Patterns of geographic variation in communication systems can provide insight into the processes that drive phenotypic evolution.Although work in birds, anurans, and insects demonstrates that acoustic signals are sensitive to diverse selective and stochastic forces, processes that shape variation in mammalian vocalizations are poorly understood. We quantified geographic variation in the advertisement songs of sister species of singing mice, montane rodents with a unique mode of vocal communication. We tested three hypotheses to explain spatial variation in the song of the lower altitude species, Scotinomys teguina: selection for species recognition in sympatry with congener, S. xerampelinus, acoustic adaptation to different environments, and stochastic divergence. Mice were sampled at seven sites in Costa Rica and Panamá; genetic distances were estimated from mitochondrial control region sequences, between-site differences in acoustic environment were estimated from climatic data. Acoustic, genetic and geographic distances were all highly correlated in S. teguina, suggesting that population differentiation in song is largely shaped by genetic drift. Contrasts between interspecific genetic-acoustic distances were significantly greater than expectations derived from intraspecific contrasts, indicating accelerated evolution of species-specific song. We propose that, although much intraspecific acoustic variation is effectively neutral, selection has been important in shaping species differences in song. K E Y W O R D S :Acoustic adaptation, bird song, character displacement, ecological selection, mammal, speciation.Understanding the origins of phenotypic diversity is a fundamental goal of evolutionary biology; few phenotypes are as diverse as signals used in intraspecific communication (Endler 1992;Bradbury and Vehrencamp 1998). Acoustic signals make particularly good models for signal evolution because variation is readily quantifiable, and population and species differences can accrue over short evolutionary timescales with significant impact

No abstract

Developmental stressors are increasingly recognised for their pervasive influence on the ecology and evolution of animals. In particular, many studies have focused on how developmental stress can give rise to variation in adult behaviour, physiology, and performance. However, there remains a poor understanding of whether general patterns exist in the effects and magnitude of phenotypic responses across taxonomic groups. Furthermore, given the extensive phenotypic variation that arises from developmental stressors, it remains important to ascertain how multiple processes may explain these responses. We compiled data from 111 studies to examine and quantify the effect of developmental stress on animal phenotype and performance from juveniles to adulthood, including studies from birds, reptiles, fish, mammals, insects, arachnids, and amphibians. Using meta-analytic approaches, we show that across all studies there is, on average, a moderate to large negative effect of developmental stress exposure (posterior mean effect: |d| = −0.51) on animal phenotype or performance. Additionally, we demonstrate that interactive effects of timing of stressor onset and the duration of exposure to stressors best explained variation in developmental stress responses. Animals exposed to stressors earlier in development had more-positive responses than those with later onset, whereas longer duration of exposure to a stressor caused responses to be stronger in magnitude. However, the high amount of heterogeneity in our results, and the low degree of variance explained by fixed effects in both the meta-analysis (R 2 = 0.034) and top-ranked meta-regression model (R 2 = 0.02), indicate that phenotypic responses to developmental stressors are likely highly idiosyncratic in nature and difficult to predict. Despite this, our analyses address a critical knowledge gap in understanding what effect developmental stress has on phenotypic variation in animals. Additionally, our results highlight important environmental and proximate factors that may influence phenotypic responses to developmental stressors.