Potential vs. realized larval dispersal: fish predation on larvae of the ascidian Lissoclinum patella (Gottschaldt)

The total swimming time of an invertebrate larva in a closed container is seldom a reliable indication of its swimming time in the field. In this paper we compare in situ observations on larval dispersal of the colonial ascidian Lissoclinum patella (Gottschaldt) to the total swimming time of the larvae in closed containers. A total of 123 out of 133 larvae were followed from release until they were consumed by predators (87%) or settled (13 %). More than three-quarters of the predation was from pomacentrid fishes [primarily Pomacentrus popei Jordan & Scale and Glyphidodontops rollandi (Whitley)]; the rest was from corals and zoanthids. In closed containers the larvae swam for > 2 h, but larvae followed in situ persisted for < 10 min. The behavior of the larva and the effect of predators greatly reduce the larva's realized dispersal, and may even limit the vertical distribution of adult L. patella. Although L. patella larvae have a potential dispersal distance of several hundred meters (depending upon local current speed), their realized dispersal within a coral reef appears to be < 10 m, and the majority of the larvae end up in fish guts.