Anemonefish recruitment: the roles of order and chance

Asterisks (*...*) surround words or phrases that are to be italicized.Near Madang, Papua New Guinea, nearly all actinians of the species *Heteractis magnifica* contain anemonefish - either *Amphiprion perderaion* or *A. percula*. There were 2-6 *A. percula* per anemone; number of fish, length of largest fish, and total length of all fish correlated with host diameter. In *A. perideraion*, with 1-5 fish /anemone, "quantity" of fish was not correlated with actinian size. For both species, size interval between fish diminished with increased number of fish in an anemone; average size of female, male, and largest juvenile tended to be larger in anemones with more fish. Variation in carrying capacity of this actinian with species of symbiont demonstrates that generalizations from one species of fish or anemone, or one host-symbiont pair, may be unjustified. Removing none, some, or all fish from an anemone allowed evaluation of factors affecting recruitment. No post-larvae settled into actinians from which fish had not been removed. Only fish of the resident species recruited to anemones occupied by fish. Fish of either species settled into an empty actinian, but the number of anemones colonized by fish of each species was proportional to the fish's abundance in the breeding population. Therefore, although settlement into an individual host was stochastic, for the population as a whole settlement was deterministic