Evolution of intraspecific variation in the advertisement call of a cricket frog (Acris crepitans, Hylidae)

Abstract

Advertisement calls of the cricket frog, Acris crepitans, show statistically significant variation among populations in all call variables measured. Call variables show strong clinal variation resulting in calls of lower frequency, longer duration and slower call rates produced by A. c. blanchardi in open habitat in the west of the range, and calls of higher frequency, shorter calls and faster call rates produced by A. c. crepitans in the pine forests in the eastern part of the range. This clinal variation does not result from pleiotropic effects of body size or any other morphological characters we measured. The two subspecies usually reside in different habitats, but some A. c. blanchardi reside in an isolated pine forest in central Texas. By comparing the calls of this subspecies in open and forest habitat, and by statistically removing the effects of clinal variation for all populations, we determined that habitat explains some of the variation in call structure; this is not true of subspecies. Our data reject several hypotheses that purport to explain the evolution of mate recognition signals. (1) We reject the notion of Paterson and others that there is strong stabilizing selection on species-specific mate recognition signals. (2) There is no support for the hypothesis that call variation is primarily due to pleiotropic effects of body size or other morphological characters over the geographic range we examined. (3) There is no evidence for reproductive character displacement. (4) Our data, as well as experimental studies of habitat acoustics, support the hypothesis that some differences in calls among habitats result from environmental selection on call structure to enhance call transmission. We suggest that the latter hypothesis does not explain the strong clinal component of call variation. This might result from the passive effects of gene flow between populations at the extremes of the range under selection generated by habitat acoustics.