Is there pattern in the butchering methods of Harpy Eagles ([i]Harpia harpyja[/i]) on their prey ?

Abstract

Understanding how prey are processed before being brought to the nest is essential for linking bone assemblages to the actual behaviors of large predators. This study aimed to identify butchering patterns in the Harpy Eagle (Harpia harpyja) by (i) quantifying how prey mass influences the probability of butchering, (ii) determining the role of adult sex in this probability, (iii) distinguishing differences among the main prey species, (iv) testing the functional contrast between primates and sloths, and (v) comparing these patterns with expected taphonomic profiles. Fourteen nests were monitored in Amazonia (Mato Grosso, Brazil) using camera traps, and 235 prey deliveries were identified and classified into seven categories of butchering according to the body parts delivered. Analyses were based on Bayesian hierarchical models, complemented by principal component analysis (PCA) and hierarchical clustering. Our results indicate that mass is the major determinant. The probability of a butchered prey delivery increases strongly for small prey and then saturates at > 0.95 for the heaviest prey. Analysis of sex differences shows that males butcher more at low masses, after which probabilities converge at intermediate and high masses. Prey species clearly structures the body part delivered, and mass increases the probability of the dominant body part. Arboreal porcupines (Coendou) lean clearly towards tail and lower body, whereas Tamanduas (Tamandua), coatis (Nasua), and Woolly Opossums (Caluromys) show slight tendencies but with high uncertainty. The comparison between primates and sloths confirms two stable profiles: primates are mostly delivered as lower bodies (with possible diversification towards tail or single member at higher masses), whereas sloths (Choloepus) are predominantly delivered as upper bodies, with a probability that increases with mass. PCA and clustering recover three functional poles (large atelids, small to medium primates, and sloths) consistent with taphonomic expectations. The observed patterns are explained by (i) load reduction during flight (elimination of heavy and low-nutrition elements), (ii) sexual dimorphism modulating behavior at low masses, and (iii) anatomical constraints according to prey species, with possible variability by eagle sex. This framework refines the attribution of behaviors to recent or fossil assemblages (notably for primates) and provides a non-invasive indicator of delivered biomass useful for conservation monitoring in tropical environments. Its application to other regions of the range and to non-breeding periods, as well as to other large forest raptors, would allow assessment of external validity and clarification of the domains of application.