Relationships between activity patterns and foraging strategies of Yellowstone grizzly bears by Albert L Harting A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Biological Sciences Montana State University © Copyright by Albert L Harting (1985) Abstract: Eleven grizzly bears (Ursus arctos horribilis) were radiotracked in Yellowstone National Park and vicinity in 1981 and 1982. Principal objectives of the study were 1: to examine the daily and seasonal activity patterns of Yellowstone grizzlies and to determine what influence certain temporal and environmental factors had on these activity patterns and 2: to examine the interrelationships of food habits, habitat use, movements, and activity patterns. Two methods for rating the quality of a bear’s occupied habitat were employed. One method considered the abundance, diversity, and relative value to grizzlies of the vegetation occurring at field-checked relocation sites. The second method utilized existing habitat maps and a spatial information computer package to identify the habitats surrounding relocation points. These habitat types were then rated according to a system of Habitat Importance Values developed by the Interagency Grizzly Bear Study. Theoretical aspects of grizzly bear foraging strategies and predatory habits were also considered. Environmental factors which had a significant effect on grizzly bear activity patterns were temperature, precipitation, and cloud cover. Some of the influence of environmental variables on bear activity could be explained according to their probable effect on olfactory perception. Temporal factors found to be important were season and time of day (diel period). Grizzlies in this study were primarily crepuscular and . nocturnal but individual bears differed significantly in their activity patterns. Individual differences in grizzly bear food habits and habitat use were reflected in their characteristic activity patterns and movements. Bears which occupied vegetatively poor habitat appeared to be more reliant on "supplemental" food sources (meat or garbage) than bears in rich mesic areas. The use of trained bear dogs to retrace grizzly bear movements proved to be a valuable adjunct to traditional research tactics.  RELATIONSHIPS BETWEEN ACTIVITY PATTERNS AND FORAGING STRATEGIES OF YELLOWSTONE GRIZZLY BEARS by Albert L. Harting, Jr. A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Biological Sciences MONTANA STATE UNIVERSITY' Bozeman, Montana March 1985 APPROVAL of a thesis submitted by Albert L. Harting This thesis has been read by each member of the thesis committee and has been found to be satisfactory regarding content, English usage, format, citations, bibliographic style, and consistency, and is ready for submission to the College of Graduate Studies. 7 . / ? Sr 6 Date chairperson. Graduate Committee Approved for the Major Department 2S hbn;ary , IcISS Date Z Ak % Head, Major Department Approved for the College of Graduate Studies Date Graduate Dean ill STATEMENT OF PERMISSION TO USE In presenting this thesis in partial fulfillment of the requirments of a master's degree at Montana State University, I agree that the Library shall make it available to borrowers under rules of the Library. Brief quotations from this thesis are allowable without special permission, provided that accurate acknowledgement of source is made. Permission for extensive quotation from or reproduction of this thesis may be granted by my major professor, or in his absence, by the Director of Libraries when, in the opinion of either, the proposed use of the material in this thesis is for scholarly purposes. Any copying or use of the material in this thesis for financial gain shall not be allowed without my written permission. Date Signal VACKNOWLEDGEMENT I am sincerely grateful to the following people for their contributions to this study. My major professor. Dr. H. D. Picton, was the source of many valuable insights which were critical to the success of this project. Dr. R. R. Knight, Interagency Grizzly Bear Study, first proposed the project and provided the research I opportunity and funds without which the study would not have been possible. Drs. R. L. Eng and R. E. Moore reviewed the manuscript and made many constructive suggestions. Ms. Georgia Ziemba, MSU Computing Services, was extremely helpful with the statistical analysis aspects of the study. I am especially grateful to Dave Mattson, IGBS, for providing me with his preliminary data pertaining to grizzly bear food habits and habitat use. The fieldworkers who assisted me with data collection, especially Terry Jones and Kevin Rhodes, deserve many thanks for remaining enthusiastic despite the long hours, lack of sleep, and oftentimes dreary conditions. S. R. Vaughan and L. Lobos deserve credit for providing inspiration as only they knew how. Finally, I wish to thank my family for their enduring patience and cooperation throughout this project. My parents and my uncle, Frank Harting, managed to give me ongoing encouragement and financial assistance. My wife, Linda, and son, Aaron, contended with the many long absences and mysterious moods which every graduate student surely knows. vi I TABLE OF CONTENTS Page LIST OF TABLES................................. '.............. viii LIST OF FIGURES............................................... x ABSTRACT...................................................... xii INTRODUCTION.................................................. I STUDY AREA.................................................... 3 Administrative Context.................. 3 Geological Background....................................... 3 Vegetation Zones............................................ 5 Study Area Sub-Units..... ................................... 7 Gneiss Creek/Hebgen Lake Sub-area........................ 8 Nez Perce Cr./Firehole River Sub-area.,.................. 9 Blackball Plateau/Washburn Range Sub-area....... 9 Climatology................................................. 10 METHODS................................. 12 Data Collection.......................................... 12 Trapping and Radio-Tracking........................ 12 Activity Monitoring...................................... 13 Habitat Use and Scat Data Collection..................... 14 Dog Tracking.................................... 15 Data Analysis............................................... 16 Activity Data............................................ 16 Community Site Data Analysis............................. 17 Computer Relocation Habitat Scans............. 20 Scat Data Analysis....................................... 23 Minimum Daily Movements and Home Range Estimates......... 24 RESULTS....................................................... 25 Activity Data............................................... 25 Temperature.............................................. 25 Precipitation and Ground Moisture.......... 28 Wind Speed................................... J.......... 28 Cloud Coyer.............................................. 30 Seasonal and Monthly Effects............................. 30 vii TABLE OF CONTENTS— Continued Page. Time of Day Effects...................................... 32 Seasonal Activity Patterns by Diel Period............ 35 Individual Bear Patterns............................. 38 Community Site Analysis........ 42 Relocation Habitat Scans.......................... 44 Scat Analysis............................... 46 Movements and Home Range Use......................... 48 Tracking Grizzlies with Bear Dogs....................... 49 DISCUSSION.................. .'................................. 52 Activity Patterns........................................... 52 Diel Patterns. .................................. 52 Seasonal Activity Levels................................. 53 Environmental (Weather) Effects............... 55 Energetic Agendas of the Primary Study Bears................ 62 Bear 59.................................................. 63 Bear 38.................................................. 66 Bear 50.......... 67 Bear 15........... 69 Bear 76.................................................. 72 Grizzly Bear Foraging Strategies............................ 75 Theoretical Considerations of Grizzly Predation on Ungulates....................... 80 Tracking Grizzlies with Trained Bear Dogs................... 82 CONCLUSIONS................................ 85 LITERATURE CITED......................... 88 APPENDICES............................. 95 Appendix A - Community Site Field Form...................... 96 Appendix B - Tables of Habitat Parameter Values and Community Site Data........ 99 viii LIST OF TABLES Page Table I. Scientific names and abbreviations for habitat types referred to in the text...................... 7 Table 2. Distribution of activity records by bear, season, and year................................ 13 Table 3. Age, sex, and monitoring period for the five primary study bears............................................ 13 Table 4. Mean values for Food Value (FV), Understory Cover (C ), Understory Species Diversity (H ), and Community Si£e Quality Index (CSQ) for Bears 15,U 38, 50, 59, and 76.... .................................................. 43 Table 5. Grizzly bear Area Food Scores (FS )^ mean amount of Edge per relocation scan circle (E), mean Habitat Diversity in scan circles ( H ) , and Relocation Habitat Richness scores (RHR)................................. 44 Table 6. Mean Scat Values (SV) scores for the primary study bears.............................................. 46 Table 7. Scat summary for primary study bears: percent "digestible" diet volume and percent frequency occurrence for important diet item groups............... 47 Table 8. Consecutive minimum daily movements (km) for the five primary study bears................................ 49 Table 9. Short term home ranges for the primary study bears... 49 Table 10. Energetic efficiencies (EE), characteristic contagiousness (A ), and monthly preference values for the most important diet items of Yellowstone grizzlies as used in the community site and scat quality analyses....................................... 100 Table 11. Monthly Food Values (FV) of the most important diet items of Yellowstone grizzlies as used in the community site analyses....... 101 ix LIST OF TABLES— Continued Page Table 12. Unit area importance values (IVU's) used to score habitat types for habitat richness analysis............ 102 Table 13. Community site scores for Food Value (FV), Understory Cover (C ), Understory Species Diversity (H ), and CommunityuSite Quality (CSQ)................. V........ 103 LIST OF FIGURES Page Figure I. Map of the study area.......... ........................ 4 Figure 2. Relationship between temperature (in C) and probability of bear activity............................ 26 Figure 3. Relationship between temperature (in C) and probability of bear activity in spring. ..... 26 Figure 4. Relationship between temperature (in C) and probability of bear activity in summer..... ......... 27 Figure 5. Relationship between temperature (in C) and probability of bear activity in fall................... 27 Figure 6. Relationship between precipitation type and probability of bear activity.......................... 29 Figure 7. Relationship between ground moisture and probability of bear activity.................. 29 Figure 8. Relationship between wind speed (in km/hr) and probability of bear activity......................... 31 Figure 9. Relationship between cloud cover and probability of bear activity......................................... 31 Figure 10. Relationship between month and and the probability of bear activity............... 33 Figure 11. Relationship between diel period and probability of bear activity annually and seasonally................. 33 Figure 12. Probability of bear activity according to hour of the day annually......... 34 Figure 13. Probability of bear activity according to hour of the day in spring.......................... 36 Figure 14. Probability of bear activity according to hour of the day in summer............................. 37 X xi LIST OF FIGURES— Continued Page Figure 15. Probability of bear activity according to hour of the day in fall. ......................................... 39 Figure 16. Overall probability of activity for the five primary study bears........................................... 40 Figure 17. Probability of activity according to diel period for all monitored bears (including all activity records from all bears from both monitoring years) and for primary study bears #76 and #59...................... 40 Figure 18. Probability of activity according to diel period for primary study bears #50, #38, and #15................. 41 Figure 19. Probability of activity according to diel period for Bear 50 annually and by season...................... . 42 xli ABSTRACT Eleven grizzly bears (Ursus arctos horribilis) were radiotracked in Yellowstone National Park and vicinity in 1981 and 1982. Principal objectives of the study were I: to examine the daily and seasonal activity patterns of Yellowstone grizzlies and to determine what influence certain temporal and environmental factors had on these activity patterns and 2: to examine the interrelationships of food habits, habitat use, movements, and activity patterns. Two methods for rating the quality of a bear’s occupied habitat were employed. One method considered the abundance, diversity, and relative value to grizzlies of the vegetation occurring at field-checked relocation sites. The second method utilized existing habitat maps and a spatial information computer package to identify the habitats surrounding relocation points. These habitat types were then rated according to a system of Habitat Importance Values developed by the Interagency Grizzly Bear Study. Theoretical aspects of grizzly bear foraging strategies and predatory habits were also considered. Environmental factors which had a significant effect on grizzly bear activity patterns were temperature, precipitation, and cloud cover. Some of the influence of environmental variables on bear activity could be explained according to their probable effect on olfactory perception, Temporal factors found to be important were season and time of day (diel period). Grizzlies in this study were primarily crepuscular and . nocturnal but individual bears differed significantly in their activity patterns. Individual differences in grizzly bear food habits and habitat use were reflected in their characteristic activity patterns and movements. Bears which occupied vegetatively poor habitat appeared to be more reliant on "supplemental" food sources (meat or garbage) than bears in rich mesic areas. The use of trained bear dogs to retrace grizzly bear movements proved to be a valuable adjunct to traditional research tactics. IINTRODUCTION Prior studies of the grizzly bear (Ursus arctos horribilis) in the Yellowstone ecosystem have contributed a wealth of data pertaining to the food habits, habitat use, and general ecology of this * population (Mealey 1975; Graham 1978; Kendall 1981; Knight et al. 1981; Craighead and Mitchell 1982; Knight et al. 1984). These data provided the framework within which present management strategies were developed. But as the welfare of the Yellowstone grizzlies appears less certain, management decisions become increasingly complex and a need for data of still finer resolution becomes apparent. Recently, expanding emphasis has been placed on bear behavior. Schleyer (1983) examined the activity patterns of Yellowstone grizzlies with respect to temporal and environmental variables. Sizemore (1980) also studied the activity patterns of grizzly bears. Two prior studies dealt partially with grizzly bear foraging strategies. Mealey (1975) found that Yellowstone grizzlies appeared to forage in three physiographically distinct feeding "economies" (lake, mountain, and valley/plateau), and felt that bears mainly occupied a protein food niche. Sizemore (1980) computed the energetic requirements of individual grizzlies and described how these bears maintained an energy balance by utilizing reserve fat to supplement the available foraging opportunity. i 2None of the studies cited above have attempted to correlate data from bear food habits and habitat use with data on bear activity patterns for individual bears. This study sought to provide an overview of the grizzly bear's activity pattern/foraging strategy complex. The specific objectives of this study were: 1. To examine the daily and seasonal activity patterns of Yellowstone grizzlies, and to determine how these patterns fluctuated between individual bears and under different environmental conditions. 2. To determine how a given bear's activity patterns, movements, food habits, and habitat use were interrelated, and to contrast individual patterns to see how variation along one parameter appeared to affect the others. 3. To develop a conceptual overview of grizzly bear foraging strategies with reference to established optimal foraging theories. 4. To explore the feasibility of using trained bear dogs in bear research. 3STUDY AREA Administrative Context The study area was located in the Greater Yellowstone Ecosystem of Wyoming and Montana and included parts of Yellowstone National Park and contiguous National Forest land (Figure I). All of Yellowstone Park and much of the remainder of the study area have been designated as "Management Situation I" grizzly habitat in accordance with the grizzlies' threatened status under the Endangered Species Act (87 stat 884, 16U.S.C. 1531-1543). This designation specifies that "grizzly habitat maintenance and improvement... and grizzly-human conflict minimization will receive the highest management priority. Management decisions will favor the needs of the grizzly bear when grizzly habitat and other land use values compete." (USFS and NPS 1979) The Park Service Grizzly Bear Policy stipulates that management policies will be designed to "I. perpetuate wild, free-ranging grizzly bear populations and, 2. minimize conflicts between humans and grizzly bears by reducing man-generated food sources and by regulating visitor distribution." Geological Background The present landscape of Yellowstone was produced by repeated episodes of sedimentation, faulting, volcanic activity and, ultimately, glaciation (Keefer 1972). Two major types of bedrock were GALLAT IN N A TL . FOR. YELLOWSTONE PARK BOUNDARY I MT. WASHBURN WEST YELLOWSTONE' OLD FAfTVFUL BR IDGER -TETON N A T L . FOR. SUB-AREAS KILOMETERS Figure I. Map of the study area 5formed during periods of exceptional volcanic action in the Cenozoic era. The Absaroka bedrock was formed from major eruptions in the early Eocene which buried Yellowstone beneath thousands of feet of ash and lava. The Absaroka rocks were primarily andesite and basalt. The other major bedrock type- the Yellowstone volcahics- was deposited during three cycles of intense pyroclastic activity in the Quaternary period. The most recent of these cycles climaxed about 600,000 years ago with a major eruption of rhyolitic pumice and ash and the accompanying formation of the 2600 square kilometer Yellowstone caldera in the central portion of the Park. Lava continued to flow from ring fracture zones encircling the caldera until roughly 75,000 years ago. These flows were principally rhyolite, the so-called Plateau Rhyolite, but some basalt flows have also been identified. Eaton et al. (1975) suggested that the present hydrothermal activity in Yellowstone may be the seminal stage of a fourth volcanic cycle rather than the final phase of the third. Yellowstone was glaciated at least three times. The most recent of these, the Pinedale glaciation, occurred 25,000 to 8,500 years ago and covered up to 90% of the present Yellowstone Park. Dams formed by the receding Pinedale glaciers eventually burst causing catastrophic flooding, the results of which are still evident in many areas. ■' Vegetation Zones Although glaciation and erosion have redistributed and altered the composition of the original Absaroka and Yellowstone deposits, the 6present vegetation appears to depend in large part on the underlying bedrock type. Despain (1973) described three major vegetation zones within Yellowstone National Park. The lodgepole pine (Pinus contorta) zone appears to be strongly associated with poor soils of Quaternary rhyolite origins. It is dominated by climax lodgepole pine with occasional spruce (Picea) and fir (Abies) occurring in favorable sites. This zone typically occurs at elevations of 2320-2560 meters (m) and receives relatively low (51-102 centimeters, cm) annual precipitation. The spruce-fir zone is positively associated with the richer Absaroka (andesitic) volcanic soils. Mature stands may be dominated by either spruce or fir or, near timberline, whitebark pine (Pinus albicaulis). Successionally young stages include both spruce and fir in the understory and, frequently, vigorous stands of lodgepole pine. This zone occurs above 2560 m and generally receives greater than 102 cm annual precipitation. The third major vegetation zone is the Douglas-fir (Pseudotsuga menziesii) zone. This, zone is characterized by Douglas-fir as the dominant forest overstory with some spruce, fir, lodgepole, and aspen (Populus tremuloides) in favorable sites. Big sagebrush (Artemesia tridentata) and mixed grasses are common in open areas. This zone overlies mixed depths of glacial till and a bedrock of Quaternary sediments and granite at elevations of 1830-2320 m. Annual precipitation is generally less than 51 cm. 7Study Area Sub-Units Grizzly bears were radiotracked throughout much of the western and northcentral portions of Yellowstone Park at various times during this study. However, many of the data were collected in three principal subareas (Figure I). Habitat types described below follow the classification of Mueggler and Stewart (1980) for grassland and shrubland and Steele et al. (1979) for forested areas as identified by Despain (1984) in Yellowstone Park (Table I). Table I. Scientific names and abbreviations for habitat types referred to in the text. Habitat types follow the systems of Mueggler and Stewart (1980) for grassland and shrubland and Steele et al. (1979) for forest types. Forest habitat types: ABLA/VAS C-VAS C ABLA/VASC-CARU ABLA/VASC-PIAL ABLA/CACA PIEN/EQAR ABLA/THOC ABLA/CAGE ABLA/LIBO-VASC ABLA/VAGL-VAGL ABLA/CARU PICO/CARO PICO/PUTR PSME/SYAL PSME/CARU PIAL/VASG Abies lasiocarpa / Vaccinium scoparium-V.scoparium A. lasiocarpa / V^ _ scoparium-Calamagrostis rubescens A. lasiocarpa / Vj_ scoparium-Pinus albicaulis A. lasiocarpa / Calamagrostis canadensis Picea engelmannii / Equisetum arvense A. lasiocarpa / Thalictrum occidentale A. lasiocarpa / Carex geyeri A. lasiocarpa / Linnea borealis-V.scoparium A. lasiocarpa / Vj_ globulare-V.globulare A. lasiocarpa / C^ rubescens Pinus contorts / Carex rossii P. contorts / Purshia tridentata Pseodotsuga menziesii / Symphoricarpos albus P. menziesii / Cj_ rubescens P. albicaulis / V. scoparium Non-forest habitat types: FEID/AGSP FEID/AGCA FEID/AGCA-GEVI FEID/DBCE DECE/Carex spp. ARTR/FEID ARTR/FEID-GEVI Festuca idahoensis / Agropyron spicatum F . idahoensis / A1. caninum F . idahoensis / A^ caninum-Geranium viscosissimum F. idahoensis / Deschampsia caespitosa D . caespitosa / Carex spp. Artemisia tridentata / F^ idahoensis A. tridentata / F . idahoensis-G. viscosissimum 8Gneiss Creek/Hebgen Lake Sub-Area This area lies along the western boundary of Yellowstone Park and adjacent parts of the Gallatin National Forest. It is bounded on the west by Hebgen Lake and on the east by the southern end of the Gallatin Range. It is dissected by several major drainages, most notably Gneiss Creek, Cougar Creek, and Teepee Creek. Most of the area is dominated by climax lodgepole pine with bitterbrush (Purshia tridentata) understory (PICO/PUTR habitat type) at elevations of 1980- 2440 m. Large marshy areas with thick stands of willow (Salix spp.) occur along the lower reaches of Gneiss Creek, Cougar Creek, and around Hebgen Lake. The northern and eastern portions of this subarea are topographically more complex with spruce and fir commonly occurring with the lodgepole. Common habitat types include ABLA/VASC (VASC and CARU phases), ABLA/THOC and other ABLA habitats. Scattered stands of aspen are found throughout the area. Mesic grass/forb meadows are located along many of the creek bottoms. Sub-xeric open areas with sparse to moderately-dense stands of big sagebrush in the ARTR/FEID habitat type (GEVI phase) are found along certain ridges and southern exposures. Bear-human conflicts are a major concern in this area. Residential and summer homes are scattered along Duck Creek and around Hebgen Lake. Both Cougar Creek and Duck Creek are popular for fishing. Several campgrounds and summer resorts are also located in the area. 9Nez Perce Cr./Firehole River Sub-area This subarea is located north of Old Faithful in the west- central part of Yellowstone Park. The majority of this area lies within the vast lodgepole-pine zone in the western reaches of the Central Plateau. The most abundant habitat types are ABLA/VASC-VASC and ABLA/CAGE. Serai stages of both types are dominated by lodgepole pine. Engelmann Spruce (Picea engelmanii) and subalpine fir are also common in the overstory. Whitebark pine is present in isolated pockets. Nez Perce Creek, Sentinel Creek and the Firehole River are the major waterways. Wet forests consisting of several habitat types occur along these streams. The ABLA/CARU habitat type is common. The Lower Geyser Basin lies in the center of this sub-area. Elevations range from about 2320 m over most of the area to a maximum of 2600 m on Mary Mountain. Human activity is restricted primarily to the thermal areas around the geyser basin and to fishing along Nez Perce Creek and the Firehole. The Grand Loop road also passes through the center of this sub-area. Blackball Plateau/Washburn Range Sub-area This is the largest of the three sub-areas. It encompasses the high peaks and alpine tundra of the Washburn Range and the rolling, grassy slopes along Antelope Creek and in the Blackball Plateau. Much of this sub-area lies within the spruce-fir zone and, along the Yellowstone River, the Douglas-fIr zone. The most abundant habitat types are ABLA/VASC-VASC and ABLA/VASC-PIAL. Other common types are the ABLA/THOC, ABLA/CARU and PSME/SYAL habitats. Whitebark pine Ioccurs at mid to high elevations in a number of locations. Large sub- xeric meadows of grasses and various forbs with scattered pockets of timber are present in the Blacktail Plateau and Antelope Creek areas. Common grassland habitats are the ARTR/FEID, FEID/AGSP and FEID/AGCA- GEVI types. The elevation ranges from 2075 m in the Blackball Plateau to 3125 m on Mount Washburn. 10 Climatology Temperature and precipitation data were abstracted from annual summaries for the "Yellowstone Drainage, Wyoming" division (with reporting stations at Mammoth, Tower Falls, and Lake in Yellowstone Park and at Crandall Creek and Clark east of the park, NOAA 1980-82). Snow survey data were collected at Canyon, Norris, West Yellowstone, Old Faithful, and Lupine Creek (SCS 1983). Snowfall for December 1980 through April 1981 (winter preceding the first field season) was 4.1% below the long term mean. Temperatures for the same period were well above normal. The snowpack (in water equivalent inches) for late winter/early spring 1981 was well below normal at all reporting stations . (means of 54% and 55% of the long-term average in March and April, respectively). Heavy rains in May and June compensated for the light snow pack so that cumulative precipitation for the primary growing season (May through July) was 28.9% above normal. Temperatures for this period also averaged slightly above normal. The remainder of the first field season (August and September) was hot and dry, with temperatures averaging 6.9% above the norm and precipitation 50% below normal. 11 Precipitation for the December 1981 through April 1982 period (winter preceding the second field season) was 37.5% above normal and temperatures averaged 8.1% below the long term means. Cumulative snowpack in March and April was 102% and 117%, respectively, of the long-term average. Precipitation for the May to July 1982 period was only slightly (3.3%) below normal so that overall available moisture for the primary growing season was at or above normal. The remainder of the second field season (August and September) was unusually wet (precipitation 32.9% above normal) with temperatures averaging 5.0% above normal. Despite the apparent contrasts between the weather patterns for the 1981 and 1982 field seasons, timing of phenological development appeared to be fairly similar for both years. Peak succulence was in late May to June and most plants were in flower or fruit by early to mid July. The similar phenology during the two years may relate to the fact that many of the important bear forage species were high altitude perennials which characteristically have a climate-species transfer function of about two years (Picton pers. comm.). METHODS Data Collection Trapping and Radio-Tracking Grizzly bears were trapped by the Interagency Grizzly Bear Study Team (IGBS) or by National Park Service (NPS) personnel using culvert traps or Aldrich foot snares. Eleven different grizzlies were radiotracked during various phases of the study. Data from all eleven bears were used in the activity pattern analyses. Five bears were selected (according to the acquired data base and the availability of prior data for each bear) for a more detailed analysis of the activity pattern/habitat use interaction (Tables 2 and 3). The location of instrumented bears was first determined from routine aerial surveys by IGBS personnel. A temporary monitoring station was subsequently established at a suitable vantage point near the "target" bear. Several different tracking systems were used. A hand-held, two-element Yagi antenna was generally used for ground radio-tracking. A truck-mounted 4-element Yagi antenna which could be elevated and rotated 5 m above the cargo bed was used during the second (1982) field season. This system proved useful while night­ tracking from roads or when slight elevation changes provided a much improved signal. 13 Table 2. Distribution of activity records by bear, season, and year. Each activity record represents one monitoring period, ranging from 10 minutes to 2 hours, during which the bear's activity status was determined. Bear No. Spr 1981 Summ * Fall Spr 1982 Summ Fall Both Years Spr Summ Fall 15 0 56 24 0 0 0 0 56 24 38 75 97 0 0 0 0 75 97 0 50 40 78 50 0 0 0 40 78 50 59 0 0 0 0 44 0 0 44 0 76 0 0 0 0 124 0 0 124 0 others 0 49 54 0 151 0 0 200 54 total 115 280 128 0 319 0 115 599 128 A Spring=March-May; Summer=June-August; Fall=September-October Table 3. Age, bears this sex, . Ages study. and monitoring period for the five primary study given represent age when bear was monitored for Bear # Age Sex Monitored Yr. and Season(s) 15 11 1982 :summ, fall'*' 38 10 F2 1981:spr,summ 50 AdJ F 1981:spr,summ,fall 59 4 F 1982:summ 76 2 F 1982:summ I Bear 15 was also tracked with trained bear dogs in 1981. 2 None of the females radiotracked during the study were accompanied by cubs. 3 Ad=Adult of undetermined age (5 yrs or older) Activity Monitoring All of the study bears except #15 were equipped with tilt collars (Telonics, Inc.) designed to indicate activity status according to movements of the head. Frequent fluctuations from one pulse mode to the other suggested that a bear was active. Bear #15 was instrumented with a standard (non-tilt) collar. His activity status was 14 ascertained via the "integrity" of the signal. Such factors as erratic signal strength, obvious directionality changes, etc., were regarded as probable indicators of activity. However, other investigators have found that this method was often an unreliable indicator of activity (Lindzey and Meslow 1976; Garshelis and Pelton 1980). Activity was occasionally determined from direct observation. At regular intervals, generally hourly during the day and bihourIy at night, the bear's signal was monitored for a minimum of 10 to 15 minutes. During each monitoring session, the bear's activity level was recorded using one of five activity designations: inactive, mostly inactive with brief intermittent activity spurts, evenly divided between active and inactive periods, mostly active with intermittent quiet spells, and entirely active (erratic mode shifts almost constantly). These were later reduced to two categories, active or inactive, to facilitate analysis. During each monitoring period, data on precipitation, cloud cover, ground cover, wind velocity, wind direction and temperature were collected. These data were later analyzed to determine their influence, if any, on bear activity patterns. Habitat Use and Scat Data Collection The precise location of the study bears was determined by close range triangulation whenever possible. During the night, field situations generally precluded triangulation fixes, thus the bear's approximate position was deduced from single bearing fixes, signal integrity, intervening topography, etc. Locations derived from these 15 latter indices were considerably less reliable than triangulation fixes and the resultant data were treated accordingly. Once the bear had vacated an area, investigators returned to the site and searched for feeding activity, scats, day beds, and any additional evidence of bear activity. A "community site analysis" (Appendix I) was conducted according to the procedures outlined by the IGBS (Knight et al. 1984) at all locations where bear activity was encountered. When no bear sign was found, but radio fixes were believed to be accurate, community site analyses were also completed. These analyses included a determination of habitat type following Mueggler and Stewart (1980) for grasslands and shrublands and Pfister et al. (1977) for forested areas. Species lists were prepared at all community sites. Cover values were estimated ocularly for each species recorded. Specimens which could not be identified at least to genus in the field were collected for later identification. Questionable specimens were verified at the Yellowstone herbarium at Mammoth. Plant nomenclature followed Hitchcock and Cronquist (1973). Relevant topographic and physiognomic features were also noted during each community site analysis. Scats of the year were routinely collected whenever encountered. Scats were analyzed for content by Montana Fish, Wildlife, and Parks personnel as described by Knight et al. (1980). Dog Tracking Trained bear dogs were used during July 1981 in an effort to obtain more precise data on bear microhabitat use. Two professional 16 bear hunters/guides from Oregon volunteered their time to help with this phase of the study. They brought eight of their hounds (Plotts and related breeds) to Yellowstone in July 1981. Two basic tracking strategies were used. One was to take the hounds to a site known to be frequented by bears and then to track a grizzly away from this site. Such a site was available where a rendering plant truck overturned along Highway 191 two years previous (Schleyer 1983). The accident left considerable animal fat in the topsoil which served as an ersatz bear lure for our purposes. In this case, the bear's identity and current location were generally unknown. The other approach was to determine the day bed location of a radio-collared bear by triangulation, and then return the following morning to retrace the bear's overnight movements starting at the bed site. This strategy was preferable since periodic radio fixes alleviated the possibility of unpleasant encounters with the study bear. The dogs were equipped with loud bells and were restrained on leashes at all times. Despite these precautions, NPS officials concluded that due to inherent risks and possible visitor annoyance, the use of bear dogs was an unacceptable approach to bear research within the confines of the Park. Consequently, the canine collaborators were, retired from active duty. Data Analysis Activity Data The influence of temporal and environmental factors (e.g., time temperature...) upon bear activity level wasof day. season, 17 determined by assigning all non-active observations a numerical value of "0" and all active observations a value of "I." Then the mean probability of activity (from 0.0 to 1.0) for a particular variable was calculated. Mean values of 0.5 indicated an equal probability of activity or inactivity while values greater than or less than 0.5 indicated probabilities of higher or lower activity, respectively. SPSS computer programs were used for all initial breakdowns and statistical analyses. One-way and two-way analyses of variance (Anova) were used to test for significant differences in activity level due to temporal, environmental, and individual bear effects. Tukey1s Honestly Significant Difference (HSD) test was used as a post- hoc procedure to assess differences in individual means. Community Site Data Analysis Five bears (#15, #38, #50, #59,and #76, Tables 2 and 3) were selected for a more in-depth analysis of the interaction between activity patterns (from activity monitoring), food habits (as evidenced by scats), and. habitat use (from aerial relocations and community site analyses). Community site analyses were routinely conducted as followups to bear relocations or whenever bear sign was encountered in the field. Only those community sites associated with the five bears mentioned above were treated as follows. Data from all remaining sites were merged with existing IGBS data and applied to other ongoing analyses. Community site "quality" was evaluated according to three criteria: food value, total understory cover, and understory species I 18 diversity. Food value was determined by the density within the site of the 22 most important bear food plants as determined from prior analysis of IGBS community site and scat data, 1977-1982. Animal matter was handled separately. The relative food value, FV^, of each of these 22 items was also previously determined (Mattson, in prep.). FV^ incorporates information on the following: 1. Intrinsic energetic efficiency (EE^) of each item (subjective evaluation of energy expended to acquire characteristic bite volume versus digestible energy per bite). 2. Monthly food item preference, PF^, for item i: PFi - (Vol1 / Freq1)/PFmax where Vol^ = % of total scat volume comprised of item i, Freq^ = % frequency of item i occurrence in scats of the sample and ^ max = maximum calculated value of (Vol^ / Freq^). -I 3. Characteristic "contagiousness" (A^ ) of a given item or group. This value expresses the tendency of a particular diet item to occur in aggregations. The premise is that a bear is capable of feeding more efficiently on a given item if it tends to occur contagiously, as do certain root foods. The relative food value of a given item is, then, FV. = EE. x PF. x A."1 i i x i Since PF^ values were calculated by month, FV^ values likewise vary by -I month. Tables of all the FV^, EE^, PF^, and A^ values used herein are given in Tables 10 and 11. A more detailed derivation of each parameter is available in Mattson (in prep.). 19 Individual food values were then weighted by their cover at each site, (X , and summed to give the composite food value, FV, at each community site: FV = ZCFV1 x C1) Total understory cover, Cy, for each community site was found by adding the cover class value for shrubs to the cover class value for herbs as recorded on the community site field forms (Appendix I). The resultant value enabled a rough comparison of cover between sites. Understory species diversity, Hyf was obtained by applying the Shannon-Weaver diversity (information measure) index to all understory species (shrubs and herbs) recorded on the community site forms. All species were included regardless of whether they were known bear foods or not. Hence: H U = - s PiC1Oge Pi) where P1 = relative abundance (cover) of species i from O to 1.0. Once the values for community site food value, understory cover, and understory species diversity were calculated for all community sites, these three values were combined into a single quantitative index of community site quality (CSQ). First, each of the raw values for FV, C^, and was converted to a proportional value by dividing by the maximum value. For feedsite x: FV = FV /FV X x max C = C /Cu,x u,x u,max H H /Hu,x u,x u ,max Of these 3 variables food value (FV) was considered the most important 20 determinant of community site "quality," hence it was doubly weighted relative to cover and diversity. Therefore, the Community Site Quality index, CSQ, was found by: CSQ (FV x 2) + (C )+(H ) U 0X ' u,x This expression was then proportionally adjusted to a 0.0-1.0 scale by dividing by the maximum value. For community site x: CSQx - CSQxZCSQnax Finally, the mean CSQ for all community sites ascribed to a given bear was used for comparison with other bears. Computer Relocation Habitat Scans The community site analysis discussed in the foregoing section provided one means of assessing the quality of habitat occupied by individual bears. As a second approach, all of the radio relocations which occurred between the first and last days of activity monitoring for the primary study bears were subjected to an additional "habitat richness" analysis. Both aerial and ground relocations were included. Habitat mapping of Yellowstone National Park was conducted by IGBS and NPS employees from 1979 through 1982 according to the classification system of Mueggler and Stewart (1980) and Steele et al. (1979). Several hundred plots were used to delineate habitat types on airphotos. Resolution of the habitat mapping was roughly two hectares. The data were transferred to 15 minute topographic quandrangle maps and subsequently digitized to facilitate computer analysis. A spatial information computer software package was developed by Wm. Hoskins for the IGBS to treat relocation habitat data. As input. 21 this program receives the UTM coordinates for a given bear relocation. It then generates a "scan circle" of 0.5 km radius with the relocation point as center. This scan circle was used because most radio fixes were obtained during the day and, therefore, a description of habitat use based entirely on the immediate habitat in which the relocation points fell would tend to overrepresent day bed sites. The 0.5 km radius partially corrects for this bias by incorporating not only the relocation point proper, but also a broader area surrounding this point through which the bear presumably entered or left the site. Within the scan circle, a sampling grid is created with sample points every 88 m (for a total of ~101 points per scan circle). For each scan circle, the number of points falling within each habitat type is output along with the amount of edge (E, interface between adjacent habitats) occurring within the scan circle. The scan circle data were pooled by season for each of the five bears to give the proportionate use of habitats by season. The next step in this analysis was to rate the overall quality of habitat used by each bear according to the relative value of each component habitat type. Mattson (in prep.) developed the concept of "unit area importance values" (IV^) to describe the relative importance of individual habitat types to Yellowstone bears. IV^'s incorporate information on characteristic food items found in habitat x, the inherent food value of these items, the consistency with which each item is available in habitat x from year to year (i.e., annual flux in availability), the apparent preference toward habitat x for feeding on these food items (extrapolated from community site analysis 2 2 data) arid the diversity of feeding opportunity in habitat x. Since I the value of many food items varies according to phonological stage, IV 's are calculated on a seasonal basis- The IV 's used hereinu u pertain only to the vegetative attributes of a habitat type: animal matter is handled separately. These IVu*s are tabulated in Table 12. A single "area food score" (FSq) was calculated for each of the five bears by multiplying the proportionate use of each habitat by the corresponding seasonal IV value. FSa - S (Px>1n v u(Xji)) where p . is the proportionate use of habitat x in season I and x, I IVy^x ^ is the corresponding unit area importance value for habitat x in season i. The diversity of habitats within each scan circle was calculated using the Shannon-Weaver diversity index: Hh = - E Px(IogePx) where px is the proportional representation of habitat x in the scan circle. The mean value for Hh from all scan circles for a given bear was used for comparisons. The mean amount of edge, E, per observation was also calculated from the scan circle data. Each of the three parameters described above (area food score, habitat diversity, and amount of edge) was adjusted to a 0.0-1.O scale by dividing the value for a given bear by the maximum value, so that for bear y: » = FS /FSa,y a,y a,max h,y E Hh,y^Hh,max E /E y max The final step in the relocation data analysis was to combine the 23 three parameters described above into a single expression for relocation habitat richness (RHR). The area food score (FS) was given double weight so that for each bear: RHR = (2 x FSa) + (E) +(H^) The resulting values were again adjusted to a 0.0 to 1.0 scale by dividing by the maximum value for RHR. For bear y: RHR =RHR /RHR y y max Scat Data Analysis Scats for the five primary study bears were analyzed for diet item content and diet item richness. Only those scats which could be positively ascribed to a particular bear were included. These were generally collected at day beds or immediately following multi-bearing radio fixes. Each scat was scored according to the energetic efficiency (EE, pl8) of individual food items found in that scat. The EE^ for every "natural" (i.e., non-garbage, see below) food item, i, found in the scat was weighted by the proportional volume of item i to derive a Scat Value (SV): SV =Z. (P1EEi) where p^ = relative proportion of item I in the scat and EE^ .= energetic efficiency for item I. The mean SV score for each bear was adjusted to a 0.0-1.0 proportional scale by dividing the raw value by the maximum value (SV/SV^ax). The adjusted values were then used to make inter-bear comparisons. The percent by volume of digestible items (i.e., after deducting 2 4 volumes of dirt, debris, etc.) belonging in certain important food groups (e.g., forb, ungulate, shrub...) was also calculated for each bear. "Garbage" referred to any items of human origin and was treated as "digestible" matter despite the obvious indigestibility of plastic, etc. Minimum Daily Movements and Home Range Estimates Minimum daily movements Were calculated from consecutive daily radio fixes. Both ground triangulation fixes and aerial relocations were used in this analysis. This method can result in serious underestimation for those bears which tended to forage over long circuitous routes and then return to preferred bed sites near the previous day's relocation. Consequently, these data are best regarded as an index of movement patterns rather than as an estimate of the actual distance travelled. Home range areas were calculated for the five primary study bears for the period when they were monitored only. Areas were based on the determinant of the recapture point covariance matrix (Jennrich and Turner 1969). The areas thus obtained were not directly comparable since there was wide disparity in the length of the monitoring period for the five different bears (from 46 days for //15 to 150 days for #50). To facilitate inter-bear comparisons, the area for each bear was divided by the number of relocations contributing to the area estimate. The resulting "area per relocation" is used for comparisons herein. 4RESULTS Activity Data 25 Temperature Activity data were grouped into 5 C blocks and analyzed on an annual and seasonal basis (Figures 2 and 3). Anova indicated that annually temperature did have a significant effect on bear activity level (p<.001). The temperature data yield a more-or-less bell-shaped curve with a peak activity level in the 10-15 C range and low activity at both temperature extremes. Tukey's HSD test showed that activity at temperatures above 20 C was significantly lower than activity in the 5-20 G range (p<.05). A two-way Anova showed that jointly both season and temperature had significant main effects (as in the univariate Anova's for both variables). There was also a significant interaction factor (p=.002) indicating a significant difference in bears' responses to temperature in the different seasons. Since the majority of the temperature observations (71.0%) were collected during the summer, the temperature histogram for summer (Figure 4) approximates the annual histogram with high activity in the 0-20 C range and lower activity in the two temperature extreme ranges. The fall data (Figure 5) followed the same general trend, although the activity level was lower. The spring data (Figure 3) suggests a response opposite to the summer and fall .PA 26 U-O I -OO .90 .80 .70 .60 >- • 50 CD < CDO OC O- • 40 • 30 20 • 10 NO- OF RECORDS 7 78 196 182 119 87 22 -SC-OC 0C-5C SC-1OC 10C-1 SC 15C-20C 20C-25C 25C-30C Figure 2. Relationship between temperature (in C) and probability of bear activity. All temperature data from both monitoring years are included. % OQ < CQOOd CL NO- OF RECORDS 25 49 16 I I 0C-5C SC-1OC IOC-1 SC 15C-20C Figure 3. Relationship between temperature (in C) and probability of bear activity in spring (May). Temperature data from both years are included. 27 Figure 4. Relationship between temperature (in C) and probability of bear activity in summer (June-August). Temperature data from both years are included. 5 I .00 .90 .80 .70 .60 .50 $ CDO0£ CL .40 30 • 20 -10 NO. OF RECORDS 0C-5C 5C-10C IOC-1 SC 15C-20C 20C-25C Figure 5. Relationship between temperature (in C) and probability of bear activity in fall (September). Temperature data from both years are included. 28 pattern: activity was greatest in the 0-5 C and 15-20 C ranges and much lower in the intermediate temperature ranges. Precipitation and Ground Moisture Sample sizes were small for all but the "no precipitation" category. Activity levels were lowest for the two ."raining" categories (i.e., rain and intermittent rain. Figure 6). For the Anova, all five precipitation classes were grouped together and compared to the no-precipitation class. This analysis indicated that bears were significantly less active during precipitation than when there was no precipitation (p=.0032). The response to precipitation was partially temperature-dependent, with greater activity during rain at warm (>10 C) temperatures than at colder (<5 C) temperatures. It was also noted that bears tended to increase their activity at the onset of a rain storm as if briefly agitated, and then their activity quickly tapered down. The amount of agitation seemed to be correlated with severity of the storm. Ground moisture had a significant effect on bear activity (p=.0431, Figure 7). Bears were most active when the ground was dry and about equally active when the ground was moist or wet. However, Tukey1s HSD post-hoc procedure showed that no two groups were significantly different at the .05 probability level. Wind Speed Anova revealed significant differences in bear activity levels at; different wind speeds Cp=ZOlS, Figure 8); however, the differences were relatively minor and it is doubtful that they reflect legitimate 29 I OO NO. OF RECORDS 717 22 22 54 10 NO PREClP MISTING RAINING INTERMlTTANT SNOW/HAIL RAIN Figure 6. Relationship between precipitation type and probability of bear activity. All precipitation data from both monitoring years are included. >- I-OO OF RECORDS 577 86 157 6 DRY MOIST WET SNOW TRACE Figure 7. Relationship between ground moisture and probability of bear activity. All ground moisture data from both monitoring years are included. 30 biological variation. Tukey's HSD test showed that no two groups differed significantly at the .05 probability level. This result may relate partially to unequal sample sizes since the range (differences in means) required for significant differences in Tukey's test becomes larger with disparate sample sizes. Cloud Cover Anova indicated that bear activity differed significantly with different degrees of cloud cover (p=t.009. Figure 9), but, as with wind speed, the differences were not great enough to be biologically persuasive. Tukey's HSD test indicated that bears were significantly less active when it was overcast than when the cloud cover was 10% or less (p<-05). Seasonal and Monthly Effects Seasonal activity data were very limited. The earliest spring observations, were roughly 45 days after den emergence and the latest, fall observations were made about 30 days before the average den entrance date. Thus, * the activity patterns described below for May and September may not accurately represent the activity program for all of spring and fall, respectively. Anova demonstrated a highly significant difference in bear activity levels by month (p<.001). Activity was highest in July (Figure 10). The activity level in July and August was significantly higher than activity in May or September (p<.05). 31 LuO 5ODO (XL O- NO. OF RECORDS 62 491 162 93 10 NO WIND 0 - 5 5 - 1 0 1 0 - 2 0 20* »ph Figure 8. Relationship between wind speed (in km/hr) and probability of bear activity. All wind speed data from both monitoring years are included. % CD < CDO CC Q- 1 .00 .90 .80 .70 60 .50 .40 30 20 10 NO. OF RECORDS 0 6 5 0 6 7 0.55 259 220 97 239 10% OR LESS SCATTERED BROKEN OVERCAST SKY COVER CLOUDS CLOUDS Figure 9. Relationship between cloud cover and probability of bear activity. All cloud cover data from both monitoring years are included. (Scattered clouds - 10-50% cloudy, broken clouds = 60-90% cloudy, overcast = >90% cloudy.) 32 Time of Day Effects Annual Activity Patterns The four diel time periods were defined as follows: sunrise= one hour before to one hour after sunrise diurnal= one hour after sunrise to one hour before sunset sunset= one hour before sunset to one hour after sunset nocturnal= one hour after sunset to one hour before sunrise Annually there was a highly significant difference (p<=001) in bear activity levels at different times of day (Figure 11). The post- hoc procedure revealed that bears were significantly (p=.05) less active during the day than during the other three diel divisions. They were most active during the two crepuscular periods with significantly greater activity at sunset than at night (p<.05). Bears also displayed highly significant differences in activity level according to hour of the day (p<.001, Figure 12). There were two primary activity peaks at 0500 and 2200 h (MDT) with an intervening period of significantly (p<.05) lower activity from early to mid afternoon. The lowest overall activity was at 1500 h. Bears often went through a transition phase of brief erratic activity in the evening. This phase usually lasted from 15 minutes to an hour depending on the individual bear and seemingly represented a period of restlessness or agitation prior to becoming consistently active. Schleyer (1983) described a similar "winding up" interval for his study bears in Yellowstone. 33 Figure 10. Relationship between month and and the probability of bear activity. Data from all bears and from both monitoring years are included. Spring Summer F= I I Yearly Ave SUNRISE DIURNAL SUNSET NOCTURNAL Figure 11. Relationship between diel period and probability of bear activity annually and seasonally. Data from all bears and from both monitoring years are included. P r o b a b i I it y o f Ac t i v i Iy I .00 .90 .80 .70 60 50 .40 30 20 -10 T i m e o f Da y ( M • D . T . ) LU L e g e n d Al I Be a r s Figure 12. Probability of bear activity according to hour of the day annually. Data from all bears and from both monitoring years are included. 35 Seasonal Activity Patterns by Diel Period Univariate Anovas indicated that both time of day (dial period) and season individually had significant effects on bear activity levels (p<.001 for both variables). A two-way Anova supported these results and showed that jointly time of day and season had significant effects (p<.001) on activity and that there was no interaction (p=.219). Bears tended to be crepuscular in all three seasons, but the period of greatest and least activity varied by season. In spring (Figure 11), bears were about equally active at sunrise and sunset (mean probability of activity, x,=.80 and .82, respectively) and least active at night (x=.38). The morning peak occurred at 0600 h and the evening peak at 2000-2200 h (Figure 13). Bear activity was relatively low during both day and night, but a brief period of increased activity occurred around 1100 h. In summer, bears were most active at sunrise (x=.94) and least active during the day (x=.60). Activity peaks occurred from 0400-0700 h and from 2100-2200 h (Figure 14). The lowest activity was at 1500 h (x=.32), but bears were often active even at this time. Bears generally became active from 1600-1700 h with activity steadily increasing until the late evening peak noted above. On the whole, the grizzlies’ activity patterns were less erratic in summer than in spring and fall. Non-active periods were less likely to be punctuated by periodic spurts of activity and, once they became active, grizzlies tended to remain active for longer periods. Some of this "evening out" effect was probably due to the greater sample size, in' summer. In P r o b a b i I i t y of A c t i v i t y I -00 ■ 90 • 80 • 70 • 60 50 - 40 ■ 30 ■ 20 • I 0 T i m e o f D a y ( M . D . I . ) WOx L e g e n d A l l B c o r s - . - S p r i n g igure 13. Probability of bear activity according to hour of bears and from both mbnitoring years are included. the day in spring (May). Data from all P r o b e b i I i t y o f A c t i v i t y I -OO 90 ■80 ■ 70 ■ 60 50 •40 30 w >4 20 • I 0 oO § O OO Oro oo O B I O O O - Oo U) o rv U A Ol <7> -Xr-\ f~\ \ « R->O O O O O V v z v V w Uo o o o o o o I i m e o f D a y ( M . D . T . ) L e g e n d A M Bears - - Summer Oooo ro u oo ivAOO Figure 14. Probability of bear activity according to hour of the day in summer (Jun-Aug). Data from all bears and from both monitoring years are included. fall, grizzlies were most active at sunset (x=.71) and least active during the day (x=.44, Figure 11). Unlike spring and summer, grizzlies wete not highly active at sunrise during the fall (x=.46). Activity was lowest from 1300-1700 h in early afternoon (Figure 15). Grizzlies became active beginning around 1800 h and activity reached a peak from 2200-2300 h. The drastic activity oscillations indicated during the late night (Figure 15) are probably attributable to small sample size rather than to actual shifts in bear behavior. Activity quickly tapered down after 1000 h. Individual Bear Patterns Anova revealed a significant difference in the mean activity level of the different study bears (p<.001. Figure 16). The post-hoc procedure, Tukey's HSD test, showed that the mean activity level of bears #50 and #38 was significantly lower than the mean activity level of bears #59 and #76. The activity level of individual bears varied according to diel period (Figures 17 and 18). In a two-way Anova, both bear and diel period had significant main effects (p<.001). There was also a significant interaction (p=.007) between these two variables indicating that individuals behaved differently over the four diel time periods. All of the five primary study bears were least active during the day (Figures 17 and 18). All but #15 were also highly active during both crepuscular periods. Three of the bears (#15, #38, and #76) were most active at sunrise. Bear 50 was most active at sunset and Bear 59 38 P r o b a b i l i t y of A c t i v i t y I .00 90 80 .70 60 • 50 •40 • 30 20 • 10 T i m e o f D a y ( M • D • T . ) UJ VD L e g e n d A I I B e a r s - • F a l I Figure 15. Probability of bear activity according to hour of the day in fall (Sep). Data from all bears and from both monitoring years are included. 2400 40 I -00 NO. OF RECORDS 172 168 12* 44 80 BEAR «38 BEAR »50 BEAR »76 BEAR »59 BEAR »15 Figure 16. Overall probability of activity for the five primary study bears. All activity records from each bear are included. All Beers Beer *76 Beer *59 SUNRISE DIURNAL SUNSET NOCTURNAL Figure 17. Probability of activity according to diel period for all monitored bears (including all activity records from all bears for both monitoring years) and for primary study bears #76 and #59. 41 was active every time she was sampled except during the diurnal period. The activity pattern of grizzly #15 departed from the usual trend in several respects. He was much less active at sunset than any other bear and more active at night than any other bear except #59. He also showed lower than average activity during the day. Schleyer's (1983) data likewise indicated that the activity level of #15 was less than (or equal to) the activity level of all other bears for the diurnal and sunset periods. However, his data did not indicate unusually high activity for #15 at night. Bear #50 was the only grizzly with data from all three seasons (Figure 19). During spring and fall, she adhered to the normal pattern of high crepuscular activity. In summer, she was more nocturnal than in the other seasons. Bea r If 50 I .in Bea r * 3 8 — — — —. — — . B e a r * 1 5 SUNRISE DIURNAL SUNSET NOCTURNAL Figure 18. Probability of activity according to diel period for primary study bears #50, #38, and #15. 42 Spring Summer Fall Yearly Ave SUNRISE DIURNAL SUNSET NOCTURNAL Figure 19. Probability of activity according to diel period for Bear 50 annually and by season. Community Site Analysis Values for community site food value (FV), total understory cover (C^), understory species diversity (Uu)» and community site quality (CSQ) for all community sites ascribed to the five primary study bears are given in Table 13. Mean values for each of these parameters are given in Table 4. Bear 38 had the highest mean CSQ (proportional value of 1.00) of the five bears. Her high FV value resulted primarily from high densities of grasses and sedges at most of her relocation community sites. Mean understory cover was also highest for Bear 38. Bear 76 had a mean CSQ value of .99, only slightly below that for //38. Like Bear 38, high densities of grasses, sedges, and forbs at 43 most of # 7 6 ' s community sites accounted for the high FV and CSQ values. Table 4. Mean values for Food Value (FV), Understory Cover ( C ) 9 Understory Species Diversity ( H ) 9 and Community Site Quality Index (CSQ) for Bears 15,38^50,59, and 76. Values for individual community sites are given in Table 13. Bear # n FV C U Hu CSQ 15 4 .38 .53 .67 .73 38 8 .67 .70 .64 1.00 50 5 .41 .50 .63 .73 59 5 . .59 .53 .68 .89 76 3 .69 .62 . 66 .99 Bear 59 had the median CSQ value (.89). Her FV scores varied widely. Several sites had medium to high densities of graminoids. V Whitebark pine nuts were available at two of # 5 9 -s July community sites; however, pine nuts are only of moderate value in mid-summer (see Table 11), and there was no direct evidence that Bear 59 was feeding on pine nuts at that time. #59 had the highest mean understory diversity of the five bears, testament to the wide variety of feeding opportunity in the mesic meadows along Antelope Creek. Bears 15 and 50 had equivalent mean CSQ values of .73. Both bears tended to occupy sites with a lodgepole overstory and a sparse understory barren of most important forage species. Grass/sedge density was very low at all but one of #15*s sites and low to moderate at all of #50's sites. Bear foods at #15's community sites consisted primarily of graminoids and mixed forbs. Forage species were slightly more plentiful at #5'0's sites, consisting of grouse whortleberry 44 (Vacclnium scoparium), several tuberous species (Claytonia and Lomatium) and various forbs and grasses. Relocation Habitat Scans Bear 76 received the highest area food score (FSq) of the five primary study bears (Table 5). Most of the relocations for #76 were made in the Blacktail Plateau area southwest of Tower Falls and in the western portion of the Washburn Range. The habitats occurring most frequently in these relocation circle scans were ABLA/VASC-PIAL (35%), .FEID/AGCA-GEVI (24%), and ABLA/VASC-VASC (18.5%). Bear 76 received the highest scores for mean habitat diversity (H^) and mean amount of edge per relocation (E). Her relocation habitat richness (RHR) score was higher than any other bear's, primarily as a result of her exceptionally high ELA score. Table 5. Grizzly bear Area Food Scores (FS ), mean amount of Edge per relocation scan circle (E), mean Habitat Diversity in scan circles (H, ), and Relocation Habitat Richness scores (RHR). All values" calculated from 0.5 km radius computer scan circles around bear relocation points. Both raw values and adjusted (proportional) values (in parentheses) are given. Bear FSaL E 5H RHR 15 0.0210 (.001) 671.4 (.32) 0.23 (.33) 0.652 (.18) 38 14.1688 (.520) 1401.1 (.66) 0.55 (.79) 2.490 (.68) 50 11.1243 (.409) 939.2 (.44) 0.39 (.56) 1.818 (.50) 59 17.5651 (.646) 2114.6 (1.0) 0.70 (1.0) 3.292 (.90) 76 27.1866 (1.00) 1662.3 (.79) 0.61 (.87) 3.660 (1.0) Bear 59 received the second highest RHR score of the five bears. The habitats with the highest frequencies in the scan circles were ABLA/THOC (23%), ARTR/FEID-GEVI (20%),. and ABLA/VASC-VASC (18%). The 45 ABLA/VASC-PIAL and ABLA/CACA habitats, both of which have high IV values, 'also had high frequencies. Bear 59 received the highest scores for habitat diversity and mean edge per relocation. Bear 38 received the median RHR score. The habitats represented most frequently in her scan circles were ABLA/VASC-CARU (25%), the PICO/PUTR (25%), and the PSME/CARU (15%). The ARTR/FEID-GEVI and FEID/AGCA-GEVI habitats were also well represented. Bear 38 had the median value for area food score, habitat diversity, and edge among the five bears. Bear 50 had a relatively low RHR score (Table 5). The three habitats occurring most frequently in her scan circles were ABLA/VASC- VASC (59%), ABLA/CARU (16%), and ABLA/CAGE (15%). Bear 50 had the second lowest values for area food score, habitat diversity, and amount of edge. Bear 15 had the lowest RHR value of the five bears as well as the lowest value for the other three estimates (Table 5). Eighty-seven percent of the habitat in his scan circles was PICO/PUTR habitat type with IV scores of only .038 and .003 in summer and fall, respectively. Eighteen of the twenty-six point scans for #15 were entirely PICO/PUTR— hence, the very low scores for habitat diversity and mean amount of edge per scan circles. Bear 15's abysmal area food score (FSq) of .001 (relative value) may slightly underestimate the feeding opportunity within his range.. Field.observations indicated that moister microsites scattered throughout the vast lodgepole plateau around Hebgen Lake had a somewhat richer understory than the area at large. Nonetheless, the overall characterization of Bear 15's 46 range derived from the RHR estimate is consistent with the observer's impressions that the area was vegetatively depauperate. Scat Analysis Scats from four bears, #15, #38, #50, and #59, were used for the Scat Value analysis; £he sample size for Bear 76 was too small to reliably indicate her food habits. Bear 59 had the highest mean SV of the four bears (Table 6). Sixty-one percent of the "digestible" scat volume (see p.24) was comprised of forbs (Table 7). The forbs with the highest representations were Claytonia lanceolata (16%) and Lomatium spp. (13%), both of which typically occur contagiously in specific microsites. Grasses and sedges comprised 34% of the scat volume and were found in 88% of the scats sampled. There was no garbage in Bear 59's scats and meat accounted for only 2% of the digestible scat volume. Table 6. Mean Scat Values (SV) scores for the primary study bears. Both raw values and adjusted (proportional) values (in parentheses) are given. Bear # n SV 15 6 0.323 (0.57) 38 12 0.505 (0.88) 50 15 0.533 (0.93) 59 17 0.571 (1.00) 47 Table 7. Scat summary for primary study bears: percent "digestible” diet volume and percent frequency occurrence for important diet item groups. Grass/ sedge Forbs Root Shrub Garbage Meat Bear 15 % Vol.J 14.6 4.2 45.8 27.1 (n=6) % Freq. 66.7 16.7 66.7 33.3 Bear 38 % Vol. 76.1 8.0 0.5 3.4 (n=12) % Freq. 100.0 25.0 8.3 16.7 Bear 50 % Vol. 8.1 19.0 16.2 48.4 (n=15) % Freq.' 26.7 33.3 20.0 60.0 Bear 59 % Vol. 33.9 61.1 0.6 1.9 (n=17) % Freq. 88.2 82.3 5.9 5.9 1 % Volume= {(Total % Item i) / [(nxlOO)-(Total % Debris,etc.)]} x 100 2 % Freq= {(Number of scats containing Item i) / (Tot. # scats)} x 100 Bear 50 had the second highest SV score of the four bears. Her diet was quite unlike that of #59, however. Forty-eight percent of the digestible scat volume was meat of which 5% was rodent and the remaining 43% was ungulate. Ungulate meat has one of the highest energetic efficiency (EE^) values of any bear food (Table 10). Sixteen percent of the scat volume was garbage or human refuse of some sort presumably discarded by hikers and fishermen along Nez Perce Creek and in the Geyser Basin. Twenty-seven percent of the volume was vegetation: 19% shrub (Vaccinium scoparium and Arctostaphylos uva- ursi) and 8% grass/sedge. The relative absence of vegetative matter in #50's diet is consistent with the low community site quality (CSQ) and relocation habitat richness (RHR) scores discussed earlier. (Recall that both of these expressions pertain only to the floral components of an area.) 48 Bear 38 fed primarily on grasses and sedges (76%) and forbs (8%) during this study. Gfaminoids appeared in all 12 scats of the sample. Animal matter (trout and ungulates) accounted for only 3% of the digestible scat volume. The relatively high SV score of .88 for #38 reflects the high EE^ value of the graminoids. The scat sample for Bear 15 included only six scats. Six additional scats were not collected in the field because they appeared to consist of nothing but indigestible plastic and trash. The results obtained from #15's scat analysis accord well with other indices of his activities (from dog tracking, movement patterns, and previous years' data). Bear 15 had the lowest SV score (.57) of the four bears. He also had the highest volume of garbage (46% of "digestible" volume and scat frequency of 67%) of any bear. Human refuse was readily obtained from the resorts, residential areas, and fishing sites around Hebgen Lake and Duck Creek. Meat (ungulate and large mammal) had a scat frequency of 33% and accounted for 27% of the digestible scat volume. Graminoids and forbs . had relatively low volumes, comprising 15% and 4%, respectively. Movements and Home Range Use Minimum daily movements for the five primary study bears ranged. from 2.7 km for Bear 59 to 5.9 km for Bear 50 (Table 8). Short term home ranges, as expressed as the "area per relocation" (p.24), ranged from 4.90 sq km for Bear 59 to 23.13 sq km for Bear 38 (Table 9). Correlations between movement patterns and home range use are treated in the discussion section. 49 Table 8. Consecutive minimum daily movements (km) for the five primary study bears. Bear # n Minimum Movement 15 15 4.1 38 5 4.8 50 7 5.9 59 16 2.7 76 9 5.1 Table 9. Short term home ranges for the primary study bears. Areas derived from method of Jennrich and Turner (1969) as calculated from all relocations between first and last days of activity monitoring for each bear. "Area per relocation" values (see p. 24) are used for comparisons herein. Total Area Area per Relocation Bear # n (sq km) (sq km) 15 27 333.25 12.34 38 21 485.81 23.13 50 42 335.58 7.99 59 33 161.58 4.90 76 29 322.06 11.10 Tracking Grizzlies with Bear Dogs Trained bear dogs were capable of accurately tracking individual bears for distances up to 15 km. Intermittent bear sign (tracks or scats) verified that they were on course. In several instances, the hounds led us directly to digs, day beds, or other activity which would have been extremely difficult to locate without their assistance. To better illustrate, one tracking session during which Bear 15 was followed will be described in detail. On July 15,1981, radio 50 triangulation placed #15 bedded in dense willows south of the Grayling Arm of Hebgen Lake. The following morning there was no signal from him in that vicinity, so the tracking party entered the area with the dogs. The dogs soon located #15's day bed from the previous afternoon and began tracking him through the willows. The grizzly's foraging followed a convoluted route through very dense brush. Twice he stopped at isolated lodgepole "islands" and left a single scat. At one of these sites, he appeared to have rolled around in the grass or possibly bedded down briefly. The bear crossed Duck Creek (~10 m wide) several times while moving through the willows. He travelled along adjacent dirt roads and cattle trails (where his tracks were very apparent) for distances up to I km. More scats were found along these stretches. Bear 15 eventually crossed Hwy. 191 and entered a xeric lodgepole area. He followed a logging road through a recently cut area and soon arrived at the primary road into a residential area. Here he walked along a driveway passing within 30 m of one home, waded across a nearby pond and headed north. He travelled over a ridge and recrossed the highway by Grayling Creek. A grizzly was observed at this point at 2:30 that morning. He then crossed Grayling Creek and headed south, again leaving a scat and signs of feeding activity along the way. . By the time the tracking session had progressed this far, it was mid-afternoon and the hounds had a difficult time following the trail. Subsequent radio triangulation placed #15 along Cougar Creek, about 4 km south of our last reliable dog-assisted location. Consecutive daily radio fixes would have estimated that #15 had moved a minimum of 7.6 km, whereas dog tracking indicated that #15 travelled at least t 51 15.03 km. 52 DISCUSSION Activity Patterns Diel Patterns Grizzlies were primarily crepuscular and nocturnal in this as in most other studies of grizzly bear activity patterns. Craighead and Craighead (1965) reported that Yellowstone grizzlies were nocturnal. Pearson (1975) found that Canadian grizzlies were most active in the early morning, late afternoon to evening, and at night. Schleyer (1983) described nocturnal activity for four of his study bears while the fifth, an older male, was primarily diurnal. His annual activity peaks of 0630 and 2400 correspond closely to my observed peaks of 0530 and 2300. None of the grizzlies in my study could be regarded as "diurnal." The literature provides no consensus regarding the seasonal perturbations of bear diel activity patterns. Schleyer (1983) observed that Yellowstone grizzlies were less nocturnal and more diurnal in spring and fall than during summer. Sizemore (1980) reported that grizzlies were more active at night than day during spring/summer (March through July) but all bears were active at all hours of the day during summer/fall (August until den entrance). This study found that crepuscular activity was most pronounced and nocturnal activity lowest in spring, a pattern similar to that described by Garshelis and Pelton (1980) for Tennessee black bears. 53 Like the grizzlies in Schleyer's study (1983), grizzlies in my study were less nocturnal in spring and fall than during summer but, in contrast to his results, diurnal activity was also lower in these seasons. Although all of the bears in my study were crepuscular and/or nocturnal, there was considerable variation between the bears in the magnitude of activity during the four dial periods and during the three seasons. I believe that the disparity, as described above, for grizzly bear diel activity patterns from one study to another primarily reflects individual differences in the bears sampled and does not result merely from contrasting sampling schemes. These differences in activity patterns seem to arise from individual and regional variation in food habits and habitat use and secondarily from I differences in sex, age class, and reproductive status. It is probably legitimate to claim that grizzlies in Yellowstone are predominantly crepuscular and, to a lesser extent, nocturnal. However, due to the considerable degree of inter-bear variation, attempts to describe peak activity times by hour or diel period on a seasonal basis may not be particularly meaningful. Seasonal Activity Levels Garshelis and Felton (1980) found that Tennessee black bears were inactive most of the time during March, the first month after emergence. Thereafter, the activity level steadily increased to a peak in June, remained high through September, and then decreased until den entrance. They postulated that the low activity level in spring was related to high use of grasses and sedges which, although easily obtained, were energetically insufficient to maintain high levels of activity. Increased activity in summer was associated with breeding activity and foraging on berries and fruits— foods which were difficult to obtain but of high caloric value. The moderate activity level in fall was due to the necessary pre-denning weight gain and to patchier food distribution which required increased searching. Schleyer (1983) found that in Yellowstone, grizzly bear activity increased from March to an overall peak in May (although he cautioned that this peak may have resulted from a sampling bias). Seasonally, his bears were most active during the summer and least active in fall. Sizemore (1980) reported that grizzlies were more.active during the late summer/fall period than during the spring/early summer period. Grizzly bears in my study were most active in summer and about equally active in spring and fall. (Monthly comparisons with the above studies are not possible since spring and fall were each represented by a single month.) Low activity levels during spring were probably related to high use of ungulate carrion (Cole 1972; Craighead and Sumner 1982; Knight et al. 1984). Schleyer reported that bears were significantly less active while using a carcass than at other times,. As described above for Tennessee black bears, increased activity in summer is due in part to breeding activity ahd in part to time-intensive foraging on grand.noids and forbs (Mealey 1975; Knight et al. 1980; Knight et al. 1984). Reduced activity levels during autumn correspond to increased use of pine nuts (Mealey 1975; Knight et al. 1984; Kendall 1981) and, to a lesser extent, increased use of 54 55 ungulates. Both pine nuts and meat provide concentrated energy sources which, once located or captured, can be utilized for several days. This explanation is not entirely satisfactory, however, since one might expect that even when high energy diet items such as pine huts and ungulates are available, the requisite weight gain concomitant with the pre-hibernation period would necessitate prolonged feeding bouts and hence elevated activity levels right up until the onset of "pre-hibernation lethargy" (Craighead and Craighead 1972). Futhermore, since most qf the meat which appears in fall scats . ' is acquired by predation, some increases in activity associated with the search and kill might be expected. Environmental (Weather) Effects , There are a number of ways in which weather factors can affect bear activity levels. Unfavorable conditions may induce discomfort or stress which can be mitigated by remaining inactive or adjusting behavior. Moen (1973) described how ungulates react to the thermal environment by making physiological and behavioral adjustments to maintain an internal state at or near the optimal condition. With bears, the situation is complicated somewhat by the fact that bears’ internal miIeau varies according to four discrete physiological/metabolic phases (hibernation, "walking hibernation" or transition, normal activity, hyperphagia) as described by Nelson, et al. (1983). Much of the observed seasonal variation in response to environmental factors assumedIy relates to these phases. For example, Garsheliq and Pelton (1980) suggested that black bears were less 56 sensitive to temperature in the fall because their "preoccupation" with foraging (hyperphagia phase) suppressed temperature effects. Certain environmental conditions might influence the probability of foraging success, either by affecting the prey directly or by affecting the bears* ability to detect or capture the prey. Any factor which augments or interferes with a bear's sensory capabilities should affect activity in this fashion. Unfortunately, not a great deal is known about the relative reliance on the distance senses in grizzly bear foraging behavior. Bear auditory capabilities are reputed to be good although experimental evidence is lacking. Kuckuk (1937, cited in Pruitt and Burghardt 1977) reported that captive brown bears responded to auditory signals at a distance of 15 meters. Grizzlies preying on elk in the fall might rely partially on auditory cues to locate bugling bulls and their harems, but vision and olfaction are of much greater importance. Bacon (1973) and Bacon and Burghardt (1976a) concluded that these two senses were highly coordinated in bear foraging behavior. Recent studies have indicated that bear vision is considerably better than once believed. Although nearsighted, black bears are able to distinguish color hues and to discriminate between simple geometric forms (Burghardt 1975; Bacon and Burghardt 1976b). The presence of a well-developed tapetum lucidum indicates that bears* night vision is fairly keen (Bacon and Burghardt 1976b; Cloudsley-Thompson 1961). Garshelis and Pelton (1980) suggested that while feeding on berries during late summer, black bears rely on color vision to locate and 57 select berries. They felt that increased nocturnal activity in fall was related to feeding on acorns which, unlike berries, were large enough to be seen at night. In general, visual stimuli are subject to less variability under different environmental conditions than are olfactory stimuli. However, such factors as cloud cover and lunar phase which affect the degree of illumination might affect grizzlies' visual perceptions and, consequently, their activity patterns. This study found that grizzlies did tend to be least active when the sky was overcast, but Garshelis and Pelton (1980) reported that black bear activity levels were not related to cloud cover. Varying the degree of sun and shade on test stimuli did not interfere with black bears' ability to discriminate hues (Burghardt 1975; Bacon and Burghardt 1976b). Schleyer (1983) examined the effect of lunar phase bn grizzly activity and found that grizzlies were most active under intermediate light conditions. He noted that grizzly eyesight did not appear to confine them to activity under maximum light conditions, such as daytime or under a full moon. Olfaction is considered to be highly developed in the grizzly bear; however, most of the supporting evidence is purely anecdotal (Russell 1979; Murie 1981). Grizzlies often travel considerable distances to ungulate carrion and their movement seems to be directed by scent (Craighead and Mitchell 1982). Nuances of scent play a major role in the success of grizzly trapping operations (Schleyer, pers. comm.). Kuckuk (1937) observed that captive brown bears relied primarily on olfaction to locate hidden foods. 58 Wright (1982) described two distinct, but related processes by which police dogs use olfaction. These processes, tracking and searching, are analogous to the ways in which bears might employ olfaction while foraging. "Tracking" refers to following the route of a mobile prey by cueing on either the residual scent of the prey itself or on the scent of disturbed earth, crushed plants, etc., left by the prey. Bears may sometimes locate ungulates and other mammalian prey in this fashion. "Searching" refers to the process by which a bear detects and locates a stationary food item by following an airborne scent trail to its source. Both olfactory processes, tracking and searching, can be affected by ambient conditions. Wright (1982), relying heavily on earlier work by Budgett (1935), described how various environmental conditions might influence olfaction. Although much of his analysis pertained to tracking mobile prey, most of the effects he described should apply to search situations as well. Moisture, whether airborne or on the ground, can affect the strength and longevity of scent. Oily components of the scent accumulate on moisture droplets and thus expose a larger surface for evaporation. Thus, based on olfactory criteria alone, grizzlies would be expected to be more active than average in conditions of high relative humidity, light rain or moist ground following a rain. Conversely, heavy rain will tend to wash the scent off of exposed surfaces and will create a negative effect for olfaction. Schleyer's (1983) observations are consistent with the predicted effects of moisture on olfaction. He found that as relative humidity 59 increased, grizzly bear activity increased, and he attributed this finding to olfactory enhancement. Schleyer also reported that grizzlies were more active than average during rain and less active than average when there was no rain. Precipitaion exceeding 1.4 cm/day did inhibit activity in his study. Contrary to the predictions, in my study grizzly activity was depressed during rain and somewhat higher (no significance at p=.05) when the ground was dry. Garshelis and Pelton (1980) found that Tennessee black bears were active less than average during rain, but activity increased shortly after a rain. In these latter two studies, diminished activity during precipitaion may have resulted from bears' attempts to avoid physical discomfort. Garshelis and Pelton noted that the response to precipitation was temperature dependent. Rain at temperatures below 7 C depressed activity more than rain at higher temperatures. Similarly, in my study, bear activity during rain increased from a mean probablity of .46 at temperatures of 0-5 C to a high of .54 at temperatures of 10.1 C and higher. This relationship is still more striking if daytime observations are excluded (to correct for the low activity during the diurnal period). The mean activity during rain then becomes 0.40 at 0-5 C, 0.64 at 5.1-10 C, and 0.73 at 10.1 C and higher. Thus, the effect of precipitation and moisture on bear activity levels may relate to both physical discomfort and olfactory considerations. Thermal factors can also influence olfaction. Increased temperatures and/or direct sunlight hasten the evaporation rate of scents. However, minor increases in temperature can have a favorable 60 effect on olfaction by creating slight updrafts which make a scent more accessible for dispersion under light wind. The temperature response curves for the annual» summer, and fall periods were bell-shaped with activity peaks at 10-15 C (annual and fall) and 5-10 C (summer) and lesser activity at temperatures above and below the peak range. Decreased activity at higher temperatures may relate to time-of-day effects (highest temperatures occurred during midday when grizzlies were least likely to be active) and/or to the negative effect of high temperature on olfactory perception. Decreased activity at lower temperatures may represent a behavioral adjustment to minimize thermal stress. Garshelis and Pelton (1980) found similar responses to temperature in summer and fall, although their fall data indicated a much broader (20 C range) plateau of high activity at intermediate temperatures. Schleyer (1983) also found maximum activity at intermediate temperatures annually and in summer, but the relationship of activity to temperature in fall was less clear. In spring, both Garshelis and Pelton (1980) and Schleyer (1983) found that activity increased as temperature increased for most of the temperature range. This pattern is in marked contrast to my observation that bears were least active at intermediate temperatures and most active in the maximum and minimum temperature ranges during spring. This result would not seem to follow from either physiological or olfactory considerations. Thermal microgradients between the ground and the air can also affect scent. Budgett (1933) found that tracking was best when the ground temperature exceeded the air temperature by a few degrees. 61 This situation occurs naturally in the early evening as air temperatures drop rapidly, but ground temperatures remain somewhat higher. The converse is true at dawn when the air warms more rapidly than the ground. Support for olfactory regulation of diel activity patterns is ambiguous. Schleyer (1983) found that activity was highest at sunset annually and during spring and fall. In summer, activity at sunset was about equivalent to activity at sunrise. The activity peaks of Tennessee black bears occurred between 1600 and 2000 in all three seasons. Grizzlies in my study were most active at sunset annually and in spring and fall, but this difference was significant only in fall; in summer they were more active at sunrise than at sunset. Hence, it appears that factors other than just olfaction are instrumental in determining grizzly diel activity patterns. Olfaction should be good in fog for two reasons; I. airborne moisture tends to capture scent, and, 2. fog typically occurs when cold air moves over warmer moist ground which, as described above, is favorable for olfaction. Data from this study were too scant to evaluate the effects of fog on bear activity; however, Schleyer (1983) found that grizzlies were active well above average when there was fog. The effect of wind on olfaction depends on its speed. Strong winds can be less favorable for olfactory searching than light winds because they move more air past the source and thereby dilute the scent. Light winds dilute scent less, but take longer to transport the scent a given distance (Wright 1982). Thus olfactory perception 62 should be best in light to moderate winds. Contrary to this hypothesis, coyotes are able to locate prey at greater distances in strong (40 km/hr) winds than in light (10 km/hr) winds (Bekoff and Wells 1980). Perhaps the scent dilution factor only becomes important at very long distances. Schleyer (1983) collected a small amount of wind data and reported that wind speed did not explain a significant portion of the total variation in grizzly activity levels and there was no significant difference in activity levels when there was no wind versus high (>50 km/hr) wind. In my study, grizzly activity generally increased as wind speed increased, but there were no significant differences in activity between any two wind speeds (p=.05). In summary, olfaction alone does not adequately explain all, or even most, of the variation in grizzly bear response to weather conditions. The observed responses probably result from a complex interaction of sensory (olfactory and visual) considerations with a number of other factors including endogenous rhythms (Cloudsley- Thompson 1961), physiological state, and security. Energetic Agendas of the Primary Study Bears The objective of this analysis was twofold: I. To develop a conceptual overview of how a given bear’s activity patterns, movements, food habits, and habitat use interrelate and to thereby gain some insight into the bear’s broad energetic "agenda." The concept of an energetic agenda is introduced here to refer to the manner in which a bear tailors its behavior and activity 63 patterns to suit its contemporaneous habitat and associated diet. ("Agenda" implies a coarse fitting of behavior to environment and is preferable to the related concept of energy budget as the latter generally pertains to a more rigorous caloric analysis.) individuals to determine how a "decision” in one arena (say a decision to feed primarily on graminoids) might influence the other variables (perhaps the overall activity level). Some of the hypotheses profferred herein are necessarily tenuous due to the small sample size, but I hope that this first order and highly subjective approach might provide some grist for further meadows along Antelope Creek and in the Washburn Range for most of the monitoring period. She tended to make short movements within a small range and returned frequently to particular ridges and meadows. Both her community site quality (CSQ) and relocation habitat richness scans (RHR) indicated that she occupied high-quality habitat in terms of vegetative structure. Her scats contained very high representations of graminoids and forbs and her scat value index (SV) was the highest of the bears sampled. Bear 59's mean activity level 2. To examine contrasting patterns exhibited by various inves" — — Bear 59 Bear 59 remained in the rich mid- to high-elevation grass/forb £o(ACIe)(Afal , was also higher than any other bear. 64 Interpretation: Mattson (1984a) has suggested that in mesic areas, such as Antelope Creek, where feeding opportunity is widespread and diverse, grizzlies may make a "default" adjustment in feeding activity toward site specific foods (i.e., those foods that characteristically occur "contagiously" in certain microsites). This default spatial adjustment occurs because, of the many food items available in mesic areas, most are abundant and well-distributed, whereas the site- specific foods, such as biscuitroot (Lomatiurn sp.) and yampa (Perideridea gairdneri), are available only in localized areas. If a bear adjusts its behavior so that site-specific foods are accessible, most other diet items are likely to be available in adjoining sites. Bear 59's behavior suggested just such a foraging pattern. She was often observed feeding along the fingerlike ridges which ran between the tributaries of Antelope Creek north of Mt. Washburn. She appeared to be selectively feeding in the rocky, sparsely vegetated microsites where Lomatium cous occurred in dense patches (note the high -I "contagiousness" value, A , in Table 10). She would comb through these sites by digging for tubers and flipping rocks until reaching the periphery of the rocky area. After several passes, she would travel to a different rocky microsite and commence feeding in a similar fashion. While travelling between these patchy islands of suitable habitat, she quickened her pace and appeared to be feeding indiscriminately on forbs and grasses in the interlying area. Although Bear 59 seemed to orient her feeding activity toward Lomatium, this spatial orientation actually resulted from a default 65 adjustment to one of the few food sources which was not ubiquitously available. Her relatively short mean daily movements (2.7 km) and small area per relocation size indicate that Bear 59 was able to satisfy her short-term energy demands without having to travel very far. Bear 59’s foraging activities were timed on four occasions . (for a total of 3 1/2 hours of observation time) as she fed on Lomatium nous north of Mt. Washburn. I found that about 65% of her time was spent actually digging for or feeding upon tubers (the two activities were indistinguishable) while the remainder of the time was devoted to searching for suitable plants. In contrast, while travelling between Lomatium microsites, she fed continuously on grasses and forbs. Thus, extraction of Lomatium tubers appeared to be costly in terms of both energy and time as evidenced by 59's very high mean activity level. The scat analysis indicated that only 13% of Bear 596s digestible scat volume consisted of Lomatium while 82.6% consisted of other forbs, grasses, and sedges. Thus, if visual observations were representative of Bear 59*s overall foraging patterns, then she seemed to be investing substantial time to acquire relatively low volumes of biscuitroot. However, the high digestibility of Lomatium, especially relative to the graminoids (Mealey 1975), may have resulted in an underestimation of its use through scat analysis alone. High use of Lomatium would not appear to be favored by a purely energetic analysis. The energetic efficiency value for Lomatium is only .27, one of the five lowest values (Table 10). It is apparent that either visual estimates substantially overestimated Bear 59's 66 orientation to Lomatium foraging or else her high activity level was related to another as yet unidentified factor. Bear 38 Synopsis: Bear 38 remained in the Gneiss Creek/Duck Creek area for most of the monitoring period. Her relocation habitat richness scans (RHR) indicated median habitat quality and her community site quality index (CSQ) was the highest of the five bears. Scat analysis indicated very high use of grasses and sedges (76% of digestible volume) and her scat value score was fairly high. Bear 38's minimum daily movements were of moderate length but her area per relocation was much larger than any other bear. Her mean activity level was low. Interpretation: Theoretically, CSQ and RUR should provide two estimates of the same parameter— the quality of a bear's occupied habitat. However, as RHR incorporates a broader area (0.5 km radius) encircling the bear's relocation points, the two estimates may differ. Since bear 38's CSQ I ■ score was the highest of the five bears and her RHR score was moderate, it would appear that she was selectively utilizing the richest microsites within a matrix of otherwise marginal habitat. Her mean daily movement length of 4.8 km and her large area per relocation (23.1 km^) suggest that #38's preferred foraging sites were scattered throughout her range and required a moderate amount of travelling between favorable sites. The community site analyses had a very high understory cover of grasses and sedges, and her scat analysis 67 reflected correspondingly high use of grand, noids (Voliime=76%; Frequency=100%). Bear 38's relatively low mean activity level of 0.55 likewise reflected the ease with which she was able to satiate herself once having arrived at these high-density foraging sites. Bear 50 Synopsis: Bear 50 remained in the Nez Perce Greek/Firehole River area for most of the monitoring period. Both her CSQ and RHR values were low (second from lowest in both cases); however, there was less disparity between her CSQ value and those of other bears than existed with her RHR value. Bear 50 had a high scat value score with a high representation of meat in the diet. Her mean daily movements were longer than any other bear, but her area per relocation was relatively small indicating considerable travelling within a small area. Bear 50's mean activity level was the lowest for the five bears. Interpretation: When considered in terms of traditional bear forage plants, #50’s habitat was certainly depauperate. Her home range (for the monitoring period only) was relatively homogenous (recall the low amount of edge and low habitat diversity values-Table 5) with a sparse understory (Table 4). I was often baffled by the monotony and apparent lack of feeding opportunity at many of Bear 50’s relocation sites. Although vegetal feeding opportunity was limited in #50's range, animal matter was plentiful. The upper Madison River and its main tributaries, the Firehole and the Gibbon, are an important elk (Cervus 68 elaphus nelson!) winter range. Estimates for the wintering Madison herd range from 600-850 elk (Craighead et al. 1972; Cole 1972; Aune 1981). A sizable herd of bison (Bison bison) also occupies the Nez Perce/Firehole area (Meagher 1973). This area receives substantial use by grizzlies attracted to carrion and winter-weakened elk and bisonjin I the spring and early summer. | A The high representation of meat in Bear 50's .1981 scats (Volume=48%; Frequency=60%) indicated that she was relying heavily on these ungulates and, to a lesser extent, rodents for sustenance rather than on vegetation. Schleyer (1983) noted that during his study #50 was efficient at finding carrion and preyed on elk both before and after the elk rutting season. He reported that in 1980 she killed two bull elk and ate a road-killed elk and a bison. Bear 50's long minimum movements suggest that her feeding activity was highly "directed" toward ungulates: i.e., she would travel considerable distances seeking elk or carrion and consuming garbage, shrubs (Vaccinium spp.), and grand.noids incidentally during her foraging bouts. The presence of elk and human refuse essentially subsidized an otherwise very marginal habitat. Mattson (1984b) has pointed out that subxeric-submesic areas (such as the Nez Perce/Firehole area) lack a diversity of feeding opportunity and are of low value to grizzlies except when associated with ungulate ranges. Bear 501s low mean activity level may also have reflected her penchant for preying on elk. A predatory bear would be expected to adhere to an energetic agenda quite unlike that of a grazing bear. A 69 bear which is adept and efficient at finding and killing elk would function in spurts of activity. While searching for vulnerable prey, the activity bouts should be relatively prolonged, but once the prey is located and dispatched and the carcass is being utilized, the periods of activity should be abbreviated. Bear 50's low diurnal activity (lower than any other bear) may be an additional predatory adaptation since one would expect ungulates to be least susceptible to predation during the day. Bear 15 spent most of the monitoring period in the West Yellowstone/Hebgen Lake area. Both his CSQ and RHR scores were lower than any other bear. Bear 15's scat quality score was the lowest of the five bears and his scats contained high volumes of garbage and meat. His mean daily movements were relatively short, and his area per relocation was slightly larger than all bears except #38. Bear 15's mean activity level was moderate (median). Interpretation: Bear 15's behavior is of particular interest for several reasons: 1. He was the only one of the study bears which appeared to be inextricably linked to unnatural (human) food sources. 2. Additional data on his habits were available from the dog-tracking sessions and from previous year's research. 3. Bear 15 was positively identified as the grizzly which attacked and killed a camper at Rainbow Point Campground on Hebgen Bear 15 Synopsis: 70 Lake in June 1983. Hence, data pertaining to his prior habits are of special relevance to Yellowstone bear management. Bear 15’s short-term home range for the monitoring period was in predominantly PICO/PUTR habitat. He also utilized the willow/sedge bottoms adjoining Hebgen Lake, Duck Creek, Cougar Creek, and the Madison River. All these areas offered very limited feeding opportunity as the associated IV values (Table 12) and Bear 151s poor CSQ and RHR scores indicate. Despite the intrinsic low productivity of this range, #15 rarely ventured into the adjoining higher elevation, mesic areas north of Hebgen Lake. Instead, he displayed a tenacious affinity for the lowlands and ricocheted from one spot to another, occasionally "camping" in the same locale for several days at a time. Examining Bear 151s movement patterns, one gains the impression that he was intimately familiar with the whole of his range. The routes followed by #15 during the dog-rtracking sessions support this impression. He seemed to jog from one road or path to another in a deliberate fashion; there did not appear to be much random, non- directed element to his foraging. His straight-line approach up a residential driveway directly to the dog food bowl and his subsequent route to the "grease pit" on Highway 191 (see p. 16) left little doubt that he had reconnoitered these circuits previously. Incidental observations on Bear 15's day beds also suggested that he was acquainted with particular sites. It was not unusual to find multiple day beds when following up on radio relocations. Out of the eight day bed sites which were examined for #15, all but two had more 71 than one bed for a mean of 14.8 beds per site. One relocation site had 62 day beds, some appearing to be several years old, within a 220 X 30 m area. These multiple bed sites were characteristically in a dense copse of trees (usually lodgepole) which afforded more cover than adjacent areas. Thus, it appeared that Bear 15 knew the whereabouts of these exceptional day bed sites and returned to them habitually. It is unlikely that Bear 15 would have been able .to reach an energetic balance in this habitat, had he not subsidized his diet with human refuse and predation on ungulates. Garbage appeared in 67% of his scats and accounted for 46% of the volume. As noted earlier, another six of Bear 15's scats contained such high volumes of non- digestible trash that they were not collected. Bear 15's thorough familiarity with his range enabled him to efficiently exploit vulnerable sources of unnatural foods, from dog-food to dumpsters, and he was a perpetual nuisance bear in the Hebgen Lake area (Etzwiler pers. comm.). After monitoring of #15 was discontinued in fall 1982, he was trapped and relocated twice after incidents at a private campground and resort. Much of Bear 15's aberrant behavior appeared to be adapted for feeding on human-related foods. His low activity level for the daytime and sunset time periods and his high nocturnal activity were well-suited for avoiding detection during the times of highest human activity. His affinity for certain dense copses of trees for day bed sites was also advantageous for a bear habitually using habitat closely associated with man. As noted in the discussion on dog­ 72 tracking. Bear 15's mean consecutive daily movements of 4.1 km may be a major underestimate of the actual distance travelled, as his forays included deliberate visits to favored food sources along a roughly circuitous route. Bear 15 also had a considerable amount of meat in his scats (Volume=27%). Schleyer (1983) reported that #15 killed an elk and used three elk carcasses in spring 1980. Knight (pers. comm., in Schleyer 1983) also observed that #15 killed elk in previous years. Thus, it appeared that #15 opportunistically preyed on elk when they were vulnerable and shifted to high use of garbage when elk were unavailable. Bear 76 Synopsis: Bear 76 occupied the rich meadows in the Blacktail Plateau and western part of the Washburn Range during the 1982 monitoring period. Her CSQ and RHR scores were both very high. Her mean daily movements were relatively long and her area per relocation was moderate (median). She had a high mean activity level. The scat sample for #76 was not adequate to reliably indicate food habits and will not be discussed herein. Interpretation: Bear 76's high activity level and long daily movements did not seem consistent with the high quality of her range. One would intuitively expect that in superior habitat where most energetic demands could be achieved relatively easily within a small area, bears I 73 would have abbreviated periods of activity and short daily movements. The explanation for #76 *s enigmatic activity patterns may relate to her age (2 year old) and inexperience. Bear 76 was trapped along with her mother near Gardiner, Montana as a yearling in September of 1981 and relocated to the Blackball Plateau. She remained with the sow until the latter died in mid- October . She then denned alone near the northern Park boundary and remained solitary for all of the 1982 field season. In contrast, most Yellowstone grizzly cubs (59-64%) den with the sow as yearlings and are not weaned until the onset of the following breeding season when they are about 2 1/2 years old (Craighead and Mitchell 1982; Craighead et al. 1974). Those bears, such as #76, which separate from the sow as yearlings would seemingly be at a disadvantage without the additional year of maternal tutelage and consequently be less efficient at locating and securing desirable food items. This observation is supported by the fact that early weaners tend to weigh less the following year than bears weaned at 2 1/2 (Craighead 1972— panel discussion). Age specific survival rates (P^) are also notably low for two-year old grizzlies regardless of the age when weaned (Knight and Eberhardt 1985; Craighead et al. 1974; National Academy of Science 1974). Hornocker (1962) reported that weaned yearlings held a low position in the . social hierarchy and tended to be quite timid and apprehensive of other bears whereas yearlings still with their sow shared her social rank. It is reasonable to assume that this low status would be manifest the following year as well. 74 There are additional ecological considerations which may have played a role in #76*s behavior patterns. Baker (1982) suggests that among those animal species in which the adults have a cerebral sense of location (i.e., a sense of spatial orientation) many of the young undergo a period of "exploratory migration." This phase involves a series of movements along unfamiliar routes to unfamiliar destinations during which time the animal assesses the relative suitabilities of the encountered habitat. Baker contrasts these exploratory migrations with the more traditional concept of "dispersal” which, unlike exploratory migration, includes no systematic appraisal of available habitats. If young bears simply "dispersed" from parental territories following dissolution of the sow:cub bond and settled in the first available habitat, one would not expect a prolonged period of enhanced activity accompanied by long movements. If, however, young bears do engage in a phase of vigorous habitat assessment, such behavior would be the norm. Interference by resident adult bears might also contribute to high activity levels and frequent movements (USFWS 1982). Several authors have found that even after an animal has completed the exploratory phase of its habitat evaluation it may continue to visit suboptimal habitat patches to confirm and/or update its relative suitability rankings (Krebs and Cowie 1976). Then, if conditions should change, the animal can immediately expand its use of the prior sub-optimal habitats (Smith and Sweatman 1974; Oster and Heinrich 1976). Baker (1982) terms this behavior "revisiting for 7 5 reassessment" (RFR). Thus, increased activity in young bears may result from either exploratory migrations, RFR, or both. Bear 76's community site analyses indicated high densities of grasses, sedges, and forbs. Were #76 feeding primarily on graminoids, food items with a fairly high energetic efficiency value (Table 10) but which must be ingested in large volumes to meet nutritive requirements, high activity levels would be expected. This did not, however, appear to be true for Bear #38 who consumed large volumes of grasses and sedges but maintained a low mean activity level. Regrettably, the lack of scat data for Bear 76 precludes any substantive conclusions regarding her diet. In summary, then. Bear 76's activity and movement patterns may have resulted from any of several factors including inexperience as an early-weaned two year old, low social status and interference from adult bears, high stress factors in the two-year-old age class (as evident in the low survival rate of two year olds), exploratory and RFR movements, and, finally, dietary considerations. Grizzly Bear Foraging Strategies One of the aims of this study was to gather data to apply the theories of optimal foraging to Yellowstone grizzlies. However, the logistical constraints inherent in trying to gather contemporaneous food habit, habitat use, and activity pattern data with a single field crew were formidable. Thus, the sample size for several variables was inadequate to address the questions of optimal foraging. In addition, there is a lack of baseline physiological data specific to bears. 7b The wide disparity in the habitat attributes and diets of individual bears, as apparent in this study, suggests that a quest for a single optimal foraging strategy applicable to all bears may not be meaningful. A more productive venture, then, may be to examine the individual strategies pursued by bears in contrasting habitats and to explore the implications of these distinctions upon bear management. Although a quantitative analysis of grizzly optimal foraging strategies was not attempted, some qualitative conclusions may be tendered. The optimal foraging literature describes three forms of optimizations which an animal can pursue (Ellis et al. 1976): 1. Time minimizers include those animals whose fitness is maximized by minimizing the amount of time spent feeding to satisfy a given energy requirement (Schoener 1971). This strategy is characteristic of animals with a fixed reproductive output and of animals which must minimize foraging time to allow time for other activities, such as mate selection or predator avoidance. 2. Energy maximizers improve their fitness when net energy for a given foraging time is maximized. Animals whose seasonal reproductive output is a variable function of body size or rate of energy acquisition are likely to be energy maximizers (Schoener 1971). 3. Nutrient optimizers include those consumers which choose their diet according to both energy and nutrient content (Emlen 1973). Biochemical composition and phenological stage are important criteria for diet item selection (Ellis et al. 1976). Characteristics of all three optimization strategies are evident in Yellowstone grizzlies. Males may temporarily become time /minimizers during the breeding season. Feeding activities are subordinate to mating as males become preoccupied with locating and courting females in estrus and, in areas of high bear density, defending her against other males (Hornocker 1962). Body weights reach an annual low due to the low energy intake and high energy expenditure during the breeding season (Craighead and Sumner 1982). Bears which occupy ranges near human habitation may also partially become time minimizers as they restrict their foraging to certain times of the day to avoid detection. Bear 15's lower than average activity level during the daytime and sunset periods was consistent with this hypothesis, although his overall mean activity level was slightly above average. Other than these special cases, the concepts of energy and nutrient optimization are more applicable to bears. Prior food habit studies have concluded that grizzly bear selection of plant food was based on available energy value. In Yellowstone Park and other grizzly ecosystems east of the Continental Divide this energy is derived mainly from succulent herbaceous vegetation (Healey 1975; Healey 1979). In areas west of the Continental Divide, grasses and succulent herbs were important in spring and early summer, and, from mid-summer through fall, sugar content of huckleberries (Vaccinium spp.), buffaloberries (Shepherdia canadensis), and mountain ash (Sorbus spp.) was critical (Martinka 1972; Busby et al. 1977; Mealey 1979; Sizemore 1980). Starch from underground portions of springbeauty (Claytonia spp.), biscuitroot, and other tuberous species was temporally important in all ecosystems (Healey 1975; Busby et al. 77 78 1977; Schallenberger and Jonkel 1979). Ungulate meat (rich in protein) and whitebark pine nuts (rich in carbohydrates and fat) are important spring and fall foods as available (Mealey 1975, 1979; Husby et al. 1977; Kendall 1981)., Evidence for diet selection based on a need for particular biochemical components (i.e., nutrient selection or optimization) is lacking for the grizzly. Mattson (pers. comm.) has speculated that in Yellowstone, the energy available from certain foods such as Equisetum arvense and Trifolium repens does not appear to adequately account for the high degree of preference demonstrated for these items. Thus, it is possible that some as yet unidentified biochemical constituent is responsible for this selection. Ellis et al. (1976) point out that the three optimization strategies described above are not mutually exclusive alternatives, but they overlap to varying degrees depending on the consumer's feeding niche. Foraging bears must actually attend to all three factors concurrently such that the energy gain per unit foraging time is maximized while also achieving an appropiate nutritional balance. A predatory bear may expend considerable energy in the search and pursuit phases of foraging so that the energetic cost!benefit ratio is critical. But having once subdued its prey, the biochemical composition of meat is generally suitable to fulfill the bear's nutrient requirements along with its energy demands. A grazing bear, on the other hand, expends only a moderate amount of energy to pursue and capture any given item. Biochemical composition, however, may vary considerably between the many available species and phenological 79 stages, and the herbivorous bear must select, the most beneficial items for digestibility, nutrient content, and energy. Optimal foraging theory distinguishes between the strategic and tactical aspects of diet selection. Selection "strategies" pertain to the time minimization, energy maximization, or nutrient maximization options described above. Selection "tactics" refer to the particular manner by which a bear endeavors to accomplish this strategy. The f food habits studies described above appear to support an energy maximization strategy for grizzlies, with protein and/or carbohydrates being the principal forms of energy packaging. However, individual bears in this study differed considerably in the tactics employed to accomplish the energy maximization strategy. The major differences in diet between, for example. Bear 38 (76% grass/sedge) and Bear 50 (8.1% grass/sedge) may be attributed to a number of factors. The availability and distribution characteristics of food items within a bear’s range and the degree of acquired preference (learned affinities) for particular items are undoubtedly important. There are, however, limitations to the latitude which any bear has in its dietary habits and foraging tactics. The distinct annual cycle of weight gain and loss associated with hibernation (Kingsley et al. 1983) and the physiological constraints imposed by the bears* monogastric digestive system (Bunnell and Hamilton 1983) bracket the dietary variability. 80 Theoretical Considerations of Grizzly Predation on Ungulates Ungulate meat, primarily in the form of carrion, constitutes a major food source for Yellowstone grizzlies. Mealey (1975) proposed that competition among grizzlies for elk meat in the spring might have a regulatory effect on the Yellowstone population. Use of carrion and "walking dead" peaks in early spring, remains high through May and then becomes relatively low throughout the summer (Mealey 1975; Knight et al. 1984). Annual variation in use of carrion is governed directly by availability (Cole 1972; Knight et al. 1980; Craighead and Sumner 1982). Actual predation on ungulates is greatest in April and May as grizzlies prey upon malnourished animals before the herds disperse to summer ranges. Some bears also prey on newborn calves in late May and June (Cole 1972). Craighead and Sumner (1972) reported that individual bears apparently recalled the locations of calving areas from previous years. The degree of selectivity shown by vertebrate predators for their preferred prey (in this case, ungulates) is a function of prey density (Rolling 1965). The selectivity curve (i.e., the exclusivity of feeding on preferred prey) reaches an asymptotic maximum that depends on the ' density of the preferred prey and on the palatability of alternative prey. Emlen (1966) found that predators may select less profitable prey items over more preferred prey if the less profitable items (grasses, forbs, etc. in this analysis) become very abundant relative to the preferred items. In this context, the increased vitality and mobility of elk in late spring and summer is analogous to an effective decrease in the density of preferred prey. Concurrently, 81 spring green-up increases the relative abundance and palatability of alternative foods. Thus, foraging theory would predict, and field studies have recorded, a shift from reliance on use of ungulates to use of less profitable, but readily available forage species. Why then do some bears, such as Bear 50, belie this prediction and continue to prey on elk in summer and fall? A number of factors are involved. They may occupy ranges which although deficient vegetatively, include substantial numbers of summering ungulates, as did #50's range. In these areas, the opportunity for directed predation or chance encounter with weakened or diseased animals should be greater than where prey are more dispersed. Acquired skill is a major factor in determining a grizzly's predacious tendencies. Knight (pers. comm.) felt that certain bears were more adept at killing large mammals than others. Both #50 and #15 were inveterate predators which had displayed an enhanced capacity to kill ungulates in prior years (Schleyer 1983). One of the more interesting patterns to emerge from the food habit/habitat analysis was the apparent fidelity between vegetatively deficient habitat, predation, and use of garbage. Both #15 and #50 had low CSQ and RHR scores. They were also the only bears to rely on supplementary food items (meat or garbage) to any extent. In some I respects, these two food groups constitute very similar food sources. Both meat and garbage subsidize the available vegetal resources, both contain a wide variety of biochemical components, and both are energetically rich. The most obvious distinction between these two food groups is that meat, if obtained via predation, requires 82 considerable energy expenditure to obtain while garbage is essentially free (disregarding the possible consequences of detection). However, a single capture of meat can feed a bear for days (Cole 1972; Craighead and Sumner 1982) whereas garbage is generally less of a prize. Conclusions based on such a small sample size are tenuous, but it is tempting to speculate that #15 and #50 were somehow predisposed either psychologically or physiologically to a reliance oh energetically dense food items. The low volumes of graminoids in both bears mid-summer diets imply a "reluctance" to utilize this food group beyond what would be expected by low availability alone. It is also possible that increasing the volume of graminoids in the diet might accelerate the passage rate enough to preclude complete utilization of any meat present in the digestive tract at the same time. Thus it may be to a bear's energetic advantage to feed exclusively on meat when meat is available (Picton pers. comm.). Tracking Bears with Trained Bear Dogs Successful bear tracking using dogs depended on a number of factors. The freshness of the scent was especially important. Tracking was optimal very early in the morning. Scent faded rapidly as temperatures rose later in the day, and then it was often necessary to "work" the dogs until they relocated a good scent. This was particularly troublesome in xeric habitat, such as sterile lodgepole or rocky areas. 83 Differences between individual dogs were most apparent when the scent was weak. In traditional bear hunts, the hounds run loose in packs and those with less sensitive olfactory endowments merely follow the leader. Although all eight of the dogs we worked with were experienced, weII-seasoned bear dogs, only two or three were consistently able to follow a trail in marginal conditions. The importance of the dogmen cannot be overemphasized. Wright (1982) observed (with reference to police dogs): ...it is the dogmaster-dog combination that operates with this degree of effectiveness, and the effectiveness depends as much upon the dogmaster's understanding of the "whats" and "hows" of smells and smelling as upon the dog's ability to do the actual sniffing. Only in this way can the dog's special capabilities be applied to full advantage and . the information he receives via his nose be transmitted to his master. It was necessary for the dogmen to be especially cautious whenever the dogs had obvious trouble following the trail. The possibility of intersecting and following the trail of a non-study bear was a serious concern. Equipping the hounds with loud bells certainly helped, but these bells were audible for a long distance and their use may have led to unnecessary disturbance of the bears and disruption of their normal routine. Despite these difficulties, the use of bear dogs proved to be an exciting adjunct to traditional research techniques. The details of bear behavior revealed by accurately tracing an individual bear's path were valuable from both an ethological. and a management perspective. Trained bear dogs can provide much finer resolution in movement and habitat use studies than is available by other means. Movement data 84 from telemetry studies are often subject to topographic aberrations and researcher error. In addition, followups to radio fixes are frequently biased toward the most conspicuous feeding activities (torn-up logs, digs, etc.) while grazing and selective foraging are easily overlooked. Tracking with dogs can increase the precision in these types of studies and show which microhabitats and which food items are being exploited. Bear dogs could be used when an intensive trapping and telemetry program was infeasible. For example, when assessing the potential impact of a proposed development on a grizzly population, many bears might need to be collared to yield sufficient data specific to the area in question. Dog tracking has the potential to provide substantial data on localized areas in relatively short time. Similarly, bear dogs might be employed when the activities of particular bears are of special concern, as in cases of depredation. CONCLUSIONS The data imply, within the limits of small sample size, that there exists a substantial interdependence among habitat characteristics, food habits, and activity patterns. Contrasting energetic "agendas" (as overtly manifested in bear behavior patterns) are likely to be associated with particular foraging tactics and these tactical programs are at least partially predictable given the contemporaneous habitat parameters. A number of outstanding questions remain to be answered by further research. 1. Seasonal shifts in the food habit/activity pattern complex were not adequately examined herein. Temporal shifts in food habits are well documented (Mealey 1975; Knight et al. 1984), as are seasonal changes in activity patterns (Schleyer 1983; Garshelis and Pelton 1980). Additional research will be required to determine the interdependence of these contemporaneous shifts. 2. To what extent were the observed patterns the result of actual preference as opposed to availability? For example; did #50 use vegetatively poor habitat by default due to a lack of more favorable habitats nearby? Or if mid- to high-elevation mesic meadows with a rich grass/forb flora were available, would she preferentially utilize these meadows over the lodgepole stands she occupied during this study? 86 3. What additional factors might be implicated in determining the energetic agenda of individual bears? The interpretations provided herein are not the only plausible explanations for the observed behavior and, in some cases, they are not thoroughly satisfactory. For example, why would Bear 38 have fed primarily on graminoids and forbs and maintained a fairly low mean activity level while Bear 59 had a similar diet yet had the highest mean activity level of any bear? Was this an artifact of the sampling or a consequence of additional variables not adequately probed by this analysis? 4. To what extent would the differences in activity patterns and, especially, food habits and habitat use, even out over time? Were the contrasting patterns merely the result of spot-sampling and, in the long term, would all bears approach the observed population wide trends? Conversely, if some bears persistently adhere to food habit and activity programs quite unlike the overall trends, this fact may have broad implications for bear management, particularly with regard to habitat preservation. Grizzly bear habitat evaluations and protective measures designed to protect areas deemed valuable to the "average" bear may neglect the welfare of certain individuals which, through acquired skills, habituation, and adjustments in activity patterns and foraging tactics, have contrived an energetically viable lifestyle within marginal habitat. Recent population projection models for the Yellowstone grizzlies have suggested a continuing long term (30 year) decline with the margin between this decline and population stabilization possibly hinging on a reduction of adult female mortality by one or two bears per year (Knight and Eberhardt 1984 and 1985). Thus, it would seem that management strategies cannot afford to overlook the well being of any individual bears. 87 LITERATURE CITED 89 LITERATURE CITED Aune, K. A. 1981. Impacts of winter recreation on wildlife in a portion of Yellowstone National Park, Wyoming. M.S. Thesis, Montana State Univ., Bozeman. 110pp. Bacon, E. S. 1973. Investigation on perception and behavior of the American black bear (Ursus americanus). Ph.D. dissertation, Univ. of Tenn. Knoxville. 161 pp. _____________, and G. M. Burghardt. 1976a. Ingestive behaviors of the American black bear. Proc. Int. Conf. on Bear Res. and Manage. 3:13-25. _____________ , and ________________ . 1976b. Learning and color discrimination in the American black bear. Proc. Int. Conf. on Bear Res. and Manage. 3:27-36. Baker, R. R. 1982. Migration: paths through time and space. Hodder and Stoughton, London. 248 pp. Bekoff, M., and M. C. Wells. 1980. The social ecology of coyotes. Sci. Am. 242:130-148. Budgett, H. M. 1933. Hunting by scent. Eyre & Spottiswoode, London. Bunnell, F. L., and T. Hamilton. 1983. Forage digestibility and fitness in grizzly bears. Proc. Int. Conf. on Bear Res. and Manage. 5:179-185. Burghardt, G . M. 1975. Behavioral research on common animals in small zoos. Pages 103-133 in Research in zoos and aquariums. National Academy of Sciences, Washington, D. C. Cloudsley-Thompson, J. L. 1961. Rhythmic activity in animal physiology and behavior. Academic Press. New York. 236 pp. Cole, G. F. 1972. Grizzly bear-elk relationships in Yellowstone National Park. J. Wildl. Manage. 36:556-561. Craighead, F. C . Jr., and Craighead, J. J. 1965. Tracking grizzly bears. Bioscience 15:88-92. __________ /____________, and '______________ . 1972. Grizzly bear prehibernation and denning activities as determined by radiotracking. Wildl. Monogr. 32. 35 pp. 90 Craighead, J. J. 1972. Panel discussion on: Ecology, population characteristics, movements, and natural history of bears. Proc. Int. Conf. on Bear Res. and Manage. 2:81-83. _________________, G. Atwell, and B. W . O'Gara. 1972. Elk migrations in and near Yellowstone National Park. Wildl. Monogr. 29. 48 pp. _________________, and J. A. Mitchell. 1982. Grizzly bear (Ursus arctos). Pages 515-556 in J. A. Chapman and G. A. Feldhamer eds. Wild mammals of North America: biology management, economics. Hopkins, Baltimore, MD. _________________ , and J. S. Sumner. 1982. Evaluation of grizzly bear food plants, food categories, and habitat. Pages 44-84 in J. J. Craighead, J. S. Sumner and G. B. Scaggs. A definitve system for analysis of grizzly bear habitat and other wilderness resources utilizing Landsat multispectral imagery and computer technology. Wildlife-Wildlands Inst. Monogr. No. I. Univ. of Montana, Missoula. _________________ , J. R. Varney, and F. C. Craighead, Jr. 1974. A population analysis of the Yellowstone grizzly bears. University of Montana Forest and Conservation Experiment Station Bull. 40. 20 pp. Despain, D. G. 1973. Major vegetation zones of Yellowstone National Park. Yellowstone Nat. Park Info. Paper 10. 4 pp. ______________ . 1984. Habitat types of Yellowstone National Park. NPS report, Yellowstone Research Off., Mammoth, Yellowstone National Park. 8 pp. Eaton, G . P., R. L. Christiansen, H. M. Iyer, A. M. Pitt, D. R. Mabey, H. R. Blank, Jr., I. Zietz and M. E . Gettings. 1975. Magma beneath Yellowstone Park. Science 188: 787-796. Ellis, J. E., J. A. Wiens, C. F. Rodell, and J. C. Anway. 1976. A conceptual model of diet selection as an ecosystem process. J . Theor. Biol. 60:93-108. Emlen, J. M. 1966. The role of time and energy in food preference. Am. Nat, 100:611-617. _____________. 1973. Ecology: An evolutionary approach. Addison- Wesley, Reading, MA. 493 pp. Garshelis, D . L. and M. R. Pelton. 1980. Activity of black bears in the Great Smoky Mountains National Park. J. Mammal. 61:8-19. 91 Graham, D. C. 1978. Grizzly bear distribution, use of habitats, food habits, and habitat characterization in Pelican and Hayden valleys, Yellowstone National Park. M. S. Thesis, Montana State Univ., Bozeman. 88 pp. Hitchcock, C. L and A. Cronquist. 1973. Flora of the Pacific Northwest. Univ. of Washington Press, Seattle WA. 730 pp. Holling, C. S. 1965. The functional response of predators to prey density and its role in mimicry and population regulation. Mem. Entomol. Soc. Can. 45:1-60. Hornocker, M. G. 1962. Population characteristics and social and reproductive behavior of the grizzly bear in Yellowstone National Park. M. S. Thesis. Montana State Univ., Bozeman. 158 pp. Husby, P. O., S. Mealey, and C. Jonkel. 1977. Food habits of grizzly bears in northwestern Montana, 1975-1976. Border Grizzly Project, School of Forestry, Univ. of Montana, Missoula. Jennrich, R. I., and F . B. Turner. 1969. Measurement of noncircular home range. J. Theoro. Biol. 22:227-237. Keefer, W . R. 1972. The geologic story of Yellowstone National Park. U . S . Geol. Surv. Bull. 1347. 92 pp. Kendall, K. 1981. Bear use of pine nuts. M. S. Thesis, Montana State Univ., Bozeman. 27 pp. Kingsley, M., J. Nagy, and R. Russell. 1983. Patterns of weight gain and loss for grizzly bears in northern Canada. Proc. Int. Conf. on Bear Res• and Manage. 5:174-178. Knight, R., and L. Eberhardt. 1984. Projected future abundance of the Yellowstone grizzly bear. J . Wildl. Manage. 48:1434-1438. ___________, and _____________ . 1985. Population dynamics of the Yellowstone grizzly bear. Ecol. Monogr. (in press). ___________, B. Blanchard, K. Kendall, and L. Oldenburg, eds., 1980. Yellowstone grizzly bear investigations. Report of the Interagency Study Team, 1978-79. USDI-NPS. 91 pp. __________ , D . Mattson, and B. Blanchard. 1984. Movements and habitat use of the Yellowstone grizzly bear. Interagency Grizzly Bear Study, Bozeman, MT. 177 pp. Krebs, J. R., and R. J. Cowie. 1976. Foraging strategies in birds. Ardea 64:98-116. 9 2 Kuckuk, E. 1937. Tlerpsychologie beobachtungen an zwei jungen braunbaren. Z. Vergl. Physiol. 24:14-41. Lindzey, F. G., and E.^C. Meslow. 1976. Winter dormancy in black bears in soutnwestern Washington. J. Wildl. Manage. 40:408-415. Martinka, C. J. 1972. Habitat relationships of grizzly bears in Glacier National Park. National Park Service Prog. Rep. 19pp. Mattson, D. 1984a. Unpublished report on factors influencing grizzly bear habitat selection. Interagency Grizzly Bear Study, Bozeman, MT. 7 pp. __________ . 1984b. Preliminary ideas on habitat analysis. Interagency Grizzly Bear Study, Bozeman, MT. 4 pp. Meagher, M. 1973. The bison of Yellowstone National Park. National Park Service Scientific Monograph Series No. I. U.S. Govt. Printing Office, Washington, D.C. 161 pp. Mealey, S. 1975. The natural food habits of free-ranging grizzly bears in Yellowstone National Park, 1973-74. M. S. Thesis, Montana State Univ., Bozeman. 158 pp. _________ _. 1979. Method for determining grizzly bear habitat quality and estimating consequences of impacts on grizzly habitat quality. Pages 79-130 in USFS and NPS. Guidelines for management involving grizzly bears in the greater Yellowstone area. Moen, A. 1973. Wildlife Ecology: an Analytical Approach. W. H. Freeman & Co., San Francisco, CA. 458 pp. Mueggler, W. F. and W. L. Stewart. 1980. Grassland and shrubland habitat types of western Montana. USDA Forest Service Gen. Tech. Rep, INT-66. 154 pp. Murie, A. 1981. The grizzlies of Mount McKinley. National Park Service Scientific Monograph Series No. 14. U. S. Govt. Printing Office, Washington, D. C. 251 pp. National Academy of Sciences. 1974. Report of committee on the Yellowstone grizzlies. Natl. Acad. Sci., Washington, D. G. 61 pp. National Oceanic and Atmospheric Administration. 1980-1982. Climatological data: annual summary, Wyoming. U.S. Dept, of Commerce, Natl. Climatic Center, Asheville, N.C. 93 Nelson, R. A., G. E . Folk, E. W. Pfeiffer, J. J. Craighead, C. J. Jonkel, and D. L. Steiger. 1983. Behavior, biochemistry, and hibernation in black, grizzly, and polar bears. Int. Conf. on Bear Res. and Manage. 5:284-290. Oster, G., and B. Heinrich. 1976. Why do bumblebees major?: a mathematical model. Ecol. Monogr. 46:129-133. Partridge, L. 1978. Habitat selection. Pages 351-376 in Krebs, C. U. and Davies, N . D. eds. Behavioral Ecology: an evolutionary approach. Blackwell, London. Pearson, A. M. 1975. The northern interior grizzly bear (Ursus arctos L.) Canadian Wildl. Ser.. Rep. Series 34, Ottawa. 86 pp. Pfister, R., B. Kovalchik, S. Arno and R. Presby. 1977. Forest habitat types of Montana. USDA Forest Service Gen. Tech. Rep. INT-34. 174 pp. Pruitt, C. H., and G. M. Burghardt. 1977. Communication in terrestrial carnivores: Mustelidae, Procyonidae, and Ursidae. Pages 767-793 in T. A. Sebeok, ed. How animals communicate. Indiana Univ. Press, Bloomington. Russel, A. 1979. Grizzly country. Borzoi, New York. 302 pp. Schallenberger, A., and C . Jonkel. 1979. Rocky Mountain East Front grizzly studies, 1978. Annual Report of Border Grizzly Project, School of Forestry, University of Montana, Missoula. Schleyer, B. 1983. Activity patterns of grizzly bears in the Yellowstone ecosystem and their reproductive behavior, predation and the use of carrion. M. S. Thesis, Montana State Univ., Bozeman. 130 pp. Schoener, T. W. 1971. Theory of feeding strategies. Ann. Rev. Ecol. Syst. 11:369-404. Sizemore, D . L. 1980. Foraging strategies of the grizzly bear as related to its ecological energetics. M. S. Thesis. Univ. of Montana, Missoula. 67 pp. Smith, J. N . M. and H. P. A. Sweatman. 1974. Food searching behavior of titmice in patchy environments. Ecology 55:1216-1232. Soil and Conservation Service. 1983. Summary of snow survey measurements in Colorado, Montana, and South Dakota, 1919-1982. SCS, Casper, Wyoming. 94 Steele, R., S. V. Cooper, D. M. Ondov, and R. D. Pfister. 1979. Forest habitat types of eastern Idaho— western Wyoming. USDA For. Ser. Intermountain For. and Range Expt. Sta., Ogden, UT. 182 pp. U . S. Fish & Wildlife Service. 1982. The grizzly bear recovery plan. Prep, in coop, with MT Dept of Fish, Wildlife and Parks, Don L. Brown: recovery plan leader. Coop, agreement no. 14-16-0006- 80-923. 195 pp- U . S. Forest Service and National Park Service. 1979. Guidelines for management involving grizzly bears in the greater Yellowstone area. 136 pp. Wright, R. H. 1982. The sense of smell. CRC Press, Boca Raton, FL. 236 pp. 95 APPENDICES 96 APPENDIX A COMMUNITY SITE ANALYSIS FIELD FORM 97 G R I Z Z L Y B E A R S T U D Y C O M M U N I T Y S I T E A N A L Y S I S F e e d S i t e No. ______________ D a t e ______________ O b s e r v e r s ______________________ F o r e s t ___ U T M ____________________________D r a i n a g e _______________________________________________ B e a r No. F l i g h t D a t e ________________ A e r i a l P h o t o No. __________________________________ E l e v a t i o n A s p e c t _______________________ ° S l o p e __________T o p o g r a p h i c P o s i t i o n ______________________ A r e a P h y s i o g n o m y (if in tim b e r , a t t a c h p o i n t c r u i s e m e a s u r e m e n t s ) ______________ H a b i t a t T y p e _____________________________________ G r o u n d P h o t o No. _____________________________________ A v e . d i s t a n c e b e t w e e n t r e e s > 3 m t a l l _________________ D i s t a n c e to o p e n o r t i m b e r ________ P l o t s i z e _____________________ C o m m u n i t y s i z e _________________ T R E E S S u b t o t a l c o v e r > 3 m t a l l _______ S u b t o t a l c o v e r < 3 m call F e e d C o v e r Prom. S p e c i e s F e e d C o v e r P r o m . S p e c i e s T o t a l c o v e rS H R U B S S u b t o t a l g r a s s / s e d g eS u b t o t a l f o rbsT o t a l c o v e rH E R B S 98 F e e d S i t e No. A n t V i a l No. 50 g S a m p l e No. of a c t i v i t y : S c a t C a r c a s s S q u i r r e l c a c h e T r a c k G o p h e r d i g T o r n l o g H a i r G o p h e r c a c h e dig T u r n e d r o c k Bed R o o t d i g T o r n a n t h i l l C l a w S t r i p p e d b a r k G r a z i n g O t h e r U n k n o w n d i g M u s h r o o m s A g e o f a c t i v i t y : E x t e n t and s i z e of f e e d i n g site: D e t a i l e d a c t i v i t y in c o m m u n i t y : A d j a c e n t o r a s s o c i a t e d a c t i v i t i e s n o t in c o m m u n i t y : R e l a t i v e f o o d s o u r c e a b u n d a n c e (.PIAL c o n e s ; b e r r i e s ; r o o t foods; c a r c a s s e s ; etc . ) : APPENDIX B TABLES OF HABITAT PARAMETER VALUES AND COMMUNITY SITE ANALYSIS DATA 100 Table 10. Energetic efficiencies (EE), characteristic contagiousness (A ), and monthly preference values for the most important diet items of Yellowstone grizzlies as used in the community site and scat quality analyses. All values were adapted from Mattson (in prep.). Food Item EE A-I Preference Apr May Jun Jul Aug Values Sep Oct Ann Ungulates .88 .99 .91 .66 .52 .54 .42 .60 .54 ,85 Rodents .27 .82 .68 .56 .52 .41 .44 .50 .60 .43 Trout .42 .91 * .31 .33 .16 Pine nuts .65 .99 .97 .82 .39 .78 .78 .88 .83 Anthills .42 .41 .02 .11 .08 .23 .16 .19 .44 .24 Graminoids .73 .82 .53 .63 .55 .22 .35 .40 .40 .48 Lomatium spp. .27 .99 .07 .42 .22 .40 .35 .43 Perideridia gairdneri .15 .99 .08 .43 .55 .68 .55 .65 Cirsium spp. .15 .74 .42 .36 .40 .40 .35 .43 Mushrooms .92 .50 .14 Potamogeton spp. .15 .99 .41 Shepherdia canadensis 1.00 .91 .27 Claytonia lanceolata .96 .99 .28 .06 .38 .33 .33 1.00 Equisetum arvense .46 .91 .06 .48 .38 .36 .17 .48 .38 Vaccinium scoparium .81 .74 .35 .25 .32 .34 Vaccinium globulare .92 .91 .64 .33 .06 .62 Trifolium repens . 46 .82 .50 .23 .32 .25 .37 .13 .32 .29 Taraxacum spp. . 46 . 66 .28 .47 .30 .27 .23 .37 Polygonum spp. .69 .66 .05 .16 .20 .22 .33 .24 Fragaria spp. .73 .74 .01 .07 .13 .30 .23 .11 .16 Epilobium spp. . 85 .58 .36 .26 .23 .38 .34 .23 Unid. Forb .40 * Missing values result from 0.0 preference or insufficient sample. 101 Table 11. Monthly Food Values (FV) of the most important diet items of Yellowstone grizzlies as used in the community site analyses. All values adapted from Mattson (in prep.). Food Item Apr May 1 Jun Jul Aug Sep Oct Ann Ungulates .79 .57 .45 .47 •37 .52 .47 .74 Rodents .15 .12 .11 .09 .03 .11 .13 .09 Trout * .12 .06 Pine nuts .62 .53 .25 .50 .50 .57 .53 Anthills .00 .02 .01 .04 .03 .03 .08 .04 Graminoids .32 .38 .33 .27 .21 .24 .24 .29 -'Lomatium spp. .02 .11 .06 .11 .09 .11 Perideridia gairdneri .01 .06 .08 .10 .08 ,10 Cirsium spp. .05 .04 .04 .04 .04 .05 Mushrooms .06 Potamogeton spp. .06 -Shepherdia canadensis .25 Claytonia lanceolata .27 .06 .36 .31 .31 . .95 Equisetum arvense .02 .20 .16 .15 .15 .20 .16 Vaccinium scoparium .21 .15 .19 .20 Vaccinium globulare .54 .28 .05 .53 Trifolium repens .19 .09 .12 .09 .14 .05 .12 .11 Taraxacum spp. .08 .14 .09 .08 .07 .11 'Polygonum spp. .02 .07 .09 .10 .15 .11 Fragaria spp. .00 .04 .07 .16 .12 .06 .09 Epilobium spp. .18 .13 .11 .19 .17 .11 * Missing values result from 0.0 preference or inadequate sample. 102 Table 12. Unit area importance values (IVU's) used to score habitat types for habitat richness analysis. Only those habitat types which occurred in the computer habitat scans for the five primary study bears are included. Forest Habitat Types: Spring Summer Fall ABLA/VASC-VASC1 .003 .152 .307 ABLA/VASC-CARU .018 .084 • 909 ABLA/VASC-PIAL .000 .467 . 964 ABLA/CACA^ .106 .213 . 660 PIEN/EQAR2 .514 .253 .000 ABLA/THOC .004 .033, .125 ABLA/CAGE .009 .010 .018 ABLA/LIBO-VASC .014 .139 .000 ABLA/VAGL-VAGL .000 .257 .000 ABLA/CARU .000 .012 .000 PICO/CARO .000 .000 .000 PICO/PUTR .009 .038 .003 PSME/SYAL .001 .002 .000 PSME/CARU .003 .016 .018 P!AL/VASC .000 .229 .343 Non-forest Habitat Types: FEID/AGSP .000 .084 .000 FEID/AGCA .688 .251 .492 FEID/AGCA-GEVI .072 .208 .246 FEID/DECE .000 .162 .000 DECE/Carex spp. .036 .124 .026 ARTR/FEID .358 .167 .000 ARTR/FEID-GEVI 3 .495 .201 .444 Dry Artemisia spp. shrubland _ .178 .100 .000 Moist Artemisia spp./Potentilla shrublandJ .356 .195 .200 Sedge bogs,marsh fens,wet areas .172 .017 .155 Alpine tundra (high elev. rocky grassland) .133 .062 .000 ScirpuS spp./Carex spp. (hot spring veg.) .688 .608 .529 Salix spp./Carex spp. (at high elev.) .000 .089 .000 Salix spp./Carex spp. (at low elev.) .273 .082 .262 1 When names of three species are included in the habitat type name, the third species gives the habitat type phase. 2 Corresponds to Despain1s (1984) "Wet Forest" habitat types: ABLA/CACA for high elevations and PIEN/EQAR for low elevations. 3 These types were mapped in the Gallatin National Forest represent groupings of several habitat types. and 103 Table 13. Community site scores for Food Value (FV), Understory Cover (Cu), Understory Species Diversity (H ), and Community Site Quality (CSQ). See text for description of variables. Proportional values are given in parentheses. Site # Use Date FV C U H U CSQ 4204 4/29 1.38 (.70) 3 (.37) 2.04 (.55) 2.32 4209 5/13 0.84 (.43) 3 (.37) 2.67 (.72) 1.95 COCO 4210 5/21 1.58 (.80) . 4 (.50) 2.51 (.68) 2.78 U 4215 5/28 1.29 (.65) 8(1.00) 2.72 (.74) 3.04 cd(U 4216 6/15 1.58 (.80) 7 (.87) 2.37 (.64) 3.11 PO 4220 7/26 0.80 (.41) 6 (.75) 2.24 (.61) 2.18 4221 7/27 1.17 (.59) 7 (.87) 3.68(1.00) 3.05 4222 7/15 1.97(1.00) 7 (.87) 0.66 (.18) 3.05 4206 5/4 0.44 (.22) 4 (.50) 0.51 (.14) 1.08O LA 4211 5/31 1.41 (.72) 4 (.50) 3.04 (.83) 2.77 U 4212 6/1 0.72 (.36) 5 (.62) 3.05 (.83) 2.17 cd . 4229 6/30 1.70 (.86) 5 (.62) 3.26 (.89) 3.23 M 4230 7/6 1.56 (.79) 7 (.87) 1.70 (.46) 2.92 QJ PP 4233 8/12 0.80 (.41) 3 (.37) 2.31 (.63) 1.82 X MONTANA STATE UNIVERSITY LIBRARIES 762 1001 4204 9 MAIN N378 H25U cop. 2