Counts of burrow entrances have been positively correlated with densities of semi-fossorial rodents and used as an index of densities. We evaluated their effectiveness in indexing densities of Townsend's ground squirrels (Spermophilus townsendii) in the Snake River Birds of Prey National Conservation Area (SRBOPNCA), Idaho, by comparing burrow entrance densities to densities of ground squirrels estimated from livetrapping in 2 consecutive years over which squirrel populations declined by >75%. We did not detect a consistent relation between burrow entrance counts and ground squirrel density estimates within or among habitat types. Scatter plots indicated that burrow entrances had little predictive power at intermediate densities. Burrow entrance counts did not reflect the magnitude of a between-year density decline. Repeated counts of entrances late in the squirrels' active season varied in a manner that would be difficult to use for calibration of transects sampled only once during this period. Annual persistence of burrow entrances varied between habitats. Trained observers were inconsistent in assigning active-inactive status to entrances. We recommend that burrow entrance counts not be used as measures or indices of ground squirrel densities in shrubsteppe habitats, and that the method be verified thoroughly before being used in other habitats.
