An evaluation of distance sampling for large-scale gopher tortoise surveys in Georgia, USA

Abstract

Gopher tortoise, Gopherus polyphemus , populations are threatened by habitat loss across their range. Management and conservation of the species has been hindered by the lack of population monitoring data. Recent studies have demonstrated that line transect distance sampling (LTDS) is efficient for estimating population parameters for gopher tortoises, yet this method has not gained wide-spread acceptance. We attempted to use LTDS to survey gopher tortoise populations on 20 protected areas across southern Georgia, USA in 2007 and 2008. We used a camera system to determine burrow occupancy at each site. The survey data were used to compare population estimates derived using LTDS (with burrow scoping) to estimates obtained with survey methods recommended in the 2007 Florida Gopher Tortoise Management Plan: area-constrained surveys of 15% of suitable habitat and a standard 50% burrow occupancy. LTDS estimates of tortoise population density ranged from 0.21 ± 0.04 tortoises/ha at Ohoopee Dunes Natural Area to 1.65 ± 0.37 tortoises/ha at General Coffee State Park. Distance sampling was generally very efficient (on average our survey rate was 0.88 km/h) and we obtained estimates of population size and density at 13 of the 20 sites. The method was much less efficient at sites with extremely low tortoise densities, and at sites where the survey area was poorly defined. Under the former circumstance, LTDS sampling would have required 88-1318 km of transect per site, which was beyond the scope of our overall project. In the latter circumstance additional ground-truthing of the habitat would have been necessary. Hence, we reported only the tortoise encounter rate and burrow occupancy estimates for these sites. Population estimates obtained with area-constrained surveys and a 50% burrow occupancy rate differed by as much as 114% from those obtained with LTDS and occupancy estimates based on burrow scoping.