The use of bioacoustics in anuran taxonomy: Theory, terminology, methods and recommendations for best practice

Abstract

Vocalizations of anuran amphibians have received much attention in studies of behavioral ecology and physiology, but also provide informative characters for identifying and delimiting species. We here review the terminology and variation of frog calls from a perspective of integrative taxonomy, and provide hands-on protocols for recording, analyzing, comparing, interpreting and describing these sounds. Our focus is on advertisement calls, which serve as premating isolation mechanisms and, therefore, convey important taxonomic information. We provide recommendations for terminology of frog vocalizations, with call, note and pulse being the fundamental subunits to be used in descriptions and comparisons. However, due to the complexity and diversity of these signals, an unequivocal application of the terms call and note can be challenging. We therefore provide two coherent concepts that either follow a note-centered approach (defining uninterrupted units of sound as notes, and their entirety as call) or a call-centered approach (defining uninterrupted units as call whenever they are separated by long silent intervals) in terminology. Based on surveys of literature, we show that numerous call traits can be highly variable within and between individuals of one species. Despite idiosyncrasies of species and higher taxa, the duration of calls or notes, pulse rate within notes, and number of pulses per note appear to be more static within individuals and somewhat less affected by temperature. Therefore, these variables might often be preferable as taxonomic characters over call rate or note rate, which are heavily influenced by various factors. Dominant frequency is also comparatively static and only weakly affected by temperature, but depends strongly on body size. As with other taxonomic characters, strong call divergence is typically indicative of species-level differences, whereas call similarities of two populations are no evidence for them being conspecific. Taxonomic conclusions can especially be drawn when the general advertisement call structure of two candidate species is radically different and qualitative call differences are thus observed. On the other hand, quantitative differences in call traits might substantially vary within and among conspecific populations, and require careful evaluation and analysis. We provide guidelines for the taxonomic interpretation of advertisement call differences in sympatric and allopatric situations, and emphasize the need for an integrative use of multiple datasets (bio-acoustics, morphology, genetics), particularly for allopatric scenarios. We show that small-sized frogs often emit calls with frequency components in the ultrasound spectrum, although it is unlikely that these high frequencies are of biological relevance for the majority of them, and we illustrate that detection of upper harmonics depends also on recording distance because higher frequencies are attenuated more strongly. Bioacoustics remains a prime approach in integrative taxonomy of anurans if uncertainty due to possible intraspecific variation and technical artifacts is adequately considered and acknowledged.