NUSSEAR, K. E.*; INMAN, R. D.; TRACY, C. R.; USGS Western Ecological Research Center; University of Nevada Reno; University of Nevada Reno: Application of modeling activity and sampling availability toward improving monitoring of an elusive species

The USFWS is currently assessing the densities of desert tortoise using data from transects and Distance Sampling to calculate population density. This approach requires weighting the data from transects by a measure of the proportion of animals that are unavailable to be sampled due to their unavailability to be sampled because they are sequestered in underground burrows or other underground shelters. Currently, the approach to quantify the proportion of animals available to be sampled is to measure the percent of a small sample of focal animals that are active outside of burrows. The proportion of animals active is typically averaged for the entire season, and also for several geographic regions. This approach to accounting for population-level animal behavior may be misleading for several reasons, including both the statistical and biological consequences of regional and temporal differences in activity patterns, and the statistical effects of the small sample sizes used to estimate activity. The activity of tortoises is generally high during the spring months, and declines steadily as the summer approaches, for example in 2004 there was a significant linear decline in activity from April to late June when transects were conducted. In addition the encounter rate on tortoise surveys declines throughout the season. Our analyses of the effect of time and space variance in availability of tortoises to be sampled, suggests that the accuracy of estimates of tortoise density is currently inadequate and can be improved by integration of more precise models of behavior.