WWWWWlmMUNHHIHHIWI - 401 IN: (DOD—I LIBRARY Michigan State University s. , This is to certify that the thesis entitled BLACK BEAR HABITAT UTILIZATION AND HABITAT MODEL VALIDATION IN MICHIGAN presented by James Glenn Hirsch has been accepted towards fulfillment of the requirements for M.S. degree inflaheriea & Wildlife Major professor Date March 8, 1990 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution I PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. DATE DUE DATE DUE DATE DUE "mu? Wm Action/Equal Opportunity Institution cmmd BLACK BEAR:HABITAT'UTIIIZAETON’AND HABITNIIKMIflhlflflJIEflICN'INWMICHIGAN BY James Glenn Hirsch A.THESIS Submitted.to ' Michigan State University in partial fulfillment of the requirements for the degree of MRSTERMOF SCIENCE Department of Fisheries and Wildlife 1990 (94‘0" 79(05 Am MEEARIIABI'IATUI'IIIZATIG‘IAND WWVAIIDATIQ‘IINMGHGAN BY James GlemHirsch 'Ihe habitat selection and movanents of 33 radio-collared black bears MM) weremonitored frunMardmtoDecarber1988, on Dnmnond Island, Michigan. An attaipt was mde to validate a Habitat Suitability Index (HSI) model designed for black bears. Habitat variables required by the RBI model were measured in appropriate vegetation types to determine HSI scores for each adult bear's home range. Adult male and female hare range sizes were 75.64 ka and 43.14 kmz, respectively. The movement of bears depended on the distribution and abundance of preferred foods. Significant correlations were not obtainedwhenl-ISI scoreswereccrrparedtormerangesizes, cublitter sizes, cub weight gains, and mean daily mvements. 'Ihe HSI score for Drummond Island did not correspond well to average age of first reproduction and mean home range overlap. The black bear RSI model shaild be revised to better reflect black bear ecology. mm Iwould liketoaclmledgethepennanentresidents ofDrurmnond Island who supplemented my field season with an abundance of friendship and laughter. Special recogniticn is due to Ray and Gineene Yates, whose lmlimited hospitality and cmtinued invitations to hunt and fish withthem, madethepressures of school seemveryfaraway. Iwould also like to acknowledge the Nortl'mod Inn and cmck's Place for providing a convenient place to make friends and have good times. I gratefully aclmcwledge my field assistants Steve Griffin and Chris Lousias whose dedication and patience made this study a success. Special thanks to Dr. Iarry Vissir and Elaine Carlsogwho provided supplemental information related to this study, and who also ensured, thatIhadnadfineryandequiprentthatwereingoodworldngorder. MydeepestappreciatimtolouisBenderardMarkOttenfortheir loyal friendship. 'Ihey frequently filled the office with both huncrous and intellectual conversation. Special thanks to Louis Bender for reviewing this manuscript, and whose cynical view of wildlife research enhanced my lmcwledge and changed my attitude towards this discipline. IwouldalsoliketothankDr. ScottR. WintersteinandPaul Padding for their guidance in statistical analyses. Recognitimareduetomycaunitteemenbers; Dr. Jonathan B. I-Iaufler, Dr. Scott R. Winterstein, and Dr. James B. Hart. Special thanks to my major professor Dr. Jonathan B. Haufler whose stubborn and ii persistent dealings with Karl Hosford (wildlife division chief for the Michigan Department of Natural Resalroes) made this study possible. I would also like to acknowledge my major professor for providing guidance and advice throlglnrt this study. Finally, I would like to admledge Paul Steblein, Nancy Mathews, Rob Bouta, Dale Garner, and Glenn Johnson for their friendship and alpportduringmyyearsatmcollege offiwiramental Scienceand Forestry. 'Ihey taught me nuch about wildlife biology and have unch to dowiththesuocessofthissuadyarrimyaccatplislments atMichigan State University. iii moroomms page usrornams v usrorrrams vi omncrrvrs ..... .. 6 VegetationSampling 14 HSIModelValidation 18 HSIModelValidation 29 DISQBSIQI OOOIOCOCOOOOOOOOCO0.0.0.000....0.000000000000000000000 37 mmmmm OOOOOOOCOOOCOOOOOIO00.00.000.000.0 37 mmimtug 0.00000000COOOOOOOOOOOOOOOOOOO000...... 40 ’51 m1- validatim OOOOOOOOOOOOOOOOOOOOOOOO0.0.0.0.00.0... 43 WWI“ OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOO 57 WW“ OOOOOOOOOCOOOOOOOOOOOOOOOOOOOOOOOOOOOO 59 mm 0.0...OOOOOOOOOOOOOOOOOOOOOOO0..OOOOOOOOOOOOOOOO 62 iv LISTOF'HKBIES H.399 Mean (S.E.) percentshrubcoverof softmastproducing species, mean (S.E.) nunber of soft mast producing species, andmean (S.E.) basal area ofhardmast preducingtrees>40yearsofageforvegetatimtypes mDnmmoni Island, Michigan, 1988 ..... ...... 24 Timing of soft mast production and vegetation type location for major shrub species used by black bears on Dnmmond Island, Michigan, 1988 ............ ..... 25 Mean (8.13.) daily movements (km/day) of adult male, adult female without cub, adult female with cub, yearling male, and yearling fenele black bears during spring, ' , and Simmer/fall on Dnmmond Island, Michigan, 1988 ......... 27 Annual hone range sizes (kmz) of adult blackbears calallatedbytheminimnncorwexpolygmnethod (MCP), andharmonicmeanmethod (1W) using95% contours, on DrlmnondIsland, Michigan, 1988 ....... 28 Meanperoentusearrineanpercentavailabilityof vegetation types by season for adult black bears on W151“, MindliW’ 1988 00.000.00.00 00000 .00 000000 30 Suitability index value for percent area in wetland vegetation types (SISP) , suitability index for sumer foods (SISU) , suitability index value for percent area in non-forested vegetation types (SIV6) , suitability index for fall foods (SIFA) , suitability index value for percent area in hard mast producing vegetation types (SIV'7), suitability index value for percent area in zones of harsh influence (SIHI) , and habitat suitability index score (HSI) for all of Dnmmond Islandandfortheamlallmerangesofadult black bears on Dr'urmord Island, Michigan, 1988 ........... 31 10 LISTOF FIGJRE location of study area in relation toMichigan .. ........ Mean mom-lily taIperatures and total mnthly precipita- tion at Chippewa (nutty, Midligan, 1988 Relationship of habitat variables and life requisites in the black bear Habitat Suitability Index Model for theUpperGreat Lakes Region ...... Spearmns rank correlation analysis of annual hare range size and overall HSI score for adult black bears on Drunmond Island, Michigan, 1988 Spearmans rank correlation analysis of cub litter size and overall IBI score for adult fanale black bears on Dnmrnond Island, Michigan, 1988 ..... . ..... Spearmans rank correlation analysis of cub weight gains and overall I-BI score of adult female black bears on W151“, maxim, 1988 0.00.00.00.00... 000000000 Spearmans rank correlation analysis of mean daily movements and overall HSI score for adult black bears mDrtmndesland, Michigan, 1988 ...... Relationship between smaller vegetation type oarposition mitabilityindexscorearflpercentareainsrmner food-producingvegetationtypes ....... Relationship between human intolerance suitability index score and primary road density for black bears Relationship between hunan intolerance suitability index score and permanent dwelling density for black bearsintheUpperGreatlakesRegion ...... vi 33 34 The black bear (HISLS W) is an carnivore which usually occupies heavily forested regions interspersed with areas of early succession (Herrero 1979) . Althongh primrily vegetarians, black bears will eat carrion and occasionally prey upon ungulate fawns (King 1967, Ballard et al. 1981, Ozoga and Venue 1982, Verspoor 1983, Mathews and Porter 1988, larson et a1. 1989) . Early spring bear diets usually corsist of gm and forbs (Hatler 1972, Kelleyhouse 1980, Grenfell and Brody 1983, Rogers and Allen 1987, Rogers et a1. 1988). These foods representasouroeofprotein, butdonotresult inweight gains (Landers et a1. 1979, Beeman and Pelton 1980) . As spring progresses, bearsoccupyuplarxiareastoforageonupland forbs, ants, andthebuds and catkins of aspen (Ms spp.) and balsam poplar (2. Esauifera) trees (Rogers and Allen 1987, Rogers et al. 1988). Bearsstarttogainweightwhentheircsrbdrydrateuptake increases in smrner (Rogers 1976, 1987: Alt et a1. 1980). 'Ihe berries of Canada buffaloeberry (mg mg) , blueberries (yaccinium app-L servioeberry (engined—er sppo). wild strawberry (Emis- Spp-). wild sarsaparilla (Malia moguls). raspberries (Ms spit»). arri cherries (2mg spp.) are preferred foods during smmner (Rogers 1987, Rogers alt] Allen 1987, Noyce and Coy 1989) . Stld‘l berries are mostabundant inregeneratingaspenstandsarriopenuplanivegetation 2 typesinmrthcentraluirmesota (NoyceandOoy1989). Roadsidesand clear-wtareaswerefanritoprodweahmdantsoftnastinthecentral Adirondacks of Nev York (Warbuton 1982) . Similarly, Lindzey and Meslow (1977) observed that soft mast producing species were 7-8 times more abmflantinclear-artareasthaninoldercaliferardalderstandsm long Island, Washington. Grasses and forbs are still utilized during sumer, but to a midi lesser extant relative to spring (Hatler 1972, Zytaruk and Cartwright 1978, eternal and Brody 1983) . Infall, bearsshifttheirdietsfrunsoftmasttohardmast. MastfrunAnericenbeedl (Eagusggrgifglia), northernredoak (m mgr-a) andotherhardmastproducingtreespredaninate fall diets (Piekielek and air-ton 1975, Rogers 1976, Iarflers et a1. 1979, Garshelis and Pelton 1981, Grenfell and Brody 1983, Rogers 1987, Rogers and Allen 1987) . Such foods have a high fat content, allowing bears to build fat reservesnecessary forwinterdormancy. Softmastcrops still persisting in early fall, such as roundleaf dogwood (gem m), highbushcranberry WM), andccnmonwinterberryholly MM) arealsoutilized. War, thesefoodsarenotas energyridlashardmastcropsandareusuallypasttheirpeakof availability (Rogers and Allen 1987) . Black bears are typically dormant fran OctoberbNovenber to April- May in cool northern climates (Jonkel and Conan 1971, Rogers 1977, 1987) . In contrast, denning activity is usually brief or non-existent insouthernregions (Landersetal. 1979). Bearstendtobequite variable in den selection and cmstruction. For exanple, most bears at Fortmmrray, Alberta, werefmnritoelmvatedenswithinyormgaspen- 3 birth stands (Fuller and Keith 1980) . By contrast, bears at Gold lake, Alberta, usuallyexcavateddersmflerroomassesinmamrestandsof aspen-spruce and grace (Tietje and Ruff 1980). Most den sites in the (htskill Man'rtains, New York, were in rock cavities in heavily forested areas averaging 1.6 km fran the nearest road (O'Pezio 1980). Bears in Michigan's Upper Peniralla usually excavate dens under stunps or logs within conifer and mixed confer-hardwood swanps (Erickson 1964a). Due to the low incidence of den re-use, the availability of den sites is never considered more limiting than the distribution and abundance of bear foods (Lindzey and Meslow 1976, Alt and Gruttadauria 1984, Rogers and Allen 1987) . Black bears camonly reach sexual maturity at 3-4 years: however, seamalmamrityaslateas7yearshasbeenreported (Kolenoskyand Strathearn 1987) . Both males and fanales mate promiswously during the miquune to mid-July breeding season (Kolenosky and Strathearn 1987, Rogers 1987) . Adult females usually give birth to 1-3 :1le during winter dormancy. Bear cubs accarpany their mother for approximately 18 months. Male bears do not participate in natal care: in fact, they will occasionally kill cubs (Rogers 1987). Overall, bears are solitary, congregating only around concentrated food sources such as garbage dumps. mnnanencroadmentmbearhabitatisoneofthemajorreasms for decreased bear numbers (Rogers 1987). Bears are secretive, avoiding hlman activity centers and dwellings unless a reliable food source (i.e. garbage dump, orchards) is available. Areas of high human activity frequently act as population sinks for bears (Rogers and Allen 4 1987) . Tinber harvesting activity also negatively inpact bear popllaticns by increasing hunting ani poaching opportunities, as a result of road ccnstructim, and decreasing hard mast availability (Rogers and Allen 1987) . However, logging can also positively affect bear habitat by providing increased soft nest production. Historically, black bears were mpnotected in Michigan until 1925, when declared a game animal by the state legislature (Stuewa 1957, Midligan Department of Natural Resources 1988) . In 1939, however, the legislature renoved statewide protection, but allowed the Natural ReswrcesOamissicntoprotectbearsincolmtiesthatrequestedit. Prior to 1980, except for a brief period in the mid 1960's, bear hmrtersneremlyrequiredtopossessadeerlicensetoharvestbears. Since 1980, aseparatebearlicensehasbeenrequiredtohmltbearin Michigan. Michigan bear hunters mnbered approximately 10,000 in 1985, ardthisrnmberhassubsequentlyincreased (MidliganDeparUnentof NaturalResources1988). 'Ihegrmringmmberofbearhm'rtershas resulted in increased interest and concern over Michigan's black bear population (Smith 1985). In addition, there has been a persistent conflict between bear hunters that utilize baits and those that utilize dogstoharvestbears. Infaceoftheseissuesardprobleis, increased lmowledge of black bear population dynamics and habitat utilization is inportanttomaintainoptimalbearmmbersardhabitat. 'Iheplrposeofthisstudywastodetermineblackbearhaterange sizes, hate range overlap, movanents, activity patterns, and habitat selection on mmmmd Island, Michigan. Previous studies conducted on Michigan's black bear pqaulatim have failed to intensively study these 5 aspects of black bear ecology (Erickson 1964b, Rogers et a1. 1976, Manville 1982) . Additialally, this study also involved validation of a recently developed Habitat Suitability Index (m1) Model for black bears occupying the Upper Great lakes Region (Rogers and Allen 1987). Since HSI models are objective and produce a numeric measure of habitat quality, they have the potential to be valuable wildlife managenent tools. Hcmever, theacalracyofmostmlnodelshasmtbeentested. Validation of the black bear HSI model is necessary so that HSI scores are accepted by wildlife managers, judges, the public, and other factions involved or influencing land use decisions. 'lhemainobjectivesofthissmdywereto: 1)determirenovanentsofbearsmadailyardseasonalbasis, 2)determinebearl‘merangesizesandrnnerangeoverlap, 3) determine seasonal habitat preferences and critical habitats for bears, 4)validate the accuracy of the black bear HSI model, and 5)make recannendations for habitat management for bears in Michigan. SIUDYAREAIISCRIPI'ICN missuldywascorductedmnr'lmnorfllsland, Michigan, located at thenorthernerdoflakemlrm, 1.6kmofftheeasterntipof Michigan's Upper Peninsula (Fig. 1). 'me island, located within caippwaoamty, is337kmzinsizeandisoccupiedby800permanent and 3000 seasonal residents (Drunmord Charter of Camerce, pers. cannm.). Drummord Island is accessible by ferry year ramd, and receives heavy recreational use. Sport fisherman take advantage of the good Yellow perch (M W). walleye (film 2m» and salmon (W spp.) fisheries offered in the vicinity of minimal Island. mite-tailed deer (M yiminiams) and small game lnmting are popular recreational activities during the fall. Bear mmtingonmnmrilslandwasclosedin1982 bemuse ofconcernsof overexploitation. Hormever, a regulated permit hunt was established in Santeriber 1988. logging and mining activities also occur on Dnmmond Island. logging activities are generally concentrated at the eastern portionofthestudyarea, whileanopenpitlimestonemineexistsat the western portion of the island. Slightly over 50% of the island is mgedbyflefldmiganWofNahmalReswrms (MEIR), Forest Managanent Division. 'Ihe climate of Dnmmond Island is considered maritime, but shifts tocontirentalpolarduringthewinterwavisardr‘rey1984). 'Ihemean Drumm0nd lslond """"""" """"""""" . n to mdfigaw e relatl area 1] . f Stu 3 1:10n ° 1003 Fig“m 1' 9 annual tarpsrature is 4.8%, with a monthly mean in July and January of 18°C and -1o°c, respectively (Hadeen 1988) . 'Ilotal anmlal precipitation averages 82 an, withmost falling as rain (Hadeen 1988). Snow is permanentlycnthegrand franmid—Novarbertomid—April (lhvisand Frey1984), resrfltirginagrowingseasmthatisslightlyoverll months long (Albert et a1. 1986). Itaperaulres are moderated, and cloudiness and precipitation irrzreased, because of lake Huron. In early fall, thepassageofcoldairmasses franthewestoverthewarm watersoflakealperiorandlakefhnmresultsinfrequentfog. Simmer rains are associated with post-frontal northwest winds. long term (1951-1980) mean monthly tarperature and precipitation data, and mean nonthly temperature and precipitation for 1988 , are given in Figure 2 (Hadeen 1988). Drtmmond Island mists of lower and middle Silurian Limestone andDolanitebedrock, withthenortherntipoftheislarflcontaining upper Ordovician Limestone and Dolanite bedrock (Davis and Hay 1984). A thin layer of glacial till can be famd over the majority of the island. Detalr stony loam, Johnswood stony loan, and Carbondale muck arecamon soils withinthe studyarea (Veatchetal. 1927). The majority of organic soils (i.e. carbondale nick) are concentrated in thecentralportimoftheislandduetothelavelevationinthat area. 'nleDetarrstonyloamsoil isfertileandhasahighmoisture content, but is excessively std-1y. 'nle Jolmswood stony loam is closely associatedwiththeDetalrstonyloam, butisless stony. 'Ihissoil is moderately fertile and fairly moist. The underlying drift in the Jolmswoodstmyloamisnotasclayeyasthenetalrstonyloam: thus, 10 15 - * ‘4 _ "' "' dud! W 1“ m 1951-1980 I ‘ I \ Precipitation (cm) Temperature (°C) Figure 2 . than monthly taIperatures and total monthly precipitation at Chippewa Oamty, Michigan, 1988. 11 itisbetterdrained. Bothsoilsterrltobetooshallowfor agricultural purposes. Smooth terrain is predominate on the island, with frequent rolling gramdmorainesandanoccasiaial largeridge. Elevationsvaryfrom 175 to 315 m. Vegetative coverage a: Dnmnmd Island consists of 42% aspen-birch, 28% cmifer, 13% upland hardwoods, 5% openings, 4% wetlands, and 4% lowland hardwoods. In addition, 3% of the island is carprised of residential, industrial, and recreational areas. 'Ihere are2 farnsontthislandwhidl, “mouthed, occupy<1%ofthe total area. 'Iheaspen—birdlvegetationtypeconsisted ofquaking (B. M) ardbiqtooth asoen (2- W). with scattered paperbirch (WW). I-Iowt-Ner, sane aspen-birch standswere exclusively paper birch. Cannon understory species within this vegetation type included red-osier (9. M) and ramdleaf dogwood, Canada buffaloeberry, beaked hazel (00mins M), serviceberry, balsam fir (ALE M), and northern white-cedar (Mia my . Upland coniferous and lowland coniferals areas were both categorized as the coniferous vegetation type. lowland coniferous areas (usually cedar swanps) were predaninately northern white-cedar with scattered swanp haeyslcltle (m filmifolia) and speckled alder (bums m) in the understory. Upland conifer-ms areaswerepredaninatelyredpine MM)wifl1Camda mffaloeberry. servioeberry. and car-non imper (Mame mtg) in theunderstory. 'nhcaliferousvegetatimtypehaever, wascanposed mostly of the lowland coniferals type. American beech and sugar maple 12 (MW)weretlhdaninateoverstoryspeciesintheupland mmodvegetatimtype.wifl1redra8pbenymsrith Americanbeednandsugarnapleinthemderstory. Openingswere typicallydaninatedbywildstrawberryardgrasses,withcatm1 WQ-W)Wirgflhperiptery. Harever,sane stardsdelirhatedasopenirgswereaculallyregerhratirgaspen-birdl stands. Slmlbswanps,n1dflats,andshallavmarsheswereall categorized as wetlands. lowland hardwoods were predaninately balsam poplar and ash (m spp.) with a diverse understory that included blackspruce (MM), red-osier dogwood, balsam poplar, ash, andspeckledalder. wring sunmer, 1986, the m, Wildlife Division trapped ard ear- tagged23 blackbearsonnnmnordlslardinanefforttosttdyblack bear populations (Visser 1987) . Capture efforts continued again the following sunner with sane captured bears fitted with radio-collars. Additionally, 11 yearlings were fitted with radio-collars as a result ofwinterdenchecks. Byspring 1988, 33bearstrmmordIslardwere equipped with radio-collars, ard albsequently monitored for this study. General location ththods Radio-collared bears were located fran the grand at rardomly selected times thralghart their daily activity period (0500 to 2300 hairs) fran March to Decaiber 1988, using a portable 'IR-z receiver with a hard-held 2 element yagi antenna (Felonies, Inc., Mesa, AZ). An attenptwasnadetolocatebearsatleastonceeveryZdaysbutmnore than once per day. The vegetation type for each location was identifiedbywalkinginonthebearorbynovingarandtheanimala mininum of 3 sides. The overstory vegetation, described by the Michigan State Forest mentions Inventory systan (Midligan State Forest Cperations Inventory 1982) , activity, ard predaninate shrub species were recorded for each location. legal descriptions were used to a mininum of 16.2 ha (40 ac) to describe locations. 13 14 Bear activity was mmitored frun 2300 to 0500 hairs during spring, m, ardfall (twiceinspring, moeinsunrner, ard4timesinfall) todetermireifbearsvereactiveatallduringtlhnight. Activity fordaytimeardnighttime locatia'lsueredetermirhdbyeitherdirect observatimorbytheintegrityoftl'hradiosignal. Bearswhoseradio signal varied in intensity were ansidered active, while bears whose radio signalswerecmstantwerecmsidered inactive. 'lhiswasa subjective measure of activity, since radio signals can be influenced by wird, vegetation, precipitation, topograrily, and observer skill (anger 1988) . Vegetaticn Sanpling Vegetation types were delineated on Drunmord Island by a digitized Systan (MIRIS), a programwithin the lard ard Water Management Division ofttmR. mismpwasdevelopedfrunthe interpretation of color infrared aerial motos taken in 1978-79 (M. Scieszka, MINR, pers. comm.) . Nunberardpercentstmlbcoverofsoftmastproducing species were determined by the line intercept method (canfield 1941, Hays et a1. 1981) for aspen-birch, uplard hardwood, ard open vegetation types. Stratified randan sanpling was caducted, with stards for sanpling selectedfranthethRISvegetatimcoveragenap. Fa1r30mlong interceptlineswererardanly locatedwithineachrardanly selected stard. 'Ihenunberof softnastproducingspecies (inclldingbeaked hazel), thatcovered>1%oftheline, andperoentcoverageofsoftmast 15 producing species were tallied. thans of measured habitat variables wereextrapolatedtoeadlvegetatimtypefruntheneansofsanpled stards. fillerequirednunber(90%confidenceintervalthatis¢20%of flaw) of replicatestardsmsdeterminedbyStein's two-stage sanplenethod(Steelard'Iorrie1980). Forothervegetatimtypes (lowlard hardwood, cmiferous, andwetlard vegetation types), callar estimates of habitat variableswerenade, astheocalrrence of soft nastproducingspeciesinthesevegetationtypeswasextrenelylav. 'lhetimingofsoftnastproductimwasnotedforshrubspeciesthat wereutilizedbybears. thanbasalareaofhardmastproalcingtreesgreaterthanwyears old,ardthemmberofhardmastproducingtreespeciesweredetermined by the Bitterlich method for the upland hardwood vegetation type. Each rardanly selected stand was sanpled with 20 rardanly located sample- points. BasalareawasmeasuredwithalO—factor tubular gauge, and themmberofhardmastproducingtreespecieswastalliedateach sanple—point. 'memeanduiatwhidlhardmasttreesweregreaterthan 40yearsoldwasdeterminedbycatparingdbhwiththemmberofammal grwthringsfandonincrenentborecores. 'Iherequiredmnnber(9o% confidence interval that is 120% of themean) of replicate stards was determinedbyStein'stvn—stagesanplemetlmd (Steelard'Iorrie1980). thvalents ard Hone Ranges Bear locations were plotted m the MIRIS vegetation coverage map using ARC/INFO (mironmental Systems Research Institute, Redfield, a) , a geographic information systen operated on a VAX 8650 (Digital 16 Equipment Corporatim, thynard, MA). Bears were categorized as adult males (n = 3), yearling males (n = 5), yearling females (n = 4), ard adult fenaleswith (n=4) ardwithartalbs (n=12). Bearsz 3years of age were considered adults, since breeding was evident at this age. Adult females with albs produced their litters in the winter of 1987- 88, httthesealblitterdatanerenotusedforvalidatirgtheI-ISI model. mecific tEI nodel validatia'l procedures are disalssed in the HSI Model Validation section. Distances noved between conseartive locations were calculated for all locations using a program created in ARC/INFO. locations 3 days or more apart were deleted fran analysis. Distances moved between consecutive locations that were 2 days apart, were not different frat distancesmovedinlday. 'nms, allmovanentdatawereexpressedas kilaneters moved between consecutive locations, without modifying the raw data. Spring, breedingtard summer/fall time periods were separated for analysispurposes. ‘nlespringtine-periodocalrredfrandenenergence tolZJune, breedingfr‘anlBJmletonJUly, ardsunner/fall from14 Julytodenning. 'mestartofthebreedingtine-periodwasdelineated mulebasisofirrzreasedsightingsoftnnnarkedbears, theoccurrence of family breakup, ard the location of adult males ard females together. Frequentdtmpvisitsbyadultnalesnarkedtheerdofthe breeding time—period. Friednan's test (Siegel 1956) ard the Friedman- type simltaneals rank test (Miller1981) were used to catpare mean dailymovanentsammgtineperiods foragivensexandageclass. ‘Ihe Kruskal-Wallis analysis-of-variance test (Siegel 1956) ard a modified 1‘7 Kruskal-Wallis-type sinultaneous rank test (Miller 1981) using mean ramswereusedtocarparemeandailynovarentsamongsexardage- classes for a given time period. 'nle selected alpha level was 0.10 for these ard all other statistical tests. ‘nle Microcarprter Program for the Analysis of Animal locations (KIPAAL) (H. Stowe ard C. E. Bldlowiak, Corserv. Res. Cent., Natl. 2001. Park, anithsonian Inst., Fra'ttRoyal, Va.) was usedtodetermine annual bane range sizes for radio-collared black bears. Annual hate range sizes were calculated with mininum convex polygons (Mohr 1947) ard 95% harmcmic mean contalrs (Dixon ard (Iranian 1980) for adult males ard females. The Mann-Whitney U test was utilized to test for significant differences between male ard fanale bane range sizes (Siegel 1956) , while the Friedman's test (Siegel 1956) was used to carparehanerangesizes estimatedbytheharnonicmeannethodwith thoseestimatedbythemininunconvexpolygonnethod. Annualhane ranges for adult male ard fanale bears were also delineated with ARC/INFOusingtheminimnnconvexpolygmnethodtodeterminehate range overlap, habitat use, ard HSI scores. SeasonalHabitatUse 'Iheproportionofvegetatimtypesineadlseasonalhcnerangewas ccnparedtotheproportimofvegetatimtypesonnnmmdlslardby the Chi-square goodnss-of-fit test, with the null hypothesis that bears select vegetation types inprroportion to their availability. Inaddition, seasmalhabitat‘usewassudied withintheannual hmerangesofadultbears. Percentuseofvegetatimtypesmre 18 determined seasmally for each bear that was successfully monitored thralghout a specific season (i.e. spring, Simmer, ard fall). Percent availability was determined fran the vegetative conposition of an adult bear's armual bane range. Subsequently, significant differences between mean percent use ard mean percent availability of specific vegetation types were determined by the Chi-square goodness-of-fit test. Seasons were divided m the basis of plant plellology, with spring coalrringfrandenetergencetoumnh, amrerfranlsa‘unetos September, ardfall fran6Septenbertoderming. 'Ihefirstsoftmast cup to ripen (Canada buffaloeberry) ard associated feeding sign definedthestartofsunmer, whiletheinitialappearanceofbeechnuts ard associated feeding sign defined the start of fall. HSI thdel Validation Overall HSI scores were calculated for Drunmord Islard ard for the amnlhaterangesofadultbearsusingprocedurosaltlinedbykogers andAllen (1987). Percent area in forested wetlards, percent area in ram-forested vegetation types, percent area in vegetation types that havehardmastproducingtrees, ardpercentareairsideazoneof (Innnan) influence were habitat variables measured using ARC/INFO (Fig. 3). 'Ihe remaining habitat variables were measured in the field as discussed in the Vegetation Sanpling section. Overall HSI scores were determined by the following equation: HSI - ((SISP + (8180 x SIV6) + (SIFA x SIV7) )/3) x SIHI whereSISP=suitabilityinexvalth forpercentarea in forested wetlard vegetation types, 19 . €an 5H2 o5 mammary cosmos goose booms ofi soc autos xuofi deflated c.3351 when 82o Bu 5 gang 8.: Be 833.8, cocoons «6 macucoflnaum . m madam 00:80.3... cocoa: loosen-E. caEaa so aecon no.2.— oEo accused «0.003 05030.... has. Po: .0 .260 Rocco RP N. illuo>o.._ ~05 «25 533025 c. 080 «coated coor— =o.... rllla£ooaa 25 3.0233 unoE Bo: co tonEaz one 3 Eco» 0+ A liluoob 05039:. «nos. 9.2. .0 onto .oaom noose page“. Eon. E 0mm m has .10.: £03300.) ocunEOwlcoc p.— 080 «coach coo. toEEamiiasooan asoaoota «aoE to» so .3532 «0.02.» usuaoota «not. too so .530 no...» «coupon. nos 553002, oceans coo. actam “.3088 a. 2:0 scouted mum—Down mm... mun—35> 53:2: 20 $133 = suitability irdex for summer foods, SIV6 :- suitability irdex value for percent area in non-forested vegetatim types, sm - suitability index for fall foods, SIV7=suitabilityirdexvalueforpercentareainhardmast producingvegetationtypes, ard SlHl = suitability irdex for percent area inside zones of human influence. \—'—\.¢ Percent area in lowlard hardwood ard lowlard conifer (i.e. cedar swanps) vegetationtypes, asdirectedbytheHSInodel, wereusedto determine the spring suitability index (SISP) . 'me suitability irdex for smaller foods (SISU) was determined for aspen-birch, uplard hardwood, open, coniferous, ard wetlard vegetation types using the following equation: SISU = (3m x SIV3)1/2 where SIV2 = suitability irdex for percent shrub cover of soft mast producing species (includes hazel), ard SIV3 = suitability irdex for number of soft mast producingqeciespresentatz 1%shrub cover. 'Iheremainirgvegetationtypeshadaamnerfood suitabilityirdexof 0.0 because of a lack of soft nest promotion. The suitability irdex for fall foods (SIFA) was determined for uplard hardwoods using the following equation: sm = (SIV4 x SIVS)1/2 where SIV4=suitabilityirdexforbasalareaofhard mastproducirgtreesz40yearsofage,ard 21 SIV5=suitabilityirdexformmberofhardnest proalcingweciespresentwithatleast onematuretreeperOAha. 'n‘lerenainingvegetatimtypeshadafall food suitabilityirdexof o.o becauseofalackofhardmastprcduction. Perwrtareainsideazahofinfluence (SID-II)wascalculatedby2 nhfllodsasdescribedbyRogersardAllen (1987). 'nlefirstmethod carpeted the area of a zone of influence arcurd specific population "sirflts" using the following equation: K zg... [M where z=areaofzoneofinfluence K=mmberofbearskilledatthesinkperyear D=densityofbearsper2.59km2 M=maxinrmalstairebleamnelnortality. Powlation "sinks"areareasofhighlnnnanusea1dlascanpgrands, croplards, ardresidences (RogersardAllen1987). ‘Ihismethodwas amliedtothetamofunnunordbecausethemmberofbearskilledper yearatthetown,thedersityofbears,ardtheneximnnsustaireble ammalnortalitywerealllmown. Beardensityardmaxilmlmalstainable amnelnortalityvereobtainedfranacornlrrentpopllationstldy (Visserl987). 'nlesecordmethoddescribedbyRogersardAllen (1987) assigned fixed-sized circular areas arand towns (102.07 kmz), croplard (38.48 m2),ardresidences(3.80 1on2). 'Ihismethodwasappliedtocroplands ardresidenoessincethemmberofbearskilledperyearatsuchsites wasmllmownard/orlessthanl.0peryear. Allzonesofinfluencewere 22 digitizedusirgARC/INPOardslbsealentlyoverlaidcnrtobearhane rangesardthetflRISvegetatimcoveragenep. Percentareaocarpiedby zones of influence was then detemined. Validation of the black bear HSI nodel was attempted by conparing HSI scores of annual home ranges with each of the following: l)home A range size, 2)1988-89 alb litter size (inclldirg 3 year old feieles that.dud.not produce cubs), 3)cub‘weight.gains (shortly after parturition to 1 year old), and 4)mean daily movanents (Inn/day) . Oarparisons were done using Spear-nan rank correlation tests (Siegel 1956). Vegetatim Sanpling thanshrubcoverof softmastproalcirgspecies, ardmeannumber of softnestproducingweciesweredetermirhdforopen, upland hardwood, ard aspen-birch vegetation types with 20, 20, ard 30 replicate stands, respectively (Table 1) . 'Ihe open vegetation type had thehighestshnlbcoverardgreatestvarietyof softnestproducing species (P < 0.10) corsisting mostly of wild strawberry, serviceberry, ardcaunonchokecherry. Oonsiderablevariance inshrubcoverwas falrd inaspen-bircharduplardhardwoodvegetationtypes. careda ‘buffaloeberry, roundleaf dogwood, ard beaked hazel were predaninately falnd within aspen-bird) stands (Table 2). Upland hardwood stards however, typically contained only red raspberry (Table 2). Ocular estinetes ofpercentshrubcoverofsoftnestproducing species for coniferous, lowlard hardwood, ard wetlard vegetation types were 1.5%, 1.5%, ard 2.5%, respectively. Ocular estimates of the number of soft mast producing species for coniferous, lowlard hardwood, ard wetlard vegetation types were 0.5, 0.5, ard 1.0, respectively. 'Ihe first ard lastshrubstoproducesoftnestforbearsmnrunnordIslardwere Canada buffaloeberry ard viburnums, respectively (Table 2). thanbasalareaofhardnestproducingtreesgreaterthanwyears ofagewasdeterminedfranl? replicatestandsofuplardhardwoods 23 24 Table 1. than (8.12.) percent shrub cover of soft mast producing species, mean (8.8.) nunber of soft nest producing species, ardmean (S.E.) basal areaofhardnestproducingtrees > 40 years of age for vegetation types on Dmmmord Islard, Michigan, 1988. % SHRUB NUMBER or men AREA W SEETFS (111319;th— open 15.2A(1.8) 2.61%.2) 0.0 aspen-birch 11.03(1.9) 1.3B(o.2) o.o upland hardwood 6.7B(2.0) 0.63(o.1) 2.1(o.2) ABValues with different letters within a column are significantly different (P < 0.10) by the Kruskal-Wallis analysis-of—variance test (Siegel 1956) ard the Kruskal-Wallis-type sinultaneals rank test (Miller 1981) . 25 Table 2. Timing of soft mast productim ard vegetation type location formajorshrubspeciesusedbyblackbearsonnnnmord Island, Michigan, 1988 . VEBEIATIONTYPE spears WE IOCM'ION Careda buffaloeberry Mid-June—Early July Aspen-birch (We Mia) Serviceberry Mid-June—late July Aspen-birch, Open We: 899-) Wild strawberry HidnJune—Early July Open (Electric 8139-) Red-osier dogwood Mid-July—Early Aug Aspen-birch (anus stolonifm) Red raspberry Mid-Jilly—Mid-Aug Uplard hardwood (has moses) Cannon dlokedlerry Early Aug—Mid-Aug Open (Ems dramas) Randleaf dogwood Mid-Aug—Early Sep Aspen-birch (games magma) Beaked hazel Mid-Aug—Early Sep Aspen-birch (mus game's) Viburnuns Early Sep—Early Oct Aspen-birch (Elam Spp-) 26 (Table 1). Amricanbeedamasthemlyhardmsttreespeciesoammly ammteredmthestudyarea. Movements and Hare lunges (he trnlsard-eight-mmdred-forty-six bear radio-locations were dartained, 451 (24%) of which involved direct ciaservations. The number of locaticmsperbearrangedfranltoBSwithaneanof 56 locations. Seascaltrendsinmeandailymvaterrtsweremtcbservedforanysex and age-class (Table 3) . However, significant differences were observedammgsexandage—classes forspringandbreedirgtineperiods (P < 0.10) (Table 3). wring spring, adult females with cubs moved significantly less per day than other sex and age-classes (P < 0.10) (Table 3). Adult males, during the breedimseason, myed thenostper __ .___.—a .. ——. ... - . .a __ u................ h... day relative to all other sex and age-classes (Table 3); homeyer, ~this figfiy 'Q’Qfiélanuy different fran male and female yearlings (p < 0.10) . Meananrmlhanerangesize foradultmlesand females (Table 4) were not significantly different. Amnzal hane ranges calculated by the harmonic mean method were significantly smaller than annual home ranges determined by the mininum convex polygon method (P < 0.10) (Table 4). lbanpercentoverlapamngadultmles, meanperomtoverlapbetween adultnalesandadult fanales, ardmeanpercentoverlapamngadult fanales were 60.25% (SE 8 22.02), 86.77% (SE I 1.12), ard 76.92% (SE = 6.26), respectively. Mean percent bears that were active, between 2300 and 0500 hours, during spring, sunmer, and fall were 9% (SE = 9), 20% (no SE), and 27% (SE = 18), respectively. These data do not include 27 Table 3. Mean (S.E.) daily mvanents (km/day) of adult male, adult fanale without cub, adult fanale with cub, yearling male, ard yearling fanale black bears during spring, breeding, and smnmer/fall on Drunmond Island, Michigan, 1988. SW 30% ADMAIE 1.9?(030) 3.5 (0.17) 1.2 (0.33) AD FEMAIE 1.895(0.25) 2.1733(0.37) 2.55A(O.26) AD ma W/wB 0.853(0.24) 1.7953(0.27) 1.65A(0.42) YR ma: - 1.333 (0.26) 1.83A(0.30) YR mma - 1.293 (0.19) 1.52A(0.39) ABValues with different letters within a column are significantly different (P < 0.10) by the Kruskal-Wallis analysis-of-yarianoe test (Siegel 1956) and a modified Kruskal-Wallis—type simultaneous rank test (Miller 1981) using mean ranks. 28 Table 4. Annual hone range sizes (1002) of adult black bears calmlatedbytheminimmomvexpolygmne’diod (MCP), and harmicmeanmethod (HMM) using 95% contours, onDrurmnond Island, Michigan, 1988. ADUDTIMAIES ADUIE'FEMALES IEEHR# 113? fill IBEHR# EKI’ Inn! 190 72.22 68.00 670 64.91 48.07 510 87.00 84.48 630 20.31 17.61 620 67.71 41.76 090 16.56 15.80 490 12.65 11.89 210 66.89 45.27 480 130.00 91.33 770 14.59 10.60 140 15.55 15.94 050 38.70 21.97 040 90.05 82.29 470 78.69 67.21 250 28.80 24.20 MEAN' 75.64 64.75 NEEN’ 48.14 37.68 SE 5.83 12.44 SE 10.91 8.30 29 bearsthatutilizedthegarbagedrmp. Individualsthatusedthe garbagedmpweretypicallymcturnal. Seasaal Habitat Use 'meprtportimofvegetatimtypeswithinseasonallmeranges werethesaneasthosefoxmdoverthestlfiyareaforalladultbears. However, useofvegetatimtypeswithinarnaalhmerangesdiffered from availability for all seasons (P < 0.10) (Table 5). The nunber of adult bears that were succsssfully nnnitored for spring, summer, and fall habitat use were 19, 18, and 16 bears, respectively (Table 5) . Yearlingbearsweremtusedforthisanalysis, andsaneadultswere not successfullyinonitored throughout all 3 seasons due to radio-collar loss, legal and illegal harvest, and natural mortality. This reduced the nunber of usable radio-locations fran 1846 to 1137. Aspen-birch, coniferous, and upland hardwood vegetation types were heavily utilized during spring, with the use of uplarri hardwoods exceeding availability (Table 5) . wring suntner, aspen-bird: and upland hardwood vegetation typeswereheavilyutilizedwiththeuseofaspen-birdiaaoeeding - availability (Table 5) . Mean percent use of wetlard and coniferous vegetation types during fall, were 13% and 36%, respectively, with the use of both exceeding availability (Table 5) . RSI Model Validation 'meHSIsoorefoanmuaidIslardaIdthemeanI-ISI sooreforthe ammal hane ranges of adult bears were not significantly different (Table 6). In addition, major suitability index values for the annual 30 Table 5. Mean percent use and mean percent availability of vegetation typesbyseasmforadultblackbearsmnnmmrllslarfl, Michigan, 1988. areas 0.15 1.91 0.91210 0.00 1.50 garbage amp 1.93“ 0.02 4.46“ 0.03 1.04 0.03 croplands 0.00 0.04 0.00 0.04 0.00 0.04 open 3.61 5.10 4.01 5.26 1.82“ 5.77 aspen-birch 44.09 42.97 58.72“ 40.49 23.77“ 40.62 upland hardwood 17.20* 14.20 14.41 14.46 20.08 15. 11 coniferous 28.69 27.51 14.22“ 28.69 35.70* 27.68 lowlandhardwood 0.80** 4.09 1.50** 4.37 4.54 4.55 wetland 2.54* 3.95 1.72“ 4.43 13.05“ 4.53 x2= 112.73“ x2= 499.78** x2: 47.95“ :Significantly different fran availability (P < 0.10). “Significantly different fran availability (P < 0.05). =Total number of locations. BN2 = Sample size of adult bears. 31 tion types (SIV7) , for fall foods (SIFA) , suitability index value for percent area in hard mast producing veg influence (SIEI) , ard habitat suitability index score (HSI) forallofDnmndeslardandfortheammalhmerangesof suitability index value for percent area in zones of human adult black bears on Drunlnond Island, Michigan, 1988. vegetaticn types (SISP) , suitability index for summer foods Inn-forested vegetation types (SIV6) , suitability index value Suitability index value for percent area in wetland (SISU) , suitability index value for percent area in Table 6. 396061980887378 14113143233431.2221 0.0.0.0.0.0.nw0.0.nmnm0.nwnu.o. 986055 JA88950.0.90.nw6. 6.5 Amnwnwnmnwnmllnmllnwnwo.nm 830503871928893 9,0.95J612332—J311. nU.nU.O.nmnmnmnmo.o.o.o.o.0.0.0. 021.519 86225701. $33232 4..6A4..9~562 nU.0.0.nU.0.0.0.0.0.nm0.0.0.nw0 nwo.o.nu.o.0.nwo.o.o.o.o.o.o.o. 1111111....11111111 210 670 630 090 490 510 480 250 470 040 050 140 190 620 770 0.76 0.30 0.06 0.02 0.48 0.07 0.28 0.01 00 10 Nfiflml S.E. 0.38 0.08 0.40 0.70 0.27 0.28 Drunmond Island 1.0 32 Innerangesofadxfltbearsweremtsignificantlydifferentfrunthe values datained for Drmunond Island (Table 6). The fall food suitability index (SEA) for Drmmri Island was the lowest value relative to other suitability indioes, while the spring food suitability index (SISP) was the highest (Table 6). In addition, bear tunerangesthathadhighSDfivaluesterfledtohavehigherHSI scores. Spearman rank correlation coefficients were not significant for cmparismsbetweenHSIsooreofammalhaterangeswitheadiofthe following: 1)armua1 hone range size (rs = 0.26) , 2)cub litter size (rs = 0.14), 3)cub weight gain (rs = 0.0), and 4)mean daily movements (rS = -0.09) (Fig. 4-7) . 33 :5 150: as J ' a, 3 g 100- m j . . I; - 0.26 or Q . g . 0 O C 01 50 2 ‘ ' j 0 2g 4 e 00. 0 4 . . . . . . a r , 0.0 0.2 0.4 0.6 0.8 1.0 HSI Score Figure 4. Spearmans rank correlation analysis of annual bane range size and overall HSI score (n = 15) for adult black bears on Drummond Island, Michigan, 1988. 34 (D .h‘ (n 5 2- Q 0 .32 —1 g: 0.14 .C) :3 1‘ . o O- O O T l I l fl 0.0 0.2 0.4 0.6 0.8 1.0 HSI Score Figure 5. Spearmans rank correlation analysis of cab litter size and overall HSI score (n = 7) for adult female black bears on Dnmmcnd Island, Michigan, 1988. 35 25' A o q é .. .5 I 0 o O 20' _,_, _ '5. .. "d3 - e . 3 15.. g = 0.0 .o - e :3 O - 10 - °. . . . 0.0 0.2 0.4 0.6 0.8 1.0 HSI Score Figure 6. Spearmans rank correlation analysis of cub weight gains and overall HSI scores (n = 6) of adult female black bears on Drummd Island, Michigan, 1988. 36 3.0 O ’8‘ ‘ o a 1 r = —O 09 J Q ° E . ' ' s :r 20.) . 5 4 . '0 E . . '4 3 ~ 0 2 A g; 1.0 0 ° + D c . 0 q 0 z . 0.0 . . s . . . ; . 1s . 0.0 0-2 0.4 0.6 0.8 1.0 HSI Score Figure 7 . Spearmans rank correlation analysis of mean daily movements and overall HSI scores (n = 15) for adult black bears on Drurmnnd Islard, Michigan, 1988. DISGJSSICN mvanentsaniI-bneRarges Seasaalvariatiminmeandailymovanentsweremtobserved for anysexandage-class. Homever,therewasate1rlencyforadultmales . ..._.... —.___ Mflfiflw--- H'- M ~~_~ __ ,.__.,.._ mum—“um...— .. totavehighdailymvgentsdm'irgthebreedingseasm “Thesmall mmberofadultmalesmnitoredinthissbadymghthaveoa'rtrimtedto the nonsignificant test result. Alt et a1. (1976) fanEQEf-MEQEEE- maledailymovgentsmregreatestdurmgthebreedingseasm,mtth1s WmmwA.‘ was also observed for solitary fgales. Ragers (1987) found that _ agglt -_. Meg-Aw... Ann-n. _‘ fl .... m‘ fgalesixcreasedflienmlymvgentswmenmestrus mvementdata fran Dnmnohd Island could not evaluate this because adult fanales are A 43-h ave-u u, .fimfim ‘fl..‘- “.thm... I!" ~~.c.‘1.¢...‘e~.u« I»: .l-wu my» :4. -'»a lhAestrusformltho3days,milemr-dailymoyeuentdatawere cmbined over a 1 month period. other researdiers, however, believe that adult females should move less per day relative to males and occupyareasjust largeernightoasgreadequatemtrition (Arrstrup ard Beecham 1976, Peltm ard Birghardt 1976) . Sex and age-specific mean daily novanents differed among Drumm'd Islandbearsduringsprirgandbreedingtineperiods. Adult females “*1” r1 um - With cubs exhibited the gallest mean daily mvanents during spring .._._.-._.__ W. .___._.........._ .-~.- m relative to adult females without moibs, and adult males. Thisis . probably the base due to the limited mobility of cubs during spring ”'1‘: 5...... . .L‘L‘k' OJbshavebeenfomfltosuppressthemcvgentsofthemotherforat 37 38 least4ncnthsafterdenanergence (Meyardbieslow1977, Altetal. 1982) . 65331131133115. had high nea_n_ daily mvgents during the breeding“ timeperiod. ‘misisanattgptbynalestomaximizetheirermmters "A... 4'":- ”mm- um MWWW- fin __ . Wmmmmwm- “a... . with receptive fgalgmmammi . .._., (Rogersl987). Adultmalelunerangesteniedtobelargerflianadultfgalehane ranges, similartovtathasbeenfomriinnmerwsshflieselsahere (Erickson and Petrides 1964, Jonkel and Cowan 1971, PcelJcer and Hartwell 1973, Alt et al. 1976, Anstrup and Beecham 1976, Lindzey and Meslow 1977, Rogers 1977, Reynolds and Beecham 1980, Garshelis and Pelton 1981, Kohn 1982, Young and Ruff 1982, Rogers 1987) . The Mm annuallmerangesize foradultmales (75.64105) wassimilartothat M “no... reported fran Wisconsin (71.5 1002) (Kohn 1982), but higher than what wasfamdintheUFperPemnglaofMidiigan (51.71002) (Ericksonand wwwww "WE'- inn—la“. .M ‘II‘va‘ MW Petrides 1964) . In addition, the mean annual hane range size for adult females (48.14 Inn?) was nuch higher than that found in Wisconsin (13. 7 m2) (Kchn 1982)"; that reported fran the Umer Peninsula of Michigan (26 m2) (Erickson and Petrides 1964) . These differences my be attrihrtedtotheseveresxmnerdmghttlatocwrreddurixgfliesufly period, which reduced the summer berry supply. Pelchat and Ruff (1986) fourrlblackbears inAlbertatohavelargerhanerangesizesduring years of food scarcity. However, differences can also be attributed to thedifferentmethodsusedbytheotherstuiiestodelineatehme range. Phlerangesizesestinatedbythehamicneannethodwere cmsistently smaller than those estimated by the mininum convex polygon 39 method. Apotentialproblanwithbothnethodsisthattheyassmean arfimalusesallareaswithintleirrmerangebomflarywrtlmretal. 1989). l-nvever,mnnnunordIsland,ahmerangedelineatedbya minimcmwexpolygmusuallyimludedmreareausedfortravelard nerewatertlenoredelixeatedbyahamicmean. mrt's (1943) definition of hme range excludes migratim mites aspart of a hme “mm 2:155. 'nms, tleharmmicneannethodprobablyrepresentsabetter estimteofrnnerangesize,hrttheminimmcorwexpolyga1nethodis usefulforcmparativepurposes. mnemwerlapwasobsemedbetteensexeeardage- n. we; “Wt-1».- xvi-1:44“? . Wye; . ..r,.,...,p.:~..:rr1 ‘-"‘ 'W’W ”‘3'“; 1V.V-..Ir3 mtvrrrrwr m. ‘mlwaqu glasses, and among sex and age-c1asses.l~lmerarge overlap estimates wereconservativeestimates sixeeallbearewittfintlestuiyareawere not radio—collared. Extensive hme range overlap was also found for black bears in New York, Idaho, Washingtm, 'Ilennessee, Ontario, and North Carolina (Sauer et a1. 1969, Amstrup and Beecham 1976, Lindzey and lbslow 1977, Reynolds and Beecham 1980, GarShelis and Pelton 1981, Kolenosky and Strathearn 1987, Powell 1987) . However, sme studies have reported territorality in adult fenales (Jonkel and Cowan 1971, mller and Keith 1980, Young and Ruff 1982, Rogers 1987) . Where home rangecverlapoccurred, mugbearsofthesanesexardage, tmporal separatimandagcnisticencamterswere, insmecases, observed. Wmidmeebytmpomlseparatimwasalsoreportedbymrffy w and Meslow (1977) and Gar'shelis and Peltolil981) . idditionally, the M “M ems... amount of hme range overlap observed could-be a ftmctim of habitat E352: Garshelis and Pelton (1981) and Rogers (1987) both suggested '- thatfoodahmdamenayinfluereetledegreeofhmerangeoverlap 40 among bears. Powell (1987) observed extensive hme range overlap in adult fanale black bears in North Carolina, and concluded tJEt increased habitat productivity results in m thegg'gitorial M “g“..mM*-W“”fl’mu ' mm behavior. However, Reynolds and Beediam (1980) suggested that the patdiyandmxpredictabledistributimof foodresourcesinaarea resultsinincreasedl'merangesize. Subsequently, thecostof deferriirgtlelmerangefrmotlerbearsirereasestoapointwere defereesbreakdownanihanerangeoverlapensues. Seasonal Habitat Use Habitat selection was not evident when the proportion of vegetation types within seasonal hme ranges (spring, sunner, and fall) werecmparedwithtteproportionofvegetatimtypeswithinthestudy area. 11115 is due to the relatively uniform distribrtim of vegetation typestlmiglnrttlestudyarea, andthelargeamomtofareabearscan traverse. Habitat use, within annual hme ranges, reflected the distrihrtion and abundance of preferred bear foods. Coniferous, aspen- birch, and upland hardwood vegetation types were heavily utilized during spring. I-lowever, the use of coniferous and aspen-birch vegetation types did not exceed availability, unlike the use of upland hardwoods. Bears in early spring forage on grasses and forbs (Hatler 1972 , Kelleyhouse 1980, Grenfell and Brody 1983, Rogers ard Allen 1987 , Rogers et a1. 1988) . 'lhese foods, on W Island, were typically fourdwittfincedarswanps. Asspringprogresses, bears forageon upland forbs, ants, andtlebudsandcatkinsofpoplartrees (Rogers 41 1987, Rogers and Allen 1987, Rogers et a1. 1988). These foods were usuallywithinmamreuplaxflharmoodstandsardmature aspen-birch stands. FbrbsfreqmitlyocmrrirginbearscatcollectedintheUpper Peninsula of Michigan were jack-in-the—pulpit W W) , sweet cioely We 899-). andwild calla (all: Raisers) (Rogers 1989). Developed areas, lowland hardmod, and wetland vegetation types didnotccntainpreferredbearfoods. 'nms, percerrtuseofthese vegetation types was simificantly less than availability. Use of the garbage dunp exceeded availability: however, mean percent use was only 2%, because garbage dunp visits by radio-collared bears were limited to 3 adult males. 'Ihese visits were usually at night when only a limited mmber of locations were obtained (up to 2300 hours). Typically adult males utilized the durnp for only several days at a time before leaving the vicinity. Bears heavily utilized upland hardwood and aspen-birch vegetation typesduringsunnertoforagemammerberries. However, itwasonly the use of aspen-birch that exceeded availability. Regenerating aspen- birdistarflsarflfliimeduplardhardwoodstandsmnnmmdlslardhad more soft nest production relative to older, un-thinned stands. Sumnerberriesverefomriinmstaspenstandsinmrthcentral Minnesota, butsoftmastwasmstabmdantinregeteratirgaspmstands aniopenuplandvegetationtypes (NoyceandOoy 1989). Lindzeyand Meslow (1977) found bears to select clear-cits, that were between 6 and 11 years old, for their abundant smuner berry supply. Moststuiieshavefomdpreferredvegetaticntypesinanmerto have an abundant berry surply (Lindzey and Meslow 1977, Fuller and 42 mm 1980, Grenfell an! Brody 1986, Peldlat and Riff 1986, Yamg and Beedzami986,mswortheta1.1989). Meanpercentshmbcoverand neannmberofeoftmastpmducingspecieswerehiglestintleopen vegetatimtype. Hmevenneanpercmtteeoftheopenvegetatimtype was only 4% which did not exceed availability. 'lhis lack of correspmdereebehveensoftnastahnflameanibearrebitatseleotion maybepartiallyattrihxtedtoinacmraciesintheMIRISvegetation coveragemap. 'IlenonOsanpledstands,delireatedastheopen vegetatimtypebythemluSvegetatimcoveragemp,wereaotually yamg(510yearsold)regereratingaspen-birdmstards. Meanpercent shrubcoverofsoftmastproducingspeciesanineanmmberofsoftmast producing species were usually higher in regererating aspen-birch stardsthanintletypicelcpenvegetatimtype. Regamtingaspen— birch stands usually contained Chnada buffaloeberry and serviceberry, withlesserannlntsofbeakedhazelarrivibtmuus. 'lheopenvegetation type,mtleotlerhard,contai1edwildstrawberry,withcormnon chokederryandserviceberryintleperiptery. 'Ihus,fortheopen vegetatimtype,ueanpercentshrubcoveranineanmmberofsoftmast producing species were inflated by the misclassifimtion of regexeratingaspen—birdistands. Garbagedmmusemringamuerexoeededavailability,bdtuse,as in spring, was limited to adult males. Adult males intensely utilized thegarbagedtmpimediatelyfollowingttebreedingseasm. 'Ihis resulted in limited movement patterns by adult neles. Eighty-four percentofadultmalelocationswerewithinZJonoftlegarbagedlmp frun14Julyt012Septmber. 'Ihegarbagedtmpwasusedrarelyafter 43 12 Septmber. Developed areas, coniferous, lowland hardwood, and wetlarrivegetatimtypesdidmtcontainanahmdantbenywpply; thus, their use was less than availability. Useof conifermsardwetlardvegetatimtypesemeeded availability during fall. Black bears on Dnmnmd Island appeared to utilizecedarswanpsandwetlandsinlatefalltoforagemtheroots andtubersof lowland forbs, andtheber'ries ofcmmonwinterberry holly. In addition, bears for-aged on beechnuts which typically occlrredwithinmamreuplardhardwoodstands. Hardmast isusually coreidered a preferred food item for bears in fall (Piekielek and mrton 1975, Rogers 1976, 1.3111218 et al. 1979, 6315112115 ard Pelton 1981, Grenfell and Brody 1983, Rogers 1987, Rogers and Allen 1987) . Mar, meanpercentuseofuplardhardwoodsdidnotexceed availability. Fall foods were not found within open and aspen-birch vegetationtypes: thus, tiesevegetatimtypeswereusedlessthan availability. HSI Model Validation ‘Ihe overall HSI score for mmmoni Island was 0.27 which inferred poor quality habitat for black bears. 'lhe low suitability index value for fall foods (SIFA) suggested that the quantity and quality of fall foods limits black bear habitat quality on Drunmond Island. Calversely, the cptimal (1.0) suitability iniex value for percent area inforestedwetlardvegetatimtypessuggestedthattlequantityand qtelity of spring foods is least limiting to black bear habitat quality. However, this suitability index value led be viewed with 44 caution since bear use of lowland hardwoods during spring was significantly less than availability. Porcent area in lowland hardwoodswascrecriteriausedtodetermiretlespringsuitability index. Inadditim, tiesprirgalitabilityirdexdoesmttakeinto cmsideratim spring use of uplani hardwoods. 'Ihe cpantity and quality of smmer' foods, vegetation type cmposition, and human influences also limit black bear habitat quality, but to a lesser degree relative to the quantity and quality of fall foods. It appears, for Dnmnend Island, that the spring food suitability index and the human intolerance suitability index strongly influenced HSI scores among adult bears. The spring food suitability index was 1.0 forall bears. Wenthisvaluewasenteredintotheoverall HSI equation, the average of spring, Sinner, and fall food life requisites was appr'oocinately 0.33 (1.0/3) . This was subsequently multiplied by the lumen intolerance suitability index which resulted in an average lBI score of 0.30 with little variatim. 'nlis pattern is from the nethod by which smmer and fall food life requisite scores are obtaired. Sunner and fall food life requisite scores are determired by mlltiplying 2 suitability index values together ($150 and SIV6 for sinner, 8m and SIV7 for fall), note of which were very high. 'Ihus, summer and fall food life requisites were irsignificant relative to the spring food mutability index. Since the wring food suitability index was constant (1.0) among adult bears, but a poor reflection of spring food availability, the lumen intolerance suitability index caused the nestvariabilityanengHSI scores. 'lhus, thepresentl-lSImodel is insensitive to charges in sunner and fall food life requisite scores on 45 Dnmmxi Islam. Black bear density is not a good iniicator of habitat quality. ‘Il'ebeardensitymnnmnaldlslardwasestimatedat 26 bears/lOOka (ireludim orbs) (L. G. Visser, MEIR, pers. cmnun.), which is higher thanmostreporteddensities intheUmerGreatlakesRegim. Kohn (1982) reported the sane bear density in Wisconsin (26 bears/100 1on2) . Rogers (1987) reported a density of 22 bears/100 1on2 in northeastern Minnesota. Erickson and Petrides (1964) observed a bear density of 11 bears/100 ion? in Michigan's Upper peninsula. However, all of these reported dereities cane frml populations subjected to taxman-caused mortality, and the level of mortality varied among these studies. In adiitim, maximum black bear densities are not kncvm for the Umer' GreatlakesRegimaniforotherpartsofthecamtryaswell. Hellgren and Vaughan (1989) rqaorted a bear density of 55-66 bears/100 km2 intheGreatDismal Swanp, butevenatthishighdensityhuman- caused nertality appeared to limit the population. It is therefore inpractical to assess habitat quality by black bear abundance alone. 'lhe level of lumen-caused mortality is difficult to quantify, but has a strong influence on population density. Averagehmerangeoverlapamongadultfenalesisapotential irriicator of habitat quality. Powell (1987) omcluded that increased habitat quality results in increased bane range overlap mag adult fanaleblackbears. Ectereivelnnerangeoverlapwasobservedamong adult females, Whidl aaggested that Drtmmond Island had high quality bear habitat. conversely, Rogers (1977, 1987) found territorality among adult fenales in poor quality habitat. However, lmman influences 46 on bear densities are likely to affect social interactions. 'Ihus, hme range overlap my be an irrlicator of habitat quality on Drurmnoni Island butotherfactorscanalsohaveaninfluencemhmerangeoverlap. Heanageoffirstrepromctimisalsoapotentialindicatorof habitat quality. The average age of first reproduction for bears on Dnmnotrl Islandwas 4.5 years (L. G. Visser, MENR, pers. cmmm.). By cartrast, tleaverageageoffirstrepr'oductimforbearsin mrtleasternMimesotaanieast—central altariowere6.3 and6.0years, respectively (Rogers 1987, Kolenosky 1989) . These areas are generally cuisideredpoorblackbearhabitat. mtheotherhand, mstbears in Pemlsylvania firstreproducedbetween 3 and4 years ofage (Alt 1989). 'll'ms, average age of first reproduction indicated that Drumond Island was not poor quality black bear habitat. Cub litter size was not a good indicator of black bear habitat quality on Dnmmond Island. Mean cub litter size for bears on Drummond Island was 2.19 (SE = 0.16) (L. s. Visser, MINR, pers. cmmun.), which was similar to that found in Michigan's Upper Poninsula (2.15), Minnesota (2.46), and Wisca'ein (2.4) (Erickson and Nellor 1964, Rogers 1976, Kohn 1982) . However, habitat quality apparently differed anong these studies. The mean HSI score for animal hane ranges was not significantly differertfrunfleoverallmlscoreformmldlslam. 'Ihisisto beexpectedsincethevegetative cmpositimwithinseascnalhme rangeswassimilartothat foundoverthewholestudyarea (Table 6). 'misisalsoduetotlestructureofthepresentHSImdel. Spring food and hmnan intolerance life requisites strongly influenced HSI 47 scores iretead of all 4 life requisites (spring foods, Simmer foods, fall foods, and himn intolerance life requisites). No significant correlatims were datained when HSI scores of ammllmerangeswerecmparedwithclbweightgain, cublitter size, hme range size, and mean daily mvme'rts. lbgers (1976) found that a scarcityofsumnerandfall foods resaltedinretardedmbweightgains and low clb production. It was assured that overall habitat quality influencedtheseparaneters. WhencmparingHSI scorewithcublitter size and HSI score with cub weight gain, a valid HSI model should yield positive correlations. However, significant correlations were not obtairedinthissuxiy, butsanplesizeswereverysmall. Additional dataareneededtotesttheHSImodel inthismanrer. Young and Riff (1982) saggested that the hme range sizes of adult fanale bears could be a means of chparing habitat quality among populations. Similarly, Sanderson (1966) suggested that there is probably an inverse relationship between hane range size and habitat quality. mislesnotbeendemonstrated forblackbearhmeranges, buthasbeenillustratedwiththeterritory sizes ofsongsparrows W M) (Ricklefs 1979) . ‘Iherefore, it was expected that a validHSInedelwouldslmaregativecorrelatimbetweenHSI scoreand annualhmerangesize. However, thiswasnctobservedforthisstudy. Correlatim analysis was also coriducted on nean daily movements (km/day) . Anstmp and Beecham (1976) found an inverse relationship betweenfoodabundameandneandailywvments forbears inIdaho. 'Ihus, itwasexpectedthatanegative correlatimvmldexistbetween lBI score and nean daily Wits. However, a significant correlation 48 was not (heaved. Inadditim, probluwereermmteredwlenusingttems vegetatim coverage map for dataining habitat variables required by the HSI model. ‘lhe classificatim scheme used for the MIRIS vegetation coveragenepfailedtorecognizeageclassesmflstockingdersitieeof forestedstands, whidmisrecessarytoreducevariancesofneamred habitat variables and gore accurately explain black bear habitat use. Apotentialsolutimtothisproblanistoimorporatestardagearxi stocking density to the MERIS classification sdieme. An additional modification, to irxzrease accuracy, is the use of an ecological land classification schme. An ecological land classification schme takes into consideration the effects of soils and site factors on the vegetative cmposition of an area. For example, the understory species cmpositim of 2 aspen-birch stands of identical age and stocking density will often be different. However, these differences can often be attributed to soils and site factors. Finally, inaccuracies were found in delireating and classifying certain vegetatim types. For exanple, 10 of 20 open stands sanpled frantleMIRISvegetatimcoveragenepwereactuallyyamg (5 10years old) regenerating aspen-birch stands. 'Ihis caused minor errors in HSI scoredetermination, sincemlysikofttestudyareawasdelireatedas theopenvegetationtype. ‘Ihus, evenwithasofiterrorratein delineatingthisvegetatimtype, mly2.5% ofthetotalareawouldbe affected. A vegetatim coverage nap with greater detail and accuracy shouldbeutilized forfutureresearohtoreducevariances ofneasured habitat variables and reduce error in 161 score determination. ‘u 49 'meblackbearlBImdelslmldbealteredinanattmptto inproveitsaccuracy. chentareainforestedwetlandvegetatim typesisapoordeterminantoftlespringfoodliferequisitescore. Bearsinqaringuseawidevarietyofvegetatimtypestoforagema diversity of foods. Bearsinearly spring forageongrassesardforbs (Hatler 1972, Kelleyhouse 1980, Grenfell and Brody 1983, Rogers 1987, RogersandAllen1987,Rogersetal.1988). Asspzringprogresses, bearsocwpyuplardareastofeedmuplardforbs,ants,anittebxis andcatkins of poplars (RogersandAllen 1987,Rogersetal. 1988). In addition, spring use of lowland hardwoods (a forested wetland vegetation type) by bears was less than availability (Table 5). Spring habitatqualityshouldnotbeneasuredsolelybypercentareain forested wetland vegetation types. Habitat variables that actually measmespringfoodabtmdareeshouldbeireorporatedintothemdel, arritlesensitivityoftlenedeltotlesefoodsrefired. 'Iherelationshipbetweenbasalareaarflhardmastabmriance(SIV4 inHSImodel)shouldalsobealtered. Itemininumoptinalbasalarea-‘t intrepresentblackbearHSInedelagpearstobetoohighforuplarxi hardwoodstanisintleUpperGreatIakesRegim. RogersandAllen (l987)basedtleirassnnedoptimalbasalareamuplardhardwoodstands in Petmsylvania. Species cmpositim and basal areas of upland hardwoodstardsinPemsylvaniaaredifferentfrunthoseoftleUpper GreatlakesRegion. Adiitionalreseardlisreededtodetermirette prqaerrelatimshipbetweenbasalareaanihardmastabmflaree. In additicn, the fall food life requisite score has little inpact on the overall HSI score because the fall food life requisite score is 50 obtairedbymltiplyinnguitability indicestogetl'er (SMarflSM). 'nepresentlBImdelalsofailstotakeintocmsideraticnotherfall fooditens. BlackbearstnmrlIslandselectedcedarswanpsand nm-forestedwetlardvegetatimtypesdurimfalltoforagemthe tubersandrootsof lowland forbs. Anethodology shouldbedeveloped that accuratelymeasures the abundance ofthese lowland foods, orthat incorporatesthesevegetatimtypaintothemdel. 'lhepresentHSImodelsuggeststhatoptimalbearhabitatis cmposed of 25 to 50% non-forested vegetation types (SIV6 in present HSImodel). 'Ihisisapoorassrmptimsireemtallnon—forested vegetaticntypesproduceanabmrianceofamrerberries (i.e. herbaceousrangelands). InanearlierdraftoftheHSImodel,itwas assmedthatoptimalbearhabitatwascmposedofSMsrmmerfood producingvegetatimtypes(Rogersetal.1986). 'Ihisimludednm- forestedaswellasforestedvegetationtypes. 'Ihisprdoably representsabettermeamreofSIV6thanwhatispresentlyusedinthe HSIrodel. 'nemdelslmldbealteredsothatitisassrmedthat ideal black bear habitat contain at least 50% m food-producing vegetationtypes (Fig. 8). Simmer food producing vegetationtypes are thosethathave>10%softmastshrubcover. 'Ihe10%shrubcover criteria was selected because this produced a suitability index score for softmastahlrdance(SIV21nI-EI model) of 0.50. 'lhealtered suitability index increased the value of SIV6 for Dnmmond Island fran 0.38toO.97,andsubseqLentlyincreasedt1eoverallHSI score frmi O.27toO.30. Measurement and mlculation of the hman intolerance life Suitability Index (SIV6) Figure 8. 51 P or .0 q 9 m 0.3 '- 9 no 0.1 '1 O O I I I I I 25 35 50 65 75 100 Percent area in summer producing vegetation types (soft most shrub cover > 10%) Proposed relationship between smuner vegetation type cmposition suitability index score and percent area in sumer food-producing vegetation types. Dotted line shows non-forested vegetation type relationship given by present black bear HSI model (Rogers and Allen 1987) . 52 reqnisite score slmld also be altered. Variables utilized in the presentmlmodeltodetermirethel'nmanintolerance life requisite score are typically difficult to obtain by most wildlife managers. Bear density, number of illegally killed bears, and naximm sustainable annual mortality are usually unknown in most manageient situations. An alternative method outlired by the HSI nnodel assigns fixed sized cirmlarareasaromdtowne, crcplards, ardresidences, whichare easily delireated, but fail to take into cmsideration regional differences in l'nlman attitudes towards bears. In addition, the data suggestthatthefimedsizedareamethodrepresentsaworstcase scenario. 'lhehnmanintoleraree za'eamndthetwnofvmnmd, as calculated by bear density and mortality rates (14.58 ka) , was much analler than its appropriate fixed sized area (102.07 1on2). This dmuetratedtlettlefixedsizedareametledcandrasticallyreducean overall I-BI score for an area. Primary road ard dwelling density are more appropriate measures of blackbearhuman intolerance. Roaddensity iseasilyneasuredardhas been used to measure habitat effeCtiveness for Rocky Mountain elk (mass m M) (Inyon et al. 1985). Dwelling density is also easilyneasured, butcareshouldbetakentotallyonlytl'iosedwellings thatareoccupiedyear—rumd. Mostblackbearmisancecmplaintsm Dnmdlslandcane franpermanentresidents, whichsuggestedthat seasorelresidentsweremretolerantofbearsthanpermanent residents. Thus, seasonal dwellings are not likely to negatively inpact bear populatiae. Suitability index canvas for road ard dwelling density were modified frmi Rogers et al. (1986) , and these 53 mitability curves can be easily modified to take into account regional differences in human attitudes towarck bears (Fig. 9-10) . The Innnan intolerance life requisite score should be calculated by determining themeanofthe2 suitability irdices. 'Iheprimaryroaddensityard dwelling density on ormnond Island was 0.66 non/2.59 km2 and 2.1 dwelling/2.59 m2, respectively. The altered as: model increased the hnmnan intolerance life requisite score for Drunnerd Islard frmn 0.70 to 0.82. The overall mI equation should be altered to better reflect black bear ecology. A proposed HSI equation is: HSI = (SISP + 2((SISU + SIV6/2) + (sm + SIV7/2))/5) x SIHI where SISP = suitability index for spring food abundance, SISU = suitability irdex for snmlner food abundance, SIV6 = suitability index for percent cmnposition of Simmer food- produClng vegetatlon W905, SIFA = suitability index for fall food abundance, SIV7 = suitability index for percent cmposition of fall food producing vegetation types, and SIHI = suitability index for human intolerance. 'neprqlosedHSquuationgivesnereweighttosnmnerardfall food life requisite scores. Summer and fall foods strongly influence cub growth, cub production, and subadult survival (Rogers 1976). Spring foods, mtheotherhard, arehighinproteinbutdonctresultin weigntgainsrecessaryforwinterdornencyandreproductivesuccess (Jcnkel andcowan 1971, Iarders etal. 1979, Kim 1982). 'neproposedHSquuationasslmesthatfoodahndanceardpercent cmposition of food producing vegetation types have equal weight when ‘ 8) 9 GI n 0.3 4 Suitability Index (SIV .0 _a l O o.o 1.'6 3.'2 4's 6.4 Ian/2.59 m2 0.0 1.0 2.0 3.0 4.0 mn/mn Primary road density Figure 9. Proposed relationship between human intolerance suitability index score and primary road density for black bears in the Upper Great lakes Region (modified frmn Rogers et a1. 1986). 55 Suitability Index (SIV9) .0 .a l O I I l 5 10 15 20 Number of permanently inhabited dwellings/2.59 km2 0 Figure 10. Proposed relationship between Imman intolerance suitability index score and permanent dwelling density for black bears in the Upper Great lakes Region (Inodified frmn Rogers et a1 1986). . 56 calallating snmner and fall food life requisite scores. 'Ihus SISU and SIV6areaveragedtodeteminetl'esnmnerfoodliferequisite, and likewisesmandSIWareaveragedtodeterminetlefall food life requisite score. 'ne present HSI model multiplies both SISU and SIV6 together, andsmardSIWtogetler, whidlserveasreduction functims, ratler than modifying functims (i.e. (0.2 x 0.2) < (0.2 + 0.2)/2). These nedifications increased the l-BI score for Drunnmond Island from 0.27 to 0.36. Average age of first reproduction, nean cub litter size, and mean hane range overlap among adult fennales indicated that Dnmnord Island should have an HSI score of approximately 0.60. Based onthefindingsofthisstnldy, tnelBImodelreedssmeadded modifications. Adiitional research on paraneters of the model will l'elp clarify necessary modifications. W108 1. Significantseasanltrerdsinmeandailymvanenrtsweremtfomd forbearstnmnondIsland. I-hvever,therewasatendencyforadult malestohavehighdailymvanentsduringthebreedingseasm. 'Ihe snallmmberofadultmalesnmitoredinthissudyprobably contributed to the nonsignificanrt test result. 2. Sex and age-specific mean daily movanents differed andng bears during spring and breeding time periods. Adult fanales with cubs mved significantly less per day relative to other sex and age classes during spring. Adultmaleshadhighdailymovementsdmingthe breeding season, but this result was only significantly different from male and female yearlings. 3. Adultmalehmerangestendedtobelargerthanadult femalehane ranges, but suall sample size appeared to keep differences fran being statistically significant. 4. Honerangesizesestimtedbytheharmmicneannethodwere significantly snaller than those estimated by the miniuum convex polygonmethod. 'meharnmicmeannethodprobablyrepresentedabetter estimteofhmeraxgesizebuttheminimmcmvexpolygonmethodis useful for carparative pirposes. 5'7 58 5. Ehrtensivetnnerangeoverlapwasdaservedbetweensexesandage— c1asses,andamorgsexandage—clm. 6. Habitat selection reflected the distribution and abundance of preferred bear foods. 7. During spring, aspen-birch, coniferous, and upland hardwood vegetatim types were heavily utilized, with the use of upland hardwoods exceeding availability. The garbage dunp was also preferred, primarily by adult males. 8. Aspen-birch and upland hardwood vegetation types were heavily utilizedduringsmmer, withtheuseofaspen-birchexceeding availability. 'Ihe garbage dunp was also preferred but the majority of use was limited to adult males. 9. Bring fall, bears utilized upland hardwood, wetland, and coniferous vegetation types, with the use of wetland and coniferous vegetation types exceeding' avail' ability. 10. PeI'cent use was less than availability for most vegetation types that lacked preferred bear foods. 11. No significant correlations were obtained for oarparison of HSI scores with each of the following: a)hane range size, b)cub litter size, c)wb weight gain, and d)mean daily mvements. In addition, the HSIscorefoandeslanddidmtoonespondwelltoaverageageof first reproductimandmeanhcnerangeoverlapammgadult females. WP W106 Habitatmanaganentpractices slmldmhancetheproductim of preferredbearfoods. Cedarswanps,mamreuplandhardwoodstands, and aspen-birdistanrkstnildbemaintairedtosupplybearewithspring foods. Cedarswanpsprovidebearswithgrassesandforbswhidmare preferredearlyspringfoods. Asspringprogresses,bearsfeedon ants, forbs,andthebtdsandatkinsofpoplartrees(Rogersand Allen 1987, Rogers et al. 1988) , which typically occur within mature aspenr-birdnanduplandhardwoodvegetatim types. Onttingssnmldbeoonductedonaspen-birchanduplandhardwood vegetatimtypestoenhanoesoftmastproduction. Regeneratingaspen— birdistandsandthinxnedmlandhar'dwoodstandsmdeslandhad moresoftmastproductionthanolder,m—thimedstands. Lindzeyand Meslow (1977) foundbearstoselect clearcuts, thatwerebetweené andllyearsold,toforageonsmnnerberries. NoyoeandOoy(1989) fomdsoftnasttobemostabundanntinregeneratirgaspenstandsard open upland vegetation types. Selective cutting of non-mast producing woodyplantsslnfldbepractioedmpermanentopeningstoemanceor mintainsoftmastproductim. Parmanenntopeningscontainedwild strawberry,withserviceberryandoaunondmokedierryfandatthe perman- Manneuplandhardwoodstandsslmldbenaintainedtoprovide 59 60 bears with beedmnts, a preferred fall food. Since upland hardwoods, mDrmnondIsland, aredaninatedbyAnericanbeednandsngarmaple, light selection ants favoring American head: can be utilized for tinber regeneration. However, thesawlogbasalareashouldnctbelessthan 4.6 m2/0.4 ha (Tubbs 1977). Mature cedar swanps and run-forested vetlanrk should be preserved to supply bears, during late fall, with roots and tubers frcm mesic plants. Ideal black bear habitat shalld be canposed of 15% spring food- producing vegetation types, 50% winner food-producing vegetation types, and35%hardnastproducingvegetationtypes (Rogersetal. 1986). Cedar swanps, aspen-birch stands, openings, non-forested wetlands, arduplandhardwoodsproducedavarietyofpreferredbear foods. Cedar swanps supplied bears with early spring grasses and forbs. In addition, during late fall, starchy roots and tubers were supplied by both cedar swamps and rum-forested wetlands. Awen-birch standsmtmlyprovidedawenleavesandcatkinsaslatewring foods, hntalsoproducedanabmdantbenyswplyduringammer, especiallyin regenerating stands. Rernanent openings, especially at the periphery, supplied additional sunrner berries. Upland hardwoods provided bear foods year-round. wring late wring, upland hardmods supplied bears withinsectsandforbs. Raspberriesandotl'ersoftmastproduoerewere available in upland hardwoods in W, usually where logging activity occurred. Finally, during early fall, beechnuts were produced within mature stands. mnnanencroadnnentmbearhabitatisaeoftlemajorreasms for decreasedbearnnmbers (Rogers 1987). Bears are secretive, and 61 generally avoid human activity centers and dwellings unless a reliable food source is available. Bears that utilize artificial food sources (i.e. garbage) aremrelikelytoamnmtomman-cnused mortality (Rogersl987,RogersandA11en1987). 'nms,htmandevelo;nent should berestrictedwherepossible. Inaddition,roadsoonstructedasa result of logging activity shmld be closed to minimize Imman dismrbanmsandpoadiingqporumities. 'IheblackbearHSInodelsnmldberevisedandre-evaluatedin ternsoftndthemdeloorrespondstowringandfallblackbear habitat preferences. 'Ihe present HSI model evaluates wring food abndandethmghpercenntareainforestedwetlandvegetationtypes. However, meanperoentuse of lowlandhardmods (a forested wetland vegetation type) by bears was significantly less than availability. In addition, fallfoodahndanceisevaluatedinthepresentHSInddelby quantifyinghardnastproductim. However, bearsduring fall, preferredrm-forestedwetlandsandcedarswanpstoforagemstardmy rootsandtubers. 'n'epresentHSImodelfailedtoaddressrootsand tubersasfallfoods. Itemdelshouldalsoberevisedintemsofits quantification. 'IheIBI score forDerlnond Island (0.27) wasvery low, anddidmtcorrespondwelltoaver'ageageoffirstreproalctimand meanlnnerangeoverlapamongadultfanales. Basedonthefindingsof this study, theHSImodel needsnodificntionetomreaoanrately reflect black bear habitat quality. mm m, D. A., S. R. 115mm, AND B. V. m. 1986. Regional landscape ecosystens of Michigan. School of Natural msouroes, Univ. of Mich., Ann Arbor. 32 pp. AIII', G. L. 1989. Sane awects of fanale black bear reproductive biology in northeastern Pennsylvania. Int. Conf. Bear Res. and Manage. (in press). , ANDJ. H. mm. 1984. miseofblackbeardens in northeastern Pennsylvania. J. Wildl. mnage. 48:236-239. , F. w. m, AND J. 8. my. 1976. Home range and activity patterns of black bears in northeastern Pennsylvania. Trans. Northeast Sec. Wildl. Soc. 33:45-46. , G. J. MA'IUIA, JR., F. W. AUI‘, ANDJ. E. LTNIEEY. 1980. Dynamics ofhanerangeardnovanentsof adultblackbears in northeastern Pennsylvania. Bear Biol. Assoc. Oonf. Ser. 3:131- 136. , R. J. AUKEY, D. E. m, AND D. C. MADL. 1982. Mcvanents and nortality of black bears relocated 100i- air-miles in Pennsylvania. Trans. Northeast Sec. Wildl. Soc. 39:16. W, S. C., AND J. BEECHAM. 1976. Activity patterns of radio collared black bears in Idaho. J. Wildl. Manage. 40:340-348. ARR-10R, S. M., W. B. W, ANDJ. R. GILEERI‘. 1989. Bane range characteristics of adult fishers. J. Wildl. Manage. 53:674-679. BAIIARD, W. 8., T. H. SHAKER, AND K. P. m. 1981. Causes of neonatal noose calf nortality in south central Alaska. J. Wildl. Manage. 45:335-342. mu, L. E., ANDM. R. mm. 1980. Seasonal foods and feeding ecology of black bears in the Snoky Mnntairs. Bear Biol. Assoc. Calf. Ser. 23:141-148. EJRSER, B. 1988. Which way did he go? wecial sensors help to quantify activity data. 'Ielcnics Quarterly 1:1. 62 63 EJRI‘, W. H. 1943. 'Ierritoriality and has range conoepts as applied to manuals. J. Manual. 24:346-352. CANFIELD, R. 1941. Amlicatim of the line interception method in sanpling range vegetatim. J. For. 39:388-394. DAVIS, R. L., AND C. J. FREY. 1984. Soil resource inventory Sault Ste. Marie and St. Ignace ranger Districts. USDA Forest Service. Hiawatha Natigxal Forest. Escanaba, MI. 284 pp. 013m, K. R., AND J. A. GIAHIAN. 1980. Karmic mean measure of animal activity. Ecology 61:1040-1044. mm, A. W. 1964a. An analysis of black bear kill statistics for Michigan. Pages 68-102 in A. W. Ericksm, J. Nellor, and G. A. Petrides. The black bear in Michigan. Mid). State Univ. Agric. mp. Sta. Res. 3.111. 4. 102 m. . 1964b. Breeding biology and ecology of the black bear in Midnigan. Ph.D. 'Ihesis. Mich. State Univ., East Lansing. 274 pp. , AND G. A. PEIRIDES. 1964. Population structure, novenents, and nortality of tagged black bears in Michigan. Pages 46-67 in A. W. Erickson, J. Nellor, and G. A. Petrides. ‘Ihe black bear in Michigan. Mich. State Univ. Agric. Exp. Sta. Res. Bill. 4. 102 pp. , AND J. NEIIDR. 1964. Breeding biology of the black bear. Pages 5-45 in A. W. Erickson, J. Nellor, and G. A. Petrides. The black bear in Michigan. Mich. State Univ. Agric. Exp. Sta. Res. mu. 4. 102 pp. FULLER, T. R., AND L. B. KEITH. 1980. Sunmr ranges, cover-type use, anddemingofblackbearsnearFortmnn'ray, Alberta. Can. Field Nat. 94:80-83. GARSHELIS, D. L., AND M. R. mm. 1980. Activity of black bears in Great Smoky Mountains National Park. J. Marmal. 61:8-19. . 1981. Movenentsofblackbears intheGreat axoky LunAnthim National Park. J. Wildl. Manage. 45: 912-925. GRENFELL, W. 13., JR., AND A. J. EDDY. 1983. Seasonal foods of black bears in 'ral‘oe National Forest, California. Calif. Fish and Game 69:132-150. GRENFEIL, W. E., AND A. J. m. 1986. Black bear habitat use in ‘I‘atoe National Forest, California. Int. Conf. Bear Res. and Manage. 6:65-72. 64 HAIEEN, K. D. 1988. Climatological data annual summary Michigan. U.S. Dept. of cameos, Natienal Oceanic and Atmospheric Administration Environmental Data Service. National Climatic Center, Asheville, NC. 42 m. m, D. J. 1972. Food habits Of black bears in innterior M3813. HAYS, R. L., C. S. W, ANDW. SEI'I‘Z. 1981. Estimting wildlife habitat variables. U.S. Fish Wildl. Serv. Biol. Serv. Program. meme-81A"). 111 pp. HELIGREN, E. C., AND M. R. VAIIEHAN. 1989. Danographic annalysis of a black bear pwulatian in the Great Distal Swamp. J. Wildl. Manage. 53:969-977. HERRERD, S. M. 1979. Black bears: the grizzly's replacenent? Pages 179-195 in D. Dirk, ed. The black bear in nodern North America. Boane and Crockett Club. Ansell Press, Clinntan, N]. am, c. J., AND I. Kim. 1971. 'Ihe black bear in the wruce- fir forest. Wildl. lbrogr. 27. 57 pp. W, D. G. 1980. Habitat utilization by black bears in northern California. Bear Biol. Assoc. Conf. Ser. 3:221-227. KING, D. L. 1967. A bear kills a fawn. Can. Field Nat. 81:149-150. mm, B. E. 1982. Status and management of black bears in Wisconsin. Wis. Dept. of Nat. Resour. Tech. Bull. 129. 31 pp. mm, G. B. 1989. Reproductive biology of black bears in east- central Ontario. Int. Conf. BearRes. andManage. (in press). , AND S. H. mm. 1987. Black bear. Pages 442-454 jg} M. Novak, J. A. Baker, M. E. Gabard, and B. Mallach (eds), Wild furbearermanagemenntandconservation inNorthAmerica, Min. of Nat. Resour” Ontario, Canada. IANIIRS, L. L., R. J. W, A. S. JCI-INSCN, ANDR. L. W. 1979. Foods and habitat of black bears in southeastern North Chrolina. J. Wildl. Mnage. 43:143-153. IARSCN, D. G., D. A. M, AND R. L. MARKEL. 1989. Causes and rate of noose mortality in the southwest Yukon. J. Wildl. Mannage. 53:548-557. IINDZEY, F. G., AND E. C. mm. 1976. Characteristics of black bear dens an Long Island, Washington. Northwest Sci. 50:236-242. ,AND _. 1977. Home range and habitat use by black bears in southwestern Washington. J. Wildl. Manage. 41: 413-425. 65 man, J. L., '1‘. N. 1m, J. P. WEIGAND, C. L. MARQJM, W. D. Em, J. D. JQIES, D. W. m, AND L. L. HIQG. 1985. Cbordinatin'g elk and timber managenent. Mmtana Dept. of Fish, Wildlife, and Parks. Bozenan, Mr. 53 pp. MANVIIIE, A. M. 1982. I-hmnaninpactgntheblackbear inMichigan's Lower Peninsula. Ph.D. thesis. Mich. State Univ., East Lansing. 188 m. mews, N. E., ANDW. F. m. 1988. Blackbearpredatign of white-taileddeerneenatesintheoentralAdinondacks. Can. J. 2001. 66:1241-1242. MILLER, R. G. JR. 1981. Sinnltaneous statistic infereoe, 2nd ed. wringer-Verlag, New York, NY. 299 pp. MICHIGAN MARIMENT OF NATURAL W. 1988. Black bear nanagement plan (draft). Mich. Dept. Nat. Resour., Wildl. Div. 27 pp. MICHIGAN SIA'IE runner CREATIONS INVENKRY. 1982. Michigan state forest operatigs inventory. Cmpartment and stand evaluation. .Manage. Div., Mich. Dept. Nat. Resanr., Iansing. Mi. ‘Iypewritten manuscript. 50 pp. m, C. C. 1947. Table of equivalent populations of North American stall namnals. Am. Midl. Nat. 37:223-249. mYCE, K. V., AND P. L. (DY. 1989. Abundanoe and productivity of bear foodweciesinforestccvertypesofnorthcentralMimnescta. Int. Cenf. BearRes. andManage. (inpress). C'PEZIO, J. 1980. Unraveling the mysteries of bear damning. New York State Dept. of Envir. Conserv. 4 m. 0m, J. J., AND L. J. VEEME. 1982. Prefiatim byblackbears on newborn white-tailed deer. J. Manual. 63:695-696. W, W., AND T. S. BJRIW. 1975. A black bear pqmlation study in northern California. Calif. Fish Game 61:4-25. mm, M. R., AND G. M. m. 1976. Black bears of the amides. Nat. Hist. 85:54-63. mm, B. 0., AND R. L. IUFF. 1986. Habitat and spatial relaticrships ofblackbears inborealmixedwoodforestof Alberta. Int. Cenf. Bear Res. and Manage. 6:81-92. POELKER, R. J., ANDH. D. ML. 1973. lbs blackhear of Washington. Washington State Game Dept. Biol. Bill. 14. 180 pp. 66 PWELL, R. A. 1987. Black bear bane range overlap in North Carolinna andtheceoeptcfhanerangeappliedtcblackbears. Int. Conf. Bear Res. and Manage. 7:235-242. REYNOHB, D. G., ANDJ. J. M. 1980. Hans range activities and reproductign of black bears in west-central Idaho. Bear Biol. m. m. We 3:181-1900 RICKlEFS, R. E. 1979. Ecology. Chircrn Press, Inc. New York, NY 966 pp. m, L. L. 1976. Effects of nest failures an survival, growth, and ve success of black bears. Trans. North Am. Wildl. Nat. Rescur. Cmf. 41:431-437. . 1977. Social relationships, novenents, and population dynamics of black bears in northeastern Minnesota. Ph.D. Thesis. Univ. of Minn., Minneapolis. 194 pp. . 1987. Effects of food supply and kinship on social behavior, novenents, and population growth of black bears in northeastern Minnesota. Wildl. Monogr. 97. 62 pp. . 1989. Progress report Great lakes bear habitat study. USDA For. Serv. North Central mp. Sta. 56 an. , and A. W. ALIEN. 1987. Habitat suitability index nodels: black bear, Upper Great lakes Region. U.S. Fish Wildl. Serv. Biol. Rep. 82(10.l44). 54 pp. , D. W. RUBIN, A. W. ERICIGw, E. M. RAE-ER, L. J. VERIE, AND J. J. CZCGA. 1976. characteristics and managenent of black bearsthat feedingarbagedunps, cannpgrounds, crresidential areas. Pages 169-175 in M. R. Pelton, G. E. Folk, J. W. lentfer, eds. Bears-their biology and management. Proc. 3rd Innt. Conf. Bear ms. andManage. , J. W. m, D. M. ELSING, E. FLEGIER, J. POSSUM, J. HAM’IIIL, J. B. HAUFIER, J. mason, B. IRVINE, M. IOIAR, B. 0111M, T. REIS, J. STUHI', S. mum, AND L. G. VISSER. 1986. Habitat suitability index nodels: black bear, Upper Great lakes Region. Review copy. 70 pp. , G. A. WIIKER, ANDA. W. ALIEN. 1988. Managing northern forests for black bears. Pages 36-42 in T. W. Hcekstra, and J. Capp eds. Integrating forest managenennt for wildlife and fish. USDA For. Serv. Gen. Tech. Rep. NC-122. 63 pp. SAW, G. C. 1966. The stndy of manual novennents - a review. J. Wildl. Manage. 30:215-235. 67 SAUER, P. R., S. L. FREE, AND S. D. m. 1969. Movenents cf tagged black bears in the Adiredacks. N.Y. Fish Game J. 16:205-223. SIEGEL, S. 1956. Neparametric statistics for the behavioral scieoes. MoGraw—Hill Book Co., New York, 312 pp. SMITH, R. P. 1985. Are we taking too nanny bears? Michigan Cut-cf- dcors 39:64-67. SIEEL, R. G. D., ANDJ. H. mm. 1980. Prinnciples and procedures of statistics, a bianetrical amroach. 2nd ed. MoGraw—Hill Book Co., 633 pp. S'IUENER, F. W. 1957. History of bear regulations. Mich. Dept. Conserv. Game Div. unpubl. rept. 9 pp. TIEI‘JE, W. D., AND R. L. HJFF. 1980. Defining behavior of black bears in boreal forest of Alberta. J. Wildl. Manage. 44:858-870. 'IUBBS, C. H. 1977. Manager's handbook for northern hardwoods in the north central States. USDA For. Serv. Gen. Tech. Rep. NC-39. 29 p. M, J. W., J. J. EEO-IAN, AND L. R. IRBY. 1989. Fanale black bear habitat use in west-central Idano. J. Wildl. Manage. 53:668- 673. VEA'IO-I, J. 0., L. R. SCI-OW, A. L. GRAY, C. S. smnms, AND 2. C. ESTER. 1927. Soil survey of Chimewa County, Michigan. U.S. Dept. of Agric. Bireau of Chen. and Soils. 44 pp. VERSPCCR, E. 1983. Black bear, 111335 anger—menus, predation on a mule deer fawn, Megs M. Can. Field-Nat. 97:114. VISSER, L. G. 1987. Drnnnnnond Island bear population. Mich. Dept. Nat. Rescur. Pittman-Robertson Prcj. Rep. W-127-R-5, Job 51.3. 5 pp. W, G. S. 1982. Contents of black bear scats fran the central Adirondacks in late sunnner. N.Y. Fish Ganne J. 29:210-213. YQJNG, D. D., ANDJ. J. BEIEGIAM. 1986. Blackbear habitat use at Priest lake, Idal'o. Innt. Cunf. Bear Res. and Mannage. 6:73-80. mum, B. F., AND R. L. RUFF. 1982. Population dynamics and novenents of black bears in east central Alberta. J. Wildl. Manage. 46:845- 860. ZY'IAKJK, B., AND D. CERIWRIGIT. 1978. Black bear novenennts and food habits in New Brunswick. Pages 227-239 in) R. D. Hngie ed. Proc. Fourth East. Black Bear Workstop. Greenville, Maine.