j. Rapor Res. 35 (3) :207-213 2001 The Rapor Research Foundaon, nc. ANALYSS OF BALD EAGLE SPATAL USE OF LNEAR HABTAT ALAN R. HARMATA Fsh & Wldlfe Program, Deparmen of Ecology, Monana Sae Unversy, Bozeman, MT 59717 U.S.A. GEORGE J. MONTOPOL Deparmen of Mahemacs, Arzona Wesern College, Yuma, AZ 85364 U.S.A. AmSTP XCT.--Several echnques are avalable f)r areal analyss of anmal locaons bu few are applcable o hose ha use lnear (.e., rparan) habas. Bald Eagles (Halaeeus leucocephalus) ofen are assocaed wh rvers and concenrae perch ses near shorelnes. Dsrbuon of cumulave proporon of perches by dsance om mos recenly acve nes ses deermned by rado rackng were used o compare spaal use among fve adul Bald Eagles breedng along he Snake Rver, Wyomng. Spaal Use ndcaors (SUs) were developed from logsc regresson parameers n aemps o: (1) undersand and model underlyng processes from whch he daa may have emerged, (2) compare wh smple descrpve sascal echnques o evaluae uly for presenng a clear, accurae represenaon of spaal use dfferences among eagles, and, (3) relae measures of eagle spaal use wh long-erm producvy of breedng areas. Dsance ndcaor (D) was he dsance from he nes ncludng 50% of all deeced perches used by a rado-agged eagle and was represenave of he sze of he range. Slope ndcaor (S) was he slope of he fed logsc regresson curve a he D (nflecon pon). S was an ndcaor of lnear dsperson of perch ses whn he breedng area. Bald Eagles assocaed wh more producve (>0.77 young per occuped nes over 11 years) breedng areas perched closer o nes ses (smlar Ds) han eagles of her respecve gender n a breedng areas of low producvy (<0.77 young per occuped nes). Male Bald Eagles n hghly producve breedng areas dspersed perch ses more evenly hroughou he breedng area (fla S) han a male n a low producon breedng area, whle he oppose was rue for females. Spaal use profles derved from analyss of mean and confdence nervals and medan and nerquarle Ranges were no as descrpve or llusrave of ndvdual or group smlares or dfferences as SUs. Logsc analyss suggesed Zone (prmary bragng zone) lms recommended n regonal Bald Eagle managemen plans may need o be exended o manan performance of hghlyproducve pars nesng along rvers. SUs derved from logsc regresson models of dsance of locaons from mporan haba componens may be ndrec ndcaors of haba qualy and useful ools for descrbng and comparng spaal use of lnear habas of oher speces. KEY WOP, Ds: Greaer Yellowsone Ecosysem; Halaeeus leucocephalus; logsc reresson; spaal use ndcaors; lnear haba; rado-rackng. Anflss espacal del uso de haba lenar del fqula calva RES)MEN.--Exsen varas acncas dsponbles para el anflss de freas y de ubcac6n de anmales. Pocos son aplcables a aquellos que ulzan hfba lneares (.e., hfbas rberefos). Las fgulas calvas (Halaeeus leucocephabus) usualmene esan asocadas a ros y se concenran en sos de perchas cerca de las lneas coseras. La dsrbuc6n acumulada de la proporc6n de percha po la dsanca del so del ndo mas recenemene acvo, (deermnado po elemera), rue ulzada para comparar el uso espacal de cnco fgulas calvas adulas que andaron a lo largo del Ro Snake, Wyomng. Los ndcadores de uso espacal (UE) rueon desarrollados a par de parfmeros de regres6n logsca con el fn de: (1) Comprender y modelar los proceso subyacenes de los cuales los daos hayan emergdo, (2) Comparar con cncas de esadsca descrpva smple y evaluar la uldad de presenar una clara y acerada represenac6n del uso espacal y sus dferenca enre fgulas, (3) Relaconar las meddas del uso espacal de las fgulas con la producvdad a largo plazo de las freas de reproducc6n. E1 ndcador de dsanca (D) rue la dsanca desde el ndo ncluyendo el 50% de odas las perchas ulzadas po un fgula con ransmsor y que fuera represenava del amafo del rango. El ndcador de pendene (P) rue la pendene de la curva de la regres6n logsca en el D (puno de nfex6n). E1 P rue un ndcador de la dspers6n lnear de los sos de percha denro del frea reproducc6n. Se esudaron las fgulas calvas asocadas a una mayor producvdad (>0.77 juvenles po ndo ocupado de mas de 11 afos), 207
208 HARMATA AND MONTOPOL VOL. 35, NO. 3 areas de andac6n y de perchas cercanas a los sos del ndo (con D smlares), en conraposc6n de agulas de su respecvo sexo en/reas de reproducc6n de baja producvdad (<0.77juvenles por ndo ocupado). Los machos de /gulas calvas en/reas de reproducc6n alamene producvas con sos de perchas dspersas en forma smlar a lo largo del/rea de reproducc6n (P plano) en conraposc6n de un macho en un area de reproducc6n con un producvdad baja, se enconr6 que lo opueso ocurr6 con las hembras. Los perfles de uso espacal que se dervaron del an/dss de la meda y de los nervalos de confanza y de la medana y los nervalos enre rangos no fueron an descrpvos n lusravos de las smlardades ndvduales o de grupo o de las dferencas como los VE. E1 an/dss logsco sugr6 que los lmes de la zona 11 (zona de forrajeo prmaro) recomendada en los planes regonales de manejo de /gula calva necesa ser exendda para manener su ocupac6n por pare de las parejas alamene producvas que andan a los largo de los ros. Los VE dervados de los modelos de regres6n logsca de las dsancas de localdades provenenes de componenes de h/ba mporanes pueden ser ndcadores ndrecos de la caldad de haba y como herramenas valosas para descrbr y comparar el uso espacal de los h/bas lneares de oras especes. [Traducc6n de Csar M rquez] Several echnques are avalable o analyze spaal dsrbuon of anmal locaons (Mohr 1947, Worden 1989, Andres e al. 1994, B6gel e al. 1995, Ke e al. 1996, Marzluff e al. 1997, Buchanan 1997). Mos nvolve consrucon of wo-dmensonal polygons or kernel ellpses ha may nclude ncreasng proporons of oal anmal locaons as descrpors of spaal use. Such mehods are approprae for anmals ha dsrbue her acves somewha unformly around acvy ceners bu may be msrepresenave when appled o anmals ha dsrbue locaons lnearly along haba corrdors such as rvers. Here, we presen a mehod for comparave analyss of spaal use of Bald Ea- gles (Halaeeus leucocephalus) usng a mosly lnear, rparan corrdor and conras he uly for descrbng lnear spaal use wh sandard sascal echnques. Analyss was confned srcly o one dmensonal spaal dsrbuon of perches whn he breedng area and quanfed relave o an mporan componen of he haba, he nes se. Characerscs of spaal use among eagles were compared and relaed o number of young produced n assocaed breedng areas, long-erm. ME llol)s From 1985 hrough 1989, movemens of nesng Bald Eagles were nvesgaed n he Greaer Yellowsone Ecosysem of norhwesern Wyomng. Bald Eagles were resden a nes ses along he 108 km free-flowng poron of he Snake Rver n Teon Couny and Grand Teon Naonal Park (see Swenson e al. 1986 for descrpon of sudy area and Harmaa e al. 1999 for descrpon of Bald Eagle populaon). Daa were colleced prmarly for developmen of nes se managemen plans advocaed n he Greaer Yellowsone Bald Eagle Managemen Plan (Greaer Yellowsone Bald Eagle Workng Group 1996). Resden eagles were randomly chosen for sudy bu ulmaely seleced based on capure success (some chosen were never caugh). Adul Bald Eagles were capured and rado-agged wh al-moun and solar backpack ransmers. Gender of four eagles was deermned by poson durng copulaon pos-release and measuremens (Borolo 1984, Garcelon e al. 1985) and sze relaonshp wh he mae for anoher. Prmary funcon of elemery was o faclae connuous vsual monorng. Transmers asssed n locang marked eagles a he naon of an observaon perod and aded relocaon when eagles moved ou of sgh. Observaon perods vared from 1 o 4 hr. Monorng schedule was desgned o provde an even dsrbuon of effor over all hour perods n each week as possble. Bald Eagle perches locaed durng monorng were ploed on U.S. Geologcal Survey 7.5-mn opographcal maps. A perch was defned as any nonflyng, durnal eagle locaon deeced, eher vsually or rangulaed by elemery. Perches were locaed on rees and logs, n waer, on he ground, or on man-made srucures. Tran- gulaed locaons were easly dsngushed as perched or flyng by sgnal characerscs. Trangulaed locaons were -+ 100 m of acual as deermned by locaons of es ransmers. Acvy (hunng, loafng, erroral sgnalng, sennel, ec.) of eagles a perch ses was no evaluaed. Ngh rooss were no ncluded, nor were perches chosen mmedaely subsequen o nfrequen (N < 5) observer-nduced movemens. Perch locaons were analyzed n relaon o proxmy o he mos recenly acve nes whn he rado-agged eagle's breedng area. The mos recenly acve nes was defned as an elevaed srucure composed of scks, suaed n a conferous or decduous ree, a whch resden adul Bald Eagles were las known o be engaged n reproducve acvy. Reproducve acvy ncluded nes repar, copulaon, ncubaon, broodng, feedng, or fledgng young. Perch and nes locaons of each rado-agged eagle were assgned coordnaes based on he Unversal Transverse Mercaor sysem whn he GSDG subprogram of GEOSCAN (Monana Deparmen of Fsh, Wldlfe and Parks 1984), a compuer-based sysem of geographc nformaon programs ha relaed anmal locaons wh geographc/haba daa. Sragh-lne dsances beween perch and nes locaons were calculaed n meers by
SEPTEMBER 2001 BALD EAGLE SPATAL USE ANALYSS 209 GEOSCAN subprogram MDPP (Mnmum Dsance Pon o Pon). Daa were compled for analyss by abulang dsance of each perch deeced from he nes se, neares o farhes for each eagle racked. Because plos of he cumulave number of perches a each dsance dsplayed sgmod shape ypcal of logsc curves, a logsc regresson model (Cox 1972, Hosmer and Lemeshow 1989) was developed for each rado-agged Bald Eagle. Use of logsc regresson was an aemp o undersand and model underlyng processes from whch he daa may have emerged. The dependen varable (p) for he logsc model: e 0+ (D ) P = 1 + e was defned as he cumulave proporon of perches deeced. The ndependen varable (Dx) was dsance from he mos recenly acve nes se. Model parameers [30 (nercep) and [31(slope) were esmaed for each dsrbuon usng eravely reweghed leas squares echoques (Mongomery and Peck 1982) so ha he models could be compared among eagles. Two parameers assocaed wh he logsc model were denfed as spaal use ndcaors (SUs): (1) he dsance from he nes ncludng 50% of all perches or dsance ndcaor (D) and, (2) he slope of he regresson curve a he D (nflecon pon) or slope ndcaor (S). D was chosen because refleced he relave sze of he breedng area, ended o mnmze effec of oulers and mahemacally, was he pon on he logsc curve where slope was he seepes. S was consdered an ndcaor of perch se dsperson whn he breedng area. Seep slope (large coeffcen) a he nflecon pon ndcaed ha perches were clusered around he D. A flaer slope (small coeffcen) was ndcave of a "lazy-s" curve, ndcang perches were more evenly dsrbued hroughou he breedng area. D was calculaed by solvng for D x n he logsc equaon wh p = 0.5. Therefore, D = -[30/[3 - S was calculaed by dfferenang he logsc equaon [dp/dx = [3 p(1- p)] wh p = 0.5, resulng n S = [3 /4. Consrucon of confdence nervals of D and S for among eagle comparsons requred esmaes of varance (Var) of boh [30 and [3. Regresson analyss provded esmaes of varance of [30[Var ([30)], varance of [31[Va ([31)], and her covarance [Covar ([30,[31)] whch were used o consruc confdence nervals. Varance of he Ss ([31/4) was calculaed by Va([31)/16. However, Ds nvolved a rao of wo parameers (- [30/[31) and varance was esmaed usng he dela mehod (Bshop e al. 1975): Var (-[3 -[ 7) - Var ([3ø) 2 [3- -3 Covar ([30, Daa also were analyzed by smple mean, medan, and pos-hoc ess. Resuls were compared o SUs o evaluae whch mehods more accuraely depced Bald Eagle spaual use. P value acceped as sgnfcan was adjused for expermen-wse error usng Bonferron crera (P < c Table 1. Perch ses and evens deeced durng monorng of adul Bald Eagles along he Snake Rver, Wyomng, 1985-89. BREEDNG MONTORED PERCHES DETECTED GENDER HOURS DAYS STES EVENTS Buler 9 169 21 16 75 Cabn Creek c 364 77 46 187 Oxbow 9 448 118 22 412 Sheep Gulch c 274 74 26 63 Schwabacher c 463 164 184 433 = 0.05/10 = 0.005). Dfferences n spaal use profles among eagles were ndcaed by nonoverlap of nerquarle Ranges (QRs) and 95% confdence nervals of Ds, Ss, and means. SU of rado-agged eagles was relaed o producvy of he breedng area and gender. Average number of young produced per occuped breedng area beween 1979 and 1990 (Greaer Yellowsone Bald Eagle Workng Group 1996) was used o classfy producvy as hgh or low for comparsons. Ths long-erm producvy was consdered more represenave of resden eagles' performance han producvy concden wh perod of radorackng because severy of early sprng weaher affeced producvy of Bald Eagles n he Greaer Yellowsone Ecosysem (Harmaa and Oakleaf 1992) and weaher vared from mld o severe beween 1985 and 1989. Addonally, some eagles n sample breedng areas were resden up o 11 years (Harmaa e al. 1999). Because Sprun e al. (1973) ndcaed a leas 0.77 young per occuped breedng area was requred for Bald Eagle populaons o manan sably, hgh producvy was consdered >0.77 young produced per occuped breedng area beween 1979 hrough 1990. Low producvy was consdered <0.77 young per occuped breedng area for he same perod. RESULTS Fve rado-agged adul Bald Eagles were monored an average of 343.6 hr (SD = 123.4) over 90.8 d (SD = 53.5) (Table 1). Monorng covered a leas wo-hrds of he nesng perod (md Febmd Jul) for all eagles. Mean number of perch evens and perch ses deeced per eagle was 234 (SD = 179) and 59 (SD = 71), respecvely. Perch evens deeced were correlaed wh hr (r = 0.917, P < 0.028) and d (r = 0.888, P = 0.044) monored. However, number of perch ses deeced was no correlaed wh eher hr (P > 0.296) or d (P > 0.117) monored. Lnear breedng range vared from 2.2 km o 10.7 km from he nes se (Fg. 1). Logsc regresson models for each eagle f he daa well,
210 HARMATA AND MONTOPOL VOL. 35, NO. 3 1.0 Ll 0.9 n,' 0.8 LL o 0.7' z o o.6 n,' 0.5 o o o.4 uj o.3 o.2 ) 0.1 0 0.0 Lo 10 20 30 40 50 60 70 80 90 100 110 DSTANCE FROM NEST (m x 100) Fgure 1. Logsc regresson curves of cumulave proporon of perches by dsance from mos recenly acve nes se for fve adul Bald Eagles monored along he Snake Rver, Wyomng, 1985-89. "H" or "L" before gender con ndcaes hgh or low reproducve hsory of he assocaed breedng area (see ex). 1.e., all R s -> 0.90 (Table 2). Two males and one female exhbed smlar Ds bu Ss dffered among all eagles (Fg. 2). Two males and a female were assocaed wh hgh producvy breedng ar- eas and one male and a female were assocaed wh low producvy breedng areas (Table 2). D of he Sheep Gulch male (low producvy breedng area) was wce as far as Ds of boh he Cabn Creek and Schwabacher males' (hgh producvy breedng areas) (Fg. 2). D of he Oxbow female (low producvy breedng area) was nearly hree mes farher han ha of he Buler Creek Table 2. Logsc regresson parameers of cumulave perch se dsrbuon, Spaal Use ndcaors, and producvy of fve rado-agged adul Bald Eagles monored along he Snake Rver, Wyomng, 1985-89. P for all regressons < 0 001. BALD EAGLE DSTANCE SPATAL USE NDCATOR 1 (GENDER) 0 [3] R e (m) 2 SLOPE PRODUCTVTY Buler Creek ( 9 ) - 1.4290 0.2341 0.95 610 0.0585 Hgh (1.73):3 Oxbow ( 9 ) - 1.6987 0.0958 0.98 1774 0.0239 Low (0.58) Cabn Creek ( c ) - 1.3122 0.0708 0.97 1854 0.0177 Hgh (1.62) Sheep Gulch ( c ) - 5.7644 0.1300 0.90 4434 0.0325 Low (0.11) Schwabacher (c ) -0.7724 0.0389 0.90 1986 0.0097 Hgh (1.23) A 50h percenle (see ex). From mos recenly acve nes. Young per occuped breedng area recorded annually 1979-90.
SEPTEMBER 2001 BA.D EAGLE SPATAL USE ANALYSS 211 5000 LOW PRODUCTVTY HGH 0.08 4000 0.07 0.01 3000 0.05 0.04 2000 0.03 0.02 looo 0.01 OXBOW SC HWABACH E R SHEEP GULCH CABN CRK 9 o', o', BALD EAGLE BUTLER CRK 0.00 Fgure 2. Confdence nerval (95%) esmaes of Dsance ndcaor (D) and Slope ndcaor (S) derved by logsc regresson of cumulave proporon of perches by dsance f-om mos recenly acdve nes. Esmaes ploed n relaon o long-erm producvy of he breedng area of rado-agged adul Bald Eagles monored along he Snake Rver, Wyomng, 1985-89. female (hgh producvy breedng area) (Table 2). Males assocaed wh more hghly producve breedng areas also had flaer Ss han he male assocaed wh a breedng area of low producvy (Fg. 1). However, S of he female assocaed wh a hghly producve breedng area was wce as seep as ha of he female assocaed wh a low producve breedng area (Table 2). Parwse comparsons of medan perch dsance revealed smlar relaonshps among eagle genderproducvy groups as dd Ds, bu dfferences were no as pronounced (cf., Fg. 2, 3). Medan perch dsance was greaer for he hghly producve female han a female exhbng low producvy (U = 4656.5, P < 0.005) bu he low producon male's medan perch dsance was dfferen (U = 3737, P < 0.001) only from he male wh he hghes producon (Fg. 3). Whle Ss were dfferen among all eagles, QRs of perch dsance suggesed dsperson of perches around medan dsance was no subsanally dfferen among eagles, excep for he Sheep Gulch Male (Fg. 3). Means and confdence nervals of perch dsances mrrored gender-producvy relaonshps llusraed by D resuls (cf., Fg. 2, 3). DSCUSSON Monorng effor affeced he number of perch evens bu no number of ses deeced, ndcang vrually all ses (dsance caegores) whn eagle ranges were deeced. ncreased monorng probably would have deeced ncreased use of prevously recorded ses only, no expanson of he breedng range or changes n SUs. SUs of all Bald Eagles were dfferen bu smlares among groups were evden. Bald Eagles assocaed wh more producve breedng areas ended o perch closer o he nes se (proxmae Ds) han eagles of her respecve gender n breedng areas wh a hsory of low producvy. Breedng areas ha perm more even dsrbuon of foragng opporunes over a larger area for males (fa Ss) ye allow females o concenrae acves close o he nes (seep Ss) also may favor hgher producvy. Because he male s he prmary provder durng ncubaon and early o mdneslng phases of he breedng cycle (Salmaser 1987), localzed, emporal (daly, seasonal) dsrupons n resource pach avalably (e.g., human recreaonal presence) would be less severe n breedng areas wh more dspersed or larger resource paches han n breedng areas where resource paches were few and concenraed. Close proxmy of he aendan female would faclae quck access o he nes, permng more effecve, mely defense agans predaors (e.g., Gorvds), shelerng kom nclemen weaher, and frequen
212 HARMAT^ AND MONTOPOL VOL. 35, NO. 3 6000[ LOW.. 5000[ 4000[ 3000[ 2000[ ooo[... ñ... O L PRODUCTVTY OXBOW SHEEP GULCH SCHWABACHER CABN CRK HGH O MEAN, 95% C [] MEDAN, 25%-75% Quarles BUTLER CRK BALD EAGLE F gure g. Measures of cenral endency and dsperson of perch dsance from he nes and assocaed long-erm producvy of he breedng area recorded for fve rado-agged adul Bald Eagles monored along he Snake Rver, Wyomng, 1985-89. feedngs. Hgher producvy would be manfes n pars wh ncreased female aenveness and more effecve provsonng by he male. Resuls mply more producve Bald Eagle breed- ]ng areas along rvers may requre a larger prmary use area (.e., Zone ), a leas along shorelnes, han recommended n regonal Bald Eagle managemen plans (e.g., Monana Bald Eagle Workng Group 1994, Greaer Yellowsone Bald Eagle Workng Group 1996). Manenance of reproducve performance of hghly producve pars may requre exenson of Zone up o 2 km from he acve nes o nclude 50% of male perches and 4 km o nclude 75% of male perches (Fg. 1). Hall e al. (1997) defned haba qualy as "he ably of he envronmeno provde (for)... populaon perssence." Gomparave SUs of resden aduhs herefore may be ndrec measures of haba qualy whn Bald Eagle breedng areas along rvers and predcve of relave producvy among poneerng pars or among pars wh unknown reproducve hsory. Furher, emporal changes n SUs may be ndcave of changng haba qualy. Analyss of use of lnear haba by Bald Eagles wh paramerc ess (means& confdence nervals) may be approprae when daa are few, approxmaely normally dsrbued, and sascal esng s desred. However, resuls of such ess were nconssen wh D resuls (cf. Fg. 2, 3). Me- dan analyss may be approprae when daa are sparse and no sascal esng for concenraon s needed. When daa are adequae (N > 25), comparsons of SUs derved from logsc regresson analyss may be more descrpve of lnear haba use han smple comparsons of medans and means and her respecve measures of varably. SU analyss provded opporunes for objecve sascal esng, possbly modeled underlyng processes and was less labor nensve and cosly han GS analyss. Gomparsons of SUs derved from logsc regresson may have applcably o oher speces assocaed wh lnear habas (e.g., escarpmens, sreams, power lne and open space corrdors). However, relaonshps among Ds, Ss, long-erm producvy, and haba qualy presened here are admedly enuous due o small sample sze and requre furher nvesgaon for confrmaon. ACKNOW,EDGMENTS Bob Oakleaf, Wyomng Game & Fsh Dep., orgnally alered us o he managemen value of he 50% level n our analyss and has appled he heory o managemen of Bald Eagles n Wyomng. Sudes from whch hs paper was derved were funded hy Wyomng Game & Fsh Dep.; Unv. of Wyomng-Naonal Park Servce Coop. Research Un; Resource Managemen Dv., Grand Teon Naonal Park; U.S. Bur. of Reclamaon, Mnadoka Projec; U.S. Fores Servce, Brdger-Teon Naonal Fores, Monana Sae Unv. Research and Creavy Program,
SEPTEMBER 2001 BAL) EAGLE SPATAL USE ANALYSS 213 Unversy of Wyomng, Basc Research Grans; and Dep. of Veerans Affars. Rado rackers ncluded A1 Bah, Ma Erckson, Mch Fnnegan, Barb Frankln, Karen Koze, Carl Mchell, Jeff Rauo, Pee Paek, and M. Whfeld. We hank John Marzluff} Mke Morrson, and Marco Resan br valuable commens on earler drafs of hs man- uscrp. LTERATURE CTED ANDRES, A.M., H. GULNGK, AND M. HERREMANS. 1994. Spaal modelng of he Barn Owl Tyro alba haba usng landscape characerscs derved from SPOT daa. Ecography 17:278-287. BSHOP, Y., S. FENBERG, AND P. HOLLAND. 1975. Dscree mulvarae analyss: heory and pracce. Massachuses nsue of Technology Press. Cambrdge, MA U.S.A. BOGEL, R., W. D'OLERE-OLTM '4NS, AND H. FRANZ. 1995. An negraed sysem fbr resource nvenory, wldlf monorng, and managemen usng GS, GPS, and ADF-elemery echnques. Pages 551-555 n J.A. Bssonee and P.R. Zffausman [EDS.], negrang people and wldlf fbr a susanable f ure. Proceedngs of he frs nernaonal Wldlf Managemen Congress. The Wldlf Socey, Behesda, MD U.S.A. BORTOLOTT, G.R. 1984. Crera for deermnng age and sex of neslng Bald Eagles. J. Feld Ornhol. 55:524-542. BUCHANAN, J.T. 1997. A spaal analyss of he burrowng owl (Speoyo cunculara) populaon n Sana Clara Couny, Calfbrna, usng a geographc nfbrmaon sysem. Pages. 90-96 n J.R. Duncan, D.H. Johnson, and T.H. Ncholls [EDs.], Bology and conservaon of owls of he norhern hemsphere; second nernaonal symposum. February 5-9, 1997. Wnnpeg, Manoba, Canada. U.S. Deparmen of Agrculure. Fores Servce. Norh Cenral Fores Expermen Saon, S. Paul, MN U.S.A. Cox, D.R. 1972. Regresson models and lf ables. J. R. Sa. Soc. Seres B. 34:187-220. GARCELON, D.K., M.S. MARTELL, P.T. REDG, AND L.C. BUOEN. 1985. Morphomerc, karyoypc, and laparoscopc echnques fbr deermnng sex n Bald Eagles. J. Wldl. Manage. 49:595-599. GREATER YELLOWS'ONE BA.D EAGLE WORKNG GROUP. 1996. Bald Eagle managemen plan fbr Greaer Yellowsone: 1995 Updae. Greaer Yellowsone Bald gle Workng Group. Wyomng Game & Fsh Deparmen, Lander WY U.S.A. HARMATA, A. AND B. OAKLEAF. 1992. Bald Eagles of he Greaer Yellowsone Ecosysem: An ecologcal sudy wh emphass on he Snake Rver, Wyomng. Wyomng Game & Fsh Deparmen, Gheyenne, WY U.S.A. HARMATA, A.R., G:J. MONTOPOL, B. OAKLEAF, PJ. HAR- MATA, AND M. RESTAN. 1999. Movemens and surwval of Bald Eagles banded n he Greaer Yellowsone Ecosysem. J. Wldl. Manage. 63:781-793. HXLL, L.S., P.R. KR US, AND M.L. MORRSON. 1997. The haba concep and a plea for sandard ermnology. Wldl. Soc. Bull. 25:173-182. HOSMER, D.W. AND S. LEMESHOW. 1989. Appled logsc regresson. John Wley & Sons. New York, NY U.S.A. F, ae, J.G., A. BA.DW N, ANDJ.C. EVANS. 1996. CALHOME a program fbr esmang home ranges. Wldl. Soc Bull. 24:342-344. MARZLUFF, J.M., S.T. KNCK, M.S. VEKASY, L.S. SHUECK, AND TJ. ZARRELLO. 1997. Spaal use and haba selecon of golden eagles n souhwesern daho. Auk 114:673-687. MOHR, C.O. 1947. Table of equvalen populaons of Norh Amercan small mammals. Am. Mdl. Na. 37. 233-249. MONTANA BALD EAGLE WORKNG GROUP. 1994. Monana Bald Eagle managemen plan July 1994. USD, Bureau of Reclamaon, Monana Projecs Offce, Bdlngs, MT U.S.A. MONTANA DEPARTMENT OF FSH, WLDLFE &: PARKS. 1984. GEOSCAN users manual. Monana Deparmen of Fsh, Wldlfe & Parks Research Bureau, Box 5, Bozeman, MT U.S.A. MONTGOMERY, D.C. AND E.A. PECK. 1982. nroducon o lnear regresson analyss. John Wley and Sons. New York, NY U.S.A. SPRUNT, A. V, W.B. ROBERTSON JR., S. POSTUPALSKY, RJ HENSEE, C.E. KNODER, AND F.L. LGAS. 1973. Comparave producvy of sx Bald Eagle populaons. Trans N. Am. Wldl. Na. Resour. Conf 38:96-105. STALMASTER, M.V. 1987. The Bald Eagle. Unverse Books New York, NY U.S.A. STATSOFT, NC. 1999. STATSTCA fbr Wndows [Compuer program manual]. SarSof, nc., Tulsa, OK U.S.A. SWENSON,J.E., K.L. ALT, AND R.L. ENG. 1986. The ecology of he Bald Eagle n he Greaer Yellowsone Ecosysem. Wldl. Monogz 95:1-46. WORDEN, B.J. 1989. Kernel mehod fbr esmang he ulzaon dsrbuon n home range sudes. Ecology 70: 164-168. Receved 13 May 2000; acceped 13 Aprl 2001