Geographic variation in the echolocation calls of the hoary bat (Lasiurus cinereus)

Use of bat detectors to perform inventories, determine activity, and assess differential use of habitats has become a generally accepted method. However, there has been vigorous disagreement as to the level of efficacy, primarily relating to the ability to distinguish certain species and groups of species. The primary explanation suggested for the inability to identify certain species is due to the magnitude of intraspecific variation resulting in overlap among species, presumably compounded by geographic variation. Lasiurus cinereus has been identified as exhibiting the greatest degree of geographic variation including recent findings of distinct variation between populations in Hawaii and Manitoba. We find that claims of geographic variation have not been proven because of small sample size and lack of adequate description of method, including the behavior of the bat and the context during which bats were recorded. Previous geographical comparisons of species have relied on standard statistical methods that do not allow a comprehensive examination of the range in variation of diagnostic call parameters. We present data from a broad range of sites throughout mainland United States and Hawaii, and compare a multivariate statistical approach with repertoire plots of characteristic frequency versus call duration. Although we demonstrated a statistical finding of geographic variation in L. cinereus, small sample size, context, and behavior could not be discounted as the proximal cause of observed variation. The perceived variation across the geographic range that we sampled did not affect our ability to identify the species by call structure. We suggest methods for future studies of geographic variation.