Computational analyses of dynamic visual displays reveal diet-dependent male signalling in Rabidosa rabida wolf spiders

Abstract

Male courtship signals vary widely due to genetic, environmental and ecological influences. This study investigates how diet manipulation affects the dynamic visual components of multimodal courtship displays in male Rabidosa rabida wolf spiders. Using a Gaussian Hidden Markov Model (GHMM), we implemented an automated bottom-up approach to identify the hierarchical structure of foreleg movements during courtship. We then integrated unary and binary similarity measures to quantitatively compare these movements. Our results show that the GHMM-derived structural organization of foreleg movements closely aligns with classifications made by human observers. Furthermore, combining unary and binary similarity measures provides a robust framework for quantifying dynamic movements in animal communication. Notably, male foraging history influences both foreleg morphology and movement during courtship displays. This study enhances our understanding of the evolution of complex male courtship signals and offers a standardized framework for future research.