Diets and Foraging Behavior of Northern Spotted Owls in Oregon

j RaptorRes. 38(3):214-230 ¸ 2004 The Raptor ResearchFoundation, Inc.

DIETS AND FORAGING BEHAVIOR OF NORTHERN SPOTTED OWLS IN OREGON

EPdC D. FORSMAN 1 U.S.Forest Service, Pacific Northwest Research Station, 3200 SWJeffersonWay, Corvallis, OR 97331 U.S.A.

ROBERT G. ANTHONY AND E. CHARLES MESLOW 2 USDI GeologicalSurvey, Oregon Cooperative Fish and WildliJbResearch Unit, Departmentof Fishe•es and WildliJb, OregonState University, Corvallis, OR 97331 U.S.A.

CYNTHIAJ. ZABEL U.S.Forest Service, Pacific Southwest Research Station, 1700 BayviewDrive, Arcata, CA 95521 U.S.A

ABSTRACT.--Wedescribe local, regional, and annual variation in diets of northern SpottedOwls (Strix occidentaliscaurina) in Oregon basedon 24 497 prey collectedat 1118 owl territoriesin 1970-2003.The sampleincluded 91.5% mammals,4.3% birds,4.1% insects,and 0.1% other prey.The diet included a131 species,including 49 mammals,41 birds,3 reptiles,1 frog, 1 crayfish,I scorpion,2 snails,and 33 speciesof insects.On average,91.9 + 0.3% (SE) of preyin the diet were nocturnalanimals, 3.3 + 0.2% were diurnal,and 4.8 + 0.2% were activeboth dayand night. Of the preycaptured, 50.5 + 0.8% were arboreal, 18.7 _ 0.7% were scansorial, 4.8 + 0.2% were aerial, and 26.0 + 0.7% were terrestrial. Mean massof prey was 116.6 _+6.5 g. Diets varied among owl territories,geographic regions, and years;but were generallydominated by four to six speciesof nocturnalmammals, including northern flying squirrels ( Glaucomyssabrinus), woodrats (Neotoma fuscipes and N. cinerea),red tree voles(Arborimus Iongicaudus), western red-backedvoles (Clethrionomyscalifornicus), deer mice (Peromyscusmaniculatus), or gophers (Thomomysspp.). Estimatesof dietary evennesswere low,indicating diets dominated by a tkw speciesof mammals.Forest management practices that producehealthy populationsof arborealand scansorial mammalssuch as flying squirrels,woodrats, and red tree volesshould benefit northern SpottedOwls in Oregon and Washington. KEYWORDS: northernSpotted Owl; Strix occidentaliscaurina; diet;prey selection; northern flying squirrel; Glaucomyssabrinus; red treevole,, Arborimus longicaudus.

DIETA Y COMPORTAMIENTO DE FORRAJEO DE STRIX OCCIDENTALIS CAURINA EN OREGON

P•SUMEN.--Describimosla variaci6n local, regional y anual en la dieta de Strixoccidentalis caurina en Oregonen basea 24497 presascolectadas en 1118territorios de losbuhos para el refiodo1970-2003. La muestraincluy6 91.5% de mamfferos,4.3% de aves,4.1% de insectosy 0.1% de otraspresas. La dieta incluy6 ->131especies, incluyendo 49 mamiferos,41 aves,3 reptiles,1 rana, I pez, I escorpitn,2 caracolesy 33 especiesde insectos.En promedio 91.9 q- 0.3% (SE) de las presasen la diem theron animalesnocturnos, 3.3 -+ 0.2% fueron diurnosy 4.8 _ 0.2% fueron activosdurante el dia y la noche. De las prcsascapturadas, 50.5 _+ 0.8% theron arboreas, 18.7 q- 0.7% fueron scansorial,4.8 m 0.2% fueron aereasy 26.0 _+0.7% fueron terrcstres.La mediade la masade las presasfue de 116.6 q- 6.5 grm. Las dietasvariaron entre los territoriosde los buhos,las regionesgeogrfificas y los aftos;pero fueron generalmentedominadas entne cuatro a seiscspccics de mamlferosnocturnos, incluyendo a ardillasvoladoras (Glaucomys sabrinus), ratas (Neotoma fuscipes y N. cinerea)y ratones(Arborimus longicaudus, Clethrionomyscalijbrnicus, Peromyscus maniculatus, y Thomomysspp.). Las estimacionesde la uniformidad de la dieta rueton bajos,indicando que la dieta fue dominadapor unas pocasespecics de mamiferos. Las prficticasforestales que producen poblacionessaludables de mamiferos como ardillasvoladoras, ratasy ratonesdeben favorecera los buhosen Oregon y Washington. [Traduccitn de CfisarMarquez]

I E-mail address: eforsman@œs.œed.us 2 Present address: 8035 NW Oxbow, OR 97330 U.S.A.

214 SEPTEMBER2004 SPOTTED OWL DIET 215

Home-range areas, population cycles, and be- ness1973). For our analysiswe subdividedthe studyarea havior of owls are greatly influenced by the distri- into sevengeographic regions (Fig. 1). Regionalbound- ariesfollowed countylines, except that we usedInterstate bution, density,and behavior of their prey. To un- Highway 5 to subdividesamples from the Coast and Cas- derstand these relationships, biologists need cadesranges (Fig. 1). The easternedge of the studyarea detailed information on the diet of the predator, correspondedwith the easternlimits of the range of the including data on local and regional variation. Ex- Spotted Owl in Oregon (Fig. 1). amination of the diet can provide many clues re- METHODS garding foraging behavior, habitat selection,and degree of dietary specialization.This information Pellets were collected below owl roosts, air-dried, and stored in labeled plastic bags until they could be ana- is particularly important for understandingwhich lyzed. Some pellets were obtained from radio-marked typesof prey are most important to a predator in owls during fall and winter (Forsman et al. 1984, Miller different regions, and for understanding the nu- et al. 1997), but mostwere collected during the breeding meric impact of the predator on its prey. season (March-August) when we visited historic nest ar- There have been numerous studies of northern eas to locate and band owls.With the exception of a few radio-markedowls (Forsmanet al. 1984), no attempt was Spotted Owls (Strix occidentaliscaurina) in Oregon, made to sample different individuals or territories ran- Washington, and California. The earliest of these domly or systematically,although many territories were studiesfocused primarily on distribution,basic life sampled in multiple years. Territories were identified based on occupancyby pairs of Spotted Owls, many of history attributes, dispersal,and habitat selection which were banded or radio-marked. of the species(e.g., Gould 1977, Barrows 1980, We basedall analyseson the estimatednumber of prey Forsmartet al. 1984, Miller 1989, Carey et al. 1992, or biomass of prey in each sample. We estimated the Miller et al. 1997, Zabel et al. 1995, Thrailkill et al. number of prey in each sample by counting skulls,man- 1997). These pioneering efforts have been fol- dibles, bones of the appendicular skeleton, or pieces of exoskeleton,whichever gave the highest count. In a few lowed by nearly 20 yr of demographic studies,in caseswe also used hair or feathers to identify prey. To which researchersused mark-recapturemethods to avoid double counting remains of large prey that ap- estimatepopulation trends of the owl (e.g., Burn- peared in several pellets, we combined remains from ham et al. 1994, Forsmart et al. 1996, Franklin et multiple pellets or pellet fragments found at the same al. 2000). During many of these studies,investiga- roost on the samedate. We used dichotomouskeys (Ma- set and Storm 1970, Vertsand Garraway1984) and a ref- tors routinely collected pellets from Spotted Owl erence collection of bird and mammal skeletons to iden- roost areas to determine the composition of the tify remains in pellets. diet. Some of these data have been published (e.g., We estimated biomassby multiplying the number of Forsmart et al. 1984, Barrows 1980, Thomas et al. individualsof each speciesin a sampleby the mean mass of the species,or by estimatingand summingthe unique 1990, Ward 1990, Zabel et al. 1995), but most of massof each prey item in the sample.The latter method the data from Oregon have never been published. wasused only for large prey suchas snowshoe hare (Lepus In 1996-2003, we contacted most of the research- americanus),brush rabbits (5•lvilagus bachmani), and ers who have studied Spotted Owls in Oregon and mountain beaver (Aplodontiarufa), because those prey solicited their assistancein compiling all of the types were mostly represented in our samplesby small .juveniles,and would have been overestimated if we used available information on the diet of the Spotted mean massfrom museum specimens.For these specms, Owl in Oregon. Here, we summarizeresults of that we estimated massbased on comparisonsof bones w•th effort, compare regional, local, and annual varia- specimensof known massin our reference collectionand tion in the diet; and estimate annual number of we made the simplifyingassumption that masswas line- arly correlatedwith the sizeof bonesin pellets.Estimates prey captured by individual Spotted Owls in differ- of mean mass for birds and mammals were obtained from ent regions. Dunning (1993) and Verts and Carraway (1998; Appen- dix). For invertebrates,snakes, lizards, and amphibians, STUDY APEX we used estimatesof mean massfrom local specimensor The study area included western Oregon and the east- from similar speciesin the published literature (Smith ern slopesof the CascadesRange (Fig. 1). With the ex- and Murphy 1973;Appendix). ception of the lowland interior valleysof westernOregon, To evaluate when and where owls were foraging, we this region is characterized by mountainous terrain cov- subdivided prey into groups based on their primary pe- ered by coniferousforests. Forest composition is predom- riod of activity (nocturnal, diurnal, or both), and thmr inantly Douglas-fir (Pseudotsugamenziesii) and western primary activityzone (terrestrial, arboreal, scansorial,or hemlock (Tsuga heterophylla)in the Coast Ranges and aerial; Appendix). These classificationswere basedon in- western CascadesRange, mixed-conifer or mixed-ever- formation in Verts and Cartaway (1998) as well as our green forestsin southern Oregon, and mixed conifer for- own observationsof animals in our study areas. For th•s estson the eastslope of the Cascades(Franklin and Dyr- analysiswe classifiedmammals as arboreal if they nested 216 FORSMAN ET AL. VOL. 38, NO. 3

CLACKAMA$ • * 6

Geographic Regions 1 - North Coast Range 2 - Central Coast Range DOUGLAS 3- South Coast Range 4- Interior Southwest 5 - Central Cascades

KLAMATH 6- North Cascades 7- Eastern Cascades

N 100 0 100 km

Figure 1. The Oregon study area, illustrating locationsof 1118 northern Spotted Owl territories from which we collected data on owl diets, and sevengeographic regions that we used for regional comparisonsof diet. Boundaries of geographic regions followed county lines except that we used Interstate Highway 5 to divide samplesfi:om the Coast Ranges and CascadesMountains north of Josephine County (solid dark line bisecting Douglas, Lane, Linn, Marion, Clackareas, and Multnomah counties).

primarily in trees, and scansorialif they spent much of of means were based on a subset of owl territories from their time climbing in trees but often nested on the which we obtained samplesof -->10prey items,regardless ground (Appendix). This analysisprovided a general of the number of yearsin which pelletswere collectedat measureof the amount of owl foraging that occurred in a particular territory. We selected 10 as the minimum different time periods or activityzones, as we knew from sample size for analysisafter exploratory analyseswith field observationsthat prey wcre not alwayscaptured in larger sample sizes(i.e., 20, 50) indicated that the choice their primary activityperiods or activityzones (Sovern et of minimum sample size made little difference to our al. 1994). conclusions.We estimated mean massof prey captured We used all prey remains to compile a list of species by each pair of owls by dividing the total biomassin the captured by northern Spotted Owls (Appendix). We es- sample by the number of prey in the sample. Regional timated mean dietary compositionby computing the diet means were then estimated by averagingamong pairs. m each owl territory, and averaging acrossterritories to For comparisons of dietary composition among and get the mean and SE fbr each prey category. Estimates within regions we subdivided the diet into 18 prey cate- SEPTEMBER2004 SPOTTED OWL DIET 217 gories and used one-wayanalysis of'variance to compare captured per owl by 2.0 for non-nestiug pairs, and 2.6 fo• mean proportionsof each prey typein the diet. Statistical nesting pairs with two young. Estimates of prey capture tests of proportional data were based on arcsine trans- for adultswith young assumedthat nesting pairs fed their formations to better meet the assumptionsof the tests young for ca. 4 mo (Forsman et al. 1984), and that the (Sokal and Rohlf 1969). However, the results were virtu- mean biomass consumed per day was the same for juve- ally identical regardlessof whether the data were trans- niles and adults. The latter assumptionwas a simplifica- fbrmed or not, so we present the untransformeddata in tion of the actual rate of daily food iutake by juveniles, all tables and figures. which was small at first, and then increased as juveniles To compare the evennessof the diet in different re- matured. gions we used the reciprocal of Si•npson'sIndex (Hill All statistical analyseswere conducted with Program 1973). Simpson'sIndex (q) is equal to the sum of the SPSS (Version 11.5, SPSSInc., Chicago, IL U.S.A.). We squaredvalues of the proportionalabundances of all spe- set the significancelevel for statisticaltests at 0t = 0 05 cies (or groups) in a sample (Hill 1973). For a given All means are expressedas 2 -+ 1 SE. sample, the range of q is l?om 1/N (all speciesequally abundant) to 1 (only one speciesin diet). Gonversely,the RESUI •TS reciprocal of Simpson'sIndex (l/q) ranges from l-N, where 1 indicates a diet composed entirely of one spe- Attributes of Prey. We identified 24497 prey cies, and Nis a diet cronposedof more than one species items from 1118 owl territories, including 547 ter- with all speciesequally represented in the diet. In our ritories with N > 10 (Fig. 1). The total sample in- case, minimum and maximum values of 1/q were 1 and cluded 91.5% mammals, 4.3% birds, 4.1% insects, 18, where 1 indicted a diet composed entirely of 1 species, and 18 indicated a diet in which all 18 prey groups and 0.1% other prey (Appendix). The sample in- were equallyrepresented. We estimatedthe mean 1/q for cluded at least 131 species;49 mammals, 41 birds, each region as (E:vl/q)/N, where N = the number of 3 reptiles, 1 frog, 1 crayfish, 1 scorpion, 2 snails, territories in the sample. and 33 insects (Appendix). On average, 91.9 _+ We used X2 teststo examine among-yearvariation in the diet at individual territories that had samplesof -->20 0.3% of prey in the diet were nocturnal animals, prey in 2 or more years.For this analysiswe lumped prey 4.8 -+ 0.2% were active both day and night, and into sevengroups: (1) Lagomorphs(Lepus americanus, 5•l- 3.3 -+ 0.2% were diurnal (N = 547). Of the prey vilagusbachmani), (2) northern flying squirrels (Glauc0- captured, 50.5 -+ 0.8% were arboreal, 26.0 -+ 0.7% myssabrinus) , (3) deer mice ( Peromyscusmaniculatus) , (4) were terrestrial, 18.7 _+ 0.7% were scansorial, and woodrats (Neotomafuscipes, N. cinerea),(5) western red- backedvoles ( Clethrionomyscalifornicus), (6) red tree voles 4.8 -+ 0.2% were aerial (N = 547). The percent of (Arborimuslongicaudus), and (7) all other prey. prey that were either arboreal or scansorial was We used linear regressionto examine relationshipsbe- 70.8% in the North Coast, 74.8% in the Central tween elevation and the numeric proportion of several Coast, 77.8% in the South Coast, 64.8% in the In- speciesof small mammalsin the diet. For this analysiswe terior Southwest, 59.2% in the Central Cascades, useda singleestimate of the overalldiet at each territory, 56.7% in the North Cascades, and 64.3% in the regardlessof when pelletswere collected,and we used a single estimateof elevationat each owl territory, based East Cascades.Mean massof prey ranged from 91- on the elevation at the nest site or primary roost area. 142 g in the sevenregions, and was 116.6 + 6.5 g We determined elevations with altimeters, topographic overall (Table 1). maps,or a 30-m-resolutiondigital elevationmap. Dietary Evenness. Mean estimates of dietary We estimated the number of prey captured per year by individual owls based on the assumptionthat a Spotted evenness(l/q) for all regionswere near the lower Owl of averagemass (610 g) consumes12% of its body end of the scaleof possiblevalues, indicating diets mass(73.2 g) of food per day,or 26 718 g/yr. The num- dominated by a few speciesin all regions (Table ber of each speciescaptured per year wasthen estimated 2). Diets were most even in the Central Cascades, by multiplying the proportional biomassof each species Interior Southwest, South Coast, and Eastern Cas- m the diet by 26 718 and dividing by the mean massof the species.The critical assumptionin this analysiswas cades regions, and least even in the North Cas- the amount of food consumed per day. We could have cades,North Coast, and Central Coast regions (Ta- used a more conservativeestimate of 56-59 g/d (Weath- ble 2). ers et al. 2001), but we chose to use a slightly higher estimate because we wanted to allow for the fact that Regional Differences in Diet. Composition of SpottedOwls often discardstomachs, intestines, tails, and the diet differed among regions for 16 of the 18 other parts of the prey that they capture. Thus, we felt prey categoriesin Table 2 (P-values• 0.05). The that our estimateof 12% of body masswas a reasonable only two categoriesthat did not differ among re- measure of the amount of prey captured per day, espe- gionswere the "Bats" category(P = 0.70) and the cially considering other data on food consumption of Spotted Owls (Forsman 1980) and other owls (Graber category "Other," which included miscellaneous 1962). To estimate the number of prey captured per year prey such as molluscs,snakes, lizards, and scorpi- m each owl territory we multiplied the number of prey ons (P = 0.21). In some cases,differences among 218 FORSMAN ET AL. VOL. 38, NO. 3

Table 1. Mean mass(g) of prey capturedby northern SpottedOwls in different regionsof Oregon, 1970-2003. N is the number of owl territoriessampled in each region.

REGION N MEAN q- SE RANGE 95% CI

North CoastRegion 9 123.6 + 10.3 75-173 100-147 Central CoastRegion 90 112.8 + 3.2 41-213 106-119 SouthernCoast Region 180 131.4 + 2.7 55-317 126-137 Interior Southwest 75 142.1 + 5.0 25-242 132-152 Central Cascades 154 108.5 + 2.7 44-209 103-114 North Cascades 4 90.7 + 14.1 67-130 46-136 East Cascades 35 106.7 ___7.3 11-247 92-121 All areas • 7 116.6 q- 6.5

Grand mean of sevenregional means.

regionswere expected becausesome prey species, mals, including northern flying squirrels,woodrats, suchas American pikas (Och0t0naprinceps) and red red tree voles, western red-backed voles, deer mice, tree voles did not occur in all regions. In most cas- or gophers (Thomomysspp.; Tables 2-3). Northern es, however,there were no obvious a priori reasons flying squirrelswere the most common animal in to expect regional variation in prey composition. the diet in most regions,averaging 28-52% of prey Although there was considerable variation numbers and 30-74% of prey biomass(Tables 2- among regions,the diet in all areaswas composed 3). Woodratscomprised 11.7 + 3.3% of prey num- mainly of four to six speciesof nocturnal main- bers, 24.9 --- 5.3% of prey biomass,and were most

Table 2. Mean percent (q-1 SE) of prey numbersin diets of northern Spotted Owls in sevendifferent geographic regionsof Oregon, 1970-2003. Sample size (number of owl territorieswith >--10prey items) is in parentheses."T" indicates trace amount (<0.05%).

NORTH CENTRAL SOUTH INTERIOR CENTRAL NORTH EASTERN COAST COAST COAST SOUTHWEST CASCADES CASCADES CASCADES PREY a (9) (90) (180) (75) (154) (4) (35)

Shrews 1.8 + 1.4 0.9 + 0.2 1.2 q- 0.2 2.8 -+ 0.6 1.9 q- 0.3 2.5 -+2.5 1.5 q- 0.5 Moles 0.2 --- 0.1 0.1 q- 0.0 2.8 -+ 0.6 1.0 q- 0.2 1.2 -+ 0.5 Bats 0.2 + 0.1 0.5 + 0.1 0.3 q- 0.2 0.4 q- 0.1 0.3 q- 0.2 Rabbits/hares 0.8 + 0.4 3.6 + 0.5 4.6 + 0.4 2.6 q- 0.4 4.8 + 0.4 4.5 + 0.8 American pika 0.1 _+0.1 0.9 q- 0.• 0.3 q- 0.2 Douglas' squirrel 1.3 q- 0.8 0.6 q- 0.2 0.3 q- 0.1 0.1 q- 0.1 0.4 --- 0.1 0.1 q- 0.1 Chipmunks 2.1 --- 1.4 0.7 q- 0.2 0.6 -+ 0.1 1.5 q- 0.3 1.4 q- 0.2 2.3 -+ 2.3 1.3 q- 0.5 N flying squirrel 48.3 q- 3.6 49.5 --+1.6 36.0 q- 1.2 28.2 + 2.0 34.6 -4--1.2 52.1 -+ 8.7 38.9 -+ 3.1 Gophers 2.6 q- 1.3 0.6 - 0.2 0.1 + 0.1 5.4 + 0.9 4.9 + 0.6 6.5 q- 1.7 Deer mouse 17.3 _4_-5.4 10.5 q- 1.1 6.2 -+ 0.6 4.9 + 0.6 6.1 -+ 0.7 2.9 -+ 0.7 Woodrats 11.1 q- 2.8 7.1 -+ 0.8 18.2 q- 0.9 27.8 -+ 2.4 9.5 q- 1.0 2.3 -+ 2.3 8.2 q- 1.8 W red-backed vole 2.2 +- 0.4 2.8 -+ 0.3 6.8 q- 0.7 11.0 -+ 0.7 26.9 q- 8.7 10.4 -+ 1.6 Red tree vole 4.8 q- 2.7 12.7 -+ 1.1 18.2 q- 0.9 2.6 q- 0.7 7.7 + 0.8 Mzcrotusspp. 1.2 q- 0.3 1.5 q- 0.2 2.5 q- 0.5 2.6 q- 0.4 1.1 q- 0.3 Other mammals 3.8 + 1.8 3.9 + 0.5 3.7 q- 0.3 1.4 + 0.3 5.1 q- 0.4 3.9 q- 0.8 Birds 3.8 q- 1.2 3.9 + 0.4 3.6 q- 0.3 5.7 + 0.9 4.1 + 0.3 13.9 q- 6.1 4.3 -+ 0.8 Insects 1.0 _+ 0.6 2.1 q- 0.7 2.4 q- 0.3 4.3 + 1.2 3.5 q- 0.6 14.5 -4--3.1 Other 0.4 q- 0.4 0.1 + 0.0 T 0.1 + 0.1 0.1 q- 0.1 0.1 q- 0.1 Sum (%) 100.0 100.0 100.0 100.0 100.0 100.0 100.0 1/qb 3.1 + 0.2 3.2 q- 0.1 3.9 -+ 0.1 3.9 -4--0.2 4.5 _+0.1 2.4 + 0.2 3.7 q- 0.3 See appendix for complete list of common and scientificnames of prey. 1/q = reciprocalof Simpson'sIndex. SEPTEMBER2004 SPOTTED OWLDIET 219

Table 3. Mean percent (_1 SE) of prey biomassin diets of northern SpottedOwls in sevendifferent geographic regionsof Oregon, 1970-2003.Sample size (number of owl territorieswith >10 prey items) is in parentheses."T" indicates trace amount (<0.05%).

NORTH CENTRAL SOUTH INTERIOR CENTRAL NORTH EASTERN COAST COAST COAST SOUTHWEST CASCADES CASCADES CASCADES PREY a (9) (90) (180) (75) (154) (4) (35)

Shrews 0.1 + 0.1 0.1 + T 0.1 + T 0.1 + 0.1 0.1 + T 0.2 + 0.2 0.1 __+0.1 Moles 0.1 + T T 1.3 + 0.3 0.6 + 0.1 0.7 + 0 3 Bats T T T T T Rabbits/hares 2.2 -+ 1.1 9.9 - 1.3 11.6 _+ 1.0 5.9 + 1.1 12.9 -+ 1.1 12.3 -+ 2 6 American pika 0.1 + 0.1 1.4 + 0.3 0.5 + 0.3 Douglas'squirrel 1.9 _+ 1.1 1.2 -+ 0.3 0.5 _+0.1 0.3 _+0.2 0.7 + 0.2 0.3 _+0.2 Chipmunks 1.7 _+1.1 0.5 _+0.1 0.4 _+0.1 1.2 +- 0.2 1.2 -+ 0.2 1.4 _+1.4 1.2 -+ 0 5 N. flying squirrel 52.3 + 3.9 58.3 + 1.6 38.6 _+ 1.6 30.2 -+ 2.5 45.5 + 1.7 74.5 + 4.5 50.7 -+ 4 2 Gophers 2.5 -+ 1.3 0.5 _+0.2 0.1 _+T 4.3 _+0.7 4.8 + 0.6 7.5 _+2 3 Deer mouse 3.8 + 1.5 2.5 -+ 0.3 1.2 -+ 0.1 1.0 _+ 0.1 1.4 _+ 0.2 0.7 _+ 0.2 Woodrats 25.8 -+ 4.7 16.1 _+ 1.5 37.1 _+ 1.6 48.5 _+ 3.4 20.7 -+ 1.8 5.0 m 5.0 18.2 -+ 3.4 W. red-backed vole 0.5 _+ 0.1 0.6 _+ 0.1 1.3 _+ 0.2 2.7 -+ 0.2 8.2 -+ 3.3 2.8 -+ 0 6 Red tree vole 1.0 _+ 0.6 3.7 _+ 0.5 4.2 -+ 0.3 0.6 _+ 0.2 2.2 -+ 0.3 Microtusspp. 0.3 -+ 0.1 0.4 _+0.1 0.6 -+ 0.1 0.9 -+ 0.1 0.4 -+ 0 1 Other mammals 4.2 -+ 2.4 2.4 -+ 0.4 2.4 -+ 0.3 0.9 _+ 0.4 1.9 _+ 0.2 1.0 _+ 0 3 Birds 4.3 m 2.3 3.9 -+ 0.5 2.8 -+ 0.3 3.5 -+ 0.6 2.8 -+ 0.3 10.7 -+ 5.9 2.9 m 0.9 Insects T T T 0.1 -+ 0.1 0.1 + T 0.7 + 0.4 Other 0.1 -+ 0.1 T T T T T Sum (%) 100.0 100.0 100.0 100.0 100.0 100.0 100.0 See appendix for complete list of common and scientificnames of prey. xmportantin the diet in the Interior Southwestand hares in pellets was 340 + 6 g (range = 50-1400 South Coast regions (Tables 2-3). Red tree voles g). American pikas occurred only in the Cascades comprised6.6 --- 2.6% of prey numbers,and were and Interior Southwestregions, where they aver- most common in the diet in the South and Central aged lessthan 1% of prey numbers(Table 2). Coast regions (Tables 2-3). Western red-backed Bats (Ghiroptera), shrews(Sorexspp.), and moles voles were uncommon in the diet in all coastal re- (Scapanusspp.) were uncommon in the diet in all gions,but comprised7-27% of prey numbersin areas except the Interior SouthwestRegion, where the Cascadesand Interior Southwestregions (Ta- the averagediet included 5.6% shrewsand moles ble 2). Deer mice comprised6.8 + 2.1% of prey (Table 2). The category "Other mammals" in Ta- numbers,with the highest occurrencein the Cen- bles 2-3 included small mammals that we could tral and Northern Coastranges (Table 2). Gophers not identify to speciesas well as a variety of small comprised5-6% of prey numbersin the Interior and medium-sized mammals that were uncommon Southwest, Central Cascades and Eastern Cascades in the diet. These included weasels (Mustela ermz- regions, but were comparativelyrare or uncom- nea, M. frenata), heather voles (Phenacomysinterme- mon in the diet in all coastalregions (Table 2). dius), white-footedvoles (Arborimusalbipes), moun- Chipmunks (Tamias spp.) and Douglas' squirrels tain beaver,western gray squirrels(Sciurus griseus), (Tamiasciurusdouglasii) together comprised 1-3% ground squirrels ( Spermophilusbeecheyi, S. lateralis), of prey numbers. spotted skunks ( Spilogalegradlis) , ringtails ( Bassar- Brush rabbits and/or snowshoehare comprised iscusastutus), and jumping mice ( Zapustrinotatus, 1-5% of the prey capturedand 2-13% of the bio- Z. princeps;Appendix). mass in the diet (Tables 2-3). Of 1010 rabbits or Of 56 mountain beaver in pellets, 55 werejuve- haresfound in pellets,826 (81.8%) werejuveniles, niles, I was a small subadult, and 48 (85%) were 63 (6.2%) were subadults, 114 (11.3%) were capturedin June or earlyJuly, when juvenile moun- adults, and 7 (0.7%) were of undetermined age. tain beaver first began to emerge from their natal The estimated mean mass of individual rabbits and dens (Lovejoy 1972). The five ringtails found in 220 FORSMAN ET AL. VOL. 38, NO. 3 pellets were all smalljuveniles with estimatedmass Table 4. Annual variation in diet at two different north- = 200-400 g. The only spotted skunk was a sub- ern Spotted Owl territories in Oregon. Numbers indicate adult, with estimatedmass ca. 500 g. percent of total prey in each annual sample.Annual sam- Birds averaged5.6 -+ 1.4% of prey numbersand ple sizesare in parentheses. 4.4 _+ 1.1% of prey biomass (Tables 2-3). Most small or medium-size birds that occurred in the OAK CREEK Brummet forests of western Oregon were taken at least oc- 1970 1972 1973 1974 1990 1991 casionally (Appendix). Of 540 birds identified to SPECIES/GROUP (174) (89) (34) (95) (45) (47) species or family, the most common were jays Rabbits/hares I 2 2 2 (17.6%), small owls (23.0%), woodpeckers N. flying squirrel 5 37 29 6 47 15 (12.8%), grouse and quail (3.7%), and Varied Deer mouse 8 15 79 5 8 Thrushes (Ixoreus naevius) or Atnerican Robins Woodrats 2 8 3 13 9 (Turdus migratorius)(24.4%). In most cases we W. red-backed vole 3 could not differentiate between skeletons of Varied Red tree vole 66 10 44 11 33 60 Thrushes and American Robins, so we combined Other prey 18 25 24 2 2 6 them for analysis. Total % 100 100 100 100 100 100 Insects were generally uncommon in the diet, except in the Eastern CascadesRegion (Table 2). Of 1005 insects identified, the two most common the Oak Creek territory, deer mice variedfrom 0% specieswere the great grig (Cyph0derrismonstrosa), of the diet in one year to 79% of the diet in an- a large scansorialcricket (41.0%), and the pon- other year (Table 4). derous borer (Ergatesspiculatus), a large woodbor- Variation in Diet with ChangingElevation. In the ing beetle (24.8%). Great grigs occurred only in CascadesMountains (Central Cascades,North Cas- the CascadesRange, where we commonly heard cades, East Cascadesregions), predation on red- them on summer nights, as they stridulated from backedvoles and gopherswas positively correlated elevated perches on tree trunks or tree limbs. with elevation(red-backed vole F•,•m = 27.7, P < Adult ponderous borers were noisy, clumsy fliers 0.001, /i? = 0.127; gophersFLm • = 17.66, P < that frequently crashed into limbs while flying 0.001, Rs = 0.085). In the Central CascadesRe- through the forest at night. gion, predation on red tree voles was negatively Other items in the diet included frogs (Rana correlated with elevation (F•,• 2 = 32.6, P < 0.001, spp.), snakes (Thamnophisordinoides, T. spp.), liz- RS = 0.177). Further examination of the data from ards (Sceloporusoccidentalis, Elgaria coeruleus),ter- the Central Cascades Region revealed that tree restrial snails ( Haplotremavancouverense, Monedenia volescomprised only 2.3 -+ 0.6% of the diet at ter- fidelis), crayfish (Pacifastacusleniusculus), and scor- ritoriesabove 975 m elevation(N= 51), compared pions (Ur0ct0nusmordax). All of these were rare in to 10.3 _+ 1.1% of the diet at territories below 975 the diet (<0.1% of total prey; Appendix). m (N = 103). The analysisof elevational limits of Local and Annual Variation in Diet. Composition tree voles was limited to the Central Cascades Re- of the diet was highly variable among owl territo- gion becausetree voleswere uncommon or absent ries within regions (all P-values <0.05). While in the diet in other regions in the Cascades,re- some of this variation was probably the result of gardlessof elevation. small sample size, we fbund similar results even Number of Prey Captured Per Year. The esti- when we limited the analysisto territories with sam- mated mean number of prey capturedper year was ples >50 (data not presented).This suggestedthat 271 _+22 fbr non-nestingindividuals, 543 _+44 for d•etary variation among territories was due to fac- non-nestingpairs, and 705 _+57 for nestingpairs tors other than sample size. with two young (Table 5). Mean estimatesfor non- Composition of the diet varied among years (P nesting individualsranged from 217-384 prey/yr < 0.05) at 25 of 56 territories where we collected among regions, or 0.6-1.0 prey items captured/d -->20 prey in 2 or more years. In most cases,the (Table 5). The only region in which the estimated &fferences were relativelysmall, but there were no- number of prey captured per year per non-nesting table exceptions. For example, at two territories owl was >300, was the Eastern Cascades, which was the percent of tree volesand flying squirrels in the the only region where the diet included large num- diet varied dramaticallyamong years (Table 4). At bers of insects. SEPTF.MgEP. 2004 St'OTTED OWL DIEq' 221 222 FORSMANET AL. VOL. 38, NO. 3

DISCUSSION an area where they were sympatric,Spotted Owls Composition of the Diet. Our resultsare similar captured larger prey on averagethan Barred Owls to previousstudies of SpottedOwl dietsin Oregon, (Strix varia; Hamer et al. 2001). Washington, and northern California, in that diets Regional Variation. All studiesthat have exam- were dominated by a few speciesof mammals, es- ined variation in diets of Spotted Owls have found pecially flying squirrels,woodrats, tree voles,red- differences among regions, territories, years, and backed voles, and juvenile lagomorphs (e.g., Fors- seasons(e.g., Forsman et al. 1984, 2001, Laymon man et al. 1984, 2001, Ward 1990, Cutler and Hays 1988, Ward 1990, Ganey 1992, Verner et al. 1992). 1991, Hamer et al. 2001). The relative frequency As discussedby Forsman et al. (1984, 2001), Bull of these mammals in the diet varied among re- and Henjum (1990), and Ward and Block (1995), gions, at least partly in responseto regional differ- there are numerous factorsthat probablycontrib- ences in their abundance or distribution (Ward et ute to this variation, including (1) annual, seasonal al. 1998, Carey et al. 1999). Flying squirrels tend or local variation in prey abundance,or availability, to be the most common item in the diet in western (2) individual variation in prey selection,(3) small Washington and northwestern Oregon, whereas sample size or unequal sampling effort, and (4) woodrats predominate in diets in southwest biased delivery of large prey to the female and Oregon and northern California (Barrows 1980, young by nesting males.All of thesefactors may be Ward et al. 1998; Tables 2-3). important, but there is compelling evidencethat In contrast to diets of northern Spotted Owls, spatialand temporal variation in prey populations diets of Spotted Owls in the southwesternU.S.A. and selectiveforaging by the owls are key factors and Mexico tend to include fewer arboreal mam- influencing the diet (Forsman et al. 1984, Ward •nals, and more woodrats, terrestrial small mam- 1990, Carey et al. 1992, Ward and Block 1995, mals, bats, and insecrus(e.g., Duncan and Sidner Ward et al. 1998). For example, a number of stud- 1990, Ganey 1992, Ward and Block 1995, Smith et ies suggestthat densitiesof dusky-footedwoodrats al. 1999). These differencesare probablydue pri- and deer mice vary considerably among years marily to regional differencesin prey availability. (Linsdale and Tevis 1951, Spevak1983, Ward and For example,flying squirrelsare uncommonor ab- Block 1995, Rosenberg et al. 2003), and among sent in the mountains of southern California and and within owl territories (Ward et al. 1998). Carey the southwestern U.S.A. et al. (1992) found that densitiesof northern flying As in our study, Smith et al. (1999) found that squirrels and woodratswere highly variable in sam- the majority of insectscaptured by Spotted Owls ple plots in different Spotted Owl territories in were large cricketsand beetles.Apparently, these western Oregon. types of insectsare easier for Spotted Owls to de- Some of the regional differences observed in tect and capture or are taken selectivelybecause of our studysuggest interesting hypotheses regarding their relativelyhigh biomassper unit effort. How- differences in abundance of small mammals. For ever,we also suspectthat small, soft bodied insects example,regional differencesin the abundanceof are generally underestimated in analysesof owl red-backedvoles in owl diets (Table 2) suggestthat pelletsbecause they are more completelydigested. red-backed voles are roughly 5-10 times more It is possiblethat the apparent specializationon abundant in the Oregon Cascadesthan in the cen- large insectsis at least partially due to this bias. tral and northern Coast Ranges. However, data Mean Prey Size. The mean mass of prey cap- from field studiesof red-backedvoles suggest that tured in this studyand in previousstudies of north- they are actuallymore common in the CoastRang- ern Spotted Owls (Ward et al. 1998, Forsmanet al. es than in the Cascades(Aubry et al. 1991). The 2001, Hamer et al. 2001) indicated that Spotted higher proportions of red-backedvoles in owl diets Owls feed on larger prey, on average, than most in the Cascades,especially at higher elevations, other large owls of the northern hemisphere. For may be due to prey switching,perhaps in response example, the much larger Great Horned Owl to lower numbers of alternate prey suchas red tree (Bubo virginianus)and Snowy Owl (B. scandiaca), voles at higher elevations (Corn and Bury 1986). while capable of taking large prey, feed primarily It is also possiblethat terrestrial specieslike the on volesand other small prey in many areaswhere red-backedvole are more difficult for SpottedOwls they occur (Watson1970, Cromich et al. 2002). In to capture in the Coast Rangesthan in the Cas- SEPTEMBER 2004 SPOTTED OWL DIET 223 cades because of the dense brush that covers the to surviveon such a low capture rate becausemany ground in many areasin the Coast Ranges. of the prey they capture are squirrels,woodrats, or Timing of Foraging.In our study,and all previ- lagomorphsthat can be stored and eaten in several ous studiesof Spotted Owls, the diet wasdominat- meals spanning 2 or more d (Forsman et al. 1984). ed by nocturnal animals, indicating that Spotted Prey Selection. Shrews, western red-backed Owls forage primarily at night (e.g., Laymon 1988, voles, deer mice, and jumping mice are abundant Cutler and Hays 1991, Ward 1990, Verner et al. mammals in many forest typesin the Pacific North- 1992, Ward and Block 1995). However, it has been west (e.g., Aubry et al. 1991, Corn and Bury 1991, well documented that Spotted Owls do forage dur- West 1991, Rosenberg et al. 2003, Gomez and An- ing the day, especiallyif they are feeding fledged thony 1998). In most areas they are much more young (Miller 1974, Laymon 1988, Sovern et al. numerous than the flying squirrels,woodrats, lago- 1994). Laymon (1988) even suggestedthat Spotted morphs, gophers,and tree volesthat form the core Owlswith fledged young traveledconsiderable dis- diet of northern Spotted Owls (Carey et al. 1992, tancesaway from their roost areasto forage during Rosenberg and Anthony 1992, Ward et al. 1998). the day. However,Forsman et al. (1984), and Sov- It is tempting to conclude, therefore, that northern et al. (1994) found that Spotted Owls moved ern Spotted Owls feed selectivelyon certain kinds very little during the day and that most diurnal of mammals, especiallylarge mammals that are ar- foraging involved opportunistic attempts to cap- boreal or scansorial (Barrows 1980, Forsman et al. ture prey near day roosts.Some diurnal prey were 1984, Verner et al. 1992). Ward et al. (1998) pre- probablyalso capturedwhen they were discovered sented evidencethat Spotted Owls in northern Cal- at night, or when owls were foraging at dawn or ifornia were preying selectivelyon large prey, es- dusk. pecially woodrats. Ward et al. (1998) further SeasonalVariation. Predation by Spotted Owls suggestedthat the energetic reward per unit effort on large mammals and birds wasrestricted primar- was higher for large prey like woodrats than for ily to the spring and summerwhen large numbers smaller prey, and that Spotted Owls were foraging of small, naivejuveniles were available(Forsman et preferentially in areas where woodratswere abun- al. 1984, 2001, this study). This suggeststhat adult dant. However, an alternative hypothesisis that the snowshoehare, mountain beaver, and grouse are owls are not selectingfor certain kinds of prey, but difficult for Spotted Owls to capture. Seasonalpre- are simply preying opportunistically on prey that dation on juvenile hare and rabbits has been doc- are easiestfor them to capture, given the particular umented in many other owls, including Tawny morphological attributes of the owl and the struc- Owls (St,ix aluco;Southern 1970), Northern Hawk tural attributes of the dense forestsin which they Owls (Surnia ulula; Rohner et al. 1995), Great Gray live. Experimental tests of these hypotheseshave Owls (Strix nebulosa; Mikkola 1983), Barn Owls not been conducted, but it is obviousthat Spotted (Tytoalba; Marti 1988), Long-earedOwls (Asiootus; Owls in the Pacific Northwest rely on a few species Marti 1976), and SnowyOwls (Watson1970). Fors- of nocturnal mammals for the majority of their man et al. (1984) found that diets of northern food, and that forest management practices that Spotted Owls during winter became increasingly produce healthy populations of these species dominated by arboreal mammalsas insects,birds, should benefit Spotted Owls. and juvenile Lagomorphs became less available, and many terrestrial small mammalseither became ACKNOWLEDGMENTS less active above the snow, or went into hiberna- This study would not have been possiblewithout the tion. cooperation and assistanceof many different biologists Numbers of Prey Captured. Basedon an analysis and technicians who helped collect, clean, and analyze pellets. In particular, we thank S. Andrews,D. Barrett, M. of the number of prey in individual pellets, Fors- Broyles,B. Casler,R. Claremont, A. Eldridge, A. Elling- man (1980) estimatedthat individual SpottedOwls son, R. Forson, L. Gangle, R. Horn, P. Loschl, G. Miller, in the central Cascadesof Oregon captured 0.7- R. Miller, J. Mires, C. Ogan, F. Oliver, I. Otto, J. Perkins, 1.05 prey/d during the fall, winter, and spring J. Reid, P. Shacklee,T. Snetsinger,B. Straub,J. Thrallkill, (September-April). These estimates,based on a E. Vorisek, and J. Zisa. S.C. Sovern wrote severalcom- puter programs that were enormously helpful in analyz- different method than we used in this study,were ing the data. S. Andrews, B. Glenn, P. Loschl, C. Ogan, nearly identical to our present estimateof 0.6-1.0 P. Ward, and an anonymous referee reviewed various prey captured/d. Northern Spotted Owls are able drafts of the manuscript. Funding and in-kind support 224 FORSMANET AL. VOL. 38, NO. 3 for this studywere provided by the Pacific Northwest and FORSMAN,E.D. 1980. Habitat utilization by SpottedOwls Pacific Southwest Research stations of the U.S. Forest Ser- in the west-centralGascades of Oregon. Ph.D. disser- wce, the Oregon State Office of the Bureau of Land Man- tation, Oregon State Univ., Gorvallis,OR U.S.A. agement, the Oregon Department of Forestry, the , E.G. MESLOW, AND H.M. WIGHT. 1984. Distribu- Oregon CooperativeFish and Wildlife ResearchUnit, the Department of Fisheries and Wildlife at Oregon State tion and biology of the Spotted Owl in Oregon. Wildl University, and the Forest and Range EcosystemScience Mono•. 87. Center of the U.S. Geological Survey. , S. DESTEFANO,M.G. RAPHAEL,AND RJ. GUTIERREZ (EDS.). 1996. Demographyof the northern Spotted LITERATURE CITED Owl. Stud. Avian Biol. No. 17. AUBR¾,K.B., MJ. CraTES,•ND S.D. WEST.1991. Regional , I.A. OTTO, S.G. SOWraN, M. TAYLOR,D.W. HAYS, patternsof smallmammal abundanceand community H. ALLEN,S.L. ROBERTS,AND D.E. SEAMAN.2001. Spa- composition in Oregon and Washington. Pages285- tial and temporal variation in diets of Spotted Owls 294 in L.F. Ruggiero, K.B. Aubry, A.B. Carey, and in Washington. J. Raptot'Res'. 35:141-150. M.H. Huff lEDS.], WildliIb and vegetation of unman- FRANKLIN,J.F. AND C.T. DYRNESS.1973. Natural vegetation aged Douglas-fir forests. USDA For. Serv. Gem Tech. of Oregon and Washington. USDA For. Serv. Gen Rep. PNW-GTR-285, Pacific Northwest Research Sta- Tech. Rep. PNW-8, Pacific Northwest Research Sta- tion, Portland, OR U.S.A. tion, Portland, OR U.S.A. BARROWS,C.W. 1980. Feeding ecologyof the Spotted Owl FRANKLIN,A.B., D.R. ANDERSON,RJ. GUTII•RREZ,AND K.P. in Cali•brnia. RaptorRes'. 14:73-77. BURNHAM.2000. Climate, habitat quality, and fitness BUI,L,E.L. ANDM.G. HENJUM.1990. Ecologyof the Great in northern Spotted Owl populationsin northwestern Gray Owl. USDA For. Serv. Gem Tech. Rep. PNW- California.Ecol. Morto,k. 70:539-590. GTR-265, Pacific Northwest Research Station, Port- GANEY,J.L. 1992. Food habits of Mexican Spotted Owls land, OR U.S.A. in Arizona. Wilson Bull. 104:321-326. BURNHAM, K.P., D.R. ANDERSON, ANt) G.C. WHITE. 1994. GOMEZ, D.M. ANt) R.G. ANTHONY. 1998. Small mammal Estimation of vital rates of the northern Spotted Owl. abundancein riparian and upland areasof five seral Pages 1-44 (Appendix J) in Final supplemental envi- stagesin western Oregon. NorthwestSci. 72:293-302. ronmental impact statementon managementof hab- GOULD,G.I., JR. 1977. Distributionof SpottedOwls in itat for late-successionaland old-growthforest related California. Western Birds 8:131-146. specieswithin the range of the northern Spotted Owl. GP,agER, R.R. 1962. Food and oxygen consumptionin Vol. 2. USDA Forest Service, Portland, OR U.S.A. three speciesof owls (Strigidae).Condor 64:473-487. CAREY,A.B., S.P. HORTON, AND B.L. BISWELL.1992. North- HAMER, T.E., D.L. HAYS, C.M. SENGER,AND E.D. FORSMAN ern Spotted Owls: influence of prey base and land- 2001. Diets of northern Barred Owls and northern scape character. Ecol.Monogr 62:223-250. Spotted Owls in an area of sympatry.J. RaptorRes'. 35 , C.C. M•GuIRE, B.L. BISWELL, AND T.M. WILSON. 221-227. 1999. Distribution and abundance of Neotoma in west- HILL, M.O. 1973. Diversityand evenness:a unifyingno- ern Oregon and Washington.Northwest Sci. 73:65-80. tation and its consequences.Ecology 54:427-432. CORN, P.S. ANB R.B. BURY. 1986. Habitat use and terres- LAYMON,S.A. 1988. The ecologyof the Spotted Owl in trial activity by red tree voles (Arborlinuslongicaudus) the central Sierra Nevada, California. Ph.D. disserta- in Oregon. J. Mammal.67:404-406. tion, Univ. California, Berkeley,CA U.S.A. -- AND --'. 1991. Small mammal communities in LINSt)ALE,J.M. ANt)L.P. TEVIS,JR. 1951. The dusky-footed the Oregon Coast Range. Pages241-254 in L.F. Rug- woodrat. A record of observations made on the Has- gicro, K.B. Aubry, A.B. Carey, and M.H. Huff [Et)s.], tings Natural History Reservation.University Califor- Wildlife and vegetationof unmanagedDouglas-fir for- nia Press,Berkeley, CA U.S.A. ests.USDA For. Serv. Gcn. Tech. Rep. PNW-GTR-285, LOVEjOY,B.P. 1972. The capture-recaptureanalysis of a Pacific Northwest Research Station, Portland, OR mountain beaver population in western Oregon U.S.A. Ph.D. dissertation,Oregon State Univ., Corvallis,OR CROMRICH, L.A., D.W. HOLT, AND S.M. LEASURE. 2002. U.S.A. Trophic niche of North American Great Horned MAUTi,C.D. 1976.A reviewof prey selectionby the Long- Owls.J. Raptot'Res. 36:58-65. eared Owl. Condor 78:331-336. CUTLER, T.L. AND D.W. HAYS. 1991. Food habits of north- ß 1988. A long-term study of food-niche dynamics ern Spotted Owls in high elevation forestsof Pelican in the common Barn Owl: comparisonswithin and Butte, southwesternOregon. Northwest.Nat. 72:66-69. between populations. Can.J. Zool.66:1803-1812. DUNGAN,R.B. ANDR. SIDNER.1990. Bats in Spotted Owl MASER,C. ANt)R.M. STORM.1970. A key to microtinaeoi pellets in southern Arizona. GreatBasin Nat. 50:197- the Pacific Northwest. Oregon State Univ. Bookstores, 200. Inc., Corvallis, OR U.S.A. DUNNING,J.B., JR. (ED.). 1993. GRG handbook of avian MItmOLa, H. 1983. Owls of Europe. Buteo Books, Vet- body masses.GRG Press,Boca Raton, FL U.S.A. million, SD U.S.A. SEPTEMBER2004 SPOTTED OWL DIET 225

MILLER,G.M. 1974. Grace M. Miller on Spotted Owls. central coast ranges. Pages432-448 in J.R. Duncan, Pages 130-150 in L.W. Walker [ED.], The book of D.H. Johnson, and T.H. Nicholls lEDs.], Biologyand owls.Alfred A. Knopf', Inc., New York, NY U.S.A. conservation of owls of the northern hemisphere. Mn.LEP,,G.S. 1989. Dispersalof juvenile northern Spotted USDA For. Serv.Gen. Tech. Rep. NC-190, North Cen- Owls in western Oregon. M.S. thesis, Oregon State tral Experiment Station, St. Paul, MN U.S.A. Univ., Corvallis, OR U.S.A. VERNER,J., R.J. GUTIERREZ,AND G.I. GOULD,JR.1992. The --, RJ. SMALL,AND E.C. MESLOW.1997. Habitat se- California Spotted Owl: general biology and ecologi- lection by Spotted Owls during natal dispersalin west- cal relations.Pages 55-77 inJ. Verner, K.S. McKelvey, ern Oregon. J. Wildl. Manage.61:140-150. B.R. Noon, RJ. Gutitrrez, G.I. Gould, Jr., and T W ROHNER, C., J.N.M. SMITH, J. STP`OMAN,M. JOYCE,EI. Beck [EDS.], The California Spotted Owl: a techmcal DOM.E, AND R. BOONSTRA. 1995. Northern Hawk-Owls assessment of its current status. USDA For. Serv. Gen in the nearctic boreal forest; prey selectionand pop- Tech. Rep. PSW-GTR-133, Pacific SouthwestResearch ulation consequencesof' multiple prey cycles.Co'r•dor Station, Berkeley, GA U.S.A. 97:208-220. VERTS,BtJ. ANDL.N. CARP,AWAY. 1984. Keys to the mam- ROSENBERG, D.K. AND R.G. ANTHONY. 1992. Characteris- mals of Oregon. Oregon State Univ. Bookstores,Inc, tics of northern flying squirrel populations in young, Corvallis, OR U.S.A. second, and old-growth forests in western Oregon. --AND --. 1998. Land mammals of Oregon. Can.J. Zool.70:161-166. Univ. California Press,Berkeley, CA U.S.A. , K.A. SWINDLE, AND R.G. ANTHONY. 2003. Influ- WARD,J.P., JR. 1990. Spotted Owl reproduction, diet and ence of prey abundance on northern Spotted Owl re- prey abundance in northwest Califbrnia. M.S. thes•s, productive successin western Oregon. Can.J. Zool.81: Humboldt State Univ., Arcata, CA U.S.A. 1715-1725. --AND W.M. BLOCK. 1995. Mexican Spotted Owl SMITH,D.C. ANDJ.R. MURPHY.1973. Breedingecology of prey ecology.In USDI Fish and Wildl. Serv.,Recovery raptors in the eastern Great Basin of Utah. Brigham plan for the Mexican Spotted Owl Vol. 2. USDI F•sh YoungUniv. St. Bull, Biol. Series18:69-76. and Wildl. Serv., Albuquerque, NM U.S.A. SMITH, R.B., M.Z. PEEP,Y, R.T. GUT•P•RREZ,AND W.S. LA- --, RJ. GUTII•RREZ,AND B.R. NOON. 1998. Habitat HAYE. 1999. The relationship between Spotted Owl selection by northern Spotted Owls: the consequenc- diet and reproductive successin the San Bernardino es of prey selectionand distribution. Condor100:79- Mountains, California. Wilson Bull. 111:22-29. 92. SOKAL,R.R. AND FtJ.ROHLF. 1969. Biometry. W.H. Free- WATSON,A. 1970.The behavior,breeding, and food ecol- man and Company, San Francisco,CA U.S.A. ogy of the Snowy Owl (Nycteascandiaca). Ibis 99:419- SOUTHERN,H.N. 1970. The natural control of a popula- 462. tion of Tawny Owls Strixaluco..]. Zool. 162:197-285. WEATHERS,W.W., PJ. Ho•)un, AN•)J.A. B•a4ESLEY.2001. SOVEP,N, S.C., E.D. FORSMAN,B.L. BISWELL,D.N. ROLPH, Thermal ecologyand ecologicalenergetics of Cahfor- ANDM. TAYI.OR.1994. Diurnal behaviorof the Spotted nia Spotted Owls. Condor103:678-690. Owl in Washington. Condor96:200-202. WEST, S.D. 1991. Small mammal communities in the SPEVAK,T.A. 1983. Population changes in a Mediterra- southern Washington CascadeRange. Pages269-283 nean scrub rodent assemblyduring drought. South- in L.F. Ruggiero, K.B. Aubry, A.B. Carey, and M.H. west. Nat. 28:47-52. Huff [EDS.], WildliIb and vegetation of unmanaged THOMAS,J.W., E.D. FORSMAN,J.B. L•N7, E.C. MESLOW, Douglas-firforests. USDA For. Serv. Gen. Tech. Rep B.R. NOON,AND J. VERNER.1990. A conservationstrat- PNW-GTR-285, Pacific Northwest Research Stanon, egy for the northern Spotted Owl: report of the In- Portland, OR U.S.A. teragency Scientific Committee to addressthe conser- ZABEL,CJ., K. MCKELVEY,AND J.P. W^Rr),JR. 1995. Influ- vation of the northern Spotted Owl. USDA Forest ence of primary prey on home-rangesize and habitat- Serviceand USDI Bureau of Land Management, Port- use patterns of northern Spotted Owls (Strix occzdez&- land, OR U.S.A. taliscaurina). Can. J. Zool.73:433-439. THP•adLKILL,J.A., R.G. ANTHONY,AND E.C. MESLOW.1997. An update on demographic estimatesfor northern Received 4 April 2003; accepted May 2004 SpottedOwls (Strixoccidentalis caurina) from Oregon's AssociateEditor: James R. Belthoff 226 FORSMAN ET AL. VOL. 38, NO. 3

Appendix. Species,common names, mean mass,activity codes, and total number of prey identified in pelletsof northern SpottedOwls in Oregon, 1970-2003.

SPECIES MEAN MASS(G) a ACTIVITYCODE b N

MAMMALS 22 421 Soricidae Sorex bendirii•Pacific water shrew 18 (1) N, T 7 Sorexmonticolus--dusky shrew 6 (1) N, T 1 Sorexbairdii (obscurus)--Baird'sshrew 9 (1) N, T 1 Sorexpalustris--water shrew 12 (1) N, T 1 Sorextrowbridgii•Trowbridge's shrew 5 (1) N, T 39 Sorexvagrans--vagrant shrew 5 (1) N, T 8 Sorexpacificus (S. yaquinae)--Pacificshrew 7 (l) N, T 17 Sorexspp.--unidentified shrew 7 (4) N, T 243 Talpidae Neurotrichusgibbsii--shrew-mole 9 (• N, T 68 Scapanuslatimanus--broad-footed mole 69 (1 N,T 7 ScapanusoraNus--coast mole 56 (1) N, T 98 Scapanusspp.--unidentified mole 56 (4) N, T 66 Chiroptera Myotislucifugus--little brown myotis 6 (1) N, F 2 Myotisyumanensis--Yuma myotis 6 (1) N, F 1 Myotisspp.--unidentified myotis 6 (4) N, F 3 Lasionycterisnoctivagans--silver-haired bat 11 (1) N, F 5 Eptesicusfuscus--big brown bat 15 (1) N, F 7 Unidentified bat 10 (4) N, F 49

Ochotonidae Ochotonaprinceps American pika 171 (1) B, T 85 Leporidae Sylvilagusbachmani•brush rabbit 50-750 (3) B, T 224 Lepusamericanus--snowshoe hare 50-1400 (3) B, T 528 Unidentified rabbit/hare 50-900 (3) B, T 258 Aplodontidae Aplodontiarufa--mountain beaver 100-550 (3) N, T 56 Sciuridae Tamiasamoenus--yellow-pine chipmunk 48 (1) D, S 2 TamiastownsendiiC--Townsend's chipmunk 83 (1) D, S 203 Tamiassenex--Allen's chipmunk 85 (1) D, S 21 2i•miasspp.--unidentified chipmunk 83 (4) D, S 26 Spermophilusbeecheyi--California ground squirrel 626 (1) D, T 2 S. lateralis--golden-mantledground squirrel 169 (1) D, T 1 Sciurusgr/seus--western gray squirrel 60-795 (3) D, A 15 Tamiasciurusdouglasii•Douglas' squirrel 221 (1) D, A 103 Glaucomyssabrinus--northern flying squirrel 130 (1) N, A 8826 Unidentified Sciurid spp. 130 (4) U, S 17 Geomyidae Thomomysbottae•Botta's pocket gopher 142 (1) N, T 2 Thomomysmazama--western pocket gopher 95 (1) N, T 435 Thomomystalpoides--northern pocket gopher 102 (1) N, T 7 Thomomysspp.--unidentified gopher 95 (4) N, T 145 Muridae--Sigmodontinae Peromyscusmaniculatus--deer mouse 22 (1) N, T 1830 Neotomacinerea•bushy-tailed woodrat 284 (1) N, S 512 SEPTEMBER2004 SPOTTED OWL DIET 227

Appendix. Continued.

SPECIES MEAN MASS(G) a ACT•ITY CODEb N

Neotomafuscipes•dusky-footed woodrat 286 (1) N, S 529 Neotomaspp.--unidentified woodrat 285 (4) N, S 2258

Muridae--Murinae Rattus rattus---black rat 250 (3) N, T 5

Muridae--Arvicolinae Clethrionomyscalifornicus•western red-backed vole 23 (1) N,T 1491 Phenacomysintermedius•heather vole 28 (1) N,T 10 Arborimusalbipes•white-footed vole 23 NT 5 Arborimuslongicaudus--red tree vole 26 (1) NA 2954 Microtuscalifornicus--California vole 43 (1) NT 19 Microtuscanicaudus•gray-tailed vole 36 (1) NT 6 Mictvtuslongicaudus•long-tailed vole 56 (1) NT 28 Microtusotegoni--creeping vole 20 NT 187 Microtus richardsoni--Richardson's vole 70 (1) N T 45 Microtus townsendii--Townsend' s vole 54 (•) N T 25 Microtusspp.--unidentified vole 30 (4) NT 143 Muridae spp.--unidentified vole/mouse 25 (4) U U 616 Dipodidae Zapustrinotatus--Pacific jumping mouse 24 (1) N,T 54 Zapusprinceps--western jumping mouse 30 (•) N,T 2 Procyonidae Bassariscusastutus•ringtail 200-400 (3) N,S Mustelidae Mustela erminea--ermin e 55 (1) N,T 43 Mustelafrenata--long-tailed weasel 110 (3) N,T 2 Mephitidae Spilogalegracilis=--spotted skunk 500 (3) N,T 1 Unidentified mammals 30-400 (3) U,U 67

BIRDS 1042

Anatidae Aix sponsa--WoodDuck 658 (2) D,F Anatidae spp.--unidentified duck 658 (4) D,F

Phasianidae Bonasa umbellus•Ruffed Grouse 100-576 (3) D,F 11 Dendragapusobscurus•Blue Grouse 500-890 (3) D,F 2 Grouse spp. 100-600 (3) D,F 7 Odontophoridae Oreortyxpictus•Moun tain Quail 224 (2) D,F 10 Columbidae Patagioenasfasdata--Band-tailed Pigeon 392 (2) D,F Strigidae Asiootus--Long-eared Owl 262 (2) N,F 1 Strix occidentalis•SpottedOwl 610 (2) N,F 3 Megascopskennicottii--Western Screech Owl 169 (2) N,F 59 Glaucidiumgnoma--Northern PygmyOwl 68 (2) N,F 21 Aegoliusacadicus--Northern Saw-whet Owl 83 (2) N,F 44 Strigidaespp.--unidentified small owl 124 (4) N,F 4 228 FORSMANET AL. VOL. 38, NO. 3

Appendix. Continued.

SPECIES MEAN MASS (G) a ACTIVITYCODE b N

Picidae Colaptesauratus--Northern Flicker 142 (2) D F 30 Sphyrapicustuber--Red-breasted Sapsucker 49 (2) D F 22 Picoidesalbolarvatus-•White-headed Woodpecker 61 (2) DF 1 Picoidesarcticus•Black-backed Woodpecker 69 (2) D F 1 Picoidespubescens--Downy Woodpecker 5O (2) DF 1 Picoidesvillosus--Hairy Woodpecker 66 (2) D F 9 Dryocopuspileatus--Pileated Woodpecker 287 (2) D F 6 Picidae spp.--unidentified Woodpecker 106 (4) D F 2 Tyrannidae Contopuscooperi--Olive-sided Flycatcher 32 (2) D,F Corvidae Cyanocittastelleri--Steller's Jay 128 (2) D,F 83 Perisoreuscanadensis--Gray Jay 73 (2) D,F 12 Paridae Poecilerufescens•Chestn ut-backed Chickadee 10 (2) D,F Certhiidae Certhiaamericana--Brown Creeper 8 (2) D,F Sittidae Sitta canadensis--Red-breasted Nuthatch 10 (2) D,F Troglodytidae Troglodytestroglodytes--Winter Wren 9 (2) D,F Regulidae Re,ffuluscalendula--Ruby-crowned Kinglet 7 (2) D,F Regulussatrapa--Golden-crowned Kinglet 6 (2) D,F

Turdidae Catharus ustulatus Swainsoh's Thrush 31 (2) D,F 2 Catharusguttatus•Hermit Thrush 31 (2) D,F 5 Ixoreus naevius•Varied Thrush 78 (2) D,F 19 Turdusmigratorius--American Robin 77 (2) D,F 12 Turdus/Ixoreusspp.--thrush/robin spp. 78 (4) D,F 101 Bombycillidae Bombycillacedr•rrum--Cedar Waxwing 32 (2) D,F Parulidae Dendroica ocddentalis--Hermit Warbler (2) D,F Dendroicaspp.--unidcntificd warbler (4) D,F Thraupidae Piranga ludoviciana--WesternTanager 28 (2) D,F Emberizidae Pipilo maculatus--SpottedTowhee 40 (2) D,F Passerellailiaca Fox Sparrow 32 (2) D,F Juncohyemalis--Dark-eyed Junco •8 (2) D,F Cardinalidae Pheucticusmelanocephalus•Black-headed Grosbeak 42 (2) D,F SEPTEMBER 2004 SPOTTED OWL DIET 229

Appendix. Continued.

SPECIES MEAN MASS (G) • ACTIVITYCODE b N

Fringillidae Carpodacuspurpureus--Purple Finch 25 (2) D, F 1 Carpodacusspp.--unidentified finch 30 (4) D, F 1 Loxia curvirostra--Red Crossbill 36 (2) D, F 4 Carduelis tristis--American Goldfinch 13 (2) D, F 1 Coccothraustesvespertinus--Evening Grosbeak 64 (2) D, F 13 Unidentified birds Large bird 300-1000 (3) U,F 6 Medium-size bird 60-299 (3) U,F 185 Small bird 5-59 (3) D,F 304

AMPHIBIANS 7 Rana spp.--frog spp. 30 (6) B,T 7 REPTILES 12 Elgaria coerulea--northernalligator lizard 35 (6) D,T 1 Sceloporusoccidentalis--western fence lizard 10 (6) D,T 2 Thamnophisordinoides--northwestern garter snake 100 (6) D,T 1 Thamnophisspp.--garter snakespp. lOO (6) D, T 2 Unidentified snake 100 (4) D,T 6 MOLLUSCS--GASTROPODA (Terrestrial snails) 7 Haplotremavancouverense 7 (6) B, T 2 Monedeniafidelis 5 (6) B, T 2 Unidentified snail 5 (4) B,T 3

INSECTS 1005 Orthoptera--Tettigoniidae (Camel crickets) Cyphoderrismonstrosa--great grig 2.0 (5) N,S 412 Orthoptera--Blattidae Cryptocercuspunctulatus---wood-feeding cockroach •.o (5) U,T Parcoblattaspp. wood roach 1.0 (5) N,T Orthoptera spp. •.o (5) U,U Hemiptera--Pentatomidae (stink bugs) Chlorochroaspp. 0.3 (5) U,U Homoptera--Cicadidae (cicadas) Okanaganaspp. 0.4 (5) U,U Neuroptera--Corydalidae (nerve-wingedinsects) Dysmichohermesdisjunctus 2.0 (5) N,U Neuroptera--Raphidiidae (snakeflies) Raphidiidae spp. 0.2 (5) U,U Coleoptera--Cicindelidae (tiger beetles) Omuscalifornicus 0.5 (5) U,U Omusdejeani 0.5 (5) U,U Coleoptera--Carabidae (ground beetles) Pterostichusamethystinus 0.3 (5) U,U Pterostichus lama 0.3 (5) U,U Pterostichus neobrunneus 0.3 (5) U,U Pterostichusspp. 0.3 (5) U,U Scaphinotusspp. 0.3 (5) U,U 230 FORSMANET ^L. VOL. 38, NO. 3

Appendix. Continued.

SPECIES MEAN MASS (G) a ACTIVITYCODE b N

Coleoptera--Scarabaeidae (dung beetles) Bolboceras obesus 0.3 (5) U, U 1 Pleocoma dubitalis 0.3 (5) U, U 10 Coleoptera--Lucanidae (stag beetles) Ceruchus striatus 0.5 (5) U,U 1 Sinodendronrugosum 0.5 (5) U,U 4 Coleoptera--Buprestidae (metallic woodborers) Buprestisaurulenta 0.3 (5) U, U 1 Coleoptera--Elateridae (click beetles) Cteniceraspp. 0.3 (5) U,U 1 Coleoptera--Tenebrionidac (darkling beetles) Iphthimusserratus 0.5 (5) U,U 1 Helopsspp. 0.5 (5) U,U 1 Coleoptera--Cerambycidae(long-horned woodborers) Ergatesspiculatus--ponderous borer 3.0 (5) N,U 249 Prionuscalifornicus--giant root borer 2.0 (5) U,U 5 A cmaeopsproteus 0.5 (5) U,U 1 Centroderaspurca 0.5 (5) U,U 2 Plectruraspinicauda 0.5 (5) U,U 1 Coleptera--Curculionidae (weevils) Dyslobuslecontei 0.3 (5) u,u 3 Dyslotmsspp. 0.3 (5) u,u 10 Panscopusspp. 0.3 (5) u,u 1 Coleoptera spp.--unidentified beetles 0.3 (5) u,u 11 Lepidoptera--unidentified moths 0.5 (5) U, U 2 Hymenoptera--Formicidae (ants) Camponotusspp. 0.1 (5) U,U 12 Formicafusca 0.1 (5) U,U 1 Formicidae spp. 0.1 (5) U,U 1 Hymenoptera--Vespidae (hornets and yellowjackets) Dolichorespulamaculata 0.1 (5) u,u 1 Unidentified large insect 2.0 (5) U,U 21 Unidentified small insect 0.3 (5) U,U 218

CRUSTACEANS Pacifastacusleniuseulus•crayfish 20 (6) B, T 2 ARACHNIDA Uroctonusmordax--scorpion 3 (5) N, T 1 a Source of massestimate is in parentheses:1 = Verts and Carraway1998; 2 = Dunning 1993; 3 = massof each individualestimated based on comparisonwith referencespecimens of knownmass; 4 = mean of'all speciesin group, 5 = means basedon estimatesfrom similar speciesin this genusor group; 6 = estimatesfrom local specimens. bFirst letter indicatesprimary period of activity(D = diurnal,N = nocturnal,B = activeboth dayand night, U = unknown). Secondletter indicatesprimary area of activity(T = terrestrial,A = arboreal, S = scansorial,F = flying or aerial animal, U = unknown). cVerts and Carraway(1998) split the "Tamiastownsendii Complex" into three species,but we treatedall Tamiasspecies west of the crest of the Cascadesas one species,because we could not tell them apart basedon bone fragmentsin pellets.