Machine learning and statistical classification of birdsong link vocal acoustic features with phylogeny

Meta-analysis of the acoustic adaptation hypothesis reveals no support for the effect of vegetation structure on acoustic signalling across terrestrial vertebrates

Acoustic communication plays a prominent role in various ecological and evolutionary processes involving social interactions. The properties of acoustic signals are thought to be influenced not only by the interaction between signaller and receiver but also by the acoustic characteristics of the environment through which the signal is transmitted. This conjecture forms the core of the so-called "acoustic adaptation hypothesis" (AAH), which posits that vegetation structure affects frequency and temporal parameters of acoustic signals emitted by a signaller as a function of their acoustic degradation properties. Specifically, animals in densely vegetated "closed habitats" are expected to produce longer acoustic signals with lower repetition rates and lower frequencies (minimum, mean, maximum, and peak) compared to those inhabiting less-vegetated "open habitats". To date, this hypothesis has received mixed results, with the level of support depending on the taxonomic group and the methodology used. We conducted a systematic literature search of empirical studies testing for an effect of vegetation structure on acoustic signalling and assessed the generality of the AAH using a meta-analytic approach based on 371 effect sizes from 75 studies and 57 taxa encompassing birds, mammals and amphibians. Overall, our results do not provide consistent support for the AAH, neither in within-species comparisons (suggesting no overall phenotypically plastic response of acoustic signalling to vegetation structure) nor in among-species comparisons (suggesting no overall evolutionary response). However, when considering birds only, we found weak support for the AAH in within-species comparisons, which was mainly driven by studies that measured frequency bandwidth, suggesting that this variable may exhibit a phenotypically plastic response to vegetation structure. For among-species comparisons in birds, we also found support for the AAH, but this effect was not significant after excluding comparative studies that did not account for phylogenetic non-independence. Collectively, our synthesis does not support a universal role of vegetation structure in the evolution of acoustic communication. We highlight the need for more empirical work on currently under-studied taxa such as amphibians, mammals, and insects. Furthermore, we propose a framework for future research on the AAH. We specifically advocate for a more detailed and quantitative characterisation of habitats to identify frequencies with the highest detection probability and to determine if frequencies with greater detection distances are preferentially used. Finally, we stress that empirical tests of the AAH should focus on signals that are selected for increased transmission distance.

The Temporal Organization of Learned Vocal Behavior Is Predicted by Species Rather Than Experience

Birdsong is hierarchically organized in time, like speech and other communication behaviors. Syllables are produced in sequences to form song motifs and bouts. While syllables are copied from tutors, the factors that determine song temporal organization, including syllable sequencing (syntax), are unknown. Here, we tested the roles of learning and species genetics in song organization. We manipulated juvenile song experience and genetics in three species of estrildid finches (zebra finches, Taeniopygia guttata castanotis; long-tailed finches, Poephila acuticauda; Bengalese finches, Lonchura striata var. domestica). We analyzed the adult songs of male birds that were: (1) tutored by conspecifics; (2) untutored; (3) tutored by heterospecifics; and (4) genetic hybrids. Song macrostructure, syllable sequencing, and syllable timing were quantified and compared within and across species. Results showed that song organization was consistent within a species and differed across species, regardless of experience. Temporal features did not differ between tutored and untutored birds of the same species. The songs of birds tutored by other species were composed of heterospecific syllables produced in sequences typical of conspecific songs. The songs of genetic hybrids showed the organization of both parental species, despite the fact that only males sing. Results indicate that song organization is predicted by species rather than experience.

Translocated wild birds are predisposed to learn songs of their ancestral population

Population differences in socially learned mating signals like oscine birdsong are particularly vulnerable to breakdown through dispersal.1 Despite this challenge, geographic variation in learned signals is ubiquitous.2 A proposed explanation for this pattern is that birds express predispositions to selectively learn and produce population-typical songs.3,4,5 While experimental studies on lab-reared birds have shown the existence of within-species learning predispositions,6,7,8,9,10 it remains unclear whether and how learning predispositions influence song acquisition in the wild. Here, we investigated innate song learning predispositions in wild pied flycatchers (Ficedula hypoleuca) by measuring the songs of individuals translocated as eggs from a Dutch population to a breeding population in Sweden. We compared the songs of the adult males hatched from these translocated eggs with those from the ancestral and receiving populations. Songs of translocated males closely resemble the local Swedish songs to which they were exposed during development, supporting the importance of social learning. However, translocated males selectively learned those local Swedish song elements that sound the most "Dutch-like." As a result, their songs are significantly shifted toward those of the ancestral Dutch population. This suggests that innate learning predispositions track ongoing song evolution in wild populations of pied flycatchers. We propose that as songs continue to diverge over time, this coevolutionary relationship between song and learning predispositions may contribute to the emergence of incipient pre-mating barriers.

Language-like efficiency and structure in house finch song

Communication needs to be complex enough to be functional while minimizing learning and production costs. Recent work suggests that the vocalizations and gestures of some songbirds, cetaceans and great apes may conform to linguistic laws that reflect this trade-off between efficiency and complexity. In studies of non-human communication, though, clustering signals into types cannot be done a priori, and decisions about the appropriate grain of analysis may affect statistical signals in the data. The aim of this study was to assess the evidence for language-like efficiency and structure in house finch (Haemorhous mexicanus) song across three levels of granularity in syllable clustering. The results show strong evidence for Zipf's rank-frequency law, Zipf's law of abbreviation and Menzerath's law. Additional analyses show that house finch songs have small-world structure, thought to reflect systematic structure in syntax, and the mutual information decay of sequences is consistent with a combination of Markovian and hierarchical processes. These statistical patterns are robust across three levels of granularity in syllable clustering, pointing to a limited form of scale invariance. In sum, it appears that house finch song has been shaped by pressure for efficiency, possibly to offset the costs of female preferences for complexity.