Linked paper: Geographic variation in call structure, likelihood, and call-song associations across subspecies boundaries, migratory patterns, and habitat types in the Marsh Wren (Cistothorus palustris) by S.A.M. Luttrell and B. Lohr, The Auk: Ornithological Advances 135:1, January 2018.
Many bird species have unique geographic signatures in their vocalizations similar to human “accents.” Most of what we know about geographic variation in bird sounds comes from studies of bird song. Song has been a rich subject for studying geographic variation because it is typically learned, allowing song to change more quickly across space and time than a purely genetic trait. Song, however, is only one type of signal in a bird’s vocal repertoire. We wanted to build a broader picture of how vocal behavior evolves and changes among populations by looking at a large repertoire of sounds at once. Most birds have multiple call types in addition to their songs. Each call or song type is an individual trait used under unique circumstances, and that means that each one may be under different selective pressures. As a result, looking at multiple vocalizations may reveal multiple patterns of geographic variation, or, if their geographic patterns are similar, it may suggest a general process of vocal evolution. There are many ways in which vocalizations might change over time or distance. For example, changes could be random—as long as the signal still sends the correct message, some aspects of its acoustic structure could drift among populations. Additionally, vocal signals may be under selection to reduce distortion caused by the habitat in which they are produced and heard. Think about how sounds are distorted differently in an open, bare hallway versus a musician’s sound booth and how the local acoustics might alter a listener’s ability to understand you. Furthermore, if some aspect of the sound is learned, then copy errors or innovations during learning can result in passing down cultural changes over time. These are just a few ways in which sounds might be altered, and no two vocalizations are necessarily influenced in the same way by the same set of selective pressures.
In order to compare vocal repertoires among populations, we looked at several subspecies of the Marsh Wren. Marsh Wrens provide a natural experiment in vocal variation due to ecological and natural history differences among the subspecies. The five eastern North America subspecies we focused on are found in two distinct habitats (freshwater marshes and saltmarshes), and they exhibit three migratory patterns (resident, partially migratory, and fully migratory). Our first challenge was to describe and classify the call repertoire for Marsh Wrens. We identified seven discrete call types. Three of the seven calls varied in acoustic properties that were consistent with differences in either migratory pattern or habitat type. Surprisingly, we also found that four calls were more common in some subspecies than others and that the differences were greatest between habitat types. This variation in call production may indicate differences in behavior or timing of breeding among the subspecies with different ecologies. Our results suggest that while not all vocal signals are changing at the same rate or in the same way, differences in habitat type and migratory behavior may be related to the biggest differences in vocal behavior. Interestingly, the calls that showed the greatest differences were calls used in mate attraction and territory defense, while calls related to alarm or distress were similar across subspecies, natural history, and habitat type. This result suggests that sexual selection could be driving or reinforcing changes between populations with different ecologies.
In the future, we are excited to explore another unusual phenomenon that we report in this paper: the use of calls as embedded elements in song. Like most songbirds, Marsh Wren males sing during the breeding season to attract mates and defend territories. Unlike those of most songbirds, we found that, depending on the subspecies, 73-93% of male Marsh Wren songs contained embedded calls. Do embedded calls confer some additional message to the song? Does this behavior vary across the breeding season? Are there specific structural rules regarding the embedded call pattern within song? Stay tuned as we untangle the structural complexity and geographic variation in songs with embedded call elements in Marsh Wrens across the rest of their range. Does your study species use calls in a song-related context? If so, contact us at firstname.lastname@example.org—we would be excited hear about it!