Ecology of a prairie mule deer population by Alan Keith Wood A thesis submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Biological Sciences Montana State University © Copyright by Alan Keith Wood (1987) Abstract: The ecology of a mule deer (Odocoileus hemionus) population inhabiting an eastern Montana prairie, was studied during 1975-1986, using periodic aerial surveys and radioed and neckbanded deer. Spring population size increased by 29%/year from 1976 to 1983, stabilized at approximately 1,080 deer during 1983-1984, and declined 61% during 1984-1986. An average of 42% of the adult females successfully weaned > 1 fawn, resulting in autumn, winter, and spring fawn:doe ratios averaging 56:100, 55:100, and 35:100, respectively. Annual mortality rates for fawns, adult females, and adult males averaged 70%, 23%, and 63%, respectively; harvest rates averaged 3%, 19%, and 59%, respectively. Deer distribution was highly aggregated across the area. Mixed groups of males and females were uncommon, though distributions of males and females overlapped significantly. Individual females exhibited 1 of 3 movement patterns: yearlong residency, spring and autumn migration, or late summer and late autumn migration. Annual home ranges averaged 4.4 km2. Females and mature, males demonstrated a high degree of fidelity to traditional herd ranges,, but yearling males did not. Areas of topographic diversity in winter, and mesic sites in summer were preferred. Interspersion of habitat complexes influenced both distribution and home range size. Environmental variability caused specific responses by mule deer, and was reflected in variable fawn recruitment rates. Population regulation seemed to be influenced by the limited availability of high,quality forage during late summer. Spring population size of white-tailed deer (0. virginianus) peaked at 335 deer. Population dynamics of both species were similar, except during 1984, when 60% of the female whitetails were harvested. Significant interspecific overlap occurred, and the 2 species demonstrated no intrinsic avoidance mechanisms. Whitetailed deer preferred areas with overhead cover adjacent to agricultural fields, used larger annual home ranges (9.98 km2) and showed much less home range fidelity than mule deer. Pronghorn antelope (Antilocapra americana) occurred on the area at densities similar to white-tailed deer, used open habitats, and had little impact on the mule deer population.  ECOLOGY OF A PRAIRIE MULE DEER POPULATION by Alan Keith Wood A thesis submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Biological Sciences MONTANA STATE UNIVERSITY Bozeman, Montana April, 1987 031% APPROVAL of a thesis submitted by Alan Keith Wood This thesis has been read by each member of the thesis committee and has been found to be satisfactory regarding content, English usage, format, citations, bibliographic style, and consistency, and is ready for submission to the College of Graduate Studies. /? OteMLlDate (/ Committee Approved for the M̂tvj or Department /6 Date Head, Major Department Approved for the Q-r 7- S zI College of Graduate Studies Dat e Graduate Dean ill STATEMENT OF PERMISSION TO USE In presenting this thesis in partial fulfillment of the requirements for a doctoral degree at Montana State University, I agree that the Library shall make it available to borrowers under rules of the Library. I further agree that copying of this thesis is allowable only for scholarly purposes, consistent with "fair use" as prescribed in the U.S. Copyright Law. Requests for extensive copying or reproduction of this thesis should be referred to University Microfilms International, 300 North Zeeb Road, Ann Arbor, Michigan 48106, to whom I have granted "the exclusive right to reproduce and distribute copies of the dissertation in and from microfilm and the right to reproduce and distribute by abstract in any format". Date / 3. V ACKNOWLEDGEMENTS The assistance of numerous people from Montana State University, Montana Department of Fish, Wildlife, and Parks, Bureau of Land Management, Livestock and Range Research Station, and many private landowners is gratefully acknowledged. R. J. Mackie directed and assisted in all phases of the project. G. L. Dusek, K. L. Hamlin, and D. : ' ' v F. Pac provided helpful suggestions during analysis and writing, K. L. Hamlin also allowed the use of his unpublished 1975-1980 data. H. E.. Jorgensen assisted in vegetation sampling and calculations of winter severity. J. P. Weigand provided administrative support for the project and reviewed a draft of this thesis. C. S. Bittinger provided assistance in the use of statistical packages. R. L. Eng and H. D. Picton provided helpful editorial comments on the thesis. Individual contributions of D. A. Bricco, B. B. Compton, A. E. Darling, K. A. Kinden, N. S. Martin, L. L. Schweitzer, R. E. Short, and R. B. Staigmiller were also appreciated. Special thanks go to all the landowners of the area, especially Mr. and Mrs. L. B. Chapman, for permitting access to their lands, for their cooperation, friendship, and assistance. The study received financial support from the Montana Department of Fish, Wildlife, and Parks, Federal Aid Project W-120-R. vi TABLE OF CONTENTS Page APPROVAL...'.......................................... ii STATEMENT OF PERMISSION TO USE............... iii VITA................................................. iv ACKNOWLEDGEMENTS........................ v TABLE OF CONTENTS.................... vi LIST OF TABLES........... ........................... . ix LIST OF FIGURES...................................... xiv ABSTRACT.......................... xvii INTRODUCTION....... I METHODS.............................................. 3 Field Methods............... 3 Analytical Methods.............................. 8 STUDY AREA.................... 15 General. Description............................. 15 Physical Features............................... 15 Geology......... 17 Climate..................................... . . . . 18 Land use................ 21 Vegetation....... 22 POPULATION CHARACTERISTICS AND DYNAMICS.......... 30 Population Size................................. 30 Historic Trends............................ 30 Recent Trends.............................. 32 Productivity.................. 32 Mortality....................................... 35 Fawns........... .......................... 3 5 Adult Females.............................. 37 Adult Males................................ 39 vii Distribution............................;...... 40 Effects of Density........................ 41 Effects of Cattle.......................... 43 Sexual Segregation....................... 47 Population Structure............................ 48 Composition.......................... 48 Age Structure.............................. 51 Group Size and Composition...................... 53 Condition Indices............................... 59Factors Influencing Population Dynamics......... 62 Density.................................... 62 Hunting.................................... 63 Climate.......... 70 Time Lags......................... 73 HOME RANGE CHARACTERISTICS AND MOVEMENTS.......... 75 General Movement Patterns....................... 75 Monthly Movement Patterns........................ 78 Seasonal and Yearlong Home Range Size........... 80 Dispersal....................................... 84Home Range Fidelity....... 87 HABITAT USE...............'....................... 8 9 Habitat Selection............................... 89 Annual..................................... 89Winter....................................... 8 9 Spring................................... 92 Summer..................................... 93 Autumn..................................... 93 Comparison of Survey and Radio Telemetry Data.... 95 Food Habits..................................... 96 Habitat Characteristics of Mule Deer Ranges...... 99 Winter................................... 99 Summe r . . .................................. 10 3 Habitat, Dispersion and Home Range Interactions.. 107 Winter............................ 108 Summer..................................... H O Comparison to Habitat Use Patterns in Other Areas 111 HABITAT RELATIONSHIPS AND POPULATION REGULATION....... 114 TABLE OF CONTENTS - Continued Page ■ viii COMPARATIVE ECOLOGY AND RELATIONSHIPS OF WHITE-TAILED DEER TO MULE DEER. .................... I 22 Population Characteristics and Dynamics.......... I 2 2 Population Size........................... 122 Productivity............................... 124 Mortal ity.................................. 125Distribution............................... 129 Age and Sex Composition.................... I 3 2 Group Size .................................. 134 Condition Indices................. ......... 139 Factors Influencing Population Dynamics.......... 14 1 Home Range and Movements....'.................... 145 Habitat Selection............................... 152 Food Habits..................................... 157 Habitat, Dispersion, and Home Range Interactions. 157 Comparison to Habitat Use Patterns in Other Areas 161 Interspecific Interactions in Habitat Use....... 164 COMPARATIVE ECOLOGY AND RELATIONSHIPS OF PRONGHORN ANTELOPE TO MULE DEER....................... ........ I 6 8 Population Trends.............................. 168 Hab itat Use............-........................ 169 Habitat Overlap with Mule Deer.................. 17 2 MANAGEMENT IMPLICATIONS............................. 175 REFERENCES CITED..................................... 181 TABLE OF CONTENTS - Continued Page APPENDIX 196 ix LIST OF TABLES Table Page 1 Dominant species. Shannon-Wiener diversity index, number of species, and coverage values for the six native cover types on the Cherry Creek study area, recorded as means +95% confidence intervals..................... 25 2 Coefficients of similarity comparing plant species occurrence among native cover types on the Cherry Creek study area................ 29 3 Mule deer population estimates on the Cherry Creek Study area from December 1975 to April 1986........................... 33 4 Age specific pregnancy rates, autumn fawn recruitment rates, and projected fawns:100 total females in autumn for marked female mule deer on the Cherry Creek study area, 1983-1985. 34 5 Finite seasonal mortality rates for mule deer on the Cherry Creek study area, 1982-1986... 36 6 Number of marked mule deer lost from the Cherry Creek study area population during 1983-1985, by cause, season, and age class........... 38 7 Observed mule deer buck:doe;fawn ratios and percent bucks> does, and fawns during autumn, winter, and spring on the Cherry Creek study area, 1975-198 6.......... ........... . 49 8 Percentage of mule deer estimated in each age class on the Cherry Creek study area during s p r i„n g, 1983-1986......................... 52 9 Group composition of mule deer on the Cherry Creek study area during July-January, 1976- 1 9 8 6 ....................... ................ 5 5 10 Number of females and fawns, and fawn:doe ratios for groups with I to 7 or more mule deer females per group on the Cherry Creek study area during 1 9 7 5-1 986 ..................... 5 8 X 11 Dressed weights, antler beam diameters, main beam length, and diastema length of. hunter killed mule deer from the Cherry Creek study a r e a ........................................ 60 12 Hunting season regulations for the hunting district that included the Cherry Creek study area, 1 9 7 5-1 9 85 ........ .................. 64 13 Estimated percentage of the preseason population of mule deer which disappeared from the Cherry Creek study area during hunting season, 1982 — 1985......................... 65 14 Estimates of average activity radius and polygon home range size for mule deer on the Cherry Creek study area, 1983-1985.... 81 15 Estimates of female mule deer home range size from this study and from the literature...... 83 16 Timing and distance of 5 documented dispersals of mule deer from their presumed birth range on the Cherry Creek study area.....,............. 85 17 Habitat use of mule deer on the Cherry Creek area based on radio locations (1983-1985) and deer observed during aerial surveys (1975- 1 9 8 5 )...................................... 90 18 Use of various plant taxa by mule deer on the Cherry Creek area, 1983-1985, based on analysis of 47 rumen samples. Data are frequency of occurrence / volume percentage within each period specified......................... 97 19. Comparison of habitat variables between seasons or migration classes for mule deer on the Cherry Creek study area, 1983-1985......... 10 4 20 Comparison of habitat variables between high and low use areas, and large and small home ranges for mule deer on the Cherry Creek area, 1983-1985 ................. . .................. 109 LIST- OF TABLES - Continued Table Page 21 Cover type composition of areas occupied by subpopulations of mule deer on the Cherry Creek study area with the extremes in spring fawn recruitment rates during 1976-1986......... 1 15 22 White-tailed deer population estimates on the Cherry Creek study area from December 1975 to April 198 6 ................................. 1 2 3 23 Age specific pregnancy rates, autumn fawn recruitment rates, and projected fawns:I00 total females in autumn for marked female white-tailed deer on the Cherry Creek study area, 1 983-1 9 8 5.......................... 125 24 Finite seasonal mortality rates for white- tailed deer on the Cherry Creek study area, 1982-1 98 6 .................................. 1 2 6 25 Number of marked white-tailed deer females lost from the Cherry Creek study area population during 1983-1985 by cause, season, and age class....................................... 127 26 Percent of 2,713, 25 hectare blocks on the Cherry Creek study area occupied exclusively by mule deer, exclusively by white-tailed deer, both species, and neither species from 1975 to 1 9 8 6 ...................... ................. ■ 131 27 White-tailed deer buck:doe:fawn ratios and percent bucks, does, and fawns during autumn, winter, and spring on the Cherry Creek study area during 1975-1986............. ....... 133 28 Percentage of white-tailed deer estimated for each age class on the Cherry Creek study area during spring 1983 and 1984.............. 135 29 Group composition of white-tailed deer on the Cherry Creek study area during July-January, 1976-198 6.................................. 1 3 7 30 Dressed weights, antler beam lengths, and beam diameters of hunter killed white-tailed deer from the Cherry Creek area.............. .. 140 xi LIST OF TABLES - Continued Table Page xii 31 Harvest rates for adult male and female, mule and white-tailed deer on the Cherry Creek study area, 1 982-1 985........................... 1 4 2 32 Estimates of average activity radius and polygon home range size for female white-tailed deer on the Cherry Creek study area, 1983-1985. 148 33 Estimates of white-tailed deer home range size from this study and from the literature..... 149 34 Habitat use of white-tailed deer on the Cherry Creek study area based on radio locations (1983-1985) and deer observed during aerial surveys (1975-1985)....................... 1 5 4 35 Use of various plant, taxa by white-tailed deeron the Cherry Creek study area, based on analysis of 25 rumen samples.............. 158 36 Comparison of habitat variables for white­ tailed deer on the Cherry Creek study area, 1983-1985...':............. 159 37 Average composition of 25-hectare habitat blocks occupied by mule deer, white-tailed deer, and jointly by both species on the Cherry Creek study area during summer and winter, 1975-1985 ........................... 165 38 Aerial survey counts of pronghorn antelope on the Cherry Creek study area during autumn, winter, and spring, 1 982-1986............. 1 6 8 39 Habitat use by pronghorn antelope on the Cherry Creek study area, 1982-1985. ... .......... 1 7 0 40 Average composition of 2 5-hectare habi tat blocks occupied by pronghorn antelope, mule deer, and jointly by both species on the Cherry Creek study area, annually and during autumn, 1982-1985 LIST OF TABLES - Continued Table Page 174 xiii Table Page 41 Observability estimates of mule and white­ tailed deer on the Cherry Creek study area (No. deer observed / population estimate x 100)during aerial survey flights............. 197 42 Plant species composition in the six native cover types on the Cherry Creek study area based on an estimated ubiquity index......... 198 LIST OF TABLES — Continued xiv LIST OF FIGURES Figure Page 1 A map of the Cherry Creek study area showing the study area boundaries, drainage divide. Big Sheep Mountain, and 50 m elevation contours.... 16 2 Average monthly temperature and precipitation patterns (1958-1985) from the Buffalo Rapids weather station located 15 km south of the Cherry Creek study area.................... 19 3 Annual precipitation and temperature patterns at the Buffalo Rapids weather station during 1975-1985, recorded as a percentage of the 28 year average........ 20 4 A general overview of the Cherry Creek study area looking west from Big Sheep Mountain, along the drainage divide..................... 23 5 Trends in mule deer harvests reported by the Montana Fish and Game Department from 1960 to 1975, over a constant geographic area which included the Cherry Creek study area....... 31 6 Winter distribution of mule deer groups on the Cherry Creek study area during low populations and high populations. ..................... 42 7 Seasonal distribution of mule deer and cattle on the Cherry Creek study area, 1982-1985... 44 8 Movements of a radioed mule deer in response to livestock grazing on the Cherry Creek study area during 1 9 84 ......................... 45 9 Average total group size of mule deer groups on the Cherry Creek study area that contained adult females, yearling males, and mature males during 1 9 7 5 - 1 9 8 5 ..................... 54 10. Number of yearling and mature mule deer in the Cherry Creek population during autumn, and total autumn harvest of adults, 197 6-1985... 68 XV 11 Relationship between winter severity based on modified Lamb and Leckenby winter severity indices, and fawn mortality rate on the Cherry Creek study area from November through March, 1975-1986....................... 71 12 Relationship between population size in spring and the exponential rate of change over the following year for mule deer on the CherryCreek study area, 1976-1986................... 74 13 Monthly average activity radius for each female migration class and adult males on the Cherry Creek area, 1983-1985......................... 79 14 Locations of 4 female mule deer on the Cherry Creek study area from January 1983 to March 1985, illustrating the fidelity of females to individual branches of a drainage system...... 88 15 Monthly habitat use by radioed female mule deer on the Cherry Creek study area, 1983-1985, relative to cover type availability........ 91 16 Number of mule deer groups observed during winter on the Cherry Creek study area using each slope aspect relative to a common wind direction..................................... 100 17 Relationship between the shelter index and unsheltered wind speed at bedding and feeding sites used by mule deer on the Cherry Creek study area during winter............. . 101 18 Alfalfa field use by mule and white-tailed deer on the Cherry Creek study area during July- October, 1983 and 1984.......... 106 19 Winter distribution of mule deer and white­ tailed deer on the Cherry Creek study area from 1975 through I 985...................... 1 30 20 Average total group size of white-tailed deer groups on the Cherry Creek study area that contained adult females, yearling males, and mature males during 1975-1985............. 13 6 LIST OF FIGURES - Continued Figure Page LIST OF FIGURES Continued 21 Monthly average activity radius for female white-tailed deer on the Cherry Creek area, 1983-1985.................................. 1 4 7 Figure Page 22 Monthly habitat use by radioed female white­tailed deer on the Cherry Creek study area, 1983-1985, relative to cover type availability. 153 23 Habitat characteristics which influence distribution and range overlap of mule and white-tailed deer on the Cherry Creek study area 166 XVl I ABSTRACT The ecology of a mule deer (Odocoileus hemionus) population inhabiting an eastern Montana prairie, was studied during 1975-1986, using periodic aerial surveys and ,radioed and neckbanded deer. Spring population size increased by 29%/year from 1976 to 1983, stabilized at approximately 1,080 deer during 1983-1984, and declined 61% during 1984-1986. An average of 42% of the adult females successfully weaned >_ I fawn, resulting in autumn, winter, and spring fawn:doe ratios averaging 56:100, 55:100, and35:100, respectively. Annual mortality rates for fawns, adult females, and adult males averaged 70%, 23%, and 63%, respectively; harvest rates averaged 3%, 19%, and 59%, respectively. Deer distribution was highly aggregated across the area. Mixed groups of males and females were uncommon, though distributions of males and females overlapped significantly. Individual females exhibited I of 3 movement patterns: yearlong residency, spring and autumn migration, or late summer and late autumn migration. Annual home ranges averaged 4.4 km^. Females and mature, males demonstrated a high degree of fidelity to traditional herd ranges,, but yearling males did not. Areas of topographic diversity in winter, and mesic sites in summer were preferred. Interspersion of habitat complexes influenced both distribution and home range size. ’ Environmental variability caused specific responses by mule deer, and was reflected in variable fawn recruitment rates. Population regulation seemed to be influenced by the limited availability of high,quality forage during late summer. Spring population size of white-tailed deer (0. virginianus) peaked at 335 deer. Population dynamics of both species were similar, except during 1984, when 60% of the female whitetails were harvested. Significant interspecific overlap occurred, and the 2 species demonstrated no intrinsic avoidance mechanisms. White­ tailed deer preferred areas with overhead cover adjacent to agricultural fields, used larger annual home ranges (9.98 km^) and showed much less home range fidelity than mule deer. Pronghorn antelope (Antilocapra americana) occurred on the area at densities similar to white-tailed deer, used open habitats, and had little impact on the mule deer population. I INTRODUCTION Rocky Mountain mule deer (Odocoileus hemionus hemionus) occur across a broad spectrum of habitats, from the Great Plains, through shrublands, woodlands and forests of the Rocky Mountains, and desert scrub of the Great Basin (Wallmo 1981)i Although various aspects of mule deer biology and ecology have been intensiveIy studied in many of these habitats, relatively little research has been directed towards mule deer inhabiting nontimbered prairie environments of the northern Great Plains. Past studies of mule deer, in prairie habitats have focused primarily on range-habitat relationships. These include evaluations of range relationships between mule deer and cattle (Dusek 1975), movements and habitat use patterns of adults (Severson and Carter 1978), habitat use and mortality rates of fawns (Steigers 1981), winter habitat use patterns (Swenson et aI. 1983), and the effects of hunting on habitat use (Swenson 1982). Studies of population dynamics as they relate to habitat characteristics are lacking. In September 1975, the Montana Department of Fish and Game initiated a series of studies to evaluate factors influencing mule and white-tailed deer populations in several different habitats, including an eastern Montana 2 prairie. Although designed as a long-term, intensive study, detailed field work was conducted in the prairie habitat only from September 1975 through May 1976 (Hamlin 1976). From May 1976 through March 1980, efforts were limited to autumn (September-October), winter (December- January), and spring (March-April) aerial surveys to monitor population trends (Hamlin 1977-1980). Field work was suspended during 1980-1981. The work was revived with my studies, under which full-time field work was conducted during July-September 1982, January-March 1983, and June I983-December 1984, and part-time through April 1986. Objectives were to evaluate population dynamics, home range, movement, and habitat use patterns of mule deer occupying this prairie habitat. Additional data were, also collected on white-tailed deer, pronghorn antelope, and cattle, which shared this range with mule deer, to evaluate their potential influences on the mule deer population or its use of the area. 3 - METHODS Field Methods Existing classification and delineation of major cover types within the study area (unpubl. rpt., BLM, Proj. 208- 76-B, 19 79) were employed with 2 minor modifications. A "grassland" type was divided into 2 cover types (mixed grass prairie and bunchgrass prairie), and the closed canopy and open canopy broadleaf tree types were combined into a single hardwood draw cover type. All native cover types were sampled to quantitatively describe the vegetation. The dover type map was arbitrarily divided into 10 units of approximately equal size, and the largest area of each cover type within each of the 10 units was identified for sampling. Sample sites were then identified on aerial photographs, based on the mapping unit description (unpubl. rpt., BLM, Proj. 208-76- B, 1979). At each sample site 40, 2 x 5 dm quadrats were spaced at 5 m intervals along a transect line, and the frequency of each plant species -in each quadrat was recorded to estimate a ubiquity index (frequency x constancy, Curtis 1959:81). In addition, percent coverage of grasses, forbs, shrubs, litter, and bare ground was determined in alternate quadrats following procedures described by Daubenmire (1959). Transects in narrow cover types were oriented at a 4 45° angle to the flow of the drainage to provide samples from the edge. When an outer edge of the type was encountered, another line was established at 90° from the previous line but still at 45° to the drainage u,ntil 200 m were covered. All other cover types were sampled along a straight 200 m line. Vegetation measurements were completed during July 1983. Plant taxonomy followed Hitchcock and Cronquist (1981). During January and February 1983 and 1984, 151 mule deer (110 females and 41 males) were captured using a drive net (Beasom et'al. 1980) or a hand held net^gun (Barrett et al. 1982). An additional 13 animals were captured with the net-gun in Feburary 1986. Deer were manually restrained, ear tagged, and marked for individual identification. Fifty-three females were equipped with radio transmitters (150-151 megahertz) the remainder with colored, plastic neckbands. Ages were determined by tooth replacement and wear (Robinette et aI. 1957). A blood sample was collected from females for pregnancy diagnosis (Wood et aI. 1986b). An additional 34 white-tailed deer (29 females and 5 males) were captured and handled in a similar manner. Sixteen females were fitted with radio transmitters, the remainder with neckbands. Aerial population surveys were conducted during autumn (September-October), winter (December-January), and spring (March-April) of each year from a Piper Super Cub. One 5 aerial survey was also conducted during summer (July 1983). Survey flights were completed during morning and evening hours and followed topographic features to ensure complete coverage of the study area. Morning flights began at daylight and were terminated when activity levels of deer decreased, usually within 2 hours after sunrise. Afternoon flights began as activity levels increased (approximately 2 hours before sunset) and were terminated shortly after sunset. During aerial population surveys, deer were classified by age and sex, locations were plotted on 1:24000 topographic maps, and cover types occupied were recorded. Individuals were classified as mature males (2-years or older), yearling males (1-year old), adult females (1-year or older), and fawns (0.05). Differences in pregnancy rates apparently accounted for differences in the number of females that successfully reared at least I fawn to weaning age (Table 4). However, fawn:100 female ratios in autumn revealed that the highest reproductive rates were by females 5-7 years of age (Table 4). The lack of fetal production data made it impossible to determine whether age-specific differences in fawn production resulted from different rates of fetal production or fawn survival. 35 Although estimates of age-specific fawn production (fawns:100 females) were based on small samples of marked females, they seemed representative bf the entire population. . The mean fawn:doe ratio calculated for marked deer during 1983 and 1984 (58 fawns:100 females) was intermediate to the 67 fawns:100 females and 55 fawns:100 females observed in autumn of those years. Increased fawn:doe ratios reported for 1977 to 1981 must have resulted from increased yearling pregnancy rates, fetal production, and/or fawn survival over summer. Mortality Fawns Seasonal fawn mortality rates (Table 5) were highest in winter (43%) and summer (41%), and lowest in autumn (21%). However, because the seasons were not of equal . length, seasonal rates were not directly comparable. On a monthly basis, mortality rates decreased as fawns matured, averaging 16% from July through September and 12% from October through April. Although fawn mortality during June was not determined, substantial losses can occur during the first month of life (Hamlin et al. 1984). Therefore, higher monthly mortality rates during summer with lower rates during autumn and winter was probably the general pattern for mule deer fawns during most years. 36 Table 5. Finite seasonal mortality rates (percent) for mule deer on the Cherry Creek area during 1982-1986. Season Total Adult Fawn Adult Female Adult Male 04/82-04/83 17 46 3 63 04/82-10/82 2 401 2 210/82-12/82 16 0 I 6212/82-04/83 0 11 0 0 04/83-04/84 28 59 13 62 04/83-10/83 2 33 I 3 10/83-12/83 24 9 9 59 12/83-04/84 4 34 4 4 04/84-04/85 50 91 42 70 04/84-10/84 4 37 3 7 10/84-12/84 45 45 37 65 12/84-04/85 6 74 6 6 04/85-04/86 40 85 35 57 04/85-10/85 I 54 0 2 10/85-12/85 32 28 27 48 12/85-04/86 11 54 . 10 15 Mean Annual 34 70 23 63 Apr.-Sep. 2 41 2 4 Oct.-Nov. 29 21 19 59 Dec.-Apr. 5 43 5 6 IRates for fawns are calculated from July-September 37 During autumn, the rate of natural mortality of fawns was probably higher than hunter harvest. Fawns comprised 16 (8%) of 2 0 5 harvested mule deer examined from 1982 to 1984 while comprising from 32-38% of the population. Based on fawn:doe ratios of deer observed in hunter's possession and population estimates, natural mortality rates for fawns during autumn were 0, 8, 23, and 21% for individual years, from 1982 to 1985; corresponding hunting harvest rates were 0, I, 22, and 7%, respectively. Adult Females Annual mortality rates of adult females in the population averaged 23% (Table 5). Hunting harvest was the greatest cause of death, averaging 19% over the 4 years of intensive study (1982-1986). Natural mortality rates throughout the rest of the year averaged 7%. The annual mortality rate for marked adult females was estimated at 20%. Twelve (20%) of 60 adult females deaths resulted from causes other than hunting (Table 6). Two of these were study related, resulting from deer getting a front foot entangled in their collars. Mortality rates related to stress induced by trapping and handling were very low for the trapping techniques used in this study (Wood et al. 19 86a),. With these 2 animals removed from the calculations, 17% of all deaths resulted from natural causes and 83% were harvest related. 38 Table 6. Number of marked mule deer lost from the Cherry Creek Study area population during 1983-1985 by cause, season, and age class. A=accident, C=cripple loss, E=emigration, H=harvest; I=Illegal kill, M = malnutri.tion, P=prelation, T = trap . related, U=undetermined. Age Class Number Alive FEMALE LOSSES Annual Autumn Winter Summer P 28 4 0 IM 3U 0 I 39 6 3H II IM IT 0 2 41 9 7H IT IE 3 42 7 4H IM IU IU 4 44 V8 7H 0 IA 5 38 7 3 C 3H IU 0 6 33 9 IC II 6H IM 0 7 21 4 3H 0 IP 8 15 . 5 4H II 0 0 9 10 3 2 H IP 0 10+ 11 2 2H 0 0 Total 322 64 48 12 4 MALE LOSSES Age Number Class Alive Annual Autumn Winter Summer 0 26 2 0 2U 0 I 32 18 2H 0 16E 2 16 6 6H 0 0 3 9 4 4 H 0 0 4+ 10 - 2 2H 0 0 Total 93 32 14 2 16 39 Causes of natural mortality among females varied. During winter, 4 cases of malnutrition and I coyote (Canis latrans) kill were confirmed. Five marked females disappeared during winter and were probably winter related mortalities. During summer, I dispersal, I coyote kill, and I automobile collision were confirmed, while I marked female disappeared (Table 6). No natural mortality of marked or unmarked deer was documented among adult females during autumn. Seven (15%) of 48 marked females which died during autumn were documented crippling losses or illegal kills which were included in the calculation of total harvest rates (Table 6). Adult Males Annual mortality rates of adult males averaged 63%. Hunting related mortality was the leading cause of death, averaging 59%. Natural mortality rates averaged 10%. The only cause of death identified for marked adult males was hunting (Table 6). Nine marked mature males were shot and reported by hunters; 3 others disappeared during hunting season and were assumed shot. Two yearling males also disappeared during the hunting season and were assumed shot. The percentage of marked males in the population increased from 7% to 14% over the 1984 hunting season, indicating a selective harvest of unmarked males by i 40 hunters. No cases of crippling loss or illegal kill were documented for marked males, although 12 unmarked males were found dead of those causes. Three yearling males were known to have dispersed, 13 others disappeared during the summer and were assumed to have dispersed. Emigrating yearling males were apparently replaced by immigrants because there were no changes in population size commensurate with a nonrepla.ced loss of 16 (67%) of 24 yearling males. Distribution Mule deer were distributed in scattered aggregations across the study area. Overall, from 1975 to 1986, deer were observed in only 3 7% of the 25 ha blocks during survey flights. Seasonally, only 20%, 19%, and 15% of the blocks were occupied during autumn, winter, and spring, respectively. Density patterns also indicated an 2aggregated distribution. Density averaged 2.3 deer/km across the entire study area in December 1983, but local concentrations ranged from 5 to 33 deer/km^. The same general pattern was observed during all seasons and years. Mule deer appeared to exist in semi-isolated ' • " /subpopulation units on the Cherry Creek study area. A few subpopulations were apparent because, they were separated by up to 7 km of unoccupied habitat. Other subpopulation units were suggested on the basis of the patchy 41 distribution of deer across the area and the distribution and movements of marked deer. Effects of Density' No major changes occurred in the distribution of deer over the study area in response to increasing population density. If deer had widely expanded their distribution into previously unoccupied habitats in response to increasing population density, a large number of 25 ha blocks unoccupied during lows would have been occupied during highs. However, this was not the case. During all seasons, deer occurred in the same blocks during population highs and lows more than expected, and in newly occupied blocks during population highs less than expected based on chance alone (P<0.05). Thus, changes in distribution resulted more from differences in the relative abundance of deer within subpopulation units across the study area than from expansion of deer distribution into previously unoccupied habitats (Figure 6). Studies of white-tailed deer have indicated that female fawns are recruited into (Nelson and Mech 1985) or adjacent to (Ozoga et al. 1982) areas used by their mothers, resulting in a traditional pattern of distribution through time. This phenomenon might also explain the consistency in distribution of female mule deer on the Cherry Creek area. Dispersal of males (Robinette 1966, Severson and Carter 1978, Nelson and Mech 1984) has the O O % O Figure 6. Winter distribution of mule deer groups on the Cherry Creek study area during low populations (+, 1976-1980) and high populations (o, 1982-1983). 43 potential to change distribution. However, the small number o f males and their high mortality rate apparently minimized any such effect. Apparently, the availability of suitable habitat was sufficiently restricted and the deer density during 1975-1976 sufficiently low to result in a nearly 6-fold increase in mule deer numbers within the originally occupied habitat. Effects of Cattle Cattle grazing did not seem to broadly influence the distribution of mule deer during any season (Figure 7). The occurrence of mule deer in occupied cattle pastures was not different from expected based on the proportion of the study area occupied by cattle (P>0.05). Cattle generally grazed drainage bottoms and rolling prairies and either avoided or made relatively minor use of the rougher terrain in badlands and along drainages which deer preferred (see habitat use). This limited the opportunity for cattle to affect mule deer distribution over an entire pasture. There was one case, however, where a deer seemed to avoid occupied cattle pastures. This involved a radio- equipped, female mule deer whose home range overlapped 3 adjacent pastures. Pastures I and 2 were part of a rest- rotation grazing system and were grazed rather intensively for 40 and 50 days in early and late summer, respectively. Pasture 3, part of another allotment, was grazed SUMMER CATTLE DISTRIBUTION AUTUMN • MULE DEER GROUP -C'•e- Figure 7. Seasonal distribution of mule deer and cattle on the Cherry Creek study area, 1982-1985. 45 continuously for 184 days by relatively fewer cattle (Figure 8). When livestock were placed in pasture I, this deer appeared to spend less time there and more time in pasture 2 which remained ungrazed. When the cattle were later moved to pasture 2, she apparently spent more time in pasture 3 which was grazed all summer at a lower rate. PASTURE I PASTURE 2 260 cattle (77/km^) I 260 cattle (46/km2) June 12 - July 21 I I July 21 - September 8 May 29 i May 10 Jun. 27 I May 17 Aug. 23 I J un. 5I Jul. 10 PASTURE 3 300 cattle (9/km2) May 15 - November 15 May 2 Jul. 20 Sep. II Sep. 19 Sep. 30 Figure 8. Movements of a radioed mule deer in response to ,livestock grazing on the Cherry Creek study area during 1984. Below livestock density and grazing periods are dates on which this deer was located in each pasture. 46 Although hot conclusive, these data suggested that deer avoided close association with cattle, and that grazing patterns and stocking rates can affect how deer respond to cattle as discussed by Mackie (1981). Results of other studies on the effects of livestock grazing on deer distribution have varied greatly. McMahan (1966), Dusek (1975), and Edwards (1977) reported deer avoid pastures occupied by livestock. Knowles (1975) suggested that, while summer and autumn cattle grazing did not exclude deer from an area, it did cause them to move more widely. Komberec (1976) found little effect on mule deer distribution resulting from winter and spring grazing. Livestock grazing has also been reported to directly increase mortality rates (McMahan and Ramsey 1965, Knowles 1975) or affect the availability of preferred feeding areas (Austin and Urness 1986). Hardwood draws, an important cover type for deer during summer (Severson and Carter 1978, and see habitat use section), are also heavily used by livestock for feed and shade. When hardwood draws, are uncommon, even lightly stocked pastures can be subject to damage by rubbing, grazing, and soil compaction (Severson and Carter 1978). This type of impact could cause decreased cover for fawn rearing even if it did not directly affect deer distribution or forage availability. Hamlin et aI. (1984) reported that fawn production in central Montana was not limited by the availability of hiding cover. However, fawns on that area used patches of dense cover on steep slopes which cattle avoided. On the Cherry Creek area most of the rugged terrain had little available vegetative cover to hide young fawns. The best hiding cover occurred in the drainage bottoms that were also used for feeding and loafing by cattle. Thus, cattle had the potential to. limit the availability of fawn rearing sites by direct exclusion, or destruction of cover and forage. Sexual Segregation Mature males and adult females were least associated in common groups during summer and autumn (see group size section) which maximized chances for a sexually segregated . distribution of deer across the study area during those seasons. However, even though they were not closely associated in common groups, their distributions overlapped. During July-September, mature males and adult females occurred in the same 25 ha blocks more than expected and in different areas less then expected compared to a random distritution (]?<0.05). Male concentration areas (more than 2-times as many mature males as females) were uncommon, comprising only 3 5% of the blocks in which mature males occurred. 48 Exclusive use of areas by adult females was observed but was not evidence for sexual segregation because females were relatively more abundant. Mature males comprised only 12% of the adult population during summer and autumn. Sexual segregation apparently occurs in other environments. Mackie (1970) reported the existence of "buck" habitats for a mule deer population in central Montana as did Bowyer (1984) for southern mule deer (0̂. h.. fuliginatus) in California. The lack of buck habitats on th§ Cherry Creek area probably resulted from the relatively few mature bucks in the population, the high mortality rate of mature males, and the concentration of all animals on only a fraction of the total study area. If lower mortality rates were maintained, buck habitats might also form on the Cherry Creek area. Population Structure Composition Fawn production was relatively high throughout most of the study. During 1975-1986, fawn:doe ratios averaged 6 9:100 in autumn and winter and ranged from 39 to 111:100 (Table 7). Fawns comprised an average 33% (range 24-45%), 36% (range 25-48%), and 32% (range 10-44%) of the population during autumn, winter, and spring, respectively. The percentage of fawns in spring generally increased from 1976 to 1982, and decreased from 1983 to 1985. Table 7. Observed mule deer buck:doe:fawn ratios and percent bucks, does, and fawns during autumn, winter, and spring on the Cherry Creek study area, 1975- 1986 . AUTUMN WINTER SPRING Ratio Percent Ratio Percent Ratio Percent Year buck:doe:fawn buck doe fawn buck:doe:fawn buck doe fawn buck:doe:fawn buck doe fawn 1975-761 30:100:43 17 58 25 7:100:48 5 64 31 7:100:48 5 64 31 1976-771 42:100:67 20 48 32 36:100:66 18 49 33 36:100:66 18 49 33 1977-781 51:100:79 22 44 34 31:100:74 15 49 36 30:100:56 16 54 30 1978-791 — — — 19:100:111 8 44 48 19:100:90 9 48 43 1979-801 17:100:97 8 47 45 12:100:102 5 47 48 12:100:80 6 52 42 1980-812 36:100:84 16 46 38 — — 40 1981-822 45:100:90 19 43 38 — — 44 1982-83 31:100:64 16 51 33 12:100:66 . 7 56 37 12:100:59 7 59 34 1983-84 45:100:67 21 47 32 . 20:100:67 11 53 36 20:100:46 12 60 28 1984-85 38:100:55 20 52 28 21:100:47 13 59 28 21:100:13 16 74 10 1985-86 27:100:40 16 60 24 19:100:39 12 63 25 18:100:20 13 72 15 D̂ata from Hamlin, 1976-1980 Ûnpub. data, Mont. Dept. Fish, Wildl., and Parks, Fed. Aid Proj. W-130-R. 50 Buck:doe ratios were relatively high in autumn, averaging 36:100 (range 17-51) but dropped to 20:100 (range 7-36) following the hunting season. Post-season buck:doe ratios were highest following the 1976 and 1977 hunting seasons and lowest following the 1975, 1979, and 1982 seasons. Population composition was estimated with fair precision. During summer, fawn:doe ratios increased steadily from mid-July to late August then stabilized through October. Replicate classification surveys from late August to mid-October (5 in 1983 and 4 in 1984) resulted in coefficients of variation ranging from 4%-6% for fawn:doe ratios and I0%-l7% for buck:doe ratios. Robinette et al. (1977) also reported greater precision in estimating fawh:doe ratios as compared to buck:doe ratios. Ratios from replicated surveys in winter and spring were also fairly consistent. The difference between the fawn:adult, fawn:doe, and buck:doe ratios determined from the air and ground, ranged from 5 to 26% of the overall mean determined for the 2 independent samples. Much of the discrepancy between air and ground classification surveys was attributable to the inisclassification of small yearling bucks as does during winter, and large fawns as yearlings during spring. Deer surveys conducted from the ground had the advantage of providing more accurate 51 classifications while aerial surveys provided more representative samples across the population. Age Structure The age structure of the female segment of the population was heavily skewed toward younger age classes during 1983-1984 when harvest rates were low, but shifted toward an older age structure concurrent with increased harvest rates during 1985-1986, as a result of decreased fawn recruitment (Table 8). Approximately 53-55% of the females were 2 years or younger during 1983 and 1984. By 1986, this group comprised only 32% of the females. The percentage of females 8 years or older increased from 6% to 12% over the same time period. The percentage of prime age females (3-7 years old) showed less variation, increasing from 41% to 56% during 1984-1986. 7The male segment of the population was more heavily skewed toward younger age classes than the female segment (Table 8). Fawns comprised 69% of all males during 1983, while bucks older than 4 years were uncommon. The lack of older males in the population was also indicated by harvest data in which only 8 (10%) of 78 antlered mule deer examined were 4 years or older, and only 5 (6%) were 5 years or older. Table 8. Percentage of mule deer estimated in each age class on the Cherry Creek study area during spring 1983-1986. Percent of males recorded in age class 4 represents all males 4 and older. YEAR FEMALES MALES Age Class 83 84 85 86 83 84 85 86 0 24 20 6 9 69 61 24 36 I 17 18 18 6 17 25 46 15 2 14 15 17 17 9 9 2 0 29 3 12 13 14 15 4 3 7 13 4 11 12 12 13 I 2 3 7 5 8 8 11 11' \ 6 4 4 8 10. 7 4 4 4 7 8 3 3 4 4 9 I I 3 3 10+ 2 2 3 5 Number of deer 760 794 472 335 323 280 123 86 53 Group SIze and Composition Average group size for a given season was not correlated with population size (£>0.05). Therefore, data on group size and composition were combined from 1975 to 1986. The average number of deer per group for all groups containing adult females, mature males, and yearling males increased through the summer, peaked in January and February, then declined (Figure 9). Males occurred in larger groups than females during all months when individuals in the group could be classified. Increased female group size through summer may have resulted from reestablishment of family groups following parturition (Hawkins and Klimstra 1970) but the continued increase from October through February was also affected by the increased use of common areas as described by Mackie (1970). Studies of black-tailed deer (£. h. columbianus) by Dasmann and Taber (1956) and white-tailed deer by Hawkins and Klimstra (1970) and Gavin et al. (1984) indicated a temporary breakup of doe groups during the rut. This phenomenon was ■ . ■■ not observed on the Cherry Creek area. Over 49% of the females occurred in groups with other females, with fawns, or alone from July through January (Table 9). Females alone were most common in July, but groups including 2 or more females with fawns were most AV ER AG E GR OU P SI ZE 54 ------- females -------mature males • • • • yearling males ------- combined MONTH Figure 9. Average total group size of mule deer groups on the Cherry Creek study area that contained adult females, yearling males, and mature males during 1975-1985. Sexes were combined from Feburary through May. 55 Table 9. Group composition of mule deer on the Cherry Creek study area during July-January, 1976-1986. Data are percentage of all females, mature males, and yearling males which occurred in each category. MONTH 7 8 9 10 11 12 I FEMALES Alone 35 20 I 6 4 2 I In groups/no fawns 17 12 13 7 8 4 I Alone with fawns 12 13 8 7 5 4 5 In female/fawn groups 10 27 40 40 39 41 43 With mature males 9 4 7 9 16 22 8 With yearling males 12 17 19 22 14 18 20 In mixed groups 5 7 6 9 14 9 22 Number observed 606 10.Ql8 1821 1085 494 1793 674 MATURE MALES Alone 21 12 12 14 6 5 6 Mature male groups 20 29 32 2 6 0 5 14 With females 27 25 14 23 60 63 29 With yearling males 17 15 17 19 0 5 10 In mixed groups 15 19 25 18 34 22 41 Number observed 144 173 285 140 50 153 79 YEARLING MALES ■ Alone 25 18 10 10 9 7 3 Yearling male groups 12 8 9 7 0 2 0 With females 38 51 47 53 57 56 50 With mature males 13 7 .12 13 0 5 8 In mixed groups 12 16 22 17 34 30 39 Number observed 122 218 303 166 44 161 77 56 prevalent by August. Even during the breeding season (November-December)j no more than 22% of all adult females were observed with mature males. Mature males were observed with females almost as frequently as with other mature males during July-October (Table 9). Most mature males were associated with other mature males or in mixed groups during September; During October, mature males were in groups with any other deer. During November and December, over 60% of the mature males were associated with females during the breeding season. The fact that less than 22% of the females were attended by mature males reflected the low male:female ratios (Table 7). From November through January, mature males were most often associated with females or in mixed groups of females, mature males, and yearling males. Geist (1981) suggested that mixed groups of mule deer should be common during winter when the availability of winter range is limiting and when the chances of males successfully reproducing in consecutive years are limited. Data from this study support that hypothesis. The majority of yearling males occurred with females from July to January. Yearling males alone, with other yearling males, or with mature males were relatively uncommon from September through January. Because yearling males begin dispersing in summer (Robinette 1966, and see home range and movements section) and may establish new 57 home ranges by November (Nelson and Mech 1984), they were probably associating with females other than their mother. Females with fawns tended to occur in smaller groups than females without fawns. The number of females per group during December and January decreased as December and January fawn: doe ratios increased (1P<0.01, r=-0.83, n=9). Also, fawn:doe ratios were highest for females that were not accompanied by other adult females and generally decreased as the number of adult females per group increased (Table 10) over all years. Female white-tailed deer that lose their fawns seek the company of other deer soon after their fawns death (Ozoga et al. 1982). Thus, it appears that does with fawns prefer isolation but, except for a brief period around parturition (Ozoga et al. 1982), and under some conditions (Miller 1974), they do not enforce their isolation through aggression. Under this hypothesis, increased group size results from nonproductive I does joining with a producing doe and possibly her fawns from the previous year. Average mule deer group size on the Cherry Creek study area was larger than in other areas.of Montana (K. L. Hamlin and D. F. P a c, pers. comm.) and Utah (Robinette et al. 1977). Geist (1981:219) suggested that large group size was an antipredator adaptation of. mule deer occupying open habitats. A similar relationship has also been postulated for bighorn sheep (0vis canadensis) (Bailey 58 Table 10. Number of females and fawns, and fawn:doe ratios for groups with I to 7 or more mule deer females per group on the Cherry Creek study area during 1975-1986. ! Females/group September-October December-!anuary Fern. Fawns Ratio Fern. Fawns Ratio I 493 363 0.74 176 188 1.07 2 600 386 0.64 312 270 0.87 3 459 231 0.50 279 206 0.74 4 396 175 0.44 368 192 0.52 5 265 115 0.43 215 118 0.55 6 186 54 0.29 252 140 0.56 7 + 507 210 0.41 865 398 0.46 Total or Mean 2906 1534 0.53 2467 1512 0.61 1980, Risenhoover and Bailey 1985). If large groups were formed as an antipredator strategy, then those deer most vulnerable to predation (i.e. does with fawns) should form larger groups. Because females with fawns occurred in smaller groups and females associated in smaller groups than males (Figure 9), the hypothesis may not be valid for mule deer in prairie habitats. Apparently, mule deer social organization is similar to that described for white-tailed deer by Hawkins and Klimstra (1970) and.Nelson and Mech (1984). Yearling males 59 typically disperse from the family unit. As female fawns mature and begin to produce fawns, they separate from their mothers and form their own matriarchal group. This hypothesis would explain why the number of females per group decreased with increasing fawn production and why average group size did not change as the population increased. Condition Indices Dressed weights of fawns shot by hunters averaged 2.0.7 and 22.7 kilograms (kg) for females and males, respectively. Weights of yearling females averaged 33.1 kg, though most were obtained in 1984 when weights of other age classes were minimal and therefore are probably lower than long-term averages. Weights of mature females (2+ years) averaged 38.5 kg (range 32-48 kg), yearling males averaged 39.3 kg (range 32-48 kg), and mature males averaged 56.5 kg. The heaviest (64 kg) and lightest (52 kg) mature males were both 3 years old. Main antler beam length of yearling males averaged 21.4 cm (range 4-28 cm), antler diameter 1.54 cm (range 0.8-2.5 cm), and diastema length 6.45 cm (range 5.9-7.5 cm). Among mature males, beam diameter averaged 2.33 cm (range 1.7-3.2 cm) and beam length averaged 36.7 cm (range 25-48 cm) (Table 11). Table 11. Hog-dressed weights (kg), antler beam diameters (cm, measured 2.5 cm above the bur), main beam length (cm), and diastema length (cm) of hunter killed mule deer from the Cherry Creek study area. Sample size is recorded in parentheses. FAWNS ADULT FEMALES ADULT MALES Age (years) Age (Years) Year Fern. Wt. Male Wt. 1.5 2+ 1.5 2+ , wt. Wt. Wt. Beam diam. Beam length Diastema Wt. Beam diam. Beam length 19751 20(1) 25(2) — 41(3) 42(4) 1.7(8) --- -- 65(3) 2.6(6) — 1982 — — — 43(2) 43(1) 1.9(5) 23.3(5) 6.5(2) 61(2) 2.7(8) 38.6(8) 1983 18(1) 20(1) 34(1) 40(8) 40(14) 1.5(18) 22.1(18) 6.6(18) 57(10) 2.2(16) 37.4(16) 1984 21(4) 22(5) 33(7) 37(21) 35(5) 1.3(10) 19.1(10) 6.2(10) 51(10) 2.1(17) 35.1(17) 1985 23(1) 24(1) — 39(7) 36(1) 1.6(1) --- 6.1(1) 68(1) 3.0(2) — MEAN 20.7 22.7 33.1 38.5 39.3 1.54 21.4 6.45 56.4 2.33 36.68 D̂ata from Hamlin 1976 61 There seemed to be a decrease in the physiological condition of adult mule deer on the study area during 1983 and 1984. Although dressed weights in 1983 were comparable to those in earlier years, antler growth had decreased (Table 11). In 1984, body weights of both sexes, and antler growth of males decreased. Weights for males in 1975 and 1982 were comparable to those collected from throughout Montana (Mackie 1964), but were 18-22% less by 1984. The percentage of yearling males that were spikes (at least I unbranched antler) increased from 1983 (38%) to 1984 (52%). Although sample sizes were small, and there was no statistically significant decrease in either weights or antler measurements, other data supported the hypothesis of a decline in physical condition. The dentary bone is a priority growth area in young mammals (Reimers 1972). Diastema measurements of yearling males decreased significantly (]P<0.05) from 1983 to 1984. Also, kidney fat indices (KFI) were very low, averaging 39% (n=9) and 66% (n=7) for adult females and males, respectively in 1984. Although samples were collected during October and November, a time when fat reserves peak (Anderson et al. 1972), they were nearly at levels where utilization of marrow fat begins (Ransom 1965). The first effects of reduced physical condition in species that typically have multiple births is a decrease 62 in the mean litter size (Caughley 1976). During July, 1982-1985, a decrease in the number of fawns-at-heel per producing female (1.50, 1.43, 1.29, and 1.27, respectively), also indicated a general decrease in the physiological condition over those years. Fawn-at-heel ratios in 1986 (1.54) indicated that the increase in body weights and antler growth of deer harvested in 1985 reflected a trend of improving condition. Although no single index was adequate to conclusively demonstrate physical condition, taken collectively they all indicated a decline in physical condition beginning in 1983 and extending to summer, 1985. Factors Influencing Population Dynamics Density The finite rate of population increase from 1976 to 1982 averaged 29% per year (SE=0.92%), indicating that population growth remained relatively constant over the period. From 1976 to 1982, the size of the population in any given year could be estimated simply by knowing the population size from the previous year and the average rate of increase (r=0.997, jP<0.001) during the period. Fawn mortality rates during summer increased as the number of fawns, females, and total deer present during summer decreased (P<0.03), which was opposite to the density dependent response normally reported for ungulates 63 (Caughley 1976). There were no other significant correlations between population size and seasonal or annual natural mortality rates (P̂ >0.2). Natural mortality rates of adults were not related to density. . The strongest correlation found was between the mortality rate of adult females in summer and female abundance from 1982 to 1986, but the relationship was not significant ( r = 0.7 7 , P/= 0.23). Adult female mortality due to natural causes remained fairly stable over the study period, similar to other ungulate populations (Caughley 1970, Skogland 1985). Skogland (1983) proposed that high density ungulate populations suffer decreased physical condition, which Hanks (1981) suggested leads to the following sequential demographic changes: I) increased juvenile mortality 2) increased age of sexual maturity, 3) declined fecundity of mature females, and 4) increased adult mortality. Mule deer on the Cherry Creek study area progressed through steps 1-3, but apparently not as a simple function of density. Hunting Hunting regulations varied from 1975 to 1985 (Table 12) and influenced population size and harvest rates (Table 13) of adult females. From 1976 through 1980, antlerless mule deer were not legal game. From 1980 to 1985, at least 64 Table 12. Hunting season regulations for the hunting district that included the Cherry Creek study area, 1975-1985. Two types of permits were issued to residents, unlimited permits that all licensed hunters could purchase, and limited permits which were issued through drawings. YEAR Unlimited Permits Limited Permits 1975 I ES1 ESp. + I ES WT 19 7 6 2 I ES WT v 50 ES MD 1977 I ES WT 200 antlered MD 1978 I ES WT or antlered MD none 1979 I ES WT or antlered MD 200 antlerless WT 1980 I ES WT or antlered MD 200 antlerless ESp. 1981 I ES ESp . 200 antlerless WT 19822 I ES ESp. none 1983 I ES ESp. 200 antierless ESp. 19 84 2 I ES ES p. +■ 5 ant IerIe s s WT 5000 antlerless ESp. 1985 I ES ESp. + I antlerless WT 4000 antlerless ESp. ^ES = either sex deer, ESp. = either species of deer, WT = white-tailed deer, MD = mule deer ^Hunting district boundaries were changed so number of permits issued was not comparable to previous years. 65 Table 13. Estimated percentage of the preseason population of mule deer which disappeared from the Cherry Creek area during the hunting season, 1982-1985. Adult Males AdultYear Total Fawns Fern. All Ma t. Yearl. ( 2pt. Spike) 1982 10 0 I 62 60 63 ( — " ) 1983 19 9 9 5 9 73 48 (87 15) 1984 45 45 37 65 71 61 (80 36) 1985 31 28 27 48 48 50 ( — — ’) some hunters could legally shoot antlerless mule deer. Regulations during 1982 resulted in a 1% harvest of adult females (Table 13) which allowed the population to increase by 26%. From 1983 to 1985, special regulations were implemented (Table 12) to increase the harvest of antlerless mule deer. During 1983, 200 hunters were allowed to shoot an antlerless deer of either species (a "B" deer tag, 6.2 tags sold/100 km^) in addition to their regular "A" tag, which resulted in a 9% harvest (Table 13). During 1984, the size of the hunting district was increased and 5000 "B" tags (27.7 tags/100 km^) were sold. A maximum of 6 tags (I "A" and 5 " B") could be purchased by each hunter which resulted in a 37% harvest. In 1985, 4000 "B" tags (22.2/100 km^) were sold resulting in a 27% harvest of adult females. Apparently, the smaller population size and 66 the extremely severe weather during the last 2 weeks of the season resulted in a decreased harvest rate compared to 19 84. The more liberal regulations during 1983-1985 increased harvest rates. The harvest rates of adult males reflected intense selection by hunters for males with the largest possible antlers as described by Taber and Dasmann (1957), Robinette et al. (1977), and Barlow and McCullouch (1984). Yearling males with at least 2 points on each antler were harvested at a rate of 80-87%, while those with at least I unbranched antler (spikes) were harvested at a rate of 15-36% (Table 13). The increase in the harvest of spikes during 1984 resulted mainly from their harvest as "antlerless" deer, because small spikes are difficult to see at long distances or on running deer. Harvest rates of adult males on the Cherry Creek study area were similar to the 56% harvest rate reported by Robinette et al. (1977) and were much higher than the 15% reported by Barlow and McCullouch (1984). Annual mortality rates of males were only slightly higher than the 57% reported by White and Bartmann (1983), and were much higher than the 24-29% reported by Taber and Dasmann (1957). Harvest rates for adult females on the Cherry Creek area were comparable only to the 19% reported in Utah (Robinette et al. 1977). Only studies of deer in enclosures have reported annual harvest rates higher than these. Nellis 67 (1968) reported a 50% annual harvest of adult deer on the National Bison Range in western Montana. The decline in the number of adults from 1984 to 1986, resulted from a decrease in the number of yearlings in the population concurrent with an increase in hunting harvests (Figure 10). Total number of yearling recruits peaked in 1982. A combination of 5% natural mortality and 24% hunter harvest of adults, along with a late spring population consisting of 28% fawns (Table 7), resulted in a stable population from 19 83 to 19 84. The fact that the 1984 harvest exceeded the number of yearling recruits in the population along with the severe decrease in yearling recruitment during 1985, resulted in a 43% decrease in the autumn 1985 population size. The 1985 harvest also exceeded recruitment, resulting in a further decrease in the population size to the spring of 1986. Eustace and Swenson (1977), and Swenson (1978) suggested that the mule deer population decline which occurred in eastern Montana from 1971 to 1975 also resulted from a combination of declining recruitment and heavy hunting pressure. Harvest rates of adult females exceeded recruitment rates of fawns into the adult population during 1984 and 1985 and was therefore not sustainable. From 1976 to 1980 the population averaged 36% fawns in spring, providing for - an annual rate of population growth of 24% in adult females. If fawn production during 1983-1985 had been 68 1250-1 3 1000 Mature Yearlings J Harvest YEAR Figure 10. Number of yearling and mature mule deer in the Cherry Creek population during autumn, and total autumn harvest of adults, 1976-1985. sustained at the same rate as 1976-1980, the average harvest rate experienced from 1982 to 1986 (19%) might also have been sustained. Autumn hunting mortality and annual natural mortality appeared to be additive for adult females, for which natural mortality tended to increase with increasing 69 harvest rates (r=0.86, £=0.07). Compensatory decreases in natural mortality rates with increasing harvests are a common assumption of wildlife management, and have been proposed for North American big game by McCullough (1979) and Houston (1982). However, even if compensatory mortality can be demonstrated it is unreasonable to assume that hunting mortality could completely eliminate natural mortality in wild ungulate populations (Caughley 197 6). The number of adult males harvested on the study area was influenced primarily by adult male density from 1982 to 1985 (r=0.987, P<0.01). From 1977 to 1982 the total mule deer harvest on a 7,750 km ̂area which included the study