| This Ph.D. thesis takes advantage of new spatial and molecular genetic technologies to examine the landscape ecology and landscape genetics of a local timber rattlesnake population, with a focus on how critical habitat features influence observed patterns of movement and genetic variation at microgeographic scales. Chapter one is an empirical study that addresses a theoretical question in population genetics, namely: how does one conceptualize groups of individuals within genetic demes that are homogenous with respect to ecological interactions, but are not homogenous with respect to genetic mixing? To address this question, I used radio-telemetry derived movement data and microsatellite-DNA derived genotype data to examine patterns of genetic variation within communal timber rattlesnake hibernacula, which had been previously hypothesized to represent discrete breeding populations or metapopulation subunits. Study results demonstrated that communal hibernacula from our sampling area did not represent genetic demes, or metapopulation subunits, but were best represented as trait groups (Wilson 1975, 1977, 1980). Results from Chapter one showed that the system of mating for the sample population was random, despite apparent avoidance of inbreeding within hibernacula. In Chapter two, I examined whether mate searching behaviors associated with avoidance of inbreeding were random or non-random, and evaluated how landscape factors may contribute to such behaviors. Although low sample sizes inhibited rigorous statistical analyses for this chapter, I confirmed earlier work suggesting that male timber rattlesnakes may employ a combination of random and non-random search strategies in locating receptive females, and that landscape-specific factors may play a role in influencing search strategies. Chapter three is a landscape ecology study that explores variation in edge proximity within different habitat classes used by sample timber rattlesnakes, using a new method that has been recommended for this type of analysis. Analyses uncovered some variation in edge proximity among habitat classes, but for most edge types within habitat classes, individuals were not significantly closer to edges than expected. This finding implies that researchers may need information about edge responses at the appropriate spatial scale and time frame when conducting studies of habitat selection. |
