This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/authorsrights

Author's personal copy

Patterns of lizard species richness within National Parks and Biosphere Reservesacross North America’s deserts

C.W. Barrows a,*, H. Gadsden b, M. Fisher c, C. García-De la Peña d, G. Castañeda d, H. López-Corrujedo d

aCenter for Conservation Biology, University of California at Riverside, 75-080 Frank Sinatra Drive, Riverside, CA 92521-0334, USAb Instituto de Ecología, A. C.-Centro Regional Chihuahua, Miguel de Cervantes 120, Complejo Industrial Chihuahua, C.P. 31109, Chihuahua, Chihuahua, MexicocUniversity of California Natural Reserve System, Philip L. Boyd Deep Canyon Desert Research Station, P.O. Box 1738, Palm Desert, CA 92261, USAd Facultad de Ciencias Biológicas, Universidad Juárez del Estado de Durango, Av. Universidad S/N, Fraccionamiento Filadelfia, C.P. 35010, Gómez Palacio, Durango, Mexico

a r t i c l e i n f o

Article history:Received 26 September 2012Received in revised form2 November 2012Accepted 24 March 2013Available online

Keywords:BiodiversityBiogeographyBiosphere ReservesConservationEndemismGlobal changeNational ParksMexicoUnited States

a b s t r a c t

Warm deserts world-wide provide habitats for rich lizard species assemblages; North American desertsare no exception, however the desert regions of the US and Mexico are experiencing increasing habitatchanges from multiple anthropogenic sources. Our objective here was to document current lizard speciesrichness patterns across the North American deserts within the existing network of conservation areas.We identified 110 lizard species occurring across one or more of the 19 sites we analyzed. Three speciesrichness hot spots were identified; a northern Baja California faunal extension into southern California inthe US, and in Mexico, two sites within the state of Coahuila, as well as high endemism in the CapeRegion of Baja California Sur. Species richness was associated with sites where desert ecoregions overlapand with insular isolation. Our uncertainty regarding how species will respond to the multifaceted as-pects of global change is such that large protected natural areas with complex topography may be themost effective strategy for protecting desert lizards along with overall biodiversity. The 19 sites weanalyzed represent the cores of a more robust conservation network that will be needed for the pro-tection of biodiversity across North American Deserts.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Documenting patterns of species richness across complexlandscapes and then fitting those patterns into explanatory frame-works is a foundation of ecology and biogeography. These patternshold clues to the breadth of habitats occupied by individual species,as well as landscape features that foster species aggregationsranging from rich to depauperate. Additionally, understanding thebreadth of habitats and regions a species occupies can provide in-sights into its resiliency to habitat changes (Broennimann et al.,2006; Reif and Flousek, 2012). Overlaying patterns of biodiversityatop conservation compatible or incompatible land uses enables“gap analyses”, focusing limited conservation resources where theycan have the largest impact. Current trajectories of global climatechange are expected to result in habitat loss and fragmentation.Increasing introductions of invasive species will only increase

stresses on biodiversity. These consequences underline the need toassess the efficacy of existing conservation networks at protectingbiodiversity “hot spots” in addition tomore broadly ranging species.

Many warm deserts world-wide provide habitats for rich lizardspecies assemblages (Pianka, 1986). Deserts of North America areno exception, however they are experiencing increased habitat lossand fragmentation from urban and agricultural development (Chenet al., 2010; Gadsden et al., 2006b; Rodríguez-Estrella, 2007; Stilesand Scheiner, 2010), wind and solar energy development (Lovichand Ennen, 2011), invasive species (Barrows et al., 2009; Olssonet al., 2012), and levels of climate change with temperatures andprecipitation predicted to have a greater departure from currentconditions than other temperate regions (Kerr, 2008). Modeledshifts in lizard species’ distributions in response to a warmingdesert have predicted levels of habitat losses corresponding to thebreadth of the species’ niche (Ballesteros-Barrera et al., 2007;Barrows, 2011; Barrows et al., 2010). Losses of suitable habitatattributed to climate change have already been documented forother lizard species (Barrows et al., 2010; Sinervo et al., 2010).While some lizards have shown resiliency with regard to urban

* Corresponding author. Tel.: þ1 92521 0334.E-mail address: cbarrows@ucr.edu (C.W. Barrows).

Contents lists available at SciVerse ScienceDirect

Journal of Arid Environments

journal homepage: www.elsevier .com/locate/ jar idenv

0140-1963/$ e see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.jaridenv.2013.03.012

Journal of Arid Environments 95 (2013) 41e48

Author's personal copy

development and habitat fragmentation (Kwiatkowski et al., 2008)others have not (Banville and Bateman, 2012; Barrows and Allen,2009; Barrows et al., 2008). The uncertainty of how species willrespond to novel changes resulting from themultifaceted aspects ofglobal change outweighs what we can confidently predict.

With only a narrow ability to foresee the trajectories of speciesin response to such changes, large protected natural areas withcomplex topography may be the most effective strategy for pro-tecting desert lizards and overall biodiversity. An important ques-tion to consider is whether or not the existing network of protectedareas in the deserts of North America is sufficient to protect thecurrent distribution of lizard biodiversity, or do gaps exist thatcould be the focus of conservation efforts. Our objective herewas todocument lizard species richness patterns across the North Amer-ican Deserts in Mexico and the US within the existing network ofconservation areas (largely National Parks, National Monuments,and UNESCO Man and the Biosphere Reserves). In reporting thesepatterns we document near continental-scale patterns in desertlizard biodiversity, develop hypotheses to explain differences inspecies richness, and provide a coarse-scale assessment of whetherthe existing conservation areas adequately protect the rich desertsaurian fauna of this region.

2. Methods and materials

2.1. Study areas

We compiled lizard species lists of indigenous species for 19areas that span the Great Basin, Mojave, Sonoran, Chihuahuan andBaja CaliforniaDeserts of NorthAmerica (Fig.1).With one exception,each site included National Parks, National Monuments, State Nat-ural Areas, regional multi-agency conservation programs, or UnitedNations (UNCESCO) designated Man and the Biosphere Reserves.The one exception was La Comarca Lagunera in Coahuila, Mexico,a site which has received extensive lizard-related research(Castañeda, 2007; Castañeda et al., 2004; Gadsden et al., 2001,2006a,b, 2012;García-De la Peña et al., 2003, 2004, 2007).While thissite has been proposed as a high priority for conservation due to itsconcentration of endemic species, it has yet to receive broad officialprotection status (Gadsden et al., 2006b, 2012). A priori desertecoregional affiliations were based on the mapped extents of theNorthAmericanDeserts in Shreve (1942),Wells andHunziker (1976)andWells (1983). These 19 areas, alongwith a general description oflocation and management entities, are listed below. The latitudeelongitude (lat./long.) coordinates are the loci of the center points foranalyzing effectiveness of conserving local lizard species.

2.1.1. Great Basin Desert (GB)

(GB1) Mono Lake Tufa State Natural Area (USA: CA-NV), lat.37.819125, long. �118.473792, includes State Natural Arealands, Federal Forest Service Scenic Areas and other Federallands ownerships. Elevation range from 3800 to 2000 m.

(GB2) Great Basin National Park (USA: NV), lat. 38.985595,long. �114.313770, Area 312 km2, elevation range 3980e1700 m.

2.1.2. Mojave Desert (MD)

(MD1) Death Valley National Park (USA: CA), lat. 36.101752,long. �116.624477, Area 13,470 km2, elevation range from2700 to 145 m.

(MD2) Lake Mead National Recreation Area (USA: AZ-NV), lat.36.143138, long. �114.721882, with elevations from 1500 mto 400 m.

(MD3) Grand Canyon National Park (USA: AZ), lat. 36.054604,long. �112.140015. The Park is 4868 km2 in area, with ele-vations ranging from 2500 m to 700 m. Although notincluded in the mapped distribution of the Mojave Desert,the park’s lower elevations include an eastward extensionof desert flora and fauna.

(MD4) Mojave National Preserve (USA: CA), lat. 35.012601,long. �115.653401, A National Park administered unit of6400 km2 in area, elevations ranging from 2500m to 300m.

2.1.3. Baja California Desert (BC)

(BC3) Parque Nacional del Desierto Central de Baja California, (MX:BC), lat. 29.156734, long. �114.010397, with elevationsranging from 1249 m to sea level.

(BC4) Reserva de la Biosfera Desierto de El Vizcaíno, (MX: BCS), lat.27.309302, long. �113.442876, with elevations ranging from1200 m to sea level.

(BC5) Reserva de la Biosfera Sierra de Laguna, (MX: BCS), lat.23.841782, long. �110.092806, with elevations from 2000 mto sea level.

2.1.4. Sonoran Desert (SD)

(SD1) Joshua Tree National Park e Santa Rosa and San Jacinto Na-tional Monument e Anza Borrego State Park (USA: CA), lat.

Fig. 1. Distribution of sites included in our assessment of patterns of lizard speciesrichness across five major divisions of North American deserts. The size of the circles(outer diameter) represents the approximate area included in our compilation of lizardspecies for each site. One site, centered on Grand Canyon National Park (MD3), appearsoutside the mapped desert distributions; however the lower elevations of the park,along the Colorado River include Mojave Desert vegetation and lizards. See text for sitenames and descriptions.

C.W. Barrows et al. / Journal of Arid Environments 95 (2013) 41e4842

Author's personal copy

33.772985, long. �116.303802. This is a transitional site be-tween the Mojave Desert, Sonoran Desert, and the BajaCalifornia Deserts, incorporating the western half of the3171 km2 Joshua Tree National Park, all of the 1088 km2

National Monument and the northern portion of the2560 km2 State Park. The Coachella Valley Multiple SpeciesHabitat Conservation Plan connects these areas. Elevationsrange from 3300 m to below sea level.

(SD2) Organ Pipe National Monument e Cabeza Prieta NationalWildlife Refuge (USA: AZ) e Reserva de la Biosfera GranDesierto, (MX: Son), lat. 32.161200, long. �113.294020, TheNational Monument includes 1323 km2, with elevationsranging from 1000 m to 100 m.

(SD3) Saguaro National Park e Ironwood Forest National Monu-ment (USA: AZ), lat. 32.273235, long. �111.146980, with el-evations ranging from 2700 m to 700 m. The National park isdivided in two units totaling 365 km2, with additional landscomprised of US National Forest, and lands set aside as partof the Sonoran Desert Conservation Plan and Pima CountyMultiple Species Conservation Plan.

(SD4) Parque Nacional Constitucion 1857 e Sierra Juarez e LagunaSalada, (MX: BC), lat. 32.371276, long. �116.095596, withelevations ranging from 1740 m to 10 m.

(SD5) Parque Nacional Sierra de San Pedro Martir, (MX: BC), lat.30.992048, long. �115.373742, with elevations ranging from3040 m to 10 m.

2.1.5. Chihuahuan Desert (CD)

(CD1) Carlsbad Caverns National Park and Guadalupe Moun-tains National Park (USA: NM, TX), lat. 32.060546,long. �104.639653. The two Park units together include533 km2, with elevations ranging from 2400 m to 1000 m.

(CD2) Big Bend National Park (USA: TX), lat. 29.243626,long. �103.305132. The park includes 3205 km2, plus anadjacent State Park, with elevations ranging from 2300 m to600 m.

(CD3) Reserva de la BiosferaMapimi, (MX: Chi, Coa), lat. 26.764055,long. �103.758492, with elevations ranging from 1000 m to2000 m.

(CD4) La Comarca Lagunera (MX: Coa), lat. 25.576997,long. �103.149553, with elevations ranging from 3100 m to1100 m.

(CD5) Cuatrociénegas Protected Natural Area (MX: Coa), lat.26.923153, long. �102.118659, with elevations ranging from3000 m to 700 m.

2.2. Analyses

A critical consideration of any spatial analysis (De Knegt et al.,2010) is scale. Because the sites varied in area, we needed acommon scale with which we could compare species richnessamong sites. To identify an objective scale we analyzed speciesearea curves for selected sites for which we had the greatest first-hand knowledge (MD4, SD1, SD4, CD2, CD3, CD4, and CD5)(Fig. 2). We chose an arbitrary center point within each site andscribed 10 km radius concentric circles extending 100 km aroundthat point, tallying cumulative lizard species within each circle.Species abundance increased with increasing radius (area) withinthese circles until stabilizing at 30e50 km radius. Althoughcontinuing to increase the radius beyond 100 km would even-tually include additional ecoregions and form a stair-step speciesaccumulation curve, it would no longer represent the character-istics of the conservation sites we evaluated in this analysis.Based on these findings we chose a 50 km radius circle

(7850 km2) which is roughly within the same order of magnitudein scale as many of the conservation areas plus a surroundingbuffer area. This scale is large enough to include many isolatedmountain ranges as well as intervening valleys, and so includesthe elevation gradients and habitats available to the lizards. The“sky islands”, isolated mountains extending high above the desertplains, if high enough to include a broad temperature and pre-cipitation gradients, may be important refugia for saxicolouslizards to escape changing climates (Gadsden et al., 2012). Amuch smaller scale would miss this topographic diversity,whereas a much larger scale would begin to include multipleecoregions and lose the site-specific character of the individualareas.

Our species lists for each site were compiled through websitesfor the individual Parks and were aided by the recent completion ofcomprehensive field guides with distribution maps (Brennan andHolycross, 2009; Grismer, 2002; Jones and Lovich, 2009; LemosEspinal and Smith, 2007a,b; Stebbins, 2003). Otherwise the listswere assembled through our own fieldwork over the past threedecades. We generally followed the taxonomy used in those above-mentioned field guides, updated using Crowther (2012) for U.S.species, for the Sceloporus magister complex we follow Leaché andMulcahy (2007), for the coast horned lizard complex we followLeaché et al. (2009), and for the Xantusia species complex we followSinclair et al. (2004).

Comparisons between areas were conducted using the JaccardCoefficient of Similarity (Mueller-Dombois and Ellenberg,1974).Wecalculated the coefficient using the formula:

CCJ ¼ cS1 þ S2 � c

where S1 and S2 are the total number of lizard species occurringwithin areas 1 & 2 respectively, and c is the number of speciescommon to both areas 1 & 2. A CCJ of 1.0 means the two sites share100% of their species whereas a value of 0.0 would mean the twosites have no species in common.

3. Results

3.1. Patterns of species richness

A total of 110 lizard species occurred in the 19 protected areasacross the deserts of North America (Table 1). There was an ex-pected gradient of increasing lizard species richness from the

Fig. 2. Species area curves for those sites for which the authors have first-handknowledge of the distribution of lizard species. Site names and descriptions are inthe text.

C.W. Barrows et al. / Journal of Arid Environments 95 (2013) 41e48 43

Author's personal copy

much cooler Great Basin Desert (9e10 species) to the warmerMojave, Sonoran (including the Baja California deserts), andChihuahuan Deserts (Table 2). Once well into the warm desertzone, most areas included 19e25 lizard species, however therewere exceptions. There was a “peninsula effect”, with decliningspecies richness heading south along the Baja California penin-sula, culminating in 17e19 lizard species at the southern end(BC4e5). There were also “hot spots” of 30 or more specieslocated in three regions: the western Joshua Tree National Park tothe northern Anza Borrego State Park in southern California(SD1), and two sites in the central Chihuahuan Desert in the stateof Coahuila, Mexico (CD4 & 5). The Coahuila (CD4) “hot spot” aswell as the Cape Region of Baja California Sur (BC5) were alsoareas with high levels of narrow endemism at 8 narrow endemicspecies per site.

In many cases the existing conservation ownerships encom-passed the same species as did the 50 km radius circle (Table 2).There were exceptions, including sites focused on high elevationsand not extending to lower elevation habitats (e.g. SD4, 5 & BC5).The sites with the greatest elevational range (SD1, 5) alsoincluded high lizard species richness; however among all sites a

species richness-elevational range pattern lacked statisticalsignificance.

3.2. Similarity between sites

In addition to tabulating species richness we examined affin-ities of lizard species assemblages between areas. We expectedlizard assemblages to group in accordance within mappedboundaries of the four desert ecoregions. To some extent that wastrue; Great Basin, Mojave and Chihuahuan Desert sites eachformed discrete clusters of similarity based on shared species,with their greatest similarities to sites within their ecoregion(Table 3, Figs. 3 and 4). However there was little similarity amongSonoran Desert sites. The Organ Pipe National Monument-CabezaPrieta National Wildlife Refuge site (SD2) is well within themapped extent of the Sonoran Desert, but showed a closer affil-iation with Mojave Desert sites to the north and west (MD2 &4).The Saguaro National Park site (SD3) is located at the northeastmargin of the mapped Sonoran Desert; it had only a weak affili-ation (0.4 similarity) with the Organ Pipe National Monument-Cabeza Prieta National Wildlife Refuge site. The SD1, 4, &5 sites

Table 1The 110 lizard species occurring on at least one of the 19 sites used in our assessment of regional biodiversity. Letter designations following the lizard name indicate the site(letter þ number) or desert region (letter without numbers if the species occurs at all the sites within that ecoregion). See text for full site names and descriptions. Letters inbold print indicate narrow endemic species.

Crotaphytus antiquus CD4 Sceloporus maculosus CD4 Phyllodactylus xanti BC5Crotaphytus bicintores GB, MD, SD1 Sceloporus magister GB1, MD, BC1e3,

SD, CD2e5Aspidocelis burti SD3, CD1, 3e5

Crotaphytus collaris MD3, SD3, CD Sceloporus merriami CD2, 3, 5 Aspidocelis exanguis CD1Crotaphytus grismeri BC1 Sceloporus occidentalis GB, SD1, BC1e3 Aspidocelis gularis CD1, 3e5Crotaphytus nebrius SD2e3 Sceloporus olivaceus CD5 Aspidoscelis hyperythra SD1, BCCrotaphytus vestigium SD1, BC Sceloporus orcutti SD1, BC1e4 Aspidoscelis inornata CDGambelia copeii BC Sceloporus ornatus CD4 Aspidoscelis marmorata CDGambelia wislizenii GB, MD, BC1e2,

SD, CDSceloporus parvus CD5 Aspidoscelis septemvittata CD2-5

Ctenosaura hemilopha BC5 Sceloporus poinsettii CD Aspidocelis sonorae SD3Dipsosaurus dorsalis MD, BC, SD Sceloporus tristichus MD3 Aspidocelis tigris GB, MD, SDSauromalus ater MD, BC, SD Sceloporus undulatus CD3-5 Aspidocelis uniparens SD3Petrosaurus mearnsi SD1, BC1e3 Sceloporus vandenburgianus SD1, BC3 Aspidocelis xanthonota SD2Petrosaurus repens BC3e4 Sceloporus zosteromus BC3e5 Aspidocelis tesselata CD1e2Petrosaurus thalassinus BC5 Callisaurus draconoides GB, MD, BC, SD Xantusia bolsone CD4Phrynosoma blainvillii BC1, SD1 Cophosaurus texanus SD3, CD Xantusia extorris CD3-4Phrynosoma cerroensis BC2e4 Holbrookia approximans CD3e5 Xantusia gilberti BC5Phrynosoma cornutum CD Holbrookia elegans SD3 Xantusia gracilis SD1Phrynosoma coronatum BC5 Holbrookia maculata SD3, CD1, 3e5 Xantusia henshawi SD1, BC1e2Phrynosoma goodei SD2 Uma exsul CD4 Xantusia vigilis MD2, 4, SD1e2Phrynosoma hernandesi GB2, MD2,

SD3, CD1Uma inornata SD1 Xantusia wigginsi SD1, BC1

Phrynosoma mcallii SD1e2, BC1 Uma notata BC1 Plestiodon gilberti MD3e4,BC1e2, SD1

Phrynosoma modestum CD Uma paraphygas CD3 Plestiodon lagunensis BC5Phrynosoma platyrhinos GB, MD1, 3e4,

SD1e2Uma rufipunctata SD2 Plestiodon multivirigatus MD2, CD1e2

Phrynosoma solare SD2e3 Uma scoparia MD3e4, SD1 Plestiodon obsoletus SD3, CDSceloporus clarkii SD2e3 Urosaurus bicarinatus CD3 Plestiodon skiltonianus GB, MD2-3,

BC1e2, SD1Sceloporus consobrinus CD3 & 5 Urosaurus graciosus MD3e4, SD1e2,

BC1e2Plestiodon tetragrammus CD2, 5

Sceloporus couchii CD5 Urosaurus lahtelai BC3 Sincella kikapoo CD5Sceloporus cowlesi SD3, CD1e2 Urosaurus nigricaudus SD1, BC Sincella lateralis CD5Sceloporus cyanostictus CD4 Urosaurus ornatus MD1e2, SD2e3,

CD1e2Anniella pulchra SD1, BC1e2

Sceloporus edbelli CD3e4 Uta stansburiana GB, MD, BC, SD, CD Elgaria kingii SD3Sceloporus graciosus GB, MH1 & 3 Uta sp. nov. CD4 Elgaria multicarinata SD1, BC1e2Sceloporus grammicus CD4, 5 Coleonyx brevis CD Elgaria panamintina MD3Sceloporus hunsakeri BC5 Coleonyx reticulatus CD2, 5 Elgaria paucicarinata BC5Sceloporus jarrovii SD3, CD4 Coleonyx switaki SD1, BC1e4 Gerrhonotus lugoi CD5Sceloporus licki BC5 Coleonyx varigatus MD, SD, BC Gerrhonotus infernalis CD2, 4Sceloporus lineolateralis CD4 Phyllodactylus unctus BC5 Barisia imbricata CD5

Phyllodactylus notiocolus SD1, BC Heloderma suspectum MD 1, 2, 4, SD2e3

C.W. Barrows et al. / Journal of Arid Environments 95 (2013) 41e4844

Author's personal copy

were most closely affiliated with the Baja California Desert,forming a discrete unit only weakly affiliated with other Sonoranand Mojave sites (Table 3, Figs. 3 and 4).

4. Discussion

4.1. Gap analysis

Rather than the common misconception of a vast emptywasteland, North American warm desert regions provide habitatsfor a rich saurian fauna, along with high levels of biodiversity inmany other taxa. With one exception (La Comarca Lagunera, inCoahuila, Mexico) our analysis focused on lizard occurrences withinexisting conservation areas with national and or internationalrecognition to assess the degree to which these species have beenincluded within existing conservation networks. An importantquestion is which desert lizard species were missed by focusingonly on these 19 sites?

Across the desert regions of the US we included 70 lizard spe-cies. Just two, Aspidocelis gypsi and Aspidocelis scalariswere missed.

A. gypsi is protected within White Sands National Monument, butwe didn’t include that park due to its small size. One obvious gapwas our lack of inclusion of additional Sonoran Desert sites withinSonora, Mexico. While Sonora’s Parque Nacional Gran Desierto delPinacate was included in our Organ Pipe National Monument-Cabeza Prieta National Wildlife Refuge site (SD2), no other na-tional parks or biosphere reserves occur within the desert regionsof Sonora. As a result a handful of Sonoran lizard species weremissed, including Crotaphytus dickersonae, Phrynosoma ditmarsi,and Phylodactylus homolepidurus.

Within Baja California and Baja California Sur, of the 47 desertlizard taxa found on the peninsula, just one, Elgaria velazquezi, hada range that did not overlap with the existing conservation landsnetwork. However, due to a focus on higher elevation lands, manyof the Baja California parks failed to include from 17 to 34% of thespecies included in our 50 km radius circles. We opted not toinclude the approximately 30 island endemic taxa within the Gulfof California dividing Baja California and Sonora Mexico as theirbiogeography has previously been addressed and the islands haveprotection (Case and Cody, 2002; Grismer, 2002).

Table 2Patterns of lizard species richness and endemism across the desert regions and sites included in our analyses. Elevation range is the difference between the highest and lowestelevations found within the 50 km radius area of our site analyses.

Protected site No. Species:50 km

No. Species:Protected site

% speciesprotected

Narrowendemics

Elevationrange (m)

GB1 Mono Lake Tufa State Natural Area, CA-NV, USA 10 10 100 0 1800GB2 Great Basin National Park, NV, USA 9 9 100 0 2280MD1 Lake Mead National Recreation Area, NV-AZ, USA 12 12 100 0 2700MD2 Grand Canyon National Park, AZ, USA 19 19 100 0 1100MD3 Death Valley National Park, CA, USA 17 17 100 1 1800MD4 Mojave National Preserve, CA, USA 16 16 100 0 2200BC3 Parque Nacional del Desierto Central de Baja California, MX 21 21 100 1 1250BC4 Reserva de la Biosfera Desierto de El Vizcaíno, Baja Sur MX 17 17 100 0 1200BC5 Sierra de la Laguna Biosphere Reserve, Baja Sur MX 19 15 83 8 2000SD1 Santa Rosa-San Jacinto Mts e Joshua Tree NP, CA, USA 32 32 100 2 3300SD2 Organ Pipe Nat. Mon., Cabeza Prieta, AZ; Gran Desierto, MX 20 20 100 2 900SD3 Saguaro National Parks, Mt. Lemmon, AZ, USA 25 25 100 0 2000SD4 Parque Nacional Constitucion 1857, Baja CA MX 29 19 66 1 1730SD5 Parque Nacional Sierra de San Pedro Martir, Baja CA MX 26 20 77 0 3030CD1 Carlsbad Caverns NM e Guadalupe Mountains NP, NM, USA 19 19 100 0 1400CD2 Big Bend National Park, TX, USA 21 21 100 0 1700CD3 Mapimi Bioshpere Reserve, MX 24 24 100 2 1000CD4 La Comarca Lagunera, Coahuila, MX 30 3 10 8 2000CD5 Cuatrociénegas, Coahuila, MX 30 30 100 2 2300

Table 3Similarity matrix (Jaccard Coefficient of Similarity) between sites included in our analyses. See text for site descriptions.

GB1 GB2 MD1 MD2 MD3 MD4 BC3 BC4 BC5 SD1 SD2 SD3 SD4 SD5 CD1 CD2 CD3 CD4

GB1GB2 0.6MD1 0.4 0.2MD2 0.3 0.3 0.5MD3 0.5 0.4 0.5 0.4MD4 0.3 0.3 0.6 0.5 0.7BC3 0.2 0.1 0.2 0.2 0.3 0.3BC4 0.1 0.1 0.3 0.2 0.2 0.2 0.7BC5 0.1 0.1 0.3 0.2 0.2 0.2 0.3 0.4SD1 0.3 0.2 0.2 0.3 0.4 0.5 0.5 0.3 0.2SD2 0.3 0.1 0.5 0.4 0.4 0.5 0.3 0.2 0.2 0.3SD3 0.2 0.1 0.3 0.4 0.2 0.2 0.2 0.1 0.1 0.2 0.4SD4 0.2 0.2 0.3 0.2 0.4 0.4 0.6 0.4 0.2 0.7 0.3 0.2SD5 0.2 0.2 0.3 0.3 0.4 0.4 0.7 0.5 0.3 0.7 0.3 0.2 0.7CD1 0.1 0.1 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.1 0.1 0.2 0.1 0.1CD2 0.1 0.1 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.0 0.1 0.2 0.0 0.0 0.6CD3 0.1 0.1 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.0 0.1 0.1 0.0 0.0 0.4 0.5CD4 0.1 0.1 0.1 0.0 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.3 0.4 0.6CD5 0.1 0.1 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.0 0.1 0.1 0.0 0.0 0.4 0.4 0.4 0.4

C.W. Barrows et al. / Journal of Arid Environments 95 (2013) 41e48 45

Author's personal copy

Within Mexico’s portion of the Chihuahuan Desert we included42 species. Using the species lists and habitat descriptions includedin Lemos Espinal and Smith (2007a, 2007b), five desert lizardspecies were not included at the sites we considered: Holbrookialacerata, Sceloporus cautus, Sceloporus cyanogenys, Sceloporus minor,

and Aspidocelis uniparens. An additional species included in the USportion of the Sonoran Desert (SD3) but not included in the Mexicolist for the Chihuahuan Desert was Elgaria kingii. It is important toremember that one of the sites we did include for the MexicanChihuahuan Desert, the La Comarca Lagunera (CD4), has almost no

Fig. 3. Patterns of species similarity (Jaccard Coefficient of Similarity) and species richness between sites included in our analyses. Higher levels of similarity are indicated by thickerarrows connecting sites; the thickest arrows indicate a similarity coefficient of 0.7, the thinnest lines indicate a similarity coefficient of 0.4. Species richness levels are indicated bythe size and shape of the symbols.

Fig. 4. Patterns of species similarity (Jaccard Coefficient of Similarity) and endemism between sites included in our analyses. Higher levels of similarity are indicated by thickerarrows connecting sites; the thickest arrows indicate a similarity coefficient of 0.7, the thinnest lines indicate a similarity coefficient of 0.4. Numbers of narrow endemics areindicated by the size and shape of the symbols.

C.W. Barrows et al. / Journal of Arid Environments 95 (2013) 41e4846

Author's personal copy

protection. We recorded eight narrow endemic species onlyoccurring within that site. At the coarse scale of our analysis, thenarrow endemics included at La Comarca Lagunera constitute 14%of the saurian fauna for Mexico’s Chihuahuan Desert.

4.2. Hypothesis explaining the “hot spots”

Another component of our analysis was to identify patterns oflizard species richness across The North American Deserts. Threespecies richness hot spots were identified; one in the northern BajaCalifornia extension into the US (SD1), and two sites within Coa-huila, Mexico (CD4 & 5). Hypotheses explaining the high speciesconcentrations differed between these areas, as well as for the highlevel of endemism in the Cape Region of Baja California. Thenorthern Baja California-Sonoran Desert site (SD1) has only twonarrow endemic species. However it occurs at the confluence ofmultiple desert ecoregions, gaining species from each, with speciesaffinities spanning both Baja California and Mojave Desert speciesassociations. Additionally SD1 includes a broad elevational range,from below sea level to 3300m, a greater range than any other sitesin our study. This topographic complexity coupled with theecoregional juxtaposition accounts for their rich lizard faunaswithin a relatively small geographic unit. This rich saurian assem-blage is paralleled with a high diversity in plant community types,ranging from below sea level saltbush scrub to stands of limberpine, Pinus flexilis, above 3000 m.

In Mexico’s state of Coahuila, La Comarca Lagunera (CD4) andthe Cuatrociénegas Protected Natural Area (CD5) are well withinthe boundaries of the Chihuahuan Desert; at a coarse scale there isno ecoregional confluence to explain the rich saurian fauna, nor istheir topographic complexity exceptional when compared to manyof the other sites. However, La Comarca Lagunera is at a contractzone between the Sierra Madre Oriental and Sierra Madre Occi-dental affiliated lizard faunas, capturing species from both regionswithin the Chihuahuan Desert. The species richness of these areasmay also be a function of a northesouth vicariance for severalreptile species within Chihuahuan desert (Flores-Villela andMartínez-Salazar, 2009). The Chihuahuan Desert sites as a groupshowed little similarity with other North American Desert lizardassociations. Only the ranges of Crotaphytus collaris, Gambeliawislizenii, and Uta stansburiana consistently overlapped with otherdesert regions. The high number of narrow endemic species is animportant component of theses sites’ species richness, especiallyfor CD4. Much of that endemism is associated with isolated lime-stone rocky massifs or inselbergs imbedded in a “sea” of fine sed-iments and aeolian sand. Movement between the inselbergs bysaxicolous lizard species is inhibited if not prevented altogether,providing an archipelago-like backdrop for speciation (Gadsdenet al., 2012). The endemism and species richness in lizards areparalleled with a similar richness and endemism in the plant familyCactaceae (Cornet, 1985; Koleff et al., 2008).

Although the Cape Region of Baja California Sur, Mexico (BC5)does not have nearly the same level of species richness as SD1, or theCoahuila region of the Chihuahuan Desert (CD4 & 5), the number ofnarrow endemic lizard species was equal to CD4. The high level oflizard endemism here parallels similar patterns for both plant(Riemann and Ezcurra, 2005) and ant (Johnson and Ward, 2002)species. The Cape Region endemism is at least in part due to ageologic history of insular isolation from 10 to 3 million years ago(Murphy and Aguirre-Léon, 2002), along with a more sub-tropicalclimate as compared to the rest of the peninsula. Despite thedifferent explanations for their high species richness and/or centersof endemism these lizard “hot spots” appear to coincide with ahigh level of overall species and community biodiversity.

4.3. Conservation implications

The 18 sites included in our analyses that currently are partiallyor wholly protected include at least one protected area for 85% ofthe desert lizard fauna of North America. Adding the as of yet un-protected La Comarca Lagunera would increase this to 91%.Assessing whether or not the individual sites are sufficient to bufferagainst the risks posed to biodiversity by the complex interactionsassociated with global change will take a much finer-scaleapproach. In a similar analysis conducted to assess the sufficiencyof the existing conservation lands in Baja California for protectingnative plant diversity, Riemann and Ezcurra (2005) determinedmost were either too small and/or did not include sufficient topo-graphic diversity. We found a similar pattern among several of theBaja California sites. Much of the protection was focused at thehigher elevation, ruggedmountains whereas the more fertile lowerelevations were still at risk.

For broadly ranging species effective conservation would entailoccupancy within multiple conservation sites in order to capture amore complete range of habitats and conditions where they occur.Broadly ranging lizard species that are currently protected at justone of these sites include Sceloporus couchi, Sceloporus jarrovii,Holbrookia elegans, Aspidocelis uniparens and E. kingii. The largeranges of these species, along with those already included atmultiple sites, indicate broad niches and perhaps a greater ability tosurvive the multiple effects of global and regional change(Broennimann et al., 2006; Reif and Flousek, 2012). Local pop-ulations may still be at risk but the broader ranging species as awhole should survive in refugia; however, they will only be able todo so if conservation networks capture the breadth of those niches.For species with smaller distributions and presumably more nar-row niches buffering against the effects of global change willrequire robust sites that include the breadth of available localtopographic complexity, along with the resources to conduct reg-ular population assessments and then respond adaptively tothreats as they are identified.

With anthropogenic changes facing desert regions at anincreasing rate and with our narrow ability to predict species inresponses to such changes, large protected natural areas withcomplex topography are the most effective strategy for protectingdesert lizards as along with overall biodiversity. The 19 sites weanalyzed represent a backbone of a more robust conservationnetwork thatwill be needed for the protection of biodiversity acrossNorth American Deserts. Together with species that have beenomitted from this conservation network and other species that areunderrepresented, the desert regions of Sonora, Mexico along withLa Comarca Lagunera in the Chihuahuan Desert represent obviousconservation gaps. We also identified important gaps in Baja Cali-fornia where many of the conservation lands are focused on highmountain habitats, leaving lower elevation sites vulnerable. Each ofthe 19 sites includes lands that are under multiple uses or aredesignated for explicitly non-conservation purposes; none are fullybuffered from the impacts of global change. Filling conservationgaps along with working with surrounding multiple use managersto ensure connectivity and access to higher elevations, ecosystemprocesses are intact or restored, and that invasive species arecontrolled, will all be vital strategies for addressing these changes.

Acknowledgements

R. Johnson provided GIS support. This collaborationwas possiblethrough a grant from UCMEXUS. M. Goode, L. Jones and R. Reppprovided their expertise in defining the distributions of lizards onthe SD3 site.

C.W. Barrows et al. / Journal of Arid Environments 95 (2013) 41e48 47

Author's personal copy

References

Ballesteros-Barrera, C., Martínez-Meyer, E., Gadsden, H., 2007. Effects of land-covertransformation and climate change on the distribution of two microendemiclizards, genus Uma, of northern Mexico. Journal of Herpetology 41, 732e739.

Banville, M.J., Bateman, H.L., 2012. Urban and wildland herpetofauna communitiesand riparian microhabitats along the Salt River, Arizona. Urban Ecosystems 15,473e488.

Barrows, C.W., Allen, E.B., Brooks, M.L., Allen, M.F., 2009. Effects of an invasive planton a desert sand dune landscape. Biological Invasions 11, 673e686.

Barrows, C.W., Preston, K.L., Rotenberry, J.T., Allen, M.F., 2008. Using occurrencerecords to model historic distributions and estimate habitat losses for twopsammophilic lizards. Biological Conservation 141, 1885e1893.

Barrows, C.W., Allen, M.F., 2009. Conserving species in fragmented habitats: pop-ulation dynamics of the flat-tailed horned lizard, Phrynosoma mcallii. TheSouthwestern Naturalist 54, 307e316.

Barrows, C.W., Rotenberry, J.T., Allen, M.F., 2010. Assessing sensitivity to climatechange and drought variability of a sand dune endemic lizard. Biological Con-servation 143, 731e743.

Barrows, C.W., 2011. Sensitivity to climate change for two reptiles at the MojaveeSonoran Desert interface. Journal of Arid Environments 75, 629e635.

Broennimann, O., Thuiller, W., Hughes, G., Midgley, G.F., Alkemade, J.M.R.,Guisan, A., 2006. Do geographic distribution, niche property and life formexplain plants’ vulnerability to global change? Global Change Biology 12, 1079e1093.

Brennan, T.C., Holycross, A.T., 2009. Amphibians and Reptiles in Arizona. ArizonaFish and Game Department, Phoenix, Arizona, p. 150.

Case, T.J., Cody, M.L., 2002. Island Biogeography of the Sea of Cortez. University ofCalifornia Press, Berkeley, California, p. 508.

Castañeda, G., García-De la Peña, C., Lazcano, D., 2004. Notes on herpetofauna 5:herpetofauna of the sand dunes of Viesca, Coahuila, Mexico: preliminary list.Bulletin of the Chicago Herpetological Society 39, 65e68.

Castañeda, G.C., 2007. Ámbito hogareño y requerimientos de hábitat de los saurios:Uma exsul, Uta stejnegeri y Aspidoscelis marmorata en las dunas de arena deViesca, Coahuila, México. Tesis de Doctorado. Universidad Autónoma de NuevoLeón.

Chen, X., Li, B., Allen, M.F., 2010. Characterizing urbanization, and agricultural andconservation land-use change in Riverside County, California, USA. Annals ofthe New York Academy of Sciences 1195, E164eE176.

Cornet, A., 1985. Las Cactaceas de la Reserva de la Biosfera de Mapimi. Instituto deEcologia, Mexico, D.F., p. 51.

Crowther, B.I. (Ed.), 2012. Scientific and Standard English and French Names ofAmphibians and Reptiles of North America North of Mexico, with CommentsRegarding Confidence in Our Understanding, seventh ed., p. 92. SSAR Herpe-tological Circular No. 39.

De Knegt, H.J., van Langevelde, F., Coughenour, M.B., Skidmore, A.K., de Boer, W.F.,Heitkönig, I.M.A., Knox, N.M., Slotow, R., van der Waal, C., Prins, H.H.T., 2010.Spatial autocorrelation and the scaling of specieseenvironment relationships.Ecology 91, 2455e2465.

Flores-Villela, O., Martínez-Salazar, E., 2009. Historical explanation of the origin ofthe herètofauna of Mexico. Revista Mexicana de Biodiversidad 80, 817e833.

Gadsden, H., Dávila-Carrasco, M.de L., Gil-Martínez, R., 2006a. Reproduction in thearenicolous mexican lizard Uma exsul. Journal of Herpetology 40, 75e79.

Gadsden, H., Estrada-Rodríguez, J.L., Leyva-Pacheco, S.V., 2006b. Checklist of am-phibians and reptiles of the “Comarca Lagunera” in Durango-Coahuila, Mexico.Bulletin of the Chicago Herpetological Society 41, 2e9.

Gadsden, H.E., López-Corrujedo, H., Estrada-Rodríguez, J.L., Romero-Méndez, U.,2001. Biología poblacional de la lagartija de arena de Coahuila Uma exsul(Sauria: Phrynosomatidae): implicaciones para su conservación. Boletín de laSociedad Herpetológica Mexicana 9, 51e66.

Gadsden, H., Ballesteros-Barrera, C., Hinojosa de la Garza, O., Castañeda, G., GarcíaDe la Peña, C., Lemos-Espina, l. J.A., 2012. Effects of land-cover transformationand climate change on the distribution of two endemic lizards, Crotaphytusantiquus and Sceloporus cyanostictus, of northern Mexico. Journal of Arid En-vironments 83, 1e9.

García-De la Peña, C., Castañeda, G., Gadsden, H., Contreras-Balderas, A.J., 2007.Niche segregation within a dune lizard community in Coahuila, Mexico. TheSouthwestern Naturalist 52, 251e257.

García-De la Peña, C., Castañeda, G., Gadsden, H., Contreras-Balderas, A.J.,Lazcano, D., 2004. Autotomía caudal de Uma exsul (Sauria: Phrynosomatidae).Boletín de la Sociedad Herpetológica Mexicana 12, 43e48.

García-De la Peña, C., Castañeda, G., Lazcano, D., Jaime-Rosales, A., 2003. Uma exsul(fringe-toed sand lizard) predation. Herpetological Review 34, 370.

Grismer, L.L., 2002. Amphibians and Reptiles of Baja California, Including Its PacificIslands and the Islands in the Sea of Cortez. University of California Press,Berkeley California, p. 399.

Johnson, R.A., Ward, P.S., 2002. Biogeography and endemism of ants (Hymenoptera:Formicidae) in Baja California, Mexico: a first overview. Journal of Biogeography29, 1009e1026.

Jones, L.L.C., Lovich, R.E., 2009. Lizards of the American Southwest. Rio Nuevo,Tucson, Arizona, p. 567.

Kerr, R.A., 2008. Climate change hot spots mapped across the United States. Science321, 909.

Koleff, P.J., Soberón, J., Arita, H.T., Dávila, P., Flores-Villela, O., Golubov, J., Halffter, G.,Lira-Noriega, A., Moreno, C.E., Moreno, E., Munguía, M., Murguía, M., Navarro-Sigüenza, A.G., Téllez, O., Ochoa-Ochoa, L., Townsend, A., Peterson, Rodríguez, P.,2008. Patrones de diversidad espacial en grupos selectos de especies, en Capitalnatural de México. In: Conocimiento actual de la biodiversidad, vol. 1. CONABIO,México, pp. 323e364.

Kwiatkowski, M.A., Schuett, G.W., Repp, R.A., Nowak, E.M., Sullivan, B.K., 2008. Doesurbanization influence the spatial ecology of Gila monsters in the SonoranDesert? Journal of Zoology 276, 350e357.

Leaché, A.D., Mulcahy, D.G., 2007. Phylogeny, divergence times, and species limits ofspiny lizards (Sceloporus magister species group) in western North Americandeserts and Baja California. Molecular Ecology 16, 5216e5233.

Leaché, A.D., Koo, M.S., Spencer, C.L., Papenfuss, T.J., Fisher, R.N., McGuire, J.A., 2009.Quantifying ecological, morphological, and genetic variation to delimit speciesin the coast horned lizard complex (Phrynosoma). PNAS 106, 12418e12423.

Lemos Espinal, J.A., Smith, H.M., 2007a. Amphibians and Reptiles of the State ofChihuahua, Mexico. Universida Nacional Autónoma de México, Comisiónnacional para el Conocimiento y uso de la Biodiversidad, México, p. 613.

Lemos Espinal, J.A., Smith, H.M., 2007b. Amphibians and Reptiles of the State ofCoahuila, Mexico. Universida Nacional Autónoma de México, Comisión Nacionalpara el Conocimiento y uso de la Biodiversidad, México, p. 550.

Lovich, J.E., Ennen, J.R., 2011. Wildlife conservation and solar energy development inthe desert southwest, United States. BioScience 61, 982e992. http://dx.doi.org/10.1525/bio.2011.61.12.8.

Mueller-Dombois, D., Ellenberg, E., 1974. Aims and Methods of Vegetation Ecology.John Whiley and Sons, New York.

Murphy, R.W., Aguirre-Léon, G., 2002. The nonavian reptiles, origins and evolution.In: Case, T.J., Cody, M.L. (Eds.), Island Biogeography of the Sea of Cortez. Uni-versity of California Press, Berkeley, California, pp. 181e220.

Olsson, A.D., Betancourt, J., McClaran, M.P., Marsh, S.E., 2012. Sonoran DesertEcosystem transformation by a C4 grass without the grass/fire cycle. Biodiver-sity Research 18, 10e21.

Pianka, E.R., 1986. Ecology and Natural History of Desert Lizards. Princeton Uni-versity Press, Princeton, New Jersey, p. 208.

Reif, J., Flousek, J., 2012. The role of species’ ecological traits in climatically drivenaltitudinal range shifts of central European birds. Oikos 121, 1053e1060.

Riemann, H., Ezcurra, E., 2005. Plant endemism and natural protected areas in thepeninsula of Baja California, Mexico. Biological Conservation 122, 141e150.

Rodríguez-Estrella, R., 2007. Land use changes affect distributional patterns ofdesert birds in the Baja California peninsula, Mexico. Biodiversity Research 13,877e889.

Shreve, F., 1942. The desert vegetation of North America. The Botanical Review 8,195e246.

Sinervo, B., Méndez-de-la-Cruz, F., Miles, D.B., Heulin, B., Bastiaans, E., Villagrán-Santa Cruz, M., Lara-Resendiz, R., Martínez-Méndez, N., Calderón-Espinosa, M.L., Meza-Lázaro, R.N., Gadsden, H., Avila, L.J., Morando, M., De laRiva, I.J., Sepulveda, P.V., Duarte Rocha, C.F., Ibargüengoytía, N., Puntriano, C.A.,Massot, M., Lepetz, V., Oksanen, T.A., Chapple, D.G., Bauer, A.M., Branch, W.R.,Clobert, J., Sites Jr., J.W., 2010. Erosion of lizard diversity by climate change andaltered thermal niches. Science 328, 894e899.

Sinclair, E.A., Bezy, R.L., Bolles, K., Caramillo, J.L., Crandall, K.A., Sites Jr., J.W., 2004.Testing species boundaries in an ancient species complex with deep phylo-geographic history: genus Xantusia (Squamata: Xantusiidae). The AmericanNaturalist 164, 396e414.

Stebbins, R.C., 2003. Western Reptiles and Amphibians, third ed. Houghon-MifflinCompany, New York, p. 533.

Stiles, A., Scheiner, S.M., 2010. A multi-scale analysis of fragmentation effects onremnant plant species richness in Phoenix, Arizona. Journal of Biogeography 37,1721e1729.

Wells, P.V., Hunziker, J.H., 1976. Origin of the creosote bush (Larrea) deserts ofsouthwestern North America. Annals of the Missouri Botanical Garden 63,843e861.

Wells, P.V., 1983. Paleobiogeography of montane islands in the Great Basin since thelast glaciopluvial. Ecological Monographs 53, 341e382.

C.W. Barrows et al. / Journal of Arid Environments 95 (2013) 41e4848