A comparison of 3 common methods for monitoring gopher tortoises

Abstract Wildlife biologists choose monitoring techniques for threatened and endangered species that optimize time, resources, data quantity, and data quality. We compared 3 commonly used methods of assessing population size in gopher tortoises ( Gopherus polyphemus ), including burrow censuses, line‐transect distance sampling (LTDS) surveys, and mark‐recapture surveys. To test how season and year might affect the inferences from surveys, we conducted surveys across multiple seasons during a 2‐year continuous interval. We found that mark‐recapture estimates yielded tight and consistent estimates of population size, regardless of the statistical model used to analyze the mark‐recapture data. Then, we compared results from the mark‐recapture study to burrow censuses and LTDS surveys. Our comparison indicated that LTDS surveys had lower variation when the estimates from 2 consecutive years were pooled within each season. Burrow censuses consistently underestimated population size by approximately 19–35%, depending on the sampling year, and did not produce results with confidence intervals. Line‐transect distance sampling surveys resulted in mean population size estimates that were consistently lower than mark‐recapture results by approximately 9%, though LTDS 95% confidence intervals consistently included the mark‐recapture estimate. Our results underscore how not all studies are optimized by using the same survey method. Our results also demonstrate the risk that single‐visit population surveys likely underestimate population size, which has implications for the mitigation and translocation management of Gopherus tortoises as well as efforts to quantify the current status of Gopherus species.