Source and Specificity of Chemical Cues Mediating Shelter Preference of Caribbean Spiny Lobsters (Panulirus argus) AMY J. HORNER 1, *, SCOTT P. NICKLES 1 , MARC J. WEISSBURG 2 , AND CHARLES D. DERBY 1 1 Brains and Behavior Program and Center for Behavioral Neuroscience, Department of Biology, Georgia State University, P. O. Box 4010, Atlanta, Georgia 30302-4010; and 2 School of Biology, 310 Ferst Dr., Georgia Institute of Technology, Atlanta, Georgia 30332-0290 Abstract. Caribbean spiny lobsters display a diversity of social behaviors, one of the most prevalent of which is gregarious diurnal sheltering. Previous research has demon- strated that shelter selection is chemically mediated, but the source of release and the identity of the aggregation signal are unknown. In this study, we investigated the source and specificity of the aggregation signal in Caribbean spiny lobsters, Panulirus argus. We developed a relatively rapid test of shelter choice in a 5000-l laboratory flume that simulated flow conditions in the spiny lobster’s natural environment, and used it to examine the shelter preference of the animals in response to a variety of odorants. We found that both males and females associated preferentially with shelters emanating conspecific urine of either sex, but not with shelters emanating seawater, food odors, or the scent of a predatory octopus. These results demonstrate specificity in the cues mediating sheltering behavior and show that urine is at least one source of the aggregation signal. Introduction Many species of palinurid lobsters display gregarious social behaviors (Atema and Cobb, 1980). In the Caribbean spiny lobster, Panulirus argus, this sociality is evident in a variety of behaviors including the formation of long, single- file migratory queues, and defensive rosettes and aggrega- tions (Herrnkind, 1969, 1970, 1980; Berrill, 1975; Atema and Cobb, 1980; Herrnkind et al., 2001). However, the most ubiquitous example of their sociality is gregarious shelter- ing (Childress and Herrnkind, 1997). After solitary noctur- nal foraging trips, spiny lobsters often aggregate with con- specifics in dens where they remain sheltered throughout the day (Herrnkind et al., 1975; Kanciruk, 1980). Both the males and females shelter gregariously (Herrnkind et al., 1975, 2001), suggesting that this form of aggregation is not a sex-specific behavior (Zimmer-Faust et al., 1985). Although spiny lobsters are often aggregated in shelters, the extent of gregarious sheltering in any particular area is variable and influenced by several factors, including con- specific density, predation levels, and the number, availabil- ity, and size of suitable shelters (Eggleston et al., 1990; Eggleston and Lipcius, 1992). However, multiple occu- pancy of a shelter occurs more often than expected by random chance (Kanciruk, 1980; Herrnkind et al., 1975). The primary benefit of gregarious sheltering is believed to be a reduction in overall predation levels, which could be accomplished in several ways: through group defense or dilution effects, or via the guide effect, which suggests that spiny lobsters can minimize the amount of time spent searching for a shelter (thus minimizing their exposure to predators) by homing in on cues released from sheltered conspecifics (Eggleston and Lipcius, 1992; Childress and Herrnkind, 1997, 2001a,b). An essential first step to understanding how aggregation occurs is to identify the proximal cues that attract spiny lobsters to sheltering conspecifics. Shelter choice assays conducted in both the field (Nevitt et al., 2000) and labo- ratory (Ratchford and Eggleston, 1998, 2000) demonstrated that shelter selection by P. argus can be mediated by chem- ical signals released from conspecifics. In these studies, * To whom correspondence should be addressed. E-mail: ahorner@ gsu.edu Reference: Biol. Bull. 211: 128 –139. (October 2006) © 2006 Marine Biological Laboratory 128