Home Range, Activity, Movement, and Habitat Selection of the Flattened Musk Turtle (Sternotherus depressus) in the Bankhead National Forest, Alabama
Jenkins, Arthur Joseph
Type of DegreeMaster's Thesis
MetadataShow full item record
The Flattened Musk Turtle, Sternotherus depressus, is an imperiled aquatic species endemic to the Upper Black Warrior watershed in Alabama. As one of the most understudied turtles in the United States, little is known about its habits. This is especially true of their spatial ecology, one of the most important fields of ecological knowledge to inform management and conservation practices. To fill this information gap, this study employs radio telemetry, trapping, habitat, and wading (visual encounter) surveys to explore aspects of S. depressus spatial ecology in Bankhead National Forest (BNF) by describing home range and areas of core use, identifying factors that affect activity and movement, and modeling habitat selection on multiple levels (second-order, or population level; third-order, or patch level; and fourth-order, or microhabitat level). Home ranges, quantified as stream length inhabited, of 21 individuals averaged 332 m, ranging from 22 to 957 m. Areas of core stream use were also quantified as stream length by kernel density estimation using the Sheather-Jones plug-in method for individual bandwidth selection. Average 95 and 50 % kernel lengths of core use for 14 individuals were 185 (varying from 43 to 772 m) and 46 m (varying from 9 to 201 m), respectively. Activity, defined as a turtle being exposed instead of under refuge cover, increased with precipitation, but not with temperature, and peaked late in the evening. Movements increased with precipitation and temperature and were greater during breeding/nesting season (April to July) as compared to postnesting season (August to October). Overall, availability of bedrock and detached rock substrate/cover were the most important factors positively affecting habitat selection across scales. Snail availability was only a significant factor at patch-scale selection. Stream width and depth were identified in top models as having a positive affect on population- and habitat-patch-scale selection, respectively, but effects were not significant. These data help to inform management questions such as what length of stream is needed to maintain a viable population, how and when is best to survey for S. depressus, and what are the features of suitable habitat? As concern for this species among agencies and organizations rises with continued declines across its historic range, our results offer key information to better advise future conservation efforts.