3D tracking reveals energy-distance trade-offs in two dominant grazers on a degraded coral reef

Abstract In-depth understanding of animal movement ecology, including energy expenditure and internal energy budgeting, is crucial for deciphering the intricate dynamics of ecosystem functioning. It essentially reveals resource partitioning and energy flows among foraging organisms within their habitats. Ecosystems under severe anthropogenic stress, such as degraded coral reefs, serve as valuable model habitats for examining how patchy resource availability impacts the foraging behavior and internal energy budgets of herbivores. In this study, we employed stereo-video measurements, Artificial Intelligence (AI)-driven object recognition, and 3D tracking techniques to explore resource partitioning and energy budgets of two dominant grazers, Brown surgeonfish Acanthurus nigrofuscus and Yellowtail tang Zebrasoma xanthurum , on a degraded coral reef in Eilat, Israel. We compared feeding preferences, bite rates, and inter bite distances to comprehend the mechanisms underlying functional trait expression and resource partitioning in these key grazers. A. nigrofuscus demonstrated a strategy that allowed a higher rate of food intake within given time frames, while Z. xanthurum exhibited a more generalist approach, traversing larger distances between food patches. However, our measurements of energy expenditure did not reveal significant differences between the two species. We found that the unique foraging strategies and feeding preferences of A. nigrofuscus and Z. xanthurum may underlie the observed energy-distance trade-offs, which were determined by factors such as resource availability and feeding niches. By applying AI-generated 3D trajectories, we achieved a granular analysis of fish movement and foraging behavior. This approach demonstrates the innovative potential of blending AI-generated 3D data with traditional stereo-video measurements, thus advancing our understanding of animal movement ecology. A detailed understanding can inform and enhance management and conservation strategies, providing insights into the adaptation of grazers to resource availability within degraded ecosystems. The approach of deriving energy expenditure from automatically generated 3D trajectories of animal movements could prove to be a novel and valuable indicator of ecosystem health.