Signal partitioning between native and introduced forest birds of Hawai‘i Island

Abstract

Birds use sound for conspecific communication. According to the acoustic niche hypothesis, they avoid signal competition with other species by partitioning sound in frequency and time. Others have observed introduced species changing the vocal behavior of native species; however, community-level effects of multiple introduced species remain unknown. Hawai‘i, with its legacy of bird extinctions followed by a surge of human introductions, offers a unique opportunity to study acoustic signal partitioning between native and introduced species. We predicted that communities with higher percentages of introduced birds would exhibit more acoustic overlap between members of different species due to shorter time frames to evolve signal partitioning. Using autonomous recording units, we recorded forest bird communities during the summer in montane primary and secondary forests as well as a low-elevation agricultural site. Random samples of recordings from each site were visualized as spectrograms, and all bird vocalizations ≥5 decibels above background noise were identified and labeled by species. Frequency range and the proportion of overlap with other species were compared between native and introduced species. We also used a null model which randomized the start time of each bird vocalization within a location over 500 iterations, then compared the amount of heterospecific signal overlap in the randomizations to that observed in the recordings. While native and introduced species generally used similar frequency ranges, native-native heterospecific species vocalization pairs had a significantly higher proportion of overlap than introduced-native and native-native pairs. Additionally, the incidence of signal overlap in the original recordings tended to be lower than in the null model randomizations, but this difference was not significant, and was not influenced by the percentage of introduced species vocalizations at the site. The lack of significant difference between observed and null model signal overlap occurrence suggests that native and introduced forest birds were not partitioning acoustic space either spectrally or temporally, and that introduced birds are not strongly influencing signal partitioning in forest bird communities of Hawai‘i. However, this may have been a consequence of recording during the non-breeding season, and future work should compare vocal activity in native and introduced birds throughout the year.