Call Characteristic Network Reveal Geographical Patterns of Call Similarity:Applying Network Analysis to Frog’s Call Research

Individual’s phenotypic traits are the results of adaptation to ecological conditions.Therefore,different selection pressures caused by heterogeneous environments may result in phenotypic difference,especially for individuals in different geographical populations.Here,we illustrated for the first time to use social network analysis(SNA)for examining whether geographical proximity predicts the similarity patterns in call characteristics among populations of an anuran species.We recorded calls from 150 male dorsal-striped opposite-fingered treefrogs(Chiromantis doriae)at 11 populations in Hainan Province and one population in Guangdong Province in China's Mainland,and we measured eight acoustic variables for each male.Mantel test didn’t show a correlation between geographical proximity and the similarity in call characteristics among populations.In addition,we failed to find correlations between a population’s eigenvector centrality and the distance to its nearest neighbor,nor between the coefficient of variation of similarity in call characteristics of a population and the average distance to all other populations.Nevertheless,three acoustic clusters were identified by the Girvan-Newman algorithm,and clustering was partially associated with geography.Furthermore,the most central populations were included in the same cluster,but the top betweenness populations were located within different clusters,suggesting that centrality populations are not necessary bridging between clusters.These results demonstrate the potential usefulness of the SNA toolbox and indicate that SNA helps to uncover the patterns that often overlooked in other analytical methods.By using SNA in frog’s call studies,researchers could further uncover the potential relationship in call characteristics between geographical populations,further reveal the effects of ecological factors on call characteristics,and probably enhance our understanding of the adaptive evolution of acoustic signals.