This research project was conducted to describe habitat use of sika deer (Cervus nippon) and white-tailed deer (Odocoileus virginianus) and possibly attribute the effects of ungulate herbivory to specific deer species, if spatial separation in habitat use could be identified. Sturm (2007) conducted an exclosure study to document the effect of feral horse (Equus caballus) herbivory, deer herbivory, and horse and deer herbivory combined on plant communities. Sturm (2007) found that ungulate herbivory reduced plant species richness, evenness, and diversity in the maritime forest and affected species composition in all habitats studied. Sturm (2007) also found that herbivory on some species could be directly attributable to either horse or deer. However, the effects of sika and white-tailed deer herbivory could not be separated via an exclosure study design because of the difficulty of passively excluding one deer species but not the other.

We captured white-tailed deer and sika deer in January–March of 2006 and 2007 throughout the Maryland portion of Assateague Island. Deer were fitted with radio-collars and their survival and locations monitored via ground telemetry. Up to four locations were acquired per deer each week during early (May–June) and late (August–September) growth periods for vegetation on the island. Also, we estimated deer locations during a dormant vegetation period (November– December 2006). We used these data to estimate survival and harvest rates, document movements, and model habitat use.

We captured and fitted 50 deer with radio-collars over the course of the study. Of these 50 deer, 36 were sika and 14 were white-tailed deer. Of the 36 sika deer, 10 were harvested, three were likely killed by hunters but not recovered, and one died of natural causes while giving birth. Of the 14 white-tailed deer, three were harvested, one was illegally killed, and two were censored because of study-related mortality.

Annual survival was 0.48 (95% CI = 0.16–0.82) for male white-tailed deer, 0.74 (95% CI = 0.44–0.91) for female white-tailed deer, 0.56 (95% CI = 0.35–0.75) for male sika deer, and 0.86 (95% CI = 0.70–0.94) for female sika deer. The harvest rate was 0.12 (95% CI = 0.04–0.27) for female sika deer, 0.44 (95% CI = 0.25–0.65) for male sika deer, 0.18 (95% CI = 0.05–0.51) for female white-tailed deer, and 0.38 (95% CI = 0.10–0.78) for male white-tailed deer. Annual survival rates for both species were similar to what has been observed in other populations. Unfortunately, small sample sizes for male white-tailed deer limited inferences about harvest and survival rates, but harvest rates of females for both species were similar to other published studies. Hunting was the primary cause of mortality, and outside the hunting season survival was 0.98–1.00 for all species and sexes.

We found that the home range area of sika deer was much greater than the home range area of white-tailed deer, but failed to detect any difference between sexes or among seasons. Sika deer also made long-distance movements and left the Maryland portion of Assateague Island. No sika deer left Assateague island during our study, but we did document the dispersal of a male whitetailed deer to the mainland. In their native range, sika deer have been able to readily expand populations and occupy vacant habitat (Kaji et al. 2000; Kaji et al. 2004). The long distance movements we observed on Assateague Island, especially relative to white-tailed deer, may reflect the ability of this species to exploit food resources that may be limited in quality or quantity, or both. However, we did not collect data to assess use of food resources by sika deer and whether this may have influenced long distance movements.

We found both species of deer were less likely to use a habitat the further it was located from cover, which was defined as tall shrub or forest vegetation. For every 10 m (32 ft) from cover each species of deer was 1.23–1.38 times less likely to use any given habitat.

Patterns in use of vegetation classes were similar across species and seasons. Relative to forest habitat, both species avoided dune herbaceous, disturbed lands, sand, and water categories. Both species neither avoided nor preferred developed herbaceous, low shrub, marsh herbaceous, and tall shrub categories compared to the forest category. However, there were consistent differences between the two species. During spring, white-tailed deer were more likely than sika deer to use forested, tall shrub, disturbed herbaceous, and sand areas, but were less likely to use all other habitats. During summer, habitat use was similar between the two species except that white-tailed deer tended to use forested habitat more. During winter, white-tailed deer were less likely to use dune herbaceous, low shrub, and forested habitats than sika deer.

Sturm (2007) identified differential browsing on plant species between horses and deer, but his experimental design did not permit detection of differential browsing between sika and whitetailed deer. Our study of habitat use did not provide information to identify plant species that may be differentially consumed by sika and white-tailed deer based on differences in habitat use. We envision two approaches to addressing the effects of deer browsing. One approach would be further research that identifies the food habits of both deer species at the plant species level. This would be similar to food habits research conducted by Keiper (1985) and others or could involve direct observation of food consumption by both species. However, both fecal analysis and direct observation would be time-consuming and not guaranteed to identify differences. 

If the goal of ungulate population management is to protect the island ecosystem, another approach involving manipulation of deer abundance and monitoring the response of plant species known to be preferentially consumed by deer would be a more direct method of assessing effects of deer herbivory (Sturm 2007). Moreover, such an approach is not predicated on detecting differences between deer species. Direct manipulation of deer abundance could be incorporated into an adaptive management program (Williams et al. 2007) and may provide greater benefits to the management of ASIS in the long term. Harvest management decisions for white-tailed deer and sika deer are made on an ongoing basis and by coupling these decisions with a vegetative monitoring program it may be possible to reduce or minimize adverse effects of ungulate herbivory. Furthermore, management of feral horses could be incorporated into the decision process.