关键词: 鸣声, 文化进化, 自然选择, 社会学习, 现代综合进化论

Abstract

Background & Aims: From Darwin’s concept of natural selection leading to biological evolution in 1859, which underpins biological evolution, to the development and formation of the theory of cultural evolution in the 1980s, research on culture evolution phenomena in non-human animals has gradually emerged. Among the diverse behaviors of birds, their highly varied vocalizations represent the most prominent and extensively studied examples in cultural evolution research. This paper summarizes and synthesizes existing research on the cultural evolution of avian vocalizations. It also provides an outlook on the latest directions and methods in this field, aiming to offer a comprehensive analysis of current progress. 

Progress: We frame the discussion within the Modern Evolutionary Synthesis. The review covers four primary dimensions: (1) the genesis of cultural variation in avian vocalizations; (2) natural and sexual selection processes acting on this variation; (3) the heritability of vocal culture; and (4) the adaptive significance of vocal culture and its influence on individual fitness. 

Perspectives: Several long-neglected issues persist in contemporary research on birdsong culture, including female song, non-reproductive functions of song, and seasonal variation in song production and use. Future studies on the cultural evolution of avian vocalization should integrate field and laboratory experiments while establishing large-scale, long-term datasets of avian vocalizations. In recent years, the incorporation of artificial intelligence and machine learning techniques has enabled rapid and accurate automated detection, localization, and classification/annotation of cultural variants in large birdsong datasets. Additionally, computer simulation approaches such as agent-based modeling (ABM) allow researchers to simulate and predict patterns of cultural transmission and diffusion. These advanced methods have substantially improved both the efficiency and depth of research on birdsong cultural evolution, enabling analysis of macro-scale datasets that were previously intractable. Nevertheless, the intrinsic variability of birdsong culture continues to pose challenges for AI-driven systems, as reflected in current limitations of species identification algorithms and the need for further methodological development in passive acoustic monitoring. Beyond summarizing the evolutionary mechanisms of birdsong culture, this review therefore seeks to offer guidance on how insights from cultural evolution can enhance the robustness, self-correction capability, and cross-context generalization of machine learning classifiers, thereby providing new theoretical and practical support for avian biodiversity monitoring and conservation.

Key words: vocalization, cultural evolution, natural selection, social learning, Modern Evolutionary Synthesis