COS 7-3 - Estimating local and range-wide occupancy dynamics for federally-listed flatwoods salamanders

Monday, August 7, 2017: 2:10 PM
B110-111, Oregon Convention Center
Katherine M. O'Donnell1, William J. Barichivich1, Kevin M. Enge2, Kurt Buhlmann3, Jonathan Chandler4, Anna Farmer2, Thomas A. Gorman5, Carola A. Haas6, E. Pierson Hill2, John B. Jensen7, Mark Mandica8, Jana Mott9, John G. Palis10, Lora L. Smith11 and Susan C. Walls1, (1)Wetland and Aquatic Research Center, U.S. Geological Survey, Gainesville, FL, (2)Florida Fish and Wildlife Conservation Commission, (3)Savannah River Ecology Laboratory, University of Georgia, Aiken, SC, (4)St. Marks National Wildlife Refuge, U.S. Fish and Wildlife Service, (5)Washington Department of Natural Resources, WA, (6)Department of Fish and Wildlife Conservation, Virginia Tech, Blacksburg, VA, (7)Georgia Department of Natural Resources, (8)The Amphibian Foundation, Atlanta, GA, (9)North Florida Program, The Nature Conservancy, Bristol, FL, (10)Palis Environmental Consulting, (11)Joseph W. Jones Ecological Research Center, Newton, GA

Frosted flatwoods salamanders (Ambystoma cingulatum) and reticulated flatwoods salamanders (A. bishopi) are federally-listed pond-breeding salamanders native to the southeastern United States. Both species breed in wetlands embedded in pine-wiregrass ecosystems and have likely declined due to widespread habitat loss and degradation caused by altered fire regimes. With species recovery plans in development, it is critical to have a thorough understanding of remaining populations. In 2013, we initiated a 3-year, multi-partner effort to survey known and potential flatwoods salamander wetlands across the combined range of the two species. These species breed in late fall (October–December); all surveys targeted larval salamanders approximately 2–3 months after breeding events (January–March). All partners incorporated repeated sampling design into survey efforts, enabling us to analyze resulting data in an occupancy-based framework. We fit Bayesian models that account for the nested sampling design (i.e., multiple wetlands within eight geographically distinct survey areas), using area and year as random effects (Miller & Grant 2015).


Through this comprehensive survey effort, we discovered several previously unknown breeding wetlands, including a new site with 5 A. bishopi ponds. However, we failed to detect salamanders at many historical breeding sites across the range. From the 2013/2014 through 2015/2016 breeding seasons, we conducted 762 surveys of 479 unique wetlands. We surveyed 201, 154, and 407 wetlands per year, respectively; of those, 33 (16.4%), 19 (12.3%), and 55 (13.5%) wetlands were occupied by the target species – an overall mean of 14.0%. St Marks National Wildlife Refuge (FL) and Apalachicola National Forest (FL) contain the largest number of currently active A. cingulatum wetlands, with 23 and 21, respectively. Eglin Air Force Base (FL) harbors the largest population of A. bishopi, with 12 active wetlands. Our results illustrate the precarious range-wide status of the two species, as the most robust populations all occur in the Florida panhandle; known populations in South Carolina, Georgia, and north-central Florida have either been extirpated or are very small. Understanding the current status of these species provides the foundation for making key decisions about species recovery, including where to prioritize salvage efforts and potential reintroduction efforts.