Brown Bear Summer Use of Alpine Habitat on the Kodiak National

BROWNBEAR SUMMERUSE OF ALPINEHABITAT ON THE KODIAKNATIONAL WILDLIFEREFUGE

GERRY ATWELL,U.S. Fish and Wildlife Service, Kodiak National Wildlife Refuge, Kodiak, Alaska 99615' DANIEL L. BOONE, U.S. Fish and Wildlife Service, Kodiak National Wildlife Refuge, Kodiak, Alaska 996152 JACK GUSTAFSON, U.S. Fish and Wildlife Service, Kodiak National Wildlife Refuge, Kodiak, Alaska 996153 VERNON D. BERNS, U.S. Fish and Wildlife Service, Kodiak National Wildlife Refuge, Kodiak, Alaska 996154

Abstract: Brown bear (Ursus arctos middendorffi)alpine summerhabitat use patternswere studied at the Kodiak National Wildlife Refuge in 1973, 1974, and 1975. Seven plant communities were describedand mappedwithin the alpine and subalpinezones of the 56.5-km2 study area. Single bears and family groups showed an almost exclusive preference for Carex macrochaeta as the primaryfood. Alpine activity areas, determinedfor 29 individuallyidentified bears, were small; those of lone adults were twice the size of those females with young, 1.70 and 0.85 km2, respectively. The averagedensity was 0.85 bear/km2but rose to 2.60 bears/km2in an area where animalsconcentrated to feed. Bears spent 5-6 weeks in the high mountains, abruptlydeparting when young Carex macrochaeta plants were no longer being produced.

In Alaska, little effort has been directed towards Grizzly bears (U. a. horribilis) also utilize alpine studying brown bear use of alpine habitat. Alpine in- vegetation. In Canada, Mundy and Flook (1973) report vestigations have been confined to denning studies on that grizzlies move from avalanche slopes and lower the Alaska Peninsula and Kodiak Island (Lentfer et al. forests in the spring and early summer to alpine 1972). Annual alpine composition counts on the meadows in midsummerand back to lower elevations Kodiak National Wildlife Refuge have revealed many in autumn. These authors furtherstate that the move- bears (U. a. middendorffi)in summermonths. A study ments seem related to the availability of naturalfoods was conducted during 1973, 1974, and 1975 to deter- and that the greater density of these bears in Glacier mine the reasons for their presence. National Park can be attributedto the highly productive Brown bears (U. a. arctos) in Europe, and Russia alpine and subalpine vegetation. (U. a. beringianus)also commonly use high mountains Brown bearhabitat in Alaskais underdisruptive pres- in summer. Curry-Lindahl(1972), referring to Euro- sures from several sources. The Alaska Native Claims pean bears, states that these animalsvisit alpine areasin Settlement Act is causing millions of hectares of wild- late summerto feed on berries. He comments on Scan- lands to be transferredfrom public to privateownership, dinavianpopulations by adding (p. 78) that some "sub- with potentialexploitation of naturalresources likely in alpine birch forests ... are characterizedby a luxurious many areas. On the Kodiak National Wildlife Refuge, vegetation, which offers the bear optimal habitats." 140,000-224,000 ha of bear habitatwill be patentedto Novikov (1956) reportedthat brown bears in interior Native villages. The searchfor oil on the Alaska Penin- Russia make regular seasonal vertical migrations into sula and in the Gulf of Alaska, with attendantconstruc- the mountains, frequenting the alpine zone to an al- tion of onshore staging facilities, and increased timber titude of 2,800 m or higher. In eastern Siberia, where harvests in coastal southcentral and southeastern the habitat resembled that of southwestern Alaska, Alaska, all pose threats to bear habitat. As man con- Kistchinski (1972:70) says that brown bears "often" tinues to infringeon these wildlands, effective manage- move up to the alpine zone at 1,500-1,800 m by July. ment schemes will have to be based on a sound know- Brown bears in the Austrian Alps (Krott 1962) and in ledge of the bears' habitat use patterns if more than Italy's Abruzzo National Park (Zunino and Herrero remnantpopulations are to be perpetuated. move into the 1972) alpine zone in late spring to graze The study objectives were (1) to define the major on the freshly sproutedgreenery of the high mountain plant communities in the study area, (2) to measure meadows. bear use of each community and determine the most importantplant species sought, and (3) to determine summer alpine activity areas (a unit of alpine habitat 'Present address: U.S. Fish and Wildlife Service, New used to fulfill normal living demands) used by family England Area Office, P.O. Box 1518, Concord, New and Hampshire03301. groups single bears. 2Presentaddress: U.S. Fish and Wildlife Service, Finley The authors wish to thank the U.S. Coast Guard National Wildlife Refuge, Corvallis, Oregon 97330. Rescue CoordinationCenter at Kodiak for 3Present helicopter address: Alaska Departmentof Fish and Game, It is doubtful that this 333 RaspberryRoad, Anchorage, Alaska 99501. support. study, which was con- 4Present address: U.S Fish and Wildlife Service, Kenai ducted in a remote and isolated area, could have been National Moose Range, Kenai, Alaska 99611. successful without Coast Guard participation.In addi- 298 BEARS - THEIR BIOLOGY AND MANAGEMENT tion, field assistance provided by R. Smith, B. Bal- METHODS lenger, and W. Donaldson of the Alaska Departmentof The study area was monitored each year with a Fish and Game, and P. Smith, U.S. Fish and Wildlife fixed-wing aircraft, commencing in late April, to de- Service biological technician, was genuinely ap- termine when bears first moved to the high country. preciated. Flights were conducted at approximately10-day inter- vals until the field crew was on site. Personnel stayed STUDY AREA in the mountainstudy area as long as bears remained. To minimize disturbance,party size was limited to 2. The KodiakIsland group (Fig. 1) is positioned in the Plant identifications were based on Hulten's (1968) Gulf of Alaska southeast of the Alaska Peninsula. Flora of Alaska and Neighboring Territoriesand were Kodiak, the largest of the 3 main islands, is ap- corroboratedby D. Murray, curatorof the University proximately 96 km by 160 km. It is typified by of Alaska herbarium.Plant communities were identified fjord-likebays penetratingas far inlandas 50 km. These by using Hjeljord's (1971) Kodiak alpine vegetation bays and associated cirques and U-shaped intermoun- analysis as a reference and were then mapped for the tain valleys were created by the scouring action of entire study area at a scale of 1:15,840. In determining Pleistocene glaciers. The result is a succession of the communities, Hanson and Churchill's (1961:66) mountainspurs flanked by fjords and valley extensions generalized definition was followed, i.e., "a group of (Karlstromand Ball 1969). Rugged mountainsrising to stands that are similar in species composition and a maximumof 1,360 m dominate the island's interior. structureand occupy similarhabitats." The termalpine Sharp crested alpine peaks protrude from the main as used in this paperrefers to the entire study area, and northeast-trendingaxis of the range, and broad ridges unless stated otherwise, includes the subalpine zone extend to the northwest. comprised of the Willow Field - Subalpine Meadow and protruding tongues of the Alder Community's The island's climate is influenced by the Japanese upper limits. Current; therefore, temperatures are mild, with ex- Bear observations were made daily, using variable- tremes at sea level seldom fluctuatingbelow -18C or power (20-45X, 15-60X) spotting scopes at distances above 27 C. Frost occurs every month of the year at of 50 m to 2 km. With little vegetation to hide them, altitudes higher than 700 m. Annual precipitation is bears were easily observed in the high mountains.The 157 cm, with drizzle and fog common. Cyclonic lows entire area was searched once a week, but loca- move in from the Aleutian Islands and cause frequent study tions where these animals concentrated,as determined precipitationand windstormsthroughout the year. from aerial and ground surveys, were monitoredevery The most characteristicvegetation of Kodiak Island, 1 or 2 days. Thus, the same bears were usually seen from sea level to the brush line at about 580 m, is an several times a week and frequently for several con- alder (Alnus sp.) - willow (Salix spp.) complex that is secutive days during each week. Bears were observed interspersed with lush grass-forb meadows. Cotton- continuously for as long as conditions allowed. Their wood (Populus trichocarpa) grows along river valleys activities were noted and coded in sequence so that an and stunted birch (Betula papyrifera var. kenaica) observationusually consisted of a series of activities. occurs on gentle slopes below 250 m. The 1 conifer Repeated observations of the same bears in open present, Sitka spruce (Picea sitchensis), is confined to countryallowed the field team to identify many animals the northeastsection of the island. individually. Size, conformation, deformities, pelage About 80 percent (7,200 km2) of Kodiak Island's shedding patterns, pelage color, and scars were distin- southwest portion constitutes the Kodiak National guishingcharacteristics. Sex, age, family groupings,and Wildlife Refuge. On one of the remote mountainspurs behaviortraits were also distinguishingfactors, leading to within the refuge, a 56-5 km2 alpine study area was high confidencein the identificationof individuals.Other selected because of heavy summer use by bears (Fig. investigatorshave used similarphysical and behavioral 1). The study area, which dips into the subalpinezone, peculiaritiesto identify individual animals (Burkholder is about 9.7 by 5.8 km, and is comprisedof mountain- 1959, Meehan 1961, Woodgerd 1964, Craigheadet al. ous terrainranging in elevation from 305 to 1,316 m. It 1974, Martinka1974). is located on the west side of Kodiak Island between Boundaries of activity areas were established by Uganik Lake and the head of the South Arm of Uganik connecting the outermost position locations, as deter- Bay, 64 km west-southwest of municipal Kodiak. mined from ground observations, with straightlines. BROWN BEAR SUMMER HABITAT* Atwell et al. 299

- e~~~~~~~~~~~~~~~~~~~~~~~$~~~~~

C44,A

XA t

i 0O

; :; .

I t

O O- ...tt :~.tloaf4

. ' f-L

Fig. 1. Kodiak Island and the alpine studSTUDY AREA I Fig. 1. Kodiak Island and the alpine study area, Kodiak National Wildlife Refuge, Alaska, 1973-75. 300 BEARS- THEIR BIOLOGY AND MANAGEMENT

The alpine zone was never completely without snow, first week of May. By the second week of June, the amount varying at any given time from year to year. snowslides had cleared some south- and southwest- In both 1973 and 1975, the study area remained com- facing slopes and plant growth occurred. Snowmelt, pletely snow-covered through the first 3 weeks of May. and consequent plant development, were 2 weeks ear- New growths of grasses, sedges, and forbs did not lier in 1974 than in 1973 and 1975. appear on snow-free areas until the first week of July. Snow in gullies, hollows, and chutes receded gradually RESULTS through mid-August, continuously releasing shoots of fresh vegetation. PlantCommunities By contrast, the spring of 1974 was unusually warm The identity of 99 different plant species was estab- and sunny; a few ridges and knolls were exposed the lished and 7 communities were recognized (Table 1).

Table 1. Plant community criteria for the alpine and subalpine zones of the study area, Kodiak National Wildlife Refuge, Alaska, 1973-75.

Most common Altitude Plant community Characteristics species range Aspect Slope

Carex-ForbMeadow A thick, nearly continuous Carex macrochaeta, 670-945 m South and 30-500 growth of Carex macrochaeta Calamagrostis canaden- southwest mixed with numerousbroad- sis, Anemonenarcis- leaved species. Ericaceous siflora, Lupinus plants are absent. Soil nootkatensis, Geranium fertility is high and the erianthum, Veratrum sites, frequentlylocated eschscholtzii, Angel- below snowbanks and rocky ica lucida, Arnica outcroppings, are well latifolia. supplied with moisture. Ericaceous Knolls Differs from Carex-Forb Rocky, exposed knolls 670-790 m No specific 0-40? and Hummocks Meadows in having overall are dominatedmainly aspects. gentler topography, lower by Empetrumnigrum, average fertility, and the Cassiope stelleriana, presence of interrupted Oxytropisnigrescens, mats of Ericaceae species. Sedum rosea, Loise- Both xeric and mesic condi- leuria procumbens, tions exist and, together Salix arctica, and with a wide variety of Arnica lessingii. soils and microenvironments, contributeto species Rolling hummocks and diversity. mixed meadow slopes between the rocky knolls are densely vegetated with Carex macrochaeta, Festuca altaica tussocks, Poa arctica, Calamagros- tis canadensis, Empe- trum nigrum, Anemone narcissiflora, Lupinus nootkatensis, Salix reticulata, and Salix barclayi. Rocky Carex-Forb Moderatelyvegetated, very Most of the same 790-1,130 m South and 50-80? Meadow precipitous, and inter- species found in Carex- southwest ruptedby extensive rocky Forb Meadow, but also outcroppings. Ordinarily some hardierplants situated immediately above found on higher-stress Carex-ForbMeadows. sites such as Erica- ceous knolls; these include Cassiope stelleriana, Empetrum nigrum, Rhododendron camtschaticum, and Geum rossii. BROWN BEAR SUMMER HABITAT* Atwvell et al. 301

Table 1, Continued Most common Altitude Plant community Characteristics species range Aspect Slope

Talus Slope and This community is charac- Scatteredclumps of IUsually above Most fre- Mainly Bare Rock terized by rocks, volcanic Luzula wahlenbergii. 9)15 m but quently 40-90? ash, and little vegetation. c()ccasionally on steep but some- It is a xeric community, is low as north-facing times even though it becomes 760 m in slopes but 30-40? snowfree late in the year. areascovered may be found by snowbanks on almost mostof the any aspect. summer. Luzula- These slopes are sparsely Luzula wahlenbergii, 760-1,220 m North 15-50? Congeliturbation vegetated, xeric, and of Sedum rosea, prostrate Slope low fertility. They are Salix sp., Saxifraga located where environmen- bronchialis, Papaver tal conditions are severe. alaskanum, Potentilla Loose volcanic ash is villosa, Oxytropis interspersedwith rock nigrescens, Carex sp., slides and barrenlate Geum rossii, Silene snow-free areas. Frost acaulis, Minuartia sp., action is manifest in soil and Luetkeapectinata. lobes with obvious down- slope creep; frost scars and miniaturestone rings may be present. Willow Field- This type encompasses the Carex macrochaeta, 490-730 m Any aspect 0-35? Subalpine Meadow transitionzone between the Calamagrostis canaden- true alpine and lower Alder sis, Salix barclayi, Community. Fertile soil and Salix glauca, Salix an abundantmoisture supply sp., Geraniumerian- make it high in plant pro- thum, Arnica latifolia, ductivity. Heracleum lanatum, Rubus spectabilis, and Alnus crispa. Alder This community is charac- Calamagrostis canaden- 580 m and Any aspect 0-65? terized by a discontinuous sis, Carex macrochaeta, below belt of Alnus crispa inter- Athyriumfilix-femina, spersed with dense meadows Rubus spectabilis, Sam- of grasses, sedges, ferns, bucus racemosa, Salix and forbs, the result of spp., Heracleum lanatum, deep, fertile soil and a Geraniumerianthum, and good moisture supply. Alnus crispa.

There was considerable variation in community size bers generally increased through the second and third (Table 2). As the largest, Ericaceous Knolls and weeks of July, then slowly declined. In the first half of Hummocks covered 40 percent of the study area. The August, the animals abruptly left the study area. Alder and Willow Field-Subalpine Meadow com- A total of 804 hours were spent, during the 3 years, munities were second at 19 percent each, with the re- observing 305 individual bears and family groups maining 4 plant aggregations accounting for a total of (Table 3). Activity sequences for this period numbered 22 percent of the land area. 2,077. An average of 29 different bears was individu- ally identified yearly. An average of another 42 was Bear Use of Habitat Alpine seen too infrequently to establish sufficient criteria for Bear tracks first appeared on the study area in early field recognition. These latter individuals were mainly May each year but were not common until late May. transient animals, but undoubtedly some bears with Bear occurrence was not continuous until the last week established identities were occasionally included in this of June or the first week of July when new vegetation category when seen for only brief periods. was available in quantity on the snow-free south- and Fifty percent of the time that bears were under ob- southwest-facing slopes. Ground team observations servation, they were feeding; plants consumed were began after continuous occupation by bears. Bear num- identified in about half the instances. These plants in- 302 BEARS- THEIR BIOLOGY AND MANAGEMENT

Table 2. Comparison of plant community sizes within the study area, Kodiak were taken before the materialwas chewed and swal- National Wildlife Refuge, Alaska, 1973-75. Water surface areas are excluded. lowed. Area Percent of As snow receded in depressions, new growths of Plant community (ha) study area Carex macrochaeta were continuously released through July. By the second or third week of August Ericaceous Knolls and 2,247 40 Hummocks (first week in 1974), the perimetersof the snowbanks Alder 1,071 19 had melted, exposing ash and scree. These areas ordi- Willow Field- 1,072 19 narily are snow-covered for so much of the year that Subalpine Meadow Carex-ForbMeadow 492 9 they are devoid of vegetation. Talus Slope and 339 6 The early development, extreme abundance,appar- Rock Bare ent palatability, and high utilization of Carex mac- Luzula-Congeliturbation 287 5 Slope rochaeta made it the most importantcomponent in the Rocky Carex-ForbMeadow 100 2 bears' alpine diet. Second to sedges, bears preferred Total 5,608 100 feeding on flowers of Angelica lucida, Heracleum lanatum, and Lupinus nootkatensis. Table 3. Total brown bear observations by year on the study area, Kodiak Eighty-twopercent of the bears' time was spent in 2 National Wildlife Refuge, Alaska, 1973-75. plant communitiesin which 88 percent of their feeding 1973 1974 1975 Total Table 4. Percentage of time brown bears were observed in each of 7 plant Wildlife Number of observations 84 120 101 305 communities within the study area, Kodiak National Refuge, Alaska, 1973-75. (may include more than one observationof same All activities Feeding animal) Plant community combined only Number of activity sequences 448 884 745 2,077 (percent) (percent) (a series of all activities documentedwhile an animal Carex-ForbMeadow 65 73 was under observation) Willow Field- 17 15 Number of different indi- 15 44 29 86a Subalpine Meadow vidually identified bears Ericaceous Knolls and 8 8 (does not include young in Hummocks company of an adult) Rocky Carex-ForbMeadow 6 2 Total observationsof indi- 46 47 86 179 Luzula-Congeliturbation 2 1 vidually identified bears Slope Bears not individually 38 73 15 126 Alder 1 1 identified Talus Slope and 1 0 Time bears were observed 164 366 274 804 Bare Rock (hours) Observations Observations total total a Only 1 animal with known identity was observed each of the 3 years. 801 hours 393 hours

cluded Angelica lucida, Carex macrochaeta, Carex occurred (Table 4). By far the most importantcom- sp., Equisetumarvense, Heracleum lanatum, Lupinus munity was the Carex-ForbMeadow, where the bears nootkatensis, and Salix sp. When bears fed on known spent 65 percent of their time and which accountedfor plant species, they were almost exclusive in their 73 of their feeding. Apparently,bears concen- choices and concentrated onCarex macrochaeta 97 percent trate in these areas because of the key alpine food percent of the time. Bears frequentedthe steep Carex- Carex macrochaeta. Although this sedge is availablein Forb Meadow Community on south- and southwest- several communities, it is only abundantin Carex-Forb facing slopes in the 670- to 945-m altituderange where Meadow and Willow Field-Subalpine Meadow com- Carex macrochaeta grew in lush and almost pure plexes (Table 5). Because of steeper slopes, which stands (Table 1). The Carex-ForbMeadow prevailed trigger snowslides earlier in the higher Carex-Forb in areas having a good moisture supply. In these loca- Meadows, spring shoots emerge about 2 weeks earlier tions, bears fed along the hillsides in a manner similar there than in the Willow Field-SubalpineMeadows. In to that of ruminants,and, as has been noted for grazing addition, Carex macrochaeta concentrations in the other grazers, they probably select the highest-quality former community are more nearly pure. forage available (Klein 1970). Carex macrochaeta was Areas pulled, rather than bitten off, and separatedfrom the Activity roots to expose whitish basal stems. Several mouthfuls In a sample of 17 individually identified family BROWN BEAR SUMMER HABITAT * Atvell et al. 303

Table5. Carexmacrochaeta occurrence by plantcommunity within the study area, Kodiak National Wildlife Refuge, Alaska, 1973-75. ~'",mlmBiw Community Carex macrochaeta ^

Carex-ForbMeadow Abundantin nearly pure stands Willow Field- Extensive and predominant, Subalpine Meadow but in mixed stands Alder Moderate, but mainly in mixed stands Ericaceous Knolls and Light, with occasional Hummocks almost pure pockets Rocky Carex-ForbMeadow Light and scattered Luzula-Congeliturbation None Slope Talus Slope and None Bare Rock .*--1973 --1974 groupscomposed of females with cubs or yearlings, the ---1975 average activity area was 85 ha (Table 6). When 1 aberrantlylarge activity area of 834 ha is included, the becomes 127 ha. For 12 adults as- Fig. 2. Alpine activity areas of brown bear No. 06-73 for 3 consecutive years average solitary (1973-75), Kodiak National Wildlife Refuge, Alaska. sumed to be mostly males, the average activity area was 170 ha, twice the average for females with young. Bears tended to congregate, as reflected in the over- In a paired t-test this difference was significant (P < lapping activity areas shown in Fig. 3. On 2 August 0.025). Apparently, the dense stands of Carex mac- 1975, 32 bears were underobservation simultaneously rochaeta made it unnecessary for bears to move long in an area roughly 2.5 by 5.0 km, for a density of 2.6 distances for food, as demonstratedby the size of al- bears/km2. Even with these high concentrations, in- pine activity areas. traspecific strife was not common. Occasionally a

Table 6. Sizes of activity areas of single adult brown bears and family groups female with cubs chased a solitary adult that wandered within the study area, Kodiak National Wildlife Refuge, Alaska, 1973-75. too close, or 1 lone adult ran at another, but no en- counters that we witnessed ever ended in con- Number of Activity area physical Sample position average size tact. size locations (ha)

Females with 10 135 94 cubs-of-the-year Females with 7 57 72 yearlings Females with young 17 192 85 (above 2 cate- gories combined) Single adults 12 195 170 29 387 120

Activity areas for succeeding years were determined for only 1 animal - No. 06-73 (Fig. 2). One hundred and seven position locations for this bear were recorded during the 3 years. Locations were concentratedin the Carex-ForbMeadow Community adjacentto 2 cirque lakes. No. 06-73 returnedto the same alpine area each year but made occasional feeding forays, of less than 3 days each, up to 1.6 km east over a low pass. The straight-linedistance of these trips increased in 1974 and 1975. As a result, the activity areas became larger: Fig. 3. Alpine activity areas of individually identified brown bears on the study 1973 - 166 ha, 1974 - 209 ha, and 1975 - 472 ha. area, Kodiak National Wildlife Refuge, Alaska, 1973-75. 304 BEARS - THEIR BIOLOGY AND MANAGEMENT

Age Composition off Bears one of the earliest plants to emerge after snowmelt. Age composition fFor the 581 bears seen during the Initial growth is rapid because of the large amountsof study is compared with aerial count data in Table 7. carbohydrates stored in the roots and rhizomes The most obvious dlifferences occur in the cub and (Johnson and Tieszen 1973). Klein (1970) has stated subadult categories. Particularly lush growths of the that the highest nutritivequality of most forage plants preferred Carex macrrochaeta may attractand concen- coincides with the early stages of growth, which is tratefamily groups, w'hich would accountfor the higher probablytrue of Carex macrochaeta. percentage of cubs. ]Females with cubs, while on the Cubs only 6-7 weeks out of the den have rapid study area, were appairently able to obtain their food in growth rates, placing heavy dietary requirementson relatively small units of alpine habitat. their lactating dams, who in turn must not only feed their young but must also recover from the vicissitudes Table 7. Composition of brown bear age classifications as determinedby aerial and ground counts, Kodiak Noational Wildlife Refuge, Alaska, 1973-75. Aerial of denning. The lush standsof Carex macrochaetamay counts were conducted annua Illy on predesignated alpine units of which the satisfy these needs. study area represents 17 perceent. The bear density was high on the study area, 0.85 Study area bear/km2,determined from individually identified ani- Aerial counts ground counts mals seen on 2 or more occasions. This Age classification density ap- bqumber Percent Number Percent proximates the 0.65 bear/km2 found by Troyer and Hensel (1964) in a 249 km2 study area at KarlukLake Adults on Kodiak where brown bears had Subadults 55 13 32 5 Island, concentrated (2.5-3.5 years old) to feed on salmon, anothereasily available and proba- Yearlings 68 17 90 15 bly nutritious food (Mealey 1975). Kistchinski Cubs 22 198 35 Lavov et al. of Total 408 (1972:69), citing (1963), quoted figures 1.2-2.0 bears/km2 for parts of Paramushir Island (KurileIslands, USSR) rich in fish, but did not include the size of the land mass involved. Migration From Stuc Area Ithestudy arearatherabruptly Activity areas were small and frequentlyoverlapped. Bears lef the study area rather Troyer and Hensel found brown bear move- typically ambruptly (1964) each year. This migration coincided with a markedde- ments to be similarly limited in the vicinity of dense crease in availability of sedge shoots. In 1973, bear spawning areas. Most animals remainedthere for sev- observations were nuimerous through 17 August but eral weeks within 1.6 km of their food supply. k only 2 sightings were recordedafterwards. The bears Brown/grizzly bears are known to congregate sea- departedearlier in 19)74 than in either of the other 2 sonally at sources of preferredfoods and in this study years but not as abrupltly. During the 1975 field season, the Carex-Forb Meadow Community was especially bears were consistenttly located in moderate numbers important.There, and to a lesser degree in the Willow through 10 August; oenly 1 animal was found after that Field-SubalpineMeadow, activity areasfrequently date. overlapped. Animals so localized would be those Salmon (Oncorhynchus spp.) began arriving in whose home ranges encompassed the feeding area, as Kodiak Island's spawining systems adjacentto the study reportedfor black bears (Ursus americanus) by Jonkel area in early June but twere most numerousduring the and Cowan (1971). One individually identified bear second half of Aug ust. In the nearest stream, the returnedeach year to the same part of the study area, Uganik River, pea k counts during the study by which indicates that others may have done the same. biologists of the Alaska Departmentof Fish and Game After 5-6 weeks of grazing in the high mountains, averaged 77,000 salnaon per year. These fish became the bears abruptlyleft. Their departurecoincided with available and were heavily utilized by ears when new the return of spawning salmon, similar to what sedge shoots were nIO longer being produced in the Kistchinski (1972) reported for eastern Siberia. alpine zone. Whether the presence of salmon at Kodiak Island at- DISCUSSION tracts the bears from the alpine areas is not known; Bears appearedto have been attractedto the study however, the bears' departureoccurs at the time salmon area each summerpri imarily by Carex macrochaeta - become readily available. Most likely, as alpine plants a localized but abunidant, fast-growing, and probably become less palatablewith age, the bears merely leave highly nutritious foo d (Mealey 1975). This sedge is to seek substitutefoods and the salmon fulfill this need. BROWN BEAR SUMMER HABITAT* Atwell et al. 305

LITERATURECITED LAVOV, M. A., V. G. VORONOV,AND Y. I. YEVDOKIMOV. BURKHOLDER,B. L. 1959. Movements and behavior of a 1963. Fauna of game mammals of the islands wolf pack in Alaska. J. Wildl. Manage. 23(1):1-11. Paramushir and Shumshu and possibilities of enriching CRAIGHEAD,J. J., J. R. VARNEY, AND F. C. CRAIGHEAD, it. Proc. Sakhalin Res. Inst., USSR Acad. Sci., No. Jr. 1974. A population analysis of the Yellowstone 14:39-47. (In Russian.) (Cited by Kistchinski 1972.) grizzly bears. Montana For. and Conserv. Exp. Stn. LENTFER, J. W., R. J. HENSEL, L. H. MILLER, L. P. Bull. 40. 20pp. GLENN, AND V. D. BERNS. 1972. Remarks on denning CURRY-LINDAHL, K. 1972. The brown bear (Ursus habits of Alaska brown bears. Pages 125-132 in S. arctos) in Europe: decline, present distribution, biology Herrero, ed. Bears their biology and management. and ecology. Pages 74-80 in S. Herrero, ed. Bears IUCN Publ. New Ser. 23. their and IUCN Publ. New Ser. biology management. MARTINKA,C. J. 1974. Population characteristics of grizzly 23. bears in Glacier National Park, Montana. J. Mammal. HANSON, H. AND E. D. CHURCHILL.1961. The C., plant 55(1):21-29. community. Reinhold Publishing Corp., New York. 218pp. MEALEY, S. P. 1975. The natural food habits of free-ranging HJELJORD,0. G. 1971. Feeding ecology and habitat prefer- grizzly bears in Yellowstone National Park, 1973-1974. ence of the mountain goat in Alaska. M.S. Thesis. M.S. Thesis. Montana State Univ., Bozeman. 158pp. Univ. of Alaska, Fairbanks. 126pp. MEEHAN, W. R. 1961. Observations on feeding habits and HULTEN, E. 1968. Flora of Alaska and neighboring ter- behavior of grizzly bears. Am. Midl. Nat. 65(2):409- ritories. Stanford University Press, Stanford, Calif. 412. 1,008pp. MUNDY, K. R. AND D. R. FLOOK. 1973. JOHNSON,P., AND L. TIESZEN. 1973. Vegetative research in D., Background for bears in the national of arctic Alaska. Arctic Inst. N. Am. Tech. Pap. 25:169- managing grizzly parks 198. Canada. Can. Wildl. Serv. Rep. Ser. 22. 35pp. JONKEL,C. J., AND I. McT. COWAN, 1971. The black bear in NOVIKOV, G. A. 1956. Carnivorous mammals of the fauna the spruce-fir forest. Wildl. Monogr. 27. 57pp. of the USSR. Keys to the Fauna of the USSR, No. 62. KARLSTROM,T. N. V., AND G. E. BALL, eds. 1969. The USSR Acad. Sci. Bull. 2. 293pp. Transl. by the Israel Kodiak Island refugium: its geology, flora, fauna and Program for Scientific Translations, Jerusalem. Toronto. history. Ryerson Press, 262pp. TROYER, W. A., AND R. J. HENSEL. 1964. Structure and KISTCHINSKI,A. A. 1972. Life of the brown bear history distribution of a Kodiak bear population. J. Wildl. Man- (Ursus arctos L.) in north-east Siberia. 67-73 in Pages age. 28(4):769-772. S. Herrero, ed. Bears- their biology and management. IUCN Publ. New Ser. 23. WOODGERD, W. 1964. Population dynamics of bighorn KLEIN, D. R. 1970. Tundra ranges north of the boreal forest. sheep on Wildhorse Island. J. Wildl. Manage. J. Range Manage. 23(1):8-14. 28(2):381-391. KROTT, P. 1962 Beitrage zur Kenntnis des Alpenbaren, ZUNINO, F., AND S. HERRERO. 1972. The status of the Ursus arctos Linne, 1758. Saugetierk, Mitt. brown bear (Ursus arctos) in Abruzzo National Park, 10(Sonderheft): 1-35. Italy, 1971. Biol. Conserv. 4(4):263-272.