The Lévy flight foraging hypothesis in a pelagic seabird.

ABSTRACT

Lévy flight foraging represents an innovative paradigm for the analysis of animal random search by including models of heavy-tailed distribution of move length, which complements the correlated random walk paradigm that is founded on Brownian walks. Theory shows that the efficiency of the different foraging tactics is a function of prey abundance and dynamics with Lévy flight being especially efficient in poor prey fields. Lévy flights have been controversial in some quarters, because they previously have been wrongly ascribed to many species through the employment of inappropriate statistical techniques and by misunderstanding movement pattern data. More recent studies using state-of-the-art statistical tools have, however, provided seemingly compelling evidence for Lévy flights. In this study, we employ these maximum-likelihood methods and their Bayesian equivalents by analysing both turning angles and move length distributions. We tested, for compliance with Lévy flight foraging, a set of 77 independent foraging trajectories of Cory's shearwaters Calonectris diomedea diomedea. Birds were tagged with high-resolution GPS loggers in two Mediterranean colonies (Linosa and Tremiti) during both incubation and chick rearing. We found that the behaviour of six birds was fitted by a correlated random walk; the movement of 32 birds was better represented by adaptive correlated random walks by switching from intensive to extensive searches; and the trajectories of 36 birds were fitted by a Lévy flight pattern of movement. The probability of performing Lévy flights was higher for trips during chick provisioning when shearwaters were forced to forage in suboptimal areas. This study supports Lévy flight foraging as an appropriate framework to analyse search tactics in this pelagic bird species and highlights that the adoption of a given search strategy is a function of biological and ecological constraints.