Tuesday, August 4, 2009 - 8:40 AM

OOS 10-3: Habitat characteristics of areas inhabited by gopher tortoises (Gopherus polyphemus) at camp shelby training site: Implications for management and future studies

Jeanne Jones, Katherine Edwards, David L. Evans, Scott Roberts, and Tyler Harris. Mississippi State University

Background/Question/Methods

Camp Shelby Training Site (CSTS) is a Mississippi Army National Guard installation located on State, Department of Defense, and U.S. Forest Service (USFS) lands in south Mississippi. The CSTS is located within the distributional range of the federally-listed gopher tortoise (Gopherus polyphemus). CSTS managers employ management for habitat enhancement for tortoises in fulfillment of the CSTS Integrated Natural Resource Management Plan, the Endangered Species Act, and Special Use Permit from the USFS. Degraded forest habitat conditions can cause tortoises to relocate into ruderal areas to the detriment of tortoises  and land managers. Habitat  management practices include longleaf pine (Pinus palustris) reforestation, control of invasive species, and use of prescribed fire.  In 2007, we conducted a study to assess habitat conditions surrounding tortoise burrows (n=90 burrow plots) and in areas that lacked burrows (n=123 random plots). We measured vegetation coverage, composition, and structure at burrow and non-burrow locations. Stepwise logistic regression was used to identify habitat variables that influenced presence of tortoise burrows. Explanatory variables included in our model were coverage of understory grasses, legumes, vines, ferns, midstory, bare ground, and detritus; midstory stem counts and species richness; percentage of overstory canopy coverage, and mean litter depth (cm).

Results/Conclusions

A significant model (P < 0.001) was developed for prediction of tortoise burrow occurrence. Explanatory variables that were positively related to  burrow occurrence included understory legume coverage, midstory woody coverage, percentage of overstory pine, and bare ground and debris coverage. With cross-validation procedures, we correctly predicted the presence or absence of burrows for 83.9 % of the observed outcomes. We suggest  that tortoises in our study existed in suboptimal habitat conditions due to dense midstory coverage and limited frequency of prescribed burning over long durations. Previous studies at CSTS that evaluated effects of dormant season and growing season burns conducted over <3 years indicated that shrub cover was initially reduced post-burn, but stem densities increased in following years. This study’s findings suggested that multiple years of prescribed fire combined with additional management measures were necessary for restoration of optimal habitat conditions for tortoises. This approach may be important considering fire hazards and burning bans implemented in 2006-2007 due to drought conditions and Hurricane Katrina’s impacts.  We suggest that habitat evaluation and modeling efforts can  be used to determine habitat management and monitoring needs over the long term.  Our findings have implications for designing future models for habitat quality assessment for tortoises.