Ecology and genetics of the Virginia opossum in an agricultural landscape

William S Beatty, Purdue University

Abstract

Habitat loss and fragmentation in the Midwestern United States has drastically altered native habitats in the region, leading to dramatic changes in ecological communities. The ecology of a species may be modified to adapt to the novel environmental conditions produced by habitat fragmentation, resulting in changes in genetic composition, resource selection, and demography of a species. In this study, I examined the ecology of a generalist mesopredor, the Virginia opossum (Didelphis virginiana), in an agriculturally fragmented landscape in north-central Indiana, USA. In Chapter 1, molecular analyses revealed low levels of genetic structure among local populations (i.e. forest patches). Relatedness within patches was highly variable with 7% of local populations exhibiting significantly higher levels of relatedness compared to random expectations. I observed no evidence of sex-biased dispersal, despite previous field studies that have documented male-biased dispersal in the Virginia opossum. Overall, low levels of genetic structure in the opossum might be attributed to low levels of site fidelity, the dispersal capabilities of both sexes, and the ability of females to travel with young in the pouch. In Chapter 2, I examined home range and resource selection with a specific focus on use of anthropogenic habitat types (e.g. agriculture, developed land, roads). Home range and core area size did not differ between males and females, but I observed a significant increase in overall home range and core area size during the post-breeding season (i.e. late summer, early fall). Resource selection functions were developed separately for each season (breeding, post-breeding, and winter) and at two different spatial scales. The second-order scale compared resource use of individuals to resources that were not used at the population level. The third-order scale compared resource use and non-use within individuals. Resource selection models revealed that both anthropogenic and natural habitats were selected throughout the year at both spatial scales examined. Interestingly, roads were significantly selected at both spatial scales throughout the year indicating that roads may provide a reliable source of carrion for opossums in agriculturally fragmented landscapes. In Chapter 3, I examined spatial and temporal variation in the demography of opossums based on a 6-year capture-mark-recapture dataset. Although density estimates were high (8.01 opossums / km2 ) compared to other studies in relatively intact landscapes, overall density and female density significantly varied both spatially and temporally. Additionally, I examined the influence of landscape and patch-level habitat attributes on opossum density, female density, and density variation. Grass coverage within forest patches had a negative relationship with overall density and female density, emphasizing the importance of woody cover for the opossum. Increased patch size had a stabilizing influence on opossum density estimates. My research illustrates the Virginia opossum is a proficient disperser that is capable of using a diverse array of habitats in an agriculturally fragmented landscape.

Degree

Ph.D.

Advisors

Rhodes, Purdue University.

Subject Area

Wildlife Management|Ecology|Zoology

COinS