. R. Krausman, M. L. Morrison, andR. M. Katting. 2000. Water and nutrient contentof forage in Sonoran pronghorn habitat, Arizona.California Fish and Game 86: 216–232.

Goodwin, S. L. 2000. Conservation connections ina fragmented desert environment: The U.S.-Mexico border. Natural Resources Journal 40:989–1016.

Grenier, D., C. Barrette, and M. Crête. 1999. Foodaccess by white-tailed deer (Odocoileus virgini-anus) at winter feeding sites in eastern Québec.Applied Animal Behaviour Science 63: 323–337.

Hervert, J. J., J. L. Bright, M. T. Brown, L. A. Piest,and R. S. Henry. 2000. Sonoran pronghorn popu-lation monitoring: 1994–1998. Arizona Game andFish Department Technical Report 162.

Hervert, J. J., J. L. Bright, R. S. Henry, L. A. Piest,and M. T. Brown. 2005. Home-range and habitatuse patterns of Sonoran pronghorn in Arizona.Wildlife Society Bulletin 33: 8–15.

Hervert, J. J., J. L. Bright, L. A. Piest, M. T. Brown,and R. S. Henry. 2001. Sonoran pronghornrecovery: Habitat enhancements to increase fawnsurvival. Proceedings of the Pronghorn AntelopeWorkshop 19: 19–127.

Hosack, D. A., P. S. Miller, J. J. Hervert, and R. C.Lacy. 2002. A population viability analysis for theendangered Sonoran pronghorn, Antilocapraamericana sonoriensis. Mammalia 66: 207–229.

Hughes, K. S. 1991. Sonoran pronghorn use ofhabitat in southwest Arizona. M.S. Thesis, Univer-

sity of Arizona, Tucson, AZ. Jarman, P. J., and M. A. Brock. 1996. Collaboration

of science and management in endangered speciesrecovery. In Back from the brink: Refining thethreatened species recovery process, edited by S.Stephens and S. Maxwell, pp. 74–78. SurreyBeatty and Sons, Chipping Norton, Australia.

Koerth, B. H., L. J. Krysl, B. F. Sowell, and F. C.Bryant. 1984. Estimating seasonal diet quality ofpronghorn antelope from fecal analysis. Journalof Range Management 37: 560–564.

Kohm, K. A. 1991. The act’s history and frame-work. In Balancing on the brink of extinction,edited by K.A. Kolm, pp. 10–22. Island Press,Washington, D.C.

Krausman, P. R., L. K. Harris, C. L. Blasch, K. K.G. Koenen, and J. Francine. 2004. Effects of mili-tary operations on behavior and hearing ofendangered Sonoran pronghorn. Wildlife Mono-graphs 157.

Krausman, P. R., J. R. Morgart, L. K. Harris, C. S.O’Brien, J. W. Cain, III, and S. S. Rosenstock.2005. Sonoran pronghorn literature review: Anannotated bibliography. USGS Open File Report2005–1265. U.S. Geological Survey, SouthwestBiological Science Center, Sonoran DesertResearch Station, University of Arizona, Tucson,AZ.

Leopold, A. S. 1959. Wildlife of Mexico. Univer-sity of California Press, Berkeley, CA.

Mearns, E. A. 1907. Mammals of the Mexicanboundary of the United States. U.S. NationalMuseum Bulletin 56. Government Printing Office,Washington, D.C.

Mills, L. S., J. M. Scott, K. M. Strickler, and S. A.Temple. 2005. Ecology and management of smallpopulations. In Techniques for wildlife investiga-tions and management, edited by C. E. Braun, pp.691–713. The Wildlife Society, Bethesda, MD.

Morgart, J. R., J. J. Hervert, P. R. Krausman, J. L.Bright, and R. S. Henry. 2005. Sonoran pronghornuse of anthropogenic and natural water sources.Wildlife Society Bulletin 33: 51–60.

Nelson, E. W. 1925. Status of the pronghorned ante-lope, 1922–24. Department Bulletin 1346. U.S.Department of Agriculture, Washington D.C.

O’Brien, C. S., S. S. Rosenstock, J. J. Hervert, J. L.Bright, and S. R. Boe. 2005. Landscape-levelmodels of potential habitat for Sonoran prong-horn. Wildlife Society Bulletin 33: 24–34.

O’Gara, B. W., and R. E. McCabe. 2004. Fromexploitation to conservation. In Pronghornecology and management, edited by B. W. O’Garaand J. D. Yoakum, pp. 41–73. University Press ofColorado, Boulder, CO.

Ozaga, J. J. 1972. Aggressive behavior of white-tailed deer at winter cuttings. Journal of WildlifeManagement 36: 861–868.

Ozaga, J. J.and L. J. Verme. 1982. Physical andreproductive characteristics of a supplementally-fed white-tailed deer herd. Journal of Wildlife


Management 46: 281–301.Phelps, J. S. 1978. Sonoran pronghorn habitat in

Arizona. Proceedings of the Pronghorn AntelopeWorkshop 8: 70–77.

Phillips, G. E., and A. W. Alldredge. 2000. Repro-ductive success of elk following disturbance byhumans during calving. Journal of WildlifeManagement 64: 521–530.

Scott, J. M., S. A. Temple, D. L. Harlow, and M. L.Shaffer. 1996. Restoration and management ofendangered species. In Research and managementtechniques for wildlife and habitats, edited by T.A. Bookhout, pp. 531–539. The Wildlife Society,Bethesda, MD.

Sinclair, A. R. E. 1991. Science and the practice ofwildlife management. Journal of WildlifeManagement 55: 767–773.

Smith, B. L. 2001. Winter feeding of elk in westernNorth America. Journal of Wildlife Management65: 173–190.

Snow, T. K. 1994. Sonoran pronghorn aerial surveysummary 1992–1994. Technical Report 51.Arizona Game and Fish Department, Phoenix,AZ.

Stüwe, M., and B. Nievergelt. 1991. Recovery ofalpine ibex from near extinction: The result ofeffective protection, captive breeding, and rein-troductions. Applied Animal Behaviour Science29: 379–387.

Taylor, A. R., and R. L. Knight. 2003. Wildliferesponses to recreation and associated visitorperceptions. Ecological Applications 13: 951–963.

Tear, T. H., J. M. Scott, P. H. Hayward, and B. Grif-fith. 1995. Recovery plans and the EndangeredSpecies Act: Are criticisms supported by data?Conservation Biology 9: 182–195.

U.S. Fish and Wildlife Service. 1967. Notices.Federal Register 32: 4001.

U.S. Fish and Wildlife Service. 1982. Sonoranpronghorn recovery plan. U.S. Fish and WildlifeService, Denver, CO.

U.S. Fish and Wildlife Service.1998. Final revisedSonoran pronghorn recovery plan. U.S. Fish andWildlife Service, Albuquerque, NM.

Valone, T. J., M. Meyer, J. H. Brown, and R. M.Chew. 2002. Time scale of perennial grassrecovery in desertified arid grasslands followinglivestock removal. Conservation Biology 16: 995–1002.

Van Auken, O. W. 2000. Shrub invasions of NorthAmerican semi-arid grasslands. Annual Reviewof Ecology and Systematics 31: 197–215.

Vargas, A., D. Biggins, and B. Miller. 1999.Etología aplicada al manejo de especiesamenazadas: El caso del turón de patas negras(Mustela nigripes). Etologia 7: 33–39.

Wagner, F. H. 1978. Livestock grazing and the live-stock industry. In Wildlife and America:Contributions to an understanding of Americanwildlife and its conservation, edited by H. P.Brokaw, pp. 121–145. U.S. Government PrintingOffice, Washington, D.C.

Wright, R. L., and J. C. deVos, Jr. 1986. Final reporton Sonoran pronghorn status in Arizona. Contractnumber F0260483MS143. Arizona Game andFish Department, Phoenix, AZ.

Yoakum, J. D. 1968. A review of the distributionand abundance of American pronghorn antelope.Antelope States Workshop Proceedings 3: 4–14.

Yoakum, J. D. 2004a. Management plans, environ-mental impact statements and guides. InPronghorn ecology and management, edited byB.W. O’Gara and J. D. Yoakum, pp. 541–569.University Press of Colorado, Boulder, CO.

Yoakum, J. D. 2004b. Distribution and abundance.In Pronghorn ecology and management, edited byB.W. O’Gara and J. D. Yoakum, pp. 75–105.University Press of Colorado, Boulder, CO.

Yoakum, J. D. and B. W. O’Gara. 2000. Pronghorn.In Ecology and management of large mammals inNorth America, edited by S. Demarais and P. R.Krausman, pp. 559–577. Prentice Hall, UpperSaddle River, NJ.

Yoakum, J. D., B. W. O’Gara, and V. W. Howard Jr.1996. Pronghorn on western rangelands. InRangeland wildlife, edited by P. R. Krausman, pp.211–226. Society of Range Management, Denver,CO.

CURRENT STATUS OF MOUNTAIN LIONS AND URBAN ISSUES IN TUCSON, ARIZONA

Kerry L. Nicholson, Lisa Haynes, and Paul R. Krausman

Mammalian carnivores, including mountainlions (Puma concolor; Leopold 1933) that arewide ranging and exist at low densities, are ofparticular concern (Wilcox and Murphy 1985;Noss et al. 1996; Gittleman et al. 2001). Typi-cally carnivores that can live sympatricallywith humans are relatively small (10 kg) andare usually not perceived as an imminent threatto humans. Larger carnivores such as coyotes(Canis latrans), bobcats (Lynx rufus), wolves(C. lupus), black bears (Ursus americanus) andmountain lions are perceived in a negative lightby the public (Kellert 1985; Riley and Decker2000; Teel et al. 2002). This negative view,however, is changing. Public opinion that onceconsidered carnivores as vermin to be eradi-cated (Russo 1964; McCulloch 1986) isshifting to one that recognizes the ecologicalrole that mountain lions play in the wildlifecommunity (Noss et al. 1996; Teel et al. 2002).For example, mountain lions were viewed asa threat to ranching and free- ranging ungulatepopulations, thus, bounties were put on moun-tain lions along with hunting campaigns toeliminate them from the landscape (Russo1964; McCulloch 1986). Some of these viewsstill exist, but changes in tolerance levelstowards mountain lions have allowed popula-tions to increase. As a result mountain lionsincreasingly are encountering human domi-nated environments that often lead to increasesin conflict between mountain lions andhumans, their pets, and livestock. However,there is still a lack of understanding of basiclion ecology by the public, which has led to

intense political pressures often based onnarrow and erroneous assumptions (Casey etal. 2002).

Mountain lions have broad geographicdistribution, from sea level to > 4,500 m inelevation (Logan and Sweanor 2001) and havethe ability to persist in almost any habitat thatoffers adequate prey and cover (CougarManagement Guidelines Working Group2005). With increasing human expansion,mountain lions are threatened with habitat loss,fragmentation, and potential inbreeding prob-lems (Beier and Barrett 1993; Beier 1993,1995). Our objective was to review the statusof mountain lions around Tucson, Arizona andevaluate lessons scientists have learned abouturban wildlife management.

METHODSWe obtained information on other currentmountain lion studies from personal contactwith scientists. We conducted a literaturesearch of urban wildlife issues near Tucson andselected pertinent information for evaluation.

RESULTSMountain Lion-Human Conflict

Mountain lions are persecuted whenever thereis an attack on humans; attacks have increasedin the U.S. since 1985 (Beier 1991), the mostrecent occurring in April 2006 in the FlagstaffMountains, Colorado. In the U.S., Beier (1991)conducted a thorough review of historical andunprovoked attacks on humans. From 1909 to1970 there were 3 attacks from lions on

294 Nicholson, Haynes, and Krausman

humans. During 1970-1991 there were 10deaths and at least 44 nonfatal mountain lionsattacks (40% of the attacks were by sub-adultsor yearlings; Beier 1991). The rise in humanattacks in North America may be due to anumber of factors, one being fragmentation oflion habitat that increases the likelihood ofencounters with humans and pets (Mansfieldand Weaver 1989; Beier 1991; Torres et al.1996). Frequent encounters could also be dueto the increased outdoor activities of humansin lion habitat (Beier 1991; Torres et al. 1996).

Attacks by sub-adult mountain lions arecommon. Young mountain lions are learningto support themselves without assistance fromtheir mothers and are often exploring unfa-miliar territories (Seidensticker et al. 1973;Beier 1991). These stresses can lead a younglion to encounter a greater number of humans(Beier 1991; Cougar Management GuidelinesWorking Group 2005).

In March 2004 in the Sabino Canyon Recre-ation Area, a popular hiking area in northeasternTucson at the foothills of the Catalina Moun-tains, mountain lions were reportedlyencountering humans (Arizona Daily Star, 3March 2004). It was theorized that pressurefrom an extreme drought and fire in 2002,supplemental feeding of collared peccaries(Pecari tajacu), and lack of hunting pressureshad resulted in mountain lions moving into thisregion and being observed during daylight hours(Perry and deVos 2005). The resulting contro-versy over actions taken by the Arizona Gameand Fish Department (AGFD) to remove thehabituated animals was covered by local news-papers over the next 9 months (Arizona DailyStar March 2004, 14 December 2004). As aresult of these incidences AGFD initiatedresearch to inform the public of the life historyand population status of mountain lions inArizona (Perry and deVos 2005). The researchfocused on urban mountain lions in the Tucson,Payson, and Prescott areas.

The recently published Cougar Manage-ment Guidelines (2005) outlines ways toreduce the chances of negative encounters andto help agencies evaluate and respond to theseencounters. One of the first recommendations

was for people to recognize and distinguishbetween natural and dangerous mountain lionbehaviors. People are directed to avoidencounters whenever possible. Agencies andmanagers are directed to have a plan of actionprior to any encounter. These plans shouldaccount for investigating reports, level ofresponse, tolerance for specific behaviors,removal, and how to respond when a lion doesattack humans (Cougar Management Guide-lines Working Group 2005).

Effects of UrbanizationThe International Union for Conservation ofNature (IUCN) Cat Specialist Group has calledfor increased research of wild cats in urbansettings, because urbanization is one of thegravest threats to wild cat species worldwide(Nowell and Jackson 1996).

Within the past decade, researchers haveinitiated several studies of mountain lions inurbanized settings throughout the western U.S.One of the first lion studies conducted in anurban setting was in the Santa Ana Mountainsin Southern California (Beier and Barrett1993). The lions in Southern Californiaencountered a variety of anthropogenic threats,(e.g., vehicular-caused mortalities, occludedand degraded corridors) and direct encounterswith humans (Beier and Barrett 1993). Otherstudies in California have examined the influ-ences of roads and topography on lionmovements (Dickson et al. 2005), the efficacyof underpasses and wildlife crossings (L.Lyren, U.S. Geological Survey, pers. comm.),mountain lion / human interactions in Cuya-maca State Park east of San Diego after a lionattacked and killed a park visitor (Sweanor etal. 2004), the effects of rodenticide (anticoag-ulant “rat” poisoning) that are killing coyotesand bobcats (Recod and Marsh 1988; Riley etal. 2007), and habitat fragmentation andcorridor use (Sauvojot and Riley 2002; Rileyet al. 2004). Research has begun to modeleffects of urbanization on mountain lionsincluding efforts to model individual-basedmovements to evaluate connectivity for moun-tain lions (J. Tracy, Ph.D. candidate, ColoradoState University, pers. comm.), and use least-

Nicholson, Haynes, and Krausman 295

cost-path analysis to model corridors andconnectivity (Newel and Beier 2004).

Other states are also beginning to researchurbanization impacts on mountain lions. InWashington state, researchers initiated ProjectCAT, which conducts field studies of mountainlions, while incorporating students from allgrade levels (K-12) in the study and analysis(Koehler and Spencer 2004). In Utah, Stonerand Wolf (2004) examined lion movementpatterns, feeding behavior and habitat use inhigh and low density planned communities. Innorthern Arizona projects are underway withthe U.S. Geological Survey (USGS) exam-ining mountain lion movements near Flagstaff(Mattson et al. 2006).

Several analyses of gene flow and metapop-ulation genetic structure in mountain lions havebeen conducted. Molecular genetics have beenused in evaluating the impacts of urbanizationon wildlife such as determining the effect oflandscape barriers on gene flow in mountainlions. A study that encompassed Arizona, NewMexico, Colorado, and Utah found evidence ofa north-south partitioning that roughly equatedto Interstate Highway 40 (McRae et al. 2005).Patterns of genetic structure in California foundevidence of historic patterns based on naturallandscape features (Ernest et al. 2003). InWyoming, there was little evidence ofmetapopulation structure, suggesting a geneti-cally mixed, panmictic population (Andersonet al. 2004). These genetic studies provide thebasis for various geographic regions that canbe used to detect changes in the future due toanthropogenic factors.

Studies that specifically examine mountainlion movements within urban landscapes arelimited and most thus far are based on sightinglocations or locations obtained on an irregularbasis. Current research is underway usingglobal positioning systems (GPS) and satellitetechnology and obtaining locations on a morefrequent basis, but results of many of thosestudies have not been published. Many statesmaintain records of human interactions/conflicts and depredations of mountain lions;while these data are informative, additionaldata will be needed.

In Florida, urban development spread andfragmented habitat so much that the geneticviability of the Florida panther (Pumaconcolor coryi) came into question. Naturaldispersal was no longer a viable option tomaintain genetic diversity in the Floridapanther population (Land and Lacey 2000).The Florida panther was listed in the endan-gered species act in 1967 (Land and Lacey2000). Concerns began to rise over genetichealth and population size due to the small(n = 30) population (Lotz 2005). Eight femalemountain lions from Texas were translocatedto Florida for genetic restoration (Land andLacey 2000). About half of the presentlyknown occupied panther range in south Floridaoccurs on private land where agricultural andurban development are increasing (Maher1990; Cox et al. 2006). Human developmenthas effectively fenced in the Florida panther,creating a small population where extinctionnow has a high probability of occurrence.Without significant management information,the population appears to occupy all areas withsuitable panther habitat and may be close toreaching its carrying capacity (Lotz 2005).Primary mortality sources for the panther stemfrom intraspecific aggression and deaths fromvehicles (Lotz 2005). The panther populationis so isolated and restricted by developmentthat if attempts are made to disperse, theamount of pavement necessary to cross toreach more suitable habitat increases the like-lihood of collision. If the panthers choose notto disperse, they compete with each other forlimited available resources. The Floridapanther has become so isolated that manipula-tion by humans, such as future translocationsis likely the only way for the population tosurvive (Land and Lacey 2000). Is this thefuture of all mountain lions adjacent to urbancenters?

The effects of development on wildlife needto be considered if wildlife are to remain andcoexist with humans. Bighorn sheep (Oviscanadensis) provide an example of what canhappen when conservation efforts do notconsider animal movement or land needs(Bleich et al. 1996). One goal for the Pusch


Ridge Wilderness, Santa Catalina Mountains,Arizona, was to provide protected habitat fordesert bighorn sheep (Krausman 1997).Society wanted bighorn sheep, and wanted tobe able to have the experience of viewingbighorn sheep. However, during developmentof management plans, little or no thought wasgiven to corridors linking other mountains,which are a necessity for the survival of sheeppopulations (Krausman 1997). Due to contin-uing development, humans created a buildingand housing barrier around Pusch RidgeWilderness, which effectively fenced the sheepand reduced available habitat (Krausman1997). The bighorn sheep conservation goalfailed. There are no sheep in the Pusch Ridgearea and none have officially been documentedin the Catalina Mountains since the late 1990s.Now that desert bighorns have declined, wemust question the management strategies forthe area. If the mountains surrounding Tucsonhad been managed by conserving numeroushabitat patches and allowing for potentialdispersal, would there be a viable populationof bighorn sheep in the Santa Catalina Moun-tains? This is the same question that needs tobe asked for any animal that society would liketo maintain. It is a pivotal question for thesustainability of mountain lions.

Tucson Mountain Lion ResearchTucson, due to its rate of development, providesan ideal location to study mountain lion move-ments and interactions with urbanization.Located in the Sonoran Desert, Tucson is situ-ated within a valley: the Santa CatalinaMountains to the northeast, the Tortolita Moun-tains to the northwest, the Tucson Mountainsto the west, the Rincon Mountains to the east,and the Santa Rita Mountains to the south.There are several parks and wildlife reserves,including Coronado National Forest andSaguaro National Park, adjacent to the greaterTucson metropolis. Tucson is a major metro-politan area with a population > 900,000 people(> 2,500 people / 1.6 km²; http://www.fedstats.gov/qf/states/ 04/0477000. html).

In 2005, an intensive study initiated byAGFD and the University of Arizona using

GPS technology attempted to answer somecrucial questions and issues about mountainlion responses to urban situations. Some of theobjectives were to examine use of urban areas,movement rates through various landscapes,least-cost-path corridor analysis, individual-based movement models, and intraspecificinteractions of mountain lions. The first lionwas captured in May 2005. Collared andreleased, it was subsequently killed by anotherlion within a week of release. In August 2005,a second lion was captured in the back yard ofa private landowner bordering the SantaCatalina Mountains and fitted with a spreadspectrum satellite collar (Telonics, Inc., Mesa,Arizona) that recorded a location every 4.25hours. The 989 locations obtained from themale lion’s collar indicated that this mountainlion used the Coronado National Forest in theSanta Catalina Mountains, the Tortolita Moun-tains, Picacho Peak Mountains, the Ninety-sixHills and the Black Mountains, surroundingand within 60 km of Tucson. We collected datafrom 10 August 2005 until 10 April 2006. Thehome range (159,000 ha) incorporated severallarge parcels of land that are slated for devel-opment, including parcels near Biosphere IIand land between the Tortolita and Blackmountains, which has been approved fordevelopment within the next 5 years (K.Baldwin, Pima Parks and Recreation, pers.comm.). The land will have approximately5,000 new homes and businesses associatedwith new community development. The moun-tain lion’s habitat already incorporates somedeveloped areas and active living communi-ties (Figure 1). Movements between the SantaCatalina Mountains and the Tortolita Moun-tains crossed Oracle Road on numerousoccasions and were on the edge of severalexpanding communities.

DISCUSSIONHuman development and habitat fragmentationis inevitable as population numbers increase.In Arizona, three interstate highways (I-10, I-8, and I-40) cut across the state. The CentralArizona Project (CAP) is a canal system fromthe Colorado River through Phoenix and south


to Tucson forming another major barrier totranslocation. There is a project to erect abarrier between Arizona and Mexico to preventillegal immigration of humans (Arizona DailyStar, 12 March 2006). These man-made struc-tures are potential barriers to wildlife as welland could eventually stop the connectivitynecessary for wildlife persistence in the borderregion. If society wants mountain lion popula-tions to persist in Tucson there will need to bechanges in decision making and management.Managers may need conservation strategiesthat go beyond traditional land acquisition bygovernment and include economic programs topreserve critical landscapes on private land.Bighorn sheep in Pusch Ridge (Krausman1997) and the Florida panther (Land and Lacey2000; Lotz 2005) are examples of wildlife situ-ations where forethought, commitment, andfollow-through may have prevented problemsthat society now encounters. Arizona has the

ability to begin planning with the knowledgegained from the mistakes in other regions of thecountry. Currently, Pima County has initiatedthe Sonoran Desert Conservation Plan whichwill integrate natural resource protection andland use planning. Also, the Arizona WildlifeLinkage Workgroup is a collaboration of agen-cies and others to identify potential landscapecorridors around Arizona, and to developdetailed plans for some of these potential corri-dors, namely ones of high importance and athigh risk of impairment by highways, urban-ization, and other threats. Models are used todesign corridors for multiple focal species, andefforts would benefit from empirical data onhow mountain lions respond to habitat featuresin their activity and travel in Arizona land-scapes (P. Beier, Northern Arizona University,pers. comm.).

Future studies need to examine lion use ofurban areas, inter-mountain movements,

Figure 1. Mountain lion locations of a single male northeast of Tucson, Arizona, August 2005-April 2006.


human interactions and response practices.Mountain lions readily and easily movebetween several mountain ranges or acrossborders and therefore unimpeded pathwaysmust be available for persistence of thisspecies. Studies that document how lionsinteract with urbanization are important, butunless action is taken to implement the find-ings, mountain lions will get cut off byurbanization and will thereafter not have aviable future.

ACKNOWLEDGMENTSWe thank S. Sprague for the lion flights and J.deVos, T. McKinney, and G. Perry for tech-nical support. We thank J. Koprowski forsuggesting the preparation of this paper. Thisstudy was funded by Arizona Game and FishDepartment, School of Natural Resources,University of Arizona, and the Arizona Agri-culture Experimental Station.

LITERATURE CITEDAnderson, C. R., F. G. Lindzey, D. B. McDonald.

2004. Genetic structure of cougar populationsacross the Wyoming Basin: Metapopulation ormegapopulation. Journal of Mammalogy 85:1207–1214.

Beier, P. 1991. Cougar attacks on humans in theUnited States and Canada. Wildlife SocietyBulletin 19: 403–412.

Beier, P. 1993. Determining minimum habitat areasand corridors for cougars. Conservation Biology7: 94-108.

Beier, P. 1995. Dispersal of juvenile cougars in frag-mented habitat. Journal of Wildlife Management59: 228-237.

Beier, P. and R. H. Barrett. 1993. The cougar in theSanta Ana Mountain Range, California. Finalreport to the Orange County Cooperative Moun-tain Lion Study.

Belich, V. C., J. D. Wehausen, R. R. Ramey II, andJ. L. Rechel. 1996. Metapopulation theory andmountain sheep: In placation for conservation. InMetapopulation and wildlife conservation, editedby D. R. McCullough, pp. 353–373. Island Press,Washington, D.C.

Casey, A. L., P. R. Krausman, W. W. Shaw, and H.G. Shaw. 2002. Knowledge of and attitudestoward mountain lions: A public survey of resi-dents adjacent to Saguaro National Park, Arizona.Human Dimensions of Wildlife 10: 29–38.

Cougar Management Guidelines Working Group.2005. Cougar management guidelines. Firstedition. Opal Creek Press, LLC, Salem, OR.

Cox, J. J., D. S. Maehr, and J. L. Larkin. 2006.Florida panther habitat use: New approach to anold problem. Journal of Wildlife Management 70:1778–1785.

Currier, M. J. P. 1976. Characteristics of the moun-tain lion population near Canon City, Colorado.M. S. Thesis, Colorado State University, FortCollins, CO.

Dickson, B. G., J. S. Jenness, and P. Beier. 2005.Influence of vegetation, topography and roads oncougar movement in southern California. Journalof Wildlife Management 69: 264-276.

Ernest, H. B., W. M. Boyce, V. C. Bleich, B. May,S. J. Stiver, and S. G. Torres. 2003. Genetic struc-ture of mountain lion (Puma concolor)populations in California. Conservation Genetics4: 353-366.

Gittleman, J. L., S. M. Funk, D. MacDonald, and R.K. Wayne. 2001. Carnivore conservation.Cambridge University Press, Cambridge,England.

Kellert, S. R. 1985. Public perceptions of predators,particularly the wolf and coyote. BiologicalConservation 31: 169–189.

Koehler, G., and R. Spencer. 2004. Ensuring publicsafety and cougar persistence through science andeducation: Project CAT and cougar awareness. InCarnivores, edited by A. Delach, p. 3. Defendersof Wildlife Conference Abstracts, Santa Fe, NM.

Krausman, P. K. 1997. The influence of landscapescale on the management of desert bighorn sheep.In Wildlife and landscape ecology, edited by J. A.Bissonett, pp. 349–367. Springer-Verlag NewYork, Inc. New York, NY.

Land, E. D., and R. C. Lacey. 2000. Introgressionlevel achieved through Florida panther geneticrestoration. Endangered Species Update 17: 100–105.

Leopold, A. 1933. Game Management. CharlesScribner’s Sons, New York, NY.

Logan, K. A., and L. L. Sweanor. 2001. Desertpuma: Evolutionary ecology and conservation ofan endangered carnivore. Island Press, Wash-ington, D.C.

Lotz, M. A. 2005. Florida mountain lion statusreport. In Proceedings of the 8th Mountain LionWorkshop, edited by R. A. Beausoleil and D. A.Martorello, pp. 73–77 Leavenworth, WA.

Maehr D. S. 1990. The Florida panther and privatelands. Conservation Biology 4: 167–170.

Mansfield, T. M., and R. A. Weaver. 1989. Thestatus of mountain lions in California. Transac-tions of the Western Section of the WildlifeSociety 25: 1–5.

Mattson, D. J., K. L. Byrd, M. B. Rutherford, S. R.Brown and T. W. Clark. 2006. Finding commonground in large carnivore conservation: Mappingcontending perspectives. Environmental Scienceand Policy 9: 392–405.

McCulloch, C. Y. 1986. A history of predator controland deer productivity in northern Arizona. The


Southwestern Naturalist 31: 215–220.McRae, B. H., P. Beier, L. E. DeWald, L. Y. Huynh,

and P. Keim. 2005. Habitat barriers limit geneflow and illuminate historical events in a wide-ranging carnivore. Molecular Ecology 14:1965–1977.

Newel, S., and P. Beier. 2004. Designing linkagesfor pumas (and other species)—a sensitivityanalysis of GIS models. In Carnivores, edited byA. Delach, Page 35 in A. Delach, editor, p. 35.Defenders of Wildlife Conference Abstracts, SantaFe, NM.

Noss, R. F., H. B. Quigley, M. G. Hornocker, T.Merrill, and P. C. Paquet. 1996. Conservationbiology and carnivore conservation in the RockyMountains. Conservation Biology 10: 949–963.

Nowell, K. and P. Jackson. 1996. Wild Cats: Statussurvey and conservation action plan. InternationalUnion for Conservation of Nature and NaturalResources. Gland, Switzerland.

Perry, G. L., and J. C. deVos, Jr. 2005. A case studyof mountain lion-human interaction in south-eastern Arizona. In Proceedings of the 8thMountain Lion Workshop, edited by R. A. Beau-soleil and D. A. Martorello, pp. 104–113.Leavenworth, WA.

Record, C. R., and R. E. Marsh. 1988. Rodenticideresidues in animal carcasses and their relevanceto secondary hazards. In 13th Vertebrate PestConference Proceedings, pp. 163–168. Universityof Nebraska, Lincoln, NE.

Riley, S. J., and D. J. Decker. 2000. Wildlife stake-holder acceptance capacity for mountain lions inMontana. Wildlife Society Bulletin 28: 931–939.

Riley, S. P. E., C. Bromley, R. H. Poppenga, F. A.Uzal, L. Whited, and R. M. Sauvajot. 2007. Anti-coagulant exposure and notoedric mange inbobcats and mountain lions in urban southernCalifornia. Journal of Wildlife Management. 71:1874–1884.

Riley, S. P. D., E. C. York, J. A. Sikich, C. Bromley,C. M. Schoonmaker, and R. Sauvajot. 2004.

Movements, activity and food habits of mountainlions in a fragmented urban landscape in southernCalifornia. In Carnivores, edited by A. Delach, p.36. Defenders of Wildlife Conference Abstracts,Santa Fe, NM.

Russo, J. P. 1964. The Kaibab North deer herd: Itshistory, problems and management. ArizonaGame and Fish Department, Phoenix, AZ.

Sauvajot, R., and S. P. D. Riley. 2002. Mountainlions, habitat fragmentation and wildlife move-ment corridors in urbanizing landscapes insouthern California. In Carnivores, edited by K.Davies, A. Delach, L. Epstein, and M. Selden, p.69. Defenders of Wildlife Conference Abstracts,Monterey, CA.

Seidensticker, Jr., J. C., M. G. Hornocker, W. V.Wiles, and J. P. Messick. 1973. Mountain lionsocial organization in the Idaho primitive area.Wildlife Monographs 35.

Stoner, D. C., and M. L. Wolfe. 2004. Cougar-human interactions: Movement patterns andbehavior along the urban-wildland interface. InCarnivores, edited by A. Delach, p. 33. Defendersof Wildlife Conference Abstracts, Santa Fe, NM.

Sweanor, L. L., K. A. Logan, J. W. Bauer, B.Milsap, and W. M. Boyce. 2004. Puma-humaninteractions in Cuyamaca Rancho State Park,southern California. In Carnivores, edited by A.Delach, p. 4. Defenders of Wildlife ConferenceAbstracts, Santa Fe, NM.

Teel, T. R., R. S. Krannich, and R. H. Schmidt.2002. Utah stakeholders’ attitudes toward selectedcougar and black bear management practices.Wildlife Society Bulletin 30: 2–15.

Torres, S. G., T. M. Mansfield, J. E. Foley, T. Lupo,and A. Brinkhaus. 1996. Mountain lion andhuman activity in California: Testing speculations.Wildlife Society Bulletin 24: 451–460.

Wilcox, B. A., and D. D. Murphy. 1985. Conserva-tion strategy: The effects of fragmentation onextinction. American Naturalist 125: 879–887

Social & Cultural Issues

MODELING AIRBORNE MINERAL DUST: A MEXICO – UNITED STATES TRANS-BOUNDARY PERSPECTIVEDazhong Yin and William A. Sprigg

Air pollution is common in the modern world.Emissions from factories, power plants, andvehicles change the natural composition ofEarth’s atmosphere constantly. Rapideconomic development and industrializationmake the problem worse. The most harmfulcomponents of air pollution are ground-levelozone and fine airborne particulates. Ozone isa colorless, odorless gas and a powerfuloxidant. Airborne particles may consist ofdifferent chemical species such as sulfate,nitrate and toxic organics. When entering thehuman respiratory system, they react withtissues, enter the blood system, and createhealth problems (Kaiser 2005). Epidemiolog-ical studies find consistent evidence thatconcentration levels of particulate matter withaerodynamic diameter less than 10 µm and 2.5µm (PM10 and PM2.5) are associated withmortality and morbidity related to all causes,especially to cardiovascular and respiratoryillnesses (Burnett et al. 2000; Chock et al.2000; Goldberg et al. 2003; Kelsall et al. 1997;Kwon et al. 2001; Moolgavkar 2000, 2003;Ostro et al. 1999; Pope et al. 2002; Roemer andvan Wijnen 2001; Samet et al. 2000; Smith etal. 2000a, 2000b; Styer et al. 1995).

In the arid and semi-arid SouthwesternUnited States, because of readily availableerodible soil, wind-blown dust contributessignificantly to air particle pollution. Airbornedust events made up 26%, 33%, 21% and 9%of the total air pollution episodes, 47, 66, 78and 81 in Texas for 2002, 2003, 2004 and2005, respectively. Thus, it is imperative to

include wind-blown dust in the study of airpollution and its health consequences in theAmerican Southwest.

Climate variability, rapid population growthand urbanization in the Southwest create aconstantly changing landscape and vegetationpattern. These changes directly affect the reser-voirs of dust available to be lifted into theatmosphere. In this study, we considered twodust storms which occurred in the borderingstates of New Mexico and Texas, UnitedStates, and Chihuahua and Coahuila, Mexico,using a numerical dust transport model, theDust Regional Atmospheric Modeling(DREAM) system (Nickovic et al. 2001). Dustsources were identified and divided into areaswithin the U.S. and Mexico in the modeldomain. For these two dust storm cases, contri-butions from source areas in the two countriesto dust particle pollution were calculated bythe model. Model results demonstrated therelative importance of U.S. and Mexico dustsources to dust air pollution in the SouthwestU.S. and Northern Mexico. Results suggestthat coordinated efforts in ecological systemprotection and resource management areneeded in the U.S. and Mexico to controlharmful particle air pollution.

MODEL AND MODEL SETUPModel

In DREAM, a dust module, simulating dustproduction, advection, diffusion, and deposi-tion, is coupled online with the NationalCenters for Environmental Prediction (NCEP)

304 Yin and Sprigg

operational weather forecast Eta model(Messinger et al. 1988; Janjic 1994). The Etamodel uses primitive equations based on thehydrostatic approximation. It can be executedwith the finest resolution of about 10 km. It isa grid-point model, with which partial differ-ential equations are represented byfinite-difference schemes. Schemes aredesigned to fulfill computational requirementsand physical constraints of the real atmos-phere. Horizontally, the model uses asemi-staggered E grid (Arakawa and Lamb1977). Using the E grid yields good perform-ance in simulating small-scale processes suchas gravity-inertia disturbances. The methodthat provides a proper behavior of the modelwith variables on the E grid is developed forstrong physical forcing such as orographyinfluence, convection, and turbulence. Verti-cally, the model uses a step-mountainrepresentation (Mesinger et al. 1988). Thevertical coordinate of the model is defined by

where

PT is the pressure at the top of the modelatmosphere, psfc and zsfc are the pressure andthe height of the model bottom boundary, prefis a reference pressure, pressure of the stan-dard atmosphere (United States Committee onExtension to the Standard Atmosphere 1976).

The mountains in the model are represented asgrib-box mountain blocks. The non-slipbottom boundary conditions used at thevertical sides of the mountains in the modelprovide efficient simulation of the mountain’sblocking/splitting/channeling effects. Thesecond-order nonlinear advection scheme(Janjic 1984) conserves important parameterssuch as mass, energy and squared vorticity.

Vertical turbulent mixing in the surfacelayer is simulated using Monin-Obukhov simi-larity theory. The Mixing above the surfacelayer is modeled using Mellor-Yamada 2.5level turbulence scheme (Mellor and Yamada1982). A nonlinear fourth order lateral diffu-sion scheme is used to control the level ofsmall-scale noise. The radiation parameteriza-tion is based on the radiative transfer modeldeveloped at the Goddard Space Flight Center(GSFC), National Aeronautics and SpaceAdministration (NASA) (Chou et al. 1999,2001). The revised Betts-Miller deep andshallow cumulus convection scheme is used torepresent moisture processes causing exces-sive precipitation events (Betts and Miller1986; Janjic 1994). The Oregon State Univer-sity (OSU) scheme is used to model landsurface processes including surface hydrology.

Airborne dust concentrations are simulatedwith a set of mass conservation equations fordust particles of four size bins. Table 1 lists thefour dust categories in the model and theircorresponding physical properties. The dustparticles of these four categories are releasedfrom their distinctive soil components as listedin Table 1. Inter-particle interactions are notconsidered in the model.

Table 1. Four dust categories and their particle properties.

Typical particle Particle density AssociatedDust category Size bin (µm) radius (µm) (kg/m3) soil component

1 0~3.4 0.73 2500 Clay2 3.4~12 6.10 2650 Small silt3 12~28 18.00 2650 Large silt4 > 28 38.00 2650 Sand

.

Yin and Sprigg 305

The amount and location of dust particleslifted into atmosphere are determined by landcover types, soil texture types, soil moistureconditions and surface wind drag. In DREAM,the land cover product from Moderate Reso-lution Imaging Spectroradiometer (MODIS)aboard the NASA satellite Terra was used asthe land cover dataset. MODIS land cover dataused in our modeling studies identifies 17 cate-gories (Table 2) in the InternationalGeosphere-Biosphere Program vegetationclassification scheme. The classification wasproduced using a supervised approach(Hodges et al. 2001). A decision tree algorithm(Quinlan 1993) and a boosting technique(Freund et al. 1995) were used to improve clas-sification efficiency and accuracies. TheMODIS data represents land cover conditionsof 2001 with 30-second spatial resolution. Thedust source areas have land cover category 16,which is barren and sparsely vegetated. Modelsoil texture was obtained from the Food andAgriculture Organization / United NationsEducational, Scientific, and AgriculturalOrganization (http://www.fao.org/ag/agl/agll/wrb/wrbmaps/htm/soilres.htm) soil map data

with 134 categories and 2-minute spatial reso-lution.

Surface vertical dust fluxes are calculatedaccording to Shao et al. (1993) wind tunnelexperiments. Where the frictional velocity u*is greater than the threshold frictional velocityu*t, a dust flux Fs is

The threshold friction velocity depends on soilwetness and dust particle sizes of each dustcategory. When the soil moisture is less thanthe maximum amount of the absorbed waterof a soil type w',

where g is the gravitational acceleration, rp isthe particle radius, ρp the particle density, andρa the air density. When w is greater than w',

The dust flux of an individual dust category iscalculated using Fs and the fraction of desertsurface, the fractions and the dust productivityfactors of the clay/sand/silt components of thesoil texture type in a grid cell.

The dry deposition of dust particles is basedon the Georgi (1986) scheme, which includesdeposition processes by surface turbulent andBrownian diffusion, gravitational settlement,and interception and impaction on the surfaceroughness elements. The wet removal of dustparticles is calculated using the model precip-itation rate.

Model SetupThe model domain covers the Southwest U.S.and Northern Mexico. The domain center islocated at (35°N, 109°W). The grid spacing innorth-south direction is 1/9 degree of latitude.From sea level to 100 hPa, there are 24 etalayers. The approximate heights above sealevel (ABS) of the layer centers are as listed inTable 3.

Table 2. MODIS land cover categories.

MODIScategory Description

0 Water1 Evergreen Needleleaf Forest2 Evergreen Needleleaf Forest3 Deciduous Needleleaf Forest4 Deciduous Broadleaf Forest5 Mixed Forest6 Closed Shrubland7 Open Shrubland8 Woody Savannas9 Savannas

10 Grasslands11 Permanent Wetlands12 Croplands13 Urban and Built-up14 Crops, Natural vegetation Mosaic15 Permanent Snow/Ice16 Barren/Sparsely Vegetated

.

306 Yin and Sprigg

Model runs were carried out with all dustsources in the domain (referred as all sourceruns hereafter) and with no dust sources inMexico (referred as no-mx source runs).

CASESDust storm events in the Southwest U.S. areusually associated with weather patternsstarting with forming of a cold front andsurface low pressure center over the PacificOcean, west of the U.S. Pacific Northwest. Thestrong cold front system pushes southeastward,and moves through the southwestern U.S. andnorthern Mexico. During Dec 08 to Dec 10,2003 (hereafter referred as dust storm Case 1)and Dec 15 to Dec 17, 2003 (Case 2), twosimilar synoptic systems passed the SouthwestU.S.

As shown in Figure 1, the cold frontstretched from Montana through Idaho to midCalifornia at 12Z (1200 Greenwich MeanTime) Dec 14, 2003. New Mexico and Arizonawere under high pressure control. The coldfront moved out of the Southwest at 12Z Dec16, 2003 (Figure 1b). The system broughtstrong gusty winds to southern New Mexico,western Texas and northern Mexico. Figure 2shows the surface Meteorological AerodromeReport (METAR) wind gusts at 19Z Dec 15,2003. Wind gusts below 11.2 m/s (25mile/hour) are not marked in this figure. Theareas of southern New Mexico and western

Table 3. Height of 24 half eta levels.

Half eta level Height (m ABS)1 15022.832 13561.763 12257.344 11079.295 10006.646 9024.087 8120.088 7285.669 6513.69

10 5798.4211 5135.1312 4519.9213 3949.5214 3421.1815 2932.5916 2481.8117 2067.1618 1687.2819 1341.0020 1027.3821 745.6522 495.2423 275.7124 86.82

Figure 1. (a) Surface map for 12Z (1200 Greenwich Mean Time) Dec 14, 2003; (b) Surface map for 12Z Dec 16, 2003 (www.weather.unisys.com).

Yin and Sprigg 307

Texas had the strongest wind gusts, more than20 m/s.

These gusty winds caused saltation andsandblasting (Alfaro et al. 1997) over the loosesoil areas and sent dust particles into theatmosphere. Satellites observed large plumesof dust in southern New Mexico, westernTexas and northern Mexico during Case 1 andCase 2 periods. Observed visibilities in thesame area dropped to less than 2 miles. A largedust plume covered the Texas panhandle.During the Case 2 period, visibilities as low asa quarter of mile were reported at Lubbock,Texas.

EVALUATION OF MODEL RESULTSModel runs with all dust sources were verifiedagainst observations to assure the quality ofmodeled data. Modeled meteorological fieldswere compared to measurements and analysisproducts from surface synoptic, surfaceMETAR and upper-air radiosonde. Themodeled dust field patterns and dust concen-trations were evaluated against satellite imagesand the surface PM2.5 and PM10 observations

from U.S. Environmental Protection Agency(EPA) Air Quality System (AQS).

Performance statistical metrics were calcu-lated in the model evaluation in addition tographical comparisons, site against site timeseries and vertical profile comparisons. Table4 lists the statistical metrics and their defini-tions.

Since both dust storms occurred in southernNew Mexico, western Texas and northernMexico, the PM2.5 data from 40 observationalsites in New Mexico and Texas were used inthe model evaluation. Table 5 lists the latitudeand longitude of these sites. Figure 3 locatessome of the sites. Unfortunately, no PM2.5measurement data were available in northernMexico. Only limited PM10 data were avail-able in New Mexico and Texas. Sincestatistical representations require sufficientdata samples, the statistical metrics were calcu-lated for surface meteorological variables suchas surface wind speed, wind direction, andsurface temperature (data from total 758 sitesavailable), and for surface PM2.5, due to theavailable amount of data.

Figure 2. METAR wind gusts at 19Z Dec 15, 2003 (m/s).

308 Yin and Sprigg

Case 1Weather patternsFigure 4a is the surface map for 12Z Dec 09,2003. The modeled precipitation pattern andlocations (Figure 4b) are in agreement withthoseof the weather radar returns shown in Figure 4a.

The low center is located at Oklahoma and Texasin the modeling domain, so is the observed lowcenter. High pressure dominates the Arizona,Utah and Nevada area in both Figures 4a and 4b.

At 500 hPa level, the observed low geopo-tential height center (Figure 5a) is located in

Table 5. PM2.5 observational sites.

Site Site Latitude Longitudeno. name (degree) (degree)

1 St. Teresa 31.86 -106.692 SPCY 31.80 -106.563 El Paso 31.77 -106.504 El Paso Sun/Metro 31.76 -106.505 Anthony 32.00 -106.356 Santa Fe 35.67 -105.957 Carlsbad 32.31 -104.278 Odessa 47 31.84 -102.349 Odessa 1014 31.87 -102.34

10 Lubbock 33.59 -101.8511 Amarillo 35.21 -101.8312 Laredo 27.60 -99.5313 CPS Pecan Valley 29.41 -98.4314 Calaveras Lake 29.28 -98.3115 Mission 26.23 -98.2916 Audubo 30.48 -97.8717 Austin Northwest 30.35 -97.7618 Corpus Christi Airport 27.77 -97.4319 Haws Athletic Center 32.76 -97.3420 Diamond Hill Fort Worth 32.81 -97.34

Table 5. continued

Site Site Latitude Longitudeno. name (degree) (degree)21 National Sea Shore 27.43 -97.3022 Denton 33.19 -97.1923 Arlington Airport 32.66 -97.0924 Grapevine Fairway 32.98 -97.0625 Midlothian Wyatt 32.47 -97.0426 Midlothian Tower 32.44 -97.0227 Dallas Hinton St. 32.82 -96.8628 Conroe 30.35 -95.4329 Houston Aldine 29.90 -95.3330 Clinton 29.73 -95.2631 Houston 29.77 -95.2232 Kingwood 30.06 -95.1933 Houston Deer Park 29.67 -95.1334 Channel View 29.80 -95.1335 Seabrook Friendship Park 29.58 -95.0236 Galveston Airport 29.26 -94.8637 Hamshire 29.86 -94.3238 Karnack 32.67 -94.1739 Carrol St. Park 30.07 -94.0840 Thomas Jefferson 29.92 -93.91

Table 4. Statistical metrics and their definitions.

Mean modeled

Mean observed

Mean Bias

Mean error

Index of agreement

Mi modeled value at each site

Oi observed value at each site

0 if perfect

0 if perfect

1 if perfect

Yin and Sprigg 309

Figure 3. Locations of PM2.5 monitoring sites.

the Texas panhandle and bordering area ofNew Mexico and Colorado. A correspondingthermal low center is slightly west of theobserved geopotential height low, located nearthe New Mexico and Colorado border. Themodel geopotential height and temperaturefields (Figure 5c) are similar to the observa-tions and the observed low centers arereproduced by the model.

Vertical profilesThe modeled wind, temperature and specifichumidity for 12Z Dec 09, 2003 at Tucsonairport are compared against radio-sonde datain Figure 6. In the figure, the dots representobserved values, and the lines modeledvalues. Modeled temperature and specifichumidity lines follow the observations veryclosely. The modeled wind speed line followsthe trend of the observations, except for

Figure 4. (a) Surface map; (b) modeled mean sea level (MSL) pressure (hPa) andprecipitation (mm) pattern for 12Z Dec 09, 2003.

310 Yin and Sprigg

Figure 5. 500hPa (a) observed geopotential height (m); (b) observed temperature (°C); (c)modeled geopotential height (m) and temperature (°C ) fields for 12Z Dec 09, 2003.

Figure 6. Modeled and observed vertical profiles at Tucson airport (32.12°N, 110.98°W)for 12Z Dec 09, 2003 (dots: observed values; lines: modeled values)

Table 6. Performance statistics of surface winds and temperatures.

Wind speed (m/s) Wind direction (degree) Temperature (K)

Mean observed 5.36 222.29 278.39Mean modeled 4.9 221.48 278.19Mean bias -1.17 -0.80 -0.22Mean error 1.88 51.50 2.46Agreement index 0.77 0.75 0.96

Yin and Sprigg 311

several observational points that are likelyoutliers. Modeled wind directions are similarto the observations. The modeled wind direc-tions below several thousand meters showdiscrepancies against measurements, largelyan artifact of the figure and plot. Wind direc-tions more than 300 degrees are not significantlydifferent from wind directions near tens ofdegree. Model biases are reasonable.

Performance statistics of modeled surfacewinds and temperatures. The performancestatistics for modeled surface winds andtemperatures were calculated using modeledand observed values at 758 sites in themodeling domain. The number of data samplesis 27,587 in calculating surface wind statistics.

As listed in Table 6, the mean model biasesof wind speed, wind direction, and tempera-ture for Case 1 are -1.17m/s, -0.80 degree, and-0.22K. The agreement indices are 0.77, 0.75,and 0.96, respectively. These numbers indicatethat the model performed well in simulatingsurface winds and temperatures.

Dust patternsThe model dust concentration distribution for22Z Dec 09, 2003 is given in Figure 7.Compared to satellite observed dust at aboutthe same time as shown in Figure 5a, the dustplume in the western Texas area is reproducedby the model.

PM2.5 and PM10 series. The modeled andobserved PM10 and PM2.5 concentration time

Figure 7. Modeled surface layer dustconcentration distribution (µg/m3) for22Z Dec 09, 2003.

Figure 8. Modeled and observed (a) PM10 concentrations at Lubbock, Texas; (b) PM2.5concentrations at Calaveras Lake, Texas (dots: observed values; lines: modeled values).

series during Case 1 at two sites are given inFigure 8. Both modeled PM10 and PM2.5 timeseries follow the observed trends well. Themodeled PM10 and PM2.5 peak hours matchthe observed peak hours. However, themodeled peak PM10 has quite a large discrep-ancy against the measured value. It is aboutone third of the observed peak. The modelPM2.5 peak concentration is in better agree-ment with the measured one.

Performance statistics for surface PM2.5.The performance statistics of modeled PM2.5concentrations at 40 measurement sites (Table5) are listed in Table 7. The mean bias of themodeled PM2.5 is -2.72. The relative bias to

312 Yin and Sprigg

the observations is about 41%. The agreementindex of the model PM2.5 is 0.43.

Case 2Weather patternsFigure 9a is the modeled MSL pressure fieldand precipitation. The modeled high pressurecenter location is the same as shown in Figure1b. The precipitation area behind the cold frontis also captured by the model. At 500 hPa, the

modeled trough of the geopotential height field(Figure 9b) influences New Mexico, Texas andnorthern Mexico, and matches 500 hPaanalysis shown in Figure 9c. The model 500hPa cold center and thermal trough (Figure 9b)are similar to the analysis in Figure 9d.

Vertical profilesThe modeled wind, temperature and specifichumidity vertical profiles at Tucson airport for

Figure 9.(a) Modeled mean surface pressure (hPa) and precipitation (mm); (b) modeled500 hPa geopotential height (m) and temperature (°C); (c) 500 hPa geopotential height(m) analysis; (d) 500 hPa temperature analysis (°C ).

Table 7. Performance statistics for PM2.5.

Mean observed Mean modeled Mean bias Mean error Agreement(µg/m3) (µg/m3) (µg/m3) (µg/m3) index

6.65 3.93 -2.72 7.05 0.43

Yin and Sprigg 313

12Z Dec 15, 2003 are compared with theobserved profiles in Figure 10. Both modeledwind speeds and directions agree with the obser-vations. The modeled temperature and specifichumidity also match with the measured values.

Performance statistics (Table 8) are calcu-lated with modeled and observed surface windand temperature at the same measurement sitesas for Case 1. The mean biases of model windspeed, wind direction, and temperature are -1.16 m/s, -1.02 degree, and 0.72 K. Theagreement indices are good, especially that ofthe modeled surface temperature.

Figure 10. Modeled and observed vertical profiles at Tucson (32.12ºN, 110.93ºW) for 12ZDecember 15 (dots represent observed values and lines represent modeled values).

Table 8. Performance statistics of modeled surface wind and temperature.

Wind speed (m/s) Wind direction (degree) Temperature (K)Mean observed 5.53 231.40 276.74Mean modeled 4.37 230.38 277.48Mean bias -1.16 -1.02 0.72Mean error 2.03 47.85 2.67Agreement index 0.75 0.76 0.95(dimensionless)

Dust patternsThe modeled dust concentration distributionfor 20Z Dec 15, 2003 (Figure 11) is compa-rable to the dust observed by the satellite.

PM10 and PM2.5 time series. The timeseries of modeled PM10 and PM2.5 concen-trations at Cameron and Odessa, Texas followthe observed trend (Figure 12). The modeledtime series reproduced dust peaks for bothPM10 and PM2.5 concentrations during Case2. The modeled PM10 peak hours occurredseveral hours later that the observed peak. ThePM2.5 peak hours are in agreement with the

314 Yin and Sprigg

measurements. Peak PM10 and PM2.5concentrations are close to the observedvalues.

Performance statistics of modeled surfacePM2.5 concentrations. The statistics listed inTable 9 show that the agreement index of themodeled PM2.5 for Case 2 is better than thatof Case 1. The agreement index increases to0.57 from 0.43. The mean bias is -3.97 µg/m3,which is about 45% of the mean observedvalues.

CONTRIBUTIONS OF MEXICO DUST SOURCES

Changes of modeled dust concentrations werecalculated from all source model results to no-Mexico source model results. Differences andchanges of episode-average concentrations,domain-average concentrations and PM2.5concentrations at 40 measurement sites (Table5) before and after Mexican sources wereexcluded in the model domain show the contri-butions of Mexican dust sources to the dustpollution during these two dust storm events.

Case 1Episode-average concentrations in the duststorm area: The episode-average modeled dustconcentrations in the dust storm area werecalculated with all source model results andwith no-Mexico sources model results, respec-tively. DREAM outputs dust concentrations offour dust categories every hour. In the

following context, total dust includes all fourdust categories. PM10 and PM2.5 include onlyparticles with diameters less than 10 µm and2.5 µm. In the episode-average concentrationcalculation, the sum of the concentrations ofall vertical layers in each grid cell in the areawas calculated first. Then the averages of theseconcentrations using hourly model outputsfrom 00Z Dec 08 to 23Z Dec 10, 2003 werecalculated. The episode-average concentrationsfor Case 2 were computed in the same manner.

Episode-average dust concentrations aresignificantly different for all source modelresults and no-mx (no-Mexico) source results.The maximum total dust in this domaindropped from 2171.71µg/m3 to 1202.28µg/m3

Figure 12. Modeled and observed (a) PM10 concentrations at Cameron site, Texas(25.89ºN, 97.49ºW) and (b) PM2.5 concentrations at Odessa 1014.

Figure 11. Surface model dust concentrations(µg/m3) for 20Z Dec 15, 2003.

Yin and Sprigg 315

from all source model results to no-mx results.The maximum PM10 and PM2.5 changedfrom 813.57µg/m3 and 227.98µg/m3m, to400.49µg/m3 and 110.46µg/m3. The meanchanges of total dust, PM10 and PM2.5 were135.65µg/m3, 57.81µg/m3 and 16.51µg/m3.The relative concentration changes of totaldust, PM10, and PM2.5 range from 100% to0%, with mean relative concentration changesin the domain as 42.04%, 41.97% and 41.85%.

Domain-average concentrations in the duststorm area: Similar to episode-average concen-trations, the modeled concentrations of allvertical layers in each grid cell in the area wereadded together first. Then the averages of theseconcentrations in all cells of this domain werecomputed for every hour from 00Z Dec 08 to23Z Dec 10, 2003.

The maximum and minimum domain-average total dust concentration differences dueto excluding Mexican sources ranged between438.00µg/m3 and 0.005µg/m3. The maximumand minimum PM10 and PM2.5 differenceswere 161.00µg/m3 and 0.002µg/m3, 44.7µg/m3

and 0.001µg/m3. The mean difference of totaldust, PM10 and PM2.5 were 138µg/m3,58.6µg/m3, and 16.8µg/m3. The average rela-tive changes for total dust, PM10 and PM2.5were 42.8%, 43.2% and 43.2%.

PM2.5 at 40 sites: The PM2.5 concentra-tion changed it’s range from 0 to near 45µg/m3.The percentage change could exceed 90%.This means that at some sites in New Mexicoand Texas, most of the PM2.5 concentrationsarrive from sources in northern Mexico.

Case 2Episode-average concentrations in the duststorm area. As in Case1, spatial patterns of dustconcentration changed when desert dust

sources in northern Mexico were excluded inthe model. The values of the concentrationsalso changed considerably. The maximum andmean concentrations of total dust changedfrom 1320.71µg/m3 and 260.77µg/m3, to1318.71µg/m3 and 173.48µg/m3. Themaximum and mean PM10 concentrationschanged from 360.98µg/m3 and 106.68µg/m3

to 360.03µg/m3 and 67.32µg/m3. Themaximum and mean concentrations of PM2.5changed from 98.28µg/m3 and 30.58µg/m3, to97.99µg/m3 and 19.12µg/m3. The mean rela-tive change of total dust, PM10, and PM2.5concentrations were 34.51%, 33.70% and33.57%.

Different than Case 1, the maximumconcentrations of total dust, PM10 and PM2.5changed very little. This demonstrates that thesources inside the U.S. contributed most to thepeak concentrations in the domain. However,Mexican sources still contributed to nearly33% of overall concentrations in the domain.

The domain-average dust concentrationsduring Case 2 had but one peak, while those ofCase 1 had two peaks. The maximum andminimum changes in total dust concentrationdue to excluding Mexican sources were360µg/m3 and 16.3µg/m3, with an averagechange of 87.8µg/m3. Maximum and minimumPM10 and PM2.5 concentration changes were136µg/m3 and 38.4µg/m3, and 11.6µg/m3 and3.7µg/m3. The average PM10 and PM2.5concentration changes were 39.6µg/m3 and11.5µg/m3. The average relative changes fortotal dust, PM10 and PM2.5 were 42%, 43%,and 43.3%.

PM2.5 at 40 sites: The PM2.5 concentra-tions at 40 sites show a slightly bigger changefrom all dust source model results to no-mxsource model results. The PM2.5 changes

Table 9. Performance statistics of modeled surface PM2.5 concentrations.

Mean observed Mean modeled Mean bias Mean error Agreement(µg/m3) (µg/m3) (µg/m3) (µg/m3) index

8.66 4.70 -3.97 8.13 0.57

316 Yin and Sprigg

range from 0 to about 60µg/m3. The relativechanges could exceed 90%.

CONCLUSIONSIn the arid and semi-arid U.S. Southwest, windblown dust is an important part of airborneparticle pollution. In this work, DREAMmodel results were used to study the contribu-tions from wind blown dust sources in northernMexico to the dust pollution in this region.

The modeled meteorological variables,weather fields, dust concentration distributions,and PM10 and PM2.5 concentrations ofDREAM runs using all dust sources were veri-fied against observational data. The evaluationshowed that the modeled meteorological fieldsand PM concentrations were reasonably good.

The changes and differences of modeleddust concentrations between model resultswith all dust sources and model resultsexcluding Mexico dust sources were calcu-lated. The differences of episode-average dust(including dust with particles of all sizes,PM10 and PM2.5) concentrations in the duststorm area, the domain-average dust concen-trations in the same area, and PM2.5concentrations at 40 monitoring sites in NewMexico and Texas, showed that sources inMexico contribute substantially to airbornedust pollution.

According to changes in episode-averageconcentrations during Case 1, excludingsources in Mexico halved maximum total dust,PM10 and PM2.5 concentrations. Becausemaximum concentrations are critical in deter-mining a pollution episode, sources in Mexicoplayed major roles in Case 1, where the meancontribution of Mexican sources to airbornedust concentrations was about 42%. DuringCase 2, maximum episode-average dustconcentrations changed very little whenMexican sources were excluded; the meanrelative contribution from Mexican sourceswas about 34%. The mean relative differencethat Mexican sources made to domain-averagedust concentrations was approximately 43%,about the same for both cases.

The modeled PM2.5 concentrations at themonitoring sites in New Mexico and Texas

ranged from 0 to near 60 µg/m3 when changingfrom all source model results to no-mx sourceresults. At some sites in New Mexico andTexas, most of the PM2.5 concentrations werecontributed from dust sources inside Mexico.

The above results show that airborne dustpollution in the U.S. Southwest andsurrounding area is linked closely to dustsources in the two countries. Working onecosystem protection and management insidethe U.S. without considering contributionsfrom Mexico is insufficient to control pollu-tion in the Southwest. People in both countriesneed to coordinate their efforts in order tocontrol air pollution in this region.

ACKNOWLEDGMENTSThis work is funded by NASA under an EarthScience Research, Education, and ApplicationsSolutions Network (REASoN) project(CA#NNS04AA19A) and Research Opportu-nities in Space and Earth Sciences (ROSES)project (NNX08AL15G). MODIS land coverdata are provided by Dr. Karl Benedict at theEarth Data Analysis Center of the Universityof New Mexico. We thank Professor JimKoermer of Plymouth State University and theU.S. EPA for providing meteorological and airquality observational data. We gratefullyacknowledge Dr. Slobodan Nickovic for hiswork in developing DREAM and for permit-ting us to use the model system.

LITERATURE CITEDAlfaro, S. C., A. Gaudichet, L. Gomes, M. Maille.

1997. Modeling the size distribution of a soilaerosol produced by sandblasting. Journal ofGeophysical Research 102(D10): 11239–11249.

Arakawa, A., V. R. Lamb. 1977. Computationaldesign of the basic dynamical processes of theUCLA general circulation model. Methods inComputational Physics, vol 17, pp. 173–265.Academic Press, New York, NY.

Betts, A. K., and M. J. Miller. 1986. A new convec-tive adjustment scheme. Part II: Single columntests using GATE wave, BOMEX, and arctic air-mass data sets. Quarterly Journal of RoyalMeteorological Society 112: 693–709.

Burnett, R. T., J. Brook, T. Dann, C. Delocla, O.Philips, S. Cakmak, R. Vincent, M. S. Goldberg,and D. Krewski. 2000. Association betweenparticulate and gas phase components of urban air

Yin and Sprigg 317

pollution and daily mortality in eight Canadiancities. Inhalation Toxicology 12 (Supplement 4):15–39.

Chock, D. P., S.L. Winkler, and C. Chen. 2000. Astudy of the association between daily mortalityand ambient air pollutant concentrations in Pitts-burgh, Pennsylvania. Journal of the Air and WasteManagement Association 50: 1481–1500.

Chou, M. D., and M. J. Suarez. 1999. A solar radi-ation parameterization for atmospheric studies.NASA Tech. Memo. 15.

Chou, M. D., M. J. Suarez, X. Z. Liang, and M. M.H. Yan. 2001. A thermal infrared radiation para-meterization for atmospheric studies. NASA Tech.Memo. 19.

Freund, Y., R. E. Shapire. 1997. A decision-theo-retic generalization of online learning and anapplication to boosting. Journal of Computer andSystems Sciences 5: 119–139.

Georgi, F. 1986. A particle dry-deposition parame-terization scheme for use in tracer transportmodels. Journal of Geophysical Research 91:9794–9806.

Goldberg, M. S., R. T. Burnett, M. F. Valois, K.Flegel, J. C. Bailar III, J. Brook, R. Vincent, andK. Radon. 2003. Associations between ambientair pollution and daily mortality among personswith congestive hart failure. EnvironmentResearch 91: 8–20.

Hodges, J. C. F., M. A. Friedl, A. H. Strahler. 2001.The MODIS Global Land Cover Product: NewData Sets for Global Land Surface Parameteriza-tion, Proceedings of the Global Change OpenScience Conference, Amsterdam, 2001.

Janjic, Z. I. 1994. The step-mountain Eta coordinatemodel: Further developments of the convection,viscous sublayer and turbulence closure schemes.Monthly Weather Review 122: 927–945.

Kaiser, J. 2005. Mounting evidence indicts fine-particle pollution, Science 307: 1858–1861.

Kelsall, J. E., J. M. Samet, S. L. Zeger, and J. Xu.1997. Air pollution and mortality in Philadelphia,1974–1988. American Journal of Epidemiology146: 750–762.

Kwon, H. J., S. H. Cho, F. Nyberg, and G.Pershagen. 2001. Effects of ambient air pollutionon daily mortality in a cohort of patients withcongestive heart failure. Epidemiology 12: 413–419.

Mesinger, F., Z. I. Janjic, S. Nickovic, D. Gavrilov,and D. G. Deaven. 1988. The step-mountain coor-dinate: Model description and performance forcases of Alpine lee cyclogenesis and for a case ofan Appalachian redevelopment. Monthly WeatherReview 116, 1493–1518.

Mellor, G. L., and T. Yamada. 1982.Developmentof a turbulence closure model for geophysicalfluid problems. Review of Geophysics and Space

Physics 20: 851–875.Moolgavkar, S. H. 2000. Air pollution and daily

mortality in three U.S. counties. EnvironmentalHealth Perspectives 108: 777–784.

Moolgavkar, S. H. 2003. Air pollution and dailymortality in two U.S. counties: Season-specificanalyses and exposure-response relationships.Inhalation Toxicology 15: 877–907.

Nickovic, S., G. Kallos, A. Papadopoulos, and O.Kakaliagou. 2001. A model for prediction ofdesert dust cycle in the atmosphere. Journal ofGeophysical Research 106: 18113–18129.

Ostro, B. D., S. Hurley, and M. J. Lipsett. 1999. Airpollution and daily mortality in the CoachellaValley, California: A study of PM10 dominated bycoarse particles. Environmental Research 81: 231–238.

Pope III, C. A., R. T. Burnett, M. J. Thun, E. E.Calle, D. Krewski, K. Ito, and G.D. Thurston.2002. Lung cancer, cardiopulmonary mortality,and long-term exposure to fine particulate airpollution. Journal of American Medical Associa-tion 287(9): 1132–1141.

Quinlan, J. R. 1993. Programs for machine learning,Morgan Kauffman, 1993.

Roemer, W. H. and J. H. van Wijnen. 2001. Dailymortality and air pollution along busy streets inAmsterdam, 1987–1998. Epidemiology 12: 649–653.

Samet, J. M., F. Dominici, F. C. Curriero, I. Coursac,and S. L. Zeger. 2000. Fine particulate air pollu-tion and mortality in 20 U.S. cities, 1987–1994.The New England Journal of Medicine 343:1742–1749.

Shao, Y., M. R., Raupach, P. A. Findlater. 1993.Effect of saltation bombardment on the entrain-ment of dust by wind, Journal of GeophysicalResearch, 98: 12719–12726.

Smith, R. L., J. M. Davis, J. Sacks, P. Speckman,and P. Styer. 2000. Regression models for airpollution and daily mortality: Analysis of datafrom Birminham, Alabama. Environmetrics 11:719–743.

Smith, R. L., Y. Kim, M. Fuentes, and D. Spitzner.2000. Threshold dependence of mortality effectsfor fine and course particles in Phoenix, Arizona.Journal of the Air and Waste Management Asso-ciation 50: 1367–1379.

Styer, P., N. McMillan, F. Gao, J. Davis, and J.Sacks. 1995. Effect of outdoor airborne particu-late matter on daily death counts. EnvironmentalHealth Perspectives 103: 490–497.

United States Committee on Extension to the Stan-dard Atmosphere 1976. “U.S. StandardAtmosphere, 1976”, National Oceanic and Atmos-pheric Administration, National Aeronautics andSpace Administration, United States Air Force,Washington D.C.

OPEN SPACE PROTECTION AS A MEANS OF URBAN CONTAINMENT: A CASE STUDY FROM COLORADODavid Pesnichak

In America, with its superabundance of cheap land,simple property laws, social mobility, mania forprofit, zest for practical invention, and Bible-drunksense of history, the yearning to escape industrial-ism expressed itself as a renewed search for Eden.America reinvented that paradise, described sobriefly and vaguely in the book of Genesis, called itSuburbia, and put it up for sale.

— Kunstler 1993, The Geography of NowhereMost of the southwestern United States isgrappling with the lure of grand prosperitywhile dreaming of a simpler, smaller past.Cheap land and a strong regional traditionemphasizing property rights over governmentregulation has directed much of the growthoutward, creating sprawling cities and far-flung suburban communities that threaten toundermine the long-term viability of the citiesthemselves (Cooper 1997). In order to shedsome light on this issue, I investigated howurban containment policies that utilize theprotection of open space have affected threecities in Colorado and analyzed how suchprograms have helped these communitiesfactor in the vision of keeping open space animportant part of their growth. More specifi-cally, I took the lessons learned from Fruita,Durango and Boulder, and compared thoselessons with the historic and present situationin Montrose in order to examine whether anurban containment program that incorporatesopen space protection could help other westerncommunities in maintaining growth withsignificant open space.

As is evidenced by the ever-escalating costsof basic goods and fossil fuels, it has become

painfully obvious that all cities are membersof the ever-growing global economy: aneconomy which is now so large that societycan no longer safely pretend it operates withina limitless ecosystem. Developing an economythat can be sustained within the finite bios-phere now requires new ways of thinking(Daly 2005). There are a myriad of bygonecities from South America, Europe, Africa andAsia, as well as North America (such as theAnasazi who once inhabited much of what isnow southwestern U.S.), that are now ruinsbecause their population outgrew their avail-able resources. It leads one to wonder whetherthey were not able to change their thinkingwith the times, were not paying attention, orthat changes just came on them too quickly. Inthis modern era, both water and oil are withoutdoubt the most finite and important resourcesto the survival of a city infected with urbansprawl in the West.

According to Kunstler (2001), cities are atthe mercy of events halfway around the world.If the international oil markets suffer evenmoderate disruptions in the years ahead, muchof the U.S. will find itself in deep trouble.Hosansky (1999) reports that more than 80percent of trips in this country are made inprivate vehicles and traffic tie-ups costmotorists at least $74 billion every year inwasted time and fuel. Montrose is no excep-tion; according to the 2000 U.S. Census over87 percent of work commutes within the Cityof Montrose were made in private vehicles andthe median commute time was among the

320 Pesnichak

highest in the study group (Figure 1). Avail-able clean water is becoming more problematicevery year, with periodic droughts happeningregularly in the West (Brown 2006).

In addition to being intrinsically tied to thepolitically and geologically sensitive energysource of oil and putting more pressure ondepleted water sources, sprawl also eats upvaluable open space, worsens air and waterpollution and destroys Americans’ sense ofcommunity. Yet, developers and land-rightsadvocates call growth management policiesintrusive social engineering and say sprawl isunstoppable — a sign of American prosperityand an efficient market responding to thegrowing demand for a piece of the Americandream (Cooper 2004).

There is a constant struggle between privateproperty rights and what is perceived as theoverall public good. Planners are caught in avise between the knowledge that growth mustoccur somewhere and the reality that addingmore people doesn’t necessarily make betterplaces (Williamson 2004). Creating betterplaces however, is what planning is all about.According to Kunstler (1993) the future willrequire us to build better places or the futurewill belong to other people in other societies.The economic status quo cannot be maintainedlong into the future. If radical changes are notmade, we face loss of well-being and possible

ecological catastrophe. The main idea behindsustainability is to shift the path of progressfrom growth, which is not sustainable, towarddevelopment, which presumably is (Daly2005). Duane, Plater-Zyberk and Speck (2000)offer that the problem with suburbia is that, inspite of all its regulatory controls, it is not func-tional: it simply does not efficiently servesociety or preserve the environment.

Cooper (1997) emphasizes that despite ahands-off approach from the federal govern-ment, state and local governments are actingon their own to conserve green space withsmart growth initiatives to limit new develop-ment, and citizen-run, non-profit land trusts aresprouting up all over the country to buy upopen land. Although such enthusiasm exists fordefining communities and preserving openspace in Colorado, without regional orstatewide planning efforts, we now have situ-ations like Boulder where, according to EricBergman of the Colorado Office of SmartGrowth, sprawl just leapfrogged outsideBoulder. So you now have a pocket of smartgrowth with the sprawl all around it. Despitethe argument that such urban sprawl is an effi-cient market responding to the growingdemand for a piece of the American dream,Boulder arguably remains one of the mostattractive places to live in the U.S. preciselydue to the easily accessible protected open

Figure 1. The commuting habits of the residents of the cities in the study as collected bythe 2000 U.S. Census.

Pesnichak 321

space harmoniously coexisting with a stimu-lating urban center—and this pattern was nota free market creation, but the result of smartgrowth policies.

In the midst of all the new developmentoccurring in Colorado it seems the staunchestland rights advocates are those who stand toprofit the greatest from urban sprawl.According to Environment Colorado, an advo-cacy group out of Denver, 10 acres of valuableopen space and agricultural land are beingdeveloped every hour. Most Coloradanscontinue to support proposals to curb devel-opment, but big developers and the ColoradoAssociation of Home Builders are pushing tofurther reduce the ability of local governmentsto enact enforceable growth plans (Environ-ment Colorado, 2005). In addition, there arehuge dollars at stake with urban containmentand open space issues, and the interests that areagainst or contrary to this kind of plan and ideaare those who are making big bucks out ofdevelopment. In the meantime however, asignificant portion of the Montrose andColorado communities are able to see a largerpicture (Ron Stewart, August 2005, pers.comm.).

In a 2004 City of Montrose HouseholdSurvey, two of the three top problemsperceived to be facing the Montrose commu-nity were traffic congestion and too muchgrowth (Moorman, 2004). In addition, in a2005 Montrose Parks and Recreational NeedsAssessment, two of the five highest unmetneeds and highest importance issues to theresidents of Montrose were a lack of NaturalAreas/Wildlife Habitats and Greenway Areasalong the Uncompahgre River (Leisure Vision2005). According to Graham Billingsley, it isimperative to ask: What have we got? What dowe want to keep? And what do we need tobring in that we are missing? In the results ofthe 2004 Household Survey and the 2005Parks and Recreation Needs Assessment, theseanswers have been clarified: less traffic, slowergrowth, unimproved open spaces, wildlifehabitat, and greenway areas along the Uncom-pahgre River. And according to Mayor EricaLewis-Kennedy, we have to keep the open

spaces around Montrose. If we are going tomaintain our quality of life, we have to haveopen space or it will just be another AnywhereUSA (July 2005, pers. comm.).

The necessity for more open space anddecreased sprawl goes even further than socialviability, livability and constituent desires.According to the Transit Cooperative ResearchProgram (TCRP) Report 74, sponsored by theFederal Transit Administration, by developingwith growth-control measures in place versustraditional or uncontrolled growth, the resi-dents of the U.S. could save $12.6 billion ininfrastructure costs and 155 million gallons ofwater and sewer demand per day between2000 and 2025. In addition, in the same period,the U.S. could save 188,300 roadway lanemiles and another $110 billion in roadconstruction costs by using growth controlmeasures (Burchell et al. 2002). A studyconducted by the University of Colorado deter-mined that future sprawling development inDelta, Mesa, Montrose, and Ouray Countieswould cost taxpayers and local governments$80 million more than smart growth between2000 and 2025. It is inevitable that the costs ofbuilding and servicing infrastructure for newsprawling development is ultimately subsi-dized by the whole community (Coyne 2003).Sprawled residential growth does not do a cityany good because in the long run it costs morethan it generates. In a Cost of CommunityServices Study (COCS) conducted by theAmerican Farmland Trust, residential devel-opment costs were found to be a dollar fifteencents for every dollar raised in taxes whereasworking and open space land costs an averageof only thirty-six cents per dollar raised.

In addition, the Governor’s Commission onSaving Open Spaces, Farms & Ranches hasreinforced these findings. Coloradoans placetremendous value on their open spaces, farmsand ranches because open space is essential tothe state’s quality of life. It is, in fact, one ofthe principal reasons many Coloradoans decideto make this their home (Harris et al. 2000).However, Colorado can also be a difficultplace when citizens can’t reach agreement onwhether or how they want to manage growth

322 Pesnichak

and/or protect open space. This leads to thephenomenon of having places like Boulder,with incredibly strict control, surrounded byareas that are not doing much at all (Cooper1997).

With the complex and decentralized systemof open space protections and growth manage-ment regimes throughout the U.S. andColorado, it is difficult to assess the cumula-tive effectiveness of the various programs.However, there has been a strong surge ofinterest in open space programs in the last fewyears, especially in rapidly urbanizing areas.Unfortunately our understanding of the impactof open space protection programs on metro-politan growth is sketchy at best. It is clear thatopen space acquisition programs clearly affectmetropolitan growth patterns, but the form ittakes is different in every metro area (Hollisand Fulton 2002).

The solutions may be complex, and notsupported by everyone, but as urban sprawland what Kunstler calls the ‘geography ofnowhere’ is unmistakably taking hold withinthe city of Montrose and even in thesurrounding countryside outside the city limits,Montrose is struggling to define itself as anenvironmentally and economically attractiveplace to live, work and play. The sense that thetask before the Montrose community is toolarge and insurmountable is overshadowed bythe responsibility of elected officials to bringthe community vision of smart growth andprotected open spaces into reality.

METHODSInterview List

This study was based on literature review andextensive interviews with local planners andpoliticians in Boulder and Boulder County;Fruita and Mesa County; Durango, Montrose,and Montrose County. The individuals and thedate of their interview were:• Steve Aquafresca, Mesa County Land

Trust; July 22, 2005;• Allan Belt, Montrose County Commis-

sioner; August 8, 2005;

• Eric Bergman, Colorado Dept. of LocalAffairs, Office of Smart Growth; August 4,2005;

• Graham Billingsley, Director of BoulderCounty Planning Dept.; August 5, 2005;

• Bennett Boeschenstein, City of FruitaCommunity Development Director; July22, 2005;

• Virginia Castro, Durango City Councilor;June 15, 2005;

• Dennis Erickson, City of Montrose ParksPlanner & President of Montrose Chamberof Commerce; July 28, 2005;

• Erica Lewis-Kennedy, Mayor, City ofMontrose; July 26, 2005;

• Renee Parsons, Durango City Councilor;June 15, 2005;

• John Schneiger, Former Montrose CityManager; July 26, 2005;

• Randy See, Community Organizer, WesternColorado Congress; July 25, 2005; and

• Ron Stewart, Director of Boulder CountyOpen Space; August 5, 2005.

DefinitionsOpen SpaceThis project utilized the definition of openspace described by Hollis and Fulton (2002):land that is not devoted to urban development,especially if that land is located in a metropol-itan region. The actual uses of lands that areset aside for open space are quite varied. Usingthis definition, open space is any land that isnot already devoted to development throughsubdivision, covenants, speculation or intent.Examples of such open spaces are agriculturallands, ranch lands, recreation areas, nationalforests, protected river ways, parks andprotected wetlands.

Urban ContainmentThis project utilized the definition of urban

containment presented by Fulton, Martin and

Pesnichak 323

Pendall (2002): a set of land-use regulationsthat prohibit urban development outside acertain boundary. This paper looks at thevarious options and proven methods of urbancontainment that simultaneously incorporate“push” and “pull” factors. According toDawkins and Nelson (2004): urban contain-ment has two fundamental purposes: (1) topromote compact and contiguous developmentpatterns that can be efficiently served by publicservices; and (2) to preserve open space, agri-cultural land, and environmentally sensitiveareas that are not currently suitable for urbandevelopment. They recommended that urbancontainment programs be based on thefollowing seven objectives. A containmentprogram should:1) Accommodate long-range urban population

growth requirements consistent with stateand local goals and policies;

2) Fulfill local needs for housing, employmentopportunities, and livability;

3) Provide public facilities and services in anorderly and economic manner;

4) Maximize efficiency for land uses in or atthe fringe of existing urban areas;

5) Consider all environmental, energy,economic, and social consequences;

6) Preserve farm, forest, and other resourceland; and

7) Ensure the compatibility of proposed urbanuses with nearby resource activities.

Dawkins and Nelson (2004) demonstratethat urban containment plans can be dividedinto four basic types that represent a combina-tion of either strong or weak boundaries andeither restricted or accommodated develop-ment. The four categories are summarized asfollows:1) Weak-Restrictive. Cities that utilize this

approach have an inward-focused growthmanagement strategy with an unclear ulti-mate urban form. These growth

management plans are composed of weakcontainment measures, but they adoptrestrictive policies toward containinggrowth. Examples of Weak-Restrictivecommunities are Aspen, Colorado, andBloomington, Indiana.

2) Strong-Restrictive. Cities that utilize thisapproach have a self-determined urbanform, but no regional strategy. Thesegrowth management plans have strongcontainment but do not place high priorityon meeting regional development needs.Examples of Strong-Restrictive communi-ties are Boulder, Colorado, and San LuisObispo, California.

3) Weak-Accommodating. Cities whichutilize this approach have limited openspace protection, weak planning statutes,and plans that minimize facility costs.These growth management plans employweak urban containment measures, princi-pally through lax management or ruraldevelopment within the county or region.These plans do, nonetheless, attempt toaccommodate development pressures.Examples of Weak-Accommodatingcommunities are Lincoln, Nebraska, andCookeville, Tennessee.

4) Strong-Accommodating.Cities that utilizethis approach have a desire to balance openspace with growth pressures and to shapemetropolitan urban form. These growthmanagement plans contain developmentthrough spatial growth limits combinedwith aggressive open space preservation.These plans also meet projected growthneeds to accommodate development pres-sures. Examples of Strong-Accommodatingcommunities are Portland, Oregon, andTucson, Arizona. According to Fulton, Martin, and Pendall

(2002), every metropolitan area in the U.S. hassome form of urban containment. AlthoughMontrose does not explicitly have an officialurban containment plan, it does have a looselycoordinated, state-required, 201 Sewer Service

324 Pesnichak

Boundary, large tracts of federally controlledland to its east and west, and a recently re-adopted land use IntergovernmentalAgreement (IGA) with Montrose County. Aswith many land use IGAs, the agreementbetween the City and County of Montroseprovides the city with a first right of refusal fordevelopment occurring near the cities bound-aries. With these regulations, it could be arguedthat the City of Montrose falls into the Weak-Accommodating category which ischaracterized by large-lot land preservationand infrastructure goals, an emphasis on infra-structure and land supply, as well as moderateintergovernmental coordination. Communitiesthat use these plans typically have low percapita income, low population density and arecommon in states with no local planningmandate.

An Effective Open Space as Urban Containment Program

Unlike Oregon, Washington, Vermont, orFlorida, Colorado does not have a statewideland use plan. In the absence of a Coloradostate land use plan to use as a benchmark, Imeasured an ‘effective’ open space program asa means of urban containment in relation to thejurisdiction’s community vision as stated intheir respective comprehensive or master plan.I also focused on the opinions of local officialswhom I interviewed to back up my analysis ofwhat makes a successful or unsuccessfulprogram. The determination of an effectiveprogram is therefore subjective.

CASE STUDIESDurango, Colorado

The community vision for Durango is acommunity living in harmony with its naturalenvironment; where residents and visitors canenjoy a historic small town atmosphere, theAnimas River, and easy access to exceptionalcultural, educational and wilderness opportu-nities (Carlisle et al. 1997). Durango is acommunity that clearly benefits from thefederal open lands surrounding the community.According to Virginia Castro, a Durango CityCouncilor, the aesthetic value and the air

quality that the open spaces and the NationalForest provide really help maintain a cleanenvironment in the area. Once those thingsstart diminishing, it just follows that theeconomic viability of the city goes with it.Renee Parsons, Durango City Councilor,agrees that the open spaces surrounding thecommunity are very attractive and economi-cally very good for the city. However,according to a May 19, 2005 debate capturedin the Durango Telegraph, Parsons explainsthat Durango is facing a threat to its quality oflife that will forever alter the landscape, thelifestyle and the community. While growth hasbrought certain benefits, the community is ata critical juncture (Wells 2005). According toGreg Hoch, Planning and Community Devel-opment Director for the City of Durango,“There are old-timers vs. newcomers, settlednewcomers vs. new newcomers, and theinherent contradictions of keeping housesaffordable, keeping regulations from becomingonerous and maintaining some sense of thecommunity as it exists. But one of the under-lying forces is city vs. county, or arguably, theattractive sprawl of low-density rural subdivi-sions vs. higher-density new urbandevelopment. As more people move in, andcities run out of room, the practice of allowingoutlying areas to fill with rural-style, low-density subdivisions is becoming a flashpoint.(Hoch, 2004).

The city of Durango, which has experi-enced economic boom and bust periods overthe past 55 years, was first incorporated in1881. Since 1950, the city’s greatest period ofpopulation expansion was from 1950 to 1960,when the population grew by 29 percent in tenyears (Figure 2). Durango has seen only oneten year period since 1950 where the popula-tion actually declined, between 1960 and 1970.By the year 2000, the U.S. Census indicatesthat Durango had a total population of 13,922,a land area of 6.8 square miles and a popula-tion density of 2047 people per square mile.By 2003, the Colorado Department of LocalAffairs (DOLA) estimates that Durango had apopulation of 15,324 with an annual growthrate of less than one percent between 2002 and

Pesnichak 325

2003. Over the years Durango, like mostColorado towns, has experienced a fairlysporadic but generally increasing population.But, in the last couple years this populationincrease appears to have been slowing on apercentage basis which has been accompaniedby a declining school enrollment.

According to the Colorado PreliminaryForecasts issued by DOLA in November 2004,“the growing number of tourist dollars as thebaby-boomers reach middle-age, the desire ofmany small businesses to move to smallerareas, and the overall number of retirees, areexpected to continue or increase” (DeGroenand Westkott 2004). In addition, Durango hasseen a tremendous influx of people who wantto recreate and have the lifestyle of spendingas much time outdoors as they want. A lot ofpeople are moving to Durango who have madea lot of money someplace else and are able toretire here. A lot of people are moving herewho have sold homes somewhere else and canafford to pay more here. And so we have seena tremendous increase in property values over

the last few years. The attraction of the surrounding open

space on the local economy is having effectsbeyond that of inflated real estate prices andan increasing population, however. The newpopulation has meant a declining primary andsecondary student enrollment population as aresult of these demographic shifts. The districthad 107 fewer students in 2004 than it had in2002, and it expected another dip in the 2005school year. Enrollment in the district was hurtby high housing prices that made it difficult forfamilies with children to afford living inDurango (Slothower 2005). In effect, landprices in the surrounding areas as well as thecity center have increased tremendously dueto development pressures brought on by newresidents looking to have a front row seat tothe open spaces surrounding Durango.

As residents search for housing they canafford, sprawling development is encouragedand decreased affordability has resulted in anincrease in residents renting and a decline inhome ownership. In 2000, 52% of the popula-

Figure 2. Population from 1950 to 2000 U.S. Census for Durango, Fruita, and Montrose,Colorado.

326 Pesnichak

tion of Durango lived in a home that theyowned, while 48% rented. And althoughhousing prices in Durango are still lower thenthat of Boulder, the ratio of homeowners torenters are roughly 50/50 in both cities. Mean-while, home ownership in Fruita and Montrosewhere housing is less expensive, are signifi-cantly higher (U.S. Census 2000). In reality,the decline in home ownership and decliningschool enrollment reflects a relocation offamily units from the city of Durango tonearby cities and outlying sprawl wherehousing is more affordable. In November2003, the median price of homes for sale inBayfield, twenty-five miles east of Durango,was $207,000. Durango’s median price was$410,000 (Hoch 2004).

Because Durango is surrounded by Bureauof Land Management (BLM), U.S. ForestService and tribal land it is limited in its avail-able annexable land. These abutting publiclands are not only containing urban sprawlfrom moving out in every direction from thecity center, they also force development tofocus on the highway corridors and mix-matching affordable housing and veryhigh-end residential development. In addition,as land prices increase from the heightenedwealth accumulating in the City of Durango,pasture land, agriculture, and previously inex-pensive home sites are being converted to highend residential subdivisions which are movingoutward from the city center.

As these economic pressures combinedwith local land use regulations that restrictdensity in the inner city core are encouragingever expanding sprawl, citizens are beginningto want growth controls. A responsible growthinitiative just barely failed in 2004. The city isnow built out so what they are looking at forthe future is core infill (Parsons). Even thoughthe Responsible Growth Initiative narrowlylost that would have required voter approvalfor new annexations to the city, an open spaceacquisition sales tax was passed and aproposed City / County Land Use Intergov-ernmental Agreement was started. Thequestion now is how Durango will ultimatelymanage, and potentially contain, its outward

growth in the future. It is predictable that suchpolicy developments will lead the communitytoward its stated vision of a community livingin harmony with its natural environment;where residents and visitors can enjoy ahistoric small town atmosphere, the AnimasRiver, and easy access to exceptional cultural,educational and wilderness opportunities.

According to the Dawkins and NelsonPlanning Advisory Service (PAS) report,Durango primarily fits into a category notpreviously mentioned and not analyzed asapart of their PAS report. Durango mainly fitsinto the category of ‘Natural Containment’,which limits cities like Honolulu, Hawaii, andJuneau, Alaska, as it is surrounded by state,federal and tribal lands. Even though Durangois limited by these physical land useconstraints, growth pressures are coming to ahead, and new development is starting to moveout along the highway corridors and may actu-ally create satellite communities (such asGrand View) if some other form of growthcontrol is not instituted.

Fruita, ColoradoThe community vision for Fruita is to preservethe rural, small town atmosphere of the FruitaCommunity, while providing quality commu-nity services for a growing population andstriving for economic development and pros-perity (City of Fruita 2000). Fruita ishistorically an agricultural community. Incor-porated in 1894, Fruita was a town withhundreds of acres of apple orchards that actu-ally won prizes for the fine quality of the applesthat had grown there (Boeschenstein). Appleorchards were not destined to stay long term inFruita, however. In the 1920s, according toBoeschenstein, Community DevelopmentDirector for the City of Fruita, the Coddlingmoth wiped out the orchards. Today the agri-culture that surrounds the town is primarily rowcrops and livestock and although agriculture isstill big in Fruita, its population has grownsignificantly since the time when appleorchards ruled the local economy.

Like all the cities in this study, Fruita hasgone through significant boom and bust cycles.

Pesnichak 327

The town had a downturn in populationbetween 1960 and 1970, but that is the onlydecade since 1950 in which Fruita has lostpopulation (Figure 2). Between 1990 and2000, the city saw the greatest increase inpopulation, going from 4,045 to 6,478, almosta 38 percent increase. Among the citiesincluded in this study, Fruita had the highestrate of population increase between 1950 and2003 at 528%. Further, since 2000, Fruita hasseen the strongest population increase of thecities in this study, growing at a rate of aboutsix percent between 2002 and 2003. The causeof this increase is primarily its location: it sitsnext to the largest city on the Western Slope,Grand Junction, and is in the Interstate 70corridor. In the mid-90s Grand Junction beganan aggressive expansion push, threatening toturn Fruita into a bedroom community. Thetown suddenly found itself in the middle of anannexation war (Jenkins 2001).

According to Boeschenstein, “we had twochoices, we could look like Orange CountyCalifornia with wall-to-wall subdivisions fromthe Book Cliffs to the Monument, or we couldlook like some other examples, which I reallyknow upsets people, but Boulder is surroundedby open space … it did not sprawl and is oneof the most attractive communities in the U.S..It has some great corporate headquarters, it hassome very high paying jobs, a strong intellec-tual base, it also is a caring community andthey actually do have affordable housing. Thatwould be another model.”

At roughly 8,000 people in 2003, Fruita iscertainly at a different point in history thanBoulder, which has a population of just fewerthan 100,000 people. In light of this growth,public concern led to two years of meetingsbetween the cities of Fruita and Grand Junc-tion and a decision in 1996 to maintain anopen-space buffer between them. Grand Junc-tion and Fruita agreed that they would notannex and they would not extend sewer linesto properties in the buffer so that urbanizationwould be slowed in that strip (Jenkins 2001).The Town of Fruita then went a step furtherand in 2002 adopted a Community Plan thatincluded a chapter on the Transfer of Devel-

opment Rights. It took about three years ofpublic meetings, futures workshops, andcommunity surveys to sort out how much thecitizens wanted the community to grow, howbig they wanted the community to be, andwhat values were most important over the next20 years. The idea was that it is important topreserve and protect open land and farm landaround Fruita, including land that is adjacentto the Colorado National Monument, BLMlands, and the Colorado Canyon NationalConservation Area (Boeschenstein).

What the Town of Fruita adopted isexplained in the Fruita / Mesa County Transferof Development Rights Users Manual.According to the manual, the purpose of theprogram is to encourage the retention of agri-cultural lands in Mesa County and to protectopen lands between Fruita and Grand Junction(Fruita/Mesa County 2003). In addition, themanual explains that beyond a public policyfoundation, the establishment of a Transfer ofDevelopment Rights / Credits (TDR/C)Program requires three fundamental compo-nents: (1) an intergovernmental agreementbetween Fruita and Mesa County providing aframework for coordinating the program; (2)updates to the City of Fruita Land Use Codeand the Mesa County Development Coderegulating the program’s implementation; and(3) a manual outlining the details on how touse the program (Fruita/Mesa County 2003).

The city and county government’s role in aTDR program is to set rules. The rules basi-cally consist of two major components: asending area and a receiving area. The programworks on a free-market principle where it issimply necessary for the local governments tolegally establish the market within the land useregulations. Once the program is establishedwithin the necessary codes and IGAs, alandowner in the Sending Area may volun-tarily sell the development rights or credits toa landowner of a receiving site, at a marketvalue established by the landowner and thebuyer (Fruita/Mesa County 2003). What isbought and sold are development rights only,which means that the seller still owns their landand can keep doing what has historically been

328 Pesnichak

done on their property. However, like the saleof water rights, that land now has limited struc-tural development potential. The purchaser ofthose rights can then place more units per acreon a development property in the designatedreceiving area.

The optimism that Boeschenstein has aboutthis program is well founded as TDR programshave been used successfully by many juris-dictions throughout the country for many years(Cooper 1997). What makes Fruita unique isthat it is not a particularly wealthy communityyet it is still able to see the benefits inconserving open space as well as valuable farmand ranch land next to the city limits. “Peoplehere wanted the small town and they wantedthe open space around them,” Boeschensteinsaid. “It is hard, especially here on the WesternSlope, because people want to do whateverthey want with their land and do not think itshould be regulated by government. But thenagain, the TDR and the conservation easementis not a government program, the governmentjust has to enable it and if it happens ithappens, if it doesn’t it doesn’t, but at least youare enabling it to happen. It is a big carrotincentive.”

To this end, Fruita has been able to workwith their neighboring jurisdictions to helpachieve the vision for their community, withthe impetus coming from the residents.Although the lack of precise control in a TDRprogram may scare many in government andplanning offices, it places the fate of the cityinto residents’ hands instead of elected officialsand enables the free-market to providepayment to land owners, which is often notbudgeted in government financial plans.

According to Dawkins and Nelson (2004),Fruita falls into the Strong-Accommodatingcategory, in which Fruita is looking to balanceopen space with growth pressures and to shapemetropolitan urban form. Based on the bene-fits that such protected open space willeventually provide Fruita economically andsocially, I think it is clear that Fruita is doingwhat it can to make sure that its communityvision is advanced while maintaining the smalltown atmosphere. Hence, I believe that Fruita’s

growth management measures can be consid-ered effective.

Boulder, ColoradoBoulder has a long tradition of communityplanning. Most of the key policies and plansthat have guided the development pattern inthe Boulder Valley have not changed since the1978 Boulder Valley Comprehensive Plan wasfirst adopted, and many of them stem fromlong-standing community values (City ofBoulder 2001). They represent a clear, articu-late vision of the development patternincluding:• Respect for the community’s unique iden-

tity and sense of place;• Recognition of sustainability as a unifying

goal to secure Boulder’s future economic,ecological and social health;

• Commitment to open space preservationand the use of open space buffers to definethe community;

• Use of urban growth boundaries to main-tain a compact city (the boundaries of theservice area have remained virtuallyunchanged since first developed in 1977);

• Growth management to regulate the rateand overall amount of residential develop-ment and redevelopment opposed tosprawl;

• Recognition of the importance of a centralarea (Downtown, University of Colorado,the Boulder Valley Regional Center) as aregional service center of the BoulderValley; and

• Commitment to a diversity of housing typesand price ranges to meet the needs of theBoulder Valley population.

The City of Boulder had a population of about98,000 people in 2003; it is also the largest cityand the city with the most significant popula-tion growth since 1950 in this study. Boulderbegan preserving open space in 1904 in orderto protect watershed resources as insurance fora stable water supply, as well as to protect the

Pesnichak 329

mountain backdrop for recreational and view-shed purposes. Initially, the purchase of openspace and growth controls were not linked. Inthe 1950s, Boulder’s population grew from25,000 to 37,000, and during the 1960s it grewby a whopping 29,000 to reach 66,000 (Figure2). Some initial efforts to manage this growthincluded the Blue Line, a citizen-initiatedamendment to Boulder’s charter in 1959 thatrestricted the extension of city water serviceabove an elevation of 1753 m (5,750 ft). Whatthis Blue Line did was preserve the Flat Irons,the magnificent mountain backdrop of Boulderfrom being developed. In the proceeding years,Boulder extended the Blue Line concept tosewer service.

Another important growth managementprogram began in 1967, when Boulder becamethe first city in the U.S. to pass a tax dedicatedto preserve open space. This open spacesystem forms the outer extent of the BoulderValley, a joint planning area between the cityand county (Pollock 1998). Since the passageof the first open space tax in 1967, the City ofBoulder has established a 40,000-acre green-belt around the city as a defense againstsprawling development. Boulder County didnot pass an open space tax until 1993. Sincethat time, Boulder County has purchased75,000 acres of open space.

Too often we hear that communitiescannot afford to ‘grow smart’ by conservingopen space. But accumulating evidence indi-cates that open space conservation is not anexpense but an investment that producesimportant economic benefits (Rogers 1999).Eric Bergman agrees, “Open space has somany values for a community. It can be abuffer between communities… and helppreserve a sense of identity and boundary.Obviously, the viewsheds, habitat, agricul-tural ranch land and farm land viability andcultural value for a community make sure thatit remains viable.” In the process ofpreserving open space for almost forty yearsnow, Boulder has become intimately familiarwith the consequences. It makes it a greatplace to live, people move here because of theculture of the outdoors, and they move here

instead of someplace else on the Front Rangebecause of the availability and easy access torecreation (Billingsley).

The open space and urban containmentprograms in Boulder also come with critics.According to Billingsley, Boulder does nothave growth management, Boulder has resi-dential growth management where they haverestricted residential growth to one percent ayear while allowing commercial growth tooccur at 7 percent a year. That is why Boulderhas traffic issues. According to Stewart,Boulder has way too much land zoned forcommercial and industrial and way too muchpotential for commercial and job growth in thecity when at the same time they do not havemuch room for population growth. The citymay grow by 10, 15, or 20 percent in popula-tion but has a potential of expanding 50, 60, or70 percent in jobs. That equation should havebeen put in better balance. According toDawkins and Nelson (2004): It is one thing to limit urban development withinboundaries but quite another to absorb the develop-ment that would have occurred in its absence.Montgomery County attempts to absorb its projectedgrowth within urban areas contained by growthboundaries, as does metropolitan Portland. Incontrast, cities such as Petaluma, California andBoulder, Colorado do not attempt to absorb largeshares of the region’s projected growth. Because oftheir sensitivity to meeting regional or sub regionalhousing needs, it may be no accident that housingprices rose less quickly in Portland and MontgomeryCounty in the 1990s than in Petaluma, Boulder, orthe counties of Ventura and Loudoun.

The use of open space as a means of urbancontainment in Boulder, as well as a system forcontrolling the rate of population growth bylimiting building permits, and a defined urbangrowth boundary managed in cooperation withBoulder County, has placed Boulder firmly inDawkins and Nelson’s category of Strong-Restrictive. As a result of these containmentpolicies, housing demand appears to bedisplaced to Boulder County’s other major city,Longmont, and into the rural areas aroundLongmont. Indeed, from 1988 through 1998,Boulder’s share of Boulder County’s newhousing fell from 29 percent to 8 percent while

330 Pesnichak

Longmont’s share rose from 10 percent to 31percent (Dawkins and Nelson 2004).

The Boulder Valley Comprehensive Planhas been successful in many areas. Its imple-mentation has helped keep the communitycompact and limit sprawl. It has helpedpreserve open lands, intensify the core area andpreserve important features of the local envi-ronment. It has also encouraged thedevelopment and use of alternative modes oftransportation. Areas where additional effortsare still needed include the availability ofaffordable housing, the growing imbalancebetween jobs and housing, and reducing trafficcongestion by providing safe and convenientalternatives to single occupant vehicles (Cityof Boulder 2001). According to Dawkins andNelson (2004), as of 2000, the ratio of jobs tohousing in the Boulder Valley was 92 to 1. Toreduce this imbalance, the city adopted aninclusionary housing program in 1999 thatrequires 20 percent of all new residentialdevelopment to be affordable to persons of lowto moderate incomes. The 2001 comprehen-sive plan supplements this program withpolicies designed to promote housing devel-opment along transit corridors and withincommercial centers. As a result, attaining thevision for Boulder has been mostly a success;however there are certain areas whereimprovement is needed and appropriate actionshave been taken to enact those improvements.Boulder has maintained a central vision of acompact city with a clear identity in the midstof a rural area. The growth management tech-niques used in Boulder may vary from thoseused in other cities, and they may be changedfrom time to time to meet local conditions, butthe vision has remained intact (Pollock 1998).

Montrose, ColoradoThe community vision of the City of Montroseand its surrounding area of influence is that thecitizens wish to achieve a balance between theprotection of their natural environment andprogressive economic development, maintaina sense of community character and quality oflife, conserve their agricultural heritage andrecreational resources, and accommodate

diverse community needs for increasedemployment, education and business opportu-nities (BRW 1998).

The City of Montrose has experienced rapidgrowth for the past decade. In 2004, the cityrecorded almost one new single-family homebuilt per day. In addition, Montrose hasannexed more than 2,191 acres from 2000 to2005 to bring the overall size of the city to over15 square miles, only ten square miles less thanBoulder. In addition, the city population grewat a rate of approximately 4% in 2003 and2004. The Colorado Department of LocalAffairs estimates that the city of Montrose hada population of 15,351 in 2004, up from12,344 in 2000 and 4,964 in 1950. Since 1950the city has grown in population by almost310% (Figure 2). If the current rates of growthcontinue, Montrose will surpass Durango inpopulation within the next few years, but popu-lation growth rates change very quickly inColorado and Montrose is no exception.Unlike Fruita and Durango, Montrose has nothad any decades since 1950 where the city haslost population and it has the lowest popula-tion density at 1,073 people per square mile.Overall concern for the growth of Montrosehas been voiced frequently at public meetingsas well as in the local newspaper. In an April2005 article published in the Montrose DailyPress, then Montrose City Councilor EricaLewis-Kennedy reinforced that one of thebiggest issues facing the city is growth(Hildner 2005), and the Montrose communityagrees: in a 2004 Household Survey conductedby the city, residents ranked traffic congestion,too much growth, and declining open spacewithin the top five issues facing Montrosetoday. Meanwhile, the 1998 City of MontroseComprehensive Plan sets its first goal to helpensure the continuation of Montrose’s smalltown rural character and high quality of life.

Dennis Erickson related to me, “My fear isthat the Western Slope is like the Denvermetropolitan area and Front Range was 80years ago — from Grand Junction to Delta toMontrose to Ridgeway is all going to becomeone metropolitan district.” Further, accordingto Montrose Mayor Erica Lewis-Kennedy,

Pesnichak 331

growth is destroying the small town character.You cannot keep the rural nature of thecommunity and have growth at the rates we arehaving them. Even with these concerns,Erickson explained that the Chamber looks atgrowth as being positive, because business isgrowing, and the area is growing and Montrosefor so many years was limited with businessopportunities.

Montrose is losing its vision for the futureby a population temporarily blinded by theedification of monetary gains and a generallyun-thoughtful search for the good life.Montrose is in transition, there are importantelements of the community who want thingsto stay wide open, without government inter-vention, and this has made it difficult for staffand council to get to where other communitiesare in the state. Montrose has had a tremen-dous amount of growth and will likelycontinue to experience growth. It is importantto have a good comprehensive plan and updatethat plan on a regular basis. Billingsley feelsthat simply having a decent comprehensiveplan is not enough. One of the failings of a lotof communities that do a good comprehensiveplan is that they do not really make peopleunderstand what is necessary to make it work.As long as it is a visionary document that doesnot have any teeth behind it, then everybody issupportive of it. So you really have to get thosediscussions going that say “if you really meanthis, then these are the consequences of whatyou are saying.” Montrose is leaning moretoward the good visionary comprehensive planwithout any teeth that everybody supports, butdoes not lead to any meaningful change.

In addition to having a good strong compre-hensive plan that has teeth and which thecommunity supports, if Montrose is going tobe able to maintain its rural, small town char-acter, it is imperative that there is agreementwith the county. In late July 2005, the city andcounty of Montrose finally reaffirmed theircommitment to a land use IntergovernmentalAgreement (IGA). The IGA has severaldifferent roles, two important ones are that itsets the urban interface boundary and it allowsthe city to have first right of refusal to devel-

opment that is happening inside this urbaninterface boundary.

The cost savings to pre-determining wheregrowth will occur and attempting to define aninfrastructure development plan is essential toproviding the community with a reasonablelevel of service well into the future. “In thelong-term, planning pays for itself. Theproblem is that most people do not really lookthat far into the future; they are kind of reac-tive. We represent all the taxpayers and as aresult we have to look at what the future costsare going to be to the city and something thatconcerns me here is that someday there maynot be the tax base to support this community”(John Schneiger).

It is clear that Montrose has made a small,yet important, first step to ensure the vision asstated in the 1998 Comprehensive Plan, withthe recent signing of the land use IGA.However the city, as an organization, has madelittle effort to protect the natural environmentand has taken a myopic view of economicdevelopment within the city. The principlepurpose for doing things like defining thecommunity and having some open space is toattract economics into the community and youwant the ones that will bring variety. So thatgood and bad economic times create stabilitywhere the good times aren’t as good and thebad times aren’t as bad — so you want asmuch diversity as you can in a place(Billingsley). In addition, as is evidenced byever increasing traffic, stagnant downtownsales tax receipts and strip commercial devel-opment along the state highways, Montrose’ssense of small town character and corridorshave arguably eroded away to be indistin-guishable from almost anywhere else in theU.S. Furthermore, Montrose’s land use codes,due to the lack of agricultural zoning orenabling preservation measures, arguablyencourage subdivision sprawl and uncontrolledgrowth which has eaten away not only at thecommunity’s agricultural base (expressed inthe vision statement as a important element toprotect), but has also decreased the diversityof the local economy which, outside RussellStover, government, the hospital and the

332 Pesnichak

school district, is primarily based on construc-tion. According to Erica Lewis-Kennedy (July2005, pers. comm.), “the new developmentsthat have come to town with the new Hastings,Blockbuster, etc… are still not professional$30,000 or $40,000 a year jobs that peopleneed to maintain houses. It is obvious that thevalue of jobs is not going up but the real estatemarket keeps increasing.” To this end,although the IGA is an important first steptoward the realization of the communityvision, it would be erroneous to conclude thatthe city has truly worked toward the vision thatthe community set out in 1998. As Montrosedoes not have a growth-control program, it isclear that a lack of governmental action in thisarea will not lead the city toward its statedvision.

The large lot development is not farmground. In addition, another recent AmericanFarmland Trust (AFT) study found thatMontrose County is among the top 25 coun-ties in the seven-state Rocky Mountain Westwith 295,040 acres of prime ranchland at riskof conversion to low-density residential devel-opment. This land is equivalent to 21 percentof all land in Montrose County and 6 percentof all of Colorado’s ‘strategic’ ranchland.Again, according to the study, nearly all of the295,040 acres is along the Highway 50corridor. The study also indicated that 11percent of all prime ranchland in the RockyMountain West is threatened by conversion toresidential development by 2020 (AFT 2000).By allowing for a TDR or even PDR (Purchaseof Development Rights) and fee-simplepurchase in the future in combination withcollaborative regional planning efforts, localgovernments can help curtail this trend.

FINDINGSToo many community leaders feel they mustchoose between economic growth and openspace protection. But no such choice is neces-sary. Open space protection is good for acommunity’s health, stability, beauty, andquality of life. It is also good for the bottomline (Rogers 1999). Urban containment and thepreservation of open space are important attrib-

utes to creating efficient, sustainable, environ-mentally responsible, economically viable andlivable cities. The need to control urban sprawland preserve open space is particularly acutein the U.S., where the federal government andmost states, including Colorado, have a hands-off approach to land use and development. Asa result, it is up to local governments toconstruct places in a smart, forward-lookingmanner.

As the case studies for Durango, Fruita,Boulder, and Montrose demonstrate, eachcommunity is unique, but how do we comparethem? Here are four possible elements of adefinition for city greatness: (1) Alloweveryone to live in well-planned communitiesand neighborhoods; (2) Save natural areas andopen space and protect cultural and historicfeatures; (3) Provide adequate infrastructure;and (4) Locate all intense, ‘attraction’ activi-ties in well-planned, mixed-use areas —downtowns and cores (Engelen 2005).Although this study focuses on open space andurban containment, it is clear that each afore-mentioned element that contributes to a city’sgreatness affects the others in some way, shapeor form. What makes a community, after all, isits inherent interconnectedness, not separate-ness.

Unfortunately, most U.S. and Coloradocities are growing in an ad-hoc, fiscally irre-sponsible manner. This is due to manyconsiderations, not the least of which include:generally poor relationships between countyand city government; little state or federaloversight requiring communication and plan-ning between county and city governments; alimited tradition of regional planning; a longwestern tradition of strong private propertyrights; and a system that is fraught with poli-tics that favor short-term economic gains for afew at the expense of the community. As aresult, cities in the West generally grow in apolitically motivated and shortsighted, mannerinstead of in a manner that is communityoriented and future aware.

As a case in point, farmland and ranchlandare the historical economic roots for Montroseand all communities in this study. However, of

Pesnichak 333

all these cities, only Montrose has no plan toprotect those roots and that economic base.Durango has just passed an open space protec-tion tax, Fruita has instituted a TDR programand Boulder uses many measures to protectrural places, farmland and ranchland.According to a recent study by the AmericanFarmland Trust (AFT), the entire corridoralong Highway 50 through Montrose Countyis considered among the highest quality farm-land, yet it is facing a severe threat ofdevelopment (AFT 2002). Dennis Ericksonbelieves that if we surround Montrose withfarm land and help the farmer farm his land, itwill be a benefit to the community.

There is also a misconception thatgeographically confining a community inflatesland and real estate prices. According toDawkins and Nelson (2004): Higher prices (especially for housing) could occur ifplanning fails to increase the supply of buildable landwithin the boundary” (which suggests that adequateplanning to increase buildable land within theboundary could contain prices). Peiser (1989)observes that urban containment boundaries are“prudent land-use policies but only when accompa-nied by policies that increase urban density andintensity. Even if housing prices were to rise despiteincreasing densities, the increase itself might reflectsavings and benefits [such as in transportation costs]realized by households because of urban contain-ment.

Instead of costing money, conserving openspace as a smart growth strategy can savecommunities money. Far from being a drain onlocal taxes, farms and other types of open landactually subsidize local government by gener-ating far more in property taxes than theydemand in services (Rogers 1999). It is impor-tant to remember that ultimately, regardless ofwhat improvements a developer mayconstruct, public funds build and supportsprawl’s far-flung infrastructure: pavement,pipes, patrols, ambulances, and the other costsof unhealthy growth are paid for by taxingdrivers and non-drivers alike, whether they arethe inhabitants of sprawl or the citizens ofmore efficient environments, such as our corecities and older neighborhoods (Duany, et al.2000). As was identified in the introduction,

Montrose could literally save tens of millionsof taxpayer dollars from going to developersubsidies by instituting some sort of growthcontrol and urban definition measures. And tothis end, the taxpayer burden is possibly largerwith the current uncontrolled low-densityurban sprawl than it would be with a tax foropen space acquisition, a strong commitmentto achieving the community vision, and high-quality planning.

If land use controls were done on a state oreven federal level which encouraged orrequired cities, counties and regions to worktogether to control sprawl and work toward astable, efficient community, cities like Boulderwould not be unusual hotspots for investment.Further, many low-density suburban commu-nities might suffer lower land values becauseof poor planning, increasing traffic, deterio-rating housing stock, and loss of exclusivity;there is no greater risk to land values than unre-strained development (Rogers 1999).

It is clear that urban containment works lesswell when pursued only at the local or munic-ipal level because of the spillover effect andthe frequent creation of satellite communities(Martin et al. 2002). Boulder County has dealtwith such spillover by creating a minimum lotsize of 35 acres. Yet the true benefit to regionalplanning is that cities across the country wouldhave the impetus to create communities thatdid not exude the geography of nowhere.Because Colorado cities are primarily fundedthrough sales tax, even if you do a really goodregional plan, the employment centers may notbe where every city gets its fair share ofemployment to make it a good regional plan.So, if one town has more bedrooms then theydo employment, and they do not get therevenue for making that sacrifice, then they arenot going to go along with it (Billingsley).

Considering the foundation toward localgovernance and away from regional and stateplanning in Colorado, it is in the financialinterest of municipalities to limit the residen-tial growth potential, especially low-densityresidential growth within their communitiesand promote commercial, industrial and agri-cultural development leaving neighboring

334 Pesnichak

communities to take over the economic burdenof residential growth. Further, even thoughconserving open space as a means to ensureurban containment will save city governmentsand tax payers money in the long run, it isironic that finding the money to purchase fee-simple open space and development rights hasdifficulty gaining acceptance. In the end it is afiscally conservative approach to physicallydefine a community and the services that willbe needed by that community in the future.

Yet, according Duany et al. (2000) becausewe have rebuilt our nation every fifty to sixtyyears, it is not too late. The choice is ours:either a society of homogeneous pieces,isolated from one another in often fortifiedenclaves, or a society of diverse and memo-rable neighborhoods, organized into mutuallysupportive towns, cities, and regions. In sucha scenario, as suburbs and sprawl redevelop,there is always the opportunity to grow the citythrough higher density, create walkable neigh-borhoods, and correct past mistakes. Suchredevelopment is already occurring in someeastern U.S. cities.

Although the Montrose community maynot be as willing as some other communitiesto favor an open space as a means of urbancontainment program, the recent HouseholdSurvey and Parks Needs Assessment Surveysuggests that there is more public support thanis often given credit. The county needs to beworking with the city as partners in this growthas the city is the largest municipality in thecounty (Belt). In addition, Belt believes thatthere is general sentiment in the county forrural preservation and agrees that the lack of agrowth management plan has not been anykind of advantage. Yet, all of my intervieweesagreed that such a plan would be an uphillbattle for the Montrose community. “I thinkthis is where strong leadership comes in froma strong council that can look well down theroad and see where acquisitions like that maymake a lot of sense” (Schneiger).

In addition, many of my interviewees hadconcerns that, in the words of CommissionerAllan Belt, “the squeaky wheels are oftentimes the only ones that the elected officials

hear. And when the rest of the folks comeunglued and you have to ask where they were.John Schneiger, former Montrose CityManager had a similar perception: The thing that always concerned me with Montrosewas when we did our annual survey; it appeared asthough there would be substantial support for plan-ning and growth management. Just look at thenumbers from the survey. But people did not comeout to meetings. There are no groups on that side thatcome out. You get the NIMBYs, that is very strong. Iwas even surprised that the local environmentalgroup did not come out, hardly at all, but when youhave the real estate community and developerscoming out, it did not take very many of them to reallychange the dynamics of an issue.

According to Ron Stewart, former BoulderCounty Commissioner, One of the things we did when I was commissionerwas at least once a year do a public opinion survey.I think we have such a misconception in our societythat when you hold a public hearing you have actu-ally heard the public, you haven’t, you have heard avery thin slice of the public sentiment. And I havealways believed in serving the broader public goodand the broader public interest. I think that withoutknowing what the broader public thinks it is hard todo that. So, we used surveys to see what the publicwanted us to do. And when the slings and arrows arecoming your way, it is always comforting to knowwhat the public really wants you to be up to.

It is clear from the surveys that Montrose hasconducted in the past year that the public iswilling to listen to a community definitionplan. Although the equations to determine theexact results of such a program are outside therealm of this study, the case studies indicatethat the benefits are likely to outweigh thecosts. At this point, Montrose is at a cross-roads: the city can either continue on the samepath as it has been on throughout its historyand be willing to cope with ever-increasingtraffic congestion, a void of local agricultureand a city with decreasing livability; orMontrose can learn from the growing pains ofother municipalities in Colorado, incorporatingwhat has worked and learning from what hasnot worked. One thing is for certain, the cityof Montrose is not living up to the communityvision with its current course of action. If thecommunity decides to stay on the growth path

Pesnichak 335

it is on, it will be necessary to update thevision, leaving out any mention of agriculture,environmental protection or small town char-acter. With this knowledge and looking into thefuture at a population that could easily bedoubling within the next fifteen years,Montrose has much to gain by institutinggrowth control measures, which are reinforcedby open space protection. There is no bettertime than now to start working toward thecommunity vision. Not only will this step helpstabilize and energize the local economy forthe long run, it will also make Montrose, firstand foremost, a fiscally responsible andincreasingly livable community for futuregenerations.

ACKNOWLEDGMENTSI thank Bob Clifton, founder of the Center forNew Directions, for having the courage tomake his vision a reality and thereby creatingthe opportunity for me to explore a new andrewarding chapter of my life. I am alsothankful to: Kerwin Jensen for all his patience,motivation and understanding; KathrynCheever for her guidance and knowledge; andof course Tony Robinson, Jan Sallinger-McBride and Jim Callard for their helpfulcommentary, knowledge and guidance. Specialthanks to Graham Billingsley, BennettBoeschenstein, Steve Aquafresca, VirginiaCastro, Renee Parsons, Allan Belt, DennisErickson, Randy See, John Schneiger, EricaLewis-Kennedy, Ron Stewart and EricBergman for their time, thoughts and consid-eration in granting the interviews.

LITERATURE CITEDAmerican Farmland Trust. 2004. Farmland Infor-

mation Center, Fact Sheet, Cost of CommunityServices Studies (COCS). Retrieved July 20,2005, from http://www.farmlandinfo.org/docu-ments/27757/fs_cocs_11-02.pdf.

American Farmland Trust. (n.d.). Farming on theEdge: Sprawling Development ThreatensAmerica’s Best Farmland. Retrieved August 1,2005, from http://farmland.org/farmingontheedge/.

American Farmland Trust. (n.d.). Strategic Ranch-land in the Rocky Mountain West: Mapping theThreats to Prime Ranchland in Seven WesternStates. Retrieved August 1, 2005, fromhttp://www.farmland.org/rocky_mountain/strategi

c_ranchlands1.htm.Brown, L. R. 2006. Plan B 2.0, Rescuing a planet

under stress and a civilization in trouble. EarthPolicy Institute, W.W. Norton & Co., New York,NY.

Burchell, R., R. Dolphin, A. Downs,C. Galley, G.Lowenstein, T. Moore, S. Seskin, and K. Still.2002. Costs of Sprawl — 2000. Transit Coopera-tive Research Program Report 74. NationalAcademy Press. Washington, D.C.

BRW. 1998. City of Montrose, Colorado: Compre-hensive Plan. Unpublished Manuscript, Montrose,CO

Carlisle, Freilich, and Leitner 1997. City ofDurango, Colorado: Comprehensive Plan. Unpub-lished Manuscript, Durango, CO.

City of Boulder. (2001. Boulder Valley Compre-hensive Plan. Retrieved on August 1, 2005, fromhttp://www.ci.boulder.co.us/planning/bvcp/pdf/bvcp2002.pdf.

City of Fruita. 2000. A Vision for Fruita. City ofFruita Planning and Zoning. Retrieved on August1, 2005, from http://www.fruita.org/planning_and_zoning.htm.

Cooper, M. 1997. Can Planners Stop UncheckedGrowth Before it’s Too Late? CQ Researcher 7(37): 865–888. Congressional Quarterly, Inc., CQPress, Thousand Oaks, CA.

Cooper, M. 1999. Are Land Conservation Efforts inthe Public Interest? CQ Researcher 9 (42): 953–976. Congressional Quarterly Inc., CQ Press,Thousand Oaks, CA.

Cooper, M. 2004. Can Managed Growth ReduceUrban Sprawl? CQ Researcher. 14 (20): 469–492.Congressional Quarterly, Inc., CQ Press, Thou-sand Oaks, CA.

Coyne, W. 2003. The Fiscal Cost of Sprawl: HowSprawl Contributes to Local Governments’Budget Woes. Environment Colorado Researchand Policy Center, Denver, CO.

Daly, D. E. 2005. Economics in a Full World. Scien-tific American September: 100–103.

Dawkins, C.,and A. Nelson. 2004. Urban Contain-ment in the United States: History, Models, andtechniques for Regional and Metropolitan GrowthManagement. Planning Advisory Service. No.520. American Planning Association. Chicago, IL.

Duany, A., E. Plater-Zyberk, and J. Speck. 2000.Suburban Nation: The Rise of Sprawl and theDecline of the American Dream. North Point,New York, NY.

Engelen, R. E. 2005. The Heart of the Matter: In theFight Against Sprawl, it’s Mixed-Use Cores to theRescue. Planning, 38.

Environment Colorado. (n.d.). Responsible GrowthProject. Retrieved November 6, 2005 fromhttp://www.environmentcolorado.org/envcogrowth.asp?id2=9075.

Fruita/Mesa County. 2003. Transfer of Develop-ment Rights, User Manual. Retrieved July 12,2005, from http://www.mtsmartgrowth.org/

336 Pesnichak

CS&Rpub/ordinances/fruita-mesaTDRManual.pdf.Gehr, D., J. Gordon, H. Hoyt, P. Pollock, and J.

Raismes. Growth Management in Boulder,Colorado: A Case Study. Retrieved April 25, 2005,from http://www.ci.boulder.co.us/cao/docu-ments/x-bgmcs1.html.

Harris, J., T. Sauer, J. Swartout, B. Weygandt, andA. Yacht. 2000. Colorado’s Legacy to its Children.The Governor’s Commission on Saving OpenSpaces, Farms & Ranches, Governor’s Office onPolicy and Initiatives. Denver, CO.

Hildner, M. 2005. Resignation has City LookingToward Future [Electronic Version]. MontroseDaily Press. Accessed: 9 May 2005.

Hoch, G. 2004)0. New Urban Chic. WesternPerspective; Insight and Analysis. http://www.headwatersnews.org/p.hoch041404.html.Accessed: November 15, 2005.

Hollis, L. and W. Fulton. 2002. Open Space Protec-tion: Conservation Meets Growth Management.The Brookings Institute Center on Urban andMetropolitan Policy, Washington, D.C.

Hosansky, D. 1999. Is the U.S. Facing PermanentGridlock? CQ Researcher 9 (32): 729–752.Congressional Quarterly, Inc. CQ Press. ThousandOaks, CA.

Jenkins, M. 2001. Fruita Draws the Line AgainstSprawl. High Country News 33 (9). RetrievedAugust 16, 2005, from http://hcn.org/servlets/hcn.PrintableArticle?article_id=10503.

Kunstler, J. 1993. The Geography of Nowhere: TheRise and Decline of America’s Man-Made Land-scape. Touchstone, New York, NY.

Leisure Vision, a Division of ETC Institute. 2005.

Parks and Recreation Needs Assessment Survey.Olathe, KS.

Martin, J., R. Pendall and W. Fulton. 2002. Holdingthe Line: Urban Containment in the U.S.. TheBrookings Institute Center on Urban and Metro-politan Policy, Washington, D.C.

Moorman, J. 2004. City of Montrose: 2004 House-hold Survey. Montrose, CO.

Pollock, P. 1998. Revolutionary Ideas in Planning:Proceedings of the 1998 National PlanningConference. AICP Press. Retrieved August 14,2005, from http://www.asu.edu/caed/proceed-ings98/Pollock/pollock.html.

Roger, W. 1999. The Economic Benefits of Parksand Open Space. Retrieved August 15, 2005,from http://www.tpl.org/tier2_rp2.cfm?folder_727.

Slothower, C. 2005. 9-R Bracing for DecreasedEnrollment. The Durango Herald. Retrieved onAugust 15, 2005, from http://durangoherald.com/asp?bin/printable_article_generation.asp?article_path=/education.

U.S. Census Bureau. 2000. Profile of GeneralDemographic Characteristics: 2000. RetrievedAugust 1, 2005, from http://censtats.census.gov/pub/profiles.shtml.

Wells, S. 2005. Growing Pains: The ResponsibleGrowth Initiative Debated. Durango Telegraph, 4(20). Retrieved on August 15, 2005, fromhttp://www.durangotelegraph.com/o4-10-14/cover_story.htm.

Williamson, C. 2004. Exploring the No GrowthOption. American Planning Association,November, 34–36.

HISTORIC TRANSPORTATION CORRIDORS IN THE SOUTHWEST, 1536 – PRESENTEric Vondy

After the Spanish conquered the Aztec empirethey looked northward and saw a vast, inhos-pitable land that was sparsely populated byindigenous peoples. It was mountainous andarid. Water was scarce and often unreliable.Mountain ranges ran north to south makingeast-west travel difficult. Initial coastalsurveys turned up little to spur rapid settle-ment of the region. Eventually, however,rumors of fabulous wealth drew explorers tothe interior of this hostile land (Lorey 1999).Utilizing native guides, the explorersfollowed native trails to locations in the north,where they hoped they would find gold.Explorers like Coronado came expecting tofind an Aztec-like empire filled with uncount-able wealth but instead found dwellings madeof mud and subsistence farming (Officer,1987).

Soon silver was found in the interiorspurring development in unplanned places.The existing trails were suitable for foottraffic, but were too rugged for horse or otheranimals. For the Spaniards to settle the norththey needed to create a road network capableof horse and wagon travel. The Spaniardsraced to turn the existing foot trails into wagonroads and built new roads to reach mine sites.Communities such as San Luis Potosi,Zacatecas, and Parral were founded becauseof silver. Towns were built along the traderoutes to help protect the ore coming from themines (Lorey 1999). What would becomenorthern Sonora and Chihuahua, however,were still distant and unsettled.

As other European powers came close tothe area, Spain planned to solidify its claim on

what would become the border region (Lorey1999). Mineral interests had not been foundthis far north and so missionaries were sentinstead. The Jesuits and the Franciscans cameto convert. They followed established nativetrails north and set up missions in or nearexisting population centers. In what wouldbecome Arizona, the routes they tookfollowed the San Pedro, the Santa Cruz, andthe Colorado Rivers.

In 1691, 150 years after the Spanishinvaded the Aztec empire, Father Kino enteredpresent day Arizona for the first time. Tenyears later he had established the missions atGuevavi and Tubac. Spanish settlers were inArizona by the end of the 1730s. Often Euro-pean trade goods preceded the arrival of anySpanish as native peoples carried goodsfurther and further north (Officer 1987). Trav-eling faster than the European goods wasdisease which devastated the native popula-tion. Spanish missionaries and farmers builtsettlements using the local population as alabor source. Native trade routes becameconduits for this expansion. However, mostroads were unsuitable for wagons due to thetopography (Lorey 1999). This problemplagued all of New Spain. Many major citiesin the interior lacked rivers to keep trans-portation costs low. Thus New Spain and laterMexico was burdened with high transporta-tion costs due a lack of navigable rivers whichin turn meant there was less money to be spenton building decent roads over the ruggedterrain (Wasserman 2000). Native trails weregiven Spanish names. One example is theCamino Real. There are at least three roads

338 Vondy

called Camino Real that run from Mexico intothe United States. One is the Camino Realwhich went from Mexico City to Santa Fe viasilver mining areas and other centers(Gonzalez 1996; Moorhead 1998). One of thetwo other Royal Roads is of note. It linkedSpanish missions from the southern tip of theBaja Peninsula up the California coast termi-nating north of present-day San Francisco.Today, it is almost a connect-the-dots ofcoastal Spanish missions.

Although the primary travel corridors ofthe Spanish were oriented north-south, therewere exceptions. The Gila Trail runs alongthe Gila River through central Arizona. JuanBaptista de Anza, head of the garrison inTubac, used the trail on his journey to Cali-fornia in the 1770s. In present day NewMexico and Texas, the Rio Grande was aneast-west corridor, as was El Camino de losTejas which leaves Mexico and crosses east-ward through what would become Texas andLouisiana. In present day northern Sonoraand southern Arizona there was also theCamino del Diablo, the infamous east-westroute across the inhospitable desert (USFWS2005). The first European to use it wasMelchior Diaz during a side-trip with theCoronado expedition. Although it began inCaborca, the toughest part of the route did notstart until leaving Sonoyta, Mexico, when itbecame a long, dry ride or walk across desertflats, leg-wrenching lava hills, and driftingsand. Much of the Camino del Diablo was nota road but rather a series of interwoven tracksfollowing the lay of the land or following oldshell trails coming from the Sea of Cortez.There was very little water along this trailwhich made it particularly deadly. It was usedas a shortcut to circumvent the tribal raidingparties and bandits that often occupied thelands around the Gila River. It was heavilyused by Mexican prospectors to reach theCalifornia gold fields during the 1840s and50s. Though it was 150 miles shorter than theroute through Tucson and via the Gila River,when choosing to take it, one was opting torisk dying of thirst rather than at the hands ofIndians and bandits.

THE NINETEENTH CENTURYAmericans began creating east-west travelcorridors when the Mountain Men exploredthe region hunting for fur in the early 1800s.Following the mountain men came settlers. Inthe years following its independence fromSpain, Mexico had liberal immigration poli-cies leading to easier settlement of its lightlypopulated northern lands by Americans(Martinez 1988). From Mexico’s perspectivethere were several reasons for encouragingAmerican immigration. First, they hoped touse the American settlers as a buffer againsthostile Indians. Secondly, they hoped to staveoff American expansionism by allowing themto immigrate freely. In reality, the distance wastoo great and the Mexican population toosparse for Mexican forces to secure the border,and there was no practical way to stop theAmericans from coming. American prospec-tors came seeking mineral wealth, farmers andranchers came seeking land, and merchantscame to strengthen their economic holdings inMexico. The U.S. became the region’s primarytrading partner rather than Mexico. It becameabundantly clear that the country who settledthe region fastest would ultimately rule thenorthern territory, and that depended on whobuilt better roads.

The Spanish had no interest in trade withthe U.S. and confiscated the goods of Amer-ican merchants with whom they came incontact. It was only after Mexico’s independ-ence that trade with the Americans began. TheSanta Fe Trail opened up the region to Amer-ican expansion and Mexican trade. The SantaFe Trail originated in Missouri and in Santa Femet the Camino Real. Thus with the creationof the Sante Fe Trail it became possible for atraveler to leave Mexico City on the CaminoReal, arrive in Santa Fe, and then travel on toMissouri on the Santa Fe Trail (Moorhead1998).

By the 1830s Mexico had becomeconcerned about the growing number of Amer-ican settlers pouring into the frontier viaAmerican-built roads. Texas in particular wasa threat as Americans were about to becomethe majority there. Only a little over a decade

Vondy 339

Figure 1. The commuting habits of the residents of the cities in the study as collected bythe 2000 U.S. Census.

later, their fears came to fruition and not longafter Texas fought to gain its independence, anall-out war broke out in 1846. During theMexican-American War, a major east-westcorridor was created by the Mormon Battalionwhich left Fort Leavenworth, Kansas, andtraced a road across what was then NorthernMexico. This laid the foundation for a majoreast-west travel corridor in what was tobecome the southern portion of Arizona. Thebattalion did little road-building during its tripbut rather focused on finding routes, leavingthe retinue of over a dozen wagons to forge theroad (Jackson 1952). When hostilities ceasedin 1848, nearly half of Mexico was ceded tothe U.S. (Martinez 1988).

The California Gold Rush caused untoldthousands to pass through the border region.The population of California grew from 14,000in 1848 to 225,000 in 1852. During the peakof the Gold Rush, San Francisco’s population

was growing by 4,000 people a month.Utilizing roads like the Camino del Diablo andthe road laid out by the Mormon Battalion(both of which were built atop old Indiantrails), the mass migration to the gold fieldsbrought new prosperity to the borderlands.Agriculture and commercial developmentfollowed the prospectors on their migration.Many travelers abandoned their quest for Cali-fornia and settled along the route (Lorey 1999).

After the Mexican-American War and theGadsden Purchase of 1853, there was a majormigration of Mexicans from the U.S. acrossthe newly formed border into Mexico. Newsettlements developed on both sides of theborder. New roads were built in addition to thetraditional routes to connect lines of commu-nication (Figure 1). Mexico having seen thatits scant population and lack of roads had ledto the loss of half its land was anxious to popu-late and build an infrastructure in its new

340 Vondy

border region to protect against further expres-sions of American manifest destiny (Lorey1999).

The growth of American east-west trans-portation corridors also grew after the war withthe creation of stagecoach lines such as theshort-lived Jackass Mail (Pearce 1969) and theButterfield Overland Stage Coach. Thesecompanies followed the trail blazed by theMormon Battalion, improving upon its routewhen more accessible passages were found.Stage stops and communities sprung up withthe development. It was now as easy to traveleast-west as it had been to travel north-souththrough the border region.

The boom period stimulated by the Cali-fornia Gold Rush came to an end with thebeginning of the U.S. Civil War. Even thoughU.S. border areas had little actual combat(Arizona, for example, had only one smallskirmish), troops were deployed back East andthe war stopped most migration. With financesand energies being focused on the war in theEast, stage coach routes fell into disuse anddevelopment of the railroad stopped eventhough routes had been chosen by the 1850s(Janus Associates 1989). Growth of the westand southwest picked up again after the warand a few years of financial recovery.

Profirio Diaz came to power in Mexico in1876. His 34-year rule brought an unprece-dented period of stability to the region but withdire consequences due to his ‘pan or palo’ —bread or club — method of governance. Whilethe borderlands advanced in some ways, inothers they remained unchanged. While atten-tion was paid to enhancing the rail network,roads, which benefited the common man, wereleft largely unchanged. In all of Mexico therewere only 400 miles of rail when Diaz tookpower. By the end of his reign there were over15,000 miles (Wasserman 2000). This growthwas brought about by Diaz encouragingforeign investment. By the twentieth century,80% of the stock in Mexican railroads wasowned by Americans.

The new rail lines dramatically loweredtransportation costs — solving a problem thathad plagued the country since Spain had

conquered it 300 years earlier. By connectingthe isolated borderlands with the Mexican inte-rior, the railroads aided in eroding theautonomy of the region (Meyers and Beezley2000). The railroad allowed the Mexicanborder region to rely more on the interior thanit had previously been able to. Land valuesskyrocketed. Unique values, work ethics, andcultural practices were irrevocably changed bythe connection to the Mexican interior. Alter-nately, connection to the U.S. also broughtchanges such as access to higher wages, expo-sure to labor unions, and new political ideas.

In 1877, on the American side of the border,Southern Pacific made a deal with Texas &Pacific to build a railroad that roughlyfollowed the Butterfield Overland route. InMarch of 1881 the Santa Fe Railroadconnected with the Southern Pacific rail linein Deming, New Mexico, thus creating thesecond transcontinental rail line in the UnitedStates. In December of that same year whenSouthern Pacific eventually linked with Texas& Pacific outside of El Paso, a third transcon-tinental link was created (Janus 1989).

As railroads stretched across the AmericanSouthwest and the Mexican Northern Frontier,linking them became a primary goal. Thislinkage represents the first major developmentin north-south transportation since the SantaFe Trail linked to the Camino Real fifty yearsearlier. With the linking of the U.S. andMexican railroad networks, growth along theborder boomed. El Paso, for example becamethe biggest smelter site in the Southwest,processing ore from Texas, New Mexico,Arizona, and Chihuahua. The prosperity fromthe railroad in San Diego led to the foundingof Tijuana in 1889. Both Nogaleses werefounded in response to the railroad along theSanta Cruz River on an old native trail. Afterthe turn of the century, Mexicali and Calexicowere founded due to railroad development(Kearney & Knopp 1995). Many Mexicanborder towns strengthened economic ties to theU.S. to the point of being once again moredependent on the U.S. than on Mexico. Withthe coming of the railroads, commodities likesilver, copper, coal, lumber, cotton, wool,

Vondy 341

cattle, and salt could reach distant markets forthe first time.

The connection of Mexican railroads toAmerican railroads allowed for the first timethe mass migration of laborers from interiorMexico to the border region and on to the U.S.(Kearney and Knopp 1995). In the 1880smigrant labor forces were brought in by rail topick cotton, work in mines and smelters, andon ranches.

The railroads often followed the old trails.In Arizona, the railroads built tracks whichroughly followed the Butterfield Stage Linewhich had roughly followed the MormonBattalion’s route across Arizona which in turnhad followed native trade routes. The railroadthat crossed the border at Nogales was on atravel corridor along the Santa Cruz Riverthat had been used for hundreds of years.There were exceptions. The Cienega CreekCorridor, for example, is an area in Arizonabetween Benson and Vail through which theMormon Battalion, the Butterfield, and theSouthern Pacific Railroad lines traveled yetthe area has little evidence of being a Spanishor prehistoric travel corridor. Until Americaninterests spurred the need for more east-westtravel routes, there had been no need totraverse this region despite it having a steadywater source.

Mining and agriculture spurred new trans-portation veins. Rail lines extended to meet theneeds of the agriculture and mining industries.Mining companies were often involved withrailroads. Phelps Dodge Company, forexample, had controlling interest in the El Paso& Southwestern Railroad (Schwantes 2000).

TWENTIETH CENTURYDuring the early decades of the 20th centurythe divergence in transportation developmentbegan between Mexico and the United States.In 1910, the Mexican Revolution began.During the Revolution, transportation wasdisrupted by fighting and bandits. The rail lineswere targeted so that by the end of the revolu-tion Mexico had lost half of its locomotives.Revolutionaries also raided crops to feed theirsoldiers. Ore shipments were seized and mines

were destroyed. During the same period, theU.S. transportation infrastructure continued togrow from mining, federal efforts at reclama-tion, development of national parks, as well asthe new oil industry in Texas and California.The population on the American side of theborder grew as well as Mexicans fled thefighting in their country.

As the Mexican Revolution ended, WorldWar I began. American copper was in muchhigher demand than Mexican silver. Mexicohad also lost a valued trading partner inGermany throughout the war years. Miningefforts in Mexico were further hindered byembargos on certain explosives to Mexico. Inpart this was due to a telegram which was sentby the German Foreign Minister to the Germanminister in Mexico, suggesting a German-Mexican alliance to regain Arizona, NewMexico, and Texas should the U.S. enter thewar (Degregorio 1984). Meanwhile Americanreclamation projects continued to open vastnew areas of desert to be irrigated and turnedinto farmland.

By the second decade of the twentiethcentury, the race was on to create road systems.Most rural roads were still unfit for automo-biles. Organizations sprung up about the U.S.promoting various automobile highwaysincluding several vying for the first ocean toocean highway. Usually these routes proposedwere very similar to each other due to the lackof roads on which automobiles could actuallytravel thus the Bankhead Highway, the OldSpanish Trail, the Coast to Coast Highway, theAll Year Southern Route, and the ScenicSunshine Route all encompassed big chunksof what would become US-80 (Finley 1997).Either San Diego or Los Angeles was usuallythe western terminus of these proposedsouthern routes. Savannah, Georgia was oftenproposed as the eastern terminus. Even afteranother highway was chosen as the firstOcean-to Ocean Highway, these organizationscontinued to promote their routes as beingopen all year since there was little winterweather with which to contend.

A major goal of those that promoted thissouthern route was to bridge all obstacles that

342 Vondy

could prevent it from being open all year.Bridges were built over rivers prone toflooding such as the Gillespie Dam Bridge onthe Gila River and a plank road was built overthe sand dunes in California (Bates, 1970;Keane and Bruder 2003).

Prohibition in the U.S. lasted from 1920 to1933. Although Mexican border towns hadbegun marketing themselves as party destina-tions in the late teens, it was prohibition thatmade them boom. During prohibition the U.S.border towns became tourist destinationsadvertising the saloons and nightclubs justacross the border in Mexico where alcohol wasstill legal. Many Mexican nightclubs wereowned by American investors. The Mexicantourism industry exploded for the first timeduring this period. Conventions and confer-ences came to U.S. border towns for the accessto Mexican nightlife. Once again, the Mexicanborder region strengthened its economic tieswith the U.S. as the Americans enhanced theirtransportation systems to carry more and morepeople to the border so that they could do inMexico what they couldn’t do in the U.S.(Kearny and Knopp 1995).

The original design for the U.S. highwaysystem came from a map created in 1922 byGeneral Jack Pershing who selected roads thatwould be important to the Department of War.The original intention of the highway systemwas for defense. Pershing, having made mili-tary incursions into Mexico in pursuit ofPancho Villa during the revolution, maderoads running along the Mexican border ahigh priority but made roads running along theCanadian border a lesser priority as he did notconsider Canada a threat to the United States.Pershing’s map contained over 78,000 milesof road deemed by the Department of War tobe strategically important to the nation’sdefense or to aid with evacuations of largecities. Bureaucrats then cut down the plan toroughly half of what Pershing had recom-mended. This became the prototype for theFederal Highway System (Moon 1994) inwhich existing roads were included, highwaynames were replaced by a numbered system,and signage was standardized.

In one sense the transportation boom endedwith the Great Depression of the 1930s.However, federal road building projects in theU.S. greatly enhanced the quality of vehiculartransportation. Hundreds of millions of dollarswas being poured into highway development.In Mexico reclamation projects, merchantassociations, and free trade zones were utilizedto offset the effects of the Depression. Whiletransportation was also utilized it did not playas large a role as it did on the U.S. side of theborder. Workers surged from the Mexican inte-rior looking for work in the United States.Matamoras’s population, for example, doubledduring the course of the decade mainly due topeople seeking work north of the border(Kearny and Knopp 1995).

World War II brought more federal aid tothe Southwest states. The high-tech anddefense industries as well as coastal shipbuilding played important roles. Hundreds ofthousands of soldiers came to the Southwestfor training at Patton’s Desert Training Center.New roads linking the U.S. to Mexico werebuilt during the war including one utilizing theold shell trails that linked the U.S. to the Seaof Cortez which would be used as an emer-gency port if the Japanese blockaded theCalifornia coast. Mexican tourism boomedagain as soldiers on leave and civilians visitedMexico’s border towns to partake in itsnightlife.

With the boom in manufacturing, thegrowth of agriculture, and the loss of workerswho left to fight in the war, there was also greatdemand in the U.S. for fresh labor — thebracero program became the answer. Initiallythis program was to be short lived and used totransport Mexican laborers to agricultural areasand to help with the U.S. war effort. Each year,tens of thousands of Mexican men came tofind seasonal work in the United States. Theprogram proved so successful that it continuedinto the 1960s. In the off season many of theworkers returned to the border region ratherthan their interior homes. Their families oftenjoined them causing the border region to boomonce again. With the bracero program theborder could be opened and shut like a valve

Vondy 343

allowing cheap labor to pour into the U.S. asneeded. During the Korean War and theVietnam War, in addition to World War II,Mexican labor was in strong demand(Martinez 1988). When the program endedMexicans continued coming.

In 1956 President Dwight Eisenhowerauthorized the Interstate Highway System. Apossibly apocryphal story involved Eisenhowerbeing so impressed with Hitler’s Autobahnduring the Second World War that it acted asthe inspiration for the Interstate HighwaySystem. While there may be some truth in thisstory, Eisenhower, in his early military career,had been a truck driver and understood theconditions of the American road network. Healso used Pershing’s map from 1922 for thenew road system (Degregorio 1984).

The Interstate system serves the dualpurpose of being designed for traffic flow andto aid in the evacuation of major cities in caseof nuclear attack using a procedure calledcontraflow wherein in-bound lanes arereversed so that all lanes lead out of a city. Theeffectiveness of this second purpose was testedduring 2005 with the evacuation of NewOrleans due to Hurricane Katrina and the evac-uation of Houston due to Hurricane Rita.

In Mexico, the 1950s also saw the devel-opment of major highway systems linkingcities on the border to major interior cities.Smaller border towns like Nogales were alsolinked to major cities if they were located onmajor transportation routes. Deep in the inte-rior, highways were created linking majorcities to the growing beach resorts on the coast.In the latter half of the twentieth century, thereclaimed lands that had made the borderregions an agricultural center were becomingsuburbs of growing cities. The old Hohokamcanal systems that were excavated whenPhoenix was founded are now mostly coveredby subdivisions. “It is striking that the growthof these cities has had little relationship to theearlier agricultural economy,” Oscar Martinezsaid. Millions have now moved into the regionthat have no relationship to the natural envi-ronment or agricultural environment that madethe region habitable (Martinez 1988).

SUMMARYWilliam Langewiesche (1993) said, “Theborder itself was no longer a distant desert, butrather a string of familiar cities closely linkedto the rest of Mexico by improved roads andpublic transport.” As north-south transporta-tion needs grew, the highway systemmimicked the rail lines which in turn werefollowing the Spanish trails that had followedthe tribal trails. The primary differences wereelevation and water. Foot travel could reachplaces that horses could not. Automobiles andtrains needed road and track on certain gradesbut could go longer without water than humanand horse.

The U.S. Highway System followed oldtrails as well. Old wagon roads were replacedby primitive dirt roads. In turn, those dirt roadswere replaced by narrow paved roads. TheMormon Battalion Road was used by theButterfield Overland Stage Coach Routewhich was then used by the Great SouthernRoad which was then used for US-80 and laterInterstate 10. You stand at certain places andsee the layers of transportation history. You canstand on the San Pedro River where Coronadofollowed old native trails and see where theMormon Battalion and the Butterfield trailsmet and you can see I-10 and the railroad andknow that US-80 met these other systems onlya couple of miles away. You can stand inNogales and see the wall that splits the railroadand the road and then follow the Santa CruzRiver and I-19 all the way to Tucson.

LITERATURE CITEDBates, J. B. 1970. The Plank Road. The Journal of

San Diego History 16 (2) [http://www.sandiego-history.org/journal/70spring/plank.htm]

Degregorio, W. A. 1984. The Complete Book ofU.S. Presidents. Gramercy Books, New York, NY.

U.S. Fish and Wildlife Service. 2005. El CaminoDel Diablo [http://www.fws.gov/southwest/refuges/arizona/diablo.html]

Finley, E .J. 1997. The Old US-80 Highway Trav-elers Guide (Phoenix to San Diego) Narrow RoadCommunications, Phoenix, AZ.

Gonzalez, M. 1996. The Camino Real: Our tie withMexico, yesterday, today, andtomorrow. ResourcesMagazine Issue XV, New Mexico State University,Las Cruces, NM. [http://www.cahe.nmsu.edu/pubs/resourcesmag/spring96/camino.html]

344 Vondy

Jackson, W. T. 1952. Wagon Roads West. Univer-sity of California Press, Berkeley, CA.

Janus Associates, Inc. 1989. Transcontinental Rail-roading in Arizona: 1878–1940. State HistoricPreservation Office, Phoenix, AZ.

Keane, M. and J. S. Bruder, 2003. Good RoadsEverywhere: A History of Road Building inArizona. Arizona Department of Transportation,Phoenix, AZ.

Kearney, M. and A. Knopp. 1995. Border Cuates.Eakin Press, Austin, TX.

Langewiesche, W. 1993. Cutting for Sign. PantheonPress, New York, NY.

Lorey, D. E. 1999. The U.S.-Mexican Border in theTwentieth Century. Scholarly Resources, Wilm-ington, DE.

Martinez, O. 1988. Troublesome Border. Universityof Arizona Press, Tucson, AZ.

Meyer, M. C. and W. H. Beezley (eds.) 2000. The

Oxford History of Mexico. Oxford UniversityPress. Oxford, England.

Moon, H. 1994. The Interstate Highway System.The Association of American Geographers, Wash-ington, D.C.

Moorhead, M. L. 1998. New Mexico’s Royal Road.University of Oklahoma Press, Norman, OK.

Officer, J. E. 1987. Hispanic Arizona 1536–1856.University of Arizona Press, Tucson, AZ.

Pearce, B. C. 1969. The Jackass Mail: San Antonioand San Diego Mail Line. The Journal of SanDiego History 15(2) [http://www.sandiegohis-tory.org/journal/69/spring/jackass.htm]

Schwantes, C. A. 2000. Vision and Enterprise:Exploring the History of the Phelps Dodge Corpo-ration. The University of Arizona Press, Tucson.

Wasserman, M. 2000. Everyday Life and Politics inNineteenth Century Mexico. University of NewMexico Press, Albuquerque, NM.

CONTRIBUTORS

Barbara N. AlbertiSuperintendentWar in the Pacific National Historic Park135 Murray Blvd., Suite 100Hagatna, Guam [email protected]

Eric W. AlbrechtSchool of Natural Resources

and the EnvironmentUniversity of ArizonaTucson, AZ 85721

Pamela AnningData ManagerSonoran Institute900 E. Hilltop Ave.Flagstaff, AZ [email protected]

C. Dustin BeckerLife Net6423 South Bascom TrailWillcox, AZ [email protected]

Craig C. BillingtonSchool of Natural Resources

and the EnvironmentUniversity of ArizonaTucson, AZ 85721

Gitanjali BodnerThe Nature Conservancy1510 E. Fort Lowell Rd.Tucson, AZ 85719

Ken BoykinUSGS Cooperative Fish and Wildlife

Research UnitNew Mexico State UniversityLas Cruces, NM 88003

Melanie BucciEcology and Evolutionary BiologyUniversity of ArizonaTucson, AZ [email protected]

Debbie C. Buecher Buecher Biological Consulting7050 E. Katchina CourtTucson, AZ [email protected]

Alejandro E. CastellanosDICTUSUniversidad de SonoraApdo. Postal No. 54Hermosillo, Sonora 83000 [email protected]

Reyna A. Castillo DICTUSUniversidad de SonoraApdo. Postal No. 54Hermosillo, Sonora 83000 Mexico

Helí Coronel-ArellanoUniversidad Autonoma de QueretaroPostgrado Regional de Recursos BioticosCentro Unversitario S/N,Colonia Las CampanasQueretaro, Queretaro CP 7600 Mexico

346 Contributors

Melanie CulverSchool of Natural Resources

and the EnvironmentUniversity of ArizonaTucson, AZ [email protected]

Kathleen DochertyU.S. Institute for Environmental

Conflict Resolution130 S. Scott Ave.Tucson, AZ [email protected]

Mac DonaldsonEmpire Ranch LLPP.O. Box 57Sonoita, AZ 85637

Douglas K. DuncanU.S. Fish and Wildlife ServiceArizona Ecological Services Office201 North Bonita, Suite 141Tucson, AZ [email protected]

Erin FernandezU.S. Fish and Wildlife ServiceArizona Ecological Services Office201 North Bonita, Suite 141Tucson, Arizona 85745

Peter F. FfolliottSchool of Natural Resources

and the EnvironmentUniversity of ArizonaTucson, AZ 85721 [email protected]

Randy GimblettSchool of Natural Resources


David GoriDirector of ScienceThe Nature Conservancy in New Mexico212 E. Marcy St.Santa Fe, NM [email protected]

Gerald J. GottfriedRocky Mountain Research StationUSDA Forest ServicePhoenix, AZ 85006

William HalvorsonUSGS Southwest Biological Science CenterSonoran Desert Research Station123 Family and Consumer SciencesUniversity of ArizonaTucson, AZ [email protected]

Lisa HaynesSchool of Natural Resources


John L. KoprowskiSchool of Natural Resources

and the EnvironmentUniversity of Arizona, Tucson, AZ [email protected]

Paul R. KrausmanWildlife Biology ProgramCollege of Forestry and Conservation32 Campus DriveUniversity of MontanaMissoula, MT 59812

Amy J. KuenziDepartment of BiologyMontana Tech of the University of MontanaButte, MT 59701

Contributors 347

Carlos A. López-GonzálezLicenciatura en BiologiaUniversidad Autonoma de QueretatoCentro Universidaria S/NColonia Las CampanusQueretaro, Queretaro 76010 [email protected]

Curtis McCaslandCabeza Prieta National Wildlife Refuge1611 North Second AvenueAjo, AZ [email protected]

Cristina MelendezComision de Ecologia y DesarrolloSustentable del Estado de Sonora Hermosillo, Sonora, [email protected]

John R. MorgartU.S. Fish and Wildlife ServiceNew Mexico Ecological Services

Field Office2105 Osuna N.E.Albuquerque, NM 87113

Kerry L. NicholsonSchool of Natural Resources

and the EnvironmentUniversity of ArizonaTucson, AZ 85712 [email protected]

Ami C. PateNatural Resources ManagementOrgan Pipe Cactus National Monument10 Organ Pipe DriveAjo, AZ [email protected]

David Pesnichak308 Beaufort Ave.Livingston, NJ [email protected]

Tony PovilitisLife Netc/o Maui Forest Birds Recovery Project2465 Olinda Rd.Makawao, HI 96768 [email protected]

Brian F. PowellProgram ManagerPima County Office of Conservation

Science and Environmental Policy201 N. Stone Ave, 6th FloorTucson, AZ [email protected]

Adrian QuijadaSchool of Natural Resources


J. Daren RiedleDepartment of Life, Earth, and

Environmental ScienceWest Texas A & M UniversityCanyon, TX [email protected]

James C. RorabaughU.S. Fish and Wildlife ServiceArizona Ecological Services Office201 N Bonita Ave, Suite 141Tucson, AZ [email protected]

Philip C. RosenSchool of Natural Resources


348 Contributors

Cecilia A. SchmidtHDMS Wildlife Data SpecialistWildlife Management Division,

Habitat BranchArizona Game and Fish Department5000 W. Carefree HighwayPhoenix, AZ [email protected]

Keith A. SchulzNatureServe4001 Discovery Drive, Suite 270 Boulder, CO [email protected]

Heather SchussmanThe Nature Conservancy114 N. San Francisco St., Suite 205Flagstaff, AZ 86001

Cecil SchwalbeUSGS Southwest Biological Science CenterSonoran Desert Research Station123 Family and Consumer SciencesUniversity of ArizonaTucson, AZ [email protected]

Ronnie SidnerEcological Consulting1671 N. CliftonTucson, AZ 85745 [email protected]

Karen SimmsBureau of Land Management12661 E. Broadway Blvd.Tucson, AZ [email protected]

William A. SpriggDepartment of Atmospheric SciencesUniversity of Arizona1118 E. 4th StreetTucson, AZ 85721 [email protected]

Thomas R. StrongWestLand Resources, Inc.4001 E. Paradise Falls Dr.Tucson, AZ [email protected]

Don E. SwannSaguaro National Park3693 South Old Spanish TrailTucson, AZ 85730 [email protected]

Kathryn A. ThomasUSGS Southwest Biological Science CenterSonoran Desert Research Station123 Family and Consumer SciencesUniversity of ArizonaTucson, AZ [email protected]

Charles van Riper IIIUSGS Southwest Biological Science CenterSonoran Desert Research Station123 Family and Consumer SciencesUniversity of ArizonaTucson, AZ [email protected]

Cora VarasSchool of Natural Resources


Eric VondyArizona State Parks1300. W. Washington StreetPhoenix, AZ [email protected]

William E. WernerArizona Department of Water Resources3550 N. Central AvenuePhoenix, AZ [email protected]

Contributors 349

Ryan R. WilsonDepartment of Wildland ResourcesUtah State UniversityLogan, UT [email protected]

Dazhong YinDepartment of Atmospheric Sciences University of Arizona1118 E 4th StreetUniversity of ArizonaTucson, AZ 85721

INDEX+ after a page number indicates a discussion of the topic through the rest of the chapter

Aabandoned mines, 93acacia, 43, 56, 112Acacia spp. 112adaptive management, 131+African cichlid, 153agave, 56Agosia chrysogaster, 153, 210, 213Agosia sp., 210agricultural fields, 11Agricultural Research Service, 1Agua Dulce, 144+air-borne particulate matter pollution, 303+aircraft noise pollution, 44air pollution, 303+airborne dust, 303+Ajo Mountain scrub oak, 43Algodones Dunes, 189Allen’s big-eared bat, 96Alligator mississipiensis, 185alligator juniper, 74Altar Valley, 283Ambrosia deltoidea, 112 Ambrosia dumosa, 172, 182Ambrosia pumila, 263Ambystoma mavortium, 184, 199Ameiurus melas, 153, 210, 213, 218American alligator, 185American badger, 232American bullfrog, 33, 200, 214, 217animal poaching, 44Animas Valley, 71annual plant production, 76anthropogenic threats, 294Antilocapra americana sonoriensis, 123Apache Highlands ecoregional, 57, 132Apalone spinifera, 186, 210

Arizona Bell’s vireo, 34 Arizona cotton rat, 231, 232, 234, 236Arizona Department of Water Resources,

162Arizona Electric Power Cooperative, 162Arizona Game and Fish Commission, 162Arizona Game and Fish Department, 1, 128,

143, 162, 164, 266, 275, 283+Arizona gray squirrel, 232Arizona mud turtle, 184Arizona Power Authority, 162Arizona rosewood, 43Arizona Sonoran Desert Museum, 110Arizona sycamore, 56Arizona white oak, 57Arizona Wildlife Linkage Workgroup, 297arrowweed, 149Arundo donax, 149Ash, 254Astragalus geyeri var. triquetrus, 164Atriplex polycarpa, 148

Bbadger, 17, 97Bailey’s pocket mouse, 96Baiomys taylori, 230, 231Baja California, 173, 177banded rock rattlesnake, 98banner-tailed kangaroo rat, 97, 230Bard Water District, 162barrel cactus, 56Barry M. Goldwater Range, 182, 283Basic Water Company, 163bats, 95+, 233, 253+beautiful shiner, 210 bermuda grass, 149, 150big brown bat, 96

352 Index

big free-tailed bat, 96big-eyed leopard frog, 210, 221bighorn sheep, 283 Bill Williams River, 165biodiversity, 225, 235black bear, 97, 275+, 293black bullhead catfish, 153, 156, 210, 218black-headed snake, 98black-necked gartersnake, 63, 204, 214black-tailed jackrabbit, 96, 232black-tailed rattlesnake, 98, 203bobcat, 97, 112, 293, 294bonytail, 161, 164, 168border security, 237Borderlands Management Task Force, 129Botta’s pocket gopher, 232boxelder, 57Brassica tournefourtii, 191Brazilian free-tailed bat, 96brittle bush, 254broom seepweed. 148brush mouse, 96, 231Buenos Aires National Wildlife Refuge, 109buffelgrass, 193bulrush, 149, 150Bureau of Indian Affairs, 162Bureau of Land Management, 111, 131, 162,

288, 326Bureau of Reclamation, 162burrobush, 43 bursage, 112Bursera confusa, 242Bursera microphylla, 172, 182Butterfield Overland Stage Coach., 340+

CCabeza Prieta Mountains, 123Cabeza Prieta National Wildlife Refuge, 109,

182, 283, 287cactus ferruginous pygmy-owl, 112cactus mouse, 96Calabazas, 24+California black rail, 164California Endangered Species Act, 168California Environmental Quality Act, 168California leaf-nosed bat, 96California myotis, 96Callisaurus draconoides, 186, 201

Camino del Diablo, 338, 339Camino Real, 337, 340Campostoma ornatus, 210, 213Canis latrans, 112, 293Canis lupus, 234, 293Canis lupus baileyi, 263canyon mouse, 96canyon spotted whiptail, 63canyon treefrog, 63, 210captive breeding, 284+Carlsbad Caverns National Park, 94+Carnegiea gigantea, 112, 171+Casa Grande Ruin National Monument, 7+Catostomus bernardini, 210, 213Catostomus latipinnis, 164Catostomus wiggensi, 210, 213Cascabel Watersheds, 71+cave biodiversity, 95+cave myotis, 96cave vertebrates, 96+caves, 93+Center for Biological Diversity, 267, 268Central Arizona Water Conservation District,

162Cercidium microphyllum, 172, 176, 242Cercidium praecox, 242Cercidium spp., 254Cervus elaphus, 265Chaetodipus baileyai, 245+Chaetodipus hispidus, 230Chaetodipus intermedius, 230Chaetodipus penicillatus, 230, 246+chaparral, 112checkered gartersnake, 204Chilomeniscus stramineus, 186, 202Chiricahua leopard frog, 123Chihuahuan Desert, 93+cholla, 56, 112, 254 Cibola Valley Irrigation and Drainage

District, 162Cichlid sp, 210Cienega Creek, 132Cienega Creek Corridor, 26.5Clark’s spiny lizard, 98, 201Coachella Valley Water District, 162coachwhip, 186, 202coati, 97Coccyzus americanus occidentalis, 164Colaptes chrysoides, 164

Index 353

Coleonyx variegatus,15.6collaborative resource managementcollared peccary, 97, 112, 265Colorado Desert fringe-toed lizard, 189Colorado pikeminnow, 161, 162, 164Colorado River, 123, 161+, 181+Colorado River Board of California, 162Colorado River Commission of Nevada, 162Colorado River cotton rat, 164Colorado River flows, 163common black hawk, 63common carp, 210, 211 common chuckwalla, 47, 184, 187, 191, 192,

201common hog-nosed skunk, 97common hoptree mixed scrub, 43 common kingsnake, 47, 63, 186, 203common porcupine, 234common side-blotched lizard, 202community vision, 322+coniferous forest, 57conservation, 181+, 263+, 283+, 295+, 329conservation management status, 82Copper Mountain, 123Coronado National Forest, 71, 233, 253Coronado National Memorial, 72+, 225+corridors, 294+Coryphantha scheeri var. robustispina, 123cotton rat, 231, 234cottonwood, 254Couch’s spadefoot, 63, 200Coues white-tailed deer, 265cougar, 112, 293+Cougar Management Guidelines Working

Group, 293+coyote, 97, 112, 293 crayfish, 217creosote bush, 43, 47, 56, 112Crotalus willardi obscures, 123Cynodon dactylon, 149Cyprinella Formosa, 210Cyprinodon eremus, 153, 210Cyprinus carpio, 210, 213

DDepartment of Homeland Security, 110, 126Davis Dam, 162, 164, 166deer mouse, 231

Dendroica petechia sonorana, 164Department of Energy, 162Department of the Interior, 162Dermochelys coriacea, 263desert bighorn, 97desert chub, 210desert cottontail, 96, 233desert dust modeling, 303+desert fishes, 31+, 143+ desert grassland, 132, 135, 140desert horned lizard, 47desert iguana, 201desert kangaroo rat, 97desert night lizard, 184, 187, 202desert shrew, 96, 230+desert spiny lizard, 98, 202desert tortoise, 98, 112, 161, 164, 184, 191+,

200desert wood rat, 97Didelphis virginiana, 232Dipodomys merriami, 230, 231Dipodomys ordii, 230, 231Dipodomys spectabalis, 230, 231Dipsosaurus dorsalis, 184, 201Douglas fir, 57drug smuggling, 124+, 130dust concentrations, 303+ dust patterns, 311, 313, 316Dust Regional Atmospheric Modeling, 303dust storm, 303+

Eeastern cottontail, 230, 233, 237eastern red bat, 96economic growth, 332ecosystem conservation, 263El Camino de los Tejas, 338elephant tree, 182elf owl, 164elk, 265Emory oak, 57, 74, 77, 78Empidonax traillii extimus, 161, 164Encelia farinosa, 142, 174, 243, 254Endangered Species Act, 123, 126, 143, 161,

168, 263, 283epidermal browning, 171+Eriogonum viscidulum, 164Erithizon dorsatum, 234

354 Index

Eruca vesicaria, 191Eumops underwoodi, 150

Ffathead minnow, 210fire, 57, 58, 62flannelmouth sucker, 164, 165, 166flat-tailed horned lizard, 164, 190-191, 194,

201Florida panther, 263forage enhancement plots, 286, 289-290Fouquieria macdougalii, 242Fouquieria splendens, 112, 142, 243fragmentation, 267, 269, 293+Fraxinus pennsylvanica, 254fremont cottonwood, 56fringed myotis, 96fulvous harvest mouse, 231

GGambel oak, 57Gambelia wislizenii, 186, 201Gambusia affinis, 153, 210, 213GAP Analysis Program, 81Gastrophryne olivacea, 150, 199, 210, 212giant reed, 149giant sacaton, 137Gila cypha, 161, 162, 164, 263Gila ditaenia, 210, 213Gila elegans, 161, 164Gila eremica, 210, 213Gila monster, 98, 184, 191-192, 201Gila River, 7+, 181+, 284, 338Gila River Basin, 218, 221Gila topminnow, 210Gila Trail, 338Gila woodpecker, 164 gilded flicker, 164 Glaucidium brasilianum, 112Glen Canyon Dam Adaptive Management

Program, 165, 166glossy snake, 202Gnathamitermes tubiformans, 171Golden Shores Water Conservation District,

162Goodding willow, 149, 254Goode’s horned lizard, 201

gophersnake, 98, 203Gopherus agassizii, 112, 161, 164, 184, 200Gran Desierto, 181+Grand Canyon, 166gray fox, 97gray wolf, 234gray-footed chipmunk, 96Great Basin conifer woodland, 43Great Plains narrow-mouthed toad, 63, 150,

199, 210green sea turtle, 187, 200green sunfish, 210, 218Guadalupe Mountains, 103Guevavi, 24+Gulf of Califorina, 181+

Hhabitat conservation, 268+habitat destruction, 205habitat loss, 267+, 293harm, 126Harris’ antelope squirrel, 96hawksbill sea turtle, 186, 187, 200Heloderma suspectum, 184, 201herbaceous understories, 71+ herpetofauna, 181+, 205+Heterotermes aureus, 171+highway mortality, 44, 48Hispid cotton rat, 97hispid pocket mouse, 230hoary bat, 96Hohokam, 8honey mesquite, 43Hoover Dam, 162-164house mouse, 229, 230, 231Huachuca Mountains, 72+, 226, 233, 238Huachuca water umbel, 123humpback chub, 161+, 263Hyla arenicolor, 210, 212Hypsiglena chlorophaea, 186, 203

Iillegal border crossers, 123+, 181, 190illegal immigrants, 124+Imperial Irrigation District, 161, 162Imperial Valley, 123, 283incidental take, 126

Index 355

inland saltgrass–rush association, 43Instituto del Medio Ambiente y el Desarrollo

Sustentable del Estado de Sonora, 143Integrated Pest Management, 12interior chaparral, 43interstate highway system, 343inventory, 7+, 23+, 4+1, 53+, 225+ ironwood, 43, 112, 241+Ironwood Forest National Monument, 109+Ixobrychus exilis hesperis, 164

Jjaguar, 34, 234, 238, 263, 267+jojoba, 43, 112Juglans major, 254Juniperus deppeana, 74

KKinosternon arizonense, 184Kinosternon flavescens, 184Kinosternon hirtipes, 210, 212Kinosternon integrum, 210, 212Kinosternon sonoriense, 200, 210, 212Kinosternon sonoriense longifemorale, 143,

184, 189kit fox, 97Kofa Mountains, 123

LLake Havasu, 162Lake Mead, 162 Lake Mohave, 162Lampropeltis 186, 203land cover, 81+land cover classes, 84, 86Landsat Enhanced Thematic Mapper, 82largemouth bass, 210 Larrea tridentata, 112, 172, 176, 182, 243Las Cienegas National Conservation Area,

131+Las Glorias Ranch, 242+Lasiurus blossevillii, 164Lasiurus xanthinus, 164Laterallus jamaicensis coturniculus, 164leatherback sea turtle, 186, 187, 200, 263leopard frogs, 206+

Lepomis cyanellus, 210, 218Leptodactylus melanonotus 210, 212Leptonycteris curasoae yerbabuenae, 96,

100, 112, 123Lepus californicus, 232lesser longnosed bat, 96, 100, 112, 123Lilaeopsis schaffneriana ssp. recurva, 123 Lobos Ranch, 243+loggerhead sea turtles, 186, 187, 200long-legged myotis, 96long-nosed leopard lizard, 186, 201long-nosed snake, 203long-tailed brush lizard, 202long-tailed weasel, 97longfin dace, 153+, 210Lophocereus schotti, 172, 174, 176, 242, 243Los Angeles Department of Water and

Power, 162lower Colorado River, 161+Lower Colorado River Multi-Species

Conservation Program, 161+lowland burrowing treefrog, 200, 210lowland desert rivers, 143+lowland leopard frog, 63, 98, 164, 200, 210Lynx rufus, 112, 293

MMacNeill’s sootywing skipper, 164Madrean alligator lizard, 63 Madrean evergreen forest, 43Madrean evergreen woodland, 43, 57management goals, 131 manzanita, 57maple, 57Margalef Index, 244, 247Martes pennanti pacifica, 263Masticophis flagellum, 186Mediterranean grass, 191Mediterranean house gecko, 201Melanerpes uropygialis, 164Merriam’s kangaroo rat, 97, 230, 231mesquite, 112, 177mesquite bosque, 149mesquite–palo verde plant community, 172metapopulation, 269 Meteorological Aerodrome Report, 306Metropolitan Water District of Southern

California, 162

356 Index

Mexican blue oak, 57Mexican dust source, 314Mexican gartersnake, 204, 210, 214Mexican leaf-frog, 210Mexican long-tongued bat, 96Mexican longfin dace, 210Mexican mud turtle, 210, 214, 221Mexican pinyon pine, 57Mexican pocket gopher, 96Mexican Revolution, 341Mexican spotted owl, 63Mexican treefrog, 210Mexican spadefoot toad, 63Mexican stoneroller, 210 Mexican wolf, 263+Mexican wood rat, 97Micrathene whitneyi, 164Micropterus salmoides, 210, 213Micruroides euryxanthus, 186, 203mines, 93+mitochondrial DNA, 277+Moderate Resolution Imaging

Spectroradiometer, 305Mohave County Water Authority, 162Mohave Valley Irrigation and Drainage

District, 162Mohave Water Conservation District, 162Mohawk Dunes, 189, 191Mojave rattlesnake, 63, 98, 203molecular genetic studies, 276Montezuma Canyon, 226, 230monitoring, 7+, 23+, 41+, 53+, 225+monitoring protocols, 41, 45Mormon Battalion, 339+mosquitofish, 153+, 210mountain lion, 97, 233. 238, 293+mountain patch-nose snake, 63, 98mountain sheep, 112mule deer, 97, 112, 150, 230, 232, 233, 237Mus musculus, 229, 230

Nnarcotics smuggling, 109National Aeronautics and Space

Administration, 304National Centers for Environmental

Prediction, 303

National Environmental Policy Act, 168National Park Service, 1, 7, 26, 41, 53, 162,

182, 225National Wildlife Federation, 163 Natural Resources Conservation Service, 1Nature Conservancy, 1, 132Neotoma albigula, 231, 245-248netleaf hackberry, 56Nevada Department of Wildlife, 163, 164New Mexico Department of Game and Fish,

269New Mexico ridge-nosed rattlesnake, 123nightsnake, 186, 203non-native animals, 12+, 23+, 45+, 64, 153+,

161, 184+, 205+, 229non-native plants, 7+, 23+, 44+, 53+, 181+,

226+North Gila Valley Irrigation and Drainage

District, 162northern crayfish, 205+northern goshawk, 63northern pygmy mouse, 230northwestern Mexico, 205+Notiosorex crawfordi, 230, 231

Ooak, 57oak savanna, 56, 71+oak woodland, 71+, 228+ocotillo, 43, 112Odocoileus hemionus, 112, 230Odocoileus virginianus couesi, 265off-highway vehicles, 189+off-road travel, 125Olive Ridley sea turtle, 186, 187, 200Olneya tesota, 112, 172, 177, 241+oneseed juniper, 43Onychomys torridus, 231, 246-248Opata sucker, 210 open space, 319+Opuntia spp., 112, 254 Orconectes cf. virilis, 218Orconectes virilis, 206+ Ord’s kangaroo rat, 230organ pipe cactus, 42+, 182Organ Pipe Cactus National Monument, 41+,

109, 143, 145, 156, 182, 283ornate tree lizard, 63, 202

Index 357

Ovis canadensis, 112, 283ozone, 303

PPachycereus pringlei, 172, 174, 176Pachymedusa dacnicolor, 210Pacific fisher, 263pallid bat, 96palo verde, 43, 56, 112, 254Palo Verde Irrigation District, 162Panthera onca, 234, 263Parker Dam, 162, 164Parkinsonia microphylum, 112particulate matter, 303Pecari tajacu, 112 Peloncillo Montains, 71, 123Pennisetum ciliare, 191Perognathus flavus, 230, 231, 242, 246-248Peromyscus boylii, 231, 249Peromycus eremicus, 249Peromyscus leucopus, 231Peromyscus maniculatus, 231, 249Pholisora gracielae, 164Phrynosoma mcalli, 164Phyllorynchus browni, 186, 203Pima pineapple cactus, 123Pimephales promelas, 210, 213pine oak forest, 57pine oak woodland, 57Piranga rubra, 164Platanus wrightii, 254PM10 concentrations, 303+PM2.5 concentrations, 303+pocketed free-tailed bat, 96Poeciliopsis occidentalis, 207, 210, 213ponderosa pine, 57pond slider, 184, 186, 201Populus fremontii, 254porcupine, 97power analysis, 134Pozo Hondo Ranch, 242+prickly pear, 56, 112, 19.2 pronghorn recovery plan, 283+Prosopis glandulosa, 172Prosopis velutina, 112, 149Ptychocheilus lucius, 161, 162Puerto Blanco Mountains, 148, 149Puma concolor, 112, 233, 293+

Puma concolor coryi, 263Pyrocephalus rubinus, 164

Qquaking aspen, 57Quercus arizonica, 74Quercus emoryi, 74Quercus toumeyi, 74Quitobaquito, 143+, 182, 184, 189Quitobaquito Hills, 148-150Quitobaquito pupfish, 45, 153, 157, 210Quitobaquito spring snail, 150 Quitobaquito Springs and Pond, 42, 45, 143,

150Quitovac, 143+, 189

Rraccoon, 97Ragged Top Mountain, 114-115ragweed, 43Rallus longirostris yumanensis, 123, 161,

164ranchland, 332-333rapid assessment survey, 205, 207, 221razorback sucker, 161red brome, 56, 57red-spotted toad, 63, 98, 199redbacked whiptail, 201redbelly tilapia, 207regal horned lizard, 201Reithrodontomys fluvescens, 231 Reithrodontomys megalotis, 231relict leopard frog, 164Reserva de la Biosfera Alto Golfo de

California y Delta del Río Colorado, 182+Reserva de la Biosfera El Pinacate y Gran

Desierto de Altar, 143+, 182+Rhinichthys osculus, 210Rincon Creek, 53+Rincon Mountains, 53+ring-neck snake, 63ringtail, 97Rio Grande, 338Rio Grande leopard frog, 186, 188, 199Río Sonoyta, 143+, 181+Río Sonoyta Valley, 184+riparian birds, 13+riparian corridors, 253-259

358 Index

riparian vegetation, 26-28, 45, 149rock gooseberry, 43rock pocket mouse, 96, 230rock squirrel, 96rodents, 241+Roskruge Mountains, 112rosy boa, 203rough-footed mud turtle, 210-211, 216-217,

221round-tailed ground squirrel, 234Russian thistle, 191

Ssabinal frog, 210Sabino Canyon, 253+, 294Sabino Creek, 254+saddled leaf-nosed snake, 186, 203saguaro, 54, 56, 112, 171+, 254 Saguaro National Park, 53+Sahara mustard, 191salad rocket, 191Salix gooddingii, 149, 254Salsola tragus, 191Salt River Project Agricultural Improvement

and Power District, 162saltbush, 43, 148saltcedar, 147+saltgrass, 148Samaniego Hills, 112San Bernardino National Wildlife Refuge,

219San Diego ambrosia, 263San Diego County Water Authority, 162San Judas Ranch, 242+San Pedro River, 233, 238Santa Catalina Mountains, 254, 294, 296Santa Cruz flood plain, 112Santa Cruz River, 24+, 123, 337+Santa Fe Trail, 338-340Saracachi Ciénega, 219+Sauromalus ater, 184Sawtooth Mountains, 112 Scaphiopus couchii, 216Schismus arabicus, 191Schismus barbatus, 191Scirpes americanus, 149Sciurus arizonensis, 232semi-desert grassland, 54-58, 226, 233-236

sidewinder, 47, 202Sierra los Tanques, 148Sierra Madre Occidental, 276+Sierra Pinacate, 181-182Sigmodon arizonae, 231Sigmodon arizonae plenus, 164Sigmodon hispidus eremicus, 164Sigmodon ochrognatus, 234Sigmodon spp., 232silky pocket mouse, 230, 231Silver Bell Mountains, 112, 116-121silver-haired bat, 96Simmondsia chinensis, 112sky island, 53, 54, 55, 62, 226, 275+Sky Island Alliance, 1small-footed myotisSmilisca baudini, 210, 212Smilisca fodiens, 210, 212Smith’s black-headed snake, 203Sonoita Valley Acquisition Planning District,

131-133Sonoita Valley Planning Partnership, 132Sonora chub, 210 Sonoran collared lizard, 201Sonoran coralsnake, 63, 186, 203Sonoran Desert Conservation Plan, 297Sonoran Desert Network, 18, 35, 47, 66Sonoran Desert pocket mouse, 230Sonoran Desert toad, 63, 186+, 200, 210Sonoran desertscrub, 11, 43, 56, 254-256Sonoran green toad, 150, 199Sonoran interior marshland, 43Sonoran mud turtle, 200, 211, 214, 216+Sonoran pronghorn, 123+, 150, 158, 283+Sonoran shovel-nosed snake, 202Sonoran whip-snake, 202Sonoran yellow warbler, 164 Sonoyta mud turtle, 143+Sonoyta River, 143+, 181+, 284Sonoyta Valley, 184+Southern California Edison Company, 162Southern California Public Power Authority,

162southern cattail, 149, 150southern cattail–chairmaker’s bulrush

association, 43southern grasshopper mouse, 231Southern Nevada Water Authority, 163Southern Pacific Railroad, 340-341

Index 359

southern pocket gopher, 230, 232Southwest Regional GAP Analysis Project,

81Southwest U.S. dust sources, 303+southwest willow flycatcher, 161, 164-165,

168southwestern deciduous riparian forest, 56Spanish colonial missions, 24speckled dace, 210speckled rattlesnake, 47, 98, 202Spermophilus spilosoma, 231Spermophilus tereticaudus, 234, 246-248spiny softshell, 186, 188, 200Sporobolus wrighti, 137spotted bat, 96spotted ground squirrel, 231spotted leaf-nosed snake, 203spring snail, 150Stenocereus alamosensis, 242Stenocereus thurberi, 172, 182, 242sticky buckwheat, 164striped skunk, 97Sueda moquinii, 148summer tanager, 164supplemental feeding, 286, 289-290swift fox, 97Sycamore, 254Sylvilagus floridanus, 230

TTanque Verde Ridge, 56Tarahumara frog, 210Taxidea taxus, 232Tayassu tajacu, 265termites, 171+Thamnophis cyrtopsis, 210, 214Thamnophis eques, 210, 214Thomomys bottae, 232Thomomys umbrinus, 230threecorner milkvetch, 164tiger rattlesnake, 203tiger salamander, 98, 184, 186, 188, 199tiger whiptail, 201Tilapia zillii, 153, 207, 210Tohono O’odham, 42, 110, 112, 151, 181Toumey oak, 74Townsend’s big-eared bat, 96Trachemys scripta, 184, 200

Trachemys yaquia, 210Trailmaster camera, 15, 17, 35, 228-232transportation corridor, 337+tree growth rates, 71+Tryonia quitobaquitae, 150Tumacacori National Historical Park, 23Typha domingensis, 149

UUma notate, 189Underwood’s mastiff bat, 96, 150undocumented aliens, 109+ United Nations Biosphere Reserve, 41, 42Universidad de Sonora, 1urban containment, 319+urban growth management, 328-329urban sprawl, 19+Ursus americanus, 275+, 293U.S. Army Corps of Engineers, 127-128U.S. Border Patrol, 109-110, 114, 123+, 181,

190, 194U.S. Department of Defense, 182U.S. Environmental Protection Agency, 312U.S. Fish and Wildlife Service, 123, 143,

162, 168, 182, 20.1, 22.1U.S. Forest Service, 1, 10.9, 19.2, 25.6U.S. Geological Survey, 1, 81, 23.3U.S. Immigration and Naturalization Service.

109U.S.-Mexico border, 109+, 123+, 181+,

225+, 283, 288, 337+

VValley pocket gopher, 96variable sandsnake, 186, 202velvet ash, 56velvet mesquite, 56, 149vermilion flycatcher, 164Vireo bellii arizonae, 164Virginia opossum, 232

Wwalnut, 254wastewater discharge, 24-26, 153, 157Waterman Mountains, 112water quality, 26-27, 157

360 Index

Wellton-Mohawk Irrigation and DrainageDistrict, 161, 162, 184

Western Area Power Administration, 162western banded gecko, 186, 201western coralsnake, 47western diamond-back rattlesnake, 63, 98,

202 western groundsnake, 63, 203 western harvest mouse, 231western least bittern, 164western lyresnake, 204western mastiff bat, 96western mosquitofish, 210 western narrow-mouthed toad, 199western patch-nosed snake, 63, 203western pipistrelle, 96western red bat, 164, 168western shovel-nosed snake, 47, 202western spotted skunk, 97western threadsnake, 203western yellow bat, 164, 168West Silver Bell Mountains, 112 white bursage, 182white-ankled mouse, 96white-footed mouse, 96, 231white oak, 74white-throated woodrat, 17, 97, 231wild turkey, 63willow, 56, 254wind-blown dust, 303+wolf, 234, 293 Woodhouse’s toad, 199

XXantusia vigilis, 184Xyrauchen texanus, 161, 162

YYaqui slider, 210, 211, 216-217, 221Yaqui sucker, 210 yellow mud turtle, 184yellow-bellied sea-snake, 203yellow-billed cuckoo, 63, 164yellow-breasted chat, 63yellow-faced pocket gopher, 96yellow-nosed cotton rat, 234yucca, 56Yuma clapper rail, 123, 161, 165Yuma County Water Users’ Association, 162Yuma Dunes, 189, 190Yuma hispid cotton rat, 164Yuma Irrigation District, 162Yuma Mesa Irrigation and Drainage District,

162Yuma myotis, 96Yuman fringe-toed lizard, 187, 189, 191,

194, 202

Zzebra-tailed lizard, 186, 201

Documents

Southwestern Desert Resources - Amazon Web Services · 2017-05-31 · This book brings together peer-reviewed research from two interagency projects that focused on Southwestern Desert