Movement and habitat use of headstarted Blanding's turtles in Michigan

Abstract

ABSTRACTCaptive breeding or headstarting is a management option to increase population numbers in threatened and endangered animals. The success of these programs depends on increasing population numbers while maintaining fitness within populations that contain some captive‐reared individuals. As part of an ongoing conservation project with Shiawassee National Wildlife Refuge (SNWR) in Saginaw, Michigan, USA, the Detroit Zoological Society (DZS) headstarted Blanding's turtles (Emydoidea blandingii) to 18 months of age. To determine survival, movements, and habitat use of these headstarted Blanding's turtles, we externally attached radio‐transmitters to 24 turtles and released them at 4 sites in release groups of 6 along a transect in a wetland within SNWR during June 2014. We located turtles weekly during the active season (May–Sep) and every 2 weeks during the inactive season (Oct–Apr) for 18 months, starting immediately after release and ending November 2015. We calculated straight‐line distances between release sites and home range centers over the study period. We examined habitat use of the turtles in relation to habitat availability. To predict movement of headstarted Blanding's turtles away from release sites, we used geographically weighted regression (GWR) with measured habitat factors and release sites as independent variables. There were differences in home range sizes across release groups but not between years. Headstarted juvenile Blanding's turtles in this study have home range sizes similar to those measured in other studies for wild‐hatched juveniles but smaller home ranges than those measured for wild‐hatched adults. Our study showed that headstarted Blanding's turtles used habitats with muskrat (Ondatra zibethica) dens and cattails (Typha spp.) more than proportionally available, and used open water, willows (Salix spp.), and lowland forest less than proportionally available. The GWR model was able to predict habitat characteristics (water depth and temperature, duckweed [Lemna minor], cattails, muskrat dens and fields, buttonbush [Cephalanthus occidentalis]) that influenced the movement of turtles away from the release sites. The habitat coefficient's influence on movement varied in relation to the current location of the turtle and as habitat characteristics increased or decreased. Turtle distance from release sites decreased as water depth, water temperature, and duckweed increased, and increased as cattails, muskrat dens and fields, and buttonbush increased. Habitat in the small spatial extent at release sites may affect movement of headstarted Blanding's turtles. Furthermore, this study uses novel methodologies for assessing headstart programs that can aid future conservation and management efforts by providing information on habitat use and movement patterns of headstarted Blanding's turtles after release. © 2018 The Wildlife Society.