5 3 9.7357 Population F2oewd ecoloey of white- 1390 tailed deer in northwestern dion t din a m iffi STATE DOCUMENTS COLLECTION f5* I t L I U * 1 u APR 2 3 1991 MONTANA STATE LIBRARY 1515 E- 6th AVE. HELENA, MONTANA cMoT\tcu\a Departnjeqf of ) tFis^,2 years for age assignment by the cementum annuli technique (Matson's Lab, Milltown, MT) . Length of the diastema was measured on all deer, and antler measurements, including length and diameter of the main beam and number of points, were taken from all males of ages >1 year. Pellet group transects were run during April-May 1990 as previously described (Dusek 1989) using the rationale of Longhurst and Connolly (1982) and the analytical procedures of Davis (1982) . Sampling intensity was increased from 25 1-milliacre circular plots to 50 per transect to increase precision (Dusek 1989) . Individual transects were placed within 13 randomly selected 1-km quadrats for a total of 650 sampled plots. Estimates of survival among radio-collared females of ages >1 year and males >2 years followed Heisey and Fuller (1985) using the software MICROMORT. Time intervals, or seasons, used previously (Dusek 1989) were modified because the daily rate of survival could not be assumed constant through the autumn archery and rifle hunting seasons. The autumn period was split between 5 early autumn (1 Sep. -15 Oct.), that included the archery season, and late autumn (16 Oct. -30 Nov.) that included the rifle season. Because documented mortality among radio-collared deer was negligible during the archery season and daily probability of survival was not different from that of summer (May-Aug.), the 2 periods were combined in the analysis. Winter (Dec. -Feb.) and early spring (Mar. -Apr.) also were combined because daily survival rates were not different, irrespective of cause. Thus, 3 periods (prehunt, rifle hunt, and posthunt) were used in which probability of daily survival was assumed constant throughout each period. RESULTS AND DISCUSSION Trap Efficiency From 10 January through 24 February 1990, a total of 172 white-tailed deer were captured over 429 trap-nights, including all recaptures and mortalities, for an overall trap efficiency of 0.40 deer caught/trap-night. There was no apparent difference in overall trap efficiency between years, but vulnerability of deer to trapping apparently differed between the 2 winter ranges in 1990 (Table 1) . The higher trap efficiency at MDWR probably reflected the fact that more than one-third of the total effort there in 1990 involved an area not trapped previously, a portion of the winter range between Fortine Creek and Dudley Slough. Also, because less time and effort were spent on MDWR, deer may also have been less wary of traps. Most trapping on the BTWR in 1990 occurred in areas also trapped in previous years, and thus, some deer probably were conditioned to avoid traps. Four deer (2% of all captures) died at the trapsite during winter 1990. All resulted directly from trap-related injuries. Fawns of both sexes were relatively abundant among 142 new deer marked during winter 1990 (Fig. 3) . Application of Remote Camera Surveys General Operation of the System Three camera survey sessions during winter 1989-90 resulted in 853 exposures over 415 camera nights at 28 sites (Table 2) . Of the total exposures, 653 (77%) were of wildlife; 604 (71%) were of white-tailed deer. During the 3 sessions, a total of 555 deer-visits were identified among all exposures involving white- tailed deer. Of 25 individually marked deer photographed at camera sites during all sessions, none made repeated visits to the same camera site within a 2-week sampling period. However, one individual was photographed at 2 different sites during session 1. During late winter-early spring of the previous year, a marked yearling male (88022) was photographed at the same site 6 Table 1. Efficiency of Clover traps for capturing white-tailed deer on 2 winter ranges, 1988-90. Month and location3 Year Total trap nights Total deer captured Total deer marked Total captures/ trap night December BTWR 1988 100 41 30 0.41 January BTWR 1989 152 53 41 0.35 1990 193 66 55 0. 34 MDWR 1989 70 27 24 0.39 1990 45 20 17 0.44 February BTWR 1989 129 65 52 0.50 1990 162 54 42 0.33 MDWR 1989 42 22 20 0. 52 1990 44 32 28 0.73 rid L CI 1 BTWR 1989 42 19 13 0.45 Total 1989 535 227 175 0.42 1990 429 172 142 0.40 a Bowser-Tally Lakes winter range (BTWR) ; Murphy-Dickey Lakes winter range (MDWR) . 7 1989 70 AGE CLASS (YRS) 1990 Figure 3. Age composition (%) by year of 317 deer captured during winter on BTWR and MDWR combined, 1989-90. 8 Table 2. Summary of remote camera surveys at Bowser-Tally and Murphy-Dickey Lakes winter ranges during winter 1989-90. Session number 12 3 BTWR MDWR BTWR Dates operated: 12/18 -1/2 2/20-3/7 2/27- 3/16 No. stations: (functional) 10 8 Grid density: ( t\c\ / Lr m ^ 1:3 .5 1: :2.8 1:4 .7 Total frames : 192 312 349 System check: 34 (18%) 33 (11%) 27 (8%) Wildlife: 132 (69%) 241 (77%) 280 (80%) TTnlen vpti phi p 26 (14%') 38 (12%) 42 (12%} Frames w/ WTD: 130 207 267 No . indiv. : 126 183 246 Daylight: 74 (59%) 133 (73%) 208 (85%) Darkness : 52 (41%) 50 (27%) 38 (15%) Total classif . : 113 169 230 No. marked: 5 (4%) 6 (4%) 13 (6%) on 6 different days after the camera had been in operation at the site for approximately 5-6 weeks. Four other marked deer were photographed at the same sites during spring 1989 with second and third visits occurring 3-4 weeks following deployment of cameras. In addition to human activity, other wildlife identified in photographs included mule deer (O^. hemionus) , elk (Cervus elaphus) , and coyotes ( Canis latrans) . The number of system check exposures was reduced from session 1 to session 3 (Table 2) as a result of increasing confidence in the system and function of individual units. Most exposures caused by unknown variables were presumed the result of animals moving out of the field of view during the 2-3 second lapse between detection by the IR sensor and shutter release. The proportion of individual deer photographed during nighttime decreased from early to late winter (Table 2) . This suggested a transition from daily activity rhythms balanced with 9 diurnal and nocturnal activity near the time of winter migration to a strongly diurnal rhythm by late winter. This probably was not a behavioral response to cameras because individual sites at BTWR differed between sessions and only one session was conducted at MDWR. Winter migration onto BTWR occurred during session 1 as determined from estimated dates of arrival among 33 radio- collared deer that wintered on the survey area. Eleven (33%) deer, including 3 yearlong residents, were on the area when camera sites were set up on 18 and 19 December 1989. Fourteen (42%) moved onto the area between 16 and 31 December coinciding with the session, and the remaining 8 (25%) either moved onto the area after 1 January 1990 or did not move onto the winter range. The relatively high rate of deer visits on day 5-9 of session 1 probably coincided with movement of deer onto the winter range (Fig. 4) . The overall lower rate of daily visitation to camera sites by deer during session 1 than during session 3 could be at least partially explained by the fact that no more than 75% of the winter population was there at the time of the early winter session. Deer-visit Index vs . Time A 3x4 contingency analysis indicated that the number of deer-visits per day at camera sites was not independent of time (X = 17.43, 6 df, P < 0.01). Deer-visits per day appeared to decrease with increasing number of days from the beginning of the session (Fig. 4) . These results are from all 3 sessions combined. Camera noise, flash, etc. may have modified animal behavior to an extent that they used alternate trails during daily travel on subsequent days. There was little apparent difference in nocturnal and diurnal exposures over a session which was not surprising because 23% of daytime exposures also used a flash. As mentioned previously, operation of camera units beyond a 2 -week period apparently allowed deer to become accustomed to units and individuals routinely used the same travel routes and were repeatedly photographed. One would have to question the desirability of elimination of the camera- influenced bias if their response to camera use were predictable from session to session. One alternative might be to determine the response of deer to sites baited with aromatic attractants compared to unbaited sites. At this point, it is uncertain as to whether it is necessary to eliminate the temporal bias to use the deer-visit per day as an index of population trend. 10 5 0 5 10 15 20 NO. DAYS SINCE BEGINNING Figure 4. Trend in deer-visits per day during 3 remote camera surveys during winter-early spring 1989-90. Population Characteristics Composition by Sex and Age Composition of the population on the BTWR during winter and in the Tally Lake District during early spring appears in Table 3. Recruitment (proportion of fawns in late winter/spring populations) appeared slightly lower than in 1989 but still higher than that reported by Mundinger and Riley (1983) during the early 1980 's. Use of cameras to classify deer may minimize some of the bias associated with differential observability between sex and age groups (Downing et al. 1977). There was little apparent change in herd composition with respect to fawns and adults from winter to early spring. However, a lower proportion of fawns in classifications during May compared to February-March probably reflected difficulty in distinguishing fawns from adults or perhaps some undetected fawn mortality occurred during or following spring migration. 11 o •I-l ■U u CD T3 C cO X CO W o p E o o T5 >-l CD o o aQ o tn O 4-> r-l r-l •• 3 cO O o (J 0) >-i 3 b0 a> ■-I i-l U cfl >" E ■u u 3 CD T3 M-i i — i < Al tfl (D U tH CD t*-l E w 3 w Z to r-l O CO 4-1 O CD VO Ox r-l 00 vO VO m m CO vo m CM 1 a\ r^ >-> >■ r-l X! r4 Xi 1 x: Xi x; rH a> a. Ov a o a a Cfl 00 cO 4-) cO 00 cO cO cfl OV CD !-( CJ\ a. ON u u }-l tfl i-l r-l bO oo o bO r-l b£) r-l i-H bO bC CD o O CTv o O O o O o O T> XI •l-l U r-l ••-I 4-1 C 4-> >-l •r-l 4J 4J CD X! O r-l o cO o CO X! o OV O r-l Cn CD XJ x: 0) X! —> X! a) X! rH X! O \ > P-i o ac \ P-t \ > P-i C C S-l O X> r^ M cO CO ID a) CO •-> SS a fa s CD bO cfl tw O r4 Cfl CD r^ 12 Reproduction Radioimmunoassay for pregnancy-specific protein B (PSPB) in serum from 212 female white-tailed deer during 1988-90 indicated that yearlings were essentially the youngest breeding age class (Table 4). Only 3 (5%) of 61 female fawns had successfully bred. Among 151 older females, 93% of the yearlings and 95% of the adults were pregnant. These rates were comparable to those reported for whitetails in the Swan Valley (Mundinger 1981) and in eastern Montana (Dusek et al. 1989) . Six females (2 fawns, 4 adults) were necropsied to determine reproductive status during the report period. Neither fawn was pregnant, whereas all those >1 year of age were pregnant. A yearling carried a single fetus, and 3 older females carried twins. Only 3 8 (14%) of 129 females classified (marked and unmarked) in and around the Tally Lake District during July/August 1989 were accompanied by fawns. Eleven does had a single fawn at-heel, and 7 were accompanied by twins resulting in an estimated ratio of fawns : producing female of 139:100. Approximately 34% of adult females had fawn(s) at-heel during September/October 1988 (Dusek 1989) . Reproductive success was determined from reobservation among 25 of 106 individually marked females (>1 yr.) associated with both BTWR and MDWR. Fifteen (60%) were accompanied by fawn(s) that included none of 5 yearling females. Three adult females were observed with twins at-heel, whereas the other 12 were accompanied by a single fawn. These observations were made from August 1989 through May 1990, but should represent a minimum proportion of females rearing fawns to an age of weaning (about 4 months) . Although it is not likely that reproductive success among adult females can be estimated from classification of deer during summer-early autumn, productivity among producing females determined during that period may serve as an index of early reproductive success. Survival /Mortality Survival rates of radio-collared deer by season and age appear in Table 5. Survivorship was not determined for fawns and yearling males in both study areas or for adult males at MDWR because radioed individuals among these groups were not available for study throughout the mortality year. Annual survival among adult females (>2 yrs., Table 5) appeared lower than that among yearling females (0.89-1.00). Throughout the mortality year (1 May-3 0 April) , lowest survival rates were observed during the period of rifle hunting (16 Oct- 30 Nov) and the posthunt period (Dec. -Apr.), respectively (Table 5) . Nine deaths were documented among radioed deer during the 13 Table 4. Age specific pregnancy rates of white-tailed deer from serum assay among 212 captured females, 1988-90. a Age Fawns Yearlings Adults Year bred n % preg. n % preg. n % preg. 1987 11 0 3 100 13 92 1988 27 7 14 100 60 95 1989 23 4 12 83 49 96 All years 61 5 29 93 122 95 a Serum samples were taken during late December through early March on the Bowser-Tally and Murphy-Dickey Lakes winter ranges . past year in addition to 8 the previous year. Two deaths included adult males taken by hunters during autumn 1989. Three of seven females died during the period of rifle hunting; 2 were taken by hunters and cause of death was undetermined for the other. The specific cause of death was not determined from 4 females that died during the pre- and posthunt periods. Annual survivorship of females >2 years was 0.84 for those associated with BTWR and 0.77 for those associated with MDWR. However, the difference between areas was not significant (Z = 0.75, P > 0.20) . Twelve nontrap-related deaths of neckbanded deer were reported during the past year of which 7 resulted from hunting. Two deaths resulted from collisions with automobiles, 2 from predation presumably domestic dogs, and the cause of one death was undetermined. Condition Parameters Whole weights determined from heart girth during January/February 1989-90 increased with age through at least 4 years among both sexes of deer (Table 6) . A relationship of heart girth and whole weight of deer in Virginia (Smart et al. 1973) slightly overestimated weights of whitetails in northwestern Montana. Scale weights of 8 deer including both sexes among all ages averaged 1.76 kg less than estimates from 14 CO cd tp c (0 1 5-1 cd ■P r-l •rH Q) £ -P 5 C •H a) U CO ro d) rH M rH (d 0 o 1 >1 o cd •H ^ T3 0 ITS -H r-l Q 1 cr> >i c x: o a 6 r-l (0 3 s rH > c •H (0 o £ CO rj o r-l a a) x: rH -P 1 >i -p •H T3 C (0 ■P (0 H 0 o CO T3 CO Q) in 0/ u CD CD n c •H x» 1 o o c XI -H (GAP Eh +J T( 0) C X 0) w (0 e r-l > Al e 0) >h > CO Al Q e CD i CO IT) co VO CM H CO VO iH H (M eg CO CM IT) CO VO • * o o o I If) o o -4< VO o o o o o o CO CO o If) o CO VO • • • H o o CO CO CO • • • o o o in • if) • O o o • rH o en O CO CD o CO o ■P • • • • • f0 o r» if) o O H o iH l-» 1 T) o H CD CO ■P • <0 H o 0 C u ■p "> 0 rH z CD 4J 1 > 0 ■p CD O 0 "rH rH o in CD rH H VO i 0 H 0 C rH > >i CD CD CO -H (0 a T3 X > £ S -P •H «J •H C -i •H -P (0 u rH X! CD T3 fd (0 X! rH -P u o (C 0 rH c CD 8 1989 10 57.7 1.5 1 70.4 1990 7 63.3 1^9 heart girth for the same deer. Dressed carcass weights from 60 hunter-killed deer examined at check stations during autumn 1989 indicated an age-related pattern similar to that of heart girth measurement (Fig. 5) . Knowlton et al. (1980) reported that wild deer continue to gain weight through 4 years of age in females and 5-6 years in males. Only data from January and February 1989 were used for comparison with 1990 when weight estimates were obtained only for those months. Data within sex and age classes were pooled for January-February because a 4 factor ANOVA indicated no change in weight (P > 0.10) through the period, but main effect differences due to sex, age, and year were significant as were the 16 interactions of sex and age (P < 0.01). Sex, age, and year accounted for 93% of the variation in weight during January and February. Deer probably experienced some weight loss between late autumn and January, a period that included the onset of cold winter weather and migration from summer to winter ranges. During January-February, weight loss may have been somewhat more gradual and also may have continued beyond the period of trapping as suggested by general appearance of deer from photos taken during March and April (Dusek 1989) . Whole weights of harvested deer were estimated from dressed weights using a dressing index, or dressed weight expressed as a proportion of the whole weight, from deer specifically collected for study and from deer that died incidental to trapping (n = 15, x = 0.72 ± 0.01 SE) . Based on these estimates for October- November and estimates based on heart girth measurement during January-February, apparent weight loss of fawns was 18 and 17% for females and males, respectively. Females older than fawns experienced weight losses of 10-13%. For yearling, 2-year-old, and older males, observed weight losses were 21, 29, and 40%, respectively. An initial, rapid loss of weight, followed by a leveling off or more gradual weight loss through winter, was 17 consistent with physiological and behavioral adaptation to winter in northern environments where forage resources occur in limited abundance and/or quality (Peek et al. 1990). Adult males normally experience weight loss during late autumn-early winter, as a result of breeding behavior, regardless of abundance and quality of forage resources (Woolf and Harder 1979) . Analysis of blood parameters from samples taken during December 1988 to March 1989 indicated that serum urea nitrogen (SUN) , packed cell volume (PCV) , and hemoglobin (Hb) varied by month, PCV and Hb differed between fawns and adults, and none showed differences between males and females (Dusek 1989) . However, all suggested a decline in dietary status from December 1988 through February 1989 despite no apparent weight loss through that period. A 4 factor ANOVA indicated that SUN values differed significantly between years (P < 0.01), but main effects due to age, sex, and month were not significant (P > 0.10). Mean values of pooled samples were 30.3, 25.4, and 32.4 mg/dl for 1988, 1989, and 1990, respectively. Diastemal length of whitetails during autumn 1988 increased at least through 4 years of age in both females and males (Dusek 1989) . Mean diastemal length of yearling males examined at_ hunter check stations in 1989 was less than that for 1990 (x = 72.3, 69.4 mm, 58 df, t = 3.27, P < 0.01). Rationale for comparing measurments among yearling males followed that of Swenson and Stewart (1982). Yearlings had experienced only 1 winter, thus minimizing cumulative effects of environmental conditions on growth and condition over several years. Yearling males also were numerically abundant (n = 29, 31) among deer examined. Mean basal diameter and mean length of the main beam, and maximum spread among males >1 year of age increased with age (Dusek 1989) . With addition of 1989 check station data, ANOVA indicated differences attributable to age (P < 0.01) and year (P < 0.05) for both main beam length and basal diameter. Mean beam length was greater among all age classes of bucks in 1988 than in 1989 particularly among deer <2 years of age (Fig. 6) . Among 32 yearling males examined during autumn 1989, both antlers among 22 (69%) consisted of a single tine, or "spike", 6 had a maximum of 2 points/side, and 4 had a maximum of 3 points. Thirteen of 19 2-year-olds had a maximum of 3 points/side, 2 were 4-pointers, and 4 were 2 -pointers. Among 21 males older than 2 years, 9 carried a maximum of 4-5 points. 18 1988 H 1989 MAIN BEAM LENGTH (cm) 1 2 3 4+ AGE Figure 6. Main antler beam length (cm) of white-tailed deer bucks examined at check stations in 1988 and 1989. Population Density A 2 4 -km2 area surveyed by helicopter, remote camera surveys, and pellet group counts in 1989 (Dusek 1989) was used to estimate population size again in 1990. All trapping effort in 1990 at BTWR was conducted within this survey area bounded on the east by the Stillwater River, on the south and west by Lost Creek, and on the north by Hanson Lake. Camera placement in the area during late winter 1990 is shown in Figure 7. Remote camera and snowmobile surveys during February/March 1990 suggested that 6-8%, respectively, of the late winter-early spring population on BTWR was marked. A minimum of 134 marked deer occurred within the survey area that included all radioed deer using the survey area during late winter, all neckbanded deer captured within the area during winter 1990, excluding known mortalities, and neckbanded deer from previous years known by reobservation or recapture to occur within the area. Mortality was assumed minimal from December through mid-March, and movement by deer into or out of the area also was assumed minimal. 19 Figure 7. Winter distribution of white-tailed deer on the Bowser-Tally Lakes winter range (•) and location of camera sites (X) during late winter 1990. 20 Monte Carlo results yielded a MLE of 2,189 and a 95% likelihood interval of 1,920-2,539 on the survey area (Fig. 8). This compared to a raw Lincoln-Peterson estimate of 2,067 deer. A Lincoln-Peterson estimate for late winter 1989 was 2,057 (Dusek 1989) . 2 ... Deer density varied from 32 to 203 deer/km among individual pellet group transects. Pellet group density was converted to deer density based on a standard daily defecation rate of 12.7 for white-tailed deer (Longhurst and Connolly 1982) over a mean 98-day period that deer occupied the winter range. Sampling intensity was less in the low density stratum along bottomlands of the Stillwater River east of the farm-to-market road (Dusek 1989) . A high density stratum included rolling uplands and a series of potholes west of the farm-to-market road. Analysis of pellet group data resulted in an estimate of 3,064 deer in the survey area with an average density of 128 deer/km compared to 133/km in 1989. One (3%) of 33 radio- collared deer captured in the survey area during previous years, wintered off the survey area in 1990. Pellet group surveys probably overestimated deer numbers on the survey area. This may be partially due to the fact that tilled sites along bottomlands of the Stillwater River were not sampled. More importantly, the commonly used daily defecation rate, «13, determined from penned deer (Eberhardt and Van Etten 1956) was approximately half that reported for free-ranging deer (Rogers 1987, Sawyer et al. 1990). Rogers (1987) also implied that daily defecation rates are subject to regional variation as well as that attributable to sex and age. There are no comparative data for white-tailed deer in the northern Rocky Mountain region. Estimates derived from pellet group counts may serve as an index sensitive to population change exceeding a level of precision at current sampling intensity. Based on a sampling effort of 650 plots in the 24-km survey area, precision was ±13% (a = 0.05). Upper and lower limits were 3,450 and 2,678 animals, respectively. Precision for the high density stratum was ±12% (550 plots) , while that for the low density stratum was ±43% (100 plots) . Even considering that all deer associated with this winter range in 1989 did not use the winter range in 1990, a measurable change in population size was not evident between the 2 years from pellet group counts. Likewise, capture-resight procedures suggested nothing more than a negligible increase in wintering population between years. 21 45- 1920 -269 (95% LI) 2189 ♦ +350 (95% LI) 2539 40- 35 30- 25- 20- 15- 10- 5- • • • • mmm mm • • »• mt mm • • • • • • • •••••• • • • mm • m • M « ••mmm mm • • • m mm mm mmm mmm « <••»•.. mm mm>mm m mm** m mmm mmmm m m m m mm* mmm mm mm m • • • • • a • • • • mmm m mmm •• • •• •••• m mm m • ■ • • a mmmm • • • • t mmm m m m mm • mm m » m mmm» • mm Raw L-P = 2067 Bailey's Binomial = 1939 ±916 (95% CI) 1 022 <1 938 < 2855 No. Tagged IDs = 134 120 IDs seen 0 times 14 IDs seen 1 times Total Sightings = 216 95% "Fraction" = 6 075 »•• • • o o un CO o o O un at 1 1 1 1 I o 1 1 | 1 1 o i i | i i O 1 ' | 1 1 O 1 1 1 1 1 o o O o o m o lO o o o CM CM CO CM CM CM CM CM CM CM 8 o O O O o o O O O o to S LO o LO O LO O ■<*■ in to CO r>- 00 CM CM CM CM CM CM CM CM CM Number of Total Deer Figure 8. Monte Carlo computation of the maximum likelihood estimate and 95% liklihood interval compared with the binomial and Lincoln-Petersen estimates. Movements and Distribution This discussion is devoted to documenting distributional patterns of deer on the BTWR and MDWR and throughout the Fortine District. That on the Tally Lake District has been summarized by Morgan (1990) . General Movement Patterns The extent and timing of movement by radio-collared deer off the BTWR in late March-early April 1990 to transitional and summer ranges followed the pattern described by Mundinger and Riley (1982, 1983). Distance between geographical centers of activity for winter and summer home ranges varied from 23 to 31 km (Dusek 1989) . Major areas of concentration during late spring through autumn included both Good Creek and Star Meadows. Radio-collared deer moved onto the winter range from October 1989 through early February 1990 although most moved onto BTWR 22 during mid-to-late December as mentioned previously. The early migrants included 3 adult females captured in Griffin Creek during August 1989. Two of these (90243, 90246) wintered in Rhodes Draw (Fig. 7) , and the other (90245) wintered on the BTWR survey area. As in other years, spring migration was not characterized by the staggered pattern of movement observed during autumn/winter 1989-90. Estimated dates of departure from the winter range by radio-collared deer suggested that spring migration occurred about a week earlier in 1990 than in 1989. In 1990 the first evidence of spring migration was observed on 30 March. All but 2 spring migrants left the winter range before 10 April. One adult female (88016) left in early May as she had in 1988 and 1989, and one adult male (90310) left BTWR in early June 1990. Among 3 females radio collared on the north side of Ashley Lake in Ma^ch 1988, one of 2 survivors (88064) remained there throughout the year; the other (88063) summered in Griffin Creek immediately south of Star Meadows a pattern similar to that in 1988. Most deer wintering on MDWR summered primarily within the Fortine Ranger District. Only one of 18 radio-collared females during spring 1989 was a yearlong resident on the winter range. However, another female (89201) , captured and radio-collared in Stewart Creek in February 1989, remained near the capture site throughout most of the year; she moved onto the more intensively used portion of the winter range behind Murphy Lake in February 1990 and remained there through March. Furthest movement from the winter range in 1990, as in 1988 and 1989, included an adult female (88043) that summered in Grand Creek at the base of Elk Mountain, a straight-line distance of 29 km from her winter home range. Although the winter trapping effort in 1990 included a larger portion of the winter range than during the previous 2 winters, distribution the following spring was still similar to that of summers 1988 and 1989. Fifteen of 19 deer occurred in the Fortine Creek drainage though all but 2 were in second and third order drainages. The meadow area at the confluence of Swamp and Lake Creeks was again a major concentration area during late May. Density Distribution Density distribution of deer on BTWR generated from pellet group counts (Fig. 9) indicated the greatest area of winter concentration occurred in an area of low ridges interspersed with potholes from Cliff Lake south to Bowser Lake, bounded by Pete Ridge on the west and on the east by the ridge just northeast of Baney Lake. Lowest densities, as mentioned previously, occurred 23 2 Figure 9. Deer density (deer/km ) contours on the Bowser-Tally Lakes winter range from pellet group counts during April 1990. 24 along bottomlands of the Stillwater River that included the Kuhns WMA (Fig. 7) . Relative abundance from remote camera surveys expressed as deer-visits/day (Table 7) suggested a similar pattern of density distribution. Remote camera surveys at MDWR (Fig. 10, Table 7) suggested comparatively high deer abundance from Murphy Lake east to Martin Lake, along Cripple Creek, and around Ant Flat. Although white- tailed deer may occasionally use the upper Summit Creek drainage as indicated from locations of radio-collared deer, only mule deer were photographed at that particular site. Table 7. Density distribution of white-tailed deer on Bowser- Tally and Murphy-Dickey Lakes winter ranges during February/March 1990 from deer-visits per day at camera sites. Winter Site No . days Deer-visits/ range No. Location functioning day MDWR 9011 N. Hagadore Lake 15 1.9 9012 Martin Lake 15 0.9 9013 Sink Creek 15 0.0 9014 S. Hagadore Lake 15 0.4 9015 Murphy Lake 15 2.7 9016 Ant Flat 15 1.2 9017 Cripple Creek 14 2.0 9018 E. Dudley Slough 14 1.0 9019 E. Dudley Slough 14 0.9 9020 E. Dudley Slough 14 0.6 BTWR 9021 S. Kuhn's WMA 16 0.4 9022 N. Kuhn's WMA 16 0.8 9023 State Forest 16 2.2 9024 Cliff Lake 14 2.0 9026 Lore Lake 15 2 . 1 9028 E. Pete Ridge 14 1.4 9029 E. Bowser Lake 9 2.9 9030 Pete Ridge 10 4.1 25 Figure 10. Winter distribution of white-tailed deer on the Murphy-Dickey Lakes winter range ( • ) and location of camera sites (X) during late winter 1990. 26 LITERATURE CITED Clover, M. R. 1954. A portable deer trap and catch net. Calif. Fish and Game 40:367-373. Davis, D. E. 1982. Calculations used in census methods. Pp. 34 4-369 in D. E. Davis, ed. Handbook of census methods for terrestrial vertebrates. CRC Press, Inc., Boca Raton, FL. Downing, R. L. , E. D. Michael, and R. J. Poux. 1977. Accuracy of sex and age ratio counts of white-tailed deer. J. Wildl. Manage. 41:709-714. Dusek, G. L. 1989. Population ecology of white-tailed deer in northwestern Montana. Job Prog. Rep., Mont. Dep. Fish, Wildl. and Parks, Helena. Fed. Aid Pro j . W-100-R-2. 26 pp. , R. J. Mackie, J. D. Herriges, Jr., and B. B. Compton. 1989. Population ecology of white-tailed deer along the lower Yellowstone River. Wildl. Monogr. No. 104. 68 pp. Eberhardt, L. , and R. C. Van Etten. 1956. Evaluation of the pellet group count as a deer census method. J. Wildl. Manage. 20:70-74. Heisey, D. M. , and T. K. Fuller. 1985. Evaluation of survival and cause-specific mortality rates using telemetry data. J. Wildl. Manage. 49:668-674. Janke, D. 1977. White-tailed deer population characteristics, movements, and winter site selection in western Montana. M.S. Thesis, Univ. Montana, Missoula. 92 pp. Knowlton, F. F. , M. White, and J. G. Kie. 1980. Weight patterns of wild white-tailed deer in southern Texas. Proc. Welder Wildl. Found. Symp. 1:55-64. Krahmer, R. W. 1989. Seasonal habitat relationships of white- tailed deer in northwestern Montana. M.S. Thesis, Univ. Montana, Missoula. 104 pp. Leach, R. H. 1982. Summer range ecology of white-tailed deer in the coniferous forests of northwestern Montana. M.S. Thesis, Univ. Montana, Missoula. 80 pp. Longhurst, W. M. , and G. E. Connolly. 1982. Deer (pellet count) . Pp. 247-248 in D. E. Davis, ed. Handbook of census methods for terrestrial vertebrates. CRC Press, Inc., Boca Raton, FL. 27 Mace, R. , T. Manley, and K. Aune. 1990. Use of systematically deployed remote cameras to monitor grizzly bears. Job Prog. Rep., Mont. Dep. Fish, Wildl. and Parks, Helena. 25 pp. Minta, S., and M. Mangel. 1989. A simple population estimate based on simulation for capture-recapture and capture- resight data. Ecol. 70:1738-1751. Morgan, J. T. 1990. Summer habitat use of white-tailed deer on the Tally Lake Ranger District of the Flathead National Forest. Pp. 30-47 in Population ecology of white-tailed deer in northwestern Montana. Job Prog. Rep., Mont. Dep. Fish, Wildl. and Parks, Helena. Fed. Aid Pro j . W-100-R-3. Mundinger, J. G. 1981. White-tailed deer reproductive biology in the Swan Valley, Montana. J. Wildl. Manage. 45:132-139. . 1984. Biology of the white-tailed deer in the coniferous forest of northwestern Montana. Pp. 275-284 in W. R. Meehan, T. R. Merrell, Jr., and T. A. Hanley, eds. Proc. fish and wildlife relationships in old growth forests. Symp. Am. Inst. Fish Res. Biol. , and S. J. Riley. 1982. Population ecology and habitat relationships of white-tailed deer in coniferous forest habitat of northwestern Montana. Pp. 50-66 in Montana deer studies. Job Prog Rep., Mont. Dep. Fish, Wildl. and Parks, Helena. Fed. Aid Proj . W-120-R-13. , and . 1983. Population ecology and habitat relationships of white-tailed deer in coniferous forest habitat of northwestern Montana. Pp. 49-63 in Montana deer studies. Job Prog Rep., Mont. Dep. Fish, Wildl. and Parks, Helena. Fed. Aid Proj. W-120-R-14. Peek, J. M. , R. J. Mackie, and G. L. Dusek. 1990. Over-winter survival strategies of North American cervidae. Proc. Int. Symp. on Moose, Syktyvkar, U.S.S.R. 3: In press. Pfister, R. D. , B. L. Kovalchick, S. F. Arno, and R. C. Presby. 1977. Forest habitat types of Montana. USDA, Forest Service, Gen. Tech. Rep. INT-34. 174 pp. Rogers, L. L. 1987. Seasonal changes in defecation rates of free-ranging white-tailed deer. J. Wildl. Manage. 51:330-333. Sawyer, T. G, R. L. Marchinton, and W. MacLentz. 1990. Defecation rates of female white-tailed deer in Georgia. Wildl. Soc. Bull. 18:16-18. 28 Severinghaus, C. W. 1949. Tooth development and wear as criteria of age in white-tailed deer. J. Wildl. Manage. 13:195-216. Smart, C. W. , R. H. Giles, Jr., and D. C. Guynn. 1973. Weight tape for white-tailed deer in Virginia. J. Wildl. Manage. 37:553-555. Swenson, J. E., and S. T. Stewart. 1982. On the use of population condition indices in deer management. Pp. 28-3 5 in C. D. Eustace, ed. Practical application of recent research. Proc. Montana Chapt., The Wildl. Soc, Billings. Woolf, A., and J. D. Harder. 1979. Population dynamics of a captive white-tailed deer herd with emphasis on reproduction and mortality. Wildl. Monogr. No. 67. 53pp. Submitted by: Gary L. Dusek 29 Summer habitat use of white-tailed deer on the Tally Lake Ranger District of the Flathead National Forest Job Objectives: 1. Determine season long and diel activity and habitat use patterns of white-tailed deer on summer ranges on the Tally Lake District of the Flathead National Forest. 2 . Determine use and importance of various serai stages of coniferous forest and riparian communities and how spatial distribution of these communities to form habitat complexes influences distribution and abundance of deer. 3 . Determine the importance of various habitat features such as: slope, aspect, elevation, vegetative structure and species composition of forest stands, and distance to cover, riparian areas, and roads. INTRODUCTION White-tailed deer summer use of coniferous forest in northwest Montana has been study previously by Leach (1982) , Mundinger (1984), and Krahmer (1989). The Tally Lake District northwest of Kalispell, initially studied in the early 1980' s, has been shown to be important as whitetail summer range (Mundinger and Riley 1982, 1983). However, extensive timber harvesting and road building on the district could potentially disrupt traditional patterns of whitetail activity and habitat use on summer ranges. This study was initiated to investigate whitetail activity and habitat use while deer occupy summer ranges on the Tally Lake District as well as diel patterns during that period. This report describes research activities during the first of 3 summer field seasons (1 May-30 November 1989) . Major effort during the period concentrated on familiarizing myself with the study area, testing summer trapping techniques to increase the number of radio-collared deer in specific areas, and obtaining relocations of deer through aerial, ground, and 24-hr telemetry efforts. 30 STUDY AREA The Tally Lake District has been described previously by Mundinger and Riley (1982) and Dusek (1989) . My study area includes that portion of the district containing all deer relocations and outlined by major physiographic points north of Ashley Mountain, east of the Flathead/Lincoln County Line, south of Martin Falls, and west of Tally Lake. The study area drains to the northeast into the Stillwater River via Good, Logan, and Martin Creeks (Fig. 1) . Elevation ranges from 1020 m at Tally Lake to 1935 m on Mount Swaney. Subalpine fir/gueen's cup beadlilly (Abies lasiocarpa/Clintonia unif lora) is the major habitat type (Pfister et al. 1977) throughout the study area. However, habitat alteration through logging, cattle grazing, and natural fires has produced a forest which is a mosaic of mature mixed conifer, large stands of lodgepole oine (Pinus contorta) , new clearcuts, various stages of regrowth, riparian meadows, and other natural openings. Major species present in the overstory include lodgepole pine, Douglas-fir (Pseudotsuga menziesii) , subalpine fir (Abies lasiocarpa) , and western larch (Larix occidentalis) . Common grass, forb, and shrub species include pine grass (Calamagrostis rubescens) , timothy (Phlem pratense) , strawberry (Fraqaria virginiana) , yarrow (Achillea millefolium) . arnica (Arnica spp.), beargrass (Xerophyllum tenax) , spiraea (Siraea betulifolia) , rose (Rosa gymnocarpa) , twinf lower (Linnaea borealis) , buf faloberry ( Shepherdia canadensis) , alder (Alnus spp.), willow ( Salix spp.), snowberry ( Symphor icarpus albus) , and huckleberry (Vaccinium spp. ) . Two areas which support large numbers of deer throughout the summer are the Star Meadows complex formed by the confluence of Logan, Griffin, and Sheppard Creeks; and the Alder, Corduroy, Good Creek complex. Star Meadows is approximately 65 km . One- third is meadow bottom consisting of a mixture of open meadowlands, willows, and scattered timber. The slopes of the complex are a mosaic of timber and cutover areas. The Alder, Corduroy, Good Creek Complex is about half the size of Star Meadows. Contrastingly, the area primarily consists of large stands of 60-70 year old lodgepole pine, a remnant of large fires during the early part of this century. This complex lacks large meadows but a number of small wet meadows are associated with each drainage. METHODS Deer relocations were obtained via fixed-wing aircraft surveys, ground triangulation using a hand-held H-antennae, 24-hr telemetry sessions using 3 truck mounted null antennae arrays, and visual sightings either incidental or purposely obtained. 31 Figure 1. Tally Lake deer project study area, Tally Lake Ranger District, Flathead National Forest. 32 Test transmitters were placed at known locations to check accuracy of relocation procedures (White and Garrott 1990) . Date, time, and UTM coordinates were recorded for each relocation and plotted on aerial photographs and/or topographic maps. Streams, lakes, and prominent peaks were digitized using program CAPTURE (Desktop Digitizing Package 1988) . These data as well as aerial and ground relocations were plotted using program SURFER (Surfer Reference Manual 1988) . Summer home ranges were calculated using program TELEM (Coleman and Jones 1988) and plotted separately. An attempt was made to trap deer in specific areas on summer range in order to increase the number of radio-collared deer in areas where 24-hr telemetry sessions were being conducted. Three Clover traps (Clover 1954) baited with salt blocks and apples were placed around Star Meadows during August 1989. Trapping and handling followed procedures previously described (Dusek 1989) . RESULTS and DISCUSSION Thirty-four radio-collared deer moved from winter range on to the Tally Lake District and occupied their summer ranges by the beginning of the report period (Table 1) . During summer 1989, 4 additional deer were trapped and 471 ground and air relocations were recorded among 38 deer summering on the district, yielding information on distribution, movements, and habitat use. Distribution Summer distribution showed clustering of deer in certain parts of the study area with little documented use elsewhere (Fig. 2) . Twenty-three deer moved to the northern portion of the study area bounded by Martin Creek and associated drainages. These deer primarily used the Good Creek bottom and south facing slopes as a travel corridor and transitional area, moving to higher elevations as the snow melted. Four deer summered in each Alder and Corduroy Creeks and 6 deer in the Adams Mountain/Miller Creek area. Fifteen deer summered in the southern portion of the study area with 4 deer in both the Logan and Griffin portions of Star Meadows and 3 in the upper portion of Sheppard Creek. Initially, many deer around Star Meadows restricted their movements to the south facing slopes as much of the bottom was flooded during the spring. In June many deer made use of the meadow for foraging and fawning, retreating back to the slopes as the meadow dried out later in the summer. 33 Table 1. White-tailed deer monitored on Tally Lake District, summer 1989. DEER NUMBER SEX AGE SUMMER LOCATION 88001 F 6 LISTLE CREEK 88010* F 3 GERGEN CREEK 88014 F 5 GOOD CREEK 88016 F 3 GOOD CREEK 88019 F 5 GOOD CREEK 88020 F 2 MARTIN CREEK 88021 F 4 MARTIN CREEK 88063 F 9 GRIFFIN CREEK 89070 F 3 CORDUROY CREEK 89083 F 5 ADAMS MOUNTAIN 89084 F 4 CORDUROY CREEK 89087 F 4 LOGAN CREEK 89089 M 3 ALDER CREEK r f\ s\ f\ 89090 M 2 SHEPPARD CREEK 89092 F 8 ALDER CREEK n A A A c 89095 M 2 GOOD CREEK 89096 F 2 ALDER CREEK 89098 F 5 MILLER CREEK 89099 M 4 LISTLE CREEK 89100 F 3 FOX MOUNTAIN 89117 F 2 NELSON CREEK 89119 t» r 4 LUKUUKUl CREEK 89124 ■c r MILLER L.REEK 89134 F 9 CORDUROY CREEK 89148 F 4 LOGAN CREEK 89149 F 2 NORTH EVERS CREEK 89157b F 3 EVERS CREEK 89163 F 3 LOGAN CREEK 89188 F 3 GRIFFIN CREEK 34 Table 1, continued. DEER NUMBER SEX AGE oUflflhK LOCATION 89189 F 4 MARTIN CREEK 89195 F 4 ALDER CREEK 89197 F 7 SHEPPARD CREEK 89230 F 7 STAR FACE 89237 F 2 STAR FACE 89243d F 4 GRIFFIN CREEK 89244M F 6 STAR FACE 89245d F 3 GRIFFIN CREEK 89246d F 1 GRIFFIN CREEK °Lost contact, July 1989 Died, cause unknown, fall 1989 ^Harvested, November 1989 Trapped, August 1989 35 36 Elevational use by deer ranged from 1100 m at the confluence of Miller and Goods Creeks to 1585 m near the peak of Adams Mountain with some intra-seasonal movement. Little use was made of uplands exceeding 1600 m within the center of the study area, the Sylvia Lake/Hand Creek area in the southeast, the upper portions of Logan Creek, or the lower portions of Logan and Good Creeks. The latter however were used as transitional areas. Home Range Once deer arrived on their summer home ranges there was very little movement until fall migration (Fig. 2) . Two deer which spent May-July around the northern portion of Star Meadows moved to Tally Mountain in Early August. Although this move was included as part of the summer home range it should more accurately be considered an accessory area. Summer home ranges of all dee^, excluding the 2 above, and based on 90% of relocations, averaged (SE) 96 ha (±70) . Habitat Use Figure 2 illustrates an association of whitetails with riparian areas within summer home ranges as that of almost every deer encompassed a creek drainage. Many drainages only flow intermittently or move underground for part of their course. However for the most part they are mesic sites as are areas within the home ranges of those deer not associated with a specific creek. A vegetation cover map of the study area is being developed with the aid of Landsat imagery and program ERDAS (ERDAS, Inc. 1987) . Through general observations however, it should be noted that most individual deer relocations fell in timbered areas although often bordering another type including cutover areas. On the larger scale summer home ranges of individual deer encompassed areas of mixed timber, stands of lodgepole, cutover, and riparian areas. The complex of vegetation communities along with topographic features may determine deer distribution. Diel Patterns Twenty-four hour telemetry sessions were conducted twice in both July and August 1989. Usable data were obtained for 3 deer in Corduroy Creek and for 6 deer on the north side on Star Meadows. Diel home ranges for these deer averaged (SE) 119 ha (±106) , larger than the seasonal home ranges for the same deer (Figs. 3 and 4) . A major reason for 24 hour monitoring was to obtain data on deer movement at night when aerial locations could not be obtained. Night locations tend too increase overall 37 84- DG AJOUST 134-aa AUGUST home range size. Diel ranges also represent 1 day's movement out of the whole summer; on that given day it is possible that a deer was making non-typical movements (eg; an exploratory move) . Finally, any inherent system-related error can skew the locations and an effort has been made to reduce this. Summer Trapping During August 1989, 6 deer were captured over 36 trap nights (17%) and radio collars were placed on 4 females. On 9 occasions traps had been tripped with no capture. Evidence suggests that 4 trips were deer caused. This effort suggested that deer can be trapped during spring/summer at fairly high efficiency and this should improve with better trap placement and setting. Migration Some deer began exhibiting home range shifts as early as late July; 2 deer moved from Star Meadows to Tally Mountain where they remained through November then moved to their respective wintering areas on Pilot Knob and Pete Ridge. Two deer trapped in Griffin Creek during August 1989 moved to wintering areas in Rhodes Draw in mid-September. For the most part however, major moves did not occur until November-December. Movements between summer and winter range often occurred between aerial relocations (10-14 days) and hence getting transitional locations was difficult. Available evidence suggests that deer summering in the northern portion of the study area traveled down Good Creek to its confluence with Logan Creek (Fig. 5) . Along this portion of the drainage the topography is less severe and from here deer could move across Short, Round, and Birch Meadows to winter ranges in the Bowser, Kuhn's, Pete Ridge area. Most Deer summering in the southern portion of the study area including Star Meadows moved either over Tally Mountain or the Reid Divide to get to wintering areas. Deer began leaving the winter range near Bowser Lake in late March 1990. Deer moving to Star Meadows did so in as little as 3 days probably following routes similar to that of their fall migration. Deer associated with summer ranges in and around Good Creek often spent a few days in the meadows immediately north of Tally Lake and along the snow free portions of Good Creek. Year long elevational changes were similar to that found by Dusek (1989) . Highest elevational use occurred in July with some variation throughout the summer (Fig. 6) . Elevations on winter range were lowest in February. Elevational change was more rapid when deer moved to summer range between March and April than to winter range between October and January. 40 Figure 5. White-tail deer movements between summer and winter ranges 1989-90. 41 ELEVATION (m) 1600r 1400 1200 1000 800 , " * " * 4- + n*49 n=86 n«73 n-68 n-44 x n'66 . T n*106 rW6 n«4e n=54 n=59 n°43 J l l I l I l l l l l I. JUN JUL AUG SEP OCT NOV DEC JAN FEB MAR APR MAY MONTH Figure 6. Elevational change (mean + se) between summer and winter ranges 1989-90. FUTURE WORK 1990-91 Field Season The 1990-91 field season will be the most intensive as far as gathering data on deer locations. The anticipated field schedule includes locating deer from the air 3 times/month. Ground relocations will be limited as their accuracy has been questioned (Schmutz and White 1990) . Twenty-four hour monitoring will be conducted monthly in the Star Meadows and Corduroy Creek areas . Summer trapping will continue and be expanded to include Corduroy Creek as well as Star Meadows. The number of deer trapped will depend on the number of radio collars available. 42 Remote camera surveys will be conducted to experiment with using this method to estimate deer numbers as well as obtain deer relocations in specific drainages. Habitat data collection will begin in the 1990-91 field season. The basic procedure will be to develop a number of habitat component layers using ERDAS (ERDAS, Inc. 1987) . Layers will include: vegetation type, riparian areas, roads, slope, aspect, and elevation. The topographic layers are already complete and useable. The road and riparian layers are being updated and corrected. The vegetation layer must be created for the specific requirements of this project. The procedure to create the vegetation layer involves separating the study area into a number of vegetation types based on differences in reflective and infrared light wavelengths picked up by the satellite. Vegetation will be separated based on its possible influence on deer behavior. This will primarily be dependent on vegetation structure, ie: foraging areas or grassy openings, hiding cover or shrub/sapling stands, and mature timber. Foraging areas will be separated as cutover areas or natural openings. Hiding cover we be separated as natural willow/shrub areas or sapling regrowth after clearcutting. Mature timber will be separated as mixed conifer or mostly lodgepole. On ground data collection will involve walking into a randomly selected group of deer relocation sites. Information on species presence and relative abundance will be obtained for various layers of the forest. Canopy and horizontal cover will be noted as well as slope, aspect, and elevation. Average age, height, trees/ha, and basal area/ha for the stand will be obtained from Forest Service records. Two deer will be collected monthly to assist in food habits analysis. Rumens will be analyzed at the Wildlife Research Lab in Bozeman. 1991-92 Field Season During the summer of the 1991-92 field season, emphasis will still be on obtaining deer relocations. The schedule of aerial relocations and twenty-four hour monitoring sessions should remain near the 90-91 level. Summer trapping will be conducted in the Star Meadows and Corduroy Creek areas. Camera surveys also will be conducted around Star Meadows. Also during this time the vegetation cover type map will be completed and transferring the appropriate GIS files from the Forest Service to the Fish, Wildlife, and Parks computer system will take place. Most data analysis will take place after the summer field season has ended. 43 Data Analysis In habitat use studies there are typically 4 questions to be answered (White and Garrott 1990) . These concern the availability of various habitat components, the degree of use, any preference shown, and whether a particular habitat component is critical to an animal's survival or presence on an area. In this study we are concerned not only with the habitat components at individual deer locations but also with the complex of components making up the area surrounding a given location or group of locations. Hence these questions need to be asked about habitat complexes as well. In this study, availability of habitat components will be determined within the whole study area and within 2 subunits (ie; Star Meadows and the Alder, Corduroy, Good Creek complex) . Because the vegetation component has been digitized in cell format as a discrete number of types the availabilty of each will be determined by the summation of cells for that type. Availability of components such as slope, aspect, and elevation will be calculated by an analysis of random points. Availability of habitat in complexes is more difficult to quantifify due to the difficulty in assigning a size to the complex. One possible method would be to consider the area within the seasonal home range of a deer as a habitat complex. However, there are probably portions of that area which are not used and should not be included as available. The method I will use is to delineate the complex as a given area around each individual deer location. From testing of aerial relocations I know that the distance between the actual and recorded location can be off 50-150 m. A circle with a 150 m radius is approximately 7 ha which will be used as the complex area. The habitat composition of this area will then be compared to random circles. Utilization will be measured based on radio telemetry relocations. Habitat use on a seasonal basis will be determined by individual aerial relocations, visual sightings, and ground location to a lesser degree. Diel monitoring will be used to look for habitat use patterns throughout 24 hour periods. There are a number of approaches for determining preference or avoidance (Manly et al. 1972, Neu et al. 1974, Johnson 1980, Marcum and Loftsgaarden 1980, Heisey 1985) . The appropriate test to be used is partially determined by the type of data available. In general, when the actual proportion of each habitat component is known the procedure by Neu et al. (1974) has done well in comparison with the others (Alldredge and Ratti 1980) and will be the method used to evaluate use of vegetation types. When availability is estimated based on random points then the Marcum and Loftsgaarden (1980) approach is appropriate and will be used 44 on topographic data, as well as nearness to roads and riparian areas. After a preference is detected it is often desired to determine whether the habitat component is critical . As discussed by White and Garrott (1990) this cannot be determined by simply looking for preference. Rather a perturbation study with a large sample size needs to be conducted to see if the animal's condition or use of an area changes after the habitat is altered. In this study there are deer concentrated in 2 areas which are apparently quite different. Star Meadows is an area with great interspersion of the major vegetation types. While the Alder, Corduroy, Good Creek complex is more uniformly uncut timber. An appropriate test may be to alter one area and measure the deer response. Since Star Meadows cannot be brought back to mature timber, cutting timber in the Good Creek area is more realistic. Such an opportunity exists with the planned timber harvests in the Alder, Corduroy, Gergen Creek areas. A number of radio-collared deer are already using this set of drainages. Through aerial and diel relocations the present study will provide a great deal of pre-cutting data. Depending on when the stands are harvested post-cutting data may be obtained during this study or if not a follow-up or continuation study is highly recommended. LITERATURE CITED Alldredge, J. R. and J. T. Ratti. 1986. Comparison of some statistical techniques for analysis of resource selection. J. Wildl. Manage. 50:157-165. Clover, M. R. 1954. A portable deer trap and catch-net. Calif. Fish and Game 40:367-373. Coleman, J. S. and A. B. Jones. 1988. User's Guide to TELEM88: Computer analysis system for radio-telemetry data. Dept. Fish, and Wildl., VPI&SU, Blacksburg, Va. 49pp. Desktop Digitizing Package. 1988. Center for Remote Sensing and Mapping Center. Univ. Ga. Athens. 31pp. Dusek, G. L. 1989. Population ecology of white-tailed deer in northwestern Montana. Job Prog. Rep., Mont. Dep. Fish, Wildl., and Parks, Helena. Fed. Aid Pro j . W-100-R-2. 2 6pp. ERDAS-MV System, Image Processing System Users Guide. ERDAS, Inc. Atlanta. 45 Heisey, D. M. 1985. Analyzing selection experiments with log- linear models. Ecology 66:1744-1748. Johnson, D. H. 1980. The comparison of usage and availability measurements for resource preference. Ecology 61:65-71. Krahmer, R. W. 1989. Seasonal habitat relationaships of white- tailed deer in the coniferous forsts of northwestern Montana. M.S. Thesis, Univ. Montana, Missoula. 104pp. Leach, R. H. 1982. Summer range ecology of white-tailed deer in the coniferous forests of northwestern Montana. M.S. Thesis, Univ. Montana, Missoula. 80pp. Manly, B. D. J., P. Miller, and L. M. Cook. 1972. Analysis of a selective predation experiment. AM. Nat. 106:719-736. Marcum, C. L. and D. O. Lof tsgaarden. 1980. A nonmapping technique for studying habitat preferences. J. Wildl. Manage. 44:963-968. Mundinger, J. G. 1984. Biology of the white-tailed deer in the coniferous forest of northwestern Montana. Pp. 275-284 in W. R. Meehan, T. R. Merrell, Jr., and T. A. Hanley, eds. Proc. Fish and Wildlife Relationships in Old Growth Forests. Symp. Am. Inst. Fish. Res. Biol. , and S. J. Riley. 1982. Population ecology and habitat relationships of white-tailed deer in coniferous forest habitat of northwestern Montana. Pp. 50-66 in Montana deer studies. Job Prog. Rep., Mont. Dep. Fish, Wildl., and Parks, Helena. Fed. Aid Pro j . W-120-R-13. , and S. J. Riley. 1983. Population ecology and habitat relationships of white-tailed deer in coniferous forest habitat of northwestern Montana. Pp. 49-63 in Montana deer studies. Job Prog. Rep., Mont. Dep. Fish, Wildl., and Parks, Helena. Fed. Aid Pro j . W-120-R-14. Neu, C. W. , C. R. Byers, J. M. Peek, and V. Boy. 1974. A technique for analysis of utilization-availability data. J. Wildl. Manage. 38:541-545. Pfister, R. D. , B. L. Kovalchick, S. F. Arno, and R. C. Presby. 1977. Forest habitat types of Montana. USDA, Forest Service, Gen. Tech. Rep. INT-34. 174pp. Schmutz, J. A. and G. C. White. 1990. Error in telemetry studies: effect of animal movement on triangulation. J. Wildl. Manage. 54:506-510. Surfer Reference Manual. 1988. Golden Software Inc, Golden, Co. 46 White, G. C. and R. A. Garrott. 1990. Analysis of Wildlife radio-tracking data. Academic Press, New York. 383pp. Submitted by: John T. Morgan 47