Shelter Selection by Spiny Lobster Under Variable Predation Risk, Social Conditions, and Shelter Size

Abstract

Shelter use patterns of den—dwelling Caribbean spiny lobster, Panulirus argus, appear to be regulated by predation risk. The risk of predation may be modified by (1) social structure, which alters the effectiveness of communal defense, and (2) the scaling between lobster size and shelter size, which enhances the protective capacity of the den. These hypotheses were tested with field enclosure experiments using artificial lobster shelters, which examined the effects of predation risk (i.e., presence or absence of a major predator, the nurse shark Ginglyostoma cirratum), spiny lobster size, social condition (i.e., presence or absence of conspecifics), and shelter size upon den choice by juvenile and adult P. argus. To corroborate the findings of the enclosure experiments we also quantified seasonal, size—specific abundance patterns of P. argus in the field by deploying artificial lobster shelters (casitas) of different sizes in two habitats that differed primarily in the potential for gregarious interactions: an inner—bay, sand seagrass flat with high lobster densities, and an outer—bay, seagrass bed adjacent to coral reefs with sparsely distributed lobsters. The experimental and observational field results were strikingly similar–social condition and the scaling of lobster size to shelter size jointly regulated den choice patterns of adult and juvenile Panulirus argus, particularly under high predation risk. When conspecific density and predation risk were low, lobsters resided primarily in shelters whose dimensions were scaled to their own; when conspecific density was high and predation risk was low, lobsters resided predominantly in large shelters offering the highest potential for gregariousness; when conspecific density and predation risk were high, lobsters shifted to gregarious habitation in smaller, scaled shelters; and, when predation risk was high and conspecific density was low, lobsters occupied smaller shelters. The frequency of gregariousness in the field was much higher at the inner—bay site, where lobsters were dense, than at the outer—bay site, where lobsters were sparse, even accounting for the difference in lobster density between sites. This study indicates that the density of conspecifics in a given habitat can enhance gregariousness in spiny lobsters, which in turn influences the relative impact of lobster size, shelter size, and predation risk upon den choice. In defining the critical determinants of den choice for P. argus, we also provide an empirical and conceptual framework for identifying how variation in the availability of resources, such as conspecifics and appropriately scaled refuges, influence the distribution and abundance of social, shelter—dwelling species.