Learning from the local: the variety and spatial pattern of vocal mimicry in songs of the invasive white-rumped shama in Taiwan.

Studying the model selection, especially in multiple heterospecific mimicry, is crucial for understanding the function of vocal mimicry. Invasive songbirds with large repertoires in novel auditory environments lacking conspecifics expand their repertoires by imitating heterospecifics, offering valuable insights into model selection. This study examines vocal mimicry in the invasive (white-rumped shama), focusing on how this species selects mimicry models in Taiwan. We recorded the songs of 256 males across 26 sites in Taiwan, and their vocal mimicry of heterospecific sources was identified. Our results revealed that at least 28 animal species were mimicked, and 68% of those model species were endemic. Regarding individual mimics, 68.6% of 242 mimics imitated more than two species and 13.2% of total mimics imitated up to 4-8 species. Most mimicry types (defined by species mimicked) exhibited a significant clumped distribution, except three mimicry types. As the number of observed at a site increased, the number of identified mimicry types increased significantly. Furthermore, as the total number of mimics of the two sites increased, the compositions of mimicry types of the two sites were more likely dissimilar. We suggest that individual differences play a crucial role in the model selection of heterospecific mimicry, and these differences may result from variations in individual learning ability or preferences, or from variations in the local auditory environment where the individuals inhabit.