Asymmetry in responsiveness to playback of geographic song variation after a recent range expansion

Birdsong variation among populations within species has been regarded as a possible precursor for variation between species and may play a role in speciation. However, acoustic variation between two populations of the same or different species can have a variable impact on mutual responsiveness, and we currently lack sufficient insight into the underlying reasons. We report here on geographic song variation and responsiveness to playback for the light-vented bulbul (Pycnonotus sinensis), a species with recent range expansion in China. We recorded songs in four historical populations in the south and six newly established populations in the north. We tested responsiveness in two southern and two northern locations to songs from southern and northern birds. Besides songs from the other distribution range (south to north and north to south), we used recordings for playback from local, nearby, and faraway dialects from within their own distribution range. We confirmed distinct dialectal differentiation in the south and a more gradual pattern of geographic song divergence in the north. Birds in the south showed very little response to almost all non-local songs. In contrast, birds in the north showed a strong response to a wide range of stimuli, just gradually fading from local, nearby, to far-away northern dialects, which may be related to song type sharing or individual mobility and aggression associated with the recent range expansion. We hereby add an asymmetric response pattern in response to the literature and review the current insights about the potential impact of song variation on avian speciation.

Methods
We recorded songs in six northern populations: Beijing (BJ), Dalian (DL), Beidaihe (BDH), Xian (XA), Zhengzhou (ZZ), and Jinan (JN), and four southern populations: Nanchong (NC), Changsha (CS), Yaan (YA), and Wuhan (WH). We recorded the spontaneous songs (not playback-triggered) of 10 males from each population. All these songs were recorded during the early phase of the breeding season and were subsequently used to create playback stimuli. We conducted playback experiments from 9 April to 30 May during the early breeding seasons in both southern and northern populations.

We selected NC and CS for the southern populations and BJ and DL for the northern populations to test local responsiveness to geographic variation in song via playback experiments. We compared response strength to songs of the local population, a nearby and far-away population of the same region (historical source or recently expanded range), and a far-away population from the other region. The song of a particular male was used twice per playback population, which means that we used four sets (own, nearby, far-away, other) of ten unique song stimuli from different males per population of origin (thus forty unique songs) to test the responsiveness of twenty unique males in twenty different territories per site (thus eighty playback trials per site). In this way, we reduced problems of pseudo-replication by restricting the repeated use of the same stimuli in the same population more than twice.

We started each series of playback with a heterospecific great tit (Parus major) song to get an experimental control of baseline behavior during playback set-up, sound playback, and observer presence. After the great tit song control, we played the four bulbul song treatments of different origins in random order. We therefore had a playback sequence of five different song stimuli, with each song played five times for 11 minutes, resulting in a total trial duration of 55 minutes per bird territory. Every playback trial consisted of three periods per song stimulus (Fig. 3): pre-playback (5 min, the time before or in between subsequent playbacks), during-playback (3 min), post-playback (3 min). The heterospecific control never elicited any obvious response, and the quantification of the relatively low level of excitement during this playback period thereby adequately served as baseline behaviour for the evaluation of any responsiveness during the following bulbul song stimuli.