Detecting trends in desert tortoise population growth: elusive behavior inflates variance in estimates of population density

Abstract

Assessing the recovery of the federally listed Mojave population of desert tortoises Gopherus agassizii requires detecting subtle changes in population size over a period of many years. The methods that have been employed by the US Fish and Wildlife Service to estimate population density of desert tortoises are inadequate for detecting modest trends in population density, partly due to a hidden variance in a parameter (g0) that corrects for the proportion of tortoises that are inactive and thus unavailable for sampling when population density is being assessed. We used small dataloggers to record the activity of tortoises throughout their active season, and derived daily esti- mates of g0, which we compared with the 2004 estimate published by the US Fish and Wildlife Ser- vice. Due to the substantial variation in animal activity that we found within and among days during the active season, we used a daily method to estimate density. We found that this method, while pro- viding a more accurate assessment of g0, translates into estimates of density with coefficients of vari- ation that are 4 larger than previously reported due to the variation in g0 and n (number of animals encountered). This discrepancy could adversely influence managers' perceptions of population recovery for desert tortoises, and could undermine any ability to monitor the efficacy of recovery actions for populations.