PROXIMATE FACTORS LIMITING POPULATION GROWTH OF WHITE-TAILED DEER IN NOVA SCOTIA BRENT R. PATTERSON, 1, 2 Department of Natural Resources, Wildlife Division, 136 Exhibition Street, Kentville, NS B4N 4E5, Canada BRUCE A. MACDONALD, 3 Center for Wildlife and Conservation Biology, Department of Biology, Acadia University, Wolfville, NS B0P 1X0, Canada BEVAN A. LOCK,4 Center for Wildlife and Conservation Biology, Department of Biology, Acadia University, Wolfville, NS B0P 1X0, Canada DON G. ANDERSON, Department of Natural Resources, Whycocomagh, NS B0E 3M0, Canada LAWRENCE K. BENJAMIN, Department of Natural Resources, Wildlife Division, 136 Exhibition Street, Kentville, NS B4N 4E5, Canada Abstract: White-tailed deer (Odocoileus virginianus) densities in Nova Scotia, Canada, declined during the late 1980s and early 1990s. We estimated population change, survival, and relative importance of mortality factors for deer in 2 geographic areas of Nova Scotia from February 1994 to January 1999. Pellet-group surveys indicated that deer densities in both study areas increased slowly (λ ~ 1.07 and 1.05 in the Queens County and Cape Breton study areas, respectively) during the study. Annual survival rates of adult deer did not differ among years or between the study areas. Annual survival rates for adult females averaged 93.9 ± 4.3% (SE) and 80.4 ± 3.1% within and outside of Kejimkujik National Park, respectively, where harvest did not occur. Annual survival rates of adult males and fawns outside of the park, where hunting occurred, were 50.7 ± 7.6% and 36.9 ± 7.4%, respectively. No marked adult males or fawns died within the park during the study (winter only for fawns). Annually, hunting (34.2 ± 8.2% and 8.2 ± 2.4% for adult males and females, respectively) and predation (7.2 ± 5.3% and 7.5 ± 2.3% for adult males and females, respectively) were the largest mortality factors for adult deer. Coyote (Canis latrans) predation (27.5 ± 8.7% during Dec–May only) was most influential for fawns. Monte Carlo simulations involving a model of adult survival and fawn recruitment supported the results of the deer pellet-group inventories and suggested low but positive rates of increase for both populations. Although predation and the unregistered harvest of adult females probably slowed the growth of deer populations following the decline, the recent establishment of a zone-based antlerless harvest quota system should allow managers to regulate deer numbers by annually adjusting the number of antlerless permits in response to estimates of hunting and nonhunting losses. JOURNAL OF WILDLIFE MANAGEMENT 66(2):511–521 Key words: Canis latrans, coyote predation, hunting, illegal harvest, Monte Carlo simulation, Nova Scotia, Odocoileus virginianus, population limitation, white-tailed deer. By definition, all mortality sources must limit population growth somewhat, and thus can be considered limiting factors. However, only those limiting factors whose magnitude increases with deer density can promote stabilization of deer abundance (i.e., act as a regulating factor). Most studies of white-tailed deer survival have concluded that hunting is the major cause of mortality in exploited populations (Nelson and Mech 1986a, Fuller 1990, Dusek et al. 1992, Van Deelen et al. 1997). As such, hunting is the mortality source targeted most often in deer management plans. Typically, agencies increase the number of available antlerless deer permits as deer densities increase. In this sense, harvest can be density dependent and can potentially regulate deer numbers (Fuller 1990, Patterson and Power 2002). Predation can be a major source of mortality for deer along the northern extent of their geographic range in North America (Nelson and Mech 1986a, Poulle et al. 1993, Whitlaw et al. Eastern Canada is the northeastern extent of the geographic range of white-tailed deer in North America. Deer densities in this area are lower than those throughout more southern areas (Halls 1984, Crête 1999). Across North America, potential rates of increase of deer populations are limited by reproductive rates and fawn survival (McCullough 1979, Fuller 1990); however, the relative roles of various factors limiting deer at low densities along the northern extent of their geographic range are less clear. 1 Present address: Ontario Ministry of Natural Resources, Wildlife Research and Development Section, 300 Water Street, Third Floor North, P.O. Box 7000, Peterborough, ON K9J 8M5, Canada. 2 E-mail: [email protected] 3 Present address: Ducks Unlimited Canada, 5017-52 Street, Yellowknife, NT X1A 1T5, Canada. 4 Present address: Stora Enso Port Hawkesbury Ltd., Box 59, Port Hawkesbury, NS B0E 2V0, Canada. 511 512 J. Wildl. Manage. 66(2):2002 LIMITING FACTORS FOR DEER • Patterson et al. 1998). Crête (1999) suggested that where wolves are present, cervid densities in northeastern North America generally are limited by predation at a density below that at which food competition becomes significant. The proportion of deer removed by coyotes in Nova Scotia decreased with increasing deer densities; thus, coyote predation is likely to destabilize, rather than regulate, deer densities (Patterson 1999). The abundance of most northern deer populations is likely regulated to some extent by forage competition (Fryxell et al. 1991, Messier 1991, Post and Stenseth 1998, Dumont et al. 2000, Patterson and Power 2002). However, because of a time lag between deer density and the resulting depression of deer population growth, this regulation may result in marked fluctuations in abundance. Although deer population growth may be limited directly by winter malnutrition losses (Potvin et al. 1981, Case and McCullough 1987), winter weather cannot directly regulate ungulate populations because it is density independent (Sinclair 1989). In Nova Scotia, and in many other regions in northeastern North America, a series of mild winters during the late 1970s and early 1980s seems to have facilitated a rapid increase in densities of white-tailed deer (Patton 1991, Parker 1995; Fig. 1). Despite an effort to limit this increase via liberal hunting regulations, deer in Nova Scotia had presumably exceeded K carrying capacity (sensu McCullough 1979) and were in poor physical condition by winter of 1987 (Patton 1991, Patterson and Power 2002). The 1987 winter was relatively severe in Nova Scotia, and a substantial decline in deer density began (Patton 1991; Fig. 1). By 1993, deer densities in Nova Scotia were at their lowest level in over 50 years (Nova Scotia Department of Natural Resources [NSDNR], unpublished data). Coyotes were still becoming established throughout Nova Scotia when the decline began during 1987 (Moore and Parker 1992, Patterson 1999). Coyote predation was unable to prevent the peak in deer density, although it probably accelerated the subsequent decline (Patton 1991, Parker 1995, Patterson 1999). During 1993, hunting was restricted to antlered (males >1 yr) deer throughout Nova Scotia. Despite legal protection of antlerless deer and mild winter conditions from 1994 through 1998 (MacDonald 1996, Patterson et al. 1998), deer densities did not increase noticeably until 1997 (Fig. 1). Our study attempted to identify and quantify major factors limiting the population growth of deer in Nova Scotia following a major decline in density. STUDY AREA Our study was conducted in 2 forested areas of Nova Scotia, Canada. The Queens County study area in central southwestern Nova Scotia (44°20′N, 65°15′W) included the eastern half of Kejimkujik National Park (KNP; approx. 200 km 2) and approximately 300 km2 of primarily forested land directly to the east of the park. This region had warm summers averaging 18 °C and cool winters averaging –5 °C during January (Dzikowski et al. 1984). Queens County received little snow during the study with accumulations generally <20 cm; thus, deer did not aggregate in yards (MacDonald 1996, Lock 1997). The second area, on Cape Breton Island (45°45′N, 61°15′W, approx. 300 km2), was centered on the 24-km2 Eden deer wintering area that typically contained approximately 200 deer from January through March (Patterson et al. 1998). Elevation rose from near sea level in the River Denys Basin area to approximately 300 m in the Creignish Mountains. The climate in Cape Breton generally is more moist than KNP, with similar summer temperatures as the Queens County study area (Dzikowski et al. 1984). High elevations in the northern section of the study area typically receive 250–300 cm of snow annually, whereas lowland areas receive 200–250 cm of snow annually (Gates 1975). Median duration of snow cover varies from 140 days on higher elevations to 130 days on lower elevations (Gates 1975). This contrasts with a median duration of snow cover of 59 days in Queens County. Information on the vegetation in each study area was presented by Patterson et al. (1998). Deer were legally protected from harvest in KNP, whereas hunting was restricted to antlered deer in all other areas during autumn 1993–1997. Limited numbers of antlerless deer permits were available in both study areas during autumn 1998. METHODS We captured deer in both study areas using box traps, a ground-based rocket net, chemical immobilization, and a net-gun deployed from a Hughes 500 helicopter (MacDonald 1996). We fit deer with VHF radiocollars containing mortality switches (Lotek Engineering, Newmarket, Ontario, Canada). We located collared deer primarily from the ground with hand-held antennas and portable receivers. We checked for mortality signals approximately twice a week from 1 April to 15 October, and 4 times/week during the rest of the year. We investigated mortality signals and determined cause of death within 24 hr of dis- J. Wildl. Manage. 66(2):2002 covery. We classified deer as a predator kill if there was positive evidence of attack or chase (i.e., blood-soaked fur or snow, bleeding observed around tooth puncture wounds). Predator species involved in mortalities were identified on the basis of tracks, bite marks, or hair or scat left by the predator. We assumed that a deer died of malnutrition when the carcass was lying intact in a natural position with low femur marrow fat content (Cheatum 1949, Verme and Holland 1973). We monitored deer until 22 January 1998 in Queens County, and 14 January 1999 in Cape Breton. We estimated weekly survival (si ) from 1 June (the assumed date of birth of the new fawn crop) through 31 May the following year using the Kaplan-Meier product-limit estimator modified by Pollock et al. (1989) to allow the staggered entry of animals. For each weekly estimate, we included only deer with collars that had been transmitting signals for >4 days. We calculated annual survival estimates (Sann) as the product of the 52 weekly estimates. We estimated the variance of survival estimates at time t as var(st ) = sˆt 2 [1 – (st )]/xt , where xt was the total number of collared deer at risk at time t (Cox and Oakes 1984). Our analyses included the exposure days of censored individuals until the date of disappearance or the end of monitoring because this method results in the least biased survival estimates over a wide range of censoring and survival probabilities (Tsai et al. 1999). We estimated annual survival for adult (>1 yr) females, adult males, and fawns. Sufficient telemetry data to estimate survival of fawns was available only from December to May. We estimated survival of fawns during June–November by comparing natality rates of road-killed does examined in each study area from 1 February through 15 May (n = 151 for Queens County, 131 for Cape Breton) with the fawn:doe ratios observed in the road-killed sample from each study area during the following December and January (n = 126 for Queens County, 67 for Cape Breton). Early winter fawn:doe ratios were adjusted to account for mortality of adult females occurring during the summer and autumn intervals (Nelson and Mech 1986a). We tested for interannual differences in survival for each age and sex class using generalized chisquare tests calculated by program CONTRAST (Hines and Sauer 1989, Sauer and Williams 1989). We pooled data from all years of study when the statistical hypothesis of homogeneity among annual survival rates was not rejected. For paired comparisons between study areas, sexes, and age LIMITING FACTORS FOR DEER • Patterson et al. 513 (fawns vs. adults only) we used log-rank tests (Cox and Oakes 1984, Pollock et al. 1989) to test the statistical hypothesis that both survival curves came from the same true underlying survival function. We used CONTRAST for all comparisons involving >2 rates to reduce the probability of a Type I statistical error. Based on a visual examination of survival curves, we divided the biological year into 3 intervals with relatively constant survival for adult deer: 1 June–12 October, 13 October–7 December (included the bow and firearm hunting seasons), 8 December–31 May. In the Queens County study area, we estimated separate survival rates for deer living within KNP because hunting was prohibited in the park. Cause-specific mortality was defined as the probability of a deer dying during a given interval from a given mortality factor. We estimated mortality rates resulting from (1) predation; (2) legal harvest (registered by licensed hunters or reported by aboriginal people); (3) unregistered harvest (including deer that were known or suspected to have been illegally harvested, abandoned, or lost [wounding loss] or legally harvested but not reported by aboriginal people); (4) malnutrition; and (5) other natural causes (old age, natural accidents, or injuries). The probability of a deer dying from any given mortality source during each week, interval, and annually was calculated following Heisey and Fuller (1985). Mortality rates estimated in this manner can be seriously biased if the assumption of a constant probability of succumbing to each mortality factor for the duration of each interval is violated (Pollock et al. 1989, Tsai et al. 1999). We minimized this bias by delineating seasonal intervals only after examining the annual survival curves and the seasonal contributions of each mortality factor. Variance of mortality estimates was calculated using the Taylor series approximation method ( Johnson 1979, Heisey 1985). We determined the relative abundance of whitetailed deer within each study area using pelletgroup counts conducted along 30 1,000 × 2-m systematic line transects during April and May 1994–1997 (Neff 1968, Patterson et al. 1998). During 1998 and 1999, we estimated deer densities in each study area based on the relative change between 1997–1998 and 1998–1999 indicated by the pellet-group inventories conducted in the counties containing each study area by the NSDNR (n = 78, 95, and 55 plots surveyed each year in southwestern Nova Scotia; and 64, 64, and 65 plots in Cape Breton). We converted the numbers 514 J. Wildl. Manage. 66(2):2002 LIMITING FACTORS FOR DEER • Patterson et al. We estimated (λ 2 ) by linear regression of loge(N ) estimated from the annual pellet group surveys following Eberhardt and Simmons (1992). We recalculated λ 2 1 0 0 0 t i m e s w i t h a r a n d o m value being drawn from a normal distribution generated using N–± SE for each area and year specific density estimate. For each method in each study area, we then compared the central 95% of the λ estimates to assess the accuracy and p r e cision of these measures of population change. of pellet groups counted/km2 in each study area each year into estimates of deer/km2 assuming an average date of leaf fall of 1 November and a daily defecation rate of 16 pellet groups/day/deer. Our assumed defecation rate was calculated based on an iterative process whereby we incorporated the size of harvest during each year (from mandatory registration stations) and the estimated rates of all significant mortality factors (this study) into a model to determine the annual population size necessary to result in observed rates of interannual population change. We estimated standard errors for deer density estimates based on the variance in the number of pellet groups counted among plots in each territory. Using 2 methods, we estimated the finite rate of increase (λ) for deer in each study area. We estimated λ1 as λ = (1 – Ma)/(1 – R ) where Ma was the finite annual rate of adult mortality (calculated as 1 – Sann, weighted to account for the ratio of adult females:males in each study area, and the respective survival estimates for each sex) and R was the number of fawns/100 does observed during late winter in each study area (Hatter and Bergerud 1991). We incorporated the standard errors of Ma (calculated from the coefficient of variation [CV] for Sann) and R (calculated from the interannual variation among sex and age ratios from 1994 to 1998) into a Monte Carlo simulation to assess the variance associated with λ. For each study area, λ was recalculated 1,000 times with a random value being drawn from a normal distribution calculated using x– ± SE for each parameter during each subsequent calculation. RESULTS We captured and radiocollared 71 deer (>6 month) in Cape Breton and 50 in the Queens County study area from February 1994 through April 1997. Most deer were captured between December and March each year. Four deer (3 in Cape Breton, 1 in Queens County) shed their collars within 24 hours of capture. Five deer died within 1 week of marking and were eliminated from the analysis. No significant differences occurred in annual survival rates among years for adult males (χ23 < 2.4, P > 0.49), or adult females (χ23 < 1.9, P > 0.38) in any study area. Thus, for subsequent analyses, we pooled data among years for adults of each sex. Survival may not have been uniform among years for fawns in Cape Breton (χ23 = 6.0, P = 0.054), but we also pooled these across all years because sample sizes were small (n = 6, 7, 4, 4 fawns collared during winters 1994–1997, respectively). Annual survival rates ranged from 46.1 ± 9.8% (SE) for adult males in CB, to 93.9 ± 4.3% for adult females in KNP (Table 1; Figs. 1, 2). Annual survival curves Table 1. Seasonal survival rates of radiocollared white-tailed deer monitored in the Queens County and Cape Breton study areas, Canada, Feb 1994–Jan 1999. Cohort Rate Summer Autumn Winter–Spring (1 Jun–12 Oct) (13 Oct–7 Dec) (8 Dec–31 May) SE n Annual No. No. No. No. radio radio radio radio weeks Rate SE Non-park adult females 0.985 0.011 58 2,506 0.884 0.027 Non-park adult males 0.923 0.059 21 523 0.662 0.074 Kejimkujik Natl Park adult females 0.970 0.029 12 610 1.0 Kejimkujik Natl Park adult males 1.0 2 95 1.0 n weeks Rate SE n weeks Rate SE n weeks 56 994 0.924 0.023 59 2,875 0.804 0.031 18 161 0.830 0.047 17 447 0.507 0.076 25 1,131 10 256 0.969 0.032 12 720 0.939 0.043 2 34 2 98 1.0 1.0 62 6,375 12 1,586 2 227 J. Wildl. Manage. 66(2):2002 LIMITING FACTORS FOR DEER • Patterson et al. 515 for adults of both sexes were not different between deer living outside KNP in Queens County and those in the Cape Breton study area (females: logrank test, χ2 = 0.13, P = 0.72; males: χ2 = 0.42, P = 0.84; Fig. 2). Thus, for each sex, we pooled data from these areas for further analyses. Survival of adult females living in KNP was higher than for adult females outside of the park (0.939 ± 0.043 vs. 0.804 ± 0.031; log-rank test, χ2 = 6.49, P = 0.011). No adult males died in KNP (Table 1). Outside of the park, annual survival rates were significantly lower for adult males than females (0.507 ± 0.076 vs. 0.804 ± 0.031; χ23 = 6.75, P = 0.009; Table 1). Seasonal Survival Rates No seasonal differences occurred in weekly survival rates for adult female deer living within KNP (χ22 = 0.36, P = 0.836; Table 1). For deer living outside the park, no difference occurred in weekly survival rates between winter and autumn for adult females (χ21 = 1.23, P = 0.268) or between summer and winter rates for adult males (χ21 = 0.944, P = 0.331). Weekly survival rates were highest during summer for both adult female (χ2 14.6, P = 0.0001) and male (χ22 = 6.89, P = 0.032) deer residing outside KNP (Table 1). Fawn survival through early winter was approximately 0.63 in Cape Breton and 0.70 in Queens County (Table 2). Fawn survival from December through May was 0.505 ± 0.082 and 0.60 ± 0.105 in the Cape Breton and Queens County study areas, respectively. Estimated annual survival rates of fawns were 0.319 and 0.422, respectively (Table 2). Survival among age and sex classes generally was most uniform during summer (Table 1). During autumn, adult females living in KNP had higher survival than females outside the park (χ21 = 9.23, P = 0.002), which in turn had higher survival than adult males also living outside of the park (χ21 = 457, P < 0.0001). Winter survival rates were not uniform among sex and age classes (χ22 = 4.80, P = 0.090; Table 1). Fawns had significantly lower winter survival than adult females (0.584 ± 0.066 vs. 0.924 ± 0.024; χ21 = 4.80, P = 0.029), but not males (0.830 ± 0.081; χ21 = 1.50, P = 0.219). The difference between winter survival of adult males and females (0.094) was not significant (χ21 = 1.24, P = 0.26; Table 1). Fig. 1. The relative abundance of white-tailed deer in the Queens County (QC) and Cape Breton (CB) study areas, 1994–1999, as estimated by pellet-group counts. The inset depicts provincial deer population trends, 1983–1999, as estimated by pellet-group counts conducted along 440 ± 3 (SE) transect lines distributed randomly throughout the province (Patton 1991). Error bars represent the standard error associated with each deer density estimate. Cause-specific Mortality We documented 46 mortalities during 112,302 animal-days of telemetry contact with deer: (1) 17 due to predation (12 coyote, 1 suspected coyote, 3 bobcat [Lynx rufus], and 1 lynx [Lynx lynx]); (2) Fig. 2. Kaplan-Meier survival curves for (a) adult female and (b) adult male deer in Queens County and Cape Breton, Nova Scotia, Canada, 1994–1998. 516 J. Wildl. Manage. 66(2):2002 LIMITING FACTORS FOR DEER • Patterson et al. Table 2. Seasonal survival of fawn white-tailed deer in the Cape Breton and Queens County study areas, Nova Scotia, Canada, 1994–1997. Cape Breton Parameter Jun Jan Fawns:doe 1.23 0.82 Number doesa 100 87.4 Number fawnsb 123 71.8 Fawn survival 0.631 c Dec–May 0.505d SE Queens County Annual 0.082 0.319 SE Jun Jan 1.21 0.93 100 86.1 121 79.8 0.054 e 0.704 c Dec–May SE Annual SE 0.60d 0.105 0.422 0.074e a Hypothetical number surviving based on a starting population of 100 and seasonal survival rates calculated from telemetry data. b Estimated from the natality of female deer struck by automobiles Feb–May, and the fawn:doe ratios observed Dec–Jan. Fawn:doe ratios in winter were adjusted to account for the number of does surviving Jun–Dec. c Estimated as the number of fawns still living in Dec divided by the number born in Jun. Our initial calculation was based on survival through the end of Dec (214 days). The rate presented here was obtained by expanding the mean daily rate during this interval (0.9975) through the end of Nov only (183 days; i.e., 0.631 = 0.9975183). d Based on telemetry data. e Based on the coefficient of variation associated with the winter survival rates only; thus, this standard error estimate is minimal. 10 due to legal (registered) harvest (including 2 females harvested by native hunters and 1 female taken with an antlerless deer permit during 1998); (3) 10 due to unregistered harvest (2 males, including 1 wounding loss); (4) 2 died of malnutrition (1 fawn, 1 adult female); and (5) 5 died of other natural causes (1 female fawn fell through ice, 1 adult male became mired in mud, 1 adult male died of a chest infection, and 2 females died of apparent old age or other natural causes). We could not determine the probable cause of death for 2 adult females during early December 1996 and 1997, respectively. Harvest-related mortality was restricted to autumn (Fig. 3) but represented the greatest source of mortality for adult male deer living outside of KNP (χ2 = 9.9, P = 0.0016; Table 3). Hunting (8.2 ± 2.4%) and predation (7.6 ± 2.3) were the most significant mortality sources for adult female deer outside the park, and the 2 rates were not different (χ2 = 0.032, P = 0.86; Table 3). Predation rates on adult females were similar within and outside of the park (6.1 ± 4.2% vs. 7.6 ± 2.3%, respectively). Coyote predation removed a proportion of adult male deer similar to the number removed by malnutrition and other natural mortality factors (0.072 ± 0.053 vs 0.066 ± 0.026; Table 3). Predation was uncommon among adult deer during summer and occurred at a relatively constant rate from autumn to spring for adult females (Fig. 3A). The 2 cases of predation on adult male deer occurred during early autumn and midwinter, respectively (Fig. 3B). Although the number of fawns monitored each winter increased as winter progressed, the 3 cases of coyote predation on collared fawns occurred during January and early February. In KNP, a bobcat killed 1 adult female during November 1996 and an unknown predator(s) killed 1 adult female during July 1996. Population Trends Fig. 3. Cumulative number of (a) adult female and (b) adult male deer dying of various causes in Nova Scotia, Canada, 1994–1998. For our models of population change based on adult survival and recruitment, late winter sex J. Wildl. Manage. 66(2):2002 517 LIMITING FACTORS FOR DEER • Patterson et al. Table 3. Cause-specific mortality rates of radiocollared white-tailed deer monitored in Kejimkujik National Park (KNP), Feb 1994–Jan 1998, and in the Cape Breton and Queens County study areas combined (CB/QCNP; excluding Kejimkujik National Park), Feb 1994–Jan 1999. Area CB/QCNP CB/QCNP CB/QCNP KNPe Cohort Adult females Predationa Registered harvestb Unregistered harvestc Malnutrition Other natural mortalityd Unknown Adult males Predation Registered harvest Unregistered harvest Other natural mortality Fawns Predation Malnutrition Natural mortality Adult females Predation Summer Autumn Winter–Spring (1 Jun–12 Oct) (13 Oct–7 Dec) (8 Dec–31 May) Rate SE Rate SE Rate SE Rate SE 0.008 0.007 0.016 0.023 0.061 0.011 0.013 0.021 0.061 0.022 0.075 0.022 0.060 0.006 0.014 0.016 0.023 0.013 0.021 0.007 0.010 0.011 0.072 0.266 0.076 0.066 0.053 0.081 0.054 0.051 0.061 0.042 0.037 0.039 0.030 0.016 0.011 0.288 0.040 0.085 0.043 0.035 0.009 0.029 0.007 0.008 0.002 0.002 0.0572 0.0532 0.108 0.016 0.275 0.022 0.098 0.087 0.008 0.016 0.031 0.032 Annual a Of the 17 instances of predation observed during this study, 12 were attributed to coyotes, 1 to a suspected coyote, 3 to bobcats, and 1 to a lynx. b Includes 2 deer taken by aboriginal hunters and 1 taken in Cape Breton in 1998 under a limited antlerless deer permit system. c Unregistered harvest indicates deer which were known or suspected to have been illegally harvested, abandoned or lost (wounding loss), or harvested and not reported by aboriginal people. d Natural mortality includes deaths from malnutrition, old age, accidents, or other naturally occurring sources (excluding predation). e We recorded no mortalities of adult males during 4.4 animal-years of monitoring in Kejimkujik National Park. and age ratios were estimated from 105 roadkilled deer examined on Cape Breton Island and 115 in southwestern Nova Scotia, during April and May, 1994–1998. The distributions of rates of increase estimated by our models were similar to those estimated by the pellet-group inventories (Queens County, λ(pellet groups) = 1.064, range = 1.02 to 1.11, vs. λ(model) = 1.048, range = 0.82 to 1.26; Cape Breton, λ(pellet groups) = 1.023, range = 0.89 to 1.16 vs. λ(model) = 1.110, range = 0.94 to 1.29; Fig. 4). In Cape Breton, the pellet-group inventories indicated a higher mean rate of increase than our model (Fig. 4B). Although the overall rate of population growth was positive in both areas during this study, pellet-group inventories suggested a decline in deer densities between 1994 and 1995 in Cape Breton but not in Queens County (Fig. 1). Including only pelletcount data from 1995 to 1999 in the simulation for Cape Breton resulted in a mean rate of increase of 1.055 with a range of 0.94 to 1.19. DISCUSSION Despite legal protection, harvest represented a significant mortality factor for adult female deer during this study. In Washington, several illegally harvested female black-tailed deer (Odocoileus hemionus columbianus) were neither recovered nor field-dressed, suggesting that misidentification of a legal target may have been involved (McCorquodale 1999). In all but 1 of the cases where we suspected an antlerless deer was illegally harvested, we recovered only the severed collar, generally from a ditch, culvert, or roadside pond. Regardless of the ultimate cause of these losses, a significant proportion of antlerless deer were harvested but not reported to the NSDNR each year. Whitlaw et al. (1998) reported that the illegal harvest of deer in New Brunswick was rare. Rates of illegal harvest of deer in other areas ranged from 9 to 16% of the legal harvest for white-tailed deer in Minnesota (Nelson and Mech 1986a, Fuller 1990) to 61% of the legal harvest for deer in the 518 LIMITING FACTORS FOR DEER • Patterson et al. Fig. 4. Distribution of rates of increase for deer estimated by 1000 replicates of a Monte Carlo simulation based on deer pellet group inventories (PGI), and a simple population model using estimates of adult mortality and fawn recruitment (Hatter and Bergerud 1991), in (a) Queens County and (b) Cape Breton, Nova Scotia, Canada, 1995–1998. Klickitat Reserve Basin of Washington (McCorquodale 1999). The illegal harvest of antlerless deer was considerable when no antlerless deer permits were available. The incentive to illegally harvest antlerless deer may decrease as provincial deer numbers and the number of antlerless deer permits available to hunters continue to increase. The annual survival rate for adult females exposed to hunting (80.4 ± 3.1% ) was similar to those of adult females from populations with light hunting pressure in northern Minnesota, Michigan, and northern New Brunswick (Nelson and Mech 1986a, Van Deelen et al. 1997, Whitlaw et al. 1998), but higher than for populations subjected to heavy hunting pressure in Montana, northcentral Minnesota, and southern New Brunswick (Fuller 1990, Dusek et al. 1992, Whitlaw et al. 1998). Given the low deer densities we observed relative to many areas in northeastern North America (Fig. 1; e.g., deCalesta and Stout 1997), mild winters (MacDonald 1996, Patterson et al. 1998), and legal protection of antlerless deer during this study, we suggest that the survival rates we documented for adult females are near maximal for deer in Nova Scotia. J. Wildl. Manage. 66(2):2002 Survival rates of adult males in this study compared to those documented in Minnesota and New Brunswick (Nelson and Mech 1986a, Fuller 1990, Whitlaw et al. 1998), but were considerably higher than for an intensely harvested population in northern Michigan where the annual survival was 0.22 (Van Deelen et al. 1997). During this study, age-specific natality rates remained high despite some skewing of the adult sex ratio during 5 years of male-only hunting (NSDNR, unpublished data). Thus, it appears that increased harvesting pressure on adult males in Nova Scotia had little effect on herd productivity. The fact that some adult females were harvested each year despite legal protection (Table 3) probably also helped prevent excessive skewing of the adult sex ratio. During winter, rates of predation and natural mortality for fawns were considerably higher than observed for older deer (Table 3). A concurrent study of coyote predation on deer suggested a more subtle difference in predation rates of fawn and adult deer during winter (Patterson 1999, Patterson and Messier 2000). Patterson and Messier (2000) relied on tracking radiocollared coyotes to kill sites during intensive winter snow tracking. These authors noted that fawns were completely consumed more rapidly than adults and acknowledged that they may have underestimated the number of fawns killed per coyote family group during winter. Although the variance of our estimates of coyote predation rates during winter was large (CV = 0.32), our study supports this contention. Our estimates of annual survival rates for fawns (approx. 37%; Table 2) are similar to those for fawns in northcentral New Brunswick (0.21–0.30; Ballard et al. 1999) and northcentral Minnesota (0.35; Fuller 1990). Poor fawn survival (due largely to predation in early summer) was believed to be a major factor contributing to declines in deer populations in both of these studies. We could not determine the causes of summer fawn mortality, but summer fawn survival in Nova Scotia was positively correlated with snowshoe hare (Lepus americanus) densities (the major alternative food of coyotes during this study; Patterson et al. 1998), suggesting that predation was a major source of mortality for fawns (Patterson and Power 2002). Black bears (Ursus americanus) and bobcats also appeared to be common in both study areas and are recognized as important limiting factors for northern deer populations (Mathews and Porter 1988, Kunkel and Mech 1994, Ballard et al. 1999). J. Wildl. Manage. 66(2):2002 Although legal hunting was restricted to antlered male deer only throughout Nova Scotia during 1993 and winters were mild during this study, there was little increase in deer numbers until winter 1997 (Fig. 1). Our estimates of positive but relatively low rates of increase for deer in both study areas, derived from our model of adult survival and fawn recruitment, were corroborated by annual deer pellet-group inventories (Fig. 4). However, despite considerable effort by the staff of the NSDNR in conducting the annual pellet group surveys, and a relatively large sample of both radiocollared and vehicle-struck deer for the demographic analyses, substantial variation occurred in both estimates of deer population change (Fig. 4). Further, the variance associated with the pellet-group counts represented a minimum estimate because we only considered variation in the number of pellets among plots and not potential differences in pellet-group detectability and deposition rates. The use of pellet group counts as an index of deer numbers has been criticized (Fuller 1991, 1992). However, pellet group counts in Nova Scotia were closely related to autumn harvest during 10 years of any age–any sex hunting from 1983 to 1992 (r 2 = 0.87, P = 0.001; Patterson et al. 1998). The annual pellet-group counts in Nova Scotia are useful for determining regional trends in deer population change. There was no indication that any harvesting of deer occurred within KNP during this study, yet deer densities in KNP were no higher than those in surrounding areas (Table 1; Patterson and Messier 2000). Similarly, it may seem curious that deer densities in Queens County were lower than those in Cape Breton (Fig. 1; Patterson et al. 1998) despite the observation that Cape Breton deer experienced more severe winter weather. Compared with Cape Breton, the forests in the Queens County study area consisted of relatively large, mature stands of mature softwood or hardwood-dominated mixed-wood (Lock 1997). This was particularly evident in KNP, where forest harvesting has not occurred since the early 1960s. Deer thrive in young, heterogeneous forests (Halls 1984, Mooty et al. 1987), and we suggest that higher densities of deer in Cape Breton resulted from generally more favorable habitat conditions. The mature, relatively unbroken forests in KNP provide relatively little browse for deer (Drysdale 1986, Lock 1997). Following the province-wide population boom in deer densities during the early 1980s, deer in KNP experienced an equal if not greater decline in densities (Drys- LIMITING FACTORS FOR DEER • Patterson et al. 519 dale 1986, Patterson 1995). Prior to the decline, this population was at high densities and in poor physical condition (Drysdale 1986), suggesting that the lack of hunting was largely compensated for by high density-dependent winter mortality or reductions in recruitment. Although we did not detect significant differences in recruitment or adult survival between Cape Breton and Queens County, such differences may become more pronounced as each area approaches its respective K or when winter severity is high. MANAGEMENT IMPLICATIONS Wolves are not present east of the Saint Lawrence River. Some researchers have suggested that coyotes have replaced wolves as a significant predator of white-tailed deer in northeastern North America (Brundige 1993, Ballard et al. 1999). Although coyote predation was secondary in importance to harvest as a limiting factor for deer during this study, Patterson (1999) concluded that coyote predation could severely limit deer population growth when deer densities were very low (<0.5/km2) or when winter severity was high. Similarly, Messier et al. (1986) concluded that coyote predation could limit deer densities in southern Québec. During our study, deer densities appeared to be well below K (Lock 1997) and winters were quite mild. This, coupled with the apparently good condition of most deer killed by predators, suggests that predation losses were largely additive. Similarly, there was no increase in any other mortality factor when deer were protected from harvest within KNP, suggesting that harvest mortality was also largely additive. Fuller ( 1990 ) demonstrated that excessive hunting is more likely to initiate or contribute to population declines when winter severity is high. Similarly, killing rates of deer by predators are positively associated with winter severity (Messier and Barrette 1985, Nelson and Mech 1986b, Patterson and Messier 2000). Following the peak in deer densities in Nova Scotia during 1986, deer densities began to decline sharply and winter severity was above average during 1986–1990 and 1992–1993 (NSDNR, unpublished data). However, for political or bureaucratic reasons, the antlered-male-only hunting regulations were not established until autumn 1993. Continued high harvests following the peak in deer densities during 1986 accelerated and prolonged the subsequent decline in deer densities (Fig. 1). The establishment of a zone-based antlerless harvest quota system during 1998 should allow managers 520 LIMITING FACTORS FOR DEER • Patterson et al. to regulate deer numbers by annually adjusting the number of antlerless permits in response to estimates of hunting and nonhunting losses. Harvest and predation were the largest proximate limiting factors for deer during this study, during a period of low deer densities and relatively mild winters. Over a longer time frame, density-dependent forage competition is the ultimate regulatory factor for deer in the Northeast (Messier 1991, Post and Stenseth 1998, Dumont et al. 2000, Patterson and Power 2002), with other limiting factors such as harvest and predation temporarily altering the equilibrium around which deer densities fluctuate. In addition to understanding the effects of major proximate mortality sources on deer population growth, managers must be aware of the status of the deer populations they manage in relation to carrying capacity. Our study demonstrates that recovery of deer numbers after a population crash can be prolonged despite apparently favorable conditions for rapid growth. ACKNOWLEDGMENTS We thank D. Banks, S. Bondrup-Nielsen, G. Boros, M. J. Boudreau, H. J. Broders, R. Charlton, C. L. Cushing, C. Doliver, A. P. Duke, T. M. Fitzgerald, C. Frail, K. Huskins, A. Kennedy, K. G. Lock, C. MacDonald, S. F. Morrison, E. M. Muntz, M. O’Brien, D. Richards, M. Robinson, D. Shaw, and G. Tatlock for logistical support and V. A. Power and A. L. Nette of the NSDNR, Wildlife Division, for providing data on vehicle-struck deer and the provincial deer pellet-group inventories. We thank the staff of the Air Services division of NSDNR for their expert assistance with the aerial tracking flights, and D. O. Joly and M. Crête for comments on an earlier draft of the manuscript. Financial and logistical support for this study was provided by the NSDNR, Parks Canada, and Forestry Canada, under the Cooperation Agreement for Forestry Development. LITERATURE CITED BALLARD, W. B., H. A. WHITLAW, S. J. YOUNG, R. A. JENKINS, AND G. J. FORBES. 1999. Predation and survival of white-tailed deer fawns in northcentral New Brunswick. Journal of Wildlife Management 63:574–579. BRUNDIGE, G. C. 1993. Predation ecology of the eastern coyote (Canis latrans) in the Adirondacks, New York. Dissertation, State University of New York, Syracuse, USA. CASE, D. J., AND D. R. MCCULLOUGH. 1987. The whitetailed deer of North Manitou Island. Hilgardia 55:1–57. CHEATUM, E. L. 1949. Bone marrow as an index of malnutrition in deer. New York State Conservationist 3:19–22. J. Wildl. Manage. 66(2):2002 COX, D. R., AND D. OAKES. 1984. Analysis of survival data. Chapman & Hall, New York, USA. CRÊTE, M. 1999. The distribution of deer biomass in North America supports the hypothesis of exploitation ecosystems. Ecology Letters 2:223–227. DECALESTA, D. S., AND S. L. STOUT. 1997. Relative deer density and sustainability: a conceptual framework for integrating deer management with ecosystem management. Wildlife Society Bulletin 25:252–258. DRYSDALE, C. D. 1986. The roadside deer of Kejimkujik National Park. Thesis, Acadia University, Wolfville, Nova Scotia, Canada. DUMONT, A., M. CRÊTE, J.-P. OUELLET, J. HUOT, AND J. LAMOUREUX. 2000. Population dynamics of northern white-tailed deer during mild winters: evidence of regulation by food competition. Canadian Journal of Zoology 78:764–776. DUSEK, G. L., A. K. WOOD, AND S. T. STEWART. 1992. Spatial and temporal patterns of mortality among female white-tailed deer. Journal of Wildlife Management 56:645–650. DZIKOWSKI, P. A., G. KIRBY, G. READ, AND W. G. RICHARDS. 1984. The climate for agriculture in Atlantic Canada. Atlantic Advisory Committee on Agrometeorology Publication ACA 84-2-500. Atmospheric and Environmental Services, Halifax, Nova Scotia, Canada. EBERHARDT, L. L., AND M. A. SIMMONS. 1992. Assessing rates of increase from trend data. Journal of Wildlife Management 56:603–610. FRYXELL, J. M., D. J. T. HUSSELL, A. B. LAMBERT, AND P. C. SMITH. 1991. Time lags and population fluctuations in white-tailed deer. Journal of Wildlife Management 55:377–385. FULLER, T. K. 1990. Dynamics of a declining white-tailed deer population in north-central Minnesota. Wildlife Monographs 110. ———. 1991. Do pellet counts index white-tailed deer numbers and population change? Journal of Wildlife Management 55:393–396. ———. 1992. Do pellet counts index white-tailed deer numbers and population change?: a reply. Journal of Wildlife Management 56:613. GATES, A. D. 1975. The tourism and outdoor recreation climate of the Maritime Provinces. Publications in Applied Meteorology REC-3-73. Atmospheric Environment Service, Environment Canada, Toronto. HALLS, L. K. 1984. White-tailed deer: ecology and management. Stackpole Books, Harrisburg, Pennsylvania, USA. HATTER, I. W., AND W. A. BERGERUD. 1991. Moose recruitment, adult mortality and rate of change. Alces 27:65–73. HEISEY, D. M. 1985. Micromort user’s guide. NH Analytical Software, St. Paul, Minnesota, USA. ———, AND T. K. FULLER. 1985. Evaluation of survival and cause-specific mortality rates using telemetry data. Journal of Wildlife Management 49:668–674. HINES, J. E., AND J. R. SAUER. 1989. Program CONTRAST: a general program for the analysis of several survival or recovery rate estimates. Technical Report 24, U.S. Fish and Wildlife Service, Washington, D.C. JOHNSON, D. H. 1979. Estimating nest success: the Mayfield method and an alternative. Auk 96:651–661. KUNKEL, K. E., AND L. D. MECH. 1994. Wolf and bear predation on white-tailed deer fawns in northeastern Minnesota. Canadian Journal of Zoology 72:1557–1565. J. Wildl. Manage. 66(2):2002 LOCK, B. A. 1997. Deer wintering habitat models for two regions of Nova Scotia. Thesis, Acadia University, Wolfville, Nova Scotia, Canada. MACDONALD, B. A. 1996. Spatial distribution, survivorship and effects of harvesting on over-wintering whitetailed deer in Nova Scotia. Thesis, Acadia University, Wolfville, Nova Scotia, Canada. MATHEWS, N. E., AND W. F. PORTER. 1988. Black bear predation of white-tailed deer neonates in the central Adirondacks. Canadian Journal of Zoology 66:1241–1242. MCCORQUODALE, S. M. 1999. Movements, survival, and mortality of black-tailed deer in the Klickitat Basin of Washington. Journal of Wildlife Management 63:861–871. MCCULLOUGH, D. R. 1979. The George Reserve deer herd: population ecology of a K-selected species. University of Michigan Press, Ann Arbor, USA. MESSIER, F. 1991. The significance of limiting and regulating factors on the demography of moose and whitetailed deer. Journal of Animal Ecology 60:377–393. ———, AND C. BARRETTE. 1985. The efficiency of yarding behavior by white-tailed deer as an antipredator strategy. Canadian Journal of Zoology 63:785–789. ———, ———, AND J. HUOT. 1986. Coyote predation on a white-tailed deer population in southern Québec. Canadian Journal of Zoology 64:1134–1136. MOORE, G. C., AND G. R. PARKER. 1992. Colonization by the eastern coyote (Canis latrans). Pages 23–37 in A. H. Boer, editor. Ecology and management of the eastern coyote. Wildlife Research Unit, University of New Brunswick, Fredericton, Canada. MOOTY, J. J., P. D. KARNS, AND T. K. FULLER. 1987. Habitat use and seasonal range size of white-tailed deer in northcentral Minnesota. Journal of Wildlife Management 51:644–648. NEFF, D. J. 1968. The pellet group count technique for big game trend, census and distribution: a review. Journal of Wildlife Management 32:597–614. NELSON, M. E., AND L. D. MECH. 1986a. Mortality of white-tailed deer in northeastern Minnesota. Journal of Wildlife Management 50:691–698. ———, AND ———. 1986b. Relationship between snow depth and gray wolf predation on white-tailed deer. Journal of Wildlife Management 50:471–474. PARKER, G. R. 1995. Eastern coyote: the story of its success. Nimbus Publishing, Halifax, Nova Scotia, Canada. PATTERSON, B. R. 1995. The ecology of the eastern coyote in Kejimkujik National Park. Thesis, Acadia University, Wolfville, Nova Scotia, Canada. ———. 1999. The effects of prey distribution and abundance on eastern coyote life history and predation on white-tailed deer. Dissertation, University of Saskatchewan, Saskatoon, Canada. LIMITING FACTORS FOR DEER • Patterson et al. 521 ———, L. K. BENJAMIN, AND F. MESSIER. 1998. Prey switching and the feeding habits of eastern coyotes in relation to white-tailed deer and snowshoe hare densities. Canadian Journal of Zoology 76:1885–1887. ———, AND F. MESSIER. 2000. Factors influencing killing rates of white-tailed deer by coyotes in eastern Canada. Journal of Wildlife Management 64:721–732. ———, AND V. A. POWER. 2002. Contributions of forage competition, harvest, and climate fluctuation to changes in population growth of northern whitetailed deer. Oecologia 130:62–71. PATTON A. 1991. Deer history. Nova Scotia Conservation 15(1):3–6. POLLOCK, K. H., S. R. WINTERSTEIN, AND C. M. BUNCK. 1989. Survival analysis in telemetry studies: the staggered entry design. Journal of Wildlife Management 53:7–15. POST E., AND N. C. STENSETH. 1998. Large-scale climatic fluctuation and population dynamics of moose and white-tailed deer. Journal of Animal Ecology 67:537–543. POTVIN, F., J. HUOT, AND F. DUCHESNEAU. 1981. Deer mortality in the Pohénégamook wintering area, Québec. Canadian Field-Naturalist 95:80–84. POULLE, M. L., M. CRÊTE, J. HUOT, AND R. LEMIEUX. 1993. Prédation excercée par le coyote, Canis latrans, sur le cerf de Virginie, Odocoileus virginianus, dans un ravage en déclin de l’Est de Québec. Canadian FieldNaturalist 107:177–185. SAUER, J. R., AND B. K. WILLIAMS. 1989. Generalized procedures for testing hypotheses about survival or recovery rates. Journal of Wildlife Management 53:137–142. SINCLAIR, A. R. E. 1989. Population regulation of animals. Pages 197–241 in J. M. Cherrett, editor. Ecological concepts. Blackwell Scientific, Oxford, United Kingdom. TSAI, K., K. H. POLLOCK, AND C. BROWNIE. 1999. Effects of violation of assumptions for survival analysis methods in radiotelemetry studies. Journal of Wildlife Management 63:1369–1375. VAN DEELEN, T. R., H. CAMPA, III, J. B. HAUFLER, AND P. D. THOMPSON. 1997. Mortality patterns of whitetailed deer in Michigan’s Upper Peninsula. Journal of Wildlife Management 61:903–910. VERME, L. J., AND J. C. HOLLAND. 1973. Reagent-dry assay of marrow fat in white-tailed deer. Journal of Wildlife Management 37:103–105. WHITLAW, H. A., W. B. BALLARD, D. L. SABINE, S. J. YOUNG, R. A. JENKINS, AND G. J. Forbes. 1998. Survival and cause-specific mortality rates of adult white-tailed deer in New Brunswick. Journal of Wildlife Management 62:1335–1341. Received: 15 September 2000. Accepted: 15 November 2001. Associate Editor: Krausman.
© Copyright 2026 Paperzz