Behind anemone lines:: factors affecting division of labour in the social cnidarian Anthopleura elegantissima

When natural clonal aggregations of the sea anemone Anthopleura elegantissima expand and encounter other clones they can form distinctive anemone-free zones, several centimetres across. Contact between isolated pairs of nonclonemate polyps often triggers a directed, aggressive response via deployment of fighting tentacles (acrorhagi), suggesting that agonistic interactions between polyps arrayed along interclonal boundaries produce the anemone-free borders. Prior studies indicated that competing clones consist of at least two morphologically and functionally distinguishable castes: (1) small, well-armed 'warrior' polyps along borders and (2) larger 'reproductive' polyps that are more likely to carry ripe gonads. Here we show that patterns of division of labour are more complex, with clones potentially consisting of up to five intergrading castes: 'scouts' (small polyps along interclonal borders); 'warriors' (larger polyps in rows immediately adjacent to scouts); 'free-edge' polyps; and in the centres of clones, large 'reproductives' and small 'reserves'. Quantitative observations of interactions between intact clones further revealed the expression of a complex array of previously unknown behaviours, including (1) incursion of 'scouts' into border areas and their subsequent death or,repulsion, (2) the occurrence of multipolyp conflicts and (3) stereotypical searching behaviour of heavily armed warriors immediately following tidal inundation. Field surveys of intact clones also corroborate the existence of substantial interclonal variation in patterns of division of labour. To distinguish environmental from genetic and ontogenetic contributions to this variation, we experimentally tested the capacity of two stimuli characteristic of interclonal contact (tentacular contact and acrorhagial attacks) to induce morphological characteristics typical of warrior castes. These stimuli significantly increased average numbers of acrorhagi per polyp for reproductive polyps, but the magnitude of response varied among clones. This clone-specific variation in responsiveness, together with the complexity of agonistic behaviours, may partially explain the expression and maintenance of interclonal variation in the frequencies of different castes and patterns of division of labour.