Reference: IUI. Hull., 146: 32-43. (February, 1974) SHELL SELECTION AND AGGRESSIVE BEHAVIOR IN TWO SYMPATRIC SPECIES OF HERMIT CRABS WILLIAM C. GRANT, JR., AND KEVIN M. ULMER 1 U'illiniiis Collcf/c. Williamstown, Massachusetts, and Mt. Desert Island Biological Laboratory There is abundant evidence that hermit crabs do not enter gastropod shells at random but select shells according to species type and associated characteristics of shape, shell covering, dimension and weight. Reese (1962a, 1963) and Orians and King (1964) concluded from investigation of shell selection behavior in several species of hermit crabs that both preference for and availability of different mollusc shells influence the frequency distribution of shell types occupied. Shell con-figuration, aperture size, shell weight/crab weight indices and shell weight/shell volume indices, etc. (Markham, 1968; Hazlett, 1970b) all provide important evaluation stimuli for selection. Shell entering behavior is elicited through tactile and proprioceptive stimuli from mechanoreceptors on chelipeds and periopods which trigger a sequential series of fixed motor patterns. Reese (1963) and Hazlett (1971) present evidence that such patterns develop in glaucothoe larvae at the time of their initial entrance into shells. Kinosita and Okajima (1968) have described a series of behaviors in the land hermit crab, Coenobita rngosus, in which "measurements" of shell aperture by probing and extension of chelipeds provide crabs with information for selection of larger shells as they grow. In aggressive encounters which may be associated with competition for suitable shells and which may result in establishing dominant-subordinate relationships, hermit crabs display characteristic patterns of behavior. These have been analyzed in detail by Hazlett (1966a. 1967, 1968a, 1968b, 1970c, 1972a) and by Hazlett and Bossert (1965, 1966). Elevation of the body and extension of chelipeds pro-vide effective signals for altering behavior in combatants. Increase in weight and visual shell size enhances the probability of a crab initiating or dominating an encounter although when size differences are pronounced larger animals tend to ignore display by smaller individuals. Shell size has no effect on the level of aggressiveness in Pagurns hirsutiusculus according to Vance (1973b) although Hazlett (1970b, 1970c) found that occupancy of inappropriately sized shells in-creased aggression in both P. bcrnhardus and Clibanarius vittatus individuals. Aggressive contact may be inter-or intra-specific depending on the species in-volved. According to Reese (1962b, 1964) attacks may be prevented if subordinate crabs assume submissive postures. The probability of a crab winning an en-counter is influenced by its past history of success or failure as shown by Hazlett (1966b) and Mainardi and Rossi, (1971). According to Rossi (1971) dominant crabs are larger, obtain more food and molt more frequently. The investigations of Courchesne and Barlow (1971) and Hazlett (1966b, 1968c) have shown that isolation increases and aggregation decreases aggression, although agonistic be-1 Present address : Woods Hole Oceanographic Institution, Woods Hole, Massachusetts 02543. 32
BioStor
