How the environment shapes animal signals: a test of the acoustic adaptation hypothesis in frogs

Abstract

Long-distance acoustic signals are widely used in animal communication systems and, in many cases, are essential for reproduction. The acoustic adaptation hypothesis (AAH) implies that acoustic signals should be selected for further transmission and better content integrity under the acoustic constraints of the habitat in which they are produced. In this study, we test predictions derived from the AAH in frogs. Specifically, we focus on the difference between torrent frogs and frogs calling in less noisy habitats. Torrents produce sounds that can mask frog vocalizations and constitute a major acoustic constraint on call evolution. We combine data collected in the field, material from scientific collections and the literature for a total of 79 primarily Asian species, of the families Ranidae, Rhacophoridae, Dicroglossidae and Microhylidae. Using phylogenetic comparative methods and including morphological and environmental potential confounding factors, we investigate putatively adaptive call features in torrent frogs. We use broad habitat categories as well as fine-scale habitat measurements and test their correlation with six call characteristics. We find mixed support for the AAH. Spectral features of torrent frog calls are different from those of frogs calling in other habitats and are related to ambient noise levels, as predicted by the AAH. However, temporal call features do not seem to be shaped by the frogs' calling habitats. Our results underline both the complexity of call evolution and the need to consider multiple factors when investigating this issue.

Neotropical frogs and mating songs: The evolution of advertisement calls in glassfrogs

“…On the other hand, call duration is not phylogenetically constrained. Previous studies have found individual, intraspecific and interspecific variation of temporal traits such as call duration, call rate and pulse rate to be correlated with environmental temperature (Bellis, ; Blair, ; Gillooly & Ophir, ; Goutte et al., ; Köhler et al., ; Zweifel, ). Unfortunately, analyses of correlations between acoustic variables and environmental temperature were precluded by (a) the lack of associated temperature data of most recordings and (b) the different sample sizes of recordings available for each species, some of which were represented by recordings spanning several dates and geographic locations, rendering an average temperature estimation uninformative.…”

“…Thus, the high‐pitched calls of some relatively large species could be an adaptation to outcompete the masking interference of torrent noise. As background noise data were missing for all calls, future studies could benefit from collecting data on the intensity and bandwidth of waterfall and rapids noise (Goutte et al., , ; Narins et al., ; Schwartz & Bee, ), as well as information about height of calling perches (Kubicki, ; Señaris & Ayarzaguena, ) and standardized measurements of call amplitude along the background noise gradient (Schwartz & Bee, ).…”

“…is usually negatively correlated with the body size of the emitter in anurans (Gingras, Boeckle, Herbst, & Fitch, 2013;Goutte et al, 2016Goutte et al, , 2018Zweifel, 1968), birds (Gonzalez-Voyer, den Tex, Castelló, & Leonard, 2013;Podos, 2001;Ryan & Brenowitz, 1985;Seddon, 2005;Wallschläger, 1980) and mammals (Barclay & Brigham, 1991;Fitch, 1997;Pfefferle & Fischer, 2006).…”

“…The few studies that have investigated the evolution of anuran advertisement calls in species-rich clades at deep evolutionary timescales have revealed complex associations between temporal call traits and shared ancestry among species (Erdtmann & Amézquita, 2009), spectral call traits and body size (Erdtmann & Amézquita, 2009;Goutte et al, 2016Goutte et al, , 2018 and acoustic specificities of noisy torrent habitats (Goutte et al, 2016(Goutte et al, , 2018. In at least one Neotropical clade (Superfamily Dendrobatoidea), the evolution of call traits is also apparently related to the occurrence of aposematic dorsal colour patterns, with calls of brightly coloured species diversifying at faster rates than calls of cryptically coloured species (Santos et al, 2014).…”

Reduced geographic variation in roars in different habitats rejects the acoustic adaptation hypothesis in the black‐and‐gold howler monkey (Alouatta caraya)

“…Sampled populations are separated with white dashed lines. Equally color populations share genetic ancestry (modified from Oklander et al 2017) [Colour figure can be viewed at wileyonlinelibrary.com] 2010; rails: Rek & Kwiatkowska, 2016 and torrent frogs: Goutte et al, 2018). Evidence gathered over the years provided partial support to this theoretical framework (reviewed in Ey & Fisher, 2009), in part because other factors beyond vegetation type and/or environmental sound come into play influencing vocalization structure.…”