Herpetological Review, 2002, 33(x), xx–xx. © 2002 by Society for the Study of Amphibians and Reptiles Scavenging Behaviors of Cottonmouth Snakes at Island Bird Rookeries HARVEY B. LILLYWHITE * COLEMAN M. SHEEHY III and MARSHALL D. MCCUE Seahorse Key Marine Laboratory, Department of Zoology The University of Florida, Gainesville, Florida 32611-8525, USA e-mail [HBL]: hbl@zoo.ufl.edu * To whom correspondence should be sent concerning this manuscript. Cottonmouth snakes, Agkistrodon piscivorus, are well known as generalist and opportunistic feeders consuming fish, amphib- ians, rodents, birds, various reptiles including snakes and conspe- cifics, and carrion (Gloyd and Conant 1990; Savitzky 1992). The Florida cottonmouth, A. p. conanti, is an abundant snake through- out much of Florida, with well-established populations on some of the offshore islands. The success of this species in varied habi- tats is related, in part, to the wide range of prey items that provide a potential energy base for populations (see also Fritts 1988). Insular populations of cottonmouths inhabiting the Cedar Keys of Florida have been studied by Wharton (1966, 1969) who fo- cused his investigations on an unusually dense population of snakes inhabiting Seahorse Key. This island is approximately 64 ha in size and was estimated to support 600 cottonmouths during the period when Wharton conducted his studies. The island has no permanent fresh water, and the cottonmouths inhabit a mixed hard- wood forest where they are entirely terrestrial. These snakes sub- sist largely or entirely on fish that are dropped or regurgitated by colonial wading birds that nest on the island in large numbers from March through September or October. At present, the colonial nesting birds are largely represented by Brown Pelicans (Pelecanus occidentaliscarolinensis ), Double-crested Cormorants (Phalacrocorax auritus), White Ibis (Eudocimus albus), and sev- eral species of herons and egrets. One of us (HL) has confirmed and extended the observations of cottonmouth behaviors recorded by Wharton. At Seahorse Key these snakes characteristically forage at night on substrate directly beneath bird rookeries where avian excreta alter the vegetation and humus layers of soil. Foraging snakes methodically investi- gate these environments where numerous objects become splat- tered with fluids that are regurgitated or excreted by birds and presumably contain fish odors. The snakes largely consume fish, although occasional young birds fallen from nests might also be ingested. It is not clear, however, whether birds are eaten for their own sake or because they bear fish odors. Dead birds that are ig- nored by snakes are frequently found in or near the rookeries. Both their shape and size appear to render them too difficult to swal- low. Foraging snakes are generally quite placid and can be ap- proached and observed at close distances. They readily eat fish when these are presented to, or thrown near, a foraging snake. Snakes are also observed to investigate and grasp objects such as plant materials, bones, or bird feathers that have contacted avian excreta or fish fluids. These are usually released following con- tact with the internal tissues of the mouth. Recently we observed some unusual scavenging behaviors in a foraging female cottonmouth estimated to be about 110 cm in to- tal body length. The snake was first seen at 1810 h EDST and was observed for 20 min while it foraged in an open area beneath a Brown Pelican rookery at Seahorse Key, 2 June 2001. The sub- strate was comprised of sandy soil covered with many fallen twigs, dried leaves, and other scattered debris. Most of these objects were white or mottled due to urates that were excreted by the pelicans nesting overhead. The area was in full sun, and the ambient tem- perature and humidity at the time of observation were 31.4°C and 61%, respectively. The snake appeared to be in good health and was somewhat distended posteriorad, suggesting possible preg- nancy. The snake crawled very slowly and deliberately over the sub- strate while investigating various objects with frequent tongue flicks. We first observed the animal through binoculars from a distance of about 8 m. Eventually, we advanced closer to the snake, which continued to forage and eventually approached us within 0.5 m distance. Although the snake appeared to be aware of our presence at this point, it continued to forage as described. Soon after we first saw the snake, it investigated the ground intensely at one location and positioned its head nearly vertically with the mouth directed downward as it grasped a dark object. When the snake lifted its head, the grasped item was seen to be soil as half of it broke away from that being held in the mouth and fell to the ground. The snake then lifted its head upward at an angle of about 30° and swallowed the soil, which was seen to be damp and loosely held together with moisture. Later inspection of the ground on which the snake foraged revealed several locations where the soil was wetted, due either to excreta or regurgitated fluids from birds overhead. Following this act, the snake continued to investigate objects that were on the ground. Next it investigated the blunt end of a wooden branch about 1 cm in diameter and bearing several stains of bird excreta. The snake grasped this branch and attempted to advance the jaws over it, but the object was soon dropped and the effort abandoned. The snake continued to crawl slowly over the substrate using rectilinear locomotion while investigating objects with searching movements of the head and tongue. Finally, the snake stopped to investigate an object roughly 4 x 4 x 1.5 cm in overall size and comprised of a partial fish skull with other bones, debris and several dry leaves adhering to it. The snake grasped this object and advanced its jaws to hold the bulk of the item firmly within its mouth (Fig. 1). Then it elevated its head further off the ground, crawled forward about 3 m and stopped with the head elevated. We believe the item would have been swal- lowed, but we disturbed the snake while advancing to take photo- graphs. The snake crawled forward and disappeared within the forest edge that bordered the rookery site. These and other observations of cottonmouth foraging behav- iors at Seahorse Key suggest that numerous, diverse objects of appropriate size, shape, and odor might be ingested by foraging snakes. Thus, feces that are defecated by wild-caught snakes some- times contain leaves that are ingested inadvertently as adhering parts of carrion. In general, Agkistrodon species appear to rely more on chemosensory information during foraging than do other crotalines (Chiszar et al. 1979; Cock Buning et al. 1981). It has been suggested that evolutionary reliance on an envenomation