Comparing context-dependent call sequences employing machine learning methods: an indication of syntactic structure of greater horseshoe bats.

For analysis of vocal syntax, accurate classification of call sequence structures in different behavioural contexts is essential. However, an effective, intelligent program for classifying call sequences from numerous recorded sound files is still lacking. Here, we employed three machine learning algorithms (logistic regression, support vector machine and decision trees) to classify call sequences of social vocalizations of greater horseshoe bats (Rhinolophus ferrumequinum) in aggressive and distress contexts. The three machine learning algorithms obtained highly accurate classification rates (logistic regression 98%, support vector machine 97% and decision trees 96%). The algorithms also extracted three of the most important features for the classification: the transition between two adjacent syllables, the probability of occurrences of syllables in each position of a sequence, and the characteristics of a sequence. The results of statistical analysis also supported the classification of the algorithms. The study provides the first efficient method for data mining of call sequences and the possibility of linguistic parameters in animal communication. It suggests the presence of song-like syntax in the social vocalizations emitted within a non-breeding context in a bat species.

Social vocalizations of big-footed myotis (Myotis macrodactylus) during foraging.

Acoustic signals play a crucial role in transmitting information and maintaining social stability in gregarious animals, especially in echolocating bats, which rely primarily on biological sonar for navigating in the dark. In the context of foraging without relying on tactile, visual or olfactory cues, acoustic signals convey information not only on food but also on ownership and defense of resources. However, studies on such information remain fragmentary. In the present study, we aim to document the social vocal repertoire of Myotis macrodactylus at natural foraging sites. Multiple acoustic analyses and spectrographic classification revealed a rich foraging vocal repertoire comprising 6 simple syllables and 2 composites. Discriminant function analyses associated with a subset-validation procedure provided an optimal method to spectrographically classify all recorded sounds into different syllable types. Multidimensional scaling of median values of multiple parameters further confirmed notable differences among these syllables in a 3-D space. In addition, Euclidean distance analysis showed that there were some spectral similarities between specific social vocal syllables and feeding buzzes, which implied a potential jamming role. Altogether, the data indicate that bats at foraging sites under natural conditions used variant social vocalizations with different functions in addition to echolocation calls, providing supporting evidence for further work on the function and vocal mechanisms of acoustic communication in mammals.