THE LOWER CLARK FORK ELK STUDY FINAL REPORT 1985-1990 R. E. Henderson B. A. Sterling T. O. Lemke Montana Department of Fish, Wildlife and Parks Fall 1993 Montana State Library 3 0864 1005 5926 2 THE LOWER CLARK FORK ELK STUDY FINAL REPORT 1985-1990 The Social Structure and Seasonal Habitat Selection of a Northwest Montana Elk Population with an Analysis of Population Characteristics, Harvest Rates, and Survey Techniques . MONTANA DEPARTMENT OF FISH, WILDLIFE & PARKS Project Numbers 5123 and 5223 Fall 1993 Prepared by: ^ ^TE^Hei^e^son Wildlife Biologist B.A. St'erlinq/^ Wildlife Biologist T . 0 . Lemke Wildlife Biologist Approved by : Af- Qoyrric^ (L/icy^x^ H. j/ Cross R-1 Wildlife Manager ,E. Firebaugh i-2 Wildlife Manager D. Childress Wildlife Division Administrator Contributors: Montana. Department of Fish, Wildlife and Parks Lolo National Forest Rocky Mountain Elk Foundation Bonneville Power Administration ABSTRACT Between 19 8 4 and 19 8 9 17 6 elk were captured and fitted with radio collars and neckbands in Hunting Districts 123 and 200 in northwestern Montana. Elk wintering in the Lower Clark Fork Study- Area had a composite yearlong range of 613 mi . Nine elk herd units were defined and seasonal movements were documented. Herd units were generally exclusive. Many cow elk wintering in HD 123 summered in HD 2 00, but cow elk wintering in HD 2 00 summered primarily in HD 200 and secondarily in HD 202 , Bull elk (2+ years- old) generally had seasonal movement patterns disimilar to cows, calves and yearling bulls. Emigration was documented for both cow and bull elk. Cows seldom emigrated and did so after turning 3 years-old. Bulls were more likely at emigrate and did so at 2 years of age. Hunting related mortality accounted for nearly 8 0% of known deaths. Annual weighted hunting season mortality rates were 40% for antlered bulls and 16% for cows. Hairvest rates for both bulls and cows was higher for elk wintering in HD 123. Harvest rates were 54% for bulls > 2 years old and 20% for yearling bulls and statistically different, based on small sample sizes. Evidence indicated that cow populations were stable. Bulls may have been harvested at rates that would cause a decline in age structure. Estimates were made of elk population numbers using both Lincoln-Peterson Index and the elk sightability methodology and model developed in Idaho. Based on elk sightability surveys 14 06+ 387 (90% C.I.) elk wintered within the study area in 1989. It was also estimat:ed that during the summer and fall about 1700 elk, or 2.77 elk/mi resided within the composite yearlong range, including HD 200 and portions of HDs 123, 201 and 202. During the study 2162 hunters ^nually harvested 343 elk. Hunter density averaged 2.9 hunters/mi in HD 123 and 4.6/mi in HD 2 00. A slight upward trend in hunter numbers was noted. Yearling bulls with antler configurations of 2 points or less on both sides comprised slightly over 50% of the harvest. Hunters spent an estimated $1.37 million while hunting elk in the study area in 1989. Replicated surveys provided more reliable estimates of population trends and herd composition than traditional single flight surveys. Elk sightability methods provided population estimates for age and sex classes that were more reliable and cost efficient that L-P estimates. L-P estimates were generally more expensive to make because of the need to trap and mark elk. Observed calf/cow ratios ranged from 28-52 calves/100 cows, and elk herds in HD 12 3 were generally more productive than those in HD 200. Pregnancy rates were over 90% for cows at least 2 years-old. Observed post-season bull/cow ratios ranged from 9-19 bulls/100 cows in HD 123 and 12-31 bulls/100 cows in HD 200. Analysis of over 3800 radio locations with Geographic Information System (CIS) computer methods revealed that bull elk preferred habitats^ with zero road density and avoided habitats with >0.5 mi of road/mi during all seasons. Cow elk preferred habitats i with zero road density and avoided habitats with >1.5 mi of road/mi during fall archery and rifle seasons. The analysis indicated that elk would reclaim habitats where roads were effectively closed. Evidence from observed mortality rates, hunter harvest surveys, herd composition, and GIS elk use/road density analysis indicated that habitat security was not adequate for MDFWP to meet the objective of maintaining a diverse age structure of bulls as stated in the Montana Elk Plan. Cooperative projects completed during the study included the BPA power line elk monitoring and mitigation study, development of timber harvest guidelines for the currently roadless Mt. Bushnell area, plans for prescribed burning of 3900 acres of winter range, development of a model to be used to analyze security in timber sale planning, a Master's thesis on elk winter habitat use, a Senior thesis on physiological condition of white-tailed and mule deer, and forest travel and timber sale planning. Management recommendations include improved population monitoring efforts, consideration of alternative harvest regulations and increased cooperative efforts with land managers and the public to increase habitat security so that elk plan objectives can be met. ii ACKNOWLEDGEMENTS The Lower Clark Fork Elk Study benefitted from the efforts and cooperation of numerous individuals and agencies. U.S. Forest Service personnel provided support and input: M. Hillis, J, Deibert, J. Lyon, G, Leighton, D. Chambers, C. Spoon, R. Lindgren, G. Barce. Other Forest Service employees of the Thompson Falls/Plains and Superior Ranger Districts helped in numerous ways. Drs. L. Metzgar, C. L. Marcum and B. W. O'Gara of the University of Montana assisted with study design and enlisted student volunteers to help with elk trapping. Dr. Metzgar also provided assistance in statistical treatment of mortality rates. D. Bennett, R. Gipe, W. Stewart and D. Rosenkranz provided contract aircraft services. D. Getz, Department of Fish, Wildlife and Parks (MDFWP) , piloted MDFWP aircraft. R. Mace, MDFWP biologist with the South Fork Grizzly Bear Study provided Geographic Information System (GIS) analysis to determine elk response to road densities. D. Palmisciano of the MDFWP wildlife research laboratory helped with blood analysis. The following landowners allowed us to trap elk on their property and provided additional assistance: F. Cavill, J. Webster, T. Bartholomew, T. McKay, J. McCrea, H. Stobie, R. Stephens, J. Warnken, and M. Guthnick. J. Canfield and M. Thompson were especially helpful, contributing in many ways to the project, as contract biologists iii for the Bonneville Power Administration (BPA) Elk Monitoring and Mitigation Study. Field assistants, whose help in the trapping and handling of elk was invaluable, were: J, Henne, M. Hurley, M. Burcham, J. Brenner, T. J, Pratt, K. Solberg, E. Priest, S. Largent, G. Miller, and F. Mensik. G. Chambers, MDFWP Game Warden, was of considerable aid throughout the project. J. Unsworth and L. Kuck, Idaho Fish and Game biologists, provided explanation, training and insight into the use of the Elk Sightablity Model. Dr. E.G. Garton, University of Idaho, used our survey data to revise coefficients in the elk sightablity model, making it suitable for use with a Bell-47 helicopter. M.A. Lonner and M. Leritz of Media Works, Bozeman produce<^^ high quality graphics of radio locations using TELDAY programing. The Rocky Mountain Elk Foundation (RMEF) and BPA provided financial support for certain aspects of the study. In addition, RMEF has provided financial support for elk winter range enhancement projects that were generated as a result of this study. We thank MDFWP Regional Wildlife Managers H.J. Cross and J.E. Firebaugh for their continuous support, guidance and patience during all phases of this research project. TABLE OF CONTENTS Page ABSTRACT , , i ACKNOWLEDGEMENTS iii TABLE OF CONTENTS , . . . . v LIST OF TABLES ix LIST OF FIGURES xii LIST OF APPENDICES , XV INTRODUCTION l GOALS AND OBJECTIVES 3 STUDY AREA 4 METHODS 8 Trapping. 8 Radio Telemetry. 8 Harvest Rates and Other Mortality 9 Population Estimates and Sex/Age Classifications 10 Elk Harvest Surveys 13 Road Density and Elk Use Patterns 13 Blood Sampling 15 RESULTS 15 Trapping. 15 Elk Distribution and Movements 19 Elk Herd Units 19 Seasonal Movements - HD 123 3 2 Cow Movements 33 Bull Movements 38 V Seasonal Movements — HD 200..... / Cow Movements 44 Bull Movements 43 Road Densities and Elk Use Patterns. , 49 ElkMortality , ^ 5g Bull Mortality , 57 HD123 , .57 HD 200 ^58 Cow Mortality , gQ HD123.. 62 HD200. 62 Hunter Harvest Survey. 63 Antler Point Distribution. 65 Population Estimates 66< HD 123 66 HD 2 00 68 Elk Observability 7]_ Calf/Cow Ratios -j^ HD 123 71 HD 2 00 72 Bull/Cow Ratios 72 •HD 123 72 HD 2 00 73 Time/Cost Comparisons of L-P vs. Sightability Surveys 74 Blood Sampling 76 VI Disease ,.o ....... . .11 Investigations in Adjoining Hunting Districts 77 Seasonal Movements - HD 202 77 Sex/Age Ratios - HD 121, 201, 202.. 78 Coordination and Interagency Activities.......... 79 DISCUSSION. 81 Seasonal Movements and Elk Habitat 82 Herd Units 82 Winter/Spring Habitat 84 Summer/Fall Habitat. 86 Fall Security Habitat 87 Response of Elk to Road Densities 89 Emigration. 90 Herd Composition and Productivity 91 Harvest and Population Management 93 Cow Management 93 Bull Management 96 Habitat Security 99 Hunting Seasons.. 101 Game Damage Problems 102 Evaluation of Survey Techniques 104 Nonconsumptive Elk Values 107 MANAGEMENT RECOMMENDATIONS 107 Population Management 108 Habitat Management HO vii Public Relationships. LITERATURE CITED APPENDICES . . • LIST OF TABLES Table Pacre 1. Geographic description of trap sites for 1986-1989.. 18 2. Description of nine elk herd units in the Lower Clark Fork Study Area. 21 3. Miles of roads in the Lower Clark Fork Study Area... 51 4. Percent of study area occurring in various road density levels 51 5. Availability and use of road density classes by bull elk in Lower Clark Fork Study Area. Open road densities only 52 6. Availability and use of road density classes by cow elk in Lower Clark Fork Study Area. Open road densities only 53 7. Availability and use of road density classes by bull elk in Lower Clark Fork Study Area. All roads included 54 8. Availability and use of road density classes by cow elk in Lower Clark Fork Study Area. All roads included .' 55 9. Annual harvest of radio-tagged bulls in HD 123 58 10. Harvest rates of 3 age-classes of bulls radio-tagged in HD 12 3 59 11. Annual weighted harvest of bulls radio-tagged in HD 200 59 ix 12. Harvest rates of 3 age-classes of bulls radio-tagged in HD 200 ........... = • 59 13. Annual harvest rates of cows radio-tagged in HD 123 61 14. Harvest rates of cows radio-tagged in eastern and western portions of HD 123 61 15. Annual harvest rates of cows radio-tagged in HD 200. 61 16. Estimated elk harvest in HD 123 from statewide hunter harvest survey, 1986-1989... 64 17. Estimated elk harvest in HD 200 from statewide hunter harvest survey, 1986-1989 64 18. Antler point distribution for bull elk harvested in HD 123 • . . 65 19. Antler point distribution for bull elk harvested in HD 2 00 66 20. Lincoln-Peterson population estimates for elk wintering in HD 123 67 21. Elk population estimate in HD 12 3 using "ELKMONT" elk sightability model in 1989 67 22. Lincoln-Peterson population estimates for elk • wintering in HD 200 .... 69 23. Elk population estimate in HD 200 using "ELKMONT" elk sightability model in 1989 69 X 24. Elk classifications for HD 123 from direct observations (1986-1988) and sightability model estimates (1989) 72 25. Elk classifications for HD 200 from direct observations (1986-1988) and sightability model estimates (1989) 73 26. Comparison of flight time needed to complete helicopter surveys 75 27. Comparison of total operations costs for Lincoln- Peterson vs. "ELKMONT" sightability surveys in HD 123 and 200.. 75 28. Elk helicopter classifications in adjoining HD's.... 78 xi LIST OF FIGURES Figure Page 1. Map of Lower Clark Fork Study Area. » ................ , 5 2. Location of trapsites for 1986-1989 ..... o 16 3. Age distribution of captured cow and bull elk....... 17 4. Composite yearlong range of cow elk captured on winter ranges in HDs 123 and 200, 1986-1990 .......... 20 5. Location of nine major elk winter ranges in HDs 12 3 and 2 00 defined by radio-collared cows and calves between the dates Jan. 1-March 31 ..........22 6. Composite yearlong range of cow elk in the Prospect Creek herd unit, 1986-1990 23 7. Composite yearlong range of cow elk in the Wilkes Creek herd unit, 1986-1990 24 8. Composite yearlong range of cow elk in the Dry Creek herd unit, 1986-1990 25 9. Composite yearlong range of cow elk in the Clark Mtn. herd unit, 1986-1990 26 10. Composite yearlong range of cow elk in the Eddy Creek herd unit, 1986-1990 27 11. Composite yearlong range of cow elk in the Swamp Creek herd unit, 1986-1990 28 12 . Composite yearlong range of cow elk in the Miller Creek herd unit, 1986-1990.... 29 13. Composite yearlong range of cow elk in the Pat's Knob herd unit, 1986-1990 30 xii Tamarack herd unit, 1986-1990, ................... 31 14. Composite yearlong range of cow elk in the Boyd- Tamarack herd unit, 1986-1990, ................... 31 15. Migration corridors for elk crossing east end of C~C Di'/ide. ••"•• = =■...■•»<>. •••....o....,, 36 16. Migration corridors for elk crossing the west end of C-C Divide 37 17. Movements of migrating bull elk captured in HD 123. . . .39 18. Calving locations (May 15-June 10) for 82 adult cows in the Lower Clark Fork Study Area, 1986-1990 40 19. Summer range locations (June 30-Sept. 30) for cows in Lower Clark Fork Study Area, 1986-1990 .......... . 41 20. Early hunting season elk security areas used from Oct. 15-Nov. 5,1986-1989 42 21. Migration routes followed by elk in the HD 2 00 Boyd-Tamarack elk herd unit. 45 22. Movements of 4 migrating bulls (1387,1361,1214, and 1360) captured in HD 200 47 23. Mean monthly elevations of radio-collared elk in study area, 1985-1990 48 24. Seasonal preferences of bull elk for habitats •with zero miles road density 54 25. Seasonal preferences of cow elk for habitats with zero miles road density. . 55 26. Causes of mortality for 56 marked elk, expressed as percent of known mortality 56 xiii 27. Weighted annual harvest rates for marked bulls and cows in HD 123, HD 200 and HDs combined 57 28, Hunting season mortality rates for 3 age-classes of bulls in the study area, 60 29. Population estimates and 95% C.I.'s for wintering elk in HD 123 70 30, Population estimates and 9 5% C.I.'s for wintering elk in HD 200. 7n xiv LIST OF APPENDICES Appendix Page 1. Capture and known fate data from elk trapped in HD 123, 1986-1989 us 2 . Capture and known fate data from elk trapped in HD 200, 1986-1989 122 3 . Capture and known fate data from elk trapped in HD 200, 1984-1985 (from Thompson and Sterling 1986), 128 4. Capture and known fate data from elk trapped in HD 202, 1987-1988.... 129 5. Capture data from 88 white-tailed and mule deer marked in HD 123 , 1986-1989 130 6. Capture data from 51 white-tailed and mule deer marked in HD 200, 1986-1987 134 7. Comparison of harvest rates for pooled samples of radio-tagged bulls in HD 123 and HD 200 during 1984-1990 hunting seasons. . 136 8. Comparison of harvest rates for 3 age-classes of radio-tagged bulls in study area during 1984-1990 hunting seasons. 135 9. Comparison of hairvest rates for cows radio-tagged in HD 123 and HD 200 during 1984-1990 hunting seasons 137 10. Comparison of harvest rates for pooled samples of cows radio-tagged in permitted (EAST) and either-sex (WEST) portions of HD 123 during 1984-1990 hunting seasons. . 137 XV 11. Decision Notice and Finding of No Significant Impact, Mount Bushnell, USDA, Forest Service. 13 xvi INTRODUCTION The Lower Clark Fork Elk Study was funded for fiscal years 1986-1989 through the State of Montana Executive Planning Process (EPP) . EPP projects directed supplemental funds toward studies of high priority management concerns. The study was a cooperative effort involving wildlife biologists from Regions 1 (Kalispell Headquarters) and 2 (Missoula Headquarters) of the Montana Department of Fish, Wildlife and Parks (MDFWP) . Preliminary findings for the first 2 years of the study were " reported in the Lower Clark Fork Elk Study. Biennial Progress Report. 1985-1987 (Lemke and Henderson 198 7) . At the turn of the century few, if any, elk occupied Mineral, Sanders, and Lincoln counties of northwestern Montana (Koch 1941) . Much of the study area was heavily burned by catastrophic wildfires in 1910. Between 1912 and 19 64, elk from Yellowstone National Park were released on 10 occasions in and near the study area (Anonymous 1976). In the 80 years since 1912, elk populations in the area have increased dramatically in both size and geographic range. Currently, Hunting Districts (HD) 123 and 200 are popular elk hunting areas, offering a variety of hunting opportunities with a relatively good chance of harvesting mature bulls. As a result, the economic impact of elk hunting in the study area is considerable. It was estimated that in 198 9 elk hunters spent $1,377,589 to hunt elk in HDs 123 and 200 (Duf field 1988, Rob Brooks MDFWP, Bozeman pers. comm.). 1 Despite the area's popularity with elk hunters, other concerns in the area included game damage complaints and potential for timber harvests affecting habitat. Little was known about elk population numbers and trends, sex and age ratios, seasonal movements, habitat use, recruitment and harvest rates prior to this study. During the 1940 -s field investigators described winter distribution and forage use, sex/age ratios, and attempted population estimates for much of the study area, often focusing on forage utilization and mange problems in the Cherry Creek Game Preserve (Rognrud 1948, McDowell 1949). After 1950 MDFWP periodically examined browse utilization and condition transects, conducted sporadic ground age/sex elk classifications in HD 2 00, and began annual, nonstandardized helicopter elk classification surveys in HD 123 in 1976. Warner (1970) reported the effects of timber harvest on winter forage production in Tamarack Creek. In 1984, the first elk were trapped and radio-collared in HD 2 00 in the Deborgia area as part of a Bonneville Power Administration (BPA) mitigation study to monitor the impacts of a new 500-KV transmission line on elk (Hammond et al. 1985, Thompson and Sterling 1986, Thompson and Dickson 1987) . Monitoring of those elk indicated substantial movement across the western half of the C-C Divide between HD 123 and HD 2 00. This information further underlined the desirability for a better understanding of population parameters and movements to direct management of elk in the area. Between 1985 and 1988, the Lower Clark Fork Elk Study and the BPA Power Line Study coordinated efforts, sharing equipment, personnel, marked elk and data. Canfield (1988) reported the impacts of the newly constructed BPA powerline on elk habitat selection and hunter opportunities. GOALS AND OBJECTIVES The Lower Clark Fork Elk Study was a management-oriented project designed to provide data to achieve three major goals: Goals (1) maintain or improve elk hunting opportunities; (2) integrate the elk resource into Lolo National Forest management planning; and (3) address local elk dsunage problems on private lands. Specific objectives within the study plan to meet these goals were: Obi ectives (1) develop techniques that account for observability bias applicable for estimating elk numbers in the study area and in similar haOsitats in western Montana; (2) determine herd unit boundaries and monitor sex/age structures of those elk herd units; (3) S2uaple harvest rates on various sex and age segments of the population; 3 (4) esamine patterns of habitat use^ seasonal movements and tt.e timing of movements for various herd units? and (5) gather information on pregnancy rates physiological condition and occurrence of common diseases. A major focus of the study was to encourage public resource agencies, private groups and individuals to become involved with the study and to ultimately use the results to improve wildlife/land management efforts in the area. The Lolo National Forest (LNF) , University of Montana (UM) , Rocky Mountain Elk Foundation (RMEF) , BPA, landowners and local sportsmen have participated throughout the study (Lemke and Henderson 1987) . STUDY AREA The study area is in southwestern Sanders and northwestern Mineral Counties of western Montana (Fig. 1) along the Idaho state border. The towns of Saltese, St. Regis, Paradise, Plains and Thompson Falls are located along those boundaries. The area encompasses 612 square miles (mi^)of timbered and mountainous terrain in HD 123 (379 mi^, 62% of total study area) and HD 200 (233 mi^ 38%) . The topography is generally steep with elevations ranging from 2,400 feet (ft) at the mouth of Prospect Creek to 7,2 00 ft at the summit of Penrose Peak. The C-C Divide is a major geological formation, running southeast from the Montana-Idaho border to the 4 Figure 1 . j^^p Lo^^j. q^^^ ^^^^ 5 j^^y Clark Fork River near Paradise, The divide is the geographical boundary between MDFWP Regions 1 and 2 , separates the Thomos Falls/Plains Ranger District from the Superior Ranger District, and is the boundary between Sanders and Mineral Counties. In the winters of 1987 and 1988, the study area was expanded to include trapping sites in Marble and 2-Mile Creeks in HD 202, to determine possible interrelationships with populations previously sampled in nearby HD 200. The vegetation in the study area is representative of relatively moist and cool habitats that characterize elk ranges in northwestern Montana. Mean annual precipitation is 2 2.5 inches at Thompson Falls (U.S. Dept. of Commerce 1980). Forest growth is extensive and lush. Coniferous species present include lodgepole pine (Pinus contorta^ , ponderosa pine (Pinus oonderosa^ , Dougla f (Psuedotsuaa menziesiil , grand fir (Abies arandls^ , western rLarix occidentalism , western red cedar (Thuja plicatal , Engelmann spruce fPicea enaelmanniim , western white pine (Pinus monticola) , western hemlock (Tsuga heteroohylls m , and mountain hemlock (Tsuga mertensiana) , Wildfires, in the early 1900 's, strongly influenced vegetation in the study area. Extensive lodgepole pine stands are legacies of those fires. On warm, dry slopes, serai brush fields replaced coniferous cover following the fires. Over 80 percent of the land is administered by LNF. Corporate timber management is concentrated in the Miller and Combest Creek area south of Plains. Small private ownerships are located at 6 lower elevations along the St. Regis and Clark Fork Rivers. Major human activities include timber management and outdoor recreation. Small scale mining, livestock grazing, and hay production also take place. Elk hunting is the primary recreational activity throughout . 2 the area. From 1933 until 1950 a 60 mi portion of the study area, the Cherry Creek Game Preserve, was off-limits to elk hunting to foster earlier transplant efforts. The Montana Fish and Game Commission removed the preserve status in 1950, following reports of over use of winter forage (Rognmd 1950). Prior to the 1950's most hunting was restricted to antlered bulls in the rest of the study area. As elk populations grew, more liberal 5-week either- sex general seasons were instituted in the 1950 's. From the mid- 1950 's through the early 1960 's, antlered bull-only hunting preceded either-sex seasons, during parts of September and October. Elk hunting seasons have varied from 5-6 weeks in length since the inid-1960's. Season types, but not season length, have become more restrictive over the past 2 0 years. Five week either-sex general seasons were open to anyone with an elk license through the 1950 's and 1960*3. However, by the late-1970's antlerless elk hunting was limited for rifle hunters to just the first 8 days of the season. In 1978, in the area south of Plains, hunting was restricted to antlered bulls only to allow the elk population to increase. Since 1983, limited numbers of antlerless permits have been issued for that area. In 198 0 antlerless elk hunting in all of Region 2, including HD 200, was limited to hunters receiving permits through 7 a computerized drawing. In 198 6 the Fish and Game Coirmiission created the current boundaries of HD 12 3 so that hunter har/est^ data could be collected for analysis in this study o That boundary included small portions of what had been HDs 121 and 122. During the study period (1985-1990) , elk hunting seasons consisted of 6 weeks of either-sex archery hunting, starting the first Saturday in September, followed by 5 weeks of general (firearms) hunting for antlered bulls, beginning the third Sunday in October. Antlerless harvest in the newly , created HD 12 3 was accomplished through 2 season types: 8 days of general either-sex hunting in the western 2/3 (west of Eddy Cr.), and antlerless permits in the eastern 1/3. In HD 200 antlerless permits regulated the cow and calf harvest. Extra permits were issued for the Boyd Mountain area to address game damage complaints. METHODS Trapping Efforts were made to trap and mark elk from all major winter ranges in the study area. Trapping operations were conducted from January to April from 1986 to 1989. Elk trapping and handling procedures, using portable modified Clover traps and a corral trap, were described by Lemke and Henderson (1987) and Thompson et al. (1989) . Radio Telemetry We attempted to aerially relocate radio-tagged elk twice a month from late spring through November and less frequently from December through April. Aerial relocation methods followed descriptions in Lemke and Henderson (1987) and Canfield (1988) . With financial support from LNF, more intensive monitoring was conducted daily (weather permitting) for the periods immediately prior to and after the beginning of the general hunting seasons. Harvest Rates and Other Mortality When possible, the cause of death of marked animals were determined and tabulated. Investigations were made following non- movement of radio-tagged animals and reports by the public of dead animals marked with neckbands and eartags. We did not attempt to located other dead elk. Harvest rates were based on the verified and assumed fates of radio-collared elk alive in the study area prior to each of the fall hunting seasons. If a radio-tagged elk, known to be alive in September, was not relocated at some time during or after the hunting season, we assumed that it had been harvested by hunters. That assmption was made regardless of whether or not hunters reported the elk harvested. Elk known not to be in the study area,, or whose radios failed prior to the hunting season, were excluded.' When calculating hunting season mortality rates, we included those elk whose radios were recovered in the field and were determined to have died as a result of wounding during the hunting seasons. There was potential for over-estimating hairvest rates by counting radio failures as hunter harvests. Because of small sample sizes in each hunting season, we pooled the nxmber of radioed elk available and harvested for all 9 the hunting seasons from 1985--1990, Some elk were alive at the beginning of more than one hunting season and were included in the pooled samples. The weighted annual per capita hunting season mortality rates were then calculated from the pooled samples, Chi- square (X^) goodness-of-f it tests were applied to compare mortality rates observed in pooled samples of elk in several age, sex and geographic categories. Population Estimates and Sex/Age Classifications We employed two methods to estimate elk populations — -Lincoln- Peterson (L-P) estimates from 198 6 through 1988 and the newly developed elk sightability method in 1989. In April 198 6, 1987 and 1988 helicopter surveys were conducted to count and classify marked and unmarked elk. The helicopter was flown from 2 0-4 0 m.p.h at an altitude of 100-3 00 ft. Flights wertl^ made in early morning for 2-3 hours, until all "open" areas, generally southern and western aspects, on winter ranges in the two HDs were completely censused (Lemke and Henderson 1987) . The same experienced observers were used each year. A 2 -passenger Bell 47 was used in HD 123 and a 3-passenger Bell-47 in HD 200. All elk seen were recorded as cows, calves, yearling bulls, branched-antler bulls, unclassified bulls or unclassified elk based on morphological characteristics. Elk marked with radio-collars and neckbands were identified and recorded. A series of flights that completely surveyed winter ranges within a HD was one "replication." Three replications were made for each HD in April each year. 10 Population estimates for each HD v/ere derived from the 3 completed aerial mark-recapture surveys. Separate modified L-P estimates (N*,-) were calculated for each replication, N*,-=[(M,. + 1) (C,- + 1)/(R,- + 1)] - 1 (Chapman 1952) where for Sctmple i, M,- is the number of previously marked elk, C, is the total number of elk observed, and R- is the number of marked elk observed. When all replications were completed the final estimate for each HD was set equal to the average of the K individual estimates, denoted by N*. The standard error (SE) was calculated from the individual estimates. N*+2 SE represented an approximate 95 percent confidence interval for the true population size (N) . Calf/cow and bull/cow ratios obtained from the replicated samples were weighted for sample size. Samples were large enough that weighted mean and mean sex/ age ratios were identical. Best count ratios were calculated from the highest number of cows, calves and bulls observed during the 3 replicated surveys conducted each year. In February 1989, helicopter surveys were conducted with the 3-passenger Bell-47 in HDs 123 and 200, utilizing the methodology for elk sightability surveys developed in Idaho (Samuel et al . 1987, Unsworth et al . 1991). Within winter ranges of each HD, 4-6 mi^ subunits were deliniated on topographic maps and stratified according to expected number of elk per subunit. We based 11 stratifications on our experiences from surveying elk in the study area during the 3 previous years. We chose to survey a random sample of stratified subunits, rather than do complete-coverage surveys, to reduce helicopter time and costs. Each randomly selected subunit was surveyed by helicopter on 100-2 00 ft contours at 15 to 40 m.p.h., and at altitudes of 100-200 ft, according to protocol described in Unsworth et al. (1991) . Data collected during sightability surveys in 19 89 included each elk group size, sex and age composition of each group, % vegetative cover, activity, and % snow cover for each observation in a sampled subunit of the winter range. Data was then entered to ASCII files (Unsworth et al. 1991) . The computer model "ELK4" had already been developed in Idaho to estimate populations from the sightability surveys conducted with a 3-passenger Hiller-12E helicopter. Since the Hiller was unavailable to us, we used the 3-passenger Bell-47 helicopter in this study. We did not know how sightability from the Bell-47 compared to the model (ELK4) developed with the Hiller-12E helicopter. Therefore, in conjunction with our sightability surveys in 1989, we used telemetry equipment to locate radio-tagged elk within each sampled subunit. For both those radioed elk that had been observed during the survey and those that had not been seen, but which were located with the aid of radio telemetry after surveying the subunit, we recorded group size, age and sex composition, activity, vegetative cover and snow cover (Samuel et al. 1987) 12 Those and similar data collected on the Blackf oot-Clearwater Wildlife Management Area were forwarded to Dr. E.O. Garton at the University of Idaho- Dr. Garton determined the need for and made the necessary adjustments in the Hiller-based model to develop the new model "ELKMONT" for use with data collected from the Bell-47. "ELKMONT", like "ELK4", is a computer model developed by forward stepwise logistic regression of the probability of a group of elk being seen and was derived from linear multiple regression of factors influencing the sightability of elk (ie. group size, % canopy cover, % snow cover, and activity) (Unsworth et al . 1991). With the survey data collected in February 1989 and the new Bell 47 based model "ELKMONT", we estimated population sizes and sex/age compositions with 90% confidence intervals in HDs 123 and 200. Elk Harvest Surveys MDFWP conducted hunter harvest surveys annually, using statewide telephone interview methods (Cada 1987) . We examined annual harvest survey summaries for trends in harvest, hunter effort, and antler point distribution for elk harvested in HDs 12 3 and 2 00. Road Density and Elk Use Patterns Over 3,800 elk radio locations were analyzed using standard Geographical Information System (GIS) computer technology to determine elk use patterns relative to road densities. Computer generated maps were overlaid on detailed digitized road maps for the entire study area (Mace 1992) . The raster (cell) resolution 13 used for analysis was 3 0 meter « At this resolution, the C™C Divide study area was composed of 1,654,318 cells. ^ Roads in the study area were assigned to 1 of 4 classes: primary open road, secondary open road, tertiary open road, and seasonal or yearlong closed road. Two road density maps were generated, one with all roads digitized and another with only open roads present- Road densities were classified into 7 levels: 0=0- 0.5, 0.5-1.0, 1.0-1.5, 1,5-2.0, 2.0-2.5, 2.5-3.0, and >3 . 0 miles of road per mi^. Elk radio locations were analyzed with respect to the road density levels within which they were positioned. Female and male elk of all ages were analyzed separately using 5 seasonal categories with inclusive dates: winter (Dec.l-Feb.29) , Spring (Mar.l-May 30), Summer (June 1-Aug.31), Archery Season (Sept. 1^ Oct. 15), and General Hunting Season (Oct. l5-Nov. 3 0). Female and male locations, stratified by season, v/ere overlaid on the 2 GIS road density maps. The number of locations occurring in each road density class was determined and constituted "elk use." The amount of each road density class within the study area was determined and constituted "available" road density acreage. Simultaneous Bonferronni confidence limits were used to statistically determine if each road density level was preferred, avoided, or used in proportion to availability. Confidence limits were constructed at 95%. Blood Sampling We drew blood samples from trapped elk and forwarded them to the Montana Wildlife Research and Veterinary Research Laboratories in Bozeman for analysis. Pregnancy was determined from the presence of Protein B and age specific fates of pregnancy were calculated. Exposure to Brucellosis, Anapl£ismosis , bluetongue and Leptospirosis was determined from antibodies present in blood samples . RESULTS Trapping During the 1986-1939 trapping seasons, winters were mild with little snowfall. As a result, elk were widely dispersed, and trapping success depended on our ability to move Clover traps to areas occupied by elk. Trapsites and elk captures were well distributed on winter ranges in the study area with the exceptions of Cherry Creek (HD 12 3) and Donlan Flats (HD 2 00) (Table 1 and Fig. 2). From 1984 through 1989, 176 elk (142 cows and 34 bulls) were captured and marked in HDs 123 and 2 00. Among the cows, 84 (60%) were adults (>3.0 years), 21 (15%) were two-year-olds, 21 (15%) were yearlings, and 16 (11%) were calves. Of 34 bulls captured, 1 (3%) was an adult (>3.0 years), 2 (6%) were two-year-olds, 12 (35%) were yearlings, and 20 (56%) were calves (Fig. 3). In HD 123 66 elk (approximately 8% of the estimated population) were captured during the study (Appendix 1) , and 110 15 PERCSHT ellc (approximately 25% of the estimated population) was captured in HD 200 (Appendix 2) . Since the last progress report, 11 elk were captured in HD 123 in 1988, and an additional 5 elk were trapped in 1989. In HD 200 28 elk were captured in 1988, while no trapping was attempted in 1989. Included in the trapped sample Fig. 3. Age distribution of were 18 elk captured during the captured cow and bull elk, expressed as percent of total summers of 1984 and 1985 for the in each sex. BPA study (Appendix 3) . An additional 10 cows were captured and kradio-tagged in HD 202 during the study (Appendix 4) . 1/2 1 W2 Z 1/2 ACE i'iltS. ) >3 I COWS BULLS During elk trapping we also captured 54 white-tailed deer (Odocoileus virginianus) and 85 mule deer (Odocoileus hemionus) between 1986 and 1989. Eighty-eight deer were marked in HD 123 (Appendix 5) , and 51 were marked in HD 200 (Appendix 6) . Hurley (1987) reported results from blood sampling, DAPA fecal analysis, and physical measurements of many of those deer. Numerous other deer were captured incidental to elk trapping and released unmarked. Subsequent hunter harvests of eartagged deer were recorded in R-1 and R-2 annual reports. 17 Table 1. Geographic description of elka trapsites for 1936-89. Trap site locations appear by number in Figure 2. M Trap Site # Trap Location HD Herd Unit UTM Lat Long 1 Swamp Cr. 123 Swamp S260S 2 Brush Gul. 123 Prospect 52692 3 Dry Cr. 123 W.Fork Dry 52642 6222 4 Sheep Gap 123 Swamp 52600 S Earth's Ranch 123 Pafs K-Tob 6632 S Therriault Gul. 123 Prospect 52670 6 1 04 7 Webster i^lanch 123 Swamp 5261 2 6540 8 Shamroclc Gul. 123 Prospect 52689 6113 9 McCrea Ranch 123 Swamp 52625 6534 10 Clark 1 123 Clark S25S9 8290 11 Clark 2 123 Clark 52643 12 Eddy 123 Eddy 52658 6423 13 Hill 7 123 Wilkes 52662 14 Wilkes Cr. 123 Wilkes 52675 81 94 15 Table Top 123 Wilkes S2687 6131 16 Clear Cr. 123 Prospect 52748 17 Kraak's Ln. 123 Prospect 52672 6123 18 IMiller Cr. 123 Millar 52557 6556 19 Boyd 200 Soyd-Tam. 52403 6385 20 Mayo 1 200 Boyd-Tam. 52440 S41 5 21 Tamarack 200 Boyd-Tam. 52461 22 Dry Fk. Tamarack 200 Boyd-Tam. 52494 6435 23 Keith 200 Boyd-Tam. 52441 6414 24 Mayo 2 200 Boyd-Tam. 52450 25 Wolf 200 Boyd-Tam. 52454 6398 26 Camel's Hump 200 Boyd-Tam. 52475 6365 27 Mayo 5 200 Boyd-Tam. 52420 6455 28 Camel's Hump 3 200 Boyd-Tam. 52479 6355 29 2-Mile Cr. 202 2-Mile 52381 63S8 Marble Cr. 202 Marble 52321 5482 « Wan da's 202 Marble 52322 6516 Bouchard Lk. 202 Marble 52322 6521 ■Irap sites not included in Figure 2. 18 Elk Distribution and Movements Seasonal movements and distribution of 119 radio-collared elk (92 cows, 27 bulls) were documented through radio telemetry for the period July 1985 through June 1990. Additional information was gathered from eartag returns and observations of neck-banded elk. Aerial telemetry efforts resulted in 3,865 relocations in HDs 123 and 200. In addition, 118 relocations were made in HD 201, HD 2 02 and Idaho. Relocations were obtained during each month for all years, but more relocations were collected in October than for any other month, because of our interest in elk responses to the general rifle season. Radio relocations were entered in dBase III computer files, and hard copies were forwarded, as collected, to LNF biologists. The study area initially encompassed 629 mi^ defined by the boundaries of HDs 123 and 200. However, examination of 2866 radio locations of 82 adult cows wintering within those HDs revealed that the composite yearlong range of elk was not conterminous with the original study area boundaries but actually occupied 613 mi^ in HD 200 and portions of HDs 123, 201, and 202 (Fig. 4 and Table 2). Elk Herd Units Based on radio telemetry we initially described 7 elk herd units within HDs 123 and 200 (Lemke and Henderson 1987) . Additional trapping and telemetry in 1988 and 1989 identified 2 other small, but distinct, herd units in Miller and Wilkes Creek. 19 Figure 4. Composite yeariong range of cow elk captured on winter ranges in HDs 123 and 200, 1986 - 1990. Table Description of nine elk herd units in Lower Clark Fork Study Area, Elk Herd Unit Yearlong Range Size' Winter Population Slze0389)^ Seasonal Movements/Herd Unit Interactions Prospect Creek (based on 8 radio-lagged cows, 328 tlxes) 281.8 km*/ lOfl.S ml2 325+ Winternorth side of Prospect(Valentlne-Antlmony Cr.). SummeriSomeremaln near winter range;manycross C-C Divide to Randolf Cr. area. Occassional summer overlap with Boyd-Tam. herd In Hemlock & Meadow Mtn. areas. Wilkes Creek (based on 5 radio-tagged cows, 156 fixes) 243.7 kmV 96.4 ml^ 290+ Winter:lower Wilkes Cr. Prospect Cr. S, Dry Cr. Summerupper Wilkes, W.Fk.Dry Cr., crossing C-C Divide to Packer Cr. & Hemlock Mtn. Occassional overlap with eastern end of Pospect herd during winter and Prospect & Boyd-Tam. herds around Hemlock Mtn. during summer. Dry Creek (based on 4 radio-tagged cows, 176 fixes) 156.1 kmV 60.3 ml'^ 150+ Winter;W.Fk.Dry-Joan Cr. SummerUpper W.Fk.Dry. crosslngC-C Divide, upper SavenacCr. S, Cruzane Mtn. Occassional summer overlap with Boyd-Tam. herd ii; Cruzane area. Clark Mountain (based on 7 radio-tagged cows, 291 fixes) 149.9 kmV 57.8 ml^ 145+ Wurter:southern aspects of Clark Mtn. to Cherry Cr.. Knox 8. Goldrush Cr. up to C-C Divide. Summerupper reaches of Knox, Goldrush, S>Twelvemlle Cr. near Mt. Bushnell. Summer overlap with the Dry Cr. herd unit near Bushnell and Boyd-Tam. herd unit In upper Twelvemlle. Eckly Creek (based on 3 radio-tagged cows, 60 fixes) 50.0 km 7 19.3 nil^ 40+ Wliiterilower sections of Quartz Cr., Poacher Qui. & Eddy Cr. Summer:upperto mid reaches of Quartz, Malone, Poacher Cr. up to Eddy Mtn. No overlap with other herd units. Swamp Creek (basedon 12 radio-tagged cows, 418 fixes) 309.6 km7 119.5 ml^ 150+ Wtnterilower Swamp Cr., Bemlsh Cr.. and E. Fk. Swamp Cr. to agricultural land along Clark Fork R. Summer.-upper Swamp Cr., Bemlsh Cr., Dee Cr., and crossing C-C Divide Into Flatrock, TamarackCr. aii Olsen Pk.Summer overlap with Boyd-Tam. herd unit In Flatrock and TamarackCr. Miller Creek (based on 4 radio-tagged cows, 49 fixes) Pafs Knob (basedon 4 radio-tagged cows, 205 fixes) 44.9 km 7 17.4 ml^ 35+ Winterjlower Miller Cr. to W. Fk. Combest Cr. Summerupper Miller Cr. and Combest Cr. to Combest Pk. Some summeroverlap with Boyd-Tam. south of C-C Divide near Combest Pk. 51.4 kmV 13.8 mi-^ 40+ Winter:north of Pat's Knob and south of Clark Fork R. between Combest and Kennedy Cr. SummerCombest Cr. to Pat's Knob and east to Sheep Cr. Some summeroverlap with Miller herd unit alongCombest Cr. Boyd^Tamarack (based on 35 radio-tagged cows, 1 182 fixes) 646.2 kmV 249.4 mi 495+ WBiter:Boyd Mtn. east to TamarackCr. Occasslonaly In Rock Cr., E. Fk. 12Mlle Cr. 8. Henderson Hill. Summer;FlatrockCr., upper 12Mlle Cr., Cruzane Mtn.. & Meadow Mtn. Some summernear UpUp Mln.(HD202) & Mill Cr.(HD201). Some summeroverlap with other herd units In Packer-Randoif, In Trapper Cabin. & In Flatrock and Tamarack areas. AH herd units combined (basedon 62 radio-tagged cows, 2866 fixes) 1589.9 kmV 613.7 mi' 1400+ Composite yearlong range Includes all of HD 200, HD 123 (excluding upper reaches of Prospect Cr.), SW corner of HD 201, and small portion of HD 202 Just S of St.Regis R. between 2mile and Deer Cr. From IbLDAY maximum home range polygon. 'Based on elk sightabllity surveys and model (Interpolation for incompletely surveyed herd units). Figure 5 . Location of nine major elk winter ranges in HDs 123 and 200 defined by radio-collared cows and calves between the dates Jan. 1 - March 31. Figure 6. Composite yearlong range of cow elk in the Prospect Creek herd unit, 1986 - 1990. Figure 7. , Composite yearlong range of cow elk in the Wilkes Creek herd unit, 1986 - 1990. Figure 8. Composite yearlong range of cow elk in the Dry Qeek herd unit, 1986 - 1990. Figure 9. ; Composite yearlong range of cow eik in the Clark Mtn. herd unit, 1986 - 1990. Figure 10. Composite yearlong range of cow elk in the Eddy Creek herd unit, 1986 - 1990. r Figure 1 1. . Composite yearlong range of cow elk in the Swamp Creek herd unit 1986 - 1990. Figure 12. Composite yearlong range of cow elk in the Miller Creek herd unit 1986 - 1990. Hunting District Boundary O^x^ County/State Line Elk Relocation Herd Unit SCALE 0 Figure 13. : Composite yearlong range of cow elk in the Pat's Knob herd unit, 1986 - 1990. Figure 14. Composite yearlong range of cow elk in the Boyd - Tamarack Creek herd unit, 1986 - 1990. Herd units were defined by radio locations of cows and calves tha^ shared a specific geographical area that met their annual habitat requirements (Edge et al. 1986). Herd units are named for the particular winter range areas that those elk had in common (Table 2 and Fig. 5) . Typically herd units do not overlap during the winter months, but may partially overlap during other seasons. Herd unit size (Table 2) and distribution (Fig. 6, 7, 8, 3, lo, 11, 12, 13, 14) varied considerably. Our data indicated that movement patterns of many adult bulls (2+years~old) were substantially different than those of cows, calves and yearling bulls with which they shared winter ranges (Lemke and Henderson 1987) . To a large degree earlier evaluations were supported by additional data collected after 1987. It appearejj that mature bulls, particularly 2-year-olds, were the most likely population segment to act as dispersers and, therefore, not suitable for defining herd unit boundaries (see Seasonal Movements^ . It was interesting to note that 8 of the 9 herd units occurred in HD 123 north of the C-C Divide, while elk wintering in HD 200 belonged to a single Boyd-Tamarack herd unit. Seasonal Movements-HP 12 3 Elk winter ranges for the 8 herd units in HD 12 3 are primarily distributed from east to west with minor overlap between the Prospect and Wilkes Creek herd units (Fig. 6,7) . In the eastern 1/3 of HD 12 3 winter ranges occur on lower elevation south-f acinic 32 slopes from Pat's Knob to Miller Creek, from Swamp Creek to Sheep Gap, and around the mouth of Eddy Creek. A large portion of the winter range is located on private agricultural land and corporate- owned forest land. In Figure 5 the southern extention of the Swamp Creek winter range was exaggerated by early spring (March) movements of one cow. By comparison, the majority of elk wintering in the western 2/3 of the hunting district (west of Eddy Cr.) are on National Forest Lands in Knox Creek, Joan Creek, Dry Creek, Wilkes Creek, Clear Creek, Mosquito Creek, and on the north side of Prospect Creek from Valentine Gulch to Therriault Gulch. Generally, elk showed strong fidelity to a particular winter range through the course of the study, returning annually to the same wintering area. A few marked elk occasionally travelled between Wilkes Cr. and Prospect Or. winter ranges. Cow movements Cow and calf migration from winter ranges to higher elevation summering areas usually occurred from early March through April and May (Fig. 15,16). The timing was dependent on snow depth in transitional areas, weather, and the emergence of new vegetation. Elk that wintered south of Plains in the Swamp Cr. and Miller Cr. areas generally moved south to upper elevations, and approximately 30-40% of the radio-collared cows crossed the C-C divide to summer ranges in tributaries of Flatrock and Tamarack Creeks. Routes heavily used by elk residing south of Plains included Miller Creek, Bemish Creek, Swamp Creek, and Airvilla Ridge (Fig. 15) . 33 During the calving season (May 15 -June 10) adult cow elk were^ in the Sheep Gap area, lower elevations in Swamp Creek, East Fork Swamp Creek, Poacher Gulch, Dee Creek, Bemish Creek, Combest Creek, Miller Creek and above the private agricultural land on the mid- elevation slopes of Pat's Knob (Fig. 18) . With very few exceptions cow elk wintering in the eastern 1/3 of HD 12 3 selected calving sites north of the C-C Divide. Summer ranges were located on mid to upper elevations on Pat's Knob and the head of drainages and upper elevations along the C~C Divide. The Eddy Creek Herd Unit summered at high elevations in Poacher Creek and Quartz Creek (Fig. 19) . Elk that wintered in the western 2/3 of HD 12 3 generally exhibited southwesterly movements to higher elevation summer ranges near the C-C Divide (Fig. 16) . Many of those elk crossed the C-d^j^ Divide to summer ranges in the western half of HD 200. However, some Prospect Creek elk in upper Crow, Wilkes, Dry, and Knox Creeks moved very little and summered on and near their winter ranges. Geographic saddles and passes are well-used migration corridors (Fig. 16). In HD 200, Brimstone, Randolph, Packer, McManus and Savenac Creeks were commonly used as sxunmer ranges for some elk wintering in the Prospect area. Many May and June calving season relocations occurred in Brush Gulch, Hill 7, Knox Creek, Trapper Cabin, and Hemlock Mountain areas. Less frequently used areas selected by cow elk were the South Fork of Wilkes Creek, West Fork of Dry Creek, Crow Creek, Clark Mountain, and Savenac Creek in HD 200 (Fig. 18). 34 Hunting season security areas were mostly in upper elevation drainages and ridges generally greater than one mile from an open road (Fig. 20) . Migration from summer range to lower elevation winter range occurred during November or December, the timing dependent on temperatures and snow accumulation. Travel corridors were similar to those used during the spring migration. Most cow elk radio-tagged in HD 123 were relocated in the study area. Some exceptions were notable. A cow (1575) from the Pat's Knob Herd Unit traveled north across the Clark Fork River and Highway 2 00 to spend one to two months in HD 122 around Paradise Gulch before returning to the Pat's Knob area. Another 2 year-old cow (0157) was shot on the Idaho side of Mullan Pass in mid- September, after being trapped the previous March in Therriault Gulch in HD 12 3. Emigrating cows, those that permanently left their herd units, were notable, because most cow elk repeatedly returned to the same winter ranges. Cow (0012), captured in Sheep Gap (HD 123) in January 1987 as a two year old, summered along the C-C divide near Drury Peak in HD 2 00 in 1987. She returned the following winter to Combest Creek (HD 123) . Then in 1988 and 1989, she was relocated in Tamarack Cr. (HD 2 00) during winters and summers, never returning to HD 12 3. A second emigrating cow (0914) , was captured in W. Fk. Dry Cr. (HD 123) in Febiruary 1987 as a 3+ year-old. The following summer she was in the Meadow Mtn. area(HD 200) , and she remained in HD 200 between Timber and Savenac Creeks that winter. During the 1988 35 Figurel5. Migration corridors for elk crossing the east end of the C-C Divide. 0 5 1 0 km. Figure 16. Migration corridors for elk crossing the west end of the C-C Divide. 3 7 summer she was relocated between Savenac and Packer Creeks, until, she was legally harvested in Savenac Creek that fall. Most remarkable is the long distance movement of a cow (1067) , captured as an adult in Therriault Gul. (HD 123) in March 1987. This animal subsequently was harvested in November 1991 near Sheep Mtn. (HD 283) , some 90 air miles from her capture site. Unfortunately, no radio locations were available to further document this rare record of a long distance emigration for an adult cow elk. Bull movements Movement of bull elk was strongly related to the individual's age. Five radio-collared bull calves and 3 radio-collared yearling bulls followed very similar movement patterns as those for the cow elk with which they wintered. One exception was a bull calf (0925) . Captured in Miller Creek in HD 123, it summered as a yearling around Combest Peak and wintered that year in the Mayo Gulch-Wolf Creek area in HD 200. In contrast, seasonal movements of older bulls (2+ years-old) often did not follow those of radioed cows with which they wintered (Fig. 17) . Adult bulls generally utilized higher elevations than cows, calves and yearling bulls throughout the year and began upward migrations to summer ranges earlier in the spring (Fig. 23) . Their unpredictable and sometimes long distance movements made relocations difficult and resulted in fewer summer range locations. Two adult bulls (0370, 0502) did move to the same sxiramer range, as 38 Figure!?. Movements of 3 migrating bull elk captured in HD 121 Figurelg. Calving season locations (May 15 - June 10) for 82 adult cows in the Lower Clark Fork Study Area, 1986 - 1990. Figure 19. 'Summer range locations (June 30 - I Sept. 30) for cows and calves in the Lower Clark Fork Study Area, 1986 - 1989. ■y"'- '^. Hunting District Boundary 'Vxx' Couniy/State Line Elk Relocations Figure 20. i Early hunting season elk security areas used from Oct. 15 - Nov. 5, 1986 - i989. those occupied by the cows with which they shared winter ranges. However, four adult bulls had different movement patterns » Bull (0700) was captured in HD 12 3 on Clark Mountain as a 2 year-old, summered south of Interstate Highv/ay 9 0 on Gilt Edge Ridge in HD 2 02 and was shot in nearby W. Fk. Big Creek. During his migration from winter to summer range, this bull left behind the summer range occupied by the cows with which he wintered, passed through the summer range occupied by cows in the Boyd- Tamarack herd unit, and spent the summer and fall with an elk herd adjoining, but socially distinct from the Boyd-Tamarack elk. Emigrating bulls were noted. A bull calf neckbanded in Combest Creek (HD 123) was shot as a 2.5 year-old near Arlee, Montana in the North Fork of Valley Creek on the Flathead Indian Reservation, 34 air-miles northwest from its capture site. Yearling bull (0890) was captured on Hill 7 (HD 123) , smiunered in Packer Creek (HD 2 00) and Crow Creek (HD 123) , was not found for most of 1989, but was relocated on winter range in Little Beaver Creek and Rock Hill in HD 121, 10 miles northwest of its original winter range. This adult bull was not found again in the summer of 1990 and was subsequently shot that fall around Mosquito Peak in HD 121, as a 4.5 year-old (Fig, 17). Bull calf (0925) was captured in Miller Creek in HD 123. As a yearling, it spent the summer around Combest Peak, then wintered in Mayo Gulch and Wolf Creek (HD 2 00) . The following summer, as a 2 year-old, it was relocated again around Combest Peak and the upper Miller Creek drainage. Although this bull's radio apparently 43 fai.led in July 1990, it was shot that Septaniber on Mineral Ridge irj HD 200, 15 air miles west of its last relocation in June (Fig„ 17) , Seasonal Movements - HP 2 00 Elk generally wintered at lower elevations on south-facing slopes from Camel's Hump to Mayo Gulch and Butler Creek, from Tamarack Creek to Sesame Creek, and on Boyd Mountain (Fig. 5) . During mild portions of the winter, a few elk were found on Henderson Hill and in Middle Rock Creek, East Fork 12-Mile, Lower Flatrock and the Burnt Flats area of Tamarack Creek. The winter range in Figure 5 was exagerated by early spring (March) westerly movements of one cow to Timber Creek. Elk generally returned to the same wintering sites. Cow movements As cold weather and snow gave way to warmer temperatures anJ the regrowth of vegetation, cows and juveniles that wintered from Boyd Mountain to Seven-mile Creek moved west and north. Heavily used areas during the calving season (May 15- June 10) were Flatrock, Middle Rock, and Mineral Mountain Creeks (Fig. 18). By early summer many were on summer ranges in Flatrock, Mineral Mountain, Trapper Cabin, and Cruzane Creeks (Fig. 19) . Most elk wintering on the eastern end of Boyd Mountain remained there during the sxxmmer in proximity to irrigated hay meadows (Fig. 19) . Migratory routes followed a southeast to northwest direction (Fig. 21) . Flatrock Creek, Middle Rock Creek and the ridge complexes between East Fork 12-Mile Creek and Breen Creek were heavily used as elk travelled between low elevation winter/spring 44 Figure 2 1 . Migration routes followed by elk in the HD2vX) Boyd Mtn.-Tamarack Creek elk herd unit. range to higher summer ranges in tlie Cruzanne Mtn. and Brooks Mtn, areas. Those movements were not confined to specific mountai^^ passes or saddles, as seen in HD 123. Cows, calves and yearling bulls wintering in HD 2 00 did not cross the C-C divide to enter HD 123. Overlap of annual herd unit home ranges of elk wintering in HDs 123 and 200 resulted from movements of HD 12 3 elk to summer ranges south of the C-C Divide. While most of the radioed elk remained east of Twin Creeks, there were exceptions. For example, 7 radio-tagged cows (17% of radioed females) journeyed 20-25 miles from winter grounds in Wolf Creek and Mayo Gulch to summer range around Cruzanne and Hemlock Mountains. Three others (7% of radio-tagged females) moved south of the St. Regis River to summer in HD 202, between Little Joe and Big Creeks. One cow, summering in Tamarack Creek, occasional. ^j^^ crossed the Clark Fork River into HD 2 01 to use the Mill Creek area. Movements from summer range to winter range typically occurred during the fall hunting season or later in December. Hunting season security was noted primarily in the Brooks Mtn. , Flatrock Cr. , Camel's Hump, and Boyd Mtn. areas (Fig. 20). Those movements of elk returning to winter ranges ware most concentrated in the Flatrock Creek and Camels Hump areas during the hunting season. One cow elk radio-tagged in HD 2 00 permanently left the Boyd- Tamarack herd unit. Adult (>3 years-old) cow (0120) was captured in Mayo Gulch in 1986, moved that summer to Big Creek in HD 202, and was relocated there for 3 years before being harvested. 46 Figure 22. Movements of 4 migrating bull elk ( 1387, 1361, 1214, and 1360) captured inHD 200. Bull movements seasonal movements of bull elk partially overlapped the' movement of cows with which they wintered (Fig. 22). Again, bull movement patterns were related to the individual's age. Bull calves (1788, 1457, 0104, 0674, 0078, 1720, 1701, and 1523) and yearling bulls (0103, 1467, 0104, 0674, 0078, 1720, 1701, and 1523) followed the same patterns described by cow elk wintering in HD 200. MEAH ELEyftTIOH (FT) 5009 nam 3Qm 2008 4 T MONTH « 2EABLING BULLS COMS & CALVES ^ flIULT BULLS Figure 23. Mean monthly elevations of radio-collared elk in study area, 1985-1990. Adult bulls (>2 years-old) generally used higher elevations than cows, calves and yearling bulls during all seasons and began upward elevational movements earlier in the spring (Fig. 23). 48 Seasonal movements of older bulls did not necessarily follow those made by cows with which they wintered (Fig. 22) „ Six adult (2+ years-old) bulls (1539, 1164, 1467, 0136, 1701, and 0674) did go to the same summer ranges. However, 4 other adult bulls which wintered in HD 2 00 summered with cows from the Prospect Herd Unit in Dry Creek (1387 and 1361) and Brimstone Creek (1214 and 1360) . Additionally, 6 bulls (1301, 0104, 0078, 1436, 1788, 1344), upon turning 2 years-old, were not relocated within the study area. Most often we lost contact with these 2 year-olds in June. Radio failure could have accounted for some of these disappearances. Nevertheless, two of those bulls (1344 and 1788) were later harvested in the St. Joe River and Sandpoint areas of Idaho, 70-90 mi from where they were radio-tagged, and the collar from a third bull (0104) was found by an Idaho hunter near 4th of July Pass, about 3 5 airmiles west of the study area. Road Densities and Elk Use Patterns A 560 mi analysis area, defined by elk relocations in only HDs 123 and 200, contained 747 mi of roads in all classes (Table 3) for an average road density of 1.33 mi of road/mi^. When seasonal and permanently closed roads were omitted, the open road density decreased to an average of 0.88 mi/mi^. Evidently, very few tertiary roads remain open. Thirty percent of the total road map and 46% of the open road map were classified as unroaded (0 mi/mi^) (Table 4) . Unroaded habitat occurred in only 3-4 large habitat 49 patches of variable size, so that road densities were higher in the. roaded areas than the above averages might imply. Male and female elk responded differently to open road densities during most seasons. Bull elk showed a greater reluctance to use higher road density habitats thoughout the year (Table 5). In all seasons, bulls avoided road densities >0.5 mi/mi^'. Bulls showed the strongest preference for zero road density habitats during the archery and general hunting seasons. The seasonal relationship for the use of roadless areas by bulls appears in Figure 24. Cow elk were more evenly distributed throughout the study area in a wider variety of road densities than bull. During the winter cows generally confined their activities to road densities <1.5 mi/mi^ (Table 6) . During spring and summer, habitats having o;^:!^^ . 2 road densities of up to 3 mi/mi were used greater or equal to availability. Although cow elk preferred to be in areas with zero road density during summer, archery and general himting seasons, they were present as expected in habitats with road densities of up to 1.5 mi/mi . Seasonal use patterns by cows of roadless areas are displayed in Figure 25. Thirty-four percent of the study area roads were closed on either a seasonal or year-long basis. By looking at elk locations in relation to the total road system, it is possible to determine, if elk were using habitats adjacent to closed roads. 50 Road Class Number miles in study area Primary Open 248.02 Secondary Open 243.79 Tertiary Open 1.15 Seasonal/yearlong closed 254.34 Total 747.3 Table 3. Miles of road in the LowBr^T^TT^^TT^^''^^ Road Density (mi/mi^) Percent of Study Area All Road Map Open Road Map 0 33.52 4^.36 0-Oi 9.87 48.47 Oi-1.0 10.03 2.42 1.0-15 14.72 1.71 1.5-2.0 9.71 0.73 2.0-2.5 7.72 0.22 2J-3.0 5.26 0.09 >3.0 9.16 0.01 Table 4 Ppi-rr'P.n+- f-i-F o-(-T,^Tr ■ .n...rn,„„rrin»„™ i iiniiiiniHiiii n„||, m, y.iHsaaEBBjaa, density levels? ^ occurring in various road 51 Road Density (nu/ini"^) ■ ' -— — Se.usonal Use by Bull Elk Winter Spring Summer Archery General n n + + + U.IMJO 0.5-1.0 1.0-15 li-2.0 2^-3.0 >3.0 + = preferred, = = used as available, - = avoided based on simultaneous Bonferroni CI. (95%) Table 5. The availability and use of road density classes by bull elk in the Lov;er Clark Fork Study Area. Open road density only. • Bull elk preferred to be in zero road density habitats when their locations were overlaid on both the open road map (Table 5) and the total road map (Table 7) . By comparing Tables 5 and 7, differences in habitat use occur in road densities of 0.0-0.5 and 1.0-1.5 mi/mi^. In table 5 it is apparent that bull elk avoid open road habitats throughout the year. However, in Table 7 it clearly shows that bulls were using the medixim road densities even during the hunting seasons. This suggests that bulls were "regaining" lost habitats found behind gated roads. 52 Cow elk also preferred habitats with zero road density (Table 6) . However, no such strong preference was shown by cows, when all roads were considered; nearly all road densities were used as available (Table 8) . Like bulls, cow elk were using habitats where roads had been closed. Road Density (mi/mi^) Seasonal Use by Ojw Elk Winter Spring Summer Archery General 0.0 + + + 0.0-OJ + + 0.5-1.0 1.0-15 1-5-2.0 2.0-2.5 2i-3.0 >3.0 + = preferred, = = used as available, - = avoided based on simultaneous BonfesToni CI. (95%) Table 6. The availability and use of road density classes by cow elk in the Lower Clark Fork Study Area. Open road densities only. We did not attempt to evaluate other factors influencing habitat selection, such as population levels, forage availability, cover, topography and proximity to water. 53 1 ==^=1 Seasonal Use by Bull Elk Winter Spring Summer Archery Genera! 0.0 = + + + 0.0-Oi = = = = = Oi-1.0 l.O-lJ li-2.0 2Jj-3.0 >3.0 based on simultaneous Bonferroni CI. (95%) Table 7. The availability and use of road density classes by bull elk in Lower Clark Fork Study Area. All roads included. EXPECTED -^- - OBSERUED PERCENT SB 6B -• 40 - 23 WINTER I SUMMER SEfiSOH SRCH ERV GENERAL Figure 24. Seasonal preferences of bull elk for habitats with zero road density, expressed as percent use of zero road density habitats . 54 Road Density (mi/ mi^) Seasonal Use by Cow Elk Winter Spring Summer Archery General 0.0 0.0-05 Oi-1.0 1.0-1.5 + li-2.0 2.0-2.5 25-3.0 >3.0 + = preferred = = used as available - = avoided based on simultaneous Bonferroni C.I. (95%) Table 8. The availability and use of road density classes by cow elk in Lower Clark Fork Study area. All roads included. 30 • 2B - IB — -- • e I 1 \ _ 1 _ — I WINTER SPHING SUMMEH ARCKERV CEJSERftL SEflSOH Figure 25. Seasonal preferences of cow elk for habitats with zero road density, expressed as precent use of zero road density habitats . 55 Elk Mortality During the 19 85-199 0 study period, the cause of death was recorded for 56 marked (eartagged, radio-tagged, or neckbanded) elk within the study area (Fig. 26) . Legal hunter harvest accounted for 40 (71%) of these. Another 4 (7%) elk were legally harvested outside the study area. Of the 2 5 marked bulls verified as being dead during the study, 3 (12%) were harvested outside the study area, after having left it as 2 year-olds. Only 1 (3%) of 31 dead, marked cows was harvested outside the study area. Other causes of death included wounding (4%) , during both bow and rifle seasons, illegal harvest (6%) , and collisions with motor vehicles (6%) . Other causes of death, including predation and winter kill, were negligible. Only one of the marked elk was known to have been harvested by bowhunters, despite the popularity of the area with archers. Members of the Confederated Salish/Kootenai Tribe hunted and harvested elk within the study area, but the only ear-tag return obtained from tribal hunters was from a bull captured in HD 123, taken by a tribal hunter west of Arlee on the Flathead Reservation MS USfl HARVEST ESS ROflC KILL 4- ^2 wouKJins ZZl PREDftTOh EZH LLiS-iL HflRVEST I I JMCtlOUIl Figure 26. Causes of mortality for 56 _ marked elk, expressed as percent of knowr|J mortality. Bull Mortality The observed annual weighted harvest rate was 40.5% (SE=13.07, Var.=1195.25) for a pooled sample of bulls radioed in both HDs (Fig. 27) . By age-class, harvest rates were 20% (N=15 elk years) of the yearling bulls, 53.8% (N=13 elk years) of the 2-year-old bulls, and 55.5% for the older bulls (Fig. 28) . Harvest rates of radio-collared bulls at least 2 years-old (54.5%) were significantly higher than for yearling (spike) bulls (20%) 1X^-4.41^ df=:l^ ,05>p>.025) (Appendix 8) . We obtained eartag returns for 26 (76%) of the 34 bulls captured in the study area (Appendix 1, 2, 3). Based on this sample, legal harvest accounted for 73% of the known mortality. Other sources of mortality included wounding, vehicle accidents, and predation. HP 12 3 Of 13 bull elk captured in HD 123, 11 (85%) were known to be dead by the end of the 4-year study. Seven were harvested by hunters in the study area. One bull calf was killed by a mountain lion in March, one month after capture. One bull died in September in Crow Creek, possibly wounded during the archery season. Another 57 PCSCEUT COUS B«LLS cons BULLS COKS BULLS HD 12! HO 298 SlUCV ftREA Figure 27, Weighted annual harvest rates for radio-collared bulls and cows in HD 12 3, HD 2 00, and both HD ' s conibined. 3 year-old bull was shot, and abandoned by hunters in Big Creek (RD W 2 02) . A nec3cbanded bull was harvested by a tribal member in North Valley Creek on the Confederated Salish-Kootenai Indian Reservation, approximately 3 5 miles from the study area. The fate of 2 neckbanded bulls was unknown. A 61.5% (N=13 elk years) hairvest rate was found in the pooled sample antlered bulls radio-tagged in HD 123 (Table 9) . The harvest rates were 50% (N=4 elk years) for yearling bulls, 60% (N==5 elk years) for 2 year-olds, and 75% (N=4 elk years) for 3-year and older bulls (Table 10) . HUNTING SEASON 1986 1987 1988 1989 1990 POOLED SAP/IP^ # ALIVE (beginning season) 1 4 3 3 2 13 # HARVESTED 1 3 1 1 2 8 % HARVESTED 100 75 33.3 33.3 100 61 J Table 9. Annual harvest of bulls radio-collared in HD 123 HD 200-Of 21 bulls captured in HD 200 between 1984 and 1988, 10 (50%) were harvested by hunters in the study area, 2(10%) were legally harvested in Idaho, 1(5%) was shot and not retrieved, 1(5%) was killed by a vehicle, and 1(5%) was a trapping mortality. The fates of 6 other marked bulls are unknown. 58 BMa.3.0 TOTAL # ALIVE-beginning season (pooled sample) 4 5 4 13 # HARVESTED 2 3 3 8 % HARVESTED SO 60 75 61-5 limiHmCT»aBiM?MM»M»miJ»i«|in»f|,>Mwj,M:M,M.iijai^^ Table 10. Harvest rates of 3 age-classes of bulls radio-tagged in HD 123. HUNTING SEASON 1984 1985 1986 1987 1988 1989 POOLEDSA MPLETOTA L # ALIVE (beginning season) 4 6 6 4 3 1 24 # HARVESTED 0 1 1 3 1 1 7 % HARVESTED 0 125 12-5 75 33.3 100 29.2 Table 11. Annual harvest of bulls radio-tagged in HD 2 00 AGE (YRS.) li 25 >3.0 TOTAL # ALrVE-beginning season (pooled sample) 11 8 5 24 # HARVESTED 1 4 2 7 % HARVESTED 9.1 50 40 29.2 Table 12. Harvest rates of 3 age-classes of bulls radio-tagged in HD 2 00, The annual harvest rate was 29.2% (N=24 elk years) for the pooled sample of antlered bulls radio-tagged in HD 200 (Table 11) , ^9 significantly lower {X^=3e58, df=l;, ,10>p>.05) than the harvest 59 rate observed for bulls radio- collared in H D 12 3 (Appendix 7) . The harvest rates were 9.1% (N=ll Figure 28, Hunting season mortality rates for 3 elk years) age-classes of bulls in the study area. for yearling bulls in this sample, 50% (N=8 elk years) for 2 year-olds, and 40% (N=5 elk years) for 3+ year-old bulls (Table 12) . Cow Mortality The annual weighted harvest rate for for the pooled sample of radio-tagged cows in the study area was 16% (SE=4.23, Var.--125.41) (Fig. 27) . We had eartag returns for 31 (22%) of the 142 cows captured during the study. Over 7 0% were legally harvested in the study area. Other sources of mortality included wounding loss, illegal harvest and collisions with motor vehicles. 60 l.S 3.5 3+ RCE OF BULLS HUNTING SEASON 1986 1987 1988 1989 1990 POOLED SAMPLE TOTAL # .AjLIVE (beginning season) 21 31 26 9 8 95 # HARVESTED 4 ^ — ^ 6 7 0 0 17 % HARVESTED 19 19,4 26.9 0 0 17.9 Table 13. Jliiatlal feaififest: of cows radio-taigf^ed in HD 12 3. EAST WEST # ALIVE-beginning season (pooled sample) 43 52 # HARVESTED 7 10 % HARVESTED 16.3 19.2 Table 14 . Har\^est rates of cows radio-tagged in eastern and western portions of HO 12 3. HUNTING SEASON 1984 1985 1986 1987 1988 1989 POOLED SAMPLE TOTAL # ALIVE (beginning season) 2 14 29 31 8 2 86 # HARVESTED 0 0 3 5 3 1 12 % HARVESTED 0 0 10.3 16.1 37.5 50 13.9 Table 15. Annual harvest of cows radio-collared in HD 2 00. 61 HP 123 Out of 53 cows, captured and marked in HD 123, 14 (26%) wer? known to have died during the study period. Causes of death included legal harvest within the study area (71%), legal harvest in Idaho (7%) , and illegal harvest (21%) in the study area. The annual harvest rate for the pooled sample (N=95 elk years) of cow elk radio-tagged in HD 123 was 17.9% (Table 13). The harvest rate for cows radioed in the eastern third of the district (HD123-EAST) , where antlerless permits were available, was 16.3% (N=43 elk years), compared to 19.2% (N=52 elk years) for the western 2/3 of the district (HD123"-WEST) , where antlerless elk were legal during an 8-day either-sex season (Table 14) . The rates of harvest in the two halves of HD 12 3 were not significantly different (X''=0.13, df=:l, e90>p>.50) (Appendix 10). HD 2 00 Of 89 cow elk captured and fitted with eartags and either neckbands or radio collars, 17 (19%) were known to have died during the study period. Causes of death were legal harvest in the study area (76%), wounding (12%), motor vehicle collision (S%) , and illegal harvest (6%) . The annual weighted harvest rate was 13.9% (N=82 elk years) for the pooled sample of cow elk rad ctagged in HD 200 and remaining in the study area during the hunting seasons (Table 15) . The harvest rate for cows radio-collared in HD 200 was lower, but not statistically different (X^=0.52, df~l, .50>p>„lo) , from that for all cows radioed in HD 123 (Appendix 9) . Even though "th^p 62 differences were greatest between the harvest rates of cows radio- tagged in the western side of HD 123 (19%) and in HD 200 (14%), this difference was not significant either (X^=a.S7^ df:^!^ . 50>p>. 10) (Appendix 9). Hunter Harvest Survey Hunter harvest statistics for HD 123 (Table 16) and HD 200 (Table 17) are based on the statewide telephone sur^/ey of elk license holders. An average of 343 elk were harvested by 2,162 hunters annually in the study area between 1986 and 1989. Each district averaged between 1,050 and 1,100 hunters annually during the study period. The average annual harvests were slightly higher in HD 200 (Table 16) . An estimated 164 (48%) elk were taken in HD 123, while 179 (52%) came from HD 200. Antlerless elk harvests in HD 12 3 were characterized by s-ubstantial fluctuation, rising and falling over 50% annually. The 1986-89 average elk harvest in HD 200 was 7% greater than the average for 1980-85. The increase in the antlerless permits accounted for the increased harvest during the study period. The 1988 hunter harvest was much higher than average in both hunting districts. The 1988 elk harvest set a statewide record of an estimated 26,211 elk. Hunter density averaged 2.9 hunters/mi"^ in HD 12 3 and 4.6 hunters/mi^ in HD 200 from 1986 through 1989. For the 1988 and 1989 seasons, an average of 3.5 and 4.7 hunters/mx used HD 123 and HD 200, respectively, reflecting a slight upward trend in the number of elk hunters in the study area. 63 HUNTERS PERCENT SUCCESS NO. PERMrre' HARVEST TOTAL BULLS COWS CALVES 1986 746 17 50 190 94 94 2 1987 958 14 75 132 78 48 5 1988 1343 14 100 194 108 72 10 1989 1337 11 100 142 101 2;; 15 AVE. 1096 15 81 164 95 60 8 'Permits valid in eastern 1/3 of HD 123; either-sex season for 8 days in western 2/3. Table 16. Estimated elk harvest in HD 123 from statewide hunter harvest survey, 1986-1989. YEAR HUNTERS PERCENT SUCCESS NO. PERMIT HARVE ST TOTAL BULLS COWS CALVES 1980 1215 16 100 157 122 27 5 1981 1060 11 100 119 95 17 6 1982 786 19 100 148 112 32 5 1983 1032 9 100 97 65 29 3 1984 1001 22 125 204 148 49 7 1985 893 13 150 114 58 48 7 1986 1057 14 150 149 94 51 4 1987 998 17 150 166 106 53 8 1988 1077 23 175 249 180 62 6 1989 1130 13 175 1.53 98 50 5 AVE. 1025 16 133 156 108 42 6 1980-85 AVE. 998 15 113 140 100 34 6 1986-89 AVE. 1066 17 163 179 120 54 6 Table 17. Estimated elk harvest in HD 2 00 from statewide hunter harvest survey, 1980-1989. 64 Antler Point Distribution Distribution of antler point configurations, as reported by hunters during harvest surveys, was similar for the two hunting districts (Table 18 and 19). During the 4 hunting seasons, 55% of the bulls ha;rvested in HD 123 had antlers with 2 points or less ANTLER PTS. 1986 (no.) 1987 (no.) 1988 (no.) 1989 (no.) TOTAL (no.) % IX 23 13 24 19 79 48 2X 4 2 2 3 11 7 3X 4 1 1 3 9 5 4X 7 2 3 5 17 10 5X 12 7 8 7 34 21 6X 3 2 3 5 13 8 7X 0 1 0 0 1 1 TOTAL 53 28 41 42 164 100 Table 18. Antler point* distribution for bull elk harvested in HD 123, 1986-1989. (configurations characteristic of yearling bulls) , while 52% of the bulls harvested in HD 200 had those configurations. In HD 123 9% of the bulls were reported to have at least 6 points on a side, and in HD 200 8% had at least 6 points. No trend in antler point distribution was apparent for that 4-year period. 65 ANTLER 1986 1987 1988 1989 TOTAL % ^ PTS. (No.) (No.) (No.) (No.) (No.) IX 20 19 29 18 86 48 2X 2 3 1 8 4 3X 0 2 5 2 9 5 4X 4 2 8 7 5X 9 8 15 8 40 22 6X 1 4 6 2 13 7 7X 1 0 0 1 2 1 TOTAL 37 37 66 39 179 99 * Antler points refer to antler side with fewest number of points. For example, IX includes 1x1, 1x2, and 1x3 configurations. Table 19. Antler point distribution for bull elk harvested in HD 200, 1986-1989. Population Estimates HD 12 3 No population estimate was attempted for HD 123 in 198 6, because of a very small number of marked animals. But for 1987 and 1988 L-P population estimates and 95% C.I.'s (p<.05) for elk wintering in HD 123 were 1035+159 and 1136+395, respectively (Table 20) . Fewer marked animals resulted in higher standard errors and larger confidence intervals in 1988. The 95% confidence intervals were ri-.latively large, representing 15% and 35% of the population estimates, respectively for the 2 years. From 3 6 to 42 marked animals available during any one survey. In 1989, using elk sightability methodology, we estimated that 911+305 elk were wintering in HD 123 (Table 25) . Stratified random sampling of 10 units out of 28 subunits in 2 strata resulted in a large confidence interval, 33% of the estimate. The populatior^JI 66 1987 .1988 1st Repl. Elk Observed 497 598 Marks Observed 21 14 Marks in Pop. 42 36 iy L'P Estimate 972 1477 2nd Repl. Elk Observed 487 502 Marks Observed 21 26 Marks in Pop. 42 39 0 / L-P Estimate 953 779 3rd Repl. Elk Observed 411 431 Marks Observed 14 14 Marks in Pop. 42 39 Observ. Index* 33 36 L-P Estimate 1180 1151 Mean Observability Index 44 47 95% Confidence Interval _tll Mean L-P Estimate 1035 1136 95% Confidence Interval _+159 ^395 •Observability Index = (#Marks Observed/#Marks in Population)xlOO. Table 20. Lincoln-Peterson population estimates for elk wintering in HD 12 3. TOTAL COW BULL CA SP RAG AD. UNC. RAW COUNT 263 188 13 58 8 4 1 5 EST. NO. 911 611 68 162 54 9 6 76 90% CI. J:.305 jt216 J:53 +60 _t51 i6 _t.81 'ELKMONT-2 strata (10 of 28 subunits) Table 21. Elk population estimate ifs Mo 123 using "ELKMONT" sightability model* in 1989. estimate resulting from the 1989 sightability survey was lower than either L-P estimate from previous years. The confidence limits about the estimates partially overlapped for all three years (Fig. 29) . 67 HP 200 Population estimates, using L-P methodology, for elk '^/intering in HD 200 were 814 + 252 in 1986, 544 + 136 in 1987, and 627+137 in 1988 (Table 22) . Lower standard errors were obtained in 1987 and 1988, when 53 to 65 marked elk were known to be in the population of 409 to 764 elk. The 95% confidence interval represented 15% of the population estimate in 1986, 12% in 1987, and 22% in 1988. In 1989, using the sightability model, we estimated that 495+82 elk were wintering in HD 200 (Table 23) . A stratified random sampling technique was used, in which 12 of 18 subunits were sampled in 3 strata. The confiflence interval was not large, 16% of the estimate. All L~P estimates were higher than that obtained with the sightability model (Fig. 30) » The 1986 estimrte was substantially higher than those for succeeding years; however, the 1986 estimate was based on a relatively low proportion of the population being marked (5%) and resulted in large standard errors and confidence limits. Although estimates resulting from the two methodologies differed, confidence intervals about the estimates partially overlapped for 1986, 1987, 1988, and 1989. The elk sightability surveys resulted in an estimated 1,406+387 elk wintering in the study area (both HDs) in Februa:tT,7 1989. If the estimated 900 overwintering cows produced 300 surviving calves the following summer (as indicated by observed calf/cow ratios) , about 1700 elk would have occupied approximately 68 . 2 613 mi (defined by 2366 radio locations of 82 cow elk) during the summer and fall of 1990 at an average density of 2,77 elk/mi^. 1986 I'M 1988 1st Rep. Elk Observed 262 133 274 Marks Observed 10 15 19 Marks In Pop. 40 57 53 Observ.Index' 25 26 36 L-P E.st!mate 979 4SS 742 2nd Rep. Elk Observed 305 1S4 195 Marks Observed 13 23 24 Marks in Pop. 40 60 61 Observ.Index* 33 38 39 L-P Estimate 3rd Rep. Elk Observed 262 269 249 Marks Observed 18 24 25 Marks in Pop. 40 62 65 Observ.Index" 45 39 38 L-P Estimate 567 679 634 Mean Obserrability Index 35 34 38 95% Confidence Interval _+& Mean L-P Estimate 814 544 627 95% Confidence Interval Ji252 -±.136 + 137 "Observability Index = (#Marks Observed/* Marks in Population)X100. tiMait!JihMwiimBiMBiamiggiB!3gffiaBmi^^ Table 22. Lincoln-Peterson population estimates for elk wintering in HD 2 00. RAW- COUNT EST. NO. 90% C.L TOTAL 294 495 -^82 COW 163 282 h62 BULL 33 69 +28 ELKM'^NT-3 strata (12 of 18 subunits sampled)" CALVES 59 100 + 23 SP 25 43 + 14 RAG 13 + 9 AD. BULL 13 + 14 UNCL 33 33 + 1 ■ - ■ — — I ..I -im. m ■■■iiirnii i..in i ii i ». i^im . f rjl..— J^->*^)^'Jjpflf ^-,n[a|-J^[;U(;W1tmMIB'»ffjf<^^»J^ Table 23. Elk population estimate in HD 200 using "ELKMONT' sightability model. 69 HUMBEIiS OF ELK lam 800 60S 40Q 2@Q _i_ iSSS Figure 29. Population estimates and 95% C.I. 's for wintering elk populations in HD 123. 1987-1988 estimates based on L-P Index. 1989 estimate based on "ELKI-IONT" sightability model. NUMBERS OF ELK 1280 i ima 608 4Qa 2m 1986 138? isea Figure 30. Population estimates and G.I. 's for wintering elk populations in HD 200. 1986-1988 estim^xtes based on L~P Index. 1989 estimate based on "ELKMONT" sightability model. 70 Elk Observability Observcibility of elk, calculated as that proportion of marked animals seen on any one complete winter range survey in each of the HD's, was highly variable (25-67%). Annual mean observ^abilities from replicated flights, however, were very similar from year to year in each HD. Mean observability values were higher for HD 12 3 than HD 200. The observability of elk in HD 123 ranged from 33% to 57% during 6 helicopter surveys made in 1987 and 1988. Mean observability values were 44% and 47% in 1987 and 1988, respectively (Table 20) . The proportion of marked elk observed in HD 2 00 varied from 25% to 45% during 9 helicopter surveys in 1986, 1987, and 1988. Mean observabilities were 35%, 34%, and 38% for the 3 years, respectively (Table 22) . The lower mean observabilities of elk in HD 2 00 was consistent with higher cover values reported for its winter ranges (Lemke and Henderson 1987) . Calf/Cow ratios HD 12 3 Observed ratios of calves/100 cows in HD 123 varied from 31 to 52 for eight surveys flown between 1986 and 1989. The ratios found in 1987 were slightly higher than those in 1986 and 1988. The lowest ratios of 27 estimated and 31 observed calves/100 cows were found in 1989 (Table 24) . 71 HP 2 00 Productivity was generally lower in HD 200. Observed ratios' of calves/100 cows varied from 21 to 42 during 10 suirveys conducted from 1986 to 1989. The observed and estimated ratios of 3 6 calves/ 100 cows for 1989 were somewhat higher than weighted mean and best count ratios for 1986 and 1987 (Table 25) . This contrasts with HD 123 where in 1989 calf/cow ratios were the lowest in 4 years. Year Survey # No.Elk Calves/100 Cows Bulls/lOOCows 1986 1 4O0 44 11 1987 1 497 51 15 1987 2 487 49 15 1987 3 411 52 9 1987 mean 51 13 1987 best count 51 15 1988 1 598 49 14 1988 2 502 46 14 1988 3 431 43 19 1988 mean 46 16 1988 best count 49 14 1989 1 263 31 7 1989 estimate* 27+6 ll_flO 'ELKMONT 90% confidence intervals IttWWKHIBE msssmmmssima Table 24. Elk classifications for HD 123 from direct observations (1986-1988) and sightability model estimate (1989) . Bull/Cow ratios HD 123 The observed ratio of bulls/100 cows varied from 9 to 19 during 8 surveys in HD 123 (Table 24) . Annual mean, best count arj 72 estimated ratios varied from 11 to 16 bulls/100 cows in the post season population. Comparison of mean, best count, and ELKMONT estimated ratios indicated very little annual variation during the study period. The "ELKMONT" estimated bull/cow ratio was similar to the directly observed ratio; however, the 9 0% bound was large, 91% of the point estimate. Year Survev# No.EIk OUIIs/ 1 vAJ\_AjWS 1986 1 262 32 31 1986 2 305 28 21 1986 3 262 21 25 1986 mean 27 25 1986 best count 28 25 1987 1 133 31 15 1987 2 184 42 20 1987 3 269 27 16 1987 mean 33 17 1987 best count 27 16 1988 1 274 40 12 1988 2 195 26 27 1988 3 249 24 13 1988 mean 30 17 1988 best count 40 20 1989 1 294 36 20 1989 estimate* 36±9 25^^11 •ELKMONT 90% confidence intervals Table 25. Elk classifications for HD 200 from direct observations (1986-1988) and sightability model estimate (1989) . HD 200 The proportion of bulls in the post-season wintering population in HD 200 was consistently higher than in HD 123. Observed ratios varied from 12 to 31 bulls/100 cows on 10 surveys flown from 198 6 to 1989 (Table 25) . Mean, best count, and 73 estimated ratios ranged from 16 to 25 bulls/100 cows during the study period, but no trend was evident. The 1989 "ELKMONT" bull/cow ratio was higher than the directly observed ratio, but the 90% bound was large, 44% of the point estimate, Time/Cost Comparisons of L-P vs. Sightabilitv Surveys An examination of L-P and sightablity surveys conducted in HD 123 revealed some differences in the inherent costs of the two methodologies (Table 26) . L-P methodology required an average of 28.6 hours of flight time per year to conduct 3 replicating surveys of winter ranges in HD 123 in both 1987 and 1988. This effort averaged 9 days each year to complete. Using the stratified random sample option for a sightability survey of this same area in 1989, only 15.4 hours and 2 days were required to complete the survey. Only 36 percent of all the units were sampled in HD 123, so the necessary flight time, costs and days needed to complete the survey were greatly reduced. Nine days and approximately 19 hours of helicopter time was required to complete the 3 replicating surveys needed for L-P estimates in HD 200. Just 2 days and nearly the same number of helicopter hours were required to complete one sightability survey. Only 4 days were required for the same personnel to complete surveys of both hunting districts in 1989. Some of the differences between costs in the two HD's were due to the means by which these surveys were conducted in HD 200. First, in 1989, even though a stratified random sample of subunits was chosen, 75% of all units were surveyed and the survey covered 74 Table 26. Comparison of flight time needed to complete helicopter sur^/'eys, HP Tvpe Hours Pays Costs' 1 O 0^ 1 lya/ 123 L-P 27.3 8 S 7923.00 1988 123 L-P 29.4 10 S 8379.00 1989 123 Sightability 15.4 2 S 4389.00 1986 200 L-P 19.1 9 S 544350 1987 200 L-P 19.4 9 S 5546.10 1988 200 L-P 18.7 9 $ 5346.60 1989 200 Sightability 19.8 2 S 5643.00 •Calculated @S285.00/hour for helicopter;oth er costs excluded SSEE Table 27. Comparison of total operations costs for Lincoln-Peterson vs. "ELKMONT" sightability surveys in HD's 123 and 200. Ave. Flight Time for L-P Survey Ave. Cost for Flight Time (L-P)' Estimated Cost of Marking Elk" HP 123 28.6hr $8,151.00 54,667.88 HP 200 19.1hr $5,44350 57,224.10 Total Operations Costs S12,818.8i $12667.60 Flight Time for Sightability Survey"' . 154hr 19.8hr Cost for Sightability Flight Time' 54,389.00 55,643.00... Total Operations Cost 54,389.00 $5,643.00 'Calculated @S285/hr for helicoptenother costs excluded. "Based on estimated costs of S111.14/marked elk; 42 marked elk in HD 123, 65 marked elk in HP 200. -"Based on a stratified sample of 36% of subunits surveyed in HD 123 and 75% of subunits surveyed in HD 200 ' . .;MHia.di,ni|lfWII all of the high density, most of the moderate density, and just 2 of the low density subunits. In reality, virtually all of the winter range was surveyed in HD 2 00. Secondly, the strict adherence to flying contours within subunit boundaries resulted in substantially better coverage and more flight time per unit area than we accomplished on L-P surveys in 1986, 1987, and 1988. Thirdly, in an attempt to standardize our survey techniques in 1986, 1987, and 1988, all flights were conducted during the first 75 few morning hours of daylight. Therefore, because 3 flights wera-^ necessary to conduct 1 complete survey of the winter range, 9 day^^ were necessary to complete 3 replicating surveys. On the other hand, the survey conducted using elk sightability methodology was flown throughout the day, beginning with first good light and ending in late afternoon with stops only for rest, refueling, and weather. Additional significant costs inherent to L-P estimates involved the cost of marking the elk for later aerial surveys. We (Thompson et al . 1989) estimated that it cost $111.14 to capture an elk during this study. Given a maximum of 42 marked elk in HD 123 and 65 marked elk in HD 2 00, the additional costs of doing L-P surveys and estimates were $4,667.88 in HD 123 and $7,224.10 in HD 200. Summing these additional costs with those for average annual flight time, we found that a more accurate estimate of costs for doing L-P estimates was $12,818.88 in HD 123 and $12,667.60 in HD 2 00 (Table 27) . Total operations costs using the "ELKMONT" sightability technique was $7,025-$8,429 less than the operations costs for the L-P Index methodology. This represents a 55-66% savings, depending on which HD was surveyed. Blood Sampling Blood samples were taken from 84 elk during the 198 6, 1987, and 1988 trapping seasons. Several samples were lost by the laboratory. Samples from 41 elk were tested for the presence of antibodies to brucellosis, bluetongue, anaplasmosis, and Leptospira 76 spp. 2 6 samples were tested for Protein B, the indicator of pregnancy. Pregnancy Ninety-two percent of older cows (>2 years-old) were determined to be pregnant with high levels of protein B. One of 3 yearling cows was pregnant. No calves nor bulls tested pregnant, using this method. Disease Antibodies for Leptospira autumnal is were found in 4 (10%) samples. One sample contained antibodies for Anaplasmosis . No samples tested seropositive for brucellosis or bluetongue antibodies. Investigations in Adjoining Hunting Districts Seasonal Movements-HD 2 02 Cow elk radio-collared in the Marble Creek area generally moved west in the summer to third order drainages on both sides of the Idaho-Montana border near Little Joe Mountain. However, 2 of 5 cows radioed in the Marble Creek area did not migrate west, but remained in close proximity to irrigated hay meadows between Marble and Dry Creeks. All 3 cows captured in 2-Mile Creek migrated to summer ranges in the Ward Peak area in both Montana and Idaho. In only two instances did any of the 8 radioed cows move east, leaving HD 202. One cow (1815) radioed in 2-Mile Creek in 1988 was located once in HD 200- in Mayo Gulch in spring 1989; she then 77 H.D. 121 121 121 121 YEAR 1986 1987 198S 1989 706 58 36 823 55 29 1033 49 22 1025 37 IS No.Elk Calves/100 Cows Bulls/IOQ Cows 201 201 201 201 1986 1987 1988 1989 165 207 135 122 31 31 28 3 10 5 5 202 202 202 202 1986 1987 198S 1989 467 276 266 342 26 43 28 27 22 27 9 Table 28. Elk helicopter classifications in HDs adjoining the Lower Clark Fork Study Area. returned to HD 2 02. The other instance was permanent emigration from HD 202 to HD 201. A 3+ year-old cow (112 6) captured in Marble Creek spent the summer of 1987 in the Simmons Creek area of Idaho, returning to her winter range in December. She again moved to the Idaho border in summer of 1988. Reversing course that summer, she returned to Marble Creek, crossed the Clark Fork River into HD 201, continued northeast and recrossed the Clark Fork River into HD 200 near Sesame Creek. She remained there until October 1988, when she re- entered HD 2 01, crossing the Clark Fork River. She spent that winter and the entirety of 1989 in the Four-mile drainage of HD 201. Sex/Age Ratios-HD 121, 2 01, and 2 02 Helicopter surveys were conducted in HD's 121, 201, and 202 during the study period. These surveys were not conducted in the 78 same standardized or intensive form as those conducted in KD's 123 and 2 00; however, calf/cow and bull/cow ratios derived from those surveys are presented for comparison in Table 28. Calf /cow ratios were generally higher in HD 121 and HD 12 3 than in HD 200, HD 2 01, and HD 2 02. Bull/cow ratios were lowest in HD 2 01, which has comparatively high open road densities, seldom reaching lo bulls/100 cows during post-season surveys. Bull/cow ratios were lowest for all HDs in February 1989, following record harvests in Fall 1988. Coordination and Interagency Activities From its inception, the Lower Clark Fork Elk Study emphasized and sought cooperation with LNF, Intermountain Forest and Range Experiment Station, the UM, and local landowners. Coordination activities included: 1. The study provided several students at UM with considerable field experience. 2. Milo Burcham completed a master's thesis of winter range habitat selection by elk under the directions of C. L. Marcum, UM, and L. Jack Lyon, USDA (Burcham 1990) . 3. The MDFWP and LNF completed the Mt. Bushnell area analysis of a 50,000-acre roadless area, very important to elk in the study area. The analysis resulted in setting long-term guidelines for road constuction and timber harvest in the area. 79 6. 7. 8 10 The MDFWP and the Superior Ranger District completed th^ 12 -Mile area analysis, resulting in protection o' security and summer range habitat. Information collected in the study guided MDFWP comments on the LNF Travel Plan and several timber sale proposals. L.J. Lyon (USDA) , J. Canfield (MDFWP) and M. Hillis (LNF) used radio relocation data collected during the study to develop "security habitat guidelines" for forest management (Hillis et al. 1991) . J.Canfield (1988) reported on the impacts of the construction of the BPA powerline on elk security and hunter opportunities. Cooperation between the 2 projects greatly improved the results of both studies.^ In exploring the various methods of population estimation, the authors met several times with Idaho Fish and Game biologists J. Unsworth and L. Kuck. We also assisted them in validation of the sightability model on the National Bison Range (Unsworth et al. 1990) , A 6-year prescribed burning program for nearly 4000 acres of elk winter range was planned and implemented with LNF. Funds were provided by the RMEF and LNF. MDFWP worked with landowners who have elk depredation problems. When possible elk were trapped, radio-tagged, and tracked to understand when and how elk used those properties. Survey information was used to set 80 antlerless permit levels in portions of districts with game damage problems. 11. The Lower Clark Fork Elk Study cooperated with the Upper Blackfoot Elk Study and Dr. E.O.Garton in modifying the Idaho elk sightability model into the "ELKMONT" model which utilizes a different type of helicopter. 12. M. Hurley completed a Senior Thesis at the UM on analysis of blood and fecal parameters in white-tailed and mule deer incidentally captured during this study (Hurley 1987) . 13. M. Thompson and the authors of this report co-authored a publication describing and analyzing the use of the modified Clover trap for capturing elk (Thompson et al. 1989) . DISCUSSION From 1984 through 1989, 176 elk were trapped and marked with neck-bands or radio-collars on all winter ranges within the study area with the exceptions of Cherry Creek and Donlan Flats. Marked samples of elk provided information on seasonal movement patterns and habitat use and were the basis for making modified L-P estimates of populations from 1986 through 1988. In 1989 population estimates were made using elk sightability methodology. An additional 10 elk were marked in the Marble Creek and 2-Mile Creek area of HD 202 to evaluate possible herd interactions. We learned that the elk wintering in the study area comprised a considerable resource of more than 1,400 individuals after the 81 fall hunting seasons. After calving in 1989, an estimated 1,700 elk occupied 613 mi^ at an estimated average fall density of 2.7^^ elk/mi^. Analysis of radio telemetry data revealed maternal elk movements to be predictable for described herd units, that herd units included habitats for year round needs and varied in size and composition, that most mortality (>80%) was human caused, and that rates of hunter harvest depended on the elk's sex, age, and hunting district. Seasonal Movements and Elk Habitat Herd Units Aerial relocations of radioed cows and yearling bulls revealed the existence of 9 herd units within the study area. Seasonal movements of those elk within each herd unit were predictable from year to year. Herd units were generally exclusive on thei:^^ particular winter range, but exhibited partial overlaps of summer ranges. Three areas of overlap were on sximmer ranges in the Trapper Cabin/Upper 12 -Mile, upper Flatrock/Tamarack, and the Cruzanne/Hemlock/Meadow Mtn. areas where cows, wintering in HD 123, had summer ranges overlapping with those of elk wintering in HD 200. Such predictability was not documented for bulls once they attained 2 years of age. However, few bulls were followed from year-to-year before radio failure or death of the animal; hence, established patterns of movement may not have been detected. It was clear, however, that bulls from one winter range often mixed with cows from another winter range during the summer and fall. 82 Possible explanations for the formation of several elk herd units in HD 12 3 and only one in HD 2 00 include differences in topography and winter range distribution. HD 12 3 is characterized by more rugged terrain with several relatively narrow deep north- south running drainages separated by high steep ridges. HD 200 is characterized by flatter topography and broader drainages separated by lower ridge lines. In HD 12 3 browse dominated winter ranges are scattered throughout the district typically on south and southwest aspects in the mid to low portions of several major drainages. In contrast, most productive elk winter range in HD 2 00 is located in low elevation "foothills" at the eastern end of the district. Topography and habitat in HD 123 appears more conducive to forming smaller segregated herd units than in HD 200. Edge et al. (198 6) suggested that knowledge of herd units should be incorporated into management decisions. In this study, an understanding of herd units based on telemetry data improved inferences about movements, key habitats, and population characteristics upon which management decisions were made. The herd unit descriptions aided the Mt. Bushnell analysis, the BPA powerline mitigation study, the 12-Mile analysis, forest travel planning, individual timber sale planning, population estimation, development of strategies for relieving game damage and our analysis of mortality. Herd units were not described for elk wintering in Cherry Creek and Donlan Flats. Therefore, we do not know whether or not those elk were part of larger herd units. Those geographic areas 83 were bordered by sampled areas and were small enough tha;^ reasonable inferences about movements could be made. It is likely that elk wintering in Cherry Creek are socially related to the Clark Mtn. herd unit and that they summer along the C-C Divide in the Cherry Creek and 12-Mile Creek drainages. It is most probable that elk wintering in the Donlan Flats area are socially related to the Boyd-Tamarack herd unit and that they summer along the C-C Divide between Patrick and Flatrock Creeks. It would be desirable to capture and monitor elk in those 2 areas in the future to check our assumptions. Relocation data revealed that virtually none of our radioed animals occupied summer ranges within the upper portion of Prospect Creek (Glidden, 23-Mile, 24-Mile) , although these areas supported large numbers of elk in the summer and fall. Those third ordd|^ drainages near the Idaho border may be part of a herd unit of elk wintering near Mullan, Idaho, or possibly wintering further west in Montana HD 121. We have no data to confirm these suppositions. Winter/Spring Habitat Winter habitat can be the most important limiting factor for elk productivity and survival (Skovlin 1982) . In this study, elk chose lower elevations with south aspects, available palatable forage, and coniferous cover. Winter habitat was available in both HD's and generally described in Fig. 4 and 12. Elk winter ranges overlapped those of white-tailed deer and mule deer, as evidenced by our ability to capture both deer species at sites where we also captured elk. 84 Primary elk winter range in the study area is synonymous with young to medium age shrub brush fields. The most important browse species are serviceberry fAmelanchier alnifolia^ , mountain maple (Acer qlabmm) , snowbrush ceanothus fCeanothus velutinua> , redosier dogwood fCornus stolonifera^ , willow (Salix sdp. 1 and chokecherry fPrunus virainiana^ . Burcham (1990) described and analyzed winter habitat selection for radio-tagged elk on 3 winter ranges in the study area for winters concurrent with our study. Climatic conditions were near long-term temperature and precipitation norms. The southeast, south, and southwest aspects accounted for 68% of elk locations. Elevations ranged from 2,640 to 5,760 ft and averaged 3,950 ft. Coniferous stands with basal areas between 10-150 ftVacre, densities between 50 and 250 trees/acre, and hiding cover of 0-50% accounted for the majority of the locations. Elk favored those habitats with little or no down and dead materials. They also favored areas of ground cover having high proportions of available and palatable forage. Those data were not compared to availability of those habitats, so while instructive, do not address habitat selection. Canfield (1988) noted that most winter elk relocations were on low elevation south facing slopes and were often near clearcuts and away from open roads. Important Land Survey Inventory (LSI) types chosen were LSI 33, 31, and 43, where LSI types represented distinct combinations of landform, habitat groups, and soils. This information should guide the LNF in planning winter habitat 85 improvement projects. Such projects may be particularly important in HD 200, where lower calf/cow ratios may result from lower forage production on winter ranges where conifer encroachment appears to be more advanced. Burcham (199 0) reported some elk chose northern aspects, often feeding on arboreal lichens (Alectoria spp. ) . until a crust-forming event, such as rain, would force them to move to southern aspects with less snow depth. He also noted that where there was little coniferous cover, such as in that portion of Prospect Creek which burned in 1973, elk were most commonly found in or near that portion of the winter range which still had mature tree stands. As photoperiod lengthened and temperatures rose in early spring, we noticed elk utilizing lower elevations and slopes with little or no coniferous cover. Those sites produced early and abundant graminoid regrowth sought by elk. Canfield (1988) found that elk preferred moist moderately steep coniferous sites in the spring and avoided roads in the Packer Creek Zone. Two LSI types (23, 53) explained some variation in elk distribution in 12-Mile Creek. In some areas such as Swamp Creek, Combest Creek, Donlan Flats, Marble Creek, and Middle Rock Creek, private hay meadows with a coniferous border were heavily used by elk. Summer/Fall Habitat As temperatures continued to rise and snow recede, most elk moved up in elevation following the emergence of succulent vegetation. Elk wintering in HD 123 often moved 8-12 miles to summer range in portions of HD 200. Cow and juvenile elk wintering 8^ in HD 2 00 typically found suitable summer range in HD 2 00 within 6- 8 miles of wintering areas. Some bulls (>2 years-old) marked in HD 200 moved north of the C-C divide to summer in HD 123. Some elk in both HD's remained at lower elevations closer to their winter range, where suitable cover and succulent forage were available. Calving areas were generally located in areas that elk occupied on their way to summer ranges. Most areas used during the calving season were in or near early successional sites producing large quantities of newly emergent grasses. During the summer, elk selected cool, moist mountain slope land-types in HD 200 (Canfield 1988) . Clearcuts of intermediate age were selected for, while those less than 5 years-old and greater than 20 years were avoided. Important LSI types were 14, 23, 34, 36, 43, 44, 45, 54 and 55. During the September-October breeding season elk chose denser timber stands and lower elevations than during the summer and avoided clearcuts. During the Mt. Bushnell analysis process, the LNF redrew "critical elk summer range" (Management Area 26) in the Forest Plan based on analysis of summer radio locations. Fall Security Habitat The lack of security, reflected in distributional responses and mortality rates, can result in elk selecting suboptimal habitat (Irwin and Peek 1983), in elk having suboptimal energy budgets (Morgantini and Hudson 1979) , in higher hunting season mortality (Irwin and Peek 1979, Hurley and Sargent 1991, Leptich and Zager 87 1991) , and ultimately in losses in hunter opportunity (Lonner and Cada 1982) . The analysis provided by Canfield (1988) documented distributional responses of elk to hunting season disturbances in a portion of our study area in HD 200. In that analysis elk often chose contiguous cover in LSI types 34, 44, and 54 during the hunting season. The amount of contiguous cover needed to maximize elk density during the pre-hunting and early-hunting seasons was 3 00 acres in the Packer Creek Zone and 500-600 acres in the 12-Mile Creek Zone. This was not the kind of cover selected by elk in the summer. Open and closed roads and trails were avoided. An examination of radio locations revealed increased movements and elevational changes by elk in response to hunting seasons. Fewer elk responded to hunting seasons in the less intensively habitat managed Packer Creek Zone compared to the Boyd Mtn. and 12-Mile Creek Zones. It was concluded that pre-hunting season habitats were not secure enough to hold elk during the hunting seasons. While Canfield 's (1988) analysis focused on a portion of HD 200, our GIS analysis of all radio-locations relative to road densities throughout the study area substantiated the conclusion that habitats were not secure enough across our study area. The harvest rates we documented also indicated that fall security habitat was inadequate for keeping hunting mortality of older bulls below recruitment in the present hunting season framework. The harvest rates of antlered bulls captured on various winter ranges differed, suggesting that amounts of security varied 88 within the study area, ie. there was less suitable elk security cover in HD 12 3 than HD 2 00. Response of Elk to Road Densities The GIS analysis of elk habitat use and road density showed dramatic and statistically significant negative responses of elk to even moderately roaded areas. Both cows and bulls tended to avoid habitats with >0.5 miles of road/mi throughout the year. Bull elk were more selective and actually showed a preference for only roadless habitats. Road density appears to be a major factor in determining elk habitat use within the study area. The implications of increasing road densities in core elk habitat are serious. Elk, particularly bulls, can effectively "lose" the use of important habitat due to the existence of road densities that are below allowable Forest Service Habitat Effectiveness Index (HEI) standards. Many Forest Plans have HEI standards of 50-70% for elk habitat. This translates into allowable road densities of 0.7-1.8 mi of road/mi^, which may not be sufficient to protect elk habitat use, according to our data. The LNF Forest Plan (1985) established a road density standard of 1.1 mi/mi^ maximum on highly productive big-game summer range and provided for closing newly built roads in areas of moderate big-game summer range- Our data indicated that these standards would not adequately accomodate elk security needs in many areas, since elk showed a strong preference for habitats with less than 0.5 mi/mi^ during most times of the year. 89 Our data indicated that closing roads may allow elk, to some extent, re-inhabit roaded habitats. The likelihood of significantly expanding road closures is uncertain, given the public's demand for vehicular access to National Forest lands. Continued education and interaction with agency personnel and the public need consideration. Emiarati on Though known to occur, elk emigration is a poorly documented and analyzed phenomenon. Emigration from our study area and from one herd unit to another within the study area was documented. Eartag returns and movements of radio-tagged elk indicated that bulls were more likely to emigrate from their maternal herd units than cows. Radio-collared bulls left maternal groups in the summer, after turning 2 years-old. Radio telemetry and eartag return data indicated that some travelled more than 50 miles before being harvested. Three of 4 bulls emigrated west into Idaho. Cows were less likely to emigrate than bulls, but radio data suggested that, if they did, they were more likely to emigrate to an adjoining herd unit. In addition, evidence from our small sample suggested that cows accomplished emigration as mature individuals >3 years-old. Geist (1982) also noted that bulls were more likely to colonize than cows. Sexually dissimilar emigration patterns may result from the different repoductive strategies employed by cows and bulls. Through frequent emigration 2 year-old bulls may accomplish a mixing of gene pools by breeding with a new, sometimes distant cow- 90 dominated herd unit and can expect to do so for several years before senility or death. Less extensive gene flow is further accomplished by adult bulls summering and rutting with cows in herd units adjoining those with which they spend the winter and with which they may or may not be related. Both adaptive behaviors were observed in this study. Infrequent adult cow emigration may insure a stable social structure within which calves learn to use the most productive seasonal habitats. Although long distance cow emigrations were rare, we observed both short and one long distance emigration by individual cow elk. Herd ComTPOsition and Productivity Herd composition varied within the study area. Calf/cow and bull/cow ratios were within the ranges reported by Rognrud and Jansen (1971) for other northwestern Montana elk herds. Annual variation in calf/cow and bull/cow ratios indicated that herd composition was dynamic. Generally, HD 123 had a lower proportion of bulls and a higher proportion of calves on winter ranges, when compared to HD 2 00. The proportion of yearlings among all bulls was higher in HD 123 than in HD 200. The lowest bull/cow ratios were recorded in HD 201, adjoining the study area, where open road densities appeared to be higher and hiding cover appeared to be lowest. All HDs overwintered fewer bulls following the 1988 hunting season, dramatizing the potential impact of one "good" hunting season on populations occupying insecure habitat. 91 The calf/cow ratios in HD 123, HD 121 and HD 201 were generally higher than those in HD 200 and HD 202. This may have resulted from lower forage production or higher densities of elk in HD 2 00. Winter range habitat improvement projects, such as prescribed burning (Leege and Hickey 1971) and well-planned timber harvests (Warner 1970) , may be needed to reduce coniferous enchroachment and increase browse production in order to maintain elk productivity at present levels. Serology exams indicated a very low incidence of exposure to leptospirosis and anaplasmosis , both of which may negatively effect production in domestic livestock. Kistner et al. (1982) , in their review of these two diseases in elk, reported that the effects of either disease on elk was not understood and that serology exams had so far proven undependable. Pregnancy rate determined from blood analysis for Protein B was very high. The >90% pregnancy rate for adult cows in the study area was consistent with pregnancy rates found in HD's 121 and HD 202 that were based on examinations of reproductive tracts collected in winter (Henderson, 1990 unpublished) . In 1983 and 1990 the peak date of conception in HD's 12 3 and 202 was mid-September with few conceptions coming later than October 1 (Henderson, unpublished) . Since pregnancy rates, sex/age ratios and habitats were similar for these adjoining herds, we can reasonably assxime that most breeding also occurred during September in our study area. Herd productivity, therefore, was not limited by conception. Habitat quality probably was the major factor| 92 influencing the health of pregnant cows and their calves ' health and survival. The proportions of bulls in the herds and the age structure of the bull populations were dependent on age differentials in mortality rates. Since most mortality occurred during the hunting season, so that the size and age structure of the bull segment were largely dependent on rates of harvest (see Hairvest and Population Management) . Harvest and Population Management Direct measurements of harvest rates were difficult to obtain - In this study, constraints on budgets and manpower resulted in small sample sizes. Our interpretations are therefore qualified. Cow management As expected, mortality rates for cow elk were much lower than for bulls. Hunting regulations have limited antlerless harvests since the 1970 's. The lower female mortality was reflected in higher eartag return rates for bulls, older age distribution for females in trapped samples, and higher hunting season mortality rates observed for radioed males than females. Through 1990, only 20% of the ear-tagged cows were known to have died, but at least 7 6% of the marked bulls had died during the study. Nvunbers of male and female calves trapped were nearly identical. Assuming no differential mortality between bull and cow calves during late winter and spring, yearling males and females would have entered the summer in nearly equal numbers. Similar proportions (12-14%) of cow calves, yearlings and 2 year-olds were 93 trapped, while 3+ year-olds represented 60%. In contrast, those same age-classes of bull elk were trapped in ever decreasing proportions (58%, 32%, and 3%, respectively), indicating strongly differential sex and age dependent mortality, trapability, or a combination of both. The overall hairvest rate for radioed cow elk in the study area was 16%, similar to that reported by Leptich and Zager (1991) for cows in the nearby Cour d' Alene drainage in Idaho. The harvest rate for cow elk in HD 2 00 was 14%, not significantly lower than a 18% harvest rate in HD 123, in spite of different harvest regulations for antlerless elk harvest in the 2 HD's. The rates were similar (14-16%) for HD123-EAST and HD 200, both of which har-vest cows on permits, and for HD123-WEST (19%) in which any elk could be harvested by any license holder during the first 8 days of the general season. The major difference between HDs was in the timing of the antlerless harvest. All antlerless harvest was concentrated in the first 8 days of the season in HD123-WEST, whereas the antlerless harvest is spread fairly evenly throughout the season in HD123-EAST and HD 200. Antlerless harvest in HD 123 may display greater variation in number because different fall weather patterns can produce good or bad hunting conditions during the first 8 days of the season, when either-sex harvest is authorized. Permitted harvest of antlerless elk over the 5-week season resulted in more predictable and stable harvests. 94 since many cows captured on winter ranges in Dry Creek and Prospect Creek utilized habitats south of the C-C Divide in the western end of HD 2 00, some were hunted on a permit basis during the first part of the season. At some time during or after the season, those cows moved north over the C-C Divide, where they could be harvested only during the first 8 days of the season. This increased the probability of lower antlerless harvests in HD 12 3 -West, when hunting conditions were poor in the first week of the season. The cow populations appeared to be stable during the study period. Confidence intervals of estimated post-season populations overlapped during the study period. Harvest rates for cows were close to the 14-20% recruitment rate, predicted from observed post- winter calf /cow ratios. Also, the estimated harvest of 109 cows in fall 1989 was similar to the 131+43 cow calves estimated to be in the study area in February 1989. While most evidence suggested that cow populations were nearly stable, other considerations suggested that this conclusion should be cautiously accepted. Other sources of mortality (i.e., tribal and illegal harvest, wounding losses, and predation) were noted and were probably underestimated by our study methods. Also, the proportion of aerially classified calves was consistently higher than the proportion of calves in the trapped sample of cow elk, suggesting the possibility of identification bias in aerial surveys or higher trapability of calves. Either could lead to over- estimating recruitment. If survival of cow calves was nearer the 95 10-14% in the trapped sample, then hairvest could have been slightl;^ higher than recruitment. Winters were generally mild during the study, and weather related mortalities were few. Since harvest rates were weighted annual means encompassing only 5 seasons, other climatic conditions could have produced both higher harvest rates and greater overwinter mortality. Because estimated mortality was so near estimated recruitment, the above considerations underline the need for conducting regular, standardized population surveys with measureable reliability. Bull management The observed annual weighted harvest rates for antlered bulls was 40% for the study area, 61% for those trapped in HD 123, and 30% for those trapped in HD 200. Those rates are based relatively small samples; however, they did fall within the range of values (31-62%) reported by Leptich and Zager (1991) for the Cour d' Alene area. In HDs 282/285 of the Blackfoot drainage, over 50% of all antlered bulls were harvested annually (Hurley and Sargent 1991) . Vore and DeSimone (1991) reported hunting mortality for all bulls in the Elkhorn Mountains to be 90% when any bull was legal, 59% during branched-antler bull (BAB) seasons, and 40%, when spikes were legal and branched bulls were taken on limited access permits. It was not clear that antlered bull numbers in our study area were declining under current harvest pressure, as hunting-season mortality sample sizes were small, previous aerial surveys wer 96 inadequate for demonstrating changes, and antler point distributions for harvested bulls showed a flat trend during the study period. Nevertheless, there was reason to be concerned that the MDFWP objectives for maintaining a diverse age structure with mature bulls (MDFWP 199 2) might not be met in the future. Sampled bull age structures, sampled harvest rates, and a comparison of estimated harvest and estimated recruitment indicated that harvest could cause a decline in the number of mature bulls. The overall harvest rate of 40% for antlered bulls was 2-3 times higher than for cows in the study area, which would be acceptable, if the number of bulls recruited into each older age class met or exceeded the number of deaths in that age class. As discussed above regarding observed cow mortality rates, our study methods probably underestimated actual mortality rates for antlered bulls. Moreover, the number of males (210) estimated to have been harvested in the study area in fall 1989 slightly exceeded the number of new bulls (131+43) estimated to be entering the population in 1989. The low proportions of 2+ year-old elk in both the trapped and aerial classification samples were also indicative of high mortality, or biases in observability and trapability. Harvest rates in our study appeared to be highest on bulls that were at least 2 years-old and lowest for yearling bulls, suggesting possible hunter selection for larger bulls and/or older bulls exhibiting behaviors predisposing them to greater vulnerability. This difference was statistically significant, but 97 based on siaall sample sizes. This finding contradicted the populaj opinion that "spike" bulls were more likely to be harvested because of their inexperience. Other studies have also reported age- dependent hunting season mortality in which yearling bulls were harvested at rates lower or equal to those found for older bulls. During hunting seasons in which all antlered bulls were legal in HD 380, 86% of the marked yearling bulls and 100% of the 2.5 year-old bulls died (Vore and DeSimone 1991) . m Idaho, harvest rates were nearly identical for yearling (spike) and older (branched) bulls (Leptich and Zager 1991) , In the Blackfoot River drainage (Hurley and Sargent 1991) , 2 year-old bulls exhibited the highest harvest rates (60%), while yearling and 3+ year-olds were harvested at similar lower rates (40 and 41%, respectively). m the Blackfoot. study, the higher harvest of 2 year-old bulls was attributed to greater hunting season movements and limited experience of that age class, while yearling bulls may have found increased security by their association with larger groups led by experienced cows. The uncertainty generated by the data we were able to collect and analyze, underscores the need for regular and reliable population surveys to monitor changes in bull numbers and age structure, as well as changes in overall population numbers. Such surveys had not been available and not in use until this study, and their future utilization depends on budget and manpower re- allocations. If future surveys demonstrate that numbers of raghorn (2 year- old) and mature bulls are declining, the MDFWP goal of maintaining 98 a diverse age structure of bulls would not be accomplished, since most mortality was related to fall hunting seasons, improved habitat security and/or adjustment of hunting season regulations would be necessary and require public support and inter-agency cooperation. Habitat Security Many studies have established that elk tend to avoid roads and that the degree of avoidance is related to traffic volume, topographic features, and road side cover (MarcUm 1975, Perry and Overly 1976, Lyon 1979). Marcum (1975) suggested that the loss of habitat resulting from elk avoidance of roads could be at least partially offset by road closures. Using GIS methodology (see above) , our analyses support earlier findings that elk, particularly bulls, choose habitats away from roads, but will reclaim roaded areas, once roads are effectively closed. In addition, roaded habitats lead to higher mortality rates where hunting is the principal cause of mortality (Irwin and Peek 1979). Leptich and Zager (1991) reported that hunting season mortality rates for antlered bulls in the Couer d' Alene River drainage in Idaho varied from 61.7% in highly roaded habitats to 31.2% in areas with few roads. The higher mortality rates led to relatively low bull/cow ratios with few bulls living to be 5 years- old. Our results (40% annual bull mortality, 16% annual cow mortality, 15-2 0 bulls/100 cows post-season) were comparable to data from "managed access" areas with security intermediate between those with few and many roads in the Cour d' Alene drainage. 99 Hillis et al. (1991) provided general guidelines for plannin(^ for habitat security, when developing timber management projects Canfield (1988) also recognized the limits of fall hunting season security and suggested that roads built for the construction of the BPA powerline would increase the likelihood of timber sales in the vicinity of the powerline corridor. Road building and reductions in cover associated with those sales would increase elk vulnerability to hunter harvest. The Mt. Bushnell Analysis (Appendix 11) recognized movement patterns and security areas and provided guidelines for future timber management actions within that 50,000 acre portion of the study area. Fall habitat security may be lower in HD 123 than in HD 200. The annual harvest rate of 62% for bulls radioed in HD 123 was twice that observed for bulls radioed in HD 200, and, whi]^ differences were statistically significant, sample sizes were relatively small. The lower bull/cow ratios observed in HD 123 also suggested a higher harvest rate for bulls wintering in HD 123. The higher proportion of yearling bulls in both aerially observed and hunter harvest samples further supported the conclusion that bulls wintering in HD 123 were more vulnerable to hunter harvest. The difference in harvest rates between the two HD's was surprising, considering hunter densities were lower in HD 123, bull harvest regulations were identical, and many bulls used both hunting districts sometime during the hunting season. Apparently, habitats in the two districts provided different levels of vulnerability to hunter harvest. An overall approach to planni||| XOQ for security is needed in the study area with emphasis on improving security for bulls wintering in HD 123, Hunting Seasons Our results indicated that hunting regulations that reduce overall bull mortality would be more appropriate than ones that favored a specific (ie. yearling) age-class. Throughout the study area yearling bulls experienced harvest rates of only 20%, while nearly 55% of the older bulls were harvested annually. Given the age-specific harvest rates found in this and other study areas, the popular BAB or brow-tine bull (BTB) seasons instituted elsewhere in the state would not be appropriate where the objective is to maintain a diverse bull age structure. If BTB seasons were applied, one could expect an even higher rate of harvest on that bull age-class (2+ year-old) which already is the most heavily harvested. The lightly harvested yearling class would receive an even smaller harvest, resulting in an ever-declining number of older bulls in the population. Although HD boundaries did not reflect elk herd unit boundaries (particularly on the western end of the study area) , the similarities of harvest rates for cow elk in the two hunting districts, regardless of the different season types, indicated that a change of district boundaries is not warranted on that basis, in addition, since a proportion of the bulls in each herd unit crossed herd unit, as well as hunting district, boundaries during the summer and fall, and, as antlered bulls were hunted throughout the 5-week season in both hunting districts, it was improbable that a 101 change of hunting district boundaries would change harvest rates on the antlered segment of the populations. Illegal harvest and wounding losses claimed an unknown proportion of the population. The small sample sizes and low probability of the public reporting marked animals dying from these causes limited our ability to do more than document their occurrence. Vore and DeSimone (1991) also reported wounding losses and illegal harvest of bulls. If game managers want to reduce the mortality rate on older age class bulls through changes in regulations, perhaps the best alternatives include shortening the season and/or going to limited access permits. Game Damage Problems Elk depredation was chronic in several parts of the study area^^ from spring through summer. Depredating elk were radio-collared and monitored in Swamp Creek (HD123) , Combest Creek (HD123) , Boyd Mtn. (HD 200) , and Marble Creek (HD 202) . While most radioed elk in the study area followed the development of green forage from low elevations in the spring to higher elevations in the summer, some remained as residents at low elevations throughout the summer and fall, often in the vicinity of hay and/or oat fields. The elk involved in depredation represented a small proportion (<10% estimated) of total elk numbers in the two hunting districts. When irrigated, grain crops and hay meadows were particularly susceptible to depredation late in the summer, when native, 102 vegetation cured. Most fields were relatively small in size (<100 acres) and bordered by hiding cover in coniferous timberlands . Crop depredation was tolerated at moderate levels for several years, before landowners lodged official complaints with MDFWP. Landowner tolerance of both hunters and elk was limited in many areas. Rural subdivisions near agricultural lands often made hunting more restricted on and near these properties than on public lands some distance away. Strategies for dealing with elk depredation in these situations have included herding, raising harvest rates on antlerless elk in the entire hunting district, raising harvest rates in those portions of the hunting districts affected by depredation, encouraging landowners to accept more hunters, improving hunter access near depredation sites, increasing elk security at higher more remote portions of the range, and authorizing early or late season damage hunts in areas affected by depredation. The relatively small numbers (20-100 elk) and limited seasonal movement of depredating elk in most cases suggested that increased harvest in those limited areas would reduce the level of depredation without seriously reducing populations elsewhere in the the HD's. Such efforts already made in Boyd Mtn. , Marble Creek, Pat's Knob, and Swamp Creek areas should be evaluated for efficacy at reducing depredation and for impacts on resident and migratory populations. 103 Evaluation of Survey Techniques Traditionally, aerial surveys in western Montana have been' conducted either from helicopters or fixed-winged aircraft during the spring "green-up". Because elk were thought to be more visible during the spring there is a greater likelihood of obseirving a larger proportion of the population during this brief "window of opportunity". Age and sex composition of the herds, as well as assumed trends in population levels, were obtained from what was most often a single flight over the winter ranges in April. Nevertheless, the reliability of this method of inventorying big game populations has been questioned privately by wildlife managers, and by researchers (Caughley 1974, Caughley et al. 1976) . Caughley (1974) stated that aerial surveys invariably contained biases, generally resulting in underestimates of the^ actual population levels. Samuel et al . (1987) developed a survey method that corrected for biases inherent in surveying elk from the air in northern Idaho. Rice and Harder (1977) noted that larger proportions of a white-tailed deer population needed to be marked to obtain reliable L-P estimates where more vegetative cover obscured deer observability. We also found that a smaller proportion of elk needed to be marked in HD 123 where there was less vegetative cover than in HD 200 to obtain similar confidence intervals, which were still very large. The results of replicated flights in 198 6, 1987, and 1988 to obtain L-P estimates of population numbers demonstrated th^ 104 difficulties involved with interpreting data collected by the traditional methods. The proportions of marked elk observed during aerial surveys fluctuated considerably from survey to survey within a 2 -week period. For example, in 198 6 2 5% of the marked animals in HD 2 00 were seen on the first survey, but 45% were seen on the third survey about 10 days later. The actual number of elk observed on surveys during the same year also varied considerably. As examples, in 1987, as few as 431 elk were seen in HD 123 on one survey and as many as 598 on another, and in HD 200 only 133 were seen on the first survey, but 2 69 were counted on the last replication. Replicated surveys for L-P Indices did yield fairly consistent annual mean observabilities and mean population estimates in each hunting district, suggesting that populations were nearly stable. In addition, where elk were more observable in the relatively more open habitats in HD 123, population estimates could be obtained with narrower confidence intervals than in the more densely vegetated habitats of HD 200. As an alternative to traditional survey methods, replicated surveys utilizing averages or highest counts would more likely reflect the actual trends in population levels and composition. More than 3 replications might be needed in some areas of western Montana with abundant coniferous cover. Replicated aerial surveys required a considerable time to complete. Given the vagaries of weather, up to 2 weeks was needed to complete 3 replications. More time would be required to conduct additional replications to provide more precise estimates. 105 Additional flights greatly increase the number of hours flown and inflate survey costs dramatically. Conducting L-P estimates was expensive. In addition to the costs for helicopter time, trapping and marking elk were required. A substantial investment of time, manpower and money (approximately $lll/elk, excluding salaries) were needed to mark enough elk to yield reliable estimates before ever conducting the aerial surveys. Those costs were nearly equal to the costs for flight time alone. As with L-P methodology, a major expense for using the elk sightability model is the cost for helicopter time. An earlier limitation was corrected by the modification of the computer model, so that data can now be collected from the 3-seat Bell-47, not just the Hiller, which is a helicopter largely unavailable in Montana. The advantages of the elk sightability model were described by Unsworth et al. (1991) . Our experience with the elk sightability survey and model in the Lower Clark Fork Study demonstrated several things: 1) complete and consistent survey methodology was important and had been lacking, 2) the method can be applied at anytime during the winter, not just during some narrow window of opportunity (ie. spring "green-up"), when observability is highest, 3) confidence intervals were obtained, so that testing for future changes can be done, 4) stratified random sampling can greatly reduced helicopter costs, 5) no costs were associated with trapping and marking animals to obtain an estimate, 6) surveys could be repeated without the loss of consistency, 7) estimates for all sex and age classes of elk may be obtained, and 8) the method 106 has direct application for many other mountainous, timbered elk habitats in northwest Montana. Non-consumptive Elk Values While the study was not designed to address non-consumptive values, we believe that the study area offers a good opportunity for viewing wild elk from public roadways, particularly during the winter and spring. Elk on winter ranges in Prospect Creek, West Fork Dry Creek, Swamp Creek, Pat's Knob, Donlan Flats, Tamarack Creek and St. Regis River are often visible from highways and forest roads, especially in the early morning or late afternoon hours . Evening rides on forest roads along the C-C Divide during the summer and fall will often encounter elk. Off-road travel in most drainages will provide opportunities to see elk or fresh elk sign throughout the year. Elk that are habituated to feeding in agricultural fields south of Plains and near St. Regis can often be viewed early or late in the day. These apparently benign activities can occassionally have negative consequences. Displacement from preferred habitats by human disturbance can result in loss of productivity. Also, landowners reported instances in which viewers had frightened elk, causing them to flee, breaking down fences and injuring themselves. MANAGEMENT RECOMMENDATIONS Both the MDFWP and LNF have long recognized the value and importance of elk populations that inhabit the Lower Clark Fork 107 study area in HD's 123 and 200 of northwest Montana. The social eoonomic, recreational and ecological values of this resour# Should not be underestimated. Ellc are very closely tied to the quality of life of the people living in this region. Many resource decisions are currently and should continue to be judged by their potential effects on elk and elk habitats. State and federal agencies have a public responsibility to protect and foster thriving elk populations through well planned management efforts. Based on the findings of the 1985-1990 Lower Clark Fork Elk Study, we recomnend that the MDFWP and the LNF consider the following management actions regarding the elk population found along the c-c Divide in HDs 123 and 200. Additional details are found in the Discussion section. Population ManafyomaT^i. 1. we recorunena that elk management goals be established for the si.e ana composition of the c-0 Divide elk population. Population management goals have been generalized over a larger 5 HD area in the statewide Blk Plan. Be suggest that it is more relevant and effective to define specific minimum/maximum population goals for elk in HD 123 and 200. The elk hunting season structure should be flexible enough so that periodic adjustments can be made to achieve and maintain the population management goals. Opportunity for adult female harvest Should be based on the annual surplus provided by net recruitment. 108 When hunter harvest and natural mortality of the antlerless population exceeds recruitment, the population will decline. 2. Population monitoring efforts should shift from early spring trend counts to the use of the "ELKMONT" sightability model on a stratified random sampling of subunits during the winter. The comparison of survey methodologies (trend counts vs. L-P Index vs. sightability model) indicates there are several advantages in adopting the "ELKMONT" model-ability to calculate sample size for desired precision level, standardized survey methods, predictability and economics. We recommend that a stratified sample of 50-75% of the winter range subunits be surveyed to reduce the size of the confidence intervals generated. The "ELKMONT" model and sightability methodology has application in other areas of northwest Montana. It would probably work well in nearby HDs 104, 121, 201, 202 and 203. 3. We are concerned about the ability of this elk population to sustain high mortality rates of older age class bulls, particularly in HD 123. It may be desirable to reduce overall bull mortality through a combination of hunting season changes and habitat management . . Wildlife managers need to decide if the current overall annual harvest levels of 41% for all antlered bulls and 56% for 3+ year- old bulls are acceptable, considering our attempt to maintain a diverse age structure of bulls; harvest rates for bulls in HD 123 are higher. Shorter bull seasons and/or limited access bull harvest should be considered to help develop a more diverse age 109 structure in the bull segment of the population. We do not believe^ that the standard 5-week BTB season would improve the bull age structure. Concurrently, habitat management that protects or develops security habitat, particularly in HD 123, should be a priority. Increasing elk security may allow more bulls to reach older age classes. 4. When time and funding allow, a small sample of elk wintering in Cherry creek and on Donlan Flats should be radio-collared and monitored to determine their annual movement patterns and herd unit boundaries. These are the only 2 potentially significant elk winter ranges that were not sampled in this study. Data from these areas would help complete the picture of elk use and herd unit interactions. Information from these areas could be used in making site specific^ project decisions. Habitat Management 1. We recommend that forest management practices emphasize the importance and perpetuation of relatively large 300-600 acres blocks of security habitat, particularly in LSI types 34, 44, and 54. Hunting season relocations indicate that 3 00-600 acre blocks of undisturbed forest provide very important security for elk, particularly bulls. Protecting these areas will help improve the survivability of older age class bulls and reduce harvest vulnerability of elk in general. Existing security areas 110 identified by relocations and similar habitat types should receive special management consideration at the project level. 2. Additional new road construction of all types be minimized in core elk habits, particularly those areas used heavily during the fall hunting seasons. We encourage further road closures in these same areas to allow elk to regain the use of habitats lost because of activities associated with roads. One of the most significant and disturbing findings of this study relates to elk non-use or "avoidance" of habitats that are moderately roaded at densities of 0.5-1.0 miles of road/mi^. As with other species, the question of road densities relating to habitat use is one basic to management. Elk appear to be quite sensitive to certain road density thresholds which are often considered acceptable as habitat objectives for elk. We urge that the LNF take this new information into account when planning future development projects in HD's 12 3 and 200. Essential elements that should be addressed are: 1. Location of seasonal high use areas; 2. Variables that could compromise security/bull carryover (ie. past logging, poorly recovered cutover areas, concentrated bull activity in hunting seasons, areas of high open road density, gentle topography or sparse vegetation, concentrations of hunters, large contiguous blocks of timber in MA's allocated to timber harvest) ; 3. Existing and desired levels of security (assuming that 300-600 acre patches of timber are required) ; 111 4. Acceptable road density; 5. Road closures needed (location, seasonal or yearlong); 6. Desired vegetative pattern and acceptable seasons for logging. 3. The importance of a systematic controlled burning program to improve and create shrub field elk winter range should be recognized and incorporated into long-term forest management plans. We encourage the LNF to actively continue with its winter range controlled burning program. The habitat program begun with the assistance of the RMEF will have long-term benefits for elk and other wildlife in HD 123 and 200. 4. We recommend that documented migratory corridors used by elk be recognized as special management areas in the course of ongoing resource development planning. Movement corridors over mJi^r ridges and saddles should be maintained with as little future disturbance as possible. During this 5-year study of 9 elk herd units, we documented several elk migration routes and corridors over topographical divides. Management planning in the 50,000 acre Mt. Bushnell roadless area used this information to protect important migration routes used by elk. We encourage similar recognition and protection of major elk travel routes throughout the study area. There is a need for a security strategy that addresses special problems that arise when bulls are concentrated in a narrow zone during hunting season, especially where the habitat is fragmented by timber harvest. The "Hillis Paradigm" may not apply 112 migration corridors, where special measures (ie. higher than normal levels of cover, additional access restrictions, or special regulations) may be required. Public Relationships 1. We recommend that MDFWP and LNF maintain a close working relationship with the people of Sanders and Mineral counties regarding the management of elk and elk related recreation. The success and public acceptance of a long-term elk management plan depends on clear two-way communication between resource agencies and the interested public, to include sportsmen's clubs, landowners, commercial outfitters, commodity interest groups, local businesses and government groups, non-consumptive users, off -road vehicle/horse groups and others. Public meetings in Thompson Falls, Plains, and Superior hosted by MDFWP and LNF could periodically update interested parties on current elk population suirveys, harvest information, habitat improvement projects, quota recommendations, travel planning changes and timber management activities related to elk. Public meetings focused on elk would allow for public input regarding elk management in their own "backyard". To be more effective, there should be good media coverage prior to and following such meetings. 113 LITERATURE CITED Anonymous. 1976. Montana big game trapping and transplant record 1910-1975. MT Dept. Fish and Game. Helena. 13pp. Burcham, M. 1990. Winter habitat selection by elk in northwestern Montana. M.S. Thesis. Univ. of Mont. Missoula. 46pp. Cada, J.D. 1987. Statewide wildlife harvest surveys, Proj . No. W- 104-R-23, Job. No.l, MT Dept. Fish, Wildlife, and Parks, Helena. 23pp. Canfield, J.E. 1988. Impact mitigation and monitoring of the BPA 500-kv Garrison-Taft transmission lion — effects on elk security and hunter opportunity ; final report. MT Dept. of Fish, Wildlife and Parks, Helena. 162pp. Caughley, G. 1974. Bias in aerial survey. J. Wildl. Manage. 38:921- 933 . ^ R. Sinclair, and D. Scott-Kemmis . 1976. Experiments in aerial survey. J. Wildl. Manage. 40:290-3 00. Chapman, D.G. 1952. Inverse, multiple, and sequential sampl^ censuses. Biometrics 8 (4) : 286-306 . ^ Duf field, J. 1988. The net economic value of elk hunting in Montana. MT Dept. Fish, Wildl. and Parks. Helena. Edge, W.D., C.L. Marcum, S.L. Olsen, and J.F. Lehmkuhl. 1986. Nonmigratory cow elk herd ranges as management units . J . Wildl. Manage. 50:660-663. Geist V. 1982. Adaptive behavioral strategies. In Elk of North America-ecology and management, ed. J.W. Thomas and D.E. Toweill, pp. 219-277. Washington D.C. Wildlife Management Institute. 698pp. Hammond, G., B. Sterling, and M. Thompson. 1985. Impact mitigation and monitoring of the BPA 500-kv Garrison-Taft transmission line—effects on elk security and hunter opportunity; annual progress report for 1984. MT Dept. Fish, Wildl. and Parks, Helena. 71pp. Henderson, R.E. 1990. Preganacy rates and dates of conception based on reproductive tracts collected from cow elk in hunting districts 121 and 202. Unpublished report. MDFWP files. Missoula. 3pp. 114 Hillis, J.M. , M.J. Thompson, J.E. Canfield, L.J. Lyon, C.L. Marcum, P.M. Dolan, D.W, McCleery. 1991. Defining elk security: the Hillis paradigm, p. 38-43 in L.J. Lyon ed. Proc. elk vulnerability symposium. Mont. State Univ. Bozeman. 330pp. Hurley, M.A. 1987. Relationships between DAPA, selected blood parameters, and other indices for evaluating the nutritional condition of deer. Senior Thesis, Univ. of MT, Missoula. 22pp. , M.A. and G.A. Sargent. 1991. Effects of hunting and land management on elk habitat use, movement patterns, and mortality in western Montana, p. 94-98 in L.J. Lyon ed. Proc. elk vulnerability symposium. Mont. State Univ. , Bozeman. 3 3 0pp. Irwin, L.L. and J.M. Peek. 1979. Relationship between road closures and elk behavior in northern Idaho, pp 199-2 04 in M.S. Boyce and L.D. Hayden-Wing, eds. North American elk: ecology, behavior and management. Univ. Wyoming. Laramie. , and . 1983. Elk habitat use relative to forest succession in Idaho. J. Wildl. Manage. 47:664-672. Kistner, T.P., K.R. Greer, D.E. Worley, and O.A. Brunetti. 1982. Diseases and Parasites, pp 181-217 in J.W. Thomas and D.E. Toweill, eds. Elk of North America: ecology and management. Stackpole. Harrisburg, PA. Koch, E. 1941. Big game in Montana from early historical records. J. Wildl. Manage. 5(4) :357-370. Leege, T.A. and W.O. Hickey. 1971. Sprouting of northern Idaho shrubs after prescribed burning. J. Wildl. Manage. 35(3) :508- 515. Lemke, T.and R.E. Henderson. 1987. Lower Clark Fork elk study. Biennial progress report, 1985-1987. MT Dept. Fish, Wildlife and Parks, Helena. 58pp. Leptich, D.J, and P. Zager. 1991. Road access management effects on elk mortality and population dynamics, p. 126-131 in L.J. Lyon ed. Proc. elk vulnerability symposium. Mont. State Univ., Bozeman. 3 3 0pp. Lonner, T.N. and J.D. Cada. 1982. Some effects of forest management on elk hunting opportunity. Proc. 1982 West. States Elk workshop. Phoenix, AZ. pp 119-128. Lyon, L.J. 1979. Habitat effectiveness for elk as influenced by roads and cover. J. For. 77:658-660. 115 Mace, R. 1992. Elk response to forest road densities in wester^ Montana. Unpublished report. 9pp. Marcum, C.L. 1975. Summer-Fall habitat selection and use by a Western Montana elk herd. Ph.D. Thesis, Univ. Mont., Missoula. 188pp. McDowell, L.E. 1949. Bitterroot-Clark Fork Survey, Big Game Project Reports, Western Montana. MT Fish and Game Dept. Helena. 51pp. Montana Department of Fish, Wildlife and Parks. 1992. Montana Statewide Elk Plan. Helena. Morgantini, L.E. and R.J. Hudson. 1979. Human disturbance and habitat selection in elk. pp 132-139 in M.S. Boyce and L.D. Hayden-Wing, eds. North American elk: ecology, behavior and management. Univ. Wyoming. Laramie. Perry, C. and R. Overly. 1976. Impacts of roads on big game distribution in portions of the Blue Mountains of Washington, p. 62-68 in S.R. Hieg, ed. Proc. Elk-logging roads symposium. Univ. Ida., Moscow. 142pp. Rice, W.R. and J.D. Harder. 1977. Application of multiple aerial sampling to a mark-recapture census of white-tailed deer. J. Wildl. Manage. 41:197-206. ' Rognrud, M. 1948. Clark Fork Unit-Inspection of the Cherry Creek Game Preserve. MDFWP files, Missoula. 2pp. M. 1950. Census and distribution of elk in the Cherrv U 1-1 ■t^T-,TT^T.TT^ , _ Creek Game Preserve. MDFWP files, Missoula. 8pp ., M. and R. Janson. 1971. Elk. pp 39-51 in T.W. Mussehl and F.W. Howell eds. Game management in Montana. MT Fish and Game Dept. Helena. 238pp. Samuel, M.D., E.O. Garton, M.W. Schlegel, and R. G. Carson. 1987. Visibility bias during aerial surveys of elk in northcentral Idaho. J. Wildl. Manage. 51:622-63 0. Skovlin, J.M. 1982. Habitat requirements and evaluations, eds. J.W. _ Thomas and D.E. Toweill. pp 369-413 in elk of North America-ecology and management. Washington D.C. Wildl. Manage . Inst . Thompson, M.J. and B. Sterling. 198 6. Impact mitigation and monitoring of the BPA 500-kv Garrison-Taft transmission line- effects on elk security and hunter opportunity; annual progress report for 1985. MT Dept. Fish, Wildl. and Parks ,< Helena. 70pp. 116 andD.C. Dickson. 1987 . Impact mitigation and monitoring of the BPA 500-kv Garrison-Taft transmission line — effects on elk security and hunter opportunity; annual progress report for 1986. MT Dept. Fish, Wildl . and Parks, Helena. 80pp. ,R.E. Henderson, T.O. Lemke, and B.A. Sterling. 1989. Evaluation of a collapsible clover trap for elk. Wildl. Soc. Bull. 17:287-290. Unsworth, J.W. , L. Kuck, and E.O Garton. 1990. Elk sightability model validation at the National Bison Range, Montana. Wildl. Soc. Bull. 18:113-115. , F.A. Leban, G.A. Sargeant, E.O. Garton, M.A. Hurley, and J.R. Pope. 19 91. Aerial survey: user's manual. Idaho Dept. Fish and Game, Boise. 54pp. US Department of Commerce. 1980. Climatological Summary. National Oceanic and Atmospheric Admin- Asheville, NC. USDA Forest Service. 1985. Lolo National Forest Plan. Lolo National Forest, Missoula, MT. Vore J. and R. DeSimone. 1991. Effects of an innovative hunting regulation on elk populations and hunter attitudes, p. 23-29 in L.J.Lyon ed. Proc. elk vulnerability symposium. Mont. State Univ. Bozeman. 3 3 0pp. Warner, R. 1970. Some aspects of browse production in relation to timber harvest methods and succession in western Montana. Master's Thesis, UM, Missoula. 74pp. 117 oo ^. , ^ APPENDICES Appendix 1. CAPTURE AND KNOWN FATE DATA FROM ELK TRAPPED IN HD 123, 1986-89. DATE TRAP SITE SEX AGE FREO. COLLAR COLOR /SYMBOL EAR EARTAG STREAME RLOOn NUMBER R SAMPr p COLOR FATE 01-10-86 M IVl call 150.826 WPVC 20621/22 Blue - Shot 01-19-86 Di ubn vjuicii r 3-10 151.563 WPVC 60233/34 Blue - Unknown r 3-10 151.045 WPVC 20645/47 Blue - Winter Kill i-V/ Ovi r 2 151.012 WPVC 21545/46 Blue - Unknown AO 00 Dry Creek F 10+ 150.782 WPVC 21535/36 Blue - Unknown v/^ — ^j — oo rsartn s Kancn M Calf Blue/ •/ •/ • 21532/33 Blue - Unknown 02-26-86 T - r 3-10 151.575 WPVC 21530/47 Blue - Unknown 02-26-86 Rarfh'Q Rnnr>fi ual 111 a lVallL/11 i; r 1 5 1 .664 WPVC 21548/49 Blue - Unknown 03-01-86 Oilccp oap t? r 3-10 151.673 WPVC 21514/15 Blue Shot xJartn s Kancn "I— • F 10+ 150.039 WPVC 21507/08 Blue Shot 03-13-86 Therriault Gul. F 2 150.157 WPVC 21518/20 Blue + Idaho Shot 03-22-86 Therriault Gul. F 3-10 150.620 WPVC None Blue + Unknown 03-23-86 Sheep Gap F 3-10 150.058 WPVC 20688/89 Blue + Shot 03-24-86 Sheep Gap F 10+ 151.270 WPVC 20686/87 Blue + Unknown 03-25-86 Webster's F 3-10 150.220 WPVC 20676/77 Blue + Poached 03- 28-86 Earth's Ranch F 04- 03-86 Webster's F 04-04-86 Brush Gulch M 04-08-86 Webster's F 04-09-86 Webster's M 04-10-86 Earth's Ranch M 04-15-86 Webster's F 04-20-86 Webster's F 04-21-86 Shamrock Gulch F 04-22-86 McCrea's Ranch F 04-23-86 Websters F H 04-24-86 Shamrock Gulch F ID 12-18-86 Clark Mtn. M 01-11-87 Sheep Gap F 01-12-87 Clark Mtn. F 01-16-87 Sheep Gap F 01-18-87 Clark Mtn. F 01-24-87 Clark Mtn. F 01-30-87 Dry Creek F 3-10 150.240 WPVC 1 150.924 WPVC Calf — Blue/ 1 150.965 WPVC Calf — Blue/.... Calf — Elue/MMM Calf 150.360 WPVC Calf 150.864 WPVC 3-10 151.060 WPVC 2 150.501 WPVC Calf ~ ORANGE/" 1 150.198 WPVC 2 150.700 WPVC 3-10 150.995 WPVC 3-10 150.333 WPVC 2 150.012 WPVC 3-10 150.459 WPVC 2 150.181 WPVC 3-10 150.845 WPVC 20678/79 Blue w llivllU Wll 20690/91 Blue + Shot 20665/66 None + Unknown 20667/68 Blue + Shot 20699/70 None + Shot 20671/72 None + Shot (FIR) 20652 Blue + Unknown 20655 Blue + Unknown 20658/59 Blue + Shot 20653/62 Blue Shot 20656/57 None + Unknown 21551/52 Blue Poached 21566/67 Red Shot 21597/98 Yellow + Shot 21564 Yellow Unknown 20571/73 Yellow + Unknown 20561/62 Yellow Shot 20559/60 Yellow + Unknown 20553/54 Yellow Unknown 01-30-87 Clark Mtn. F 3-10 151.224 01-31-87 Sheep Gap M 1 151.034 01- 31-87 Therriault Gul. M Calf 150.370 02- 02-87 Sheep Gap M Calf 150.733 02-02-87 Therriault Gul. F 2 150.017 02-05-87 Clark Mtn. F 3-10 150.578 02-08-87 Clark Mtn. F 1 150.091 02-09-87 Clark Mtn. F 10+ 02-11-87 Dry Creek F 3-10 150.914 02-21-87 Eddy Creek F 3-10 151.180 02- 28-87 Clark Mtn. F 2 150.637 03- 03-87 Eddy Creek F 2 150.292 03-06-87 Therriault Gul. F 1 151.067 03-19-87 Eddy Creek F 1 151.062 03-22-87 Clark Mtn. F 3-10 01-07-88 Wilkes Creek F 3-10 150.312 01-08-88 Wilkes Creek M 1 150.890 01-09-88 Wilkes Creek M I 150.502 01-10-88 Wilkes Creek F 2 150.349 WPVC 20551/52 Yellow + Unknown WPVC 21576/77 Red Shot WPVC 21578/79 Red Shot WPVC 21582/83 Yellow Shot WPVC 20536/37 Yellow + Shot WPVC 20534/35 Yellow Unknown WPVC 20532/33 Yellow + Unknown YELLOW/ ■■■ 20530/31 Yellow + Unknown WPVC 20528/29 Yellow + Unknown WPVC 20547/48 Yellow Unknown WPVC 20584/87 Yellow + Unknown WPVC 20517/18 Yellow +Shot WPVC 20521/22 Yellow Shot WPVC 20511/12 Yellow + Unknown YELLOW/XXX 21436/37 Yellow + Unknown WPVC 21472/73 Unknown + Unknown WPVC 21489/90 Unknown Shot WPVC 21476/77 Unknown Shot WPVC 21463/64 Unknown + Unknown 01 -14- -88 Wilkes Creek F 3-10 150.059 WPVC 21465/66 Unknown Unknown 01 -19- -88 Wilkes Creek F 3-10 150.966 WPVC 21438/39 Unknown + Unknown 01 -21- -88 Wilkes Creek F 2 _ _ ORANGE/ 21443/44 Unknown Unknown 01 -21- 88 Wilkes Creek F Calf ORANGE/o 0 0 21445/46 Unknown Shot 01 -26- -88 Wilkes Creek F Calf ^ _ ORANGE/ 21491/92 Unknown Unknown 02 -04- -88 Kraak's F Calf 151.046 WPVC 21453/54 Unknown Shot 02 -04- ■88 Clear Creek M Calf 150.996 WPVC 21451/52 Yellow Mt. Lion Kill 02 -05- -88 Clear Creek F 1 150.660 WPVC 21455/56 Yellow + Unknown 01 -18- -89 Miller Creek F Calf 150.137 WPVC 21467/68 Yellow Unknown 01 -20- -89 Miller Creek F 3-10 150.734 WPVC 21563 None Unknown Ul -20- on -89 Miller Creek r J- 1 u 1 JU.oZ / YY r V INUIIC T T n n r\ \n n \J iiivUU WU 01 -21- -89 Miller Creek F 2 150.997 WPVC 21459/60 Orange Unknown 01 -25- -89 Miller Creek M Calf 150.925 WPVC 21461/62 Red Shot (REF: 5123.BS) WPVC = White PVC pipe collar -f = Blood Sample Collected Age = At capture Fate = Through June 1991 Appendix 2. CAPTURE AND KNOWN FATE DATA FROM ELK TRAPPED IN HD 200, 1986-88. EAR DATE TRAP SITE SEX AGE FREO. COLLAR COLOR /SYMBOL EAR TAG NUMBER STREAME RCOLOR BLOOD SAMPLE FATE 01-10-86 Boyd M Calf 150.562 WPVC 20801-02 Blue Shot 01-14-86 Boyd F 1 150.475 WPVC 20803 Blue - Unknown 01-14-86 Mayo 1 F 1 150.716 WPVC - Unknown V/ 1 1 ^ o\j nTamarafk' 9 F 2 150.437 WPVC 20815 Blue Unknown V I- 1 0 — oO Duyu M 1 RR w/hlack har"! 20805-06 Unknown M Calf 150.103 WPVC w/blue Stripes 20810-1 1 Blue Roadkill 01-16-86 Boyd F 3-10 - OB w/open black 20822-23 - - Shot A 1 to 0/C Tamarack 2 r J- lU 1 JU.HVH Wr V zuo 1 y jjjue ollul 02-01-86 Dry Fork 1^ r J-IU I ju.ouo Wr VC ZUOJJ- JO ijiue wounaingioss 02-08-86 Dry Fork F 2 151.489 WPVC 20832-31 Blue Shot 02-08-86 Dry Fork F 3-10 151.098 WPVC Poached 02-12-86 Boyd F 3-10 150.877 WPVC 20829-30 Blue Unknown 02-12-86 Boyd F 1 OB w/black bars 20827-28 Blue Unknown 02-13-86 Dry Fork F 3-10 150.542 WPVC 20883-84 Blue + Unknown 02- -14- -86 Dry Fork 02- -19- 86 Dry Fork 02- 19- 86 Tamarack 2 02- -19- 86 Keith 02- 19- ■86 Boyd 02- -20- ■86 Dry Fork 02- -21- ■86 Keith 03- -04- -86 Dry Fork 03- -06- -86 Mayo 2 03- -06- -86 Boyd 03- 10- -86 Mayo 2 03- -10- ■86 Dry Fork 03- -15- -86 Dry Fork 03- -20- ■86 Boyd 03- -25- 86 Dry Fork 03- ■27- 86 Tamarack 2 01- -07- 87 Boyd 01- -07- ■87 Boyd 01- ■07- 87 Boyd F Calf 151.467 F 1 151.438 F 3-10 150.765 F 8-10 151.453 F 10+ F Calf M 1 151-301 F 3-10 150.429 F 3-10 150.120 F 3-10 150.080 F 10-h 150.520 F 10+ M 1 151.360 F 3-10 F 1 150.280 F 1 F Calf F 3-10 151-425 F 3-10 151.253 WPVC & blue band WPVC & blue band w/black stars WPVC WPVC OB w/black H's OB w/solid black circles WPVC w/blue stripe WPVC WPVC WPVC WPVC OB w/black ='s WPVC w/blue stripe OB w/black bull WPVC OB w/black X's YB w/Black T WPVC WPVC 20877-78 20885-86 20880-82 20887-79 20837-38 20839-40 20845-47 20846 20843-89 20890-91 20895-94 20900 20855 20860-61 20872-74 20865-66 20446-47 20444-45 20464-65 Blue Blue Blue Blue Blue Blue Blue Blue Blue ■f Yellow Yellow Shot Unknown Unknown Roadkill Unknown Unknown Unknown Shot Shot Unknown Unknown Trap Mort. Unknown Unknown Unknown Unknown Unknown Unknown Unknown 01-08-87 Dry Fork F Calf 01-21-87 Boyd F 3-10 151.626 01-21-87 Boyd F 2 01-21-87 Boyd M Calf 150.103 01-22-87 Tamarack M Calf 01-23-87 Tamarack F 3-10 151.157 01-24-87 Tamarack F 1 — 01-30-87 Wolf F 3-10 150.318 02-04-87 Camels Hump F 6-8 151.655 02-05-87 Dry Fork M 2 150.136 02-05-87 Camels Hump F 3-10 150.337 02-17-87 Camels Hump F 10+ 151.554 02-19-87 Wolf 2 F 3-10 151.739 m on 87 rioya E" r i-lu 1 JU.OJJ 02-20-87 Boyd F 3-10 m on ftT xJOyQ \A M 1 Kf\ f,1 A 1 jU.O /4 02-20-87 Boyd F 10+ 02-24-87 Wolf 2 F Calf 02-26-87 Mayo 5 F 3-10 150.039 YB w/black bars WPVC YB w/black + WPVC RB w/white + WPVC YB w/black • WPVC WPVC WPVC WPVC WPVC WPVC WPVC YB w/black moons WPVC w/red stripe YB w/black solid diamonds YB w/black arrows WPVC 20436-477 20443-42 20460-61 20440-41 20456-57 20454-88 20458-59 20705-06 20702-01 20474-75 20433-34 20466.67 20788-89 20432-69 20784-85 20782-83 Yellow Yellow Yellow Red Red Yellow Yellows Yellow Yellow Yellow Yellow Yellow Yellow Yellow Red Yellow + + + + + 20755-56 Yellow 20757-58 Yellow Unknown Unknown Unknown Unknown Unknown Unknown Unknown Unknown Shot Shot Unknown Woundingloss Unknown Shot Shot Woundingloss Shot Unknown Unknown AO uZ-Z 1-0 / iviayo J r 1-10 03-03-87 Dry Fork F 3-10 - 03-10-87 Mayo 5 F 3-10 - 03-11-87 Camels Hump 3 F 3-10 151.194 03-20-87 Mayo 5 F 3-10 - 03-21-87 Boyd F 10+ - 03-21-87 Camels Hump F 10+ 151.565 03-24-87 Mayo 5 F 3-10 - 03-25-87 Mayo 5 F 6-8 - 03-26-87 Dry Fork M 1 151.360 03-26-87 Mayo 5 F 10+ - 03-27-87 Mayo 5 F 3-10 - 03-29-87 Dry Fork F 3-10 150.844 01-12-88 Dry Fork F Calf 01-13-88 Camels Hump 4 F 6-8 01-13-88 Dry Fork M Calf 150.078 01-14-88 Dry Fork F Calf 01-20-88 Tamarack 2 F 1 I r> w/ uiaOK upcii diamonds ■60 \J\j Yellow TTnlcnnwn YB w/black S 20761- 62 Yellow + Unknown YB w/black P 20765- 66 Yellow + Shot WPVC 20780- 81 Yellow + Unknown YB w/black + 20703- 04 Yellow + Unknown YB w/black ./././ 20773- -74 Yellow + Unknown WPVC 20767- -68 Yellow - Unknown YB w/black 0 20463 Yellow + Unknown YB w/black » 20769- -70 Yellow + Unknown WPVC w/red stripe 20771- -72 Red + Shot YB w/black ZZ 20751- -52 Yellow + Unknown YB w/black V|V| 22942 -43 Yellow - Shot WPVC 20792- -93 Yellow - Unknown Rpd w/white 2 20720- -21 Orange Unknown Red w/white * cioseu triangles 22948 Z J 1 4^ Orange Unknown WPVC w/red zigzag 20717 20764 Orange - Unknown Orange w/white ■ closed rectangle 20794- -5 Orange Unknown Orange w/white stars & bars 20724 20800 Orange + Unknown 01-20-88 01-:!0-88 01-20-88 01-:.6-88 01-27-88 01-27-88 01- 29-88 ^ 01-30-88 cn 02- 02-88 02-02-88 02-03-88 02-03-88 02-04-88 02-04-88 Dry Fork Mayo 5 Camels Hump 4 Dry Fork Dry Fork Camels Hump 4 Dry Fork Dry Fork Dry Fork Dry Fork Mayo 5 Camels Hump 5 Dry Fork Camels Hump 4 F 6-7 F 4-5 F 3 F Calf F Calf F 3 F 1-2 F 2-3 F 3-4 F 3-4 F 3-4 F 3-10 M Calf F 3-10 Red w/white 3 20725 22990 Orange + Unknown Blue w/white stars & bars T*r| 22991-2 Orange + Unknown Orange w/white zigzag 22988-9 Orange + Unknown Orange w/white squiggle 22995-6 Orange - Unknown Orange w/white zeros O 22999 23000 Orange Unknown Orange w/white □ open rectangle 22976 22977 Orange + Unknown Orange w/white M Orange w/white bar & slash -/-/ 22985 22986 22978 22979 Orange Orange + Unknown Unknown Orange w/white stripe 22912 22980 Orange + Unknown Orange w/white bars III 22983-84 Orange Unknown Orange w/white chevrons & dots <.> 22927-28 Orange + Unknown Red w/white arrows - Orange Unknown WPVC 22931-33 Orange Unknown Re^lp/white p>.05 15.0 Appendix 7. Comparison of harvest rates for pooled samples of radio- tagged bulls in HD 123 and HD 200 during the 1984-1990 hunting seasons. Observed and expected values. AGE CLASS (YRS) OBSERVED VALUES SURVIVED HARVESTED TOTAL EXPECTED VALUES SURVIVED HARVESTED 1.5 12 15 8.9 6.1 10 12 22 13-1 8.9 TOTAL 22 15 37 22 15 X 4.41, df=l, .05>p>.025 Appendix 8. Comparison of harvest rates for 3 age-classes of radio- tagged bulls in the study area during 1984-1990 hunting seasons, Observed and expected values. 136 OBSERVED VALUES TOTAL EXPECTED VALUES CTTRVIVED HARVESTED SURVIVED HARVESTED HD 12 3 (ALL) 1 o 17 95 7Q R ±0 • ^ HD 200 74 12 86 72.2 13 . 8 TOTAL 152 29 181 152 . 0 29 HD 123 (WEST) 42 10 52 43.7 8.3 HD 200 74 12 86 72.3 13 .7 TOTAL 116 22 138 116 22 HD 123(ALL) VS. HD 2 OO-X'^rO . 52 , df=l, .50>p>.10 HD 123 (WEST) VS. HD 200-X =0.67, df=l, .50>p>.10 Appendix 9 . Comparisons of harvest rates of cows radio-tagged in HD 123 and HD 200 during the 1984-1990 hunting seasons. Observed and expected values. OBSERVED VALUES TOTAL EXPECTE D VALUES SURVIVED HARVESTED SURVIVED HARVESTED EAST 36 7 43 35.3 7.7 WEST 42 10 52 42.7 9.3 TOTAL 78 17 95 78 17 X'-=0.13, df=l, .9 0>p>.50 Appendix 10. Comparison of harvest rates for pooled samples of cows radio-tagged in permitted (East) and in either-sex (West) portions of HD 123 during the 1984-1990 hunting seasons. 137 DECISION NOTICE * P P 8 n d i x 11 FINDING OF NO SIGNIFICANT IMPACT MOUNT BUSHNELL USDA, FOREST SERVICE PLAINS/THOMPSON FALLS AND SUPERIOR RANGER DISTRICTS LOLO NATIONAL FOREST It is my decision during the next 10 year period to construct up to 8l miles oi road, reconstruct approximately 9 miles of existing road and initiate tSber management activities on an estimated 3,200 acres. initiate txmber Based upon an analysis of cumulative effects by an Inte-discinlinarv T^.m ' composed of Lolo Forest Officials and Wildlife BioloSs-f^rom thrM t lllTZT °f l'^'^' '''''''' '^^'^ (BFW^K tSs^'dLlsLTiip eSn s"^L Lolo National Forest Plan on National Forest land^ hot-h n^^^K "^pxemencs cne Cabinet-Couer-d Alene Divide from Gold Rusfcie^fto he Eas't F^rk'of LI Pac£r°? t Pl--/Tl^ompson Falls Ranger District and from ^„In CreLs to Packer Creek on the Superior Ranger District. ^^-eeKs zo Tlie Interdisciplinary Teams' analysis encompassed approximately '50 330 acres Thxs decision adaresses actions connected with timber harvest activitLf on ' and a variety of harvest treatments including clearcut se^dtrP^ IhJll \, and selection silvicultural systems during tL n:;^?^^"^ pS^d ^Sis a^L will provide a variety of forest products for local use. The Management Areas contained within the study area as definpri t iIP Qth^r^K 1:10,19 and 27 are represented which reflect resource valuP^ l^Z x^a™r:^1rSc:! £f - --It Of habitafinvStSL reflB^; fh= » critical elk suMiet- range were refined and mapoed to N^So^niitrSrpTa^ isrpo'rLrBS'jrAprSiosi'sr ^= " implement the standards and guidellnL o? tje ForesfpiS?''"^'' '""""^ PUBLIC ISSUES AND Cn^JCKH^J■9 oove'rTf^iheltS; ^T' "^'^ '"^ vegetative quSi" °P!>°"™"ie= provided are generally perceived to be of high vary between Si ReSns tte^e ar^SL'^' S"""?^ "^^^ regulations and guides active within hhf',.^ ? ° * "'^ commercial outfitters .on-f^sidLts fir r™r?eS'of uses'"' ""^"^ ^^""^ '° 138 DECISION NOTICE 2 A significant potential exists for large scale value and volume loss due to the southerly migration of the mountain pine beetle epidemic in northwestern Montana. Some of the public issues and management concerns identified in the Mt. Bushnell analysis are listed below: 1. How will the cumulative impacts of timber management activities in the area affect security cover and long term harvest rates of elk? 2. How will timber management activities in the area affect the productivity of elk summer and winter range? 3- How will road construction and timber harvest within the area affect traditional regional values, community feelings and local lifestyles? k. How can the area be managed to maintain the current quality of recreational use? 5. How can the area be managed to accommodate the full range of DFWSP objectives? 6. How will resource development impact the currently established commercial interests in the area? 7. How can the lodgepole pine stands be managed to reduce the impacts of large scale mountain pine beetle epidemics? Alternatives Considered: Should the mountain pine beetle epidemic enter the study area, analysis and recent history of the epidemic indicate that the volume and value loss cannot be avoided. Accelerated harvest and road construction schedules that could salvage this material would create cumulative impacts on watershed, elk productivity and recreation that would violate Forest Plan standards. Based upon the issues and concerns identified through a public involvement process, the Interdisciplinary Team developed a range of alternative management strategies that were responsive to each. Each alternative addressed the issues and concerns and meets the objectives of the Lolo National Forest Plan and those defined by the team specific to the Mt. Bushnell study area. During the analysis process, alternatives that were developed which did not meet Forest Plan standards were eliminated and received no further consideration . Following is a descriptive list of the alternatives considered, but not selected: Alternative 1 maintains recreational use in its present condition as long as possible. Harvest units will be selected to reduce risk ox mountain pine beetle loss but with minimum impacts to recreation and commercial interests. Reentry into initial project areas will be on a 15 year frequency. 139 DECISION NOTICE 3 Alternative 2 has an objective to maintain security/cover and productivit^Bjr elk. Rate of timber harvest is commensurate with the Lolo National Fores^^ assigned sell quantity (ASQ) . Alternative 2B is the same as alternative 2 except the rate of harvest is reduced by 23% to reflect the current situation on the Forest. Alternative 2N defers timber resource development activities for the 10 year decision period. This alternative is the No Action alternative. Alternative 6 places an emphasis on treatment of lodgepole stands with harvest rates accelerated in the near term for mountain pine beetle risk reduction. Dispersed recreation and elk productivity are provided for at minimal levels. However, escape cover of each herd unit will be maintained. Alternative 6B has the same objective as alternative 6 with a scheduled rate of harvest on a even- flow basis throughout the planning period. Alternative 11 schedules timber harvest at a rate similar to 6B while emphasizing the need for long term elk security/cover. RATIONAL FOR THE DECISION I chose to implement alternative 12 because recreational opportunities, cover/forage ratios and substantial amounts of security cover are maintained or enhanced in every drainage of the anauLysis area. |B Road management provides for adequate big game security needs by incorporating a closed system' to motorized vehicles. Road systems are designed for within drainage flow patterns to minimize disturbance levels and impacts to adjacent security cover areas. Impacts on commercial Outfitters and Guides are minimized during the 10 year period of this decision. This alternative maintains the full range of herd management options within the analysis area and allows for changing conditions and DFWiP's input. While significant loss of lodgepole pine to the beetle is unavoidable, this alternative will initiate harvest activities in moderate and high hazard areas along the perimeter of the study area. Under Alternative 12, resource development in the Mt. Bushnell area reflects the management standards and goals of the Lolo National Forest Plan. The cumulative effects expected by implementation of this alternative will not significantly impact the existing resource values in the analysis area. This alternative accepts additional loss of lodgepole pine value and results in a moderate reduction in total planned harvest volume during the 10 year decision period. However, expected benefits include improvement of cover/ forage ratios, enhancement of elk productivity, incorporation of DFW^^ objectives and maintaining quality recreational opportunities. 140 DECISION NOTICE Within the 10 year decision time-frame, development activities in proiert area<, idLti? ^-^1- at?a?hed m'ap Identify project area locations. 5 Yr. Est. Acres Project Area Period Treatment I 1 182 II 1 305 III 2 452 IV 2 294 V 1 228 VI 1 400 VII 2 309 VIII 1 309 IX 2 244 X 2 360 Est.MBF Est. Mi. Est.Mi.Rd Volume Const. Reconst . 2,184 3.3 ♦ 3,355 6.8 4.972 11.5 2.0 3.234 5.7 2.964 11.1 6.223 12.5 » 3,708 7.4 3,984 8.2 3.5 2,960 5-9 3.0 4.866 8.6 * Reconstruction needs for these project areas are outside analysis boundary. Considering the above analysis, I have found this to be a non-simificant actxon and not requiring a environmental impact statement. ■. This decision will have no effect on flood plains, wetlands, threatened or endangered speciSs lrlT''^/.^'°''l'=^ inventories and evaluation will be conductef prior project disturbance activity. Proper protection or mitigation aSions wSI be J^^?. The project w?ll havj no sig^iJic^t impacts on the quality of the human environment. sxgniticant The project file may be reviewed at the Plains/Thompson Falls Ranger District llT"' f 59859 or the Superior Ranger District, Box 460. SupeSor S 59872* SticT ?^is"defi.T ' ""'''h' t-^- -Pon the date o? this'D^is ol 2n la' " subject to a 45 day review period pursuant to 36 CFR ^11. la. Questions concerning this project may be addressed to the Forest Supervisor. Lolo National Forest. Building 24. Fort Missoula MissoSa MT ORVILLE L. DANIELS DATE Forest Supervisor 140 . v.-:V ■ l-r- '■' ' ■ r/X-^v /-'V^^ /■ M •.^/^^TT^r^^^-) — Lo i ,. • • p • ci3,h „ ■ ci3ih ill ■ , . , ,-■ MOUNT BUSHNELL Alternative 12 Environmentaf Analysis Ten year projecl area locaWons and eludy area vicinity map I f -,V':<-:. I ' '// ' r -.1- -I * -ri\' .''-1 ^- . •