Journal of Mammalian Evolution, Vol. 11, Nos. 3/4, December 2004 ( C  2004) Are Megabats Big? James M. Hutcheon 1,3 and Theodore Garland Jr. 2 Traditionally, bats (Order Chiroptera) are divided into two suborders, Megachiroptera (“megabats”) and Microchiroptera, and this nomenclature suggests a consistent difference in body size. To test whether megabats are, in fact, significantly larger than other bats, we compared them with respect to average body mass (log transformed), using both conventional and phylogenetic statistics. Because bat phylogeny is controversial, including the position of megabats, we employed several analyses. First, we derived two generic-level topologies for 101 genera, one with megabats as the sister of all other bats (“morphological” tree), the other with megabats as the sister of one specific group of microbats, the Rhinolophoidea (“molecular” tree). Second, we used a recently published “supertree” that allowed us to analyze body mass data for 656 species. In addition, because the way body mass has evolved is generally unknown, we employed several sets of arbitrary branch lengths on both topologies, as well as transformations of the branches intended to mimic particular models of character evolution. Irrespective of the topology or branch lengths used, log body mass showed highly significant phylogenetic signal for both generic and species-level analyses (all P ≤ 0.001). Conventional statistics indicated that megabats were indeed larger than other bats (P ≪ 0.001). Phylogenetic analyses supported this difference only when performed with certain branch lengths, thus demonstrating that careful consideration of the branch lengths used in a comparative analysis can enhance statistical power. A conventional Levene’s test indicated that log body mass was more variable in megabats as compared with other bats (P = 0.075 for generic-level data set, P ≪ 0.001 for species-level). A phylogenetic equivalent, which gauges the amount of morphospace occupied (or average minimum rate of evolution) relative to topology and branch lengths specified, indicated no significant difference for the generic analyses, but did indicate a difference for some of the species-level analyses. The ancestral bat is estimated to have been approximately 20–23 g in body mass (95% confidence interval approximately 9–51 g). KEY WORDS: Body size, comparative method, Chiroptera, disparity, echolocation, independent contrast, Megachiroptera, Microchiroptera, phylogenetic signal, supertree, systematics. INTRODUCTION Dobson’s subordinal division of the Chiroptera into Megachiroptera (Pteropodidae) and Microchiroptera (16 other families) represents one of the most fundamental taxonomic dis- tinctions in any order of mammals (Dobson, 1875). The morphological differences between the two suborders have been well documented, and are listed exhaustively in Pettigrew et al. (1989). Although body size per se has not been used as a character in phylogenetic analyses, 1 Department of Biology, Georgia Southern University, Statesboro, Georgia, USA. 2 Department of Biology, University of California, Riverside, California, USA. 3 To whom correspondence should be addresse at Department of Biology, P.O. Box 8042, Georgia Southern University, Statesboro, Georgia 30460-8042, USA. E-mail: jhutcheo@georgiasouthern.edu 257 1064-7554/04/1200-0257/0 C  2004 Springer Science+Business Media, Inc.