Nesting Ecology of Burrowing Owls Occupying Black-Tailed Prairie Dog Towns in Southeastern Montana

j RaptorRes. 35 (4) :296-303 ¸ 2001 The Raptor Research Foundation, Inc.

NESTING ECOLOGY OF BURROWING OWLS OCCUPYING BLACK-TAILED PRAIRIE DOG TOWNS IN SOUTHEASTERN MONTANA

MARCO RESTANI 1 Fishand Wildli•[bProgram, Department of Biology,Montana State University, Bozeman, MT 59717 U.S.A.

LARRY R. RAu Bureauof Land Management,III GarryowenRoad, Miles'City, MT 59301 U.S.A.

DENNIS L. FLATH MontanaFish, Wildlife,and Parks,PO. Box 173220, Bozeman,MT 59717 U.S.A.

ABSTR^CT.--Detailedinvestigations of the relationshipbetween BurrowingOMs (Athenecunicularia) and black-tailedprairie dogs (C}n0my,ludovitianu,) are rare, but such information is necessaryto manage the population declinesof oMs reported throughout much of the western United States.In 1998 we studied nest-siteselection, productivity, and food habits of Burrowing Owls breeding on prairie dog towns in southeasternMontana. We located 13 breeding pairs, seven of which nested on private land. Nesting density(1 pair/110 ha) on prairie dog townswas low comparedto densitiesin other regions.Few habitat characteristicsdifibred between nest sitesand random points, but power in statisticaltests was low. Nesting densityand habitat use suggestedthe population of owlswas well below carrying capacity.Productivity was 2.6 young/pair. OMs fed on iuvertebrates(mainly grasshoppers and beetles),mammals (mice and voles), birds (blackbirdsand buntings),and amphibians(fi•ogs). Plague (Yersiniapestis), poison, and habitat con- versionhave fragmentedprairie dog habitat and potentiallythreaten owl persistencein our studyarea. KEYWORDS: BurrowingOwl; Athene cunicnlaria; blach-tailedprairie dom Cynomysludovicianus; plague,, Yersiniapestis; food habits;habitat selection; Montana.

Eco16giadel anidamiento de Bahos Cavadoresocupando poblados de perros de la pradera de cola negra en el sudestede Montana RtsSUmEN.--Investigacionesdetalladas de la relacitn entre BahosCavadores (Athene cunicularia) y perros de la pradera de cola negra (Cynomysludovicianus) son raros, pero tal informacitn es necesariapara manejar el descensode la poblacitn de bahos reportado en la mayorfa del occidente de los Estados Unidos.En 1998 nosotrosestudiamos la seleccitndc sitiosnido, productividad,y habimsalimenticios de Bahos Cavadoresreproducifindose en coloniasde perros de la pradera en el sudestede Montana. Local- izamos 13 parejasreproductoras, siete de las cualesanidaban en terrenos privados.I•a densidad de anidamiento (1 parcja/110 ha) en lospoblados de perrosde la praderaam bajaen comparacitna densidades de otras regioncs.Pocas caracteristicas del habitat difErian entre los sitMsnido y puntos al azar, pero el podcr de las pruebas estadfsticasrue bajo. La densidadde anidamiemo y el nso de habitat sugiereque la poblacitn de bdhos estababien por deb•oode la capacidadde carga.La productividadfi•e 2.6 jtvenes/ ]):u'tj:t.Los bahos se alimentaronde inverwl)•a(h•s (l n incipalmentesaltmmn•tcs y t,sc,u'al):tj(•s), mamffcros (ratonesy campafioles),aves (mirlos y verdetones),y antibios (ranas). I,a peste ( Yersiniape. qi.9, el veneno y la transformacitn del habitat ha fragmentadoel habitat de los perros dc la pradcra y potencialmente ha puesto bajo amen•a la persistenciade los bahos en nuestra area de esmdio. [Traduccitn de Victor Vanegasy (;•sar Marquez]

The recent finding that the petition to list the der the U.S. Endangered SpeciesAct is "warranted black-tailedprairie dog (Cynomysludovicianus) un- but precluded" drawsnational attention to the status and management of a declining speciessub- • Present address:Department of BiologT,Rocky Moun- jected to poisoning campaigns,recreational shoot- tain College, Billings, MT 59102 U.S.A. ing, and introduced plague (Yersinia pestis). E-mail address:testahim@rocky. erin Decreases of prairie dog populations and their

296 DECEMBER 2001 BURROWIN(; OWl, ECOI,OGY IN MONTANA 297

habitat are thought to be responsible for similar rolling to fiat (elevation 680-865 m). Vegetation was declines of closelyassociated species (Miller et al. dominated by grasses(Agropyron smithii) and shrubs (Ar- temisiatridentata and A. cana). Riparian areas supported 1994, Samson and ICuopf 1994), most notably the scatteredcottonwood (Populustremuloides) and willow (Sa- black-footed i•rret (Mustelanigripes) and mountain lix spp.), while open standsof ponderosapine (Pinuspon- plover (Charadrius montanus). Burrowing Owls derosa)and juniper (Juniperusscopulorum) dominated hilly (Athene cunicularia) in the Great Plains south of terrain. Climate was semi-arid. The study area was an Canada also rely on prairie dog habitat (Butts and even mixture of public (federal and state) and privatc land that supportedlivestock grazing. Recreational shoot- Lewis 1982, Plumpton and Lutz 1993a, Desmond ing of prairie dogsoccurred year round, but wasconcen- et al. 1995), and many statesreport recent declines trated during spring and early summer. m owl abundance (James and Ethier 1989, Marti Beginning in mid-May 1998, we used spotting scopes and Marks 1989,James and Espie 1997). (15-45x) and binoculars (10x) to survey prairie dog Although Burrowing Owls nest extensively in townsIbr Burrowing Owls.We made no attempt to search tbr owlsoff of prairic dog towns.Wc scanncdtowns from prairie dog burrows, few studies have reported a vehicle or on foot, concentrating efibrt in early morn- habitat characteristicsimportant in nest-site selec- ing (0500-1000 H) or late atternoon (1700-2200 H), the tion or factors influencing owl densitywithin prai- daytime periods when owls are most active and visible rie dog towns. In Colorado, differences between (Haug and Oliphant 1990). Presenceof territorial pairs, nest burrows and random burrows in surrounding whitewash,cast pellets, molted feathers,and prey remains were used to identify nest burrows. We used a GPS re- burrow density,town size, and distanceto road var- ceiver (Carmen XL 12) to plot the locations of nest bur- ied from year to year (Plumpton and Lutz 1993a); rows on USGS 7.5 min topographic maps. Individual however, owls favored areas with lower vegetation towns were visited repeatedly (every 2 wk) throughout than wasavailable at random on prairie dog towns. the field season(May-August) to minimize the possibility In Nebraska, owls nested in loose colonies within of overlooking secretive or non-breeding owls and to monitor nesting success. larger prairie dog towns, but spaced themselves At nest burrows (N = 13), we measured elevation randomly within smaller towns (Desmond et al. (nearest m with a GPS receiver) and percentage slope 1995). Density of prairie dog burrows did not af- with a clinometer. From the nest burrow, we used a tape fect spacingpatterns of nesting owls,and a positive to measure distance to the nearest active and inactive relationship existedbetween town sizeand number prairie dog burrows (+0.05 m), nearestedge of the prairie dog town (+0.5 m), and nearest road (+0.5 m). We of nesting pairs (Desmond and Savidge 1996). also counted active (presence of fresh diggingsand/or Plague, poisoning, and habitat conversion have scat) and inactive prairie dog burrows within a 30-m ra- reduced and fragmented prairie dog towns across dius of the nest burrow (0.28 ha circle) to index prairie the Great Plains, including Montana (Flath and dog activity (Biggens et al. 1993). Size of prairie dog Clark 1986), but how these processesaffect nest- towns was obtained from habitat mapping with a GPS receiver conducted from July-September 1996 (K. Wlt- site selection and population ecologyof Burrowing tenhagen,Jr. and D. Tribby unpubl. data). We useda GPS Owls remains unknown. In 1998 we initiated a receiver to measure distance (+50 m) from the nest bur- study in southeastern Montana to elucidate nest- row to the nearest neighboring nest burrow. site selection of Burrowing Owls occupyingblack- We also measured these same habitat variables at 13 tailed prairie dog towns. We also estimated pro- burrowsselected haphazardly from prairie dog townsnot occupied by nesting owls.We selectedburrows by divid- ductivity and quantified food habits. We selecteda ing randomly selectedprairie dog townsinto progressive- study area previouslymapped for prairie dogs be- ly smaller quadrants bisected by the cardinal directions cause presence/absenceof Burrowing Owls had (numbered 1-4, chosen using a random numbers table). been recorded during visits (R. Richardson, D. The number of quadrants required to narrow down to a Tribby, K. Wittenhagen, Jr. unpubl. data). Thus, single potential nest burrow depended upon the size of some data were available to determine the popu- the prairie dog town. We picked only thoseburrows with openings large enough tbr nesting owls. lation trend of owls. Terminology describing reproductive successand pro- ductivityof Burrowing Owls tbllowed Steenhof (1987). STUDY AREA AND METHODS successthlpairs fledged at least one young, and produc- The study area in southeasternMontana (Custer and tivity estimatesincluded both successfuland thiled pairs Prairie counties;46ø44'N, 105ø38'W) encompassedap- We assumedevery owl pair attempted to breed (i.e., laid proximately400 km2, of which 1425 ha (3.6%) was oc- eggs). Multiple visits (10-20) to individual nest sites cupied by black-tailedprairie dogs.We surveyedthe prai- throughout the breeding seasonpermitted accurate de- rie dog complex within the Custer and Harris Greek termination of nesting success(young fledged per pmr). watersheds,areas being considered for black-footed fer- We estimated nesting chronology from age of emerged ret reintroduction. The badland topography was gently youngbased on plumage (Priest1997), assumingan in- 298 BIOLOGY VOL. 35, NO. 4

u) 35 We log transformed data prior to analyses(SPSS 1998) to achieve normal distributions (Kolmogorov-Smirnov • • Occupiedtest) and homogeneityof variances(Levene's test). How- • • Unoccupied_ ever, we have presented untransformed data (x ñ SE) in 0 25 this paper to facilitate interpretation. Becauseof the rel- atively small number of nesting pairs (N = 13) and con- • 20 cerns of low statisticalpower, we opted to reduce Type II errors by assigningstatistical significance at P <- 0.10 •- 15 when comparinghabitat variablesbetween occupied and random sites. • lO

RESULTS • 5 z o Prairie dog towns on our study area averaged

0-9 30-39 60-69 90-99 120-129 150-159 180-189 19.5 _+ 3.6 ha (range : 0.4-198.3 ha, N = 73). Size of prairie dog town (ha) Most occupied prairie dog habitat surveyedwas on private land (65%), followedby federal (30%) and F•gure 1. Size (ha) of black-tailedprairie dog townsoc- state (5%) lands. Prairie dog townsaveraged 11.0 cupied and unoccupiedby Burrowing Owls in southeast- _+1.9 ha on privatelands (N = 30) and 17.3 _+5.3 ern Montana, 1998. ha on public lands (N = 19; t = 0.64, df = 47, P : 0.53). cubation period (first egg to first hatch) of 30 d and BurrowingOwls nested on 12 of 73 (16%) prai- fledging at 40 d (Haug et al. 1993). rie dog towns that we surveyedin 1998. We found We collectedpellets and prey remainsopportunistically from May-August while visiting nest sitesand surround- 13 breeding pairs of Burrowing Owls on ca. 1425 mg perching and feeding areas. Entomologistsat Mon- ha (1 pair/110 ha) of prairie dog townswithin the tana State University, Bozeman, used a dissectingscope 400 km2 studyarea. No singleadult owlswere ob- (6.4-10X) to sort and identify invertebrate remains to served. Size of prairie dog towns did not diffkr be- family, and relied on museum specimensand Borror et al. (1989) tbr classificationto genus. To save space, we tween towns occupied by owls and townsunoccu- have presented invertebrate taxa to only the family level. pied by owls (t = 1.24, df = 71, P = 0.22; Fig. 1). We checkedremains of vertebrateprey againstpellet con- Burrowing Owls neither prefkrred nor avoided tents collected during subsequentvisits to the samenests nesting on prairie dog towns subjected to recrea- to minimize duplication.We used a dissectingscope (7- 30X) to identify vertebrate prey, and relied on museum tional shooting (X• = 0.00, df = 1, P = 1.00) or to specimensand Hoffman and Pattie (1968) for classifica- grazing (X2 = 0.16, df = 1, P = 0.69). Sevenpairs uon to species.Number of both invertebrate and verte- nested on private land, with three pairs each on brate prey were determined conservativelyby presence f•deral and state land. of diagnosticbody parts (e.g., legs,mandibles, skulls). We Most habitat characteristics did not diffbr fbr oc- calculated percentage biomassusing estimatesfrom Mar- t• (1974), Rodriguez-Estrella(1997), and museumspeci- cupied Burrowing Owl nest sites and random mens. points (Table 1). Occupied nestswere closerto ac-

Table 1. Itabitat characteristics(x ñ SE) of Burrowing Owl nest sites(N = 13) and random sites (N = 13) on black- tailed prairie dog townsin southeasternMontana, 1998.

OCCUI'II'5I) RANI)()M HABITAT VARIABLE SITE SITE t P

Elevation (m) 749 ñ 51 752 ñ 58 0.15 0.88 Percentageslope 2.6 ñ 0.5 2.3 ñ 0.4 0.27 0.79 Nearest active burrow (m) 14.6 -+ 7.1 21.8 ñ 6.4 1.81 0.08 Nearest inactive burrow (m) 6.7 ñ 1.9 7.7 ñ 2.8 0.68 0.50 Number of active burrows l l ñ 2 9 ñ 2 0.74 0.47 Number of inactive burrows 32 ñ 3 30 ñ 3 0.56 0.58 D•stanceto town edge (m) 111 ñ 36 73 ñ 17 0.85 0.40 Town size (ha) 27.3 ñ 10.1 25.4 ñ 7.1 0.96 0.35 Nearest road (m) 227 ñ 98 280 ñ 110 1.29 0.21 Nearest neighbor (kin) 2.2 ñ 0.5 3.3 -+ 0.7 1.28 0.21 DECEMBER 2001 BURROWING OWL ECOLOGY IN MONTANA 299

Table 2. Prey of Burrowing Owls based on remains Table 2. Continued. fbund at nest and perch siteson black-tailedprairie dog

townsin southeasternMontana, 1998. Prey are expressed PER- PER-

m number of items (N), percentagefrequency, and per- CENT CENT centage biomass. Unidentified items were not included FRE- BIO- m biomass estimates. Invertebrates were identified to TAXON N QUENCY MASS family, vertebratesto genus or species. Aves PER- PER- Sturnellaneglecta 18 1 25 CENT CENT Calamospizamelanocorys 22 2 11 FRE- BIO- Undetermined 1 <1 TAXON N QUENCY MASS Mammalia

Chilopoda Spermophilusrichardsonii 6 • 1 18 Scolopendromorpha <1 <1 Perognathusfasciatus 4 <1 <1 Peromyscusspp. 36 3 11 Arachnida Onychomysleucogaster 2 <1 1 Scorpiones < 1 < 1 Microtusspp. 12 1 8 Araneae < 1 < 1 Zapushudsonius 2 < 1 < 1 Non-insectarthropod <1 <1 Unknown rodents 16 1

Insecta Total 1202 100 100 Odonata Family undetermined 1 <1 Orthoptera tive prairie dog burrowsthan were random points. Acrididae 311 26 9 Neither number of active prairie dog burrows nor Gryllacrididae 4 <1 <1 total number of burrows (inactive + active) cor- Gryllidae 8 <1 <1 relatedwith town size (P > 0.30, N = 26). Statis- Hemiptera tical power was 0.35 for each of the two contrasts Belostomatidae 3 <1 <1 with low and nonsignificantP-values (i.e., nearest Reduviidae 1 <1 <1 roadand nearest neighbor). Goleoptera Burrowing Owls produced 2.6 + 0.4 young/pair Garabidae 337 28 3 (• + SE,N = 13pairs). Twelve pairs (92%) each Silphidae 72 6 <1 producedatleast one fledgling. One pair failed for Hydrophilidae 3 <1 <1 Histeridae I <1 <1 unknown reasons. Productivity did not correlate Scarabaeidae 127 11 <1 with number of active or inactive prairie dog bur- Elateridae 1 <1 <1 rows (P > 0.30, N = 13) within a 30-m radius of Tenebrionidae 81 7 <1 the nest. Productivity also did not correlate with Meloidae 1 <1 <1 size of prairie dog towns (P = 0.57, N = 13). Owls Gerambycidae 42 3 <1 nesting on prairie dog towns subjected to recrea- Ghrysomelidae 5 <1 <1 tional shooting(N -- 6) had productivitysimilar to Gurculionidae 12 1 <1 thosenesting pairs not exposedto shooting(N = Diptera 7) (2.3 young/pair versus2.9 young/pair, respec- Asilidae 1 <1 <1 tively;t = 0.65,df = 1, P = 0.54). Bybackdating Muscoidea 1 <1 <1 fromyoung of knownage (N = 7 nests),we esti- Lepidoptera mated a • layingdate of 20 May (+_1 d), • hatching Sphingidae 1 <1 <1 dateof 19June (_+1 d), and• fledgingdate of 29 Hymenoptera July (_ 1 d). Sphecidae 7 <1 <1 We identified 1053 invertebrateand 149 verte- Eumenidae Formicidae 164 <11 <1 brateprey remains (Table 2). Themost common invertebrate prey were grasshoppers (Orthopo- Undetermined Hymenoptera 2 tera) and beetles(Goleoptera). Amphibian prey in- Amphibia cluded northern leopard frogs (Rana pipiens)and liana pipiens 28 2 10 plainsspadefoot (Scaphiopus bombifivns). Only two Scaphiopusbombifrons 2 <1 <1 speciesof birdswere taken: Lark Bunting (Cala- 300 BIOLOGY VOL. 35, NO. 4 mospiza melanocorys)and Western Meadowlark cupied towns in Nebraska (Desmond et al. 1995). (Sturnellaneglecta). Mice (Peromyscusspp.) and voles Burrow densities of prairie dog towns in south- (Microtusspp.) were the most common mammalian eastern Montana and Colorado (Plumpton and prey. Most important prey items in terms of bio- Lutz 1993a) were similar, but were 3X higher than mass were meadowlarks (25%), mice + voles in Nebraska (Desmond et al. 1995). Therefore, (20%), and Richardson's ground squirrels ($per- prairie dog towns occupied by nesting owls in mophilusrichardsonii; 18%). southeasternMontana were relativelysmall and active, habitat conditions that should have minimized DISCUSSION the probability of badger predation. Badger pre- Burrowing Owl Use of Prairie Dog Towns. Two dation of owls did not occur during our study,sup- observationssuggested the Burrowing Owl popu- porting this hypothesis. lation was well below carrying capacityof nesting Historically,Burrowing Owl occupancyof prairie habitat on black-tailedprairie dog townswithin our dog townson our studyarea washighest in 1978- study area. First, density of Burrowing Owls (1 79 (27%, 15 of 55 towns;C. Knowlespers. comm.), pair/110 ha) was low compared to densities in intermediate in 1991 (14%, nine of 66 towns; R. Oklahoma (1 pair/0.19 ha, Butts 1971), and / Richardsonand D. Tribby unpubl. data), and low- nearest-neighbordistance on our study area (2.2 est in 1996 (4%, three of 73 towns;K. Wittenhag- kin) greatly exceeded that in Nebraska (0.11-0.13 en, Jr. and D. Tribby unpubl. data). We recorded kin, Desmond and Savidge1996). In fact, only one an increaseto 16% (12 of 73 towns)occupancy in prairie dog town supported more than one pair of 1998. Fluctuating population size of Burrowing owls. Second, the habitat characteristics we mea- Owls over the past 20 yr may have reflected the sured did not differ between nest sites and random impact of plague. Plague was first confirmed on points. Prairie dog townsunoccupied by owlswere our study area in 1986, and by the late 1980s had vacant apparently for reasons other than habitat significantly reduced prairie dog populations in suitability,perhaps indicating an owl population in southeasternMontana (cJ. Knowlesunpubl. data). decline (Schmutz 1997). However, conclusions re- Owl occupancyshould lag behind fluctuatingprairie garding habitat suitability remain preliminary be- dog populations(which it did) if townsdecimated by causesome comparisonslacked adequatestatistical plague provide nestinghabitat for 3-4 yr before in- power to detect differences between occupied and activeburrows collapse or fill in with soil (Buttsand random sites. Lewis 1982, Desmond et al. 2000). The only habitat attribute that appeared to dif- In this study area, rodents and birds composed fer between nests and random points was distance most of the Burrowing Owl diet by percent bio- to the nearest burrow occupied by prairie dogs, mass,whereas insectsdominated percent frequen- which was lessfor nest sites.Why owlsnested near cy. Owls nesting on prairie dog townsin Colorado active prairie dog burrows remains unknown, but and Wyoming exhibited similar prey use (Marti two previously noted patterns imply an anti-pred- 1974, Thompson and Anderson 1988, Plumpton ator benefit. Owls nesting in areas of highest bur- and Lutz 1993b). Use of prey varied seasonally,as row densityin Nebraska suffered lessbadger (Tax- mammalian prey were most important to owlsearly •dea taxus) predation than did other nesting owls in the nesting period before insectsbecame avail- (Desmond et al. 2000). Badgerpredation on black- able (Marti 1974, Green and Anthony 1989, tailed prairie dogs also correlated positively with Schmutz et al. 1991). Owls appeared to forage for town size in Wyoming (Campbell and Clark 1981). mammals mostlyat night and concentratedon in- Density of prairie dog burrows did not correlate sectsduring daylight.In Saskatchewanhome range with town size in both Wyoming and southeastern size decreasedsignificantly once insectsbecame Montana. The relationships between badger pre- abundant (Haug and Oliphant 1990). dation and (1) burrow density and (2) town size Management Implications. Productivity and ju- imply that highest predation of owls should occur venile and adult survivorshipact in concert to de- on large townswith low burrow density,assuming termine the trend of Burrowing Owl populations badger predation on owls occurs under the same (James et al. 1997, Johnson 1997, Clayton and conditionsas predation on prairie dogs. Schmutz 1999). Some of the mechanisms that af- Prairie dog towns occupied by Burrowing Owls fect demographyincluded habitat availability,pre- 111 southeastern Montana were half the size of oc- dation, and food availability.Productivity and pop- DECEMBER 2001 BURROWING OWL ECOI,OGY IN MONTANA 301 ulation size of Burrowing Owls in southeastern by owls during the nesting season (e.g., mice and Montana during 1998 was low and we did not es- voles,meadowlarks, grasshoppers, and beetles). timate survivorship.Populations in neighboring Finally, managementto benefit Burrowing Owls Saskatchewanand Alberta with similar or higher should consider historically-basednegative atti- productivity showed significant annual decreases tudes toward prairie dogs because nesting owls over the past decade (Hjertaas 1997, Wellicome were evenlydistributed across both public and pri- 1997b). Our preliminary resultssuggested the owl vate lands. Many state agricultural agencies,includ- population in Montana may have increasedwithin ing Montana's, continue to consider prairie dogs the past 5 yr as prairie dogs rebounded from "vertebrate pests" requiring systematic"suppres- plague epizootics.However, future monitoring is sion" (Sections 7-22-2207 [6] and 80-7-1101 Mon- warranted because productivity and density were tana Code Annotated). The acrimonious debate both low, and becausesignificant owl declinescon- betweenagricultural and conservationinterests im- tinue nearby in Canada. pedes effectivewildlife managementßConservation Management of Burrowing Owls in southeastern of prairie dog habitat can only proceed through Montana must consider population ecology and partnershipsbetween private citizens and govern- habitat selectionof black-tailedprairie dogs.Man- ment (Samson and Knopf 1994, Holroyd et al. aging plague is the greatest challenge to prairie 2001). To address both economic and conserva- dog conservation,and has similar potential to chal- tion concerns, the Montana Prairie Dog Working lenge managementof Burrowing Owlsin the Great Group is developingand implementing a statewide Plains. Plague moved through prairie dog townsin conservationplan for black-tailedprairie dogs.In- southeastern Montana during the mid-1980s (G. centivesto maintain prairie dog habitat on private Knowles unpubl. data), and reduced prairie dog lands are an important component of the plan, as populations to a level where plans to reintroduce is the goal to maintain viable populations of asso- black-footed ferrets were halted. Whether size or ciated species,such as the Burrowing Owl. distribution of prairie dog townsinfluences epizo- ACKNOWLEDGMENTS otic severityor movement of plague remains un- We thank the following ranchers for graciouslygrant- known. However, because Burrowing Owls select ing accessto private lands: A. Haughian, M. Haughian, the best,not the largest,remaining habitatpatches P. Haughian, Q. Haughian, T. Haughian, and D. Reed C. Stuart and M. Ivie identified invertebrate prey re- (Butts and Lewis 1982, Warnock and James1997), mains, andJ. Rozdilskykindly loaned mammal specimens plague may severely reduce owl populations in from the Burke Museum, University of Washington, Se- Montana. attle. G. Holroyd, B. Olenick, J. Sidle, and T. Wellicome Burrowing Owls did not avoid nestingon prairie provided helpful commentson the manuscript.The U.S. Bureau of Land Management and Montana Fish, Wild- dog townssubjected to recreational shooting,and life, and Parks provided funding for this study in coop- productivity of pairs nesting on or off shooting ar- eration with Montana State University,Bozeman. eas was similar. Although owls have been shot in LITERATURE CITED other areas (Butts 1973), we found no evidence of shooting mortality in our area. Nonetheless, rec- BIGGrNS,D.E., B.J. MmLrR, L.R. HANEBURY,B. OAFd•EAF, reational shooting may have disrupted daytime for- A.H. F;mMrR, R. CRETE, AND A. Door). 1993. A tech- nique for evaluatingblack-looted ferret habitat. Pages aging by adults and thus produced subtlenegative 73-88 inJ.L. Oldemeyer,D.E. Biggens,and B.J.Miller effects.For example, owlsfed extensivelyon diur- [Er)s.], Proceedings of the symposiumon the man- nal prey (e.g., birds and grasshoppers)when the agemerit of prairie dog complexes for the reintroduc- food demand of owl broods in southeastern Mon- tion of the black-looted ferret. USDI Fish and Wildl. tana washighest, in mid- to late-July(see also Haug Serv., Biol. Rep. ]3, Washington, DC U.S.A. and Oliphant 1990). Food limits Burrowing Owl BORROR,D.J., C.A. TRIPLEHORN,AND N.E JOHNSON.1989. productivityduring the nestling stage (Wellicome An introduction to the study of insects,6th Ed. Saun- 1997a, 2000), so abovegroundcounts of juveniles ders Gollege Publ., Philadelphia, PA U.S.A. BUTTS,K.O. 1971. Observationson the ecologyof Bur- •n Montana would have underestimatednestling rowing Owls in western Oklahoma: a preliminary re- mortality if starvationhad occurred belowground. port. Proc.Okla. Acad. Sci.51:66-74. In addition to maintaining nesting habitat, re- ß 1973. Life history and habitat requirements of source managersmust ensure that grasslandsand Burrowing Owls in western Oklahoma. M.S. thesis, shrublandssupport the primary prey speciestaken Oklahoma State Univ., Stillwater, OK U.S.A. 302 BIOLOGY VOL. 35, No. 4

-- •ND J.C. LEw•s. 1982. The importance of prairie --AND T.J. ETHIER. 1989. Trends in the winter dis- dog towns to Burrowing Owls in Oklahoma. Proc. tribution and abundance of Burrowing Owls in North Okla. Acad. Sci. 62:46-52. America. Am. Birds 43:1224-1225. CAMPBELL,T.M., III •ND T.W. GLARK.1981. Golony char- --, T.J. ETHIER,AND M.K. TOUTLOFF.1997. Parame- acteristics and vertebrate associates of white-tailed and ters of a declining Burrowing Owl population in Sas- black-tailed prairie dogs in Wyoming. Am. Midi. Nat. katchewan.Pages 34-37 in J.L. Lincer and K. Steen- 105:269-276. hof [EDs.], The Burrowing Owl, its biology and GIAYrON,K.M. ANDJ.K. SGHMUTZ.1999. IS the decline of management including the proceedingsof the first Burrowing Owls Speotytocunicularia in prairie Ganada international Burrowing Owl symposium.J. Raptor linked to changes in Great Plains ecosystems?Bird Res. Report 9. Cons. Intl. 9:163-185. JOHNSON,B.S. 1997. Demographyand populationdynam- DESMOND,M.J. ANDJ.A. S^WDC,F•. 1996. Factors influenc- ics of the Burrowing Owl. Pages28-33 inJ.L. Lincer ing Burrowing Owl (Speotytocunicularia) nest densities and K. Steenhof [EDs.], The Burrowing Owl, its and numbers in western Nebraska. Am. Midi. Nat. 136: ology and management including the proceedingsof 143-148. the first international Burrowing OM symposium.J --, AND K.M. ESKRIDGE. 2000. Correlations Raptor Res. Report 9. between Burrowing Owl and black-tailedprairie dog MART•, G.D. 1974. Feeding ecology of four sympatric declines:a 7-year analysis.J. Wildl. Manage.64:1067- owls. Condor 76:45-61. 1075. --AND J.S. M2kmcs.1989. Medium-sized owls. Pages , --, AND T.F. SIEBERT.1995. Spatial patterns 124-133 in B.A. Giron-Pendleton [ED.], Proceedings of Burrowing Owl (Speotytocunicularia) nests within of the Western Raptor Management Symposium and black-tailedprairie dog (Cynomyslucovicianus) towns. Workshop. Nat. Wildl. Fed., Washington,DG U.S.A. Can.J. Zool.73:1375-1379. MILLER,B., G. CEBAI,LOS,AND R. READING.1994. The prai- FLATH, D.L. ^ND T.W. CLARK. 1986. Historic status of rie dog and biotic diversity.Cons•v. Biol. 8:677-681. black-looted ferret habitat in Montana. Great Basin PLUMPTON,D.L. AND R.S. LUTZ. 1993a. Nesting habitat Nat. Mere. 8:63-71. use by Burrowing Owls in Colorado.J. RaptorRes. 27. GREEN,G.A. AND R.G. ANTHONY.1989. Nesting success 175-179. and habitat relationshipsof Burrowing Owls in the -- AND--. 1993b. Prey selectionand food habits Columbia Basin,Oregon. Condor91:347-354. of Burrowing Owls in Colorado. GreatBasin Nat. 53' HAUG, E.A. AND L.W. OLIPHANT. 1990. Movements, activ- 299-304. ity patterns, and habitat use of Burrowing Owls in Sas- PPdEST,J.E. 1997. Age identification of nestling Burrow- katchewan.J. Wildl. Manage.54:27-35. ing Owls. Pages 125-127 inJ.L. Lincer and K. Steen- --, B.A. MILLSAP, AND M.S. MARTEI.L. 1993. Burrow- hof [EDS.], The Burrowing Owl, its biology and man- ing Owl (Speotytocunicularia). In A. Poole and F. Gill agement including the proceedings of the first [EDS.], The Birds of North America, No. 61. The the international Burrowing Owl symposium.J. Raptor Academy of Natural Sciences, Philadelphia, PA and Res. Report 9. the American Ornitholo•sts' Union, Washington,DC RODPdGUEZ-ESTRELLA,R. 1997. Nesting sitesand feeding U.S.A. habits of the Burrowing Owl in the BiosphereReserve HJEP,T•.AS, D.C. 1997. Operation Burrowing Owl in Sas- of Mapimi, Mexico. Pages 99-106 inJ.L. Lincer and katchewan.Pages 112-116 inJ.L. Lincer and K. Steen- K. Steenhof lEDs.], The Burrowing Owl, its biology hof lEDS.I, The Burrowing Owl, its biology and man- and management including the proceedings of the agement including the proceedings of the first first international Burrowing Owl symposium.J. Rap- international Burrowing Owl symposium.J. Raptor tor Res. Report 9. Res. Report 9. SPSS, INC. 1998. SPSS Base 8.0 for Windows. User's t{tWFMAN,R.S. ANI) D.L. PATTIE.1968. A guide to Mon- Guide, SPSS,Inc., Chicago, IL U.S.A. tana mammals: identification, habitat, distribution, SAMSON, F. AND F. KNOPF. 1994. Prairie conservation in arid abundance. Univ. Montana, Missoula, MT U.S.A. North America. BioScience 44:418-421. } IOLROYr), G.L., R. RODRK;UEZ-ESTRELLA,AND S.R. SnEE- SCnMUTZ,J.K. 1997. Selectedmicrohabitat variables near E•ELD.2001. Conservation of the Burrowing Owl in nests of Burrowing Owls compared to unoccupied western North America: issues,challenges, and rec- sites in Alberta. Pages 80-83 in J.L. Lincer and K ommendations.J. RaptorRes'. 35:399-407. Steenhof [EDs.], The Burrowing Owl, its biology and JAMES,P.C. ANDR.H.M. ESmE.1997. Current statusof the management including the proceedings of the first Burrowing Owl in North America. Pages 3-5 in J.L. international Burrowing Owl symposium.J. Raptor Lincer and K. Stcenhof [EDs.], The Burrowing Owl, Res. Report 9. its biology and management including the proceed- --, G. WOOD,^ND D. WOOD. 1991. Spring and sum- ings of the first international Burrowing Owl sympo- mer prey of Burrowing Owls in Alberta. BlueJay 49' sium. J. Raptor Res. Report 9. 93-97. DECEMBER 2001 BURROWING OWL ECOLOGY IN MONTANA 303

STEENHOF,K. 1987.Assessing raptor reproductive success WELI,[COME,T.I. 1997a. Reproductiveperformance of and productivity.Pages 157-170 in B.A. Giron-Pendle- BurrowingOwls ($)beotytocunicularia): effects of sup- ton, B.A. Millsap, K.W. Cline, and D.M. Bird [Eos.], plemental food. Pages 68-73 in J.L. Lincer and K Raptor management techniques manual. Nat. Wildl. Steenhof leos.], The Burrowing Owl, its biology and Fed., Washington, DC U.S.A. management including the proceedings of the first THOMPSON,C.D. ANDS.H. ANDERSON.1988. Foragingbe- international Burrowing Owl symposium.J. Raptor havior and food habits of Burrowing Owls in Wyo- Res. Report 9. ming. Prairie Nat. 20:23-28. ß1997b. Status of the BurrowingOwl (Speotyto•.un- W^RNOCK,R.G. ^ND P.C.J^MES. 1997. Habitat fragmen- icularia hypugaea)in Alberta. Alberta Environmental tation and Burrowing Owls (Speotytocunicula•a) in Protection,Wildl. Manage.Div., Wildl. StatusRep. No Saskatchewan.Pages 477-486 in J.R. Duncan, D.H. 11, Edmonton, AB Canada. Johnson and T.H. Nicholls [EDS.], Biology and con- --. 2000. Effectsof food on reproductionin Burrow- servation of owls of the northern hemisphere: 2rid ing Owls (Athenecunicularia) during three stagesof international symposium.USDA Gem Tech. Rep. NC- the breeding season.Ph.D. dissertation,Univ. Alberta, 190, St. Paul, MN U.S.A. Edmonton, AB Ganada.