SOCIETY

D

FE

HE

WILDLIFE MONOGRAPHS(ISSN:0084-0173)

A Publication of The Wildlife Society

DISTRIBUTION AND BIOLOGY OFTHE SPOTTED OWL IN OREGON

byERIC D. FORSMAN, E. CHARLES MESLOW, AND HOWARD M. WIGHT

APRIL 1984

NO. 87

DISTRIBUTION AND BIOLOGY OFTHE SPOTTED OWL IN OREGON

ERIC D. FORSMANCooperative Wildlife Research Unit, Oregon State University, Corvallis, OR 97331

E. CHARLES MESLOWCooperative Wildlife Research Unit, Oregon State University, Corvallis, OR 97331

HOWARD M. WIGHT'Cooperative Wildlife Research Unit, Oregon State University, Corvallis, OR 97331

Abstract: We studied the distribution, habitat, home range characteristics, reproductive biology, diet, vocal-izations. activity patterns, and social behavior of the spotted owl (Sift occidentalis) in Oregon from 1969through 1980. Spotted owls were located at 636 sites, including 591 (93%) on federal lands. The rangeincluded western Oregon and the east slope of the Cascade Range. Most pairs (97.6%) were found in unloggedold-growth forests or in mixed forests of old-growth and mature timber. No owls were found in forestsyounger than 36 years old. Paired individuals tended to occupy the same areas year after year and use thesame nests more than once. Mean nearest neighbor distances were 2.6 km west of the crest of the CascadeMountains and 3.3 km on the east slope of the Cascades. From 1969 to 1978, the population declined at anaverage annual rate of 0.8%. The principal cause of site abandonment was timber harvest.

Home range areas ranged from 549 to 3,380 ha. Seasonal home ranges averaged largest during fall andwinter. Home range overlap averaged 68% for paired individuals and 12% for individuals occupying adjacentterritories. Old-growth forests were consistently preferred for foraging and roosting by all owls studied. Inmost cases, use of younger forests was either less than or not significantly different than expected. There wasno apparent tendency to concentrate foraging activity in "edge" areas such as along clear-cut boundaries orroadsides. Owls roosted significantly higher in the forest canopy during cold, wet weather than during warmor hot weather.

Of 47 nests located. 30 were in cavities in old-growth conifers and 17 were on platforms of sticks or otherdebris in mature or old-growth conifers. None of the owls built its own nest. Average nest height was 27.3m (range 10.0-55.3 m, SE .■ 1.67). Forty-two nests were in unmanaged old-growth forests. Canopy closureexceeded 70% in most nest areas.

On the average, 62% of the pairs checked each year attempted to nest; 81% of all nesting attempts weresuccessful. Egg laying began between 9 March and 19 April (f •• 2 April). Average clutch size was 2.Incubation lasted 30 ± 2 days. Eggs hatched between 8 April and 20 May. Females alone incubated andbrooded the young. Males provided all food for the female and young until the young were approximately2 weeks old. The young left the nest at 34-36 days of age. After leaving the nest, owlets were fed by theadults until late August or September. The young became independent and began dispersal in October.Owlet mortality during the first summer of life was 35%. Before they were killed by predators in midwinter,2 radio-tagged juveniles traveled maximum straight line distances of 10.1 and 16.4 km from their respectivenatal sites.

The diet varied seasonally and included a variety of mammals, birds, and insects. Mammals comprised92% of all prey taken. Foraging activity usually began shortly after sunset and ceased shortly before sunrise.Calls given by spotted owls were classified into 13 types. Juveniles did not give adult vocalizations until lateSeptember or October of their first year.

WM. MONO= 87, 1-84

1 Deceased:

6 WILDLIFE MONOGRAPHS

CONTENTS

INTRODUCTION 8Acknowledgments 7

STUDY AREA 8METHODS 10

Location of Owls 10Habitat Evaluation 12Radiotelemetry Data Collection and Analysis.__ 12Dietary Analysis 14

RESULTS 15Distribution & General Habitat Characteristics_15

Habitats Occupied 16Spacing of Pairs 17Population Trends 17Land Ownership at Sites Where Owls were

Located 18Site Tenacity 19

Home Range Characteristics 19Total or Cumulative Home Range Size_20Seasonal Differences in Home Range Use_21Influence of Topographic Features on Home

Range Shape 22Horne Range Overlap Between Paired

Owls 22Home Range Overlap Between Neighbors _22

Habitat Selection for Foraging 24Forests 24Talus Outcrops and Rocksfides 25Roads and Miscellaneous Cover Types_25Seasonal Differences in Habitat Selection for

Foraging 29Habitat Selection for Roosting 29Habitat Selection for Nesting 30

Reproductive Biology 30Nests 30Timing of Nesting 32Clutch S9e and Rate of Egg Production —32Incubation Per8d 33Percent of Population Attempting to Nest -33Nest Success 33Copulation and Nest Selection 34Adult Behavior During Incubation and Brood-

ing 35Defense of the Nest and Young 35Development and Behavior of Nestlinp-36Development and Care of Fledglings 37Juvenile Mortality and Natural Enemies-38Dispersal of Juveniles 38

Food and Foraging Behavior 40Composition of the Diet 40Seasonal Variation in the Diet . 40Annual Variation in the Diet • , 42Foraging Behavior 42Food Caching 44A Comparison of Prey Selection by Males and

Females 45Vocal Displays 45Activity Patterns 51

Timing of Activity 51Rate of Movement While Foraging 51

Social Behavior 52DISCUSSION 53MANAGEMENT IMPUCATIONS seUTERATURE CITED 57APPENDIXES 60

INTRODUCTION

The spotted owl is a medium-sized,nocturnal owl inhabiting forested areas inwestern North America (Fig. 1). It occu-pies a variety of forest types, from humidconiferous forests in the northern part ofits range to xeric pine-oak woodlands inthe mountains and canyonlands of Mexicoand the American Southwest (Bent 1938).It is an extremely tame species, often al-lowing humans to approach within 2-3 mbefore flying away.

When the present study was initiatedin 1969, the spotted owl was considered arare or uncommon species in Oregon (Ga-brielson and Jewett 1940, Marshall 1969).There were only 24 previous records ofthe owl in the state, no information wasavailable on nest characteristics, and theonly information on the diet was based onan examination of 4 stomachs (Marshall

1942, Maser 1965). Although the specieswas believed to be threatened by loss ofhabitat (Marshall 1969), there was no in-formation to confirm this belief. A similarlack of information prevailed throughoutthe Pacific Northwest and British Colum-bia.

In the study described in this report theobjectives were:

To determine the distribution and nu-merical status of the spotted owl in Or-egon and evaluate population trends.To determine the types of habitat oc-cupied by spotted owls in Oregon andevaluate habitat preferences for for-aging, roosting, and nesting.

3. To describe the life history of thespecies, including home range charac-teristics, reproductive biology, diet, vo-cal repertoire, activity patterns, and so-cial behavior.

I= Q CAURINA EM2 OCCIDENTAUS- MILLI/GM

THE SPOTTED OWL IN OREGON —Forman et al. 7

4. To develop specific management rec-ommendations for the species based onthe information obtained.ACkflUtavagrsietiie. —We are deeply

indebted to many biologists and foresterswho helped -locate spotted owls. These in-dividuals and the agencies with which theywere Zffifited included: Charles Bruce,James Collins, Robert Mace, Ralph Opp,and Harlin Scott (Oregon Department ofFish and Wildlife); James Harper, Rich-ard King, Wayne Logan, Jerry Mires,Kenneth Munson, William Neitro, Law-rence Scofield, Douglas Smithy, CharlesThomas, and Frank Wagner (Bureau ofLand Management); Gerald Astrella,Margaret Bell, Terry Bryan, Warren Cur-rent, Jerry Diebert, William Dugas, Ste-phen English, Edward Harshman, Wil-liam Heckel, John Hillis, Kirk Horn, BruceHotchkiss, Phillip Janik, Paul Moroz, Wal-ter Parker, Richard Reynolds, Gene Sil-ovsky, Robert Summerfield, and LeeWebb (U.S. Forest Service); William Chil-cote (Oregon State University); Ron Nuss-baum (Michigan State University); WesPike (University of Arizona); James Blais-dell (National Park Service); Keith Com-stock (Georgia Pacific Lumber Compa-ny); and Robert Anderson (WeyerhaeuserCorporation). If we have left anyone out,we apologize.

For their participation in developing aspotted owl management plan for Ore-gon, we would like to acknowledge thefollowing individuals: Charles Bruce, Wil-liam Haight, Robert Maben, and FrankNewton (Oregon Department of Fish andWildlife); Barbara Holder, Kirk Horn,Leon Murphy, Richard Pederson, andGene Silovsky (U.S. Forest Service); DaveLuman: William Neitro, Douglas Smithy, •and Charles Thomas (Bureau of LandManagement); and David Marshall andPeter Stein (U.S. Fish and Wildlife Ser-vice).

For his support of this study, we wouldlike to thank jack Ward Thomas of theU.S. Forest Service Pacific Northwest For-est and •ange Experiment Station, La-Grande,70regon (Range and WildlifeHabitat Research Project USDA-FW-

Rg. 1. Distribution of the 3 &Suomi.. of spotted owl in NorthAmite.. The distribution In Colorado and Metioo is poorlydocumented. Map compiled from nunwrOus sources in the lit-erature.

PNW-1701). We also acknowledge An-drew Carey and Len Ruggiero of the U.S.Forest Service Pacific Northwest Forestand Range Experiment Station, Olympia,Washington, for their support of ongoingstudies of the spotted owl.

Chris Maser of the Bureau of LandManagement helped identify mammal re-mains in pellets and generously providedhis unpublished data on body weights ofOregon mammals. Loren Russell of Ore-gon State University identified most of the

8 WILDLIFE MONOGRAPHS

Fig. 2. Mayor geographic regions vidhin the Oregon studyarea (area west of the heavy dashed Ine). RsdizteMinistry studyareas are indicated by square symbols.

nonvertebrate prey remains in owl pellets.Monica Forsman contributed to almostevery aspect of the study, including trap-ping and radio tracking owls, caring forcaptive owls, collecting and dissecting pel-lets, conducting calling routes, measuringnest trees, and typing manuscripts.

Special thanks are due Charles Thomasof the Bureau of Land Management(BLM) who arranged the funding and lo-gistic support for a radiotelemetry studyon BLM lands, His enthusiasm and sup-port were greatly appreciated. John Beu-ter, Richard Clark, John Crawford, CarlMarti, Richard Reynolds, Thomas Scott,Helen Schultz, Robert Storm, and jackWard Thomas reviewed drafts of themanuscript; Alma Rodgers and CharlotteVickers typed several drafts of the manu-script.

The study was conducted under theauspices of the Oregon Cooperative Wild-life Research Unit: Oregon Department ofFish and Wildlife, Oregon State Univer-sity, U.S. Fish and Wildlife Service, andthe Wildlife . Management Institute coop-erating. Poftionftif this study were pre-sented by the senior author in partial ful-fillment for the M.S. and Ph.D. degrees inWildlife Science at Oregon State Univer-sity. This is Oregon Agricultural Experi-ment Station Technical Paper 6251.

STUDY AREAThe study area included the western

one-third of Oregon (Fig. 2). This area ischaracterized by rugged mountains and,together with western Washington, com-prises the most densely forested region inthe United States (Franklin and Dyrness1973). Based on climate and vegetation,the study area was divided into 3 majorsubregions: (1) western Oregon (which in-cluded the Coast Ranges, Willamette Val-ley, and west slope of the Cascade Rangein Fig. 2), characterized by mild, wet win-ters and warm, dry summers, (2) theKlamath Mountains (southwestern Ore-gon), characterized by cool, dry wintersand hot, dry summers, and (3) the eastslope of the Cascade Range, characterizedby cool winters and warm, dry summers.Elevations ranged from sea level to 3,452m. Most of the area was below 1,500 m.

Vegetation of western Oregon wasdominated by extensive temperate forestsof Douglas-fir and western hemlock (sci-entific names of plants are in Appendix 1)(Fig. 3). Other major forest types in thisregion included a narrow zone of sitkaspruce along the coastal headlands and arelatively narrow subalpine forest zone athigh elevations on the west slope of theCascades, where Pacific silver fir andmountain hemlock forests predominated.

In the Klamath Mountains, temperateforests of Douglas-fir and western hem-lock were largely replaced by mixed ev-ergieen and mixed conifer forests (Frank-lin and Dyrness 1973). In the latter foresttypes, Douglas-fir occurred in mixed

t

THE SPOTTED OWL IN OREGON—Forslnan et al. 9

Rg. 3. A typal Wyman forest of Douglas& and western hemlock on the H. J. Andrews study area. Oregon.

stands with white fir, grand fir, sugar pine,ponderosa pine, incense-cedar, Port-Ox-ford cedar, and associated broad-leavedevergreen trees such as tanoak, Pacificmadrone, canyon live oak, California lau-rel, and golden chinkapin. Above 1,500 min this region, mixed evergreen and mixedconifer forests were largely replaced byforests of white fir and Shasta red fir, withlesser amounts of Douglas-fir, sugar pine,and western white pine. The east slope ofthe Cascades was characterized by forestsof Pacific silver fir, subalpine fir, Shastared fir, lodgepole pine, or mountain hem-lock at higher elevations, mixed forests ofwhite fir, Douglas-fir, ponderosa pine, andincense-cedar at midelevations, and for-ests of ponderosa pine or lodgepole pineat lower elevations.

In all 3 subregions, extensive areas ofold-growth and mature forest had beenharvested within the 19th century (Fig.4). As a result- the vegetation was distrib-uted in re-mosaic of recent harvest units,

young second-growth stands, and residualstands of mature and old-growth timber.Typical old-growth forests were 230-350years old in the Coast Ranges and 250-500 years old in the Cascades (Franklin etal. 1981).

Radiotelemetry studies were conductedon 2 areas in Lane County in western Or-egon (Fig. 5). The H. J. Andrews (HJA)study area was located on the west slopeof the Cascade Mountains, 8 km northeastof the town of Blue River. The Bureau ofLand Management (BLM) study area waslocated in the Coast Ranges, 16 km westof the town of Lorane. Both study areaswere characterized by mountainous ter-rain and numerous small streams andrivers. Elevations ranged from 380 to 1,630m on the HJA study area and 150 to 590m on the BLM study area. Slope gradientson both areas commonly ranged from 40to 70% and occasionally up to 100%.

The climate and vegetation of both ra-diotelemetry study areas were typical of

fi

10

WILDLIFE MONOGRAPHS

Ft. 4. Mrial photo of the southern half of the BLM study arse. illustrating extensive cubmer areas and tomment stands ofold•rowth and mature forest. Asterisks Mica* nest sites or malt y roost area of 6 pairs of spotted owls Weed an thisponce of the study area.

western Oregon, as described earlier. The2 areas differed, however, in the amountof harvest that had taken place. On theHJA study area about 55% of the land areawas covered by old-growth forests andabout 10% was covered by 80-200-year-old forests. The remainder of the area wasmostly covered by early successionalvegetation on recent clear-cuts (Fig. 6).On the BLM area, about 70% of the landarea had been clear-cut within the pre-vious 40 years. The remaining blocks ofold-growth forest were widely spaced andcovered only about 20% of the land area.A detailed description of vegetative covertypes on the HJA and BLM study areas isprovided in Appendix 2.

METHODS

Location of Owls

To locate spotted owls we imitatedspotted owl calls in forest areas at night,either vocally or with a tape recorder. Thisincited the resident owls to call and reveal

their presence. Most nighttime calling in-ventories were conducted by driving alongforest roads, stopping at 0.3-0.8-km inter-vals to call and listen for owls. Inventoriesalso were conducted by calling every 15-30 seconds as we walked continuous tran-sects along roads or trails (Foramen et al.1977). Caning inventories were conductedbetween March and September when owlswere most responsive. ---- --

Numerous biologists from federal, state,and private agencies assisted in the owiinventory by conducting- calling routes.Because most of these cooperators werefederal employees, the greatest inventor)effort was concentrated on federal landsHowever, BLM and Forest- Service biolo-gists also inventoried extensive areas o:private land in areas where federal ancprivate lands were intermixed. Meatnearest neighbor distances were deter-mined by measuring between nests or ma•jor roost areas of neighboring pairs.

Daytime roosts were found by hominEin on radio-tagged owls. Roosts of un

t

THE SPOTTED OWL IN OREGON—FM/Ilan et al. 11

Fix 5. bi:J.Andrsws (top) and WA (bottom) mdiotstartistry study sums In western Ong=•

12 WILDLIFE MONOGRAPHS

Rg. 6. A typical 12-year-old cleazort on the H. J. Andrewsstudy area. (Photo by Jerry Franklin).

marked owls were found by calling dur-ing the day in areas where owls had beenlocated at night. When owls responded,they were approached on foot and visu-ally located in their roost trees.

Nest trees were located during the dayby (1) homing in on radio-tagged individ-uals, (2) observing adults during the pe-riod of nest site selection, (3) imitatingspotted owl calls in suspected nest areasuntil females left their nests to confrontthe suspected intruder, or (4) tethering livemice on the ground in view of roostingadults and then watching as the owls cap-tured the mice and carried them to thenest. The latter method also was effectivefor locating owlets after they left the nest.As we became familiar with individualpairs, it became easier to relocate themeach year because the same nest and roostareas were often used year after year. Ob-servations on behavior during the day weremade by sitting or standing quietly nearoccupied roosts or nests and watching owlswith binoculars or spotting scope. The owlswere so unafraid of humans that conceal-ment was unnecessary.

In addition to observations on free-ranging owls, we raised 2 female spottedowls in captivity and recorded informa-

tion on their development, molt, and be-havior (Forsman 1981b). One of the cap-tive owls laid eggs every year from 1975to 1980. Vocalizations were recorded on aUher 4000L tapes recorder and analyzedon a Kay 601A sonograph.

Habitat EvaluationIn all habitats occupied by spotted owls,

we recorded species composition of over-story and understory trees and stand age.Stand age was determined from core sam-ples or by counting annual growth ringson stumps in harvested areas adjacent tostands occupied by spotted owls. Variablesmeasured at nest sites included percentslope, aspect, canopy closure, and distanceto the nearest stream or spring. Canopyclosure at nests was estimated at 4 pointsaround the nest tree and averaged. The 4points were determined by walking a ran-dom number of steps (1-9) away from thenest tree along north, south, east, and westcompass bearings.

Radiotelemetry Data Collectionand Analysis

To capture owls we set mist nets or bownets near the roosting owls during the dayand baited the nets with live mice. Theowls almost invariably attempted to cap-ture any small prey placed near their roastsand were thus easily captured. Radiotransmitters were placed on 8 adult spot-ted owls (5 females, 3 males) on the HJAstudy area in 1975 and 6 adults (3 females,3 males) on the BLM study area in 1980.Each of the radio-tagged owls was a resi-dent that occupied an area with its mate.In 6 cases transmitters were placed on bothmembers of a pair. In 2. cases, we puttransmitters on females but were unableto capture the males with which they nest-ed. In addition to the 14 adults radio-tagged, 4 owlets were radio-tagged in Au-gust 1975.

Radio transmitters were attached witha backpack harness of tubular teflon rib-bon (Dunstan 1972). Transmitters weighed18-22 g including battery and harness and

13THE SPOTTED OWL IN OREGON--Forsman et al.

operated for 111-383 days before theyeither stopped transmitting or were re-moved. Transmitter frequency was 164Mhz. Signal pulse rate was 56-120 ppm.

Radio 1m:civil% equipment includedan AVM Model LA 12 receiver, a whipantenna 'mounted on a vehicle, and ahand-held 4-element yagi antenna. To de-termirreNfe. location of an owl by trian-gulation,.the observer obtained a bearingwith the • hand-held yagi, then moveddown thb road or trail to obtain at least 2additional bearings. If the owl moved be-fore at least 3 bearings were determined,a new series of bearings was obtained assoon as movement ceased. Movementcould be detected from fluctuations in sig-nal strength and changes in signal direc-tion relative to the observer. The problemof owls moving before a location could beaccurately determined was relatively mi-nor because of the manner in which theowls foraged. Except for occasional pe-riods of movement, the owls usually spentconsiderable time at each foraging loca-tion, permitting the observer to obtain nu-merous bearings before movement oc-curred.

The accuracy of locations generally de-clined as the distance between the observ-er and signal source increased. For thisreason we always moved as close as pos-sible to radio-tagged owls before triangu-lating (in most cases we were able to getwithin 400 m). Locations that were notsufficiently accurate to determine the typeof cover occupied were not used in habitatutilization analyses.

Locations were plotted on a cartesiancoordinate system overlaid on an ortho-photo of the study area. Coordinates werethen recorded on Fortran computer sheetsalong with the time (Pacific StandardTime), date, activity type, and type ofcover occupied.

Seven types of activity were recog-nized: foraging, roosting, moving, calling,territorial interaction, incubation orbrooding; and unknown. At night whenadults were not calling or moving from 1place to another, they were considered tobe foraging even if they did not move for

long periods of time (except, of course, forincubating or brooding females). This ap-proach was adopted because it was ex-tremely difficult to distinguish betweenforaging and resting behavior at night. Forexample, an owl might be immobile (i.e.,"resting") for some time and yet still be '-activelyactively watching for prey. When track-ing owls we always tried to stay withinhearing distance so that we could detectcalling activity or territorial interactions.On both study areas, owls were trackedan average of 5 nights/week. On the HJAstudy area we attempted to locate eachowl at least once each night. This "spotcheck" method was alternated with con-tinuous tracking sessions in which 1 or 2individuals were observed continuously forup to 12 hours at a time. During contin-uous tracking sessions, locations were re-ported every 15-30 minutes. Between No-vember and February on the HJA area,deep snow limited access to the homeranges of 3 owls at higher elevations, andwe were forced to concentrate observa-tions on 5 individuals. During this periodthe 3 individuals at high elevations werelocated at least once each week.

On the BLM study area a different ob-servation schedule was followed. Eachnight 1 pair of owls was observed fromearly evening until midnight, and anotherpair was observed from midnight untildawn. When both pair members could notbe tracked simultaneously, we eithertracked 1 member of the pair continu-ously or alternated between the 2 individ-uals at intervals of 1-2 hours. Locationswere determined at 5-minute intervals in-stead of 15-30-minute intervals. A rotat-ing schedule was established so that ob-servations were distributed equally withrespect to pairs and time of night. Theobservation schedule used on the BLMstudy area was initiated because it pro-duced more data per unit time and pro-vided a more continuous record of owlmovements than the method used on theHJA area.

The minimum convex polygon and el-lipse methods (Jennerich and Turner 1969)were used to calculate home range areas.

14 WILDLIFE MONOGRAPHS I •

Maps of owl movements were generatedon a computerized plotter. The homerange was defined as the area used by anowl for foraging, roosting, and nesting.

Based on the age, structure, and speciescomposition of the vegetation, we recognized 16 cover types on the HJA studyarea and 10 cover types on the BLM studyarea (Appendix 2). Cover types were de-termined by ground reconnaissance andwere plotted as overlays on orthophotos ofeach study area. The area of each covertype within the home range (minimumconvex polygon method) of each owl wasdetermined with a dot grid or polar pla-nimeter. The preference analysis de-scribed by Neu et al. (1974) was used todetermine which cover types were pre-ferred for foraging.

To determine when spotted owls wereactive, we either watched them in theirday roosts or monitored them remotely,listening until the transmitter signal indi-cated that movement had begun or ceased.When possible, owls were located visuallyor via triangulation in the morning, thenrelocated in the late afternoon to see ifthey had changed roost areas.

Roost site characteristics were deter-mined by visually locating 1-6 radio-tagged individuals in their roosts each day.Variables recorded at roosts on the BLMstudy area were: stand age, perch heightabove ground, dbh of roost tree, speciesand age of roost tree, aspect, distance tonearest spring or stream, and overheadperch protection. Overhead perch protec-tion was a ranked variable, indicatingwhether an owl was roosting under somesort of cover where it could avoid gettingwet during rainstorms or was roasting ina position where it was not well protectedfrom rainfall. Variables measured at roostsites on the HJA study area were the sameas on the BLM study area except that as-pect was not recorded.

Data collected at roosts were dividedinto 5 categories, depending on the pre-vailing weather conditions on the day ofuse: (1) clear, hot day with temperature>28 C, (2) clear, warm day with temper-ature between 10 and 28 C, (3) overcast

day with little or no precipitation, (4)overcast day with steady or intermittentrain or snow, or (5) clear, cold day withtemperature <10 C. Roost characteristicswere then compared to determine if roostselection was influenced by weather con-ditions.

Dietary Analysis

The diet of spotted owls was deter-mined from prey remains inregurgitatedpellets. Freshly killed prey staired in ornear roost trees were also recorded. Pelletscollected during each visit to a roost werecombined in a single sample, unless somepellets were markedly older than others,in which case samples were further sub-divided into "old" vs. "recent" pellets.Vertebrate prey in pellets were Identifiedfrom skeletal remains, fur, feathers, orscales (see Appendix 3 for scientific namesand weights of prey species). Insects wereidentified and enumerated from mandi-bles and fragments of the exoskeleton.Numbers of vertebrate prey were deter-mined by counting skulls, pairs of Jaws, orother skeletal remains (whichever gave thehighest count) in each sample. The bio-mass of each prey species in the diet wasestimated by multiplying the number ofindividuals times the mean body weightof the species. In the case of snowshoehares, brush rabbits, and mountain bea-vers, however, the body w . ht of eachindividual was estimated on the sizeof skeletal fragments and then summed toobtain total biomass.

To determine seasonal and sexual dif-ferences in prey selection on the radiote-lemetry study areas, we marked roost treesoccupied by individual owls of known sexand then returned to those roosts the nextday to collect any pellets that had beenregurgitated. Less frequently, pellets werecollected after we saw radio-tagged owlsregurgitate them. If pellets were collectedseveral days in a row from the same owl,those pellets were combined into a singlesample. This was done because remains ofa single large vertebrate often appearedin different pellets regurgitated 1 or more

THY SPOTTED OWL IN OREGON—Forsman et al. 15

days apart. Pellets collected under roostsoccupied by more than 1 owl were notused to compare sexual differences in preyselection. Also, pellets regurgitated bynesting fg." 11°C %War° n►t Ilaad to comparesexual differences in prey selection, be-cause=nesting females received most oftheir food from males. However, some ra-dio-tag/Ali- pairs did not nest during theradiotelqmetry studies; we were able tocompare sexual differences in prey selec-tion in those pairs during the spring andsummer.

For all statistical tests the maximumprobability accepted for statistical signif-icance was P = 0.05. Probabilities belowP 0.01 were considered highly signifi-cant.

RESULTSDistribution and GeneralHabitat Characteristics

Spotted owls were found at 636 loca-tions during the study (Fig. 7). At 322 lo-cations both a male and female were ob-served. Single individuals were observedat 172 locations. At 142 locations the num-ber of owls observed was not reported.Repeated sampling of 124 locations indi-cated that resident pairs were present atmost sites where single owls respondedduring a first census. We believe, there-fore, that most of the areas where singleowls (or unspecified numbers of owls) werereported were occupied by resident pairs.Sightings <1.5 km apart were consideredas originating from the same locations (i.e.,the same pair of owls), unless it was ver-ified that individuals from adjacent terri-tories were involved. To verify the exis-tence of 2 pairs in close proximity requiredthat the pairs be incited to call simulta-neously (or nearly simultaneously) at night,or that they be located during the day.

The range of the species included all ofthe mountains of western Oregon and theeast slope of the Cascade Range (Fig. 7).Only 2 spotted owls were reported fromareas eat of the Cascades. One was astarving -individual observed at Dufur,Wasco County, where it subsequently died

Fig. 7. Western hall or Oregon, illustrating 636 canons wherespotted owls were located (1969-80).

(Kirk Horn, pers. commun.). The otherwas an apparently healthy individual pho-tographed 6 km east of Klamath Lake,Klamath County, in September 1974 (TexWilliams, pers. commun.). We were un-able to locate spotted owls in the latterarea on subsequent visits.

Elevations at which owls were locatedranged from 24 m above sea level nearthe coast up to 1,340 m in the northernCascades, 1,524 m in the central Cas-cades, 2,010 m in the southern Cascades,1,830 m in the .Klamath Mountains, and1,130 m in the Coast Ranges. The upperlimits of the range generally correspondedwith the upper edge of the ecotone sepa-rating midelevation forests of Douglas-fir,western hemlock, white fir, grand fir, pon-derosa pine, or Shasta red fir from subal-

16 WILDLIFE MONOGRAPHS

Pg. 6. Odqowth stand of white fir and of:exhume' pine inthe southern Oregon Cascade Range near Lake of the Woods.Klamath County. A pair of spotted owls nested in Me areashown.

pine forests of Pacific silver fir, noble fir,subalpine fir, lodgepole pine, whitebarkpine, or mountain hemlock.

Habitats Occupied.—Habitat charac-teristics were reported for 595 of the 636sites where spotted owls were located.Spotted owls were found in all of the ma-jor coniferous forest associations that oc-curred in the study area, except for sub-alpine forests and nearly monospecificforests of ponderosa pine, lodgepole pine,or sitka spruce. Most (98.3%) of the siteswhere owls were located were forestedwith old-growth (>200 years old) conifersor mixtures of mature (100-200 years old)and old-growth conifers. Seven pairs(1.2%) occupied second-growth coniferforests in which very small patches of old-growth were present, and 3 (0.5%) occu-pied second-growth forests in which littleor no old-growth was present. No owls

were located in forests younger than 36years old.

The most consistent feature of old-growth forests occupied by spotted owlswas the presence of an uneven-aged, mul-tilayered canopy (Figs. 3, 8). Overstorytrees in these fatiests were typically 230-600 years old. Understory layers weregenerally dominated by younger (30-200years old) shade-tolerant trees such aswestern hemlock, western redcedar, Port-Orford cedar, grand fir, white fir, shastared fir, western yew, vine maple, canyonlive oak, California laurel, or tanoak. Al-though the density and closure of individ-ual canopy layers within these forests var-ied considerably, composite canopy closureusually averaged moderate to high (65-80%) as a result of the layered structure.Old-growth forests were also character-ized by moderate to high numbers of largetrees with broken tops, deformed limbs,and heart rot. Spotted owl nests were usu-ally located in such trees (see Reproduc-tive Biology).

Of the pairs found in old-growth ormixed old-growth and mature forests, 4occupied stands that had been selectivelylogged prior to the initiation of the studyand 9 occupied stands that were selective-ly logged after we first located them. Se-lective logging is a harvest method inwhich canopy density is reduced by re-moving some of the overstory trees. Theunderstory is either left intact or thinned.Of the 4 pairs occupying previously loggedsites, 3 nested in stands that had beenlogged 30-40 years earlier. The nest ofthe fourth pair was in an unharvested old-growth stand, 5 m from the edge of anarea that had been selectively logged about10 years earlier. In the former stands,young trees had since filled in many ofthe openings created by harvest, resultingin multilayered stands that were similarto unlogged old-growth stands except thatthe density of overstory trees was re-duced.

Of the 9 sites that were selectivelylogged after owls were located, 7 weresubjected to relatively light overstory re-moval, and 2 were heavily thinned. At 6

THE SPOTTED OWL IN OREGON--Forsman et al. 17

of the 7 sites subjected to light overstoryremoval, timber sales were laid out so thata small patch (2.6-10 ha) of unlogged, old-growth was left around the nest tree ofthe resident cr.. •!s. On the latter sites, 3pairs subsequently used their old nest treesin 1-or-snore years following harvest. Theother 3 pairs remained in the same gen-eral timer after harvest, but we were un-able to document nesting. The nest of theseventh pair was not located prior to har-vest, but we suspected it was located inthe area harvested (fledglings were ob-served in • the harvest unit prior to har-vest). In this case the owls responded toharvest by moving to a new nest tree inan unharvested old-growth stand 1.2 kmfrom the suspected location of the old nest.The new nest was subsequently used for2 years before the owls disappeared fromthe area completely.

On the 2 sites where overstory andunderstory trees were heavily thinned, thespacing between trees in the owl nest areaswas increased to 10-20 m as a result ofharvest. Canopy closure was reduced to<50%. One of these pairs subsequentlydisappeared. The other pair shifted theiractivities to an unlogged old-growth standbordering the harvested area; a new nestwas located in the unlogged area 4 yearsafter harvest of the original nest area.

Of 10 pairs located in predominantlysecond-growth forests, 7 occupied forestsin which 3-20-ha patches of old-growthand scattered individual old-growth treeswere intermixed with extensive 40-80-year-old forests. The other 3 pairs occu-pied relatively uniform 45-60-year-oldforests where no old-growth was present.Second-growth forests occupied by spot-ted owls differed structurally from old-growth in that the trees were smaller andmore uniform in age and size. Canopieswithin these stands were closed (65-80%closure) but were not strongly multilay-ered. Of the 10 pairs in second-growthforests, we were able to verify nesting by2, both of which occupied stands in whichsmall patches of old-growth trees were in-termixed with-the second-growth (second-growth tries at both sites were 70-90 years

old). One of the pairs did not nest suc-cessfully in the 3 years that we observednesting attempts. One year the nest failedfor unknown reasons, the next year theyoung were taken from the nest by apredator, and the next year the eggs were _somehow destroyed. This pair disap-peared in 1980. The other pair nested in2 consecutive years; we observed youngin the nest both years but did not deter-mine if they fledged.

Spacing of Pairs.—Mean nearestneighbor distances were 2.6 km west ofthe Cascades (N 47, SE .• 0.130, range1.6-5.2 km) and 3.3 km on the east slopeof the Cascades (N 18, SE 0.266,range 1.6-6.4 km). The minimum dis-tance between 2 active nests was 1.9 km,but in several instances we suspected thatpairs nested even closer together becausetheir principal roost areas were only 1.6-1.8 km apart. We did not locate the nestsof the latter pairs. Our observations on thespacing of pairs support Marshall's (1942:67) observation that where "suitable hab-itat prevails, pairs (of spotted owls] can beexpected at intervals of one to two miles."

In many areas, particularly the CoastRanges and Klamath Mountains, federaland private lands were distributed in anirregular checkerboard pattern with ap-proximately every other square mile sec-tion (259 ha) privately owned (Fig. 5). Thedistribution and density of spotted owls insuch areas was strongly influenced by thedistribution of federal lands. Most privatelands had been cutover, whereas federallands often contained residual tracts of old-growth and mature forest that were oc-cupied by spotted owls. Where these tractsof older forest were spaced close together,owl densities were high. In some areas thathad been almost entirely logged within theprevious century (e.g., the extreme north-ern end of the Coast Ranges), we had dif-ficulty finding any owls at all (Forsman etal. 1977).

Population Trends.—From 1972 to1978, 98 sites occupied by pairs of spottedowls were checked at 1-3-year intervalsto determine if the sites were still occu-pied. Sites were considered occupied as

18

WILDLIFE MONOGRAPHS

Table 1. Code mums. duration of tracking periods. and home range size of radio-tagged adult spotted owls an the H. J.Andrews and BLM study areas. Oregon.

Steer arma

code name Tricking periodNo. of days in

tracking period

Home range area (ha)Minimum crewelpolygon setunate Mae eeieuari

HJA areaIA male 4 May 1975-5 May 1976 367 3.254 4,8551A female 8 June 1975-14 May 1976 332 2,179 3,1692A male 25 May 1975-3 Apr 1976 314 1,272 1,1562A female 12 Jul 1975-3 Jun 1976 327 1,220 9363A female 12 Jul 1975-9 May 1976 302 920 1,0024A female 26 Jul 1975-1 Jun 1976 311 1,376 1,2395A male 25 May 1975-11 Jun 1976 383 950 6845A female 14 Sep 1975-11 Jun 1976 271 1,324 1,173

Mean for all HJA owls' 1,177 1.032Mean for males" 1.111 920Mean for females+ 1,210 1,087

BLM area1C male 2 Apr-22 Jul 1980 111 3,380 3.7761C female 13 Apr-26 Aug 1980 135 2,140 1.8232C male 2 Apr-15 Aug 1980 135 549 6022C female 13 Apr-27 Aug 1980 136 3,376 3.4813C male 12 Apr-27 Aug 1980 135 997 7963C female 16 Apr-1 Sep 1980 138 1,038 995

Mean for all BLM owls 1,913 1.912Mean for males 1,642 1,725Mean for females 2,185 2,100

For calculation i.e fennerich and TUTOR 1969.*swags do not include data from lA nuke and female bemuse that pair did not occupy the same home range arum far*, demise of the

stdy

long as a pair could be located within a1.5-km radius of the historical nest orprincipal roost area. Eight (8%) of the siteswere no longer occupied by 1978, indi-cating an 0.8% rate of population declineper year. In 7 cases, sites were abandonedafter nests or major portions of the forestsurrounding nests were harvested. Onepair disappeared after the female was ap-parently killed by a predator. She disap-peared from the nest overnight and wasnot seen again. The male eventually leftthe area. Although we could not discountthe possibility that displaced pairs simplyrelocated elsewhere, this was unlikely, be-cause all of the displaced pairs occupiedheavily cutover regions where there waslittle habitat in which to relocate.

Portions of the forest on at least 54 ofthe 98 sites monitored from 1972 to 1978were harvested during the period withoutcompletely displacing the resident owls.When part of the forest area occupied by

a pair of spotted owls was harvested, theowls usually responded by shifting theirarea of activity into the remaining uncutforests. When the remaining-forests con-sisted of extensive areas of young second-growth or heavily thinned timber, how-ever, the owls frequently disappeared.

Another indication that the owl popu-lation was declining was that owls couldnot be relocated in 7 (35%) of the 20 areaswhere they had been observed in Oregonprior to 1970. Because accurate locationsof most historical sightings were not avail-able, however, we were unsure in somecases whether we inventoried the exactareas where the historical sightings oc-curred. For this reason we view the resultsof this index as being only roughly indic-ative of population trends. Gould (1974,1977, 1979) studied occupancy rate's ofhistorical sites in California and reportedsimilar trends and problems with analysis.

Land Ownership at Sites Where Owls

•THE SPOTTED OWL IN OREGON -Forsman et al. 19

Table 2. Percent of locations of radio-tagged owls on the H. J. Andrews and ELM study areas according to activity type.

Stud .: 17code name

Number oflocations

Activity type

Foraging Rotating Moving CallingTerritorialinteraction

Incubationor brooding Unknown

HJA areaIA male 708 54.0 36.4 5.8 2.4 1.1 0.3IA lernile - 659 55.2 38.1 2.0 1.7 2.6 0.42A male 732 49.7 41.1 3.6 4.2 0.7 0.72A laiipilffl. 773 53.8 33.1 5.7 6.8 0.3 0.33A female 582 59.3 29.2 1.9 • 1.7 0.5 6.9 0.54A female 669 54.0 25.0 5.5 2.4 0.1 12.0 1.05A male- 417 51.1 36.7 5.8 3.6 0.2 2.65A female 387 41.7 24.8 5.9 1.0 22.5 4.1

BLM area1C male 1,986 92.7 4.0 0.2 1.8 0.1 1.21C female 1,903 92.7 4.7 0.4 1.4 0.82C male 1,754 92.6 5.3 0.7 0.4 0.92C female 1,977 92.4 5.0 0.5 0.9 0.1 1.23C male 2,042 91.5 4.6 0.6 2.6 0.73C female 2,283 93.2 4.2 0.2 1.6 tr• 0.84C male' 7 71.4 28.6tr e0.05%.

b The 4C male was an unmarked adult that occiusonall y rotated with the RC female. Hs ma was determined from vocalizations (see vocaldisplays).

were Located.-Fifty-three percent (314)of the sites where owls were located wereon lands administered by the U.S. ForestService, 40% (240) were on lands admin-istered by the Bureau of Land Manage-ment, 6% (34) were on privately ownedlands, 0.3% (2) were in national parks, and0.8% (5) were on lands owned by the stateof Oregon. Land ownership was unknownat 41 sites. The low incidence of occur-rence on private and state-owned landswas partly due to the fact that most co-operators were federal employees whosearched for spotted owls primarily onfederal lands. More important, however,was the fact that the majority of privateand state lands in western Oregon hadbeen cutover or burned within the pre-vious 70 years and were no longer occu-pied by spotted owls (Forsman et al. 1977).Recent inventories in Washington andCalifornia indicated that spotted owls alsowere uncommon on cutover private landsin those states (Gould 1974,1977; Postovit1979).

Site Tenacity.--Of 14 radio-taggedadults, 12 . remained on the same homerange areas foi the duration of the track-ing period (3-13 months). Two that were

identifiable because their transmitterswere not removed were still present in thesame areas 26 and 37 months after initialcapture, respectively. These results indi-cate that individual adults occupy thesame home ranges for long periods of time(probably for life), unless displaced byhabitat destruction. Miller (1974) reacheda similar conclusion based on observationsof a pair of spotted owls during a 7-yearperiod. The continuous occupancy ofmany sites by pairs of spotted owls duringour study was apparently made possiblebecause individuals that died were quick-ly replaced. This was verified on 5 occa-sions when marked individuals or individ-uals with distinctive markings disappearedand were replaced by the next spring (seealso Miller 1974).

Home Range Characteristics•

On the HJA study area, 261 nights and268 days were spent tracking owls, and atotal of 4,927 owl locations was obtained(Tables 1, 2). The 5 pairs studied on theHJA area were distributed along the 2major stream drainages (Lookout Creekand Blue River) that crossed the area (Fig.

20

WILDLIFE MONOGRAPHS

Fig. 9. Outlines of home ranges used by 8 radio-tagged adult spotted owls on the H. J. Andrews study area (May 1975-)un1976). Asterisks indicate locations of nests used by each pair.

9). An intensive search indicated that therewere no other pairs occupying areas be-tween the radio-tagged pairs. There were,however, at least 4 additional pairs occu-pying areas immediately adjacent to thestudy area. The latter pairs were not stud-ied intensively although some data on theirfood habits and spacing relative to the ra-dio-tagged pairs were collected.

On the BLM study area, 92 nights and111 days were spent tracking owls, and atotal of 11,952 owl locations was obtained(Tables 1, 2). The 3 pairs studied in thisarea confined most of their activities tothe Siuslaw River drainage and its tribu-taries (Fig. 10). On this study area there

were also several pairs of unmarked spot-ted owls occupying areas adjacent to thehome ranges of the marked owls. The un-marked owls were not studied except tonote their presence and their main areasof activity.

Total or Cumulative Home RangeSize.—Based on the minimum convexpolygon method, home ranges averaged1,177 ha on the HJA study area and 1,913ha on the BLM study area (Table 1; Figs.9, 10). The jennerich and Turner (1969)elipse method indicated a- mean homerange of 1,032 ha on the HJA area and1,912 ha on the BLM area (Table 1)..Sizesof male and female home ranges were not

THE SPOTTED OWL IN OREGON—Forsman et al. 21

Fig. 10. Outlines of home ranges used by 6 radio-tagged adult spotted owls on the BLM study area (Apr-Aug 1980). AsterisksIndicate location of nests used by 2 of the On in years preceding the study.

significantly different on either area (HJAarea t = 0.54, df = 4, P > 0.05; BLM areat 0.49, df = 4, P > 0.05) (Table 1). Inthe following sections, minimum convexpolygon estimates of home range size areused for all comparisons.

Home range size was at least partiallya function of how long an individual wasobserved. It was assumed that the totalhome range had been determined when aplot of home range size as a function ofthe length of the observation periodreached an asymptote (Fig. 11). On theaverage, 131 days (range .• 75-150 days)were required to determine 80% of thetotal home range used by each owl on theHJA area, and 165 days (range = 105-204days) were required to determine 90%(Fig. 11). After owls had been observedfor 5-6 months, movements outside thehome range boundaries already delineat-ed became uncommon. The relatively longperiod required to determine the total (ornear total) home range reflected the factthat owls visited some portions of theirhome ranges infrequently.

In spite of the fact that owls on the BLMstudy area. were only observed for 4

months, their home ranges averaged larg-er than the home ranges of owls on theHJA area where owls were observed for9-13 months (Table 1). Had owls on the2 areas been tracked for equal periods, thedifference in mean home range size wouldlikely have become even larger.

Seasonal Differences in Home RangeUse.—Home range data from the BLMarea were not subdivided seasonally be-cause data were only collected for part ofthe year. On the HJA .area home rangeuse changed seasonally (Table 3, Fig. 12).The most noticeable differences in homerange use on the HJA area occurred be-tween the breeding season (late Feb-Sep)and post breeding season (Oct-Jan). Dur-ing the breeding season, both nesting andnon-nesting individuals confined most oftheir activities to within 2.6 km of theirtraditional nests (Fig. 12). During the postbreeding season, the owls became solitaryandn to wander more extensively,gradub:g expanding their home ranges inmost cases (Table 3, Fig. 12). Apparently,use of areas located far from nest sites wasenergetically more feasible during winterwhen the owls did not have to return to

WILDLIFE MONOGRAPHS •22

100r

0z2 SO

0

a

z 404.;

04-,,, 20

0 30 60 90 120 ISO ISO 210 240 270 300 330 360OBSERVATION 9E9100 (DAYS)

Fig. 11. Size of total home range as • function of the lengthof the observation penod (data averaged for 4 owls on the H.J. Andrews study area). During the observation period theowls were observed approximately 6 days and 5 nights eachweek.

their nest areas 1 or more times each night.The use of larger home ranges in wintermay also have been a response to gradu-ally declining prey populations (Clark1975).

Influence of Topographic Features onHome Range Shape.—Several home rangeboundaries on the HJA study area corre-sponded roughly with high elevation(1,000-1,500 m) ridge crests, suggestingthat such ridges formed natural barriersto movement (Fig. 9). It is possible, how,ever, that avoidance of high elevationareas was due to patterns of vegetationdistribution rather than physical barriers.Vegetation on the higher ridges on theHJA area was dominated by mature for-

ests and subalpine brushfields of Sitka al-der. Owls may have avoided such areasbecause they preferred to forage in old-growth stands, which were more abun-dant at lower elevations (see Habitat Se-lection for Foraging).

Owls occupyinp,home ranges adjacentto Blue River Reservoir on the HJA studyarea rarely crossed the reservoir exceptnear its upper end where it was <150 mwide. We observed only 1 instance inwhich an owl made a continuous flight ofnearly 400 m across the main body of thereservoir.

Home Range Overlap Between PairedOwls.—Home ranges occupied by pairedindividuals overlapped by 50-73% on theHJA area and 40-93% on the BLM area(2 •. 68%) (Figs. 9,10). These estimates didnot include 2 pairs (1A and 2C pairs) inwhich there was some question aboutwhether the owls were established pairs.In all cases, the areas of overlap betweenpaired individuals included the nest area,major roost sites, and the areas that wereused most intensively for foraging. Thecombined home range areas used bypaired individuals ranged from 1,149 to4,225 ha (2 •• 2,144 ha).

Although their home ranges overlappedconsiderably, paired individuals used thesame foraging locations (X, Y coordinates)only 4-10% of the time. This suggestedthat there was little competition for foodbetween pair members.

Home Range Overlap Between Neigh-bors.—On the average, any 2 owls occu-pying adjacent territories on the HJA area

SO

Table 3. Home range size (ha) of 6 radio-tagged adult spotted owls during different periods of the year on the H. J. Andrewsstudy area (May 1975-Jun 1976).

owl

1975 19786 May-Jul Aug-Sep OM-Nov Dac-Jaa Feb-Mar Age-10 Jua

2A male 661 593 646 824 6042A female 336 1.047 1,048 449 3353A female 332 695 693 2244A female 1.062 911 1,154 318 2275A male 172 262 386 160 834.5A female 596 1.054 227 810Means 416 530 790 930 330 552

THE SPOTTED OWL IN OREGON —Forsman et al. 23

1-1:17---41 MILE

89.12. Computer maps illustrating seasonal changes In home range use by the 3A female on the H. J. Andrews study area(Aug 1975-Apr 1976). Dashed lines indicate boundaries of seasonal areas used and solid lines Indicate total home range.Location of nest used in 197618 indicated by an asterisk.

shared 12% of their home ranges (range3-25%, N •• 6). Shared areas were gener-

- ally near-the periphery of adjacent homeranges where the owls spent a relativelysmall proportion of their time (Fig. 9).Home range overlap between neighbor-

ing individuals did not differ greatly de-pending upon sex (2 11% between fe-males, 14% between males and females).Overlap between males could not be cal-culated because the radio-tagged males onthe HJA area did not occupy adjacent

24

WILDLIFE MONOGRAPHS

Table 4. Use of cover types for foraging by 8 radio-tagged adult spotted owls on the H. J. Andrews study area (May 1975—Jun 19761.

Cover type

Percentcoverage on

home ranges'

Percent oftime spentforagingin type'

No. of individuals using cover type

Significantlymorkthanexpected

Notsignificantly

differentthan expected

SipeScantlylas thanespected

Old-growth conifer forest 33-66 85-99 "*- 0 0Selectively logged old-growth

conifer forest tr-1 0-tr 0 2 481-200-year-old conifer forest 3-27 0-16 1 3 3Mature hardwood forest 0-2 2 0 1 061-80-year-old conifer forest tr-10 0-2 0 3 231-60-year-old conifer forest 1-2 0—tr 0 1 2_21-30-year-old conifer forest 3-19 0—tr 0 0 7--5-20-year-old conifer forest 10-31 0-6 0 0 8Recent clear-cuts covered by

brush or grass-forb associations 2-10 0-tr 0 0 6Rock talus tr 5 1 0 0Reservoir 2-6 0 0 0 4-Other' tr-7 0-1 0 3 5

• Test used to evaluate habitat we was described by Neu et al. (1974). AU 8 owls were not tested for each cover type because some WNW typesdid not occur in all home ranges or were so uncommon in some home ranges that statistical tests would not have been meeningfuL

b tr <0.5%. -e As indicated by the percent of owl locations in the cover type.d Other cover types present on the area are listed in Appendix 2.

home ranges. The 1A male and femalewere not included in home range overlapcalculations because it was not clearwhether they were an established pair.

Total overlap between the home rangeof 1 individual and the home ranges of 2adjacent pairs could only be calculated forthe 4A female. She shared a total of 39%of her home range with the 1A and 5Apairs (Fig. 9).

There was little overlap among thehome ranges of the 3 pairs on the BLMarea. The absence of overlap was not un-expected in the case of the 1C pair, be-cause that pair was located nearly 8 kmfrom the 2C and 3C pairs (Fig. 10). Thelack of appreciable overlap between thehome ranges of the 2C and 3C pairs mayhave been due to the fact that there wasa wide (0.5-1.6 km) corridor of clear-cutland separating the 2 pairs.

Several unmarked owls on the BLMstudy area had home ranges that consid-erably overlapped the home ranges of themarked individuals. These unmarked in-dividuals were heard or seen within thehome ranges of the radio-tagged owls ona number of occasions, including 3 in-

stances when unmarked owls called nearthe center of the area occupied by the 3Cpair. The 2C female regularly traveledback and forth between the home rangesof 2 different males, roosting occasionallywith each of them.

Habitat Selection for ForagingForests.—All of the radio-tagged owls

on the HJA and BLM study areas showeda strong preference for foraging in un-logged old-growth forests (Tables 4, 5;Figs, 13-15; Appendix 2). Use of old-growth stands that had been selectivelylogged was either not significantly differ-ent than expected, or significantly less thanexpected (Table 4), suggesting that selec-tive logging decreased the suitability ofold-growth stands for foraging. Use ofolder second-growth and mature forests(61-200 years old) was variable, rangingfrom significantly less than expected inmost cases, to significantly more than ex-pected in a few cases (Tables 4,5). Use ofyoung second-growth forests (25-60 yearsold) was in all cases significandy less thanor not significantly different than expect-

No. of individuals using cover typePercent of Nottime spent Significantly significantly Significantlyforaging more than different ins thanin typeo expected then expected expected

Percentcoverage onhome rangesr•••••••

THE SPOTTED OWL IN OREGON—Forsman et al. 25

Table 5. Use of cover types for Imaging by 6 radio-tagged adult spotted owls on the BLM study area (Apr–Aug 1980).8

Old-growth conifer forest 20-54 64-98 6 0 081-200.yettr-old conifer forest 0-2 1-2 0 1Mature ash and oak forest 0-tr tr-1 1 1 061-80-Naz-g1d conifer forest 1-45 tr-36 1 1 436-60-yea -old conifer forest 0-2 • 0-tr 0 0 225-35-year-old conifer forest 0-47 0-31 0 1 35-24-year-old conifer forest 2-26 0-1 0 0 6Recent cLear-cuts covered by

brush or grass-forb associations 9-23 0-tr 0 0 6Other 0-2 0 0 0 4

8 Test used to evaluate habitat use was described by Neu it al. (1974). A116 owls were not tested for each cover type because some cover typesdid not occur in all home ranges or were so uncommon in some home ranges that statistical tests would sot have been meaningful.

`As indicated by. the percent of owl locations in the cover type. tr • <0.5%.e Other cover types present on the area Ste listed in Appendix 2.

ed (Tables 4, 5). Areas that had been clear-cut or burned within the previous 20 yearswere rarely used for foraging (Tables 4,5).

Although second-growth forests werenot preferred for foraging in most in-stances, some individuals did spend asmuch as 36% of their time foraging insuch forests (Table 5). This suggested thatyoung forests provided at least marginalforaging habitat after reaching 25-35years of age. It should be noted, however,that little intensive management had beenconducted in any of the young forests oneither study area. We do not know if thesestands would have been used for foraginghad they been intensively managed.

Although recent clear-cuts and burnedareas were rarely used for foraging, mostowls occasionally made long flights (up to1.5 km) across such areas en route fromonepatch of older forest to another. Ifpossible, however, owls usually avoidedcrossing these open areas by travelingthrough corridors of uncut timber aroundcutover areas (Figs. 13-15). This is not en-tirely obvious from the computer maps ofowl movements in Figs. 13-15, becausemany of the lines across clear-cuts con-nected sequential locations determinedseveral hours or even days apart (and thusdid not necessarily represent the actualflight path): -

Talus Outcrops and Rockslides.—Nat-ural talus outcrops and areas of rocky rub-ble along road cuts were absent on theBLM area and uncommon (<0.3% cov-erage) on the HJA study area. The onlyindividual that used areas of rocky rubbleor rock talus significantly more than ex-pected was the 2A male (Table 4). Be-tween June and September, this male wasobserved on 8 different nights, perched atpoints overlooking rockslides. He was ap-parently hunting for pikas, which wereactive both day and night in such areas.Foraging in rock talus ceased entirely af-ter about 15 October, probably becausepikas curtailed their above-ground activ-ities during winter.

Roads and Miscellaneous CoverTypes.—Narrow secondary gravel roadswere common on both study areas. All ofthe radio-tagged owls foraged at least oc-casionally in forests bordering such roadsand appeared undisturbed by the occa-sional passage of vehicles. However, therewas no apparent tendency to concentrateforaging activity near roads or other typesof "edge" areas such as boundaries be-tween clear-cuts and old-growth forests(Figs. 13-15).

Miscellaneous cover types on the HJAstudy area included 7 small rock quarries,Blue River Reservoir, a golf course, a for-ested public campground, a powerline

/ Atir F: ?I • ‘.1 _tviii944 -t- s ittpit 14* di%)

4.1f t14.1 1ri4"1/114111.114 11 lee

useisArgesraNktIC ire4•

*aeliteurrAr /11r _pi stP.

• - vt.t. %it At,,,T, IN91,

I3

ti•

• 1 • I

" 1/1,

A

8

45%

4'1111 044" 0

0 KM I

Omni? I.

ktfl%

tilt* A .43,p

IZ3trot ESSi

ilk/ 011r 340

OLO-ORONITH FOREST 2 6 1 -TO 200-YE AR-OLO FOREST

MATURE RIPARIA N FOREST Me TALUS SLOPES. OUTCROPS

0-TO SO-YEAR-OLD CLEAR•CUTS.OURNS EM SLUE IL ors.

Pall .A1,;)ide ipr. . *It:•ieze..0.- . ,.....,N9 ...00.e0.6...10:

13; bowler generated mip Illustrating home rmcs Ind hebltel use al the 2A ode an the (i. J. Anckews study lee (25 M T5 5 Apray 15- 157!). &Mire •Ymbols fonfesent

toMg.

rso! locations end Mangles represent roosting bunions. ' ' '

rte

Filtifgr*Cielrooteotii

rokorro ‘t

4411.44414P45111141 J

p) 14415r",110114.VIC ir

*. '74.11.1'm M!!4441 ET]

141.41 C::3 TALUS SLOPES. OUTCROPS:tar

1.1.;gr 0 I KM

iOLD-GROWTH FOREST 1,1MATURE RIPARIAN FOR T0-TO 30-YEAR-OLD CLE f1-CUTt. BURP S61-TO 200-YEAR-OLO FOREST '

I MILE

•ko,3,riarp

..o4,14.440:0!

ft16.01'

Fig. 14. Computer generated map Illustrating home range and habitat use of the 4A female on the H. J. Andrews study area (28 Jul 1975-1 Jim 1976). The 2 concentrations oflocations lo the center of the home range represent locations of nests used in 1975 and 1978. Square symbols represent foraging locations end Mangles represent roosting locations.

OLD-GROWTH FOREST

61-TO 200-YEAR-OLD FOREST

36-TO 60-YEAR-OLD FOREST 4•1

25-TO 35-YEAR-OLD FOREST

0-TO 24-YEAR-OLD CLEAR-CUTS

AGRICULTURAL LAND (CLEARED)

*4"AMP

Voi:tGe***7.?j.

i KM

psystIv NI 4„..

• I

11* 11" Cs"thilte """Isd ther) thetheeeg °thee Rthge e th.thfe °I the 2C thee" °1! th• PY4 !bid! r's 03 Apr-2? &s! rem. Won srbob «waging

Wan" and Mangles represent vacating locations.' , , ,

THE SPOTTED OWL IN OREGON—Forman et a/. 29

Table 6. Percent occurrence of roosting in different cover Table 7. Percent of roosts In different tree or shrub speciestypes by radio-tagged spotted owls on the H. J. Andrews study under different weather conditions on the BLM study area (2area (May 1975-Jun 1976) and 81A4 study area (Mar-Sep Mar-1 Sep 1980). Roost sample size is in parentheses.1980). Number of roosts is in parentheses.

Cover typeHA BLM

Old-grdwat-conifer forest 97.6 91.0Mature Oregon white oak and

Oregescsli- forest 4.081-200-year-old conifer forest 1.361-80-year-old conifer forest 0.5 4.025-60-yearLold conifer forest 0.6 1.0

right of way, an agricultural field, and a9.4-ha area of old-growth Douglas-fir thathad been harvested by the shelterwoodMethod (Appendix. 2). On the BLM studyarea, miscellaneous cover types included2 livestock pastures and a 1-ha-gravel stor-age area. Foraging was observed in themiscellaneous cover types on only 2 oc-casions, once when the 2A male foragedin a small rock quarry and once when the3A female foraged in a public camp-ground.

Seasonal Differences in Habitat Selec-tion for Foraging.—Because owls on theBLM study area were only observed for 4months, we did not attempt a seasonalcomparison of habitat selection. On theHJA study area, seasonal differences inhabitat selection for foraging were rela-tively small. During all seasons, foragingwas limited primarily to forested areas,especially old-growth forests. Duringmidwinter (Dec-Feb), when approxi-mately 85% of the diet consisted of arbo-real mammals (see Food and ForagingBehavior), the owls foraged almost exclu-sively in forested areas. During the rest ofthe year, rock talus and recent clear-cutswere used occasionally, but over 90% offoraging occurred in forests >30 years old.

Habitat Selection for Roosting

On the HJA and BLM study areas, welocated 1,098 and 555 owl roosts, respec-tively. Of these, 645 were visually locatedby homing in on radio-tagged owls, and1,008 Were located remotely by triangu-

Tree species

Weather conditionsInter-

mittentor

steadyrain(61)

Over-cast,no

precipi-tattoo(40)

Clear-warm.10<temp

<28 C(160)

Clearandhot. ,

temp /-avi c

(32)

Douglas-fir 57 20 9Western hemlock 8 30 26 31Western red cedar 10 7 11 10Vine maple 2 10 22 28Pacific dogwood 5 13 16 28Incense-cedar 1Other broad-leaved

tree • or shrubs 18 20 14 3Log or limb on

ground 1Included Oregon white oak. western hazel. entombed' oceenspray,

Pacific madrone. bigleid maple. Oregon ash. red alder. golden chin-kapin. and an muchentsied willow.

•

lation. Of the 645 owls observed in dayroosts, 640 were perched on limbs in treesor large woody shrubs, and 5 were perchedon limbs or logs on the forest floor. Allroosts were in forests. Roosting in tree cav-ities was not observed, except in the caseof females about to lay eggs (see Repro-ductive Biology). Over 90% of all roostswere in old-growth forests, indicating thatsuch forests were strongly preferred forroosting (Table 6).

The type of roost tree used and perchposition within the forest canopy were in-fluenced by weather conditions on bothstudy areas. During warm or hot weather,the owls usually roosted low in the forestunderstory in small trees or shrubs, there-by reducing their exposure to solar radia-tion and high temperatures (Tables 7, 8).Spotted owl use of cool roosts during hotweather has previously been described byForsman (1976, 1980,1981a), Barrows andBarrows (1978), and Barrows (1980).When it was cold, raining, or snowing,owls roosted significantly higher in theforest overstory, usually in old-growth ormature conifers (Tables 7, 8) (HJA areat 12.75, df 191, P < 0.005; BLM areat 12.22, df .• 251, P < 0.005). In addi-

Intennittemor steady rain

Overcea. noprecipitation

Clear and Gear-warm,cold. temp . 10.c temp Claw.hot.

s <28 C temp sob C

30

WILDLIFE MONOGRAPHS

Table 8. Characteristics of roosts used by spotted owls under different weather amditions on the H. J. Andrews study area(May 1975-Jun 1976) end SW study area (Mar—Sep 1980).

Weather conditions and area

HJA BLM HJA BLM HJA BLM HJA BLM HJA BIM

No. of roosts measured 103 61 69 40 60 0 78 160 12 32Mean perch height (m) 23.5 14.9 18.7 7.3 20.4 10.1 5.2 3.6 4.3Mean dbh of roost trees (cm) 115 84 85 33 88 45 25 18 15Roosts protected from rain (90. 58.3 51 23.2 5 7.2 1.3 3.0 0 0Roosts in trees >80-years-old (%) 81.6 62 56.5 10 59.4 16.7 6.0 0 0

• A roost was protected from rain (and mow) if the owl was perched close against the trunk of a large tree. either cm the underside of a leaningtrunk or under a series of large overhanging limbs

tion, they tended to roost close against treetrunks, under some sort of overhead pro-tection, such as a leaning trunk or a clumpof overhanging limbs, where . they weresheltered from rain and snow (Table 8).One result of this study that seemed tocontradict the hypothesis that spotted owlsselected the coolest possible roosts duringhot weather was that the majority of roostson the BLM study area were on south as-pects, ragardless of weather conditions(Table 9). We could not explain this re-sult.

Habitat Selection for NestingForty-seven nests were located during

the study, all in trees. Forty-two (90%)were in multilayered old-growth forests,2 (4%) were in a stand dominated by 100-140-year-old Douglas-fir, with a few old-growth Douglas-fir trees interspersed, and3 (6%) were in 70-80-year-old forestsdominated by Douglas-fir, with residualold-growth trees scattered among theyounger trees (0-5 old-growth trees/ha).In most cases, nest sites on the HJA andBLM study areas were centrally locatedwithin home ranges (Figs. 13-15).

Canopy closure was measured at 26nests and averaged 69% (range 85-91,SE 2.65). Only 2 nests were located instands with canopy closure less than 55%.Both were located at the edge of small dirtinging roads, such that the canopy on 1side of the nests was entirely open. The

effect of the logging roads was the sameas if the nests had been located at the edgeof small clearings in the forest.

There was no significant preference forany particular exposure for nesting (Fig.16) (x2 5.66, cif .• 4, P > 0.05) (to con-duct a x2 test we assumed that slopes onnorth, east, south, and west aspects wereequally available). It was our impressionthat aspect was of minor importance innest site selection as long as a suitable nesttree and a closed canopy forest were pres-ent.

Nest trees were found on slopes rangingfrom 0 to 85%. Most were located on slopesunder 75% (Fig. 16). The majority (57%)of nests were located on the lower half ofhillsides. Thirteen (28%) were on the up-per half of hillsides, 2 (4%) were on ridge-tops, and 5 (11%) were on relatively fiatground in areas of moderate topographicrelief.

The majority (84%) of nests locatedduring the study were within 250 m of astream or spring (range =-15-1,41:7 m)(Fig. 16). We could not determine wheth-er there was a significant preference fornest sites close to water, however, becausewe did not determine mean distances be-tween randomly selected points and thenearest source of water in each nest area.We suspect, however, that the presence ofwater is an important factor in habitat se-lection by spotted owls because we fre-quently observed both captive and wildspotted owls drinking and bathing.

1816141210a64

18

16

14

12

10

a64

Ifaito NA

DISTANCETO H20(M)

THE SPOTTED OWL IN OREGON —Forsman et al. 31

Table 9. Percent of spotted owl roosts on different aspectsunder different weather conditions co the KM study area (2Mar-1 Sep 1980). Roost sample size is in parentheses.

`"-s•her conditions

Inter- Over- Clear- Clam,-mittent east. no warm. 10< hot.

- - or steady prompt- temp temprain titian <28 C L211 C

Aepect (68) (42) (171) (34)

Flat 31 21 9 6North (315-40 9 10 22 15East (45-1341 - 5 17 13 20South (135-2,24• 36 33 40 41West (225-3141 19 19 16 18

Reproductive Biology

Nests.—Of 47 nests, 30 (64%) were incavities and 17 (36%) were on platformsof sticks or other debris on limbs. Forty-one (26 cavities, 15 platforms) were inDouglas-fir trees, 4 (3 cavities, 1 platform)were in white fir, 1 (cavity) was in a pon-derosa pine, and 1 (platform) was in a sug-ar pine. All 30 cavity nests were in old-growth trees. Nine of the platform nestswere in old-growth trees, 7 were in 100-200-year-old trees, and 1 was in an 80-year-old tree. Forty-five nests were in liv-ing trees, and 2 were in dead trees. Nestheight above the ground ranged from10.0-55.3 m (2 = 27.3, SE 1.67) (Table10).

Of the 30 cavity nests, 26 were in cav-ities that formed when tops of old-growthtrees broke off, exposing the hollow inte-rior of the tree (Fig. 17). Four were incavities that formed when large limbsripped loose from tree trunks (Fig. 18).Cavities in treetops were typically shapedlike a stovepipe, open at the top, with thewalls formed by the hardened shell of thebroken trunk. All but 2 trees in whichtreetop nests were located had live sec-ondary crowns that overtopped the bro-ken trunks and shielded nest cavities fromrain and direct sunlight (Fig. 17). The 2exceptions were dead trees.

Of 17 platform nests, 9 were in abnor-mally dense clusters of limbs ("brooms")caused by dwarf mistletoe (Arceuthobiumspp.) infections. _Eight were on relativelyhealthy limbs". At least 9 were constructedby other species, including goshawks (1),

Fig. 16. Placement of 47 spotted owl nests located in Ore-gon from 1970 through 1980. with respect to aspect. slope.and distance to water. North 315-44•, East 45-134•.South ■■ 135-22 •. West 225-314•. Distance to water wasunknown at 3 nests.

Cooper's hawks (1), red-tailed hawks (1),woodrats or squirrels (3), and unknown (3).The other platforms consisted of naturalaccumulations of debris (twigs, coniferneedles, cones, lichens, etc.) that had be-come entrapped by limbs. Nests in treesinfected by dwarf mistletoe were fre-quently situated in such dense tangles oflimbs that it was difficult to see into thenests. Although most (65%) platform nestswere located against tree trunks, somewere located asfar as 2.8 m out on limbs.

The nesting substrate in all cavity nestsconsisted of natural accumulations of rot-ted wood and, frequently, many old co- -

nifer needles, cones, and small twigs. Plat-form nests were generally lined with acompacted mass of old conifer needles andsmall twigs. Nests that were reused severaltimes also had a powdery mixture of bonesand old pellets intermixed with the sub-strate. None of the owls that we observedattempted to build or repair nests. Eachfemale simply scraped out a shallowdepression in the existing debris. Thedepression eventually became lined witha small amount of down from the broodpatch. In addition, 1 nest contained a few

32 WILDLIFE MONOGRAPHS

Table 10. Characteristics of 47 spotted owl nests and nest trees located In Oregon (1970-79).

MeasurementNests in cavities Naa in platforms►

Nb Mean SE flange N Mean SE Range

Nest height (m) 30 30.1 1.78 11.7-55.3 17 22.0 1.58 10.0-37.5Tree height (m) 28 38.1 2.37 18.6-64.0 16.- 42.0 3.42 23.5-68.6Bole height (m)' 20 15.7 1.14 7.3-24.4 14 13.6 1.48 6.7-28.6Dbh (cm) 28 135 6.03 74-205 16 106 11.93 36-179

. Tree diam. at nest height (cm) 22 78 4.83 51-105 8 75 11.99 23-12.6No. of secondary tops in nest tree 23 3 0.50 0-8Avg. nest diam. (em)" 20 50 0.93 30-112 8 62 1.32 41-76Cavity depth (cm) 22 26 7.39 0-123Width of cavity entrance (cm) 18 30 4.11 10-77

• Included stick platforms constructed by other birds or mammals as well as platforms that formed naturally when accumulations ef Was.mesa and other debris collected on top of dense tangles of limbs.

b All variables were not measured at each nest because all nest trees were not climbed and because some measurements did not apply to allnests. Also. some nest trees were cut or fell down before w obtained a full set of measurements.

Bole height is height above ground of first live limb.II Average diameter of cavities or platforms was determined by measuring at right angles moss the widest and narrowest alma of the seek

summing the 2 values. and dividing by 2.

green sprigs of Douglas-fir needles liningthe nest scrape. It appeared that the fe-male had added these to the nest. Bent(1938:185) reported that female barredowls sometimes lined their nests with li-chens or "fresh, green sprays of whitepine." Based on our observations, we be-lieve that reports of nest building by spot-ted owls (Bendire 1882, Heller 1893) wereerroneous. Both Bendire and Heller sawspotted owls using platform nests and as-sumed that the owls had built the plat-forms.

Eighty-one percent of all nests in north-western Oregon were in cavities, com-pared to only 50% in the Klamath Moun-tains and on the east slope of the Cascades.These differences appeared to reflect re-gional differences in availability of the dif-ferent nest types. Dwarf mistletoe infec-tions in Douglas-fir (and the numerousdebris platforms that were associated withdwarf mistletoe infections) were commonin the mixed conifer forests of the Klam-ath Mountains and the east slope of theCascades, but did not occur in westernOregon.

Of 25 nests that were checked in 2 ormore years, 17 were used more than once.The longest period of occupancy was anest that was used 6 times in an 8-yearperiod (in 2 years owls were present butdid not nest). The maximum number of

nests used by a single pair was 5 during a6-year period. In the latter case, 2 nestswere in the same tree and all 5 were with-in 50 m of each other. One of the 5 nestswas used twice. Distances between alter-nate nest sites used by individual pairsranged from the minimum just describedup to 1.2 km.

The attrition rate of nest trees was high.Eight (17%) of the nests examined from1970 through 1978 were no longer usableby the end of 1978. Three were cut down,4 fell down, and the bottom of 1 rottedthrough. In addition, 2 nest trees died butremained standing. One was not subse-quently reused. The other was reused atleast twice after it died.

Timing of Nesting.—The mean date ofclutch initiation (calculated by backdat-ing from the date when owlets left thenest) was 2 April (N 21, range Mar-19 Apr). Owls at lower elevations in south-western Oregon generally nested_ 1-2.weeks earlier than owls in the Coast Rangeand Western Cascades and 2-4 weeks ear-lier than owls on the east slope of the Cas-cades.

Clutch Size and Rate of Egg Produc-tion.—Clutch size was 2 at each of 4 nestsexamined during the incubation period.In lieu of information on clutch size atmost nests, we used the number of youngleaving the nest as an index of minimum

Tim Stu TED OWL IN OREGON —Forsman et al. 33

fig. 17. A typical spotted owl nest cavity in the broken topof an cid•growth Douglas-fr. Note large hobs that have groomupward, forming a secondary crown.

clutch size. Broods of 2 were most com-mon, but broods of 1 were also common(Table 11). Broods of 3 were uncommon.The mean number of young produced persuccessful nest during all years combinedwas 2.0, again suggesting a modal clutchof 2. Bent (1938:204) reported that "Thespotted owl lays two or three eggs, usuallyonly two, and very rarely four . ." Weobserved no broods of 4. Bendire (1892)described a nest containing 4 young, andDunn (1901) reported a nest with 4 eggs.

One of our captive owls laid clutches of2-3 eggs every year from 1975 through1980. In 2 years she laid a second clutchafter the first was removed from her nestbox. Regardless of clutch size, the intervalbetween laying of successive eggs was al-aways 72 ± 6 hours. We were unable toobtain comparable data from wild birdsbut suspected that the rate at which eggs

Fig. 18. A spotted owl nest In a large cavity that formedwhen a limb ripped loose from the bunk of an old-growthDouglas-fir.

were laid was similar in both wild andcaptive individuals.

Incubation Period.—In 1976, one of ourradio-tagged females abandoned her 2eggs after incubating for approximately28 days. Assuming that 1 of the eggs waslaid 72 hours after the other, then the 2eggs had been incubated for 28 and 25days, respectively. We recovered the eggsand found that both contained healthyembryos that were nearly full term. Basedon the development of these embryos weestimated that the eggs would havehatched in another 2-5 days. This wouldindicate an incubation period of 30 ± 2days.

Percent of Population Attempting toNest.—On the average, only 62% of thepairs checked each year attempted to nest(range 16-89%) (Table 11). Why somepairs did not nest in some years was un-known. Local or regional fluctuations inprey abundance may have influencedbreeding, a phenomenon that has beenobserved in other owls (Pitelka et al. 1955,Southern 1970, Busch et al. 1972).

Nest Success.—Eighty-one percent ofall nesting attempts were successful (Ta-ble 11). Nest failures were attributable toa variety of causes. Some pairs acted as ifthey were going to nest (displaying at thenest) but then apparently never laid eggs.Known causes of failure after eggs werelaid included abandonment of eggs con-taining healthy embryos (3 nests), destruc-

34 WILDLIFE MONOGRAPHS

Table 11. Summary of nesting data for spotted owls in Oregon.

Year1972 1973 1974 1973 1976

No. pairs checked 28 31 37 14 20Percent nesting 89 16 = 46 57 757Percent of nesting pairs fledging young 92 49._ 72 75 87No. of broods observed (by size class)

1 young 8 1 7 0 0 •2 young 12 1 6 1 6t-3 young 3 0 0 0 1.Undetermined 0 0 0 5. 6_

Mean no. young fledged/successful nest 1.87 1.50 1.46 2.00 2.14Mean no. young fledged/nesting attempt 1.72 0.60 1.11

tion of eggs or nestlings by predators (2nests), death of nestlings from disease orexposure (1 nest), destruction of eggs whena platform nest collapsed (1 nest), andpremature departure of owlets from thenest (3 nests). Of the 3 cases where fe-males abandoned eggs, it appeared thatthe female was killed at 1 nest (she dis-appeared overnight) and that incubationwas disrupted at another nest when a largeoverhanging piece of rotten wood fell intothe nest cavity. In the third case the fe-male (a radio-tagged individual) simplystopped incubating.

Copulation and Nest Selection.—Be-tween October and January, adult spottedowls lived a largely solitary existence,roosting together and calling infrequent-ly. In February or early March, 33-68 daysbefore the eggs were laid, the residentmale and female on each territory beganto roost together near the eventual nestsite (usually the same nest used in pre-vious years). They also began to call al-most every night, especially at dusk justbefore they began to forage and again atdawn when they rejoined near the nest.We did not determine when males beganto feed females, but both pair memberscontinued to forage until about 12 daysbefore eggs were laid. Thereafter, femalesbecame increasingly sedentary, seldommoving more than a few hundred metersfrom the nest. During the last 5-7 daysbefore the eggs were laid, females spent

most of their time at night perched in thevicinity of the nest tree, waiting for malesto arrive with food.

Visual observations of pair behaviorduring the prenesting period were madeat 2 nests, beginning 13 and 17 days, re-spectively, before the first eggs were laid.At both nests, observations were conduct-ed each evening from about 1.hour beforesunset to 2 hours after sunset. Both pairscopulated on the first evening they wereobserved, indicating that copulation be-gan at least 2-3 weeks before the eggswere laid. Copulation and the behaviorassociated with it were relatively stereo-typed. Each night the first copulation usu-ally occurred at dusk just ter the owlsleft their roosts. Before copulation oc-curred, the male usually parled near thefemale and gave the 4-Note Location Callor Agitated Location Call at irregular in-tervals (see Vocal Displays). The femaleresponded with the Contact Call at inter-vals of 5 seconds to several minutes. Thesepreliminary vocalizations continued forseveral minutes or more, after which themale flew to the female and they copu-lated. During copulation the femaleperched crosswise on a limb and gave theFemale Copulatory Call. The male archedhis wings over his back and fluttered themrapidly until copulation was completed.In addition, the male gave either the Male

gr11latory Call or the Agitated Locationduring copulation. Occasionally, be-

THE SPOTTED OWL IN OREGON—Forsman et al. 35

fore copulation occurred, the male flewinto the nest tree and perched near thenest while giving the Nest Call.

After copulation occurred the maleusually flew to a perch near the nest andgave the Nest Call for several minutes be-fore frying off to hunt. The female eitherwatched this performance from a nearbytrerbr followed the male into the nesttree. During this display, females some-times entered the nest and gave the NestCall for several minutes.

The above sequence of events, with onlyminor variations, was repeated on mostevenings during the last 13 days beforeeggs were laid. The maximum number ofcopulations observed in 1 evening was 2.The minimum time between consecutivecopulations was 35 minutes. Both pairsceased copulation within 4 days aftercompleting their clutches.

Adult Behavior During Incubation andBrooding.—Several days before the firstegg was laid, females began to roost in thenest or on limbs near the nest during theday. A female observed in 1974 was typ-ical. On 4 April, at least 2 days before shelaid the first egg, she began roosting in her.nest cavity. During the first 2 days thatshe roosted in the cavity we always foundher perched upright looking out of thecavity. On the evening of 6 April she waslying flat in typical incubation posture andwe suspected she was about to lay or hadalready laid the first egg.

Incubation was performed entirely byfemales. and was initiated soon after thefirst egg was laid. During the first 48 hoursof incubation females occasionally left thenest for up to 2 hours at night, possibly toforage near the nest. By the time the sec-ond egg was laid, however, females in-cubated continuously except for occasion-al 10-20-minute periods during the nightwhen they left their nests to regurgitatepellets, defecate, .preen, or receive foodfrom the male. During incubation, all for-aging was done by the male. Males typi-cally roosted within 200 m of their nesttrees during the day and began to forageshortly. alter sunset. Occasionally, how-

ever, radio-tagged males roosted up to 1.1km from their nests.

When arriving at the nest with food,males usually announced their arrival bycalling quietly (Contact Call or 4-NoteLocation Call) from a tree near the nest.Females usually responded by leaving thenest to accept the food. Food transfer wasfrom beak to beak. If the female did notleave the nest, the male would usually car-ry the food directly to the nest.

At most nests, eggs hatched between 8April and 20 May. For 8-10 days there-after, females brooded their young almostconstantly, leaving their nests for onlybrief periods during the night. When theyoung were 2-3 weeks old, females beganforaging for progressively longer periodseach night. Two radio-tagged femaleswere first observed foraging considerabledistances (up to 2.6 km) from their nestswhen their young were approximately 16and 13 days old, Tespectively-.-Up to-thattime the foraging activities of both fe-males had been restricted to within a fewhundred meters of the nest. After they be-gan to forage farther from their nests bothfemales regularly left their young unat-tended for periods of 1-4 hours and oc-casionally up to 6 hours on clear warmnights.

Male behavior during the nestling pe-riod was essentially the same as during theincubation period except that males car-ried food directly to the nest and left itthere when females were away from thenest. It appeared that the food was simplyleft in the nest for the female to feed tothe young. Males were not observed feed-ing nestlings.

During the day, females continued toroost in their nests until 3-6 days beforethe young left the nest. Thereafter, fe-males were often found roosting in a treenear the nest, maintaining vocal contactwith their young by giving the ContactCall at irregular intervals during the day.

Defense of the Nest and Young.—Adult spotted owls behaved aggressivelytowards us only if we climbed nest treesor approached owlets that had left the nest

ik.er• - Pa;• •

14440.

4.;

36 WILDLIFE MONOGRAPHS

Fig. 19. Juvenile spotted melt, Illustrating development during the first summer. (A) 42-46 days old, 10 days after leaving thenest: (B) ca. 110 days old In August: (C) ca. 150 days old in late September.

but could not fly. Some individuals wereextremely pugnacious, striking repeatedlywith their talons when their nest trees wereclimbed or their young were approachedon the ground. Others flew around us aswe climbed their nest trees, but never ac-tually hit us. At most nests both males andfemales participated in nest defense.

Two instances were observed in whichspotted owls defended their young frompredators other than man. In 1 encounter,a female spotted owl dived repeatedly at2 ravens that were hopping around onlimbs just outside her nest cavity. This in-cident was instigated when we called thefemale out of her nest during the day. Al-though the owl seemed to be holding herown against the ravens, we frightened theravens away rather than risk losing a clutchbecause of human disturbance. On anotheroccasion a male spotted owl attacked anddrove off an adult female Cooper's hawkthat was in the process of attacking a re-cently fledged owlet.

Development and Behavior of Nest-lings.—When they hatched, owlets werecovered by pure white natal down, andtheir eyelids were closed. Until their eye-

lids opened 5-9 days later, the owlets wererelatively inactive. When owlets were 10-20 days old, the juvenile (mesoptile)plumage, which was pale brown andbarred, began to replace the natal downon the wings, back, and top of the head(Forsman 1981b). At this age owlets beganto sit upright in the nest and became moreactive. At night they could sometimes beheard begging in the nest, especially whenadults called or entered the nest area.

Most owlets left the nest when they were34-36 days old, between 15 May and 24June. Several days before fledging, someowlets began to perch on limbs adjacentto the nest. Others peered from nest en-trances, but did not venture out until theyfledged. Nine owlets that were raised inplatform nests fell or jumped from the nestprematurely when they were 15-25 daysold. Of these, 7 were killed by the fall ordisappeared before reaching the flyingstage. No owlets were lost in this mannerfrom cavity nests, suggesting that cavitynests provided a more secure environmentfor the young. Zarn (1974) was misin-formed when he stated that spotted owlsleave the nest when only 1 week old.

THE SPOTTED OWL IN OREGON—Forsman et al. 37

Fig. 20. Contrast between juvenile minces (A) and feeble's acquired in subsequent molts op. Juvenile rectrioes were sharp-pointed and had a clear white temiinal band. Rectrices acquired in later molts were more rounded and had the terminal bandmottled with brown. -

When owlets left the nest at the normal34-36 days of age, replacement of the na-tal down by the juvenile plumage wasnearly complete, but the remiges wereonly one-half to two-thirds developed (Fig.19). As a result, fledglings were weak fliersand often fell to the ground when theyleft the nest. Some managed to flutterdownward into trees below the nest.

Within 3 days after leaving the nestmost owlets were able to fly or climb intoelevated perches. However, individualsthat left the nest prematurely (15-25 daysold) sometimes spent as long as 10 dayson the ground. Juvenile spotted owls wereclumsy but persistent climbers and couldclimb almost any tree trunk to reach anelevated perch. To climb trees, owletsgrasped the rough bark with their talonsand then walked upward while flappingtheir wings. When they became tired theystopped moving and draped their out-spread wings against the trunk to holdthemselves in place. Occasionally, the beakwas also used to grasp limbs or flakes ofbark. Once strong enough to get on theground, owlets usually roosted on limbs intrees, or occasionally, on logs or other de-bris on the ground.

Development and Care of Fledg-lings.—After leaving the nest, siblings

stayed together for the rest of the sum-mer, usually remaining near-the nest. Of17 broods that were checked at regularintervals during summer, 13 were stillwithin 250 m of the nest in late August,and 2 were 490 and 670 m from the nest,respectively. Two broods could not be re-located in August. When last observed inJuly they were 365 and 1,050 m from thenest, respectively. Of the 15 broods locat-ed in August, at least 7 were still within200 m of the nest in early September.

The mobility and foraging skills of owl-ets improved gradually during summer.Within a week after leaving the nest, mostowlets were able to make short, clumsyflights between trees and began to exhibitincipient foraging behavior, pouncing onleaves, pieces of moss, or small twigs andtearing at them with the beak or talons.Three weeks after leaving the nest owletswere able to hold and tear up prey ontheir own, and by late July they becameproficient at pouncing on relatively im-mobile objects such as crawling insects. Bymid-August they were able to capture livemice that we tethered in roost areas.

As the owlets matured, the adults roost-ed with them less frequently. On 49 oc-casions when we located owlets during theday in August or September, adults were

38 WILDLIFE MONOGRAPHS

-= •

present only 23% of the time. By compar-ison, owlets observed in May and Junewere attended by 1 or both adults 81% ofthe time.

Although adults infrequently roostedwith their young in late summer, theycontinued to feed them at night until lateAugust or early September. Wheneverthey were hungry, owlets gave the Beg-ging Call loudly and persistently, espe-cially so if an adult appeared with food orcalled nearby. Begging behavior declinedin late August and ceased by late Septem-ber or early October. By this time owletswere apparently self-sufficient, althoughthey had not yet begun to disperse.

Replacement of the juvenile plumageby the Basic I plumage began at age . 47-56 days and was complete by the end ofSeptember or early October (Fig. 19).During this molt all juvenile feathers ex-cept the remiges, rectrices, and greaterprimary coverts were replaced. Owlets inthe Basic I plumage were indistinguish-able from adults except that their rectriceshad clear white, sharp-pointed tips. Thesedistinctive rectrices were retained for thefirst 26 months of life (Forsman 1981b).Tips of rectrices acquired in later moltswere more rounded and were mottled withbrown (Fig. 20).

Juvenile Mortality and Natural Ene-mies.—In 1972, we observed 29 owlets atweekly or biweekly intervals from the timethey left the nest until late August. Only19 (65%) were still alive at the end of Au-gust. The cause of mortality was deter-mined in only 1 case when an owlet waskilled by a great horned owl. In all otherinstances, owlets simply disappeared andwere presumed dead. Predation was thesuspected cause of mortality in most in-stances because owlets appeared healthyshortly before they disappeared.

One other case of juvenile mortality ob-served during the study was attributableto a great horned owl (see Dispersal ofJuveniles). The fact that great horned owlswere the cause of mortality in both casesin which the predator could be identified,suggested that great horned owls were animportant source of juvenile mortality.

Great horned owls were heard or seen inmost areas occupied by spotted owls.

Because they were so secretive duringthe day, spotted owls suffered little mor-tality from diurnal raptors. This was in-dicated by the fact thgt there were at least10 occasions during the study when spot-ted owls nested successfully within 400 m.of active goshawk nests. In 1 instance, nestsof the 2 species were only 125 m. apart.As mentioned earlier, we did see 1 in-stance in which a female Cooper's hawk.attempted to capture a recently fledgedowlet.

Dispersal of Juveniles.—In August 1975we radio-tagged 2 owlets each on the HJAstudy area and on the east slope of theCascades near Abbot Butte, DeschutesCounty. Before the radio-tagged owletsleft their parental nest areas, 1 of thetransmitters at Abbot Butte stopped trans-mitting, and 1 of the owlets on the HJAstudy area died. Movements of the other2 owlets are described below.

After leaving the nest in June the owleton the HJA area spent the entire summerwithin a 35 ha area around the nest (Fig.21). The owlet at Abbot Butte was onlychecked at 2-3-week intervals during thesummer, so the full scope of its move-ments was unknown. Each time it was lo-cated during the summer, however, it waswithin 300 m of the nest tree, so we as-sumed that it did not travel far from thenest area.

On 12 October the owlet on the HJAarea suddenly left the parental nest areaand within 2 days moved 4.4 km to thesoutheast. For the next 45 days it wan-dered about over an area of approximate-ly 2,900 ha, just east of the parental homerange (Fig. 21). The farthest straight-linedistance traveled from the parental nestduring this period was 10.1 km. On 26November the owlet was killed, appar-ently by a great horned owl. A large owlpellet containing remains of the owlet wasfound near the discarded transmitter.

We were unable to determine exactlywhen the owlet at Abbot Butte left theparental nest area, but on 18 October itwas located 16.4 km southeast, in an old-

Fig. 21. Movements of the 4A owlet on the H. J. Andrews study ores (141 Jul-26 Nov 1975). Before It became Independent from Its parents the owlet confined Its activities to a

35.be area around the parental nest (are inside dashed line). Thereafter. It wandered over a large area. The home ranges of the 4A female (the owlet's mother) and 5A pair are

Indicated by solid Ines.

co

40

WILDLIFE MONOGRAPHS

Fig. 22. Areas where data on fool habits of spotted owlswere collected in Oregon from 1970 through 1980. NumbersIndicate which tables in text should be coreulted for each area.

growth forest of Douglas-fir, white fir, andponderosa pine. It had apparently trav-eled across extensive areas of open pon-derosa pine forest to arrive at this location.When we checked this owlet again on 6December, it had been killed and eatenby an unknown predator. Its remains werefound 12 km southeast of the parental nestsite.

Food and Foraging Behavior

Composition of the Diet.—From April1970 to September 1980, 4,527 prey items

were identified from pellets or kills of 62pairs of spotted owls, including 19 pairsin the Coast Ranges, 23 pairs in the Klam-ath Mountains, and 20 pairs in the Cas-cade Range (Fig. 22). The diet included31 species of mammals, 23 species of birds,2 species of reptiles, a crayfish, a terres-trial snail, 29 genera of insects, and severalunidentified spiders (Appendix 3). Nine-ty-two percent of all prey and 98% of totalprey biomass was comprised of verte-brates. Mammals comprised over 90% ofthe biomass consumed in all areas (Tables12-15). Mean prey weights calculated for10 different areas ranged from 54 to 150g (Tables 12-15).

The composition of the diet variedamong regions and forest types. In gen-eral, the diet was dominated by flyingsquirrels and red tree voles in forests ofDouglas-fir and western hemlock, and bydusky-footed woodrats in mixed coniferforests in the Klamath Mountains (Tables12-15). Other prey that were importantin terms of numbers or biomass were deermice, western red-backed voles, snowshoehares, bushy-tailed woodrats, and pocketgophers. Pocket gophers and western red-backed voles were most common in thediet at higher elevations in the CascadeRange and Klamath Mountains (Areas B,C, D, E, G in Table 15). Birds comprised5.3% of the diet on the average (range -2.6-9.3%). Insects were uncommon in thediet except in a few areas, and in all areascomprised only .a small fraction of the to-tal prey biomass (Tables 12-15).

Seasonal Variation in the Diet.—Onthe HJA study area, flying squirrels com-prised over 60% of all prey taken duringfall and winter (Table 16). The proportionof flying squirrels in the diet began to de-cline in late March and April, and by mid-summer flying squirrels comprised only27% of the diet. As the proportion of flyingsquirrels in the diet declined during springand summer, snowshoe hares, shrews,pocket gophers, western red-backed voles,small birds, and insects became increas-ingly common in the diet (Table 16). Al-though the percentages differed some-what, seasonal changes in the diet on the

THE SFOTTED OWL IN OREGON-Forsman et al. 41

Table 12. Diet of spotted owls occupying forests of Douglas-fir and western hemlock in the Oregon Coast Ranges. Prey itemsample size is in parentheses..

Percent of total preyPercent

of biomass

1970(45)

1972(694)

1973(126)

1974(33)

1975(60)

1980(256)

Allyears

combined(1.214)

Allyears

combined(100.961 g)

MammalsN. IlyinereNfiel 51.1 31.6 37.1 45.5 58.3 34.4 35.2 47.2Red tree vole 20.0 21.0 34.7 6.1 11.7 9.4 19.1 6.0Deer mouse - 4.5 15.6 10.5 30.3 5.0 2.3 11.7 3.0W. red-backed vole 6.3 4.9 8.3 4.7 5.5 1.5Woodrat spp.' 6.7 5.2 2.4 6.1 6.3 4.9 15.3Snowshoe hare or brush rabbit 2.2 2.7 2.4 3.0 8.3 9.0 4.3 16.5Microtus spp. 1.2 1.7 0.7 0.3Sore: spp. 1.4 1.6 1.7 0.4 1.2 0.1Other mammals 2.2 3.3 3.0 6.2 3.3 4.2Unidentified mammals 4.2 0.8 3.0 1.7 1.6 3.0 3.1

Birds 2.7 3.2 3.3 4.7 3.0 2.5Reptiles 2.2 0.8 0.3 0.1Insects and arachnids 11.1 4.8 1.6 3.0 21.1 7.8 0.2

No. of pairs from which pellets were collected each year ranged from 4-8. x s for differences in diet among years 151.86. df 10. P < 0.005(test based on numbers of prey). Only years in which N > 50 were compared.

Prey species not specifically identified in the table are listed in Appendix 3, along with scientific names of all prey.

BLM study area closely paralleled thechanges observed on the HJA study area(Table 17).

On all study areas, predation on brush

rabbits and snowshoe hares was restrictedprimarily to small juveniles capturedbetween May and August. Seventy-fourpercent of the individuals captured

Table 13. Diet of spotted owls in forests of Douglas- it and western hemlock on the H. J. Andrews study area. Oregon. Preyitem sample size is in parentheses.•

Percent of total preyPercent

of biomes

1971(1112)

1975(99)

1978(472)

1977(3S)

19711(26)

0111.1=1(617)

call=(69.946 g)

MammalsN. flying squirrel 27.5 32.3 50.0 50.0 34.7 42.4 57.5Red tree vole 12.1 12.1 13.3 21.1 15.4 13.3 4.2W. red-backed vole 1.7 7.1 11.9 7.9 23.1 9.2 2.5Deer mouse 17.0 10.1 5.7 7.9 8.7 2.3Snowshoe hare 4.9 6.1 0.9 3.8 2.5 15.0Bushy-tailed woodrat 2.2 2.0 2.1 2.6 3.8 2.2 6.9Sorer spp.b 6.0 1.0 2.1 2.7 0.3W. pocket gopher 6.0 10.1 1.7 3.6 3.6Microtus app. 3.3 1.0 1.7 5.3 2.0 0.5Townsend's chipmunk 0.6 4.0 1.1 • 1.2 1.1Other mammals 3.3 5.1 1.1 2.6 7.7 2.3 2.6Unidentified mammals 2.8 5.1 3.6 3.3 1.5

Birds 5.0 1.0 2.5 2.6 7.7 3.1 1.9Reptiles - as 0.1 0.1Insects and arachnids 7.1 3.0 2.3 3.8 3.4 tr

No. of pain from which pellets were collected each year ranged from a to 7. its for differences in diet among yews ,. 91.57. df 10.9 < 0.005(test based on numbers of prey). Only years in which N s• 50 were compered.

th Prey species ski specifically identified in the table are listed in Appendia 3. along with ocientifie memo of all prey.tr c0.0511.7

42

WILDLIFE MONOGRAPHS •

Table 14. Diet of spotted owls occupying mixed cotter end mixed evergreen forests In the Kiwnath Mountains. Oregon. Preyitem sample size is in parentheses,

Percent of total preyPerwmi

of biomes

1972(30)

1973(155)

1974(97)

1976(97)

1977(175)

1978(100)

Li =d(651)

=

(96.167 g)

Mammals•••• M.

Dusky-footed woodrat 43.3 40.7 39.2 46.5 33.7 37.0 39.0 69.9N. flying squirrel 6.7 18.7 17.5 26.8 14.9 16.0 17.7 13.6W. red-backed vole 3.3 7.8 7.2 4.1 10.9 19.0 9.5 1.4Deer mouse 20.0 5.2 4.1 7.2 3.4 2.0 5.0 0.7Mfcrotus spp.• 1.9 4.1 6.2 12.6 5.0 6.1 1.5Red tree vole 10.3 5.2 1.0 3.4 4.0 4.9 0.9Brush rabbit 3.2 1.0 1.0 3.4 4.0 2.6Other mammals 10.0 3.2 9.3 1.0 4.0 3.0 4.3 1.5Unidentified mammals 6.7 3.2 2.1 4.6 1.0 2.8 1.5

Birds 3.3 5.2 10.3 4.1 6.3 8.0 6.4 2.8Insects 6.7 0.6 2.1 2.8 1.0 1.7 tr.-

No. of pain from which pellets were collected each year ranged from 2 to 11. x l for differences in diet among yearn =Ma < aces(test Wed on number of prey). Only years in which N > 50 were compared.

', Prey species not specifically identified in the table are heed in Appendix 3. along with scientific names of all prey.tr <0.05%.

weighed <400 g and 92% weighed <700g. The largest lagomorph captured was asnowshoe hare weighing approximately1,500 g.

Pocket gophers were commonly takenby owls on the HJA study area during Mayand June when the ground was clear ofsnow, but were uncommon in the diet atother times of the year (Table 16). Simi-larly, predation on coast moles on the eastslope of the Cascade Range was limited tomid-April-early July, with a peak about20 June. Gigef (1965) suggested that coastmoles were most vulnerable to predationby owls between April and June when ju-venile moles were dispersing above-ground. On both the HJA and BLM studyareas, predation on western red-backedvoles peaked in April (Tables 16, 17).

The few mountain beavers taken byspotted owls were small juveniles (150-300 g) captured from 1 to 25 June (Table17). Apparently, young mountain beaverswere susceptible to predation by spottedowls for only a few weeks after leavingtheir natal burrows in June.

Predation on insects was restricted pri-marily to summer and early fall on allareas. Two insects that were especiallycommon in the summer diet in some areas

were Ergates spiculatus (a_ large flyingbeetle) and Cyphoderris monstrosa (alarge arboreal cricket) (Areas B, E, G inTable 15, 17). Marshall (1942) also re-ported predation on Cyphoderris by spot-ted owls.

Annual Variation in the Diet.-An-nual variation in the diet was examinedin 3 areas where the largest samples ofdata were collected (Tables 12-14). Di-etary composition differed significantlyamong years in all 3 areas. It was not clear,however, whether the year to year varia-tion in the diet was due to changes in preypopulations or to sampling bias (pelletswere not always collected at the same timeor from the same pairs each year).

Although statistically significant, di-etary changes among years- did not in-volve a complete restructuring of the diet.In all areas the numerical ranking of - themajor prey in the diet (from most to leastcommon) changed only slightly or not atall from year to year (Tables 12-14). Someprey comprised a remarkably stable pro-portion of the diet, e.g., the dusky-footedwoodrat in the Klamath Mountains (Table14) and the red tree.vole on the HJA studyarea (Table 13).

Foraging Behavior.-Nine predation

=

THE SPOTTED OWL IN OREGON-Forsman et al. 43

Table 15. Diet of spotted owls in 7 areas in Oregon in which data were not subdivided annuity because of ernes sample sizein some years. Totai number and biomass (g) of MY n each sample are in parentheses!

Arm A Area B Arm C Area D

Per-centof

total

elrfl)

Per-centof

totalbiomass(9.641

g)

Per-Centof

total

la)

Per-cental

totalbiomass(9.654(9g

Per-centof

total

(171)

Per-centof

totalbiomass(17.398

g)

Per-centof

totalprey

Per-cent

totalbiomass(6.292

g)

N. flying squirrel 17.8 25.1 12.0 27.4 46.8 53.5 26.4 42.0Red tree vole 49.1 16.2 3.7 2.0 14.9 5.6Woodrat spp.b 10.2 33.0 1.6 8.2 8.1 29.3 3.4 12.6W. red:backed vole 0.9 0.2 14.7, 6.7 20.8 4.8 31.0 9.9Deer mouse 6.8 1.8 2.6 1.1 3.5 0.8 3.5 LOW. pocket gopher 9.4 16.2 0.6 0.5 &O 9.7Snowshoe hare or brush rabbit 2.5 10.4 2.1 10.9 1.7 4.9 2.3 9.5Townsend's chipmunk 1.6 2.5 0.6 0.5 2.3 2.5Other mammals 3.4 6.6 11.0 15.3 2.8 0.4 1.2 0.4Unidentified mammals 5.2 5.6 2.9 0.7 1.2 4.0

Birds 9.3 6.7 3.1 2.4 5.8 4.5 3.5 2.7Reptiles 0.5 0.5Insects 32.5 1.2 6.4 0.1 2.3 0.1

Area E Area F Area G

Ter-centof

total

tra)

Per-centof

totalbiomass(19,102

g)

Per-centof

totsl

12)

Per-centof

totalbiomes(30.314

g)

Per-centof

total

121

Per-centof

totalbiomass(30.310

g)

MammalsN. flying squirrel 19.7 18.1 13.8 29.4 25.1 31.8Red tree vole 38.2 19.1Woodrat app. 26.3 56.3 4.9 24.3 7.0 20.6W. red-backed vole 5.3 1.0 1.1 0.4 5.4 1.4Deer mouse 13.1 2.3 22.1 9.0 8.5 2.0W. pocket gopher 5.3 3.6 5.0 4.8Snowshoe hare or brush rabbit 4.6 9.4 0.7 3.3 5.4 24.9Townsend's chipmunk 2.0 ' 1.3 0.1 1.4Coast mole 0.5 0.5 3.2 1.9Other mammals 1.4 0.2 5.7 4.0 6.8 5.5Unidentified mammals 2.0 2.4 1.2 1.1 1.8 2.2

Birds 8.5 5.3 7.9 8.6 5.8 4.0Reptiles 0.5 0.2Crayfish 0.2 0.1Insects 11.8 0.1 3.2 tr. 25.9 0.5

Location of arms shown in Fig. 18. Teen of data collection. number of pairs ermined and dominant *smuttiest in the different areaswere:Area A: Pellets collected in 1972 and 1974 from 2 pairs occupying old-powth formal Douglas-Br. sneak. California laurel, weans hemlock.

and cant redwood. 10 km met of BrookIngs on the southern Oregon coast.Arm B: Pellets collected in 1970 and 1972 from S pain oscupyim old-growth keen, of Dougleo-fir. Pacific silver fir. and western hemlock

n the west elope of the Cascade Rum ear Tombstone Summit. Linn County. elevation w 973-1,400 ns.Ann C: Pellets collected in 1972. 1973, and 1974 from 3 pain occupying high elevation 11.460-1.700 m) mired forests of white fit. siesta red

fir. Douglas-fir. and ponderosa pine in the southern Cascade Range near Lido of the Woods. ffloseath County.Ares D: Pellets collected in 1972 and 1974 from pairs occupying high elevation forms of white Br. Doughe►fir, aid sham red Br in the

ICkmeds Mountains near Oregon Caves National Monument. Josephine County.Area E Pellets collected In 1972 and 1973 from 3 pain on:upping mired aseciatious of Douglas-fir. while Ir. ponderosa pine, and incense.

• cedar In the southern Cascade Range near Parker Mountain. Klamath County.Ana F: Pollen collected in 1970. 1972. 1973. and 1974 from 2 pain occupying mimed foram of Dangles-6r. grand fir. and Oregon white oak

at the edged the Willamette Valley. 4 km northwest of Corvallis. Benton Comm.Ana G: POW collected in 1972. 1973, 1973. 1978. 1977. and 1978 from 3 pairs canopying mired forests of Dougles-fir. white Sr, ponderosa

pies. end incerorecedar on the OM elope of die Cswasde Range new Abbot Butte. Jefferson County.b Prey omens net specifically identthed in the table are listed in Appendix & along with seleatilie senses of all prey.i ts <0.05%.

,..-_ ....

44

WILDLIFE MONOGRAPHS

Table 16. Seasonal companson of the diet of radio-tagged spotted owls in the H. J. Andrews study area. (May 1975-Jim1976). Number of prey in each sample is m parentheses.

Percent of total preyMay(82)

ful•(58) Ser281 )ec

Feb(16)

Mar(73)

Apr(104)

MammalsN. flying squirrel 39.0 26.7 57.2 71.7 67.0 44.2Red tree vole 12.2 12.5 7.1 15.2 13.7 15.4W. red-backed vole 9.8 7.1 8.7 11.0 16.3Deer mouse 12.2 17.8 3.6 5.8Snowshoe hare 7.1 0.9Townsend's chipmunk 1.2 7.1 1.4 0.9W. pocket gopher 2.4 5.4Bush-tailed woodrat 4.9 1.8 7.1 4.0Shrews 4.9 3.6 1.4 0.9Other mammals' 4.9 5.4 4.4 2.7 3.6Unidentified mammals 2.4 5.4 17.9 2.2 1.4 4.0

Birds 1.2 3.6 2.2 1.4 4.0Insects and arachnids 4.9 3.6

• Some months were combined because of small sample size. Also. data from the same months in different yens were combined. No data weavailable for August.

6 Included chicken.. creeping vole. pike. shrew -mole, coast mole. Pacific *taping mouse. water vole. Scientific names of all prey me OVIMAppendix 3.

attempts by spotted owls were observed,including 7 attempts (1 successful) to cap-ture squirrels or birds in trees, and 2 at-tacks (1 successful) on mice or squirrels onthe ground. Whether the intended preywas on the ground or in a tree the usualmethod of attack was to dive upon theprey from an elevated perch. When at-tacks on squirrels in trees were unsuccess-ful, owls often hopped or flew from limbto limb after the fleeing animal. On 1 oc-casion, when 2 chipmunks retreated intoa woodpecker cavity in a rotten snag, afemale spotted owl flew to the cavity andclung to the lip of the cavity with her feetwhile attempting to reach inside with herbeak. As she did so, the chipmunks es-caped through a hole on the opposite sideof the snag.

Insects were captured either on theground or on limbs in trees. The usualmethod of attack on insects was to pounceon them with the feet or to land besidethem and pick them up with the beak. Wesaw no evidence of aerial pursuit of in-sects.

Upon capture, vertebrate prey werekilled by seizing and immobilizing themwith the feet and then tearing at the base

of the cranium with the beak, crushingthe posterior cranium or breaking theneck, or both. The heads of vertebrate preywere usually eaten first, a typical behaviorin hawks and owls (Mumford and Zusi1958, Collins 1963, Mikkola 1970, Wing1972, Balgooyen 1976). Small prey weresometimes swallowed whole, but animalslarger than deer mice were usually at leastpartially dismembered. Adult spotted owlsusually removed and discarded the stom-achs of mammals larger than deer mice.Tails of squirrels and woodrats were alsodiscarded, presumably because they hadlittle food value. Owlets were- consider-ably less selective with respect to what theyate, often swallowing squirrel and wood-rat tails.

Food Caching.-Spotted owls_regularlycached excess food, retrieving it later. Weobserved 24 instances of this behavior inwild owls and many instances in captiveowls. Prey remains were cached on top oflimbs in trees or on the ground beside logs,trees, or large rocks. Caching behavior wasobserved during all seasons, whenever owlscaught more than they could eat. On 1occasion when we tethered 5 live mice Infront of a roosting adult male, he killed

(1)MayISM

Mammals _. - ■-•N. flying squirrel 60.5 50.0

Red tree vole 4.7 12.5W. red-baW9d-vole 23.2Deer mouse 4.7 3.6Snowshoe hare or brush rabbit 7.1Bush-tailed or dusky-footed woodrat 2.3 8.9Mountain *averTownsend's chipmunk 3.6W. pocket gopher 1.8Other mammalsUnidentified mammals 2.3 5.4

Birds 2.3 5.3Insects' 1.8

Percent of total prey

1:51 (47)

12.8 14.810.6 8.24.3

17.1 3.34.3 3.3

4.3 1.6

4.3 1.61.6

10.6 3.331.9 62.3

42.38.9

4.420.1

4.48.92.22.24.4

2.2

kits I ED OWL IN ORECON-Forsman et al. 45

Table 17. Seasonal comparison of the diet of radio-tagged spotted owls on the BIM study area (Apr-Aug 1980). Number ofprey in each sample is in parentheses.

Included ermine. chickaree. big brown bat. shrew,. Scientific names of all prey an given in Appendix 3.b Errata spteelseus comprired 93% of all Insects taken.

and ate the first 2, then killed the othersand cached them on a nearby limb. Thiswas the only instance in which small preywere cached and in which stored preywere not decapitated and partially eaten.Other prey cached by free ranging owlsincluded 13 flying squirrels, 4 woodrats, 5snowshoe hares, 1 chipmunk, and 1screech owl. When caching prey, owlspushed and pulled the prey into positionwith the beak, wedging it securely inplace.

A Comparison of Prey Selection byMales and Females.-In most hawks andowls, including the spotted owl, males aresmaller than females, a condition referredto as reversed sexual size dimorphism. Apopular theory concerning the evolutionof reversed sexual size dimorphism is thatdivergence in body size permits males andfemales to partition the prey resource, col-lectively feeding upon a wider range ofprey (Selander 1966, Storer 1966, Reyn-olds 1972, Snyder and Wiley 1976). The-oretically, the smaller male should feed onsmaller prey than the female.

To compare prey selection by male andfemale spotted owls, we used data fromthe HJA and-BLM study areas where wewere able to mark roost trees occupied by

individual owls of known sex and then re-turn to the same roosts the next day tocollect pellets that had been regurgitated.For x2 comparisons of dietary composi-tion, uncommon prey were lumped intogroups (Table 18). On the HJA area welumped all mammals listed below deermice in Table 18 into 1 group and birdsand insects into another. On the BLM areawe lumped deer mice and red-backedmice into 1 group and all mammals listedbelow woodrats in Table 18 into another.On both areas, there was no significantdifference in the composition of the dietof males and females (HJA area x2 ma 7.9,df 5, P > 0.05; BLM area x2 la 11.13,df = 7, P > 0.05) (Table 18) or the meanweight of prey taken by males and fe-males (Table 19)•. We concluded that malesand females did not partition the food re-source on the basis of size or species oneither study area.

VOcal Displays

Calls given by spotted owls included avariety of hooting and loud barking callsas well as clear whistling calls. The hoot-ing calls of females were consistentlyhigher pitched than those of males, mak-

46 WILDLIFE MONOGRAPHS

Tabie 18. Diet (• composition) of male and female spotted owls on the H. J. Andrews study area (May 1975-Jun 1976) andBLM study area (Apr-Aug 1980). Number of prey in each sample is in parentheses.

HJA study area BLM study areaMalesU2S)

Females(66)

Males(MI

Females(76)

MammalsN. flying squirrel. 56.0 54.5 28.5 27.7Red tree vole 11.2 12.5 11.9 11.8W. red-backed vole 4.0 9.1 3.6Deer mouse 8.8 9.1 3.6 2.6Snowshoe hare or brush rabbith 1.6 1.1 14.3 9.2Dusky-footed or bushy-tailed woodrat 1.6 1.1 2.4 14.5W. pocket gopher 3.2 1.3Chickaree 0.8 1.2Townsend's chipmunk 0.8 2.3 4.8Pika 2.4Shrew-mole 0.8Creeping vole 3.2Water vole 0.8Pacific jumping mouse 1. 1Sorel app. 1.2Long-tailed weasel 1.2Mountain beaver 5.3Big brown bat 1.3

Birds 1.6 4.6 8.3 3.9Insects 3.2 4.6 19.0 22.4

Scientific names are in Appendix &b Included mostly snowlhoe hares on the HJA study area and a mixture of snowshoe hares and brush rabbits on the BLM study amm AB rabbits.

harm and mountain Woven captured were juveniles.

ing sex identification by call relativelyeasy. With the exception of copulatorycalls, all types of calls were given by bothsexes. However, some call types were giv-en more frequently by 1 sex than the oth-er. In the descriptions that follow, calls aredivided into 13 different types. Some calls,particularly the hooting and barking calls,

were characterized by considerable vari-ation in cadence and number of notes. Thefollowing descriptions include only themore common patterns, but calls that wereparticularly variable will be so indicated.

Four-note Location Call: This is thehooting call most frequently mentioned inthe literature (Ligon 1926, Bent 1938). It

Table 19. Mean weights of prey captured by male and female spotted owls on the H. J. Andrews study area (May 1975-Jun1976) and BLM study area (Apr-Aug 19801.

Males FemalesMean MIon Signifies/me

of differenceStudy area and seasonal period Pay

werSE Pasty war SE between mean.

HJ A AreaSep-Feb (fall-winter) 18 108.4 12.2 52 86.8 6.9 P > 0.05Mar-Apr (spring) 58 99.7 5.6 21 84.5 9.1 P > 0.05May-Jul (summer) 52 75.6 12.2 20 74.6 39.0 P> 0.05Overall average 128 91.1 5.9 93 84.7 10.2 P> 0.05

BLM AreaApr-Aug (late spring-summer) 85 90.5 10.2 76 -114.7 12.9 r) 0.05

HUI.

•••■•••as Oita - OM*

I .* 4.1 04 44 *14 104ts to is

Allipeka.. •

-11, "Nib

.6■14P

to A as go

•••• I• • 44,

to do 35

I I

is Ongoninigaina

io 25

421

THE SPOTTED OWL IN OREGON—Forsman et al. 47

Plat

A A A

IF&. 4b. AL A. Ai. 41. ab is do

SECONDS

Fig. 23. Sonograms of hooting and barking vocalizations commonly given by spotted owls. (A) Male 4-Note Location CaN; (B)Female 4-Note Location Cali: (C) Series Location Call--unevenly spaced series: (0) Series Location Cal—evenly spaced series;(E) Bark Series: (F) Nest Cal.

most commonly had 4 notes in a series: asingle introductory hoot followed by ashort pause, 2 closely spaced hoots, anotherpause, and then a final hooting note trail-ing off at the end (Fig. 23A,B). A phoneticrendition was hoo - - - hoo-hoo - - - - h000.The introductory note was sometimesomitted. Ligon (1926:423) described thiscall as coo - - - coo-coo - - - - coo-o-o. Thiscall was given by both sexes in a varietyof situations and seemed to function as ageneral location call between paired birdsand as a territorial challenge. It was givencommonly when paired birds called backand forth on their territories and duringterritorial disputes. Males often gave a verysubdued version of this call when they ar-

rived near the nest with food. During ter-ritorial disputes or before and after cop-ulation, the 4-Note Location Call wasfrequently alternated with other types ofcalls described below. The 4-Note Loca-tion Call was the principal call used forlocating and censusing spotted owls. Un-der good conditions this call could beheard over 1 km away.

Agitated Location Call: This call wasgiven in situations of. apparent sexual oraggressive excitement. It usually began likethe 4-Note Location Call but was moreintense and ended in a loud ow! note. Aphonetic rendition of the most commonversion was Iwo - - - hoo-hoo - - - ow!Frequently, this call was given several

48

WILDLIFE MONOGRAPHS•

times in a rapid series, with the introduc-tory note left out after the first 4-notephrase. Both sexes gave this call duringterritorial disputes and males sometimesgave it immediately before, after, or dur-ing copulation. Females did not usuallygive this call in association with copula-tion.

Series Location Calls: This category in-cluded a variety of calls characterized by7-15 hooting notes in series. The arrange-ment of notes was extremely variable bothwithin and among individuals. A commonpattern was an introductory series of 5-7evenly spaced notes, followed by severalnotes or pairs of notes at longer intervals(Fig. 23C). Sometimes all notes in a serieswere evenly spaced (Fig. 23D). These callswere commonly given during territorialdisputes. Series Location Calls were alsogiven occasionally by paired individualsas they called back and forth within theirterritories.

Bark Series: This call was a rapid seriesof 3-7 loud barking notes given at the rateof 2-3 notes/second (Fig. 23E). We pho-neticized it as owl-owl-owl-owl-ow!, orsometimes yenk!-yenkl-yenkl-yenkl. Thiscall was given by females (less frequentlyby males) during territorial disputes. It alsoappeared to function as a long-distancecontact call between paired birds.

Nest Call: This call was given by bothsexes when they called from the nest treeduring prenesting displays. It was heardinfrequently in other situations. The callwas a steady series of evenly-spaced, sub-dued hoots, sometimes given for severalminutes or more with only minor breaks(Fig. 23F). The individual notes were giv-en at the rate of about 3/second. Malesusually gave this call while perched onlimbs just outside their nests whereas fe-males usually gave it from within the nest.

Contact Call: This call was a hollowwhistled note ending in an upward inflec-tion (Fig. 24A). We phoneticized it asc000-weep!. Females gave this call whenthey were roosting near their mates oryoung during the day and when the maledelivered food to the nest area. Each eve-ning during the 2-3 weeks before eggs

were laid, nesting females usually gave theContact Call repeatedly (at 15-45-secondintervals) from their day roosts until themale flew to them and copulated. In allof the above situations, the call apparentlyserved to inform tjie male or young of thefemale's location, thereby facilitating foodexchanges, copuation, and general coor-dination of activities. Males sometimesgave the Contact Call when they arrivednear the nest with food. The Contact Callmight also have been termed the "AdultBegging Call" because in the course ofdevelopment it evolved directly from thejuvenile Begging Call (see below). Be-cause the adult version played a moregeneral role in communication than sim-ply "begging," we chose the more generalterm of Contact Call. Southern (1970) usedthe term "Contact Call" in reference to avery similar call given by the tawny owl.

Agitated Contact Call: The AgitatedContact Call was a high intensity versionof the Contact Call. Instead of the mellowwhistled quality of the Contact Call, thecoon-weep! notes were given loudly andshrilly, often with a raspy or grating qual-ity (Fig. 24B). This call was given by fe-males during territorial disputes and, in-frequently, in other situations. Males gavethis call infrequently.

Wraaak► Call: This call consisted of aloud, grating, 2-syllabled note that wassometimes given during territorial inter-actions. We phoneticized it as toraaald ormask! (Fig. 24C). On 1 occasion during aterritorial interaction, a female began giv-ing this call at 2-5-second intervals, thengradually changed to the Agitated Con-tact Call as the interaction proceeded.

Chatter: This call consisted of a. rapidseries of low chittering notes (Fig. 24D).It was given by adults and owlets whenthey were handled, apparently indicatingfear or irritation. It was also given by fe-males during copulation and occasionallyby males or females when they were en-gaged in allopreening (Foramen andWight 1979). Clark (1975) described asimilar call in the short-eared owl.

Female Copulatory Call: This call wasslightly variable but the basic pattern con-

THE SPOTTED OWL IN OREGON —Forsman et al. 49

Fig. 24. Sonograms of spotted owl vocakzations charactertzed by whistling, chittering, grating, or sibilwit sounds. (A) ContactCap: (B) Agitated Contact Call: (C) Wraaakl (D) Chltter: (E) Cooing Caps; (F) Juvenile Bagging Cal

sisted of a rapid series of chittering notesas the female was mounted that gradedinto a high pitched, wheezy, prolongedwhistled note as copulation occurred. Thecall ceased as the male dismounted.

Male Copulatory Call: Immediatelybefore copulating, males usually flew to-wards the female giving a • mixture of4-Note Location Calls and Agitated Lo-cation Calls. After they mounted, how-ever, some males switched to a series ofemphatic single hoots during copulation.We phoneticized the latter call as hool-hool-hool-hoolhool-hool. This call wasusually followed by Agitated LocationCalls as the male dismounted and flewaway.

Alarm CitlirThis call was a terse, single-

syllabled warning given in response to ap-proaching predators. It was given veryquietly and was audible only at closerange. This call defied phonetic represen-tation but resembled a nasal grunt orgroan. We heard the Alarm Call given inresponse to the presence of red-tailedhawks, Cooper's hawks, and goshawks. Inaddition, I of our captive owls frequentlygave the Alarm Call when dogs or unfa-miliar humans walked by her cage.

Cooing Calls: This category included avariety of subdued' whistles and cooingsounds given at close range between adults(Fig. 24E). These calls were given in manycontexts, most commonly when pairedadults were roosting together or allo-preening.

APR MAYJUN JUL AUG SEP OCT NOV DEC JAN FEB MAR

011•••0500

0400 • ••0 • oo

0 •••• ••000

00300

2100

2000

1900

1000

$700

1600Ho

001"

ono

0600

50

WILDLIFE MONOGRAPHS

Fig. 25. Activity periods of spotted owls on the H. J. Andrews and BLM study areas. Stars indicate when activity began withrespect to sunset. and circles indicate when activity ceased with respect to sunrise. Solid and open symbols represent the H.J. Andrews and BLM areas. respectively. Each point represents 1 observation.

Juvenile Begging Call: Owlets under 6months of age gave only 3 different calls,2 of which were uncommon. The uncom-mon calls included a low chittering sound(similar to the adult Chitter) uttered whenowlets were handled or otherwise dis-tressed, and a barely audible gruntingsound (similar to the adult Warning Call)that was given when owlets were startled.The only call that was commonly givenby all owlets was the Begging Call, araspy, high-pitched whistle that was givenpersistently when they were hungry (Fig.24F). This call was similar in duration andinflection to the adult Contact Call (Fig.24A) but was much higher pitched andentirely lacked the clear whistled qualityof the adult call. We phoneticized it assweeeeet!. Owlets were easily induced tobeg if the call of an adult was imitated

near them. The begging call was not loud,carrying <200 m under most conditions.

One of the owlets raised in captivitygave only typical Begging Calls until 27September, when it was approdmately 6months old. After that date a gradualchange in the quality of the Begging Callwas noted. The raspy quality of the callbegan to diminish gradually, and the callbegan to assume the clear whistled qualityof the adult Contact Call. By T Octoberthe transition was complete. Thereafter,the owlet gave the typical adult ContactCall with only an occasional hint of raspi-ness. This same owlet was first heard giv-ing the typical 4-Note Location Call on 7February and the Bark Series on 13 Feb-ruary. Both of the latter calls were givenclearly but at reduced volume.

In the wild the earliest date that we

THE SPOTTED OWL IN OREGON--F0737ritiri et al. 51

Table 20. Average horizontal distance moved (nthour) by wooed owls whNe foraging under different weather conditions andduring different seasons on the H. J. Andrews study area (May 1975-Jun 1976).•

Weather commonsPeriod

MausMar-Apr May-Jun Jul-Sep Oct-Feb

Clear. no_metpitation 203 532 283 197 304Overeast7neprecipitation 530 316 192 329Intermittent rain or snow 286 297 312 231 244Steady iiffrek-Inow 178 302 345 194 242Rain or snow and strong winds

(>32 knithour) 32Means - 209 488 293 206 247

• Data compiled from 8 adults.

heard a young of the year attempt to giveadult vocalizations was 18 October whena radio-tagged juvenile gave the 4-NoteLocation Call in response to our imitationof adult calls. The quality of this juvenile'scall was very poor. Although the cadencewas normal, the volume was so low andthe notes were so raspy that the call couldbarely be heard 75 m away. These obser-vations indicate that juvenile spotted owlsdevelop most of the adult vocal repertoireduring fall and winter of their first year.

Activity Patterns

Timing of Activity.—On the averageon the HJA study area, owls left their dayroosts to begin foraging 14 minutes aftersunset (range a. —54 to +60 min) andstopped foraging 21 minutes before sun-rise (range = —45 to +6 min) (Fig. 25).Activity generally began earlier (in rela-tion to sunset) during spring and earlysummer than during fall and winter. Ac-tivity periods on the BLM study area werepractically identical (Fig. 25).

During the day the owls spent most oftheir time roosting in trees, occasionallymoving in place, but generally remaininginactive. On 9 occasions, however, we sawspotted owls attempt to capture diurnal orcrepuscular mammals that wandered intoroost areas (2 attempts successful). We alsofound that spotted owls Would readilyleave their day roosts to capture live micethat we tethered on the ground beneathroost trees:- They also made occasionalshort flights during the day to retrieve

cached prey, change roost trees, or drinkor bathe at streams near their roosts. Therelative infrequency of diurnal activitywas indicated by the low proportion ofdiurnal animals in the diet (4.5% on HJAarea, 7.4% on BLM area).

Rate of Movement While Foraging.—Spotted owls foraged at night by movingfrom perch to perch, watching and listen-ing for prey. Time spent at individualperches ranged from only a few secondsto several hours. The average rate ofmovement (in terms of horizontal dis-tance) while foraging in this manner was247 m/hour on the HJA study area and323 m/hour on the BLM study area(range = 0-3,371 m/hour). Our estimatesof distance moved were undoubtedly con-servative, because it was impossible to de-tect all of the short flights that radio-taggedowls made while foraging.

On the HJA area, the rate of movementwhile foraging was highest during springand summer and lowest during fa andwinter (Table 20). This difference wasmost. pronounced in nesting individuals,

wasas characteristic of non-nesting in-dividuals as well. The increased rate ofmovement of nesting individuals duringspring and summer was undoubtedly re-lated to increased energy demands. Re-gardless of their nesting status, however,spotted owls had less time to forage dur-ing the short spring and summer nightsthan during winter and may have com-pensated by spending less time resting atnight during spring and summer.

On the average, owls moved less when

N 0 .) F M A IA157$

52

WILDLIFE MONOGRAPHS

Fig. 26. Percent of days on which adutt spotted owls on theH. J. Andrews study area roosted together as pairs (dashedline) or near (within 400 m) their traditional nest sites (solidline). Percentages were calculated on a monthly basis fromMay 1975-Jun 1976. Data for August 1975 rare inadequateto calculate pair association.

it rained or snowed (Table 20). However,when the data were examined on a sea-sonal basis the only consistent relationshipbetween weather conditions and rate ofmovement was that movement ceased al-most altogether when heavy rains wereaccompanied by strong winds (Table 20).It appeared, therefore, that seasonalweather patterns were responsible for lit-tle if any of the seasonal differences inrate of movement.

Nightly patterns of movement were un-predictable. For instance, at any time ofthe night it was common for individualsto stop moving and stay at the same lo-cation for up to several hours. Thesequiescent periods did not seem to fit anyrecurring pattern, probably because 2 ofthe major variables affecting movement atnight (prey encounters and prey capture)did not occur on a predictable schedule.When owls captured large prey (e.g.,squirrels, woodrats, or hares), they com-monly cached what they could not eat andthen roosted in the vicinity until the preywas consumed. This response to prey cap-ture was verified on a number of occasionswhen owls stopped moving during thenight and were found roosting the nextmorning near a cached, partially eatenprey animal.

Social Behavior

During midwinter (Nov-Jan) when theowls were largely solitary, they called in-frequently, and only 1 territorial interac-tion was observed. After the resident maleand female otreacirterritory began to roosttogether in late February or March (Fig.26), they became more vocal and beganto interact more frequently with neigh-boring pairs. The increased level of socialactivity continued through summer, andthen declined in October.

During 353 nights of observation on theHJA and BLM study areas, we witnessedonly 17 territorial interactions (defined asany instance in which at least 2 individ-uals from different areas called withinhearing distance of each other). The rel-ative infrequency of interactions preclud-ed the identification of distinct territorialboundaries. In fact, we concluded that.such boundaries probably did not exist.The owls appeared to space themselves far.enough apart to avoid frequent contactwith each other. Instead of distinctboundaries between territories, there werebroad areas of home range overlap thatwere contested occasionally by the adja-cent parties, but defended consistently bynone.

In the few instances in which residentindividuals detected intruders, they usu-ally responded with the 4-Note LocationCall or with a mixture of hooting, barand whistling calls. The nature of thesequent interaction was dependent uponthe sex of the individuals involved and thelocation of each individual relative to itshome range area. In general, if both in-dividuals were of the same sex ancLif onewas well inside the home range of the oth-er, the intruder would stop calling and re-treat. If the interaction occurred in an areaof home range overlap and both birds wereof the same sex, they sometimes ap-proached each other and then pursuedeach other from tree to tree, calling in anagitated manner. Eventually 1. or bothbirds would leave the area. One interac-tion of this type between 2 males lastedover 2 hours. Territorial interactions be-

THE SPOTTED OWL IN OREGON —Forsman et al. 53

hymn birds of the same sex did not alwaysescalate into encounters at close range; onat least 1 occasion, the 3A pair and the IAfemale called from opposite sides of BlueRiver Reservoir witheut ever approachingcloser than 300 m to each other.

Resident .1:individuals did not appearnearly as aggressive in rebuffing membersof the oppsajtelex encountered within thehome range. Five interactions were ob-served involving a male and female fromadjacent home ranges. In all 5 cases theowls called excitedly but did not retreatfrom each other, even • when 1 individualhad wandered well inside the home rangeof the other. Another indicator that malesand females did not behave aggressivelytowards each other was that 3 radio-taggedowls (2 females, 1 male) occasionallyroosted with individuals of the oppositesex other than their regular mates.

Because interactions between male andfemale spotted owls from adjacent homeranges did not involve the aggressive chas-ing and retreating behavior characteristicof interactions between individuals of thesame sex it was not clear whether male-female interactions should be called "ter-ritorial" or not. It was also not clearwhether spotted owls formed a perma-nent • bond extending beyond a singlebr g

•

season or whether mate constan-cy from year to year occurred becausemales excluded (as much as possible) othermales from the home range and femalesexcluded other females. This pattern ofhome range defense, combined with thetendency of residents to remain withintraditional home ranges, made it virtuallycertain that the same individuals re-pairedeach spring, as long as each individual wasable to exclude others of its sex from thehome range. Mate constancy, therefore,may have been more a function of theattachment to a traditional home rangethan attachment to a particular mate.

DISCUSSION

The spotted owl population in Oregonis undoubtedly. larger than indicated byour sample bicause we did not inventory

all areas of suitable habitat in the state.We estimate that only 50-60% of the po-tential spotted owl habitat was searchedfor owls. If we assume that populationdensities were similar in the searched andunsearched areas, then the statewide pop-ulation was roughly 1,000-1,200 pairs ofowls.

Studies of spotted owls in California andWashington suggest that populations aredeclining there also. Gould (1974, 1977)checked numerous historical sites wherespotted owls had been reported in Cali-fornia and was able to relocate owls atonly 72% of the sites. Because many ofthe historical sightings were poorly docu-mented, however, Gould considered hisresults as being only suggestive of a de-cline, rather than an accurate indicator oftotal decline. Data on population trendsin Washington are limited, but Postovit(1979) found few owls in second-growthforests in western Washington. The pau-city of spotted owls on private lands in ourstudy was also noted in California andWashington (Gould 1974, 1977; Postovit1979). The majority of owls found byGould and Postovit were in old-growthforests or mixed forests of old-growth andmature trees.

Most home ranges of spotted owls onthe HJA and BLM study areas were largerthan home ranges reported for barred andgreat horned owls (Nicholls 1970, Nichollsand Warner 1972, Fuller 1979). These re-suits do not support the theory that raptorbody size and home range size are posi-tively correlated (Schoener 1968). We sus-pect that the relationship between raptorbody size and home range size is not con-sistent because it is confounded by factorssuch as habitat quality and regional andlocal, differences in the types of prey uti-lized.

Why spotted owls in Oregon consis-tently preferred old-growth forests forforaging was unknown. It is • thatthe principal prey utilized . the owls(flying squirrels) were most numerous inolder stands. No data were available onthe abundance of flying squirrels on thestudy area, but studies by Gashwiler (1959)

•

54 WILDLIFE MONOGRAPHS

and Hooven (1972) indicated that flyingsquirrels were eliminated from recentlyclear-cut areas. Flying squirrels probablydo not begin to repopulate clear-cut areasuntil at least 20-30 years after harvest,when trees become large enough to sup-port arboreal nests.

It was also possible that the owls for-aged selectively in older forests becausethe biomass of nocturnal prey within thesize range preferred by the owls (20-300g) was greater in older forests than in re-cent clear-cuts and young second-growth.Although populations of some nocturnalmammals (notably deer mice and somevoles) increase after clear-cutting (Gash-wiler 1959, Hooven 1972), it is not clearif such increases compensate for the re-duction in biomass associated with theelimination or reduction of populations offlying squirrels, red tree voles; western red-backed voles, and snowshoe hares.

Recent clear-cuts and young second-growth forests also may have been avoid-ed because foraging perches were absentthere. In addition, the dense brush andherb layers that developed on many clear-cut areas may have made it difficult forowls to attack prey on the ground. For-aging in open areas also could have in-creased the risk of predation byerpredators such as the great horned ow!.

Because as much as 77% of the forestswithin the home ranges of some individ-uals had been clear-cut within the pre-vious 40 years, it was obvious that the owlswere able to tolerate considerable reduc-tions in the amount of older forests withintheir home ranges. There were indica-tions, however, that they were able to doso only by increasing the size of their homerange areas, thereby encompassing addi-tional areas of older forest. This was sug-gested by the fact that owls in more heavi-ly cutover areas (BLM study area) hadlarger home ranges on the average thanowls in less heavily cutover areas (HJAstudy area) (Table 1). Similar relationshipsbetween home range size (or territory size)and habitat quality have been reported inother birds (Stenger 1958, Seastedt andMacLean 1979).

We believe that older forests were pre-ferred for roosting because they providedthe most protective environment forroosting under most weather conditions.Old-growth forests had deep, multilay-ered canopies that' produced numerouspatches of shade during warm weatherand also had numerous large trees suitablefor roosting in cold, wet weather. By com-parison, young stands had relatively shal-low canopies (measured from top to bot-tom) and few large trees of the typepreferred for wet weather roosts.

One factor that undoubtedly causedspotted owls to select old-growth forestsfor nesting was that suitable nests weremost common in older forests. Large cav-ities and heavy dwarf mistletoe infesta-tions were generally uncommon or absentin young forests, but were common in oldforests. The fact that the owls showed a ,strong preference for roosting in old-growth stands also suggested that suchstands provided an optimal environmentin which to raise the young after they leftthe nest.

In contrast to spotted owls in Oregon,which nest almost exclusively in trees,spotted owls in the mountains and can-yonlands of southern California and thesouthwestern U.S. frequently nest in cavesor potholes in cliffs (Peyton 1910, Dickey1914, Ligon 1926). These regional differ-ences in nest types are undoubtedly dueto differences in the types of habitat oc-cupied. There are few cliffs or caves'suit-able for nesting in the humid forests ofOregon and Washington, whereas cliffsand caves are abundant in the mountainsand canyonlands of southern California,Arizona, and New Mexico.

The tendency of spotted owls to use atraditional nest area or nest tree for manyyears has been reported by other observ-ers. Miller (1974) observed iLpair of spot-ted owls that remained in the same areafor at least 7 years, nesting in 5 years. Pey-ton (1910) located a nest in 1908 that wasstill being used by spotted owls 21 yearslater in 1929 (Bent 1938).

Timing of nesting in Oregon did notdiffer appreciably . from dates given by

TIIE SPOTTED OWL IN OREGON---Forstrian et al. 55

Bent (1938) for southern California. Helisted 15 records of eggs observed or col-lected between 1 March and 10 May.Eight of the records fell within the period27 March-1 Ap:P., "ir.e.!cating the heightof the season" (Bent 1938:207).

The disphy - sequence in which malespotted owls called from a perch near thenest just ire or after copulation is com-mon in owls (Haverschmidt 1946, Jansson1964, Konig 1965, Ligon 1968). Ligon(1968) found that male elf owls called frominside the nest cavity and that females fol-lowed them inside. He interpreted this be-havior as an attempt by the male to enticethe female into the nest.

Except for females that were about tolay eggs, spotted owls were never ob-served roosting in cavities. Ligon (1968)found that female elf owls began to roostin their nest cavities 1-2 weeks before lay-ing eggs. He suggested that this behaviormay have been adaptive because it in-sured that other birds or mammals did notusurp the nest and because it precondi-tioned males to feed females at the nest.

Apparently, most young strigids leavethe nest before they are fully capable offlight (Dement'ev et a1. 1966, Hoglund andLansgren 1968, Southern 1970, Dunstanand Sample 1972). Dunstan and Sample(1972) suggested that this might occur be-cause nestlings were unable to exercisetheir wings within cramped nest cavities.The fact that young great gray owls andspotted owls (which frequently occupyopen platforms) also leave their nests be-fore they can fly well seems to contradictthis explanation. Hoglund and Lansgren(1968) suggested that young great grayowls left the nest before they could flybecause they became increasingly uncom-fortable as the season advanced and thesun began to shine more directly into nests.This explanation hardly seemed to applyto spotted owls because they were usuallyprotected from direct sunlight by nestcavities or dense foliage.

A more likely reason for leaving the nestas early as possible is to escape increasingparasite numbers in the nest. As pellets,fecal mattePr and prey remains accumu-

lated in nests, and as warm temperaturesbegan to predominate in late May andJune, spotted owl nests began to attractincreasing numbers of flies. Three nest-lings in 2 nests during this period werebitten repeatedly around the eyes bybloodsucking flies (family Hipoboscidae),causing enough hemorrhaging to glue theeyelids shut with dried blood and debris.Similarly, Fitch et al. (1946) and Fitch(1947) reported that Eusim/tum clarum,a bloodsucking fly, caused heavy mortal-ity among nestling red-tailed hawks andcaused scabs to form on the eyelids ofnestling great horned owls. Bohm (1978)and Tirrell (1978) indicated that parasitesmay cause a significant reduction in nest-ling vigor and growth rate in some rap-tors. If the reduction in fitness associatedwith parasitism is greater than that asso-ciated with leaving the nest at an earlyage, then selection should favor leavingthe nest as early as possible.

The climbing ability of young spottedowls is typical of many species of owls,including great horned, barred, screech,great gray, tawny, and hawk owls (Smith1963, Hoglund and Lansgren 1968, Dun-stan and Sample 1972, Scherzinger 1980,R. Palmer, pers. commun.). The tendencyof young spotted owls to leave the nestbefore they can fly well and to climb intoelevated perches has led some authors(e.g., Dickey 1914, Karalus and Eckert1974) to mistakenly suggest that adultspotted owls physically transport theiryoung from the nest.

Our observations on 2 dispersing owletswere too limited to determine much aboutdispersal. The relatively short distancestraveled by both owlets during the first 2months of independence did not appearunusual in light of the relatively short dis-persal distances that have been docu-mented for other sedentary owl species(Southern 1970, Houston 1978).

The mixture of arboreal and terrestrialspecies in the diet of spotted owls in Or-egon indicated that the owls foraged at alllevels in the forest, from the ground to theupper canopy. The preponderance offlying squirrels and woodrats in the diet

56

WILDLIFE MONOGRAPHS

suggested either that the owls searched se-lectively for squirrel or rat-sized mam-mals or were most adept at capturing suchmammals, or both. The preponderance offorest-dwelling mammals in the diet re-flected the strong preference of the owlsfor foraging in forest areas as opposed toclear-cuts and other large openings. Smallamounts of information on the diet of thespotted owl in other parts of its range in-dicate that spotted owls depend on wood-rats or northern flying squirrels for thebulk of their food in most areas, but alsoforage opportunistically on a variety ofsmall mammals, birds, and insects (Dick-ey 1914; Ligon 1926; Marshall 1942,1957;Smith 1963; Beebe and Schonewald 1977;Kertell 1978; Barrows 1980).

The absence of significant differencesin prey selection by male and female spot-ted owls was not surprising consideringthat female spotted owls are only about10% larger than males (Earhart and John-son 1970). Nevertheless, if owls that areonly slightly dimorphic do not partitiontheir prey on the basis of size, then theremust be some more fundamental reasonfor the evolution of reversed sexual sizedimorphism.

The prey caching behavior of spottedowls is typical of many owls, hawks, andfalcons (Pitelka et al. 1955, Mumford andZusi 1958, Schnell 1958, Collins 1963,Mueller 1974, Balgooyen 1976, Collopy1977, Phelan 1977). As discussed by Bal-gooyen (1976), Callopy (1977), and Phe-lan (1977) prey caching is adaptive forseveral reasons, the most obvious beingthat by storing and returning to excessprey, a predator reduces the amount ofenergy it would expend if it did not com-pletely use all prey captured.

Radiotelemetry studies of other suppos-edly nocturnal owls have indicated activ-ity patterns both similar to and noticeablydifferent than observed in this study. Theactivity cycle of a saw-whet owl studiedby Forbes and Warner (1974) was vir-tually identical to that of spotted owls onthe HJA and BLM study areas. Greathorned owls studied by Fuller (1979) usu-ally began activity within 30 minutes of

sunset and ended activity within 30 min-utes of sunrise. Barred owls studied byNicholls (1970) and Fuller (1979) weremost active during the night but were alsofrequently active during the day; diurnalactivity increased eansiderably when theowls were feeding young. Like spottedowls, great horned and barred owls wereactive for longer periods during winterthan summer (Fuller 1979).

In contrast to our findings, Miller (1974:115) reported that a female spotted owl"foraged at all hours of the day or night"when feeding young. However, Miller didnot differentiate between opportunisticattempts to capture prey in roost areas andlong distance foraging flights away fromroost areas. If the foraging behavior Millerreferred to was of the former type, then-differences between her observations andours may have been more semantic thanreal.

MANAGEMENT IMPLICATIONS

The most serious threat to the spottedowl in Oregon is the gradual eliminationof its preferred habitat (old-growth andmature forests). To alleviate this threat werecommend that a system of old-growthmanagement areas be established on fed-eral forest lands throughout the range ofthe spotted owl in Oregon. When possible,such areas should be selected on the basisof occupancy by spotted owls. Old hmanagement areas should be s at in-tervals of 2-10 km so that pairs of owls orsegments of the population do not becomereproductively isolated. To achieve aspacing of 2-10 km between old-growthmanagement areas throughout the rangeof the spotted owl in Oregon would re-quire that approximately 400 areas be es-tablished, assuming that the areas are ir-regularly spaced at 2-10-km Intervals.

Within each old-growth managementarea, 400 ha of old-growth forest shouldbe maintained in a relatively naturalunlogged) condition within a.2:5-km. ra-dius of the nest of the resident pair ofowls. The 400 ha figure is based on thefact that all of the radio-tagged pairs we

Thz Srorm) OWL IN OREGON—Forsman et a/. 57

studied had more than 396 ha of old-growth within their home ranges (thehome range of a pair included the com-bined home ranges of the male and fe-male). We 1-.:Izvc th.lt of old-growth will provide adequate habitat forindividual: pairs, assuming the most ex-treme scenario hi which all surroundingareas are-arsertd by intensively managedsecond-growth forests <80 years old (alikely situation in the future). It is possiblethat less extensive areas of old-growth maybe needed if extensive areas of matureforest or older (>80 years old) second-growth forest are available as alternative(albeit less optimal) foraging habitat.However, until this has been documented,adherence to the 400-ha old-growth min-imum should be maintained.

Old-growth forests within each man-agement area should include some patchesthat are at least 40-100 ha in area andshould be connected by travel corridors ofmature or old-growth forest if possible.Old-growth habitat in management areasshould not be harvested until replacementold-growth stands are regenerated on ad-jacent areas (this includes salvage loggingof logs and dead or damaged trees). Theabove recommendations closely parallelrecommendations made by the Oregon-Washington Interagency Wildlife Com-mittee, a group of biologists representingthe U.S. Forest Service, Bureau of LandManagement, U.S. Fish and Wildlife Ser-vice, Oregon Department of Fish andWildlife, and Washington Department ofGame (Oregon-Washington InteragencyWildlife Committee 1980).

Assuming that a management plan sim-ilar to the above is implemented, thereremains the problem of how to managepairs outside the old-growth managementareas. We assume that most of these pairswill disappear in the long term, as theirhabitat is eliminated. To slow the rate atwhich such pairs are eliminated, however,we recommend that harvest of known nestsites or principal roost areas be avoided.We also recommend that forest managersenhance the -suitability of selectivelylogged stands by leaving 2-5-ha patches

of old-growth trees scattered within logged.stands.

The most important factor to considerin the establishment of a spotted owl man-agement plan in Oregon is the rate atwhich the remaining old-growth forestsare being harvested. In the Coast Rangesin particular, most old-growth forests havealready been cut, and only isolated islandsof old-grOwth will remain by the year 2000(Beuter et al. 1976). As a result, optionsfor management are rapidly being lost,particularly options for maintaining thedesired distribution of old-growth man-agement areas. It is imperative, therefore,that management be initiated as soon aspossible. If future studies demonstrate thatthe proposed management plan is unnec-essarily conservative (i.e., retains more old-growth than is necessary), then the excessold-growth can always be harvested atsome future date. In contrast, it will bevirtually impossible to adjust manage-ment in the opposite direction if the de-cision to manage the owl is delayed, andit eventually becomes apparent that notenough old-growth has been retained.Once harvested, it is extremely unlikelythat old-growth forests will ever be regen-erated.

LITERATURE CITED

BALCOOYEN, T. C. 1976. Behavior and ecology ofthe American kestrel (Falco sparoarius L) in theSierra Nevada of California. Univ. Calif.. Berke-ley, Publ. Zool. 103:1-83.

BARROWS, C. 1980. Summer roost selection byspotted owls: an adaptation to heat stress. M.S.Thesis. California State Univ., Long Beach. 51pp.

, AND K. BARROWS. 1978. Roost character-istics and behavioral thermoregulation in thespotted owl. West. Birds 9:1-8.

BEEBE, G. D., AND J. SCHONEWALD. 1977. Spottedowls near Palomarin. Point Reyes Bird Obser-vatory Newsl., Point Reyes. Calif. 43:6-7.

BENDIRE, C. E. 1882. The. spotted owl. Ornithol.and Oologist 7:99.

1892. Life histories of North Americanbirds. U.S. Natl. Mrs. Special Bull. 1.446pp.

BENT, A. C. 1938. Life histories of North Americanbirds of prey. Put 2 Smithson. Inst.. U.S. Natl.Mus. Bull. 170.482pp.

BEUTER, J. H.. K. N. JOHNSON, AND H. L SCHEUR-MAN. 1976. Timber for Oregon's tomorrow: an

58 WILDLIFE MONOGRAPHS

analysis of reasonably possible occurrences. Oreg.State Univ. Sch. For. Res. Bull. 19. 11Ipp.

Bocx, C. E., AND ). F. LYNCH. 1970. Breeding birdpopulations of burned and unburned conifer for-est in the Sierra Nevada. Condor 72:182-189.

BOHM, R. T. 1978. Protocalltphora infestations ingreat horned owls. Wilson Bull. 90:297.

CATLING, P. M. 1972. Food and pellet analysisstudies of the saw-whet owl (Aegolius acadicus).Ont. Field Biol. 26:1-15.

CLARK, R. J. 1975. A field study of the short-earedowl, (Aso flatnmens) (Pontoppidan). in NorthAmerica. WW1. Monogr. 47. 67pp.

COLUNS, C. T. 1963. Notes on the feeding behav-ior, metabolism, and weight of the saw-whet owLCondor 65:528-530.

COLLOPY, M. W. 1977. Food-caching by femaleAmerican kestrels in winter. Condor 79:63-68.

DEMENT 'EV, G. P., N. A. GLADKOV, E. S. Prostmixo,E. P. SPANGENBERG, AND A. M. SUDILOVSKAYA.1966. Birds of the Soviet Union. Vol. I. IsraelProg. Sci. Trani., Jerusalem. 704pp.

DICKEY. D. R. 1914. The nesting of the spottedowl. Condor 16:193-202.

DUNN, H. H. 1901. The spotted owl (Syrnium oc-cicientale). Oologist 18:165-167.

DUNSTAN. T. C. 1972. Radio-tagging falconiformand strigiform birds. Raptor Res. 6:93-102.

, AND S. D. SAMPLE. 1972. Biology of barredowls in Minnesota. Loon 44:111-115.

EARHART, C. M., AND N. K. JOHNSON. 1970. Sizedimorphism and food habits of North Americanowls. Condor 72:251-265.

FrrcH, H. S. 1947. Predation by owls in the Sierranfoothills of California. Condor 49:137-151.

F. SWENSON, AND D. F. TILLOTSON. 1946.Behavior and food habits of the red-tailed hawk.Condor 48:205-237.

FORBES. J. E., AND D. W. WARNER 1974. Behaviorof a radio-tagged saw-whet owl. Auk 91:783-795.

Fonn4AN, E. D. 1976. A preliminary investigationof the spotted owl in Oregon. M.S. Thesis. Ore-gon State Univ., Corvallis. 127pp.

1980. Habitat utilization by spotted owlsin the west-central Cascades of Oregon. Ph.D.Thesis. Oregon State Univ., Corvallis. 95pp.

1981a. Habitat utilization by spotted owlson the Eugene District of the Bureau of LandManagement. Unpubl. Tech. Rep., U.S. Dep. In-ter. Bur. Land Manage., Portland. Oreg. 63pp.

1981b. Molt of the spotted owl. Auk 98:735-742.

, E. C. MESLOw, AND M. J. STAUB. 1977.Spotted owl abundance in young versus old for-ests, Oregon. Wildl. Soc. Bull. 5:43-47.

AND H. M. WIGHT. 1979. Allopreening inowls: what are its functions? Auk 96:525-531.

FRANKLIN, J. F., K. CROMACK, JR., W. DENISON, A.MCKEE, C. MASER, J. SEDELL. F. SWANSON, ANDG. JUDAS. 1981. Ecological characteristics ofold-growth Douglas-fir forests. U.S Dep. Agric.For. Serv. Gen. Tech. Rep. PNW-118. 48pp.

, AND C. T. DYRNESS. 1973. Natural vege-tation of Oregon and Washington. U.S. Dep.Agric. For. Serv. Gen. Tech. Rep. PNW-8.417pp.

FULLER, M. R. 1979. Spatiotemporal ecology offour sympatric raptor species. Ph.D. Thesis. Univ.of Minnesota, Minneapolis. 220pp.

GABRIELSON, I. N., AND S. G. JEWETT. 1940. Birdsof Oregon. Oregot"--State College, Corvallis.650pp.

GASHWILER, J. S. 1959. Small mammal study inwest-central Oregon. J. MammaL 400:129-139.

GIGER, R. D. 1965. Surface activity of moles as.indicated by remains in barn owl pellets. Mur-relet 46:32-36.

Comm, C. 1., JR. 1974. The status of the spottedowl in California. Unpubl. Tech. Rep.. Calif.Dep. Fish and Game. Sacramento. 36pp.

. 1977. Distribution of the spotted owl inCalifornia. West. Birds 8:131-146.

1979. Status and management of elf and.spotted owls in California. Pages 86-97 in P. P.Schaeffer. and S. M. Ehlers, eds. Owls of thewest: their ecology and conservation. Proc. of aNatl. Audubon Soc. Symp. at Calif. Acad. Sat,San Francisco. Available from Natl. AudubonSoc. West. Educe Cent., Tiburon, Calif. 97pp..

GRINNELL, J., J. DIXON, AND J. M. LINSDALE. 1930.Vertebrate natural history of a section of north-ern California through the Lassen Park region.Univ. Calif., Berkeley, Pubi. ZooL 594pp.

HAVERSCHMIDT, F. 1946. Observations on thebreeding of the little owl. Ardor 34:214-248.

HELLER, E. 1893. Article without title. Oologist 10:236.

HOGLUND, N. H., AND E. LANSCREN. 1988. Thegreat grey owl and its prey in Sweden. Vilftevy(Stockholm) 5:364-421.

HOOVEN. E. F. 1972. Effect of clearcutdng a.Douglas-fir stand upon small mammal popula-tions in western Oregon. Ph.D. Thesis. OregonState Univ., Corvallis. 170pp.

HOUSTON, C S. 1978. Recoveries of Saskatchemn-banded great horned owls. Can Field Nat.• 92:61-66.

JANSSON, E. 1964. Anteckinger rorande hackandespannuggla (Glaucidium • passerinum). Vas Fa-gelvarld 23209-222.

JENNRICH, R. 1., AND F. B. TURNER. 1969. Mea-surement of non-circular home range. j. Theor.BioL 22:227-237.

JONES, J. K., JR., D. C. CARTER. AND K H.GENOWAYS. 1973. Checklist of North Amen-can mammals north of Mexico. Ocees. Pap., Mus.Texas Tech. Univ., Lubbock. 12:1-14.

KARALUS, K. E., AND A. W. ECZERT. 1974. Theowls of North America. Doubleday 4 Co.. Inc.,Garden City, N.Y. 278pp.

ICERTELL, D. 1978. The spotted owl at Zion Na-tional Park, Utah. West. Birds 8:147-150.

KILGORE, B. M. 1971. Response of breeding.bird

MI6Am. Midl. Nat. 85:135-152.population to habitat changes in a giant sequoia

ICONIC, C. 1965. "Zeigen" der Brutbohle bei Sper-

THE SPOTTED OWL IN OREGON-Foreman et al. 59

lingskauz (Glaucidium passerinum) and Zwer-gohreule (Otus stops). J. fur Omithologie 106:349-350.

LIGON, J. D. 1968. The biology of the elf owl (Mi-crathene whitneyi). Univ. Mich. Mus. ZooL Misc.Publ. 136. 70pp.

LIGON, J. S. 1926. Habits of the spotted owl (Syr-mum occiderialc). Auk 43:421-429.

MARSHALL, D. B. 1969. Endangered plants and an-imals of .,,egon. III. Birds. Oreg. State Univ.,Agric. Exptn. Spec. Rep. 278. 23pp.

MARSHALL, J. T.. JR. 1942. Food and habitat of thespotted owl. Condor 44:66-67.

. 1957. Birds of pine-oak woodland in south-ern Arizonl and adjacent Mexico. Pac. CoastAvifauna 32:78-79.

MASER. C. 1965. Spotted owl preys on dusky three(sic) mice. Murrelet 46:46.

MnocoLA, H. 1970. On the activity and food of thepygmy owl (G/aucidium passerinum) duringbreeding. Ornis Fenn. 47:10-14.

MILLER, G. M. 1974. Grace M. Miller on spottedowls. Pages 130-146 in L. W. Walker, ed. Thebook of owls. Alfred A. Knopf, Inc., New York.N.Y. 255pp.

MUELLER. H. C. 1974. Food caching behavior inthe American kestrel (Falco tparverius). L Tier-psychol. 34:105-114.

MUMFORD, R. E., AND R. L Zust. 1958. Notes onmovements, territory, and habitat of winteringsaw-whet owls. Wilson Bull. 70:188-191.

NEU, C. W., C. R. BYERS, AND J. M. PEEK. 1974.A technique for analysis of utilization-availabil-ity data. J. WildL Manage. 38:541-545.

NICHOU-S. T. IL 1970. Ecology of barred owls asdetermined by an automatic radio-tracking sys-tem. Ph.D. Thesis. Univ. of Minnesota, Minne-apolis. D

D. W. WARNER. 1972. Barred owlhabitat use as determined by radiotelemetry. J.WildL Manage 36/13-224.

OREGON-WASHINGTON INTERAGENCY WILDLIFECOMMITTEE. 1980. Oregon interagency spot-ted owl management plan. Oregon Dep. Fishand WildL, Portland. 3pp.

PEYTON, L 1910. Nesting of the spotted owl innortheastern Los Angeles County, California.Condor 12:122-123.

PHELAN. F. J. S. 1977. Food caching in the screechowl. Condor 79:127.

PnELEA, F. A., P. Q. TOMICH. AND G. W. TREICHEL.1955. Ecological relations of jaegers and owlsas lemming predators near Barrow, Alaska. EcoLMonogr. 25:85-117.

Posrovrr, H. 1979. A survey of the spotted owl(Stria occidentalis) in northwestern Washing-ton. UnpubL Rep., Natl. For. Prod. AMC.,Washington, D.C. 15pp.

REYNOLDS, R. T. 1972. Sexual dimorphism in ac-cipiter hawks: a new hypothesis. Condor 74:191-197.

Ruscm D. H.. E. C. MESLOW, P. D. DOERR, AND L.B. KEITH. 1972. Response of great horned owlpopulations to changing prey densities. J. WildLManage. 36:282-296.

SCHERZINGER, W. 1980. Zur ethologie der fort-pflanzung and jugendentwicklung des ha-bichtskauzes (Stria uralensis) mit vergleichenzum waldkauz (Sens aluco). Bonn. ZooL Mono-gr. 15. 66pp.

SCHNELL, J. H. 1958. Nestling behavior and foodhabits of goshawks in the Sierra Nevada of Cal-ifornia. Condor 60:377-403.

SCHOMER, T. W. 1968. Sizes of feeding territoriesamong birds. Ecology 49:123-141.

SEASTEDT. T. lip AND S. F. MACLEAN. 1979. Ter-ritory size and composition in relation to re-source abundance in lapland longspurs breedingin Arctic Alaska. Auk 96:131-142.

SELANDER„ R. K. 1966. Sexual dimorphism and dif-ferential niche utilization in birds. Condor 68:113-151.

SMITH, C. C. 1963. First breeding record of thespotted owl in British Columbia. Condor 65:440.

SNYDER, N. F. R., AND J. W. WEXY. 1976. Sexualsize dimorphism in hawks and owls of NorthAmerica. Ornithol. Monogr. 2Q 96pp.

SOUTHERN, IL N. 1970. The natural control of apopulation of tawny owls (SW: absco). J. Zool.London 162:197-285.

STENGER, J. 1958. Food habits and available foodof ovenbirds in relation to territory size. Auk 75:335-346.

9roftER, R. W. 1966. Sexual dimorphism and foodhabits in three North American accipiten. Auk83:423-436.

T1RRELL, P. B. 197& Protocalliphora aoium (Dip-ten) myiasis in great horned owls, red-tailedhawks. and Swainson s hawks in North Dakota.Raptor Res. 1221-27.

ZARN, M. 1974. Habitat management series forunique or endangered species. Report No. 10.Spotted owl (Stria occickntatis). U.S. Dep. In-ter.. Bur. Land Manage., Denver Serv. Cent.,Denver. Cob. 22pp.

Received 17 May 1982.Accepted 8 November 1983.

MM.

60 WILDLIFE MONOGRAPHS

APPENDIXES

Appendix 1. Vernacular and scientific names of plants mentioned in this report. Nomenclature follows Franklin and(1973).

a

Bigleaf maple Acer macrophyllum Ponderosa pine ,jnus ponderosaCalifornia-laurel Umbellidaria call-

forntcaPort-Orford-cedar Charnaecyparis

lawsonianaCanyon live oak Quercus chrysolepis Red alder Alnus rubraCoast redwood Sequoia sempervirens Shasta red fir Abies magnificaCreambush oceanspray Holodtscus discolor var. shastensis

(Roch spiraea) Sitka alder Alnus sfituataDouglas-fir Pseudotsuga mensiesii Sitka spruce Pieta sitchensisDwarf mistletoeGolden chinkapin

Arceuthobium spp.Castanopsis chryso-

Subalpine firSugar pine

Abies lasiocarpePious lambertiana

Grand firphylla

Abies grandisTanoak Lithocarpus densi-

AnusIncense-cedar Libocedrus decurrens Vine maple Acer cirri/WarnLodgepole pine Pious contorta Western (Cali- Corylus cornutaMountain hemlock Tsuga mertensiana fornia) hazelNoble fir Abies procera Western hemlock Tsuga haterophyllaOregon ash Frazinus laiifolia Western redcedar Thuja plicateOregon white oak Quercus garryana Western white pine Pious monticolaPacific dogwood Cornus nutfallit Western (Pacific) yew Tazus brevifoliaPacific madrone Arbutus menziesii White fir Abies concolorPacific silver fir Abies amabilis

Appendix 2. Cover types used for habitat selection analyses on the H. J. Andrews and BLM study areas.

Covet type

Area in whichcarer typeoccurred Dakuption

Multilayered old-growth associations of Douglas-fir, western hem-lock, and western redcedar with understories dominated by west-ern hemlock, vine maple, Pacific dogwood. bigleaf maple, andwestern yew. Also included mixed associations of old-growth Pa-cific silver fir, Douglas-fir, and western hemlock on the- HJAstudy area. On the HJA study area, old-growth overstory treeswere typically 275-450 years old, 50-70 m tall, and 100-150 cmdbh. On the BLM study area, old-growth overstory trees weretypically 220-350 years old, 50-65 m tall, and 100-170 cm dbh.Canopy closure in most stands averaged between 70 and 804%.

HJA Same as above except that some removal of logs and dead or dam-aged trees had occurred, resulting in a slight reduction in canopyclosure. On the BLM study area this cover type was uncommonand was lumped with unlogged old-growth conifer forest.

HJA, BLM Unlogged mature forests of Douglas-fir with variable amounts ofwestern hemlock, vine maple, and Pacific dogwood in the under-story. Also included small areas of mixed associations of Pacificsilver fir, Douglas-fir, and western hemlock on the HJA studyarea. Trees typically 40-65 m tall and 40-130 cm dbh.

This type included a 4-ha mixed stand of large Oregon asb andOregon white oak on the BLM study area On the HJA studyarea it included riparian mixtures of bigleaf maple, red alder,western redcedar, western hemlock and Douglas-fir.

Selectively loggedold-growth coni-fer forest

81-200-year-oldconifer forest

Old-growth conifer HJA, BLMforest

Mature hardwoods HJA. BLM

4THE SMUTTED OWL IN OREGON—Forstnan et al. 61

Appendix 2. Centime&

Cover typ,e

Araw in whichCOW typeoccurred Deaeription

61-80-year-old co-nifer forest

-

31-60-year-old co-nifer forest

•

21-35-year-old co-nifer forest

5-20-year-old coni-fer forest

Recent clear-cutsor burns

Rock talus outcropsand rockslides

Rock quarries orgravel storageareas

Shelterwood har-vest unit

Blue River Reser-voir

HJA, BLM Unlogged older second-growth forests dominated by Douglas-fir.Trees typically 20-40 m tall and 25-75 cm dbh. Understory de-velopment in these stands was variable, ranging from few under-story trees and little ground cover to a well-developed layer ofvine maple, Douglas-fir, Pacific dogwood, and western hemlock.In most stands overstory .trees were relatively uniform in size.

HJA, BLM Included 31-60-year-old forests of Douglas-fir on the HJA studyarea and 36-60-year-old forests of Douglas-fir on the BLM studyarea (unlogged stands). Overstory trees typically 18-40 in talland 20-60 cm dbh.

HJA, BLM Included 21-30-year-old stands of Douglas-fir on the HJA studyarea and 25-35-year-old stands of Douglas-fir on the BLM studyarea. These sites had been clear-cut between the late 1930's andearly 1950's. Trees typically 10-20 cm dbh and 15-46 m tall.Along streams on the BLM area these stands often consisted ofmixtures of red alder and Douglas-fir.

HJA, BLM Stands of Douglas-fir that were 5-26 years old on the HJA studyarea and 5-24 years old on the BLM study area. Growing onsites clear-cut or burned between 1950 and 1973.

HJA, BLM Clear-cut or burned areas overgrown by mixtures of woody shrubssuch as vine maple, Pacific dogwood, and salal. or by grass-forbassociations.

HJA Areas of coarse rocky rubble.

HJA, BLM

HJA

HJA

Small (0.2-1 ha) areas where rock was quarried or stored.

A mixed stand of old-growth Douglas-fir and western hemlock inwhich all trees had been harvested except for individual old-growth trees that were spaced 20-30 m apart.

This type included the reservoir and grassy slopes fanning a nar-row border around the reservoir. Reservoir was drained duringwinter months.

A public campground characterized by a mixed stand of matureand old-growth Douglas-fir. bigleaf maple, western redoedar,and western hemlock that had been selectively thinned to openup campsites and parking areas.

The Tokatee golf course bordered the southern edge of the studyarea and fell within the home range area used by the 1A male.

Cleared area beneath a high voltage powerline near the southernboundary of the study area.

Areas cleared for grazing by domestic livestock.

Blue River Reser- HJAvoir campground

Golf course HJA

Powerline right of HJAway

Pastureland

BLM

62 WILDLIFE MONOGRAPHS

Appendix 3. Vemacuiar and scientific names of animals mentioned in this report. ft a apanes warred in the diet Its meanweight or range of weights is also green.

Vernacular. Scientific Dame

Meta weight(g) at nine

of weights usedfoe biotin*caleulatiots

Sourcefor

weighteat-web

TAW. ambers(from Mt) hiwhich species

oosurredIn diets

MammalsBig brown bat Eptestcus fuscus 16 3 13Brush rabbit Sylvtiagus bachmant 150-900 1 12; 14; 15FBushy-tailed woodrat Neotoma mnerea 265 2 12; 13; 14; 15B,

C. D. E.F,CCalifornia vole Microtus c.altfornicus 39 3 14; 15C, EChickaree Tamtascturus douglasts 208 2 12; 13; 15A, B,

D, F, GCoast mole Scapanus amnia 53 2 12; 13; 14; 15F,

Creeping vole Microtus °morn 19 2• 12; 13; 14; 15A,B. C. E..F

Deer mouse Peromyocus mantculatus 22 2. all areasDusky-footed woodrat Neotoma *wipes 269 2 12; 14; 15A, FDusky shrew Sorer °Lacuna 9 2 15BErmine Mustela ermines 41 13Heather vole Phenacomys internzedtus 27 2 15GLong-tailed weasel 'Mustela frauds; 200 2 14; 15FMarsh shrew Sorer bendini 18 2 13Mazama pocket gopher Thomomys mazama 87 2 14; 15B, C.

D, E, GMountain beaver Aplodontia ftlfa 120-400 1 12; 15ANorthern flying squirrel Glaucomys sabrintss 115 2 all areasPacific shrew Sorer pactficus 12 2 13;15FPika Ochotona princeps 158 2 15B, GRed tree vole Phenacomys lonocaudus 27 2 12; 14; 15A. B.

D. FShrew mole Neurotrichus ebb** 9 2 12; 13; 14; 15C,

CSilver-haired bat Lasionycterts nocttvagans 10 12Snowshoe hare Lepus omericanos 150-1,500 1 12; 13; 15A. B.

C. D. E,F.GTownsend's chipmunk &star:nits townsman 80 2 12;13; 14; 15B,

D, E.F,GTownsend's vole Microtus townsendti 54 3 12Trowbridge shrew Sewer trobtagtt 5 2 12; 13; 15B, E. GVagrant shrew Sorer vagrans 5 2 12Water shrew Sorer palustris 11 2 15BWater vole Aron:01a richardront 70 2 13; 15B, GWestern jumping mouse Zapus tnnotatus 24 3 12; 13; 15B, FWestern red-backed vole Clethrionomys occidentalis 23 2 all areasUnidentified shrew 10 4 12; 13; 14; 15A,

C. D. F, GUnidentified squirrel 265 4 12;13; 14; 15B.

D. E, F, GUnidentified woodrat 100 4 12; 13; 14; L5GUnidentified vole/mouse 25 4 12;13; 14;15B.

C D. E. F. GBirds

Barred owl Stn.: voilaBlack-backed three-toed

woodpeckerPtcotdes arcticus 69 10 156

Blue grouse Dendragapus obscures 500 1 12;15C, FBrown creeper Certhia familiaris 8 6 12;14

•

Tun SrorrED OWL IN OREGON—Forsman et al. 63

appendix 3. Confined.

Venneularr Wend& name

Mean weight(g) or range

of weights usedfor biomasscalculation

Sourcefor

weightsate-

mateb

Table numbers(from text) inwhich mimes

occurredIn diets

Chestnut: backed chickadee Parus rufescens 10 4 12; 13Cooper's hatI - Accipiter cooperitDark-eyed junco Junco oreganus 18 6 12; 15FElf owl 'ilk' . Micrathene whitneytEvening grqsbeak Hespertphona vespertina 64 6 12; 13; 15FGolden-crowned kinglet Regulus satrapa 6 6 12Goshawk Accipiter gentilisGray jay Perisoreus canadensisGreat gray owl Striz nebulosaGreat horned owl Bubo virgintanusHairy woodpecker Dendrocopus villosus 70 6 15EHawk owl Surnta ululaKestrel Falco :parvenusLong-eared owl Asto otus 262 5 15GMountain quail Oreortyx pictus 244 6 15EOlive-sided flycatcher Nuttallornis borealis 34 6 15APygmy owl . Glaucidius gnoma 67 5 12; 13; 14; 15FRaven Corvus corazRed-shafted flicker Coloptes safer 146 6 12; 14; 15C, GRed-tailed hawk Buteo jamaicensisRobin Turdus migratorius 81 14; 15FSaw-whet owl Aegoltus acadicus 83 5 12; 13; 15E, F. GScreech owl Otus asto 140 4 12; 13; 14; 15ESteller's jay Cyanocitta stelleri 107 6 12; 13; 14;15C,

D, E, F, GTawny owl Strix alucoThrush spp. Hylocichla spp. 26 6 13Varied thrush Izoreus naevius 79 7 12; 14; 15A, B, EWestern tanager Piranga ludoviciana 31 6 15C, EWhite-headed woodpecker Dendrocopus albolarvatus 58 9 15CWinter wren Troglodytes troglodytes 9 6 13Yellow-bellied sapsucker Sphyrapicus vartus 46 6 12; 13; 15A, F, GUnidentified medium-sized

bird70 4 12; 13; 14; 15A,

F. GUnidentified small bird 30 4 all areasUnidentified small owl 90 4 12; 13; 15F, G

ReptilesGarter snake Thamnophis spp. 50 8 12; 13Western fence lizard Sceloporus occidentalis 10 8 12

CrustaceansCrayfish Pasifastacus spp. 20 4 15F

InsectsProphalangopsidae

Cyphoderris monstrosa 2 4 13; 15B. C. E. CBlattidae

Gryptocercus punctulatus 1 4 15D, EPentatomidae

Chlorochroa spp. 0.5 4. 12Mediae

Okanagana spp. 0.5 4. 14; 15ECorydalidae

Dysmicoltermes disjunctus 2 4 15BRaphidiidae

Genus unknown 0.5 4 15G

bwriktwall11: tvt 0341110.

64 WILDLIFE MONOGRAPHS

Appencex 3. Continued.

Vermicular. Scientific name

Mean weight(g) or range

of weights usedfor Manuseatsiatioll....

Sourcefor

weightesti-

match

Table number(from text) inwhich species

occurredin diete

Insects (cont.) CerambycidaeErgates spiculatus 2 4 13; 14; 15C.

D. E., GPrionus caltfornicus 2 4 15BOrtholeptura oalida 0.5 4 15B

CerambicidaeOrtholeptura spp. 0.5 4 15CCentrodera spurca 0.5 4 12; 15GMonochmus oregonensis 0.5 4 15CArhopalus productus 0.5 4 13Acmaeops spp. 0.5 4 15GPlectrura spinicauda 0.5 4 13

CarabidaeScaphinotus marginatus 0.5 4 12Pterostichus lama 0.5 4 13; 15FPterostichus herculaneus 0.5 4 12

- Pterostichus neobrunneus 0.5 4 15GPterostichus amethystinus 0.5 4 13

CupedidaePriacma serrata 0.5 4 15G

ScarabacidaePleocoma dubitalis 0.5 4 15FBolboceras obesus 0.5 4 15GPolyphylla decemlineata 0.5 4 13

TenebrionidaeUnidentified spp. 0.5 4 12

ElateridaeCtenicera spp. 0.5 4 15B

CicindelidaeOmus californicus 0.5 4 12.

LucanidaeCeruchus striatus 0.5 4 15BSinodendron rugosum 0.5 4 12.

CurculionidaePanscopus spp. 0.5 4 14Unidentified spp. 0.5 4 13Dyslobus lecontei 0.5 4 13Dyslobus spp. 0.5 4 13; 15G

FormicidaeFormica fusca 0.2 4 15GCamponotus spp. 0.3 4 13; 15B. C,

L GUnidentified ant 0.2 4 14

Unidentified insects 0.5 4 12; 13; 14; 15CSpiders

Unidentified spider 0.5 4 12; 13Gastropods

Unidentified snail 2.0 4 12

s Vernacular and scientific names of mammals follow /ones at al. 1973.• Sources were (I) weights tenanted based on size of skeletal fragments in individual penes. (2) Chris Maser (unpubl. deal. (3) speeeneor

Oregon State Universit y collections. (4) estimated average not substantiated by speeenere. (5) Earhart and Johnson (1970). (6) Kilgore (1971). (7)unpublished data from records at Point Reyes Bird Observatory, California. (8) Ron Nussbaum (unpubl. data). (9) Bock and Lynch (19701. sad(10) Grinnell at al. (1030). -

e Mess are described in each table.

Penalised by USDA ForestSHOW fit official use