Movements and habitat interactions of white-tailed deer: implications for chronic wasting disease management.
Chronic wasting disease (CWD) is a fatal neurodegenerative disease that significantly affects the survival and ecology of white-tailed deer (Odocoileus virginianus, Williams et al., 2002; Joly et al., 2003). CWD has the potential to diminish the overall deer population (Gross and Miller, 2001; Joly et al., 2003; Dulberger et al., 2010) and has had significant economic impact by reducing hunter participation (Bishop, 2004; Needham et al., 2004; Vaske et al., 2004). The disease was first detected in Wisconsin's white-tailed deer herd in 2002, and the Wisconsin Department of Natural Resources (WDNR) responded to the discovery of CWD by attempting to reduce herd size in the area where disease prevalence was highest, defined as the Disease Eradication Zone (DEZ). This strategy was based on the expectation that increased harvest would reduce deer density (Blanchong et al., 2006) and reduced density would lead to less transmission of the disease (Habib et al., 2011). We also expect reduced density and increased harvest will not influence home range size in a manner that causes disease prevalence or transmission to increase, but this assumption has not yet been tested empirically.
It is essential to understand deer home range dynamics if we are to understand disease transmission and the success of its management. Females are typically highly philopatric; they are characterized by stable home ranges and show substantial individual overlap within the same social group (Mathews and Porter, 1993; Aycrigg and Porter, 1997; Schauber et al., 2007). However, deer exhibit a great deal of behavioral plasticity under different habitats and conditions (Miller, 1997; Laseter, 2004; Comer, 2005), and their site fidelity may change or break down under conditions of high habitat fragmentation, altered herd density, and elevated hunting pressure (Nixon et al., 1991; Miller, 1997; Habib et al., 2011). Because CWD persists in environmental reservoirs (Johnson et al., 2006, 2007), it is important to determine the area within which an individual may shed prions, which is likely concentrated within the home range, where species spend the majority of their time (Clements et al., 2011; Walter et al., 2011). Local behavioral studies are critically needed in areas where disease prevalence is highest to evaluate how social cohesion and landscape features affect movement.
Ungulates in fragmented habitat with various configurations of agricultural and forest land often exhibit altered home range size and movement behavior (Oyer et al., 2007; Storm et al., 2007; Skuldt et al., 2008; Walter et al., 2009; Felix et al. 2010). The amount of forest cover (Nixon et al., 1991; McKeefry, 2003; Long et al., 2005; Kjaer et al., 2008), forest edge (Kie et al., 2002; Skuldt, 2005), and habitat diversity (Loft et al., 1984; Beier and McCullough, 1990) all play a role in determining habitat use. Studies that relate deer density to home range size have been equivocal and reveal both positive (Kilpatrick et al., 2001) and negative (Bertrand et al., 1996) influences. Although CWD prevalence occurs at elevated rates when densities are high (Joly et al., 2003), the behavioral responses that occur as densities change, and their implications for disease management, are poorly understood (Schauber and Woolf, 2003). In addition many studies of white-tailed deer populations have demonstrated females disperse from or expand their home ranges in response to increased hunting pressure (Sparrowe and Springer, 1970; Kufeld et al., 1988; Naugle et al., 1997; VerCauteren and Hyngstrom, 1998; Kilpatrick and Lima, 1999). However, in cases where high harvest rates lead to more orphaned fawns, the animals exhibit smaller home ranges (Giuliano et al., 1999). Given such variation, the regional differences in habitat characteristics, deer demographics, and hunting pressure might be expected to have an impact on social behavior and transmission patterns or rates of disease.
Our goal was to assess the effect of variation in deer density, harvest intensity, and landscape patterns on home range size within the areas of highest disease prevalence in south-central Wisconsin. Based on previous findings of home range fidelity, we hypothesized home range size would not change relative to changes in density or harvest intensity but would vary based on measureable landscape features such as forest area. To date, few research efforts have related empirical information on white-tailed deer movement behavior in this disease zone to its impact on disease spread and transmission. Although demographic, statistical, and genetic modeling can provide insight on disease spread, empirical data is essential if we are to understand the local dynamics that influence the progression of the disease across the landscape.
We conducted our research in south-central Wisconsin, U.S.A. from 2002-2008 within the CWD Disease Eradication Zone (DEZ), as designated by the Wisconsin Department of Natural Resources (WDNR). Land within the DEZ in south-central Wisconsin is a mosaic dominated by forests mixed with agriculture and grassland. The average winter temperature is -8.3 C and in summer, 19.3 C. Mean monthly rainfall varies from 2.5 cm to 10.8 cm; average snowfall is 133.1 cm (Wisconsin State Climatology Office). We defined two 78 [km.sup.2] landscapes (1 and 2) in Dane and Iowa County, both within the 1809 [km.sup.2] area of the 2002 DEZ (Fig. 1, 43.006, -89.734). We selected these two landscapes to overlap areas with the highest proportions of CWD-positive deer within the DEZ (Wisconsin Department of Natural Resources, unpublished data). Within each landscape, we chose six study sites (Skuldt, 2005). These were selected to represent both high and low ratios of agriculture to forest (the dominant land use types), and we ensured each site was separated by at least 3 km to decrease the likelihood of capturing deer from the same social groups. Our study sites were on private land and we obtained permission from all landowners before initiating research on their property. All data were taken within 15 miles of Mt. Horeb, Wisconsin (43.0064[degrees]N, 89.7342[degrees]W).
TRAPPING AND HANDLING PROCEDURES
We captured white-tailed deer from Dec.-Apr. during 2002-2006 with modified Clover and Stephenson box traps, rocket nets (Hawkins et al., 1968), drop-nets (Ramsey, 1968), and darting. We aged them as fawns (<1 y), yearlings ([greater than or equal to] 1 y, <2 y), and adults ([greater than or equal to] 2 y) by tooth wear and replacement (Severinghaus, 1949) and subsequently collected incisors from recovered animals for age verification. We chemically immobilized captured deer (Skuldt, 2005). They were tested for CWD-status based on tonsillar biopsy (Wolfe et al., 2002), and CWD positive animals were later found and culled, and as such not included in this study. The University of Wisconsin-Madison (UWM), College of Agriculture and Life Sciences' Animal Care and Use Committee (Permit A-3368-01), UWM Research Animal Resources Center (A01088309-02), and the WDNR (Scientific collector's permit SCP-SCR-018-0202), approved our capture and handling methods.
We triangulated locations of radio-collared deer using three to five azimuths collected from fixed telemetry stations and obtained locations during rotating start times based on a 20 h clock, with start time varying by 4 h each day. In general we located radio-collared deer at least three times/w, though in some cases weather or other factors made this impossible (Range: 1-6, Mean: 3.2). We estimated locations using Location of a Signal (LOAS), Version 2.09 (LOAS, 2003). We accepted for analysis locations that consisted of azimuths obtained within 20 min or less and that resulted in positional errors under 0.05 [km.sup.2]. We used locations obtained at least 6 h apart to minimize the potential for autocorrelation. We defined seasons as gestation (1 Jan.-9 May), parturition (10 May-30 June), prerut (1 July-9 Oct.), and rut (10 Oct.-31 Dec.) based on breeding phenology for Wisconsin (Skuldt, 2005). Analysis of radio-telemetry location accuracy (average and standard deviation of the difference between estimated and actual locations) was assessed for the first 3 y; it was 190.84 m ([+ or -] 62.99 m) in 2003, 190.19 m ([+ or -] 99.33 m) in 2004, and 199.38 m ([+ or -] 13.85 m) in 2005 (Skuldt, 2005). In the accuracy analysis, there were nine observers in 2003, ten in 2004, and four in 2005. Inter-observer reliability tests were conducted to ensure all observers used similar methodology, and observer locations in all cases differed <5%.
HOME RANGE SIZE
We used a smoothed bootstrap approach to determine if the number of telemetry locations was sufficient to calculate a home range (Kernohan et al., 2001). Based on this analysis, we determined that bias and variance approached an asymptote at 15 or more telemetry relocations, and deemed this many locations sufficient. We calculated home ranges for all deer with at least 15 telemetry relocations during a season (Range: 15-42). For comparison we also calculated annual home ranges for all deer with over 15 telemetry relocations within a year (Range: 21-167). For the purpose of this study, we identified and removed all exploratory and dispersal movements (defined as point locations at least 1.5 home range diameters beyond this range) and recalculated the home range. Although these long-distance movements may have implications for disease dynamics, the scope of this project was to determine the factors influencing local home range size. We estimated home range using the 95% fixed kernel method (Worton, 1989) in the animal movement extension for ArcView[R] GIS, Version 3.2 (Environmental Systems Research Institute, Redlands, California, USA). We then used least square cross validation to estimate the smoothing parameter for kernel estimates (Silverman, 1986). We used log (In) transformed home ranges to achieve normality.
We included only data taken within the home range. Preliminary results showed that inclusion of data from buffer areas around the home range provided no additional predictive power, even when multiple buffers were utilized (Skuldt, 2005). We plotted home ranges for individual deer on the land cover classification LANDFIRE, which is derived from satellite imagery with 30 m resolution (LANDFIRE, 2007). We reclassified land cover into eight habitats: forest, agriculture, grassland, open water, wetland, barren, shrubland, and urban. We combined forested wetlands with forest habitat and lowland shrub with shrubland habitat for our analysis. To identify landscape predictors of home range size, we quantified the pattern of the study area using five landscape indices, identified based on importance in previous research (Nixon et al., 1991; Kie et al., 2002; McKeefry, 2003; Skuldt, 2005; Oyer et al., 2007; Storm et al., 2007; Kjaer et al., 2008; Skuldt et al., 2008; Walter et al., 2009; Felix et al. 2010): (1) amount of forest area ([km.sup.2]); (2) median forest patch size ([km.sup.2]); (3) standard deviation of forest patch size ([km.sup.2]); (4) ratio of agricultural area to forest area; and (5) forest edge density (km forest edge/[km.sup.2] forest area). We selected only forest edge density and the ratio of agriculture area to forest area to include in home range models due to high correlation ([greater than or equal to] 0.80) among other landscape indices and to retain a small model set for analysis. We also examined correlation matrices for all of the independent variables to assess multicollinearity, and included only independent variables that had low correlation (r < 0.40) in the analysis.
Using data from WDNR aerial surveys, we calculated average post-harvest deer density (deer/[km.sup.2]) within each home range. The WDNR counted deer in these surveys from January-March each year within randomly selected sections (range of number of sections sampled; 100-399, [bar.x] = 240.67, [sigma] = 123.75) within the DEZ "between 2003 and 2008 (Blanchong et al., 2006). The WDNR flew surveys within sections (2.6-[km.sup.2]) using a stratified random block design (Siniff and Skoog, 1964). This design incorporated section-sampling rates based on percent deer range. Surveys incorporated data from both public and private lands. For our purposes, we assumed estimates of observability from these surveys were accurate. The numbers of deer counted within sections in the DEZ during the aerial surveys, after accounting for observability, were highly variable between 2003 and 2008 (range = 0-301.25 deer per section, [sigma] = 176.12).
Density data were not available for the entire study area, so we used kriging to create a smoothed surface of deer density values in the study area, based on results from surveyed sections and percent of deer habitat (as estimated by the WDNR). To determine the best model to use for kriging density in this system, we fitted models for each year using generalized least squares, and developed error correlation structures for each model using the "nlme" package of the R statistical program (Venables and Ripley, 2002). We assumed that deer density would depend on percent of deer range within sections, as well as x and y coordinates, if there is a linear trend in density. We initially considered three models for each year: percent of deer range within each section, location within the DEZ (x and y section coordinates), or both. We also fitted each of those models with either a spherical or exponential spatial correlation structure. Consequently, six models were fitted for each year. We selected the optimal model for each year among the six based on Akaike's Information Criteria (Burnham and Anderson, 2002) and used this model to create the kriged surface. Therefore, in each year, the model that fit best was used to create smoothed layers of deer density at a spatial resolution of one section (2.6 [km.sup.2]). The deer density associated with each home range was the weighted average of the estimated densities in the sections the home range overlapped, with weighting based on the proportion of the home range overlapping each section. We matched density estimates for each year to the seasonal home ranges in the same year.
DEER HARVEST INTENSITY
We quantified the number of deer harvested within each section of the CWD DEZ between 2003 and 2008 using harvest records collected by the WDNR during hunting seasons (Sep.-Mar.). The WDNR required hunters to register every deer harvested within the DEZ, and assigned harvest locations to the nearest section. Each home range was assigned a harvest intensity value proportional to the area of each section that it overlapped. We matched harvest results from annual hunting seasons with seasonal home ranges such that both overlapped as close in time as possible: the 2002-2003 hunting season was linked to gestation and parturition in 2003; the 2003-2004 hunting season was linked to prerut and rut in 2003 and gestation and parturition in 2004, and so on. Our results reflect actual harvest rates that occurred as a result of normal hunting practices and enhanced management for CWD. Our design was not experimental.
We used linear mixed-effects models (Pinheiro and Bates, 2000) to analyze individual seasonal home range size using the "nlme" package of the statistical program R (Pinheiro et al., 2009). To investigate the impact of independent variables on home range size, fixed effects included forest-edge density, the ratio of agriculture to forest, the season (gestation, parturition, prerut, and rut), deer density, and harvest intensity. We included year and the identification number for each deer as separate random effects. We performed model selection for seasonal home ranges, and separated the data for analysis into categories: adult male, adult female, yearling females (1-2 y old), and yearling males.
For each sex/age class subset, we fitted each possible combination of the five fixed variables (32 models) to determine which combination best explained individual home range size (Table 1). When the number of models is small compared to the sample size (Range of n in data subsets: 106-498), this type of model choice is unlikely to lead to spurious selection of unimportant variables (Burnham and Anderson, 2002). We did not include models that included interactive terms in order to maintain a reasonably small model set, and because we had no a priori biological reason to expect particular interactions between covariates that would impact home range size. We used Akaike's information criterion adjusted for small sample sizes ([AIC.sub.c]) to select the best explanatory models, with the number of parameters in each model estimated by Program R (Version 3.1.1, R Development Core Team, 2008) and verified manually. We used variable importance weights, that is, the sum of the AIC weights of each model that contains a given variable, to determine the relative explanatory power of each variable. We used model averaging to generate parameter estimates of each independent variable (Burnham and Anderson, 2002). We calculated semi-variograms on the residuals of the regression models to test for autocorrelation using the "geoR" package of the statistical program R. We included spatial models for the errors in the mixed-effects models to account for spatial dependence if present (Pinheiro and Bates, 2000), using spherical or exponential structure.
TRAPPING AND TELEMETRY
Between January 2003 and April 2005, we captured 173 individuals (113 females, 60 males), of which five tested CWD-positive (4 females, 1 male). Of these, 157 had sufficient telemetry to be included in the analysis. We located 33 deer via radio-telemetry in 2003, 86 in 2004, 99 in 2005, 55 in 2006, 68 in 2007, and 19 in 2008, with some deer (93) located in more than 1 y. A total of 33,340 telemetry locations were used to calculate home ranges for this study.
The top model used to create a kriged surface for 2003 and 2006 included the parameters for x & y coordinates and deer range, and a spherical spatial correlation structure. For 2004, 2007, and 2008, the optimal model replaced the spherical structure with an exponential one. The best model for 2005 included only the percentage of deer range and an exponential error structure. Annual trends in average deer densities within home ranges show a reduction for the first few years, from 46.84 deer/[km.sup.2] (se = 2.08 deer/[km.sup.2]) in 2002 to 33.08 deer/[km.sup.2] (se = 0.76 deer/[km.sup.2]) in 2004, but densities rebounded in the final years of the study to levels comparable to the first year (48.57 deer/[km.sup.2] in 2008, se = 1.34 deer/[km.sup.2]). Although overall mean densities varied little, the range of densities within home ranges in the study area in any given year was broad (0-115 deer/[km.sup.2]).
In general overall harvest levels declined following the first year of the study, from 19.99 deer/[km.sup.2] in 2002 to 17.88 deer/[km.sup.2] in 2005, and settling at approximately 16.5 deer/[km.sup.2] for the remainder of the study. In particular harvest was somewhat lower in all years than it was in 2002-2003. The overall range of harvest rates in sections during the study was 1.1-39.5 deer/[km.sup.2].
HOME RANGE SIZE
Home range sizes for all sex and age classes over all seasons and years ranged from 0.7 to 3.8 [km.sup.2] (Table 2). Semi-variogram results indicated no significant spatial dependence remained after incorporating a spherical correlation error structure.
For adult females, the best model included the terms for forest edge density, agriculture to forest ratio, and season (Table 3). This model held over 99% of the total model weight, indicating a high likelihood that home range size was best explained by those variables. Yearling female home range sizes were not well explained by the independent variables we tested. Although the best model contained only one term, deer harvest, that model held just 27% of the total model weight. No variable included in the analysis for yearling females held over 33% of the variable importance weight (Table 4), indicating low explanatory power throughout (Burnham and Anderson, 2002). The best model for adult males included only season; whereas, the next best model included both season and agriculture/forest ratio. Together these models contained over 90% of total model weight, indicating a high degree of probability these two variables best explain adult male home range size. Home range sizes of yearling males were best predicted by season and agriculture/forest ratio (96.4% total model weight for the model using those variables). Parameter estimates for top models are provided (Table 5).
An understanding of the influence of landscape pattern, deer density, and harvest intensity on deer home range size as it relates to potential CWD transmission can provide insights into disease spread and potential management strategies. Such strategies may include a more refined, spatially explicit approach to reducing deer densities on a localized level in areas where CWD is present. In this study we predicted deer, and particularly adult females, would remain philopatric to home ranges, such that individual home range size would be independent of density and harvest intensity. Despite large variations in density and harvest intensity within our study area, the ratio of agriculture to forest, forest edge density, and phenological season were the only important predictors of home range size for adult females in this analysis. It is likely that socio-spatial factors (i.e., fidelity) and availability of food and cover resources, rather than hunting pressure or reductions in density, have a strong influence on home range size (Skuldt, 2005). This finding also appears to hold true for all age and sex classes. No age or sex class demonstrated a variable importance weight for the effect of density higher than 0.14, indicating that density had very low explanatory power for home range size (Burnham and Anderson, 2002). Similarly, only yearling females demonstrated a potential impact from harvest, and this relationship was not strongly supported.
We predicted home range size would be partially dependent on elements of the surrounding landscape. We found the relationship between home range size and forest edge density was negative for females, whereas males and yearling females did not exhibit a strong association with forest edge. However, even when statistically nonsignificant, the estimated co- efficient for forest edge density was negative for all sex/age classes. Forest edge density has been found to be closely linked to home range size for deer and other cervids as well (Kie et al., 2002; Said and Servanty, 2005; Walter et al., 2009; Masse and Cote, 2012). Edge in this study area is mostly due to the interface of forest and agriculture. Therefore, it is likely edge represents areas where food and cover resources are adjacent to one another, such that deer can acquire necessary resources while still occupying small home ranges. Females have smaller home ranges and may therefore be particularly sensitive to forest edge, as they need to secure food and cover within a reduced area.
Male and adult female deer showed strong positive relationships between home range size and the ratio of agriculture to forest within the home range. Therefore, the home ranges with a higher proportion of agricultural cover tended to be larger compared to those that were made up predominantly of forest cover. One possible explanation for this pattern is that only a small amount of forested land is needed for cover, but when agricultural land is available, deer browse over a much wider area to find high quality forage. It is also possible that the disturbance associated with agriculture leads deer to make frequent movements to avoid humans. Previous studies have found the extent of forest cover often most limits deer distribution and abundance (Smith, 1987; Nixon et al., 1991; McKeefry, 2003; Habib et al., 2011), particularly in mixed forest-agricultural landscapes where the size and number of forest patches are reduced (McKeefry, 2003). Although high proportions of agricultural land are associated with larger home ranges in our study system, forest edge density, usually representing forest agriculture/interfaces, is associated with smaller home ranges for adult female deer, suggesting agricultural land does still appear to provide important resources in this area.
The highly stable nature of female home range size among seasonal and annual ranges suggests highly philopatric life history strategies consistent with those previously reported (Tierson et al., 1985; Mathews, 1989; Mathews and Porter, 1993; Comer et al., 2005). The steadily increasing home range for males over longer time periods is likely a result of males making more frequent movements, particularly as yearlings (Skuldt et al., 2008; Clements et al., 2011). Home range sizes for all age classes, sexes, and seasons were slightly smaller than most reported in other mixed forest-agricultural areas throughout North America but were within the range reported throughout the Midwest (range: 0.33-5.18 [km.sup.2]; e.g., Kilpatrick and Spohr, 2000; Grand et al., 2002; Brinkman, 2003; Brinkman et al., 2005).
Our results indicate home range sizes are independent of density. This finding holds despite fluctuations of density that occurred during the study. Aerial survey results suggest deer density estimates of 0-115 deer/[km.sup.2] during our study within the core area of the DEZ, from 2003 to 2008 It appears average densities were reduced in the early part of the study, likely because of increased harvest instituted in an attempt to control the spread of CWD in the area (Blanchong et al., 2006). However, by the end of the study, densities had returned to approximately their original values. Deer densities reported in this study were those measured within home ranges in our study area and may not be representative of the wider landscape. Previous studies of white-tailed deer have related both larger (e.g., Beier and McCullough, 1990; Holzenbein and Marchinton, 1992), and smaller (Kilpatrick et al., 2001) home ranges to density, and attributed these patterns to food availability. The lack of an observed relationship between home range size and density may result because food and cover resources appear to be relatively abundant in our study area. This study represents 6 y of data collection, so if there is a causal relationship between density reduction and home range size in this system, the time lag is apparently considerable (Oyer and Porter, 2004). We conclude it is unlikely deer density has a major impact on home range size in this study, at least within the spatial (section-level) and temporal (~6 y) resolution of our data and the range of values we observed.
Our study indicates deer remain philopatric to home ranges and maintain their home range size independent of harvest intensity. Three of four sex/age classes showed no evidence of any relationship between home range size and harvest intensity in the range of harvests observed in this study area. Only yearling females showed a weak negative trend. The range of harvest rates throughout the study area was relatively wide (1.1-39.5 deer/ [km.sup.2]), probably because our study area included landowners with varying opinions on deer management strategies, some of whom did not implement intensive harvest strategies. Despite previous findings that indicated deer make wider movements in response to harvest (Sparrowe and Springer, 1970; Kufeld et al., 1988; Naugle et al., 1997; VerCauteren and Hyngstrom, 1998; Kilpatrick and Lima, 1999), we found no such relationship, and indeed yearling females displayed a negative correlation between harvest and home range: as harvest intensity increased, their home range size decreased. In general female deer in this study had their smallest home ranges during the rut season, when hunting occurred. These data assume accurate reporting of harvest locations, and represent coarse spatial resolution (section-level). Therefore, they should be examined cautiously.
High harvest pressure and high harvest rates of females (resulting in a younger female age structure) have been implicated in a breakdown in the degree of philopatry, thereby limiting the formation of cohesive social groups. A high annual female harvest combined with an elevated female dispersal rate (25%), may have limited the formation of persistent, cohesive social groups in South Carolina (Comer, 2005). Similar to the Comer study, females represented 46-54% of the total annual harvest within the core area of the Wisconsin DEZ between 2002 and 2008. Harvest data suggest that here as well, the female age structure within our study area is young However, despite high female harvest rates and a young female age structure, deer in our study system maintained their home range sizes independent of harvest intensity. Further, although we observed exploratory movements (occasional, short-term movements outside of the existing home range) among 33% (16 of 48) of yearling females and 23% (21 of 91) of adult females, we detected only one permanent dispersal of a yearling female (Oyer et al., 2007) and no confirmed dispersals of adult females. Although rates of male dispersal were comparatively high with 53% (31 of 58) engaging in exploratory behavior, and 17 individuals (29 percent) showing evidence of dispersal (Skuldt et al., 2008), male home range sizes were also unrelated to rates of harvest. We also note that changes in density or harvest beyond those we monitored have potential for a greater impact, but our study suggests that harvest regimes to date, and the resultant decreases in density, do not visibly influence deer home range size.
Our findings support the notion that localized deer reductions may create areas of low density without changing the behavior patterns of nearby deer. This phenomenon is an essential premise for future localized reduction in deer density to succeed as a disease management strategy, as it verifies that harvesting efforts will not cause deer to spread disease across a wider area. The home range estimates we provide should be useful in deriving appropriate spatial management units (Webb et al., 2010). We suggest future management will require a more refined spatial approach to density reduction in areas where CWD is present. This is particularly important where "hot spots" of CWD exist, such as mineral licks or deer scrapes, with their potential for contamination through soils (e.g., VerCauteren et al., 2007). Future research should continue to examine deer behavior, including home range size, shifts in home ranges, and movements, as densities fluctuate within the CWD management zones in south-central Wisconsin in order to gain a more comprehensive perspective about CWD transmission and its control.
Acknowledgments.--We gratefully acknowledge D. Anderson, J. Zhu, N. Keuler, and T. Van Deelen, and particularly V. St-Louis for their data analysis support. G. Allez provided valuable comments on the manuscript. We thank Drs. R. MacLean, D. Grove, W. Delanis, and T. Hoffman for their veterinary expertise. We thank R. Rolley, T. Sickley, and J. Bartelt for contributions to study design and implementation; V. Greene, J. Isabelle, and M. Lorenz for field leadership; our field technicians and volunteers; and many other Wisconsin Department of Natural Resources (WDNR) and University of Wisconsin--Madison personnel for field or analytical support. We also thank the landowners for their support and allowing us access on their property. Our study was jointly funded by the WDNR, Whitetails Unlimited, the National Beef and Cattlemen's Association, and the Northcentral Agricultural Experiment Station Hatch program.
SUBMITTED 16 JANUARY 2014
ACCEPTED 2 DECEMBER 2014
Aycrigg, J. L. and W. F. Porter. 1997. Sociospatial dynamics of white-tailed deer in the central Adirondack Mountains, New York. J. Mammal., 78:468-482.
Beier, P. and D. R. McCullough. 1990. Factors influencing white-tailed deer activity patterns and habitat use. Wildl. Monogr., 109:1-51.
Bertrand, M. R., A. J. DeNicola, S. R. Beissinger, and R. K. Swihart. 1996. Effects of parturition on home ranges and social affiliations of female white-tailed deer. J. Wildl. Manage., 60:899-909.
Bishop, R. C. 2004. The economic impacts of chronic wasting disease (CWD) in Wisconsin. Human Dimensions of Wildlife, 9:181-192.
Blanchong, J. A., D. O. Joly, M. D. Samuel, J. A. Langenberg, R. E. Rolley, and J. F. Sausen. 2006. White-tailed deer harvest from the chronic wasting disease eradication zone in South-central Wisconsin. Wildl. Soc. Bull., 34:725-731.
Brinkman, T. J. 2003. Movement and mortality of white-tailed deer in southwest Minnesota. Thesis, South Dakota State University, Brookings, South Dakota, USA.
--, C. S. DePerno, J. A.Jenks, B. S. Haroldson, and R. G. Osborn. 2005. Movement of female white-tailed deer: effects of climate and intensive row-crop agriculture. J. Wildl. Manage., 69:1099-1111.
Burnham, K. P. and D. R. Anderson. 2002. Model selection and multimodel inference: a practical information-theoretic approach. Springer-Verlag, New York, New York.
Clements, G. M., S. E. Hygnstrom.J. M. Gilsdore, D. M. Baasch, M. J. Clements, and K. C. Ver-Cauteren. 2011. Movements of white-tailed deer in riparian habitat: implications for infectious disease. J. Wildl. Manage., 75:1436-1442.
Comer, C. E. 2005. An analysis of spatial and genetic population structure in white-tailed deer with implications for management. Ph. D. Dissertation, University of Georgia, Athens, Georgia.
--, J. C. Kilgo, G. J. D'Angelo, T. C. Glenn, and K. V. Miller. 2005. Fine-scale genetic structure and social organization in female white-tailed deer. J. Wildl. Manage., 69:332-344.
Dulberger, J., N. T. Hobbs, H. M. Swanson, C. J. Bishop, and M. W. Miller. 2010. Estimating chronic wasting disease effects on mule deer recruitment and population growth. I. Wildl. Dis., 46:1086-1095.
Giuliano, W. M., S. Demarais, R. E. Zaiglin, and M. L. Sumner. 1999. Survival and movements of orphaned white-tailed deer fawns in Texas. J. Wildl. Manage., 63:570-574.
Gross, J. E. and M. W. Miller. 2001. Chronic wasting disease in mule deer: disease dynamics and control. J. Wildl. Manage., 65:205-215.
Grund, M. D., J. B. McAninch, and E. P. Wiggers. 2002. Seasonal movements and habitat use of female white-tailed deer associated with an urban park. J. Wildl. Manage., 66:123-130.
Habib, T. J., E. H. Merrill, M. J. Pybus, and D. W. Coltman. 2011. Modelling landscape effects on density-contact rate relationships of deer in eastern Alberta: implications for chronic wasting disease. Ecol. Model, 222:2722-2732.
Hawkins, R. E., L. D. Martoglio, and C. G. Montgomery. 1968. Cannon-netting deer. J. Wildl. Manage., 32:191-195.
Holzenbein, S. and R. L. Marchinton. 1992. Spatial integration of maturing male white-tailed deer into the adult population. J Mammal, 73:326-334.
Johnson, C. J.,J. A. Pedersen, R. J. Chapell, D. McKenzie, and J. M. Aiken. 2007. Oral transmissibility of prion disease is enhanced by binding to soil particles. PLoS Pathog., 7:874-881.
--, K. E. Phillips, P. T. Schramm, D. McKenzie, J. M. Aiken, and J. A. Pedersen. 2006. Prions adhere to soil minerals and remain infectious. PLoS Pathog., 2:1-7.
Joly, D. O., C. A. Ribic, J. A. Langenberg, K. Beheler, C. A. Batha, B. J. Dhuey, R. E. Rolley, G. Bartelt, T. R. van Deelen, and M. D. Samuel. 2003. Chronic wasting disease in free-ranging Wisconsin white-tailed deer. Emerging Infect. Dis., 9:599-601.
Kernohan, B. J., R. A. Gitzen, and J. J. Millspaugh. 2001. Analysis of animal space use and movements, p. 125-166. In: J. J. Millspaugh and J. M. Marzluff (eds.). Radio Tracking and Animal Populations. Academic Press, San Diego, California, U.S.A.
Kie, J. G., R. T. Bowyer, M. C. Nicholson, B. B. Boroski, and E. R. Loft. 2002. Landscape heterogeneity at differing scales: effects on spatial distribution of mule deer. Ecology, 83:530-544.
Kilpatrick, H. J. and K. K. Lima. 1999. Effects of archery hunting on movement and activity of female white-tailed deer in an urban landscape. Wildl. Soc. Bull., 27:433-440.
-- and S. M. Spohr. 2000. Spatial and temporal use of a suburban landscape by female white-tailed deer. Wildl. Soc. Bull., 28:1023-1029.
--, --, and K. K. Lima. 2001. Effects of population reduction on home ranges of female white-tailed deer at high densities. Can. J. Zool., 79:949-954.
Kjer, L. J., E. M. Schauber, and C. K. Nielsen. 2008. Spatial and temporal analysis of contact rates in female white-tailed deer./. Wildl. Manage., 72:1819-1825.
Kufeld, C. F., D. C. Bowden, and D. L. Schrupp. 1988. Influence of hunting on movements of female mule deer. J. Range. Manage., 41:70-72.
LANDFIRE: LANDFIRE National Existing Vegetation Type Layer. 2006, September--last update. U.S. Department of Interior, Geological Survey. [Online], Available: http://gisdata.usgs.net/ website/landfire/ [2007, February 8].
Laseter, B. R. 2004. Sociospatial characteristics and genetic structure of white-tailed deer in the Central Appalachians of West Virginia. Dissertation, University of Georgia, Athens, Georgia, U. S.A.
Location of a Signal[TM]. 2003. Ecological software solutions. Schwagalpstrasse 2, 9107 Urnasch, Switzerland. Version 2.09. URL: http://www.ecostats.com/index/htm.
Loft, E. R., J. G. Kie, and J. W. Menke. 1984. Seasonal movements and summer habitats of female black-tailed deer. J. Wildl. Manage., 48:1317-1325.
Long, E. S., D. R. Diefenbach, C. S. Rosenberry, B. D. Wallingford, and M. D. Grund. 2005. Forest cover influences dispersal distance of white-tailed deer. J. Mammal., 86:623-629.
Masse, A. and S. D. Cote. 2012. Linking alternative food sources to winter habitat selection of herbivores in overbrowsed landscapes. J. Wildl. Manage., 76:544-556.
Mathews, N. E. 1989. Social structure, genetic structure, and anti-predator behavior of white-tailed deer in the central Adirondacks. Dissertation, State University of New York, College of Environmental Science and Forestry, Syracuse, New York, USA.
-- and W. F. Porter. 1993. Effect of social structure on genetic structure of free-ranging white-tailed deer in the Adirondack Mountains. J. Mammal., 74:33-43.
McKeefry, J. F. 2003. Landscape structure and deer abundance: forest land cover and edge relationships across Wisconsin (USA). Thesis, University of Wisconsin--Madison, Madison, Wisconsin.
Miller, K. V. 1997. Considering social behavior in the management of overabundant white-tailed deer populations. Wildl. Soc. Bull., 25:279-281;
Naugle, D. E., J. A. Jenks, B. J. Kernohan, and R. R. Johnson. 1997. Effects of hunting and loss of escape cover on movements and activity of female white-tailed deer, Odocoileus virginianus. Can. Field Nat., 111:595-600.
Needham, M. D., J. J. Vaske, and M. J. Manfredo. 2004. Hunters' behavior and acceptance of management actions related to Chronic Wasting Disease in eight states. Human Dimensions of Wildlife, 9:211-231.
Nixon, C. M., L. P. Hansen, P. A. Brewer, and J. E. Chelsvig. 1991. Ecology of white-tailed deer in an intensively farmed region of Illinois. Wildl. Monogr., 118:1-77.
Oyer, A. M., N. E. Mathews, and L. H. Skuldt. 2007. Long-distance movement of a white-tailed deer away from a chronic wasting disease area./. Wildl. Manage., 71:1635-1638.
-- and W. F. Porter. 2004. Localized management of white-tailed deer in the central Adirondack Mountains, New York. Journal of Wildlife Management, 68:257-265.
Pinheiro, J. C., D. Bates, S. DebRoy, D. Sarkar, the R Core Team. 2009. nlme: linear and nonlinear mixed effects models. R package version 3.1-93.
-- and D. M. Bates. 2000. Mixed-effect models in S and S-PLUS. Springer, New York, New York, USA.
R Development Core Team. 2008. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www. R-project.org.
Ramsey, C. W. 1968. A drop-net deer trap. J. Wildl. Manage., 32:187-190.
Said, S. and S. Servanty. 2005. The influence of landscape structure on female roe deer home-range size. Landsc. EcoL, 20:1003-1012.
Schauber, E. M., D. J. Storm, and C. K. Nielsen. 2007. Effects of joint space use and group membership on contact rates among white-tailed deer. J. Wildl. Manage., 71:155-163.
-- and A. Woolf. 2003. Chronic wasting disease in deer and elk: a critique of current models and their application. Wildl. Soc. Bull., 31:610-616.
Severinghaus, C. A. 1949. Tooth development and wear as criteria of age in white-tailed deer. J. Wildl. Manage., 13:195-216.
Silverman, B. W. 1986. Density estimation for statistics and data analysis. Chapman and Hall, London, UK.
Siniff, D. B. and R. O. Skoog. 1964. Aerial censusing of caribou using stratified random sampling. J. Wildl. Manage., 28:391-401.
Skuldt, L. H. 2005. Influence of landscape pattern, deer density, and deer harvest on white-tailed deer behavior in south-central Wisconsin. M.S. Thesis, University of Wisconsin, Madison, Wisconsin, USA.
--, N. E. Mathews, and A. M. Oyer. 2008. White-tailed deer movements in a chronic wasting disease area in South-central Wisconsin. J. Wildl. Manage., 72:1156-1160.
Smith, W. P. 1987. Dispersion and habitat use by sympatric Columbian white-tailed deer and Columbian black-tailed deer. J. Mammal., 68:337-347.
Sparrowe, R. D. and P. F. Springer. 1970. Seasonal activity patterns of white-tailed deer in eastern South Dakota. J. Wildl. Manage., 34:420-430.
Storm, D. J., C. K. Nielsen, E. M. Schauber, and A. Woolf. 2007. Space use and survival of white-tailed deer in an exurban landscape. J. Wildl. Manage., 71:1170-1176.
Tierson, W. C., G. F. Mattfeld, R. W. Sage, Jr., and D. F. Behrend. 1985. Seasonal movements and home ranges of white-tailed deer in the Adirondacks. J. Wildl. Manage., 49:760-769.
Vaske, J. J., N. R. Timmons, J. Beaman, and J. Petchenik. 2004. Chronic wasting disease in Wisconsin: hunter behavior, perceived risk, and agency trust. Human Dimensions of Wildlife, 9:193-209.
Venables, W. N. and B. D. Ripley. 2002. Modem applied statistics with S. Fourth edition. Springer-Verlag, New York, New York, USA.
VerCauteren, K. C., P. W. Burke, G. E. Phillips, J. W. Fischer, N. W. Seward, B. A. Wunder, and M. J. Lavelle. 2007. Elk use of wallows and potential chronic wasting disease transmission. J. Wildl. Dis., 43:784-788.
-- and S. E. FIygnstrom. 1998. Effects of agricultural activities and hunting on home ranges of female white-tailed deer. J. Wildl. Manage., 62:280-285.
Walter, W. D., D. M. Baasch, S. E. Hygnstrom, B. D. Trindle, A. J. Tyre, J. J. Millspaugh, C. J. Frost, J. R. Boner, and K. C. VerCauteren. 2011. Space use of sympatric deer in a riparian ecosystem where chronic wasting disease is endemic. Wildl. Bio., 17:191-209.
--, K. C. VerCauteren, H. Campa III, W. R. Clark, J. W. Fischer, S. E. Hygnstrom, N. E. Mathews, C. K. Nielsen, E. M. Schauber, T. R. Van Deelen, and S. R. Winterstein. 2009. Regional assessment on influence of landscape configuration and connectivity on range size of white-tailed deer. Landsc. Ecol.
Webb, S. L., S. Demarais, and D. G. Hewitt. 2010. Size of home ranges and movements determine size and configuration of management units and potential spread of disease in white-tailed deer (Odocoilms virginianus). Southwest Nat., 55:488-492.
Williams, E. S., M. W. Miller, T. J. Kreeger, R. H. Kahn, and E. T. Thorne. 2002. Chronic wasting disease of deer and elk: a review. J. Wildl. Manage., 66:551-563.
Wolfe, L. L., M. M. Conner, T. H. Baker, V. J. Dreitz, K. P. Burnham, E. S. Williams, N. T. Hobbs, and M. W. Miller. 2002. Evaluation of antemortem sampling to estimate chronic wasting disease prevalence in free-ranging mule deer. J. Wildl. Manage., 66:564-573.
Worton, B. J. 1989. Kernel methods for estimating the utilization distribution in home- range studies. Ecology, 70:164-168.
SETH B. MAGLE (1)
Urban Wildlife Institute, Lincoln Park Zoo, 2001 N. Clark Street, Chicago, Illinois 60614
LESA H. KARDASH
Wisconsin Department of Natural Resources, 473 Griffith Avenue, Wisconsin Rapids 54494
ANNE OYER ROTHROCK
New York State Department of Environmental Conservation, 182 East Union Street, Suite 3, Allegany, 14706
JEROMY C. CHAMBERLIN and NANCY E. MATHEWS
Nelson Institute for Environmental Studies, University of Wisconsin, 550 N. Park Street, Madison, 53706
(1) Corresponding author: Telephone: (312) 742-7215; FAX: 312-742-7220; e-mail: SMagle@lpzoo.org
TABLE 1.--Thirty-two candidate models tested to determine best predictors of home range size of white-tailed deer (Odocoileus virginianus) in southern Wisconsin, U.S.A. from 2003-2008 using Akaike's Information Criteria. Separate analyses were conducted for adult male, adult female, yearling male, and yearling female white-tailed deer Model # Variables 1 Agriculture/Forest Ratio + Forest Edge Density + Deer Density + Harvest Intensity + Season 2 Agriculture/Forest Ratio + Forest Edge Density + Deer Density + Harvest Intensity 3 Agriculture/Forest Ratio + Forest Edge Density + Deer Density + Season 4 Agriculture/Forest Ratio + Forest Edge Density + Harvest Intensity + Season 5 Agriculture/Forest Ratio + Deer Density + Harvest Intensity + Season 6 Forest Edge Density + Deer Density + Harvest Intensity + Season 7 Agriculture/Forest Ratio + Forest Edge Density + Deer Density 8 Agriculture/Forest Ratio + Forest Edge Density + Harvest Intensity 9 Agriculture/Forest Ratio + Forest Edge Density + Season 10 Agriculture/Forest Ratio + Deer Density + Harvest Intensity 11 Agriculture/Forest Ratio + Deer Density + Season 12 Agriculture/Forest Ratio + Harvest Intensity + Season 13 Forest Edge Density + Deer Density + Harvest Intensity 14 Forest Edge Density + Deer Density + Season 15 Forest Edge Density + Harvest Intensity + Season 16 Deer Density + Harvest Intensity + Season 17 Agriculture/Forest Ratio + Forest Edge Density 18 Agriculture/Forest Ratio + Deer Density 19 Agriculture/Forest Ratio + Harvest Intensity 20 Agriculture/Forest Ratio + Season 21 Forest Edge Density + Deer Density 22 Forest Edge Density + Harvest Intensity 23 Forest Edge Density + Season 24 Deer Density + Harvest Intensity 25 Deer Density + Season 26 Harvest Intensity + Season 27 Agriculture/Forest Ratio 28 Forest Edge Density 29 Deer Density 30 Harvest Intensity 31 Season 32 (Intercept Only Model) TABLE 2.--Summary of average seasonal home range size ([km.sup.2]) for white-tailed deer (Odocoileus virginianus) in southern Wisconsin, USA from 2003 to 2008. x is the mean value, n is the sample size, and standard errors are represented by SE Gestation Parturition Prerut Sex/Age Class x SE n x SE n x SE n x Female Adult 1.14 0.18 176 0.88 0.20 123 0.96 0.25 130 0.62 Yearling 0.96 0.15 43 0.89 0.13 33 0.70 0.11 36 0.62 Male Adult 1.66 0.17 29 1.53 0.37 13 1.09 0.14 50 1.84 Yearling 2.07 0.30 32 1.66 0.36 21 1.11 0.38 29 2.23 Rut Annual Sex/Age Class SE n x SE n Female Adult 0.05 69 0.93 0.11 244 Yearling 0.08 35 1.09 0.26 48 Male Adult 0.31 34 1.61 0.17 63 Yearling 0.33 24 3.83 0.84 38 TABLE 3.--Results of model selection process using independent variables to predict white-tailed deer (Odocoileus virginianus) home range sizes in Wisconsin's Chronic Wasting Disease Deer Eradication Zone, from 2003 to 2008. Displayed are the top 5-7 linear mixed-effects models which cumulatively make up >95% of the model weight based on Akaike's Information Criterion, and the null model (intercept only). K represents the number of parameters in the model 3.1 Adult females [DELTA] Model Model K AIC weight Ratio + Edge Density + Season 10 0.000 0.992 Density + Ratio + Edge Density + Season 11 9.871 0.007 Harvest + Ratio + Edge Density + Season 11 16.291 0.000 Ratio + Edge Density 7 19.137 0.000 Harvest + Density + Ratio + Edge Density + 12 20.572 0.000 Season Null 5 23.897 0.000 3.2 Yearling females [DELTA] Model Model K AIC weight Harvest 6 0.000 0.270 Edge Density 6 1.260 0.144 Ratio + Season 9 1.309 0.140 Ratio 6 1.460 0.130 Density 6 1.480 0.129 Season 6 2.020 0.098 Harvest + Ratio 7 3.764 0.041 Null 5 7.832 0.001 3.3 Adult Males [DELTA] Model Model K AIC weight Season 8 0.000 0.487 Ratio + Season 9 0.321 0.415 Ratio 6 4.435 0.053 Edge Density + Season 9 7.261 0.013 Harvest + Season 9 7.861 0.010 Null 5 15.675 0.000 3.4. Yearling Males [DELTA] Model Model K AIC weight Ratio + Season 9 0.000 0.964 Ratio + Edge Density + Season 10 7.641 0.021 Harvest + Ratio + Season 10 9.501 0.008 Density + Ratio + Season 10 10.221 0.006 Harvest + Ratio + Edge Density + Season 11 17.194 0.000 Null 5 19.378 0.000 TABLE 4.--Variable importance weights for the variables used to predict the home range sizes for white-tailed deer (Odocoileus virginianus) in Wisconsin's Chronic Wasting Disease Deer Eradication Zone from 2003 to 2008 Adult Adult Yearling Yearling Variable female male male female Harvest 0.000 0.018 0.009 0.338 Density 0.007 0.004 0.006 0.143 Ratio 1.000 0.481 1.000 0.164 Edge Density 1.000 0.023 0.021 0.317 Season 1.000 0.939 1.000 0.273 TABLE 5.--Parameter estimates for models predicting home range sizes for white/tailed deer (Odocoileus virginianus) in Wisconsin's Chronic Wasting Disease Eradication Zone from 2003 to 2008. Estimates are derived from model averaging of all 32 models tested for each age/sex combination. The gestation season is used as the baseline reference for the "season" effect. "CI" refers to the 95% confidence interval. Asterisks indicate the variable had a variable importance weight >0.5 during model selection Adult females Yearling females Lower Upper Lower Upper Estimate 95% CI 95% CI Estimate 95% CI 95% CI Ratio 0.428 * 0.269 0.588 0.095 -0.289 0.478 Edge -0.030 * -0.041 -0.018 -0.012 -0.032 0.007 Density Season Parturition -0.187 * -0.279 -0.183 -0.148 -0.387 0.092 Prerut -0.274 * -0.365 -0.183 -0.463 -0.773 -0.152 Rut -0.225 * -0.335 -0.115 -0.332 -0.643 -0.022 Harvest 0.004 -0.005 0.012 -0.007 -0.042 0.028 Density -0.003 -0.007 0.001 0.009 -0.003 0.021 Adult males Yearling males Lower Upper Lower Upper Estimate 95% CI 95% CI Estimate 95% CI 95% CI Ratio 0.278 0.021 0.535 0.854 * 0.544 1.164 Edge 0.005 -0.018 0.027 -0.014 -0.033 0.004 Density Season Parturition -0.543 * -0.176 0.190 -0.314 * -0.612 -0.016 Prerut -0.805 * -0.542 -0.279 -0.828 * -1.128 -0.528 Rut -0.236 * -0.524 0.051 0.019 * -0.294 0.332 Harvest -0.009 -0.036 0.018 -0.004 -0.028 0.019 Density 0.002 -0.005 0.010 0.006 -0.006 0.018
|Printer friendly Cite/link Email Feedback|
|Author:||Magle, Seth B.; Kardash, Lesa H.; Rothrock, Anne Oyer; Chamberlin, Jeromy C.; Mathews, Nancy E.|
|Publication:||The American Midland Naturalist|
|Date:||Apr 1, 2015|
|Previous Article:||Habitat characteristics within a zone of separation between the ranges of two species of pocket gophers.|
|Next Article:||Effects of white-tailed deer (Odocoileus virginianus) on the maritime forest of Bald Head Island, North Carolina.|