Movement behavior preceding autumn mortality for white-tailed deer in central New York

Journal of Mammalogy, 99(3):675–683, 2018 DOI:10.1093/jmammal/gyy023 Published online March 16, 2018 Movement behavior preceding autumn mortality for white-tailed deer in central New York Brigham J. Whitman, W. F. Porter, Amy C. Dechen Quinn, David M. Williams, Jacqueline Frair, H. Brian Underwood, and Joanne C. Crawford* Department of Environmental and Forest Biology, State University of New York College of Environmental Science and Forestry, Syracuse, NY 13210, USA (BJW, JF) Department of Fisheries and Wildlife, Michigan State University, East Lansing, MI 48824, USA (WFP, ACDQ, DMW, JCC) USGS Patuxent Wildlife Research Center, State University of New York College of Environmental Science and Forestry, Syracuse, NY 13210, USA (HBU) * Correspondent: A common yet largely untested assumption in the theory of animal movements is that increased rates and a wider range of movements, such as occurs during breeding, make animals more vulnerable to mortality. We examined mortality among 34 white-tailed deer (Odocoileus virginianus) wearing GPS collars during the autumn breeding season of 2006 and 2007 in a heavily hunted, forest-agricultural landscape of central New York state. We evaluated whether individuals having higher rates of movement incurred higher rates of mortality and whether mortality risk was higher when deer were in less familiar areas. We used a Cox proportional hazards model to analyze how mortality risk changes with movement rates measured over 3 time periods: < 1 day, up to 2 weeks prior to death, and 3–4 weeks prior to death. Overall, deer increased their movement rates as autumn progressed, males more so than females. However, deer that died moved at a slower rate relative to surviving deer up to 2 weeks prior to death (β = −2.22 ± 0.81; 95% confidence interval [CI] = −3.91 to −0.51) and a slower rate on their day of death compared to deer that survived (β = −1.77 ± 0.73; 95% CI = −3.19 to −0.33). Site familiarity was not significantly related to mortality risk. Deer were equally likely to die within their 50% core use area as elsewhere within their autumn home range. We hypothesize that increased sociality associated with breeding may make animals more vulnerable to harvest mortality. Our findings contradict general assumptions about the influences of movement behavior on mortality risk, suggesting that patterns may be sensitive to the spatiotemporal context of the movement analysis. Key words: breeding season, Cox proportional hazards, harvest, mortality risk, movement rate, Odocoileus virginianus, survival, white-tailed deer A major principle defining the relationship between animal behavior and population ecology is that movement rate affects survival. More specifically, an increasing rate of movement or movement into less familiar areas increases exposure to a host of mortality factors. Increased movement generally increases the time an animal is at risk of predation because more time spent foraging, or foraging over a wider area, exposes an animal to more predators (Daly et al. 1990; Lima and Dill 1990; Werner and Anholt 1993; Norrdahl and Korpimaki 1998; Brown and Kotler 2004). In human-dominated landscapes, moving more often increases mortality risk by increasing encounters with mortality sources such as road traffic and hunters (Casagrandi and Gatto 1999; Flather and Bevers 2002; Fahrig 2007). The notion that increased rates of movement, as well as movements in unfamiliar areas, carry a higher mortality risk has been tested empirically in only a few studies. These studies indicate that cervids exhibit considerable behavioral plasticity in their response to mortality risks (Kilgo et al. 1998; D’Angelo et al. 2003; Frair et al. 2007; Ciuti et al. 2012; Little et al. 2016). Several studies have noted changes in movement behavior associated with the presence of predators or human hunters (D’Angelo et al. 2003; Laundre et al. 2010; Little et al. Published by Oxford University Press on behalf of American Society of Mammalogists 2018. This work is written by (a) US Government employee(s) and is in the public domain in the US. 675 676 JOURNAL OF MAMMALOGY 2016; Simoneaux et al. 2016). Ciuti et al. (2012) found that harvested elk (Cervus canadensis) used more open habitats and had higher movement rates than elk that survived the hunting season. Cleveland et al. (2012) also found that elk in Montana moved faster during hunting seasons. As observed for other cervids, greater movement rates by whitetailed deer (Odocoileus virginianus) during the autumn breeding season, or rut, are expected to increase vulnerability to mortality from a number of causes, including hunter harvest and deervehicle collisions (DVCs—Conover et al. 1995; Kilgo et al. 1998; Sudharsan et al. 2006; Little et al. 2014). Most males expand their home ranges during the rut or make excursions outside of their home range in an effort to find and court estrus females and breeding behavior may cause them to spend more time in less familiar areas (Larson et al. 1978; Marchinton and Hirth 1984; Nixon et al. 1991, 1994; Karns et al. 2011; Foley et al. 2015). Younger, subordinate bucks disperse in autumn due to breeding competition with older, more dominant bucks (Kammermeyer and Marchinton 1976; Tierson et al. 1985), exposing them to unfamiliar areas and presumably greater mortality risk. Females, in contrast to males, may be less vulnerable because they are philopatric and tend to display greater fidelity to their home range during autumn than do males (Nelson and Mech 1992; Mathews and Porter 1993; Aycrigg and Porter 1997), presumably lowering their mortality risk. The combination of breeding behavior and hunting disturbance can further elicit increased movements (Gleason and Jenks 1993; Naugle et al. 1997), leaving deer vulnerable to other sources of mortality (Puglisi et al. 1974; Swenson 1982; Nelson and Mech 1984; Nixon et al. 1994). Landscape composition and configuration also influence movement behavior and mortality risk. Forests generally provide security cover for deer, but the amount and arrangement of forest cover, and its proximity to roads, can affect an animal’s susceptibility to predation and human harvest (Sparrowe and Springer 1970; Nixon et al. 1991; Foster et al. 1997). Deer in homogenous landscapes tend to range more widely to obtain required resources, whereas deer living in fragmented landscapes tend to move less and maintain smaller home ranges (Beier and McCullough 1990; Etter et al. 2002; Kie et al. 2002; Dechen Quinn et al. 2013). Roads and road networks increase mortality risks directly through DVCs (Etter et al. 2002; Ng et al. 2008) and indirectly by providing access to hunters (Fuller 1990; Kilgo et al. 1998; Frair et al. 2008). White-tailed deer alter their movement behavior to avoid roads, open earlysuccessional habitats, and hunting zones, as well as the timing and speed of movements, in the presence of hunters (Kilgo et al. 1998; Little et al. 2014; (...truncated)