Donald E. Kroodsma

Reproductive development in a female songbird: differential stimulation by quality of male song.

Female canaries exposed to playback of large repertoires of male songs built nests faster and laid more eggs than did those females exposed to smaller song repertoires: females are attentive to attributes of male song, and their choices have played a role in the evolution of oscine singing behaviors.

Repertoires and Geographical Variation in Bird Song

Publisher Summary This chapter deals with repertoires and geographical variation in bird song. Three methods have been used to investigate the function of a behavior pattern: (1) interspecific comparisons in which differences between species are related to ecological and life history variables; (2) intraspecific comparisons relating variations between individuals to variations in survival, reproductive success, or success in acquiring resources; and (3) direct experimental tests of the effect of a behavior. All three methods have been applied to the study of how selection favors repertoires. The chapter emphasizes micro-geographical variation, and examines the two factors that most significantly influence local spatial variation in bird song: (1) the extent and accuracy of imitation, and (2) the site of song imitation with respect to the site of breeding. It discusses the relationship between genetic separation of populations and song variations, and the influence of song repertoire size on the ability to establish or maintain sharp discontinuities in song traditions between neighboring populations. The usefulness and definition of the term “dialect” and the functional significance of vocal imitation and its influence on song dialects are also discussed.

Design of Playback Experiments: The Thornbridge Hall NATO ARW Consensus

Playback is an experimental technique commonly used to investigate the significance of signals in animal communication systems. It involves replaying recordings of naturally occurring or synthesised signals to animals and noting their response. Playback has made a major contribution to our understanding of animal communication, but like any other technique, it has its limitations and constraints.

A re-evaluation of song development in the song sparrow

Song sparrows were hand-reared from the egg in auditory isolation from adult conspecifics and maintained either together with other species or individually in sound chambers. Songs developed by the males contained several normal species-specific characteristics, but on the whole differed both qualitatively and quantitatively and were therefore readily distinguished from wild-type songs. The data are largely consistent with those of Mulligan (1966), but the normal aspects of song development have been over-emphasized in recent literature.

Organizing and archiving private collections of tape recordings.

Many readers of this journal have accumulated large numbers of recordings of wild animals that now sit unused on their shelves. We would like to argue that these recordings have continued value and we outline a way by which they can be shared with the greater research, education and outreach communities. The arguments for sharing these recordings are very straightforward. Everyone who has attempted to record the auditory signals of animals in the wild is aware of how difficult this task can be. Despite the difficulty, the many articles in this and related journals on animal acoustical signals attest to the success that can be achieved with careful technique and proper equipment. The cost of quality recording is that any one of us is likely to record only a limited number of species. Clearly, there are many broad questions about signal evolution, function, phylogeny and ontogeny that can be solved only if researchers have access to a broader range of recordings than they can themselves collect. The obvious answer is shared archiving of our individual recording collections with one of the natural sound collection centres. These institutions have as their function the preservation of representative acoustical signals from as many extant taxa as possible and the sharing of these archived materials with researchers, wildlife management staff, educational programmes and the media. Many researchers hesitate to provide copies of their recordings to one of these centres for one or both of two reasons: (1) they do not think their recordings are needed or valuable; or (2) they think it will be too complicated or time consuming to make such a contribution. In this note, we hope to persuade all of our colleagues that both suppositions are false. Are my recordings superfluous? The answer is a strong no. Given the abundance of articles on bioacoustical topics and the many television programmes featuring wild animal signals, one might think that existing archives had no need of more material. This is totally false. Not only are many taxa not yet represented in any collection, but we have only begun to provide representative recordings of each component in the repertoires of even commonly recorded species. Data on geographical variation and ontogenetic stages are very poorly represented for the majority of taxa. There is thus a very high probability that your recordings will fill critical gaps in current archives. There are other reasons why archiving your recordings centrally is an important thing to do. With the rapid loss of species world-wide, recordings from each species should be archived in sites where there is a guarantee that they will be preserved for the long term. Once extinct, there will be no chance to record rare species. Second, there are ethical reasons for central archiving. For example, where studies of ontogeny require hand-reared animals, central archiving is more likely to alert researchers that recordings needed for a study are already available, and thus reduce the numbers of animals that need be manipulated. Finally, sound recordings are the basic substrate on which many publications are based, and many journals now encourage authors to make copies available on Web sites. While we applaud making these raw data available for inspection by other researchers, storage is then decentralized across the many journal Web sites. It is very difficult to compare sounds even if from the same taxon and studied by the same researcher when articles are published in different journals and sample sounds are stored on multiple Web sites. It seems much more useful to deposit all voucher samples and raw data in one of the centralized archives Correspondence: Jack Bradbury or Gregory Budney, Library of Natural Sounds, Laboratory of Ornithology, Cornell University, 159 Sapsucker Woods Road, Ithaca, NY 14850, U.S.A. (email: jwb25@cornell.edu). D. W. Stemple is at the Department of Computer Science, University of Massachusetts, Amherst, MA 01003, U.S.A. D. E. Kroodsma is at the Department of Biology, University of Massachusetts, Amherst, MA 01003, U.S.A.

Birdsong ‘performance’ studies: a sad commentary

Birdsong performance studies have embraced the significance of a scatterplot between trill rate and frequency bandwidth, using the distance that a given sound plots from an upper bound (i.e. ‘vocal deviation’) as a measure of how difficult it is to perform that sound; the relative difficulty of the performance then becomes a means by which to assess the quality of the song and the singer (review in Kroodsma, 2017). Those who promote this literature defend it by its wide acceptance: ‘ ... vocal deviation has been used widely as a composite index of vocal performance ... ’ (Podos et al., 2016, page 204) and ‘Vocal deviation... has indeed been adopted widely in tests of song function’ (Goodwin & Podos, 2015, page 1). ‘Studies of trilled vocalizations’, with their inherent ‘tradeoff between syllable repetition rate and frequency bandwidth’, are hailed as ‘a premiere illustration of how performance constraints shape the evolution of mating displays... [with] sexual selection favoring high performance trills’ (Wilson, Bitton, Podos, & Mennill, 2014, page 214). In reality, however, no available scientific data support these ideas. The considerable literature that has developed on this topic is instead a premiere illustration of how highly flawed methods have been used repeatedly to tell an intuitively appealing but unsupported story, all of which has until now escaped a critical evaluation. The three responses to my Forum (Cardoso, 2017; Podos, 2017; Vehrencamp, de Kort, & Illes, 2017) do not refute the big-picture and methodological problems that I raise. Podos and Vehrencamp et al. distract from the big issues largely by addressing minor, secondary issues or nonissues altogether. Cardoso does not really address the main problems I discuss, but instead offers advice on how to disentangle ideas and facts in performance-related research. Below, I offer a few final thoughts on this topic.