PS 1-18
The influence of wetland characteristics and spatial configuration on occupancy of larval Reticulated Flatwoods Salamanders (Ambystoma bishopi)

Monday, August 10, 2015
Exhibit Hall, Baltimore Convention Center
Houston C. Chandler, Department of Fish and Wildlife Conservation, Virginia Tech, Blacksburg, VA
Thomas A. Gorman, Washington Department of Natural Resources, WA
Kelly C. Jones, Department of Fish and Wildlife Conservation, Virginia Tech, Blacksburg, VA
Carola A. Haas, Department of Fish and Wildlife Conservation, Virginia Tech, Blacksburg, VA
Background/Question/Methods

Population declines in pond-breeding amphibians are commonly attributed to degradation of breeding sites and reproductive failure due to changing hydrologic conditions.  Reticulated Flatwoods Salamanders (Ambystoma bishopi), a federally endangered species, breed exclusively in ephemeral wetlands embedded within longleaf pine forests.  Altered fire regimes have shifted the state of vegetation characteristics at many breeding sites used by flatwoods salamanders.  Further, shortened hydroperiods documented during recent years have resulted in complete reproductive failure at many of the last known breeding sites.  We examined how a suite of wetland and environmental characteristics (including wetland depth, hydroperiod, herbaceous vegetation cover, distance to the nearest occupied wetland, and Palmer Hydrologic Drought Index [PHDI]) related to occupancy, detection probability, extinction probability, and colonization probability of larval flatwoods salamanders.  We monitored water levels and sampled for larval flatwoods salamanders at 18 previously identified breeding wetlands on Eglin Air Force Base in the panhandle of Florida during seven breeding seasons from 2008-2014.

Results/Conclusions

Larval occupancy was best predicted by the distance to the nearest occupied breeding wetland and the PHDI.  Breeding wetlands that were closer to a wetland occupied at least once from 2003-2006 had a higher probability of being occupied than sites that were more isolated.  The probability of a site being occupied was nearly three times lower for sites >200m from a site occupied from 2003-2006.  Further, the probability of a site being occupied decreased in years when PHDI was negative (indicating drought conditions).  The detection probability for flatwoods salamander larvae varied among breeding seasons (Mean = 0.48; Median = 0.60).  The extinction probability of breeding wetlands remained constant across years and occupied wetlands had a relatively low probability of extinction (0.09).  Similarly, the probability of a wetland being colonized was low (0.03).  Our results, indicate that connectivity among sites may be a critical component to recovery of this endangered species.  Further, these data combined with results from other ongoing work suggest that this species may be at higher risk of failed reproductive attempts in future years if drier conditions persist during the breeding season as predicted by climate change models.