Identification Of Bat Species By Echolocation Calls

Bat identification can be mainly devided intotwo approaches,that is,traditional morphological identification and molecular identification.The morphological identification based on body and skull parametershas been the dominant method in past years,yet the accuracy largely depends onresearchers' experience and ability.In addition,the amazing species diversity in bats,together withnocturnal habit and powered flight,yieldingthe practical difficulties forfieldcapture and sampling.It has been shown that molecular sampling fromliving bats hascertainnegative influence on the survival of bats.Overall,both morphological and molecular identifications have various limitations.As a consequence,researcherssought to explore other methods to identify species of bats.In particular,echolocation calls ina variety of bats exhibit remarkably species-specific in spectro-temporal parameters,which have been successfully used for identifying bats distributed in Europe and north America.In this study,we focus on 10 typical echolocating species,which produce constant-frequency and frequency-modulated signals,in Yunnan province,China.By recordingtheecholocation sounds,we establish a wide call database and then test the validilityof acoustic identification approach.Specifically,study objects encompass 6 species fromRhinolophidae,2 species in Vespertilionidae,2 species of Hipposideridae.Our measure of mahalanobis distance,together withdiscriminant analysis,indicates considerable interspecific variation in call parameters,and such species-level variation is remarkably greater than sexual and context variation within species.Based on the templates of echolocation pulses,cross correlation analyses of template further document higher accuracy of acoustic identification approach,with a classification accuracy of 99.50%—100%.In all,this is the first paper on acoustic identification of bats using template cross correlation.It revealsthe basic characteristics of echolocation callsinsome bats in Yunnan province,laying the foundation for future research regardingevolution of echolocation calls.Our results provide important implications forbats' field investigation,classification,and conservation.