The ecology of coral larvae: settlement patterns, habitat selection and the length of the larval phase

Abstract

Underestimating the capacity of larvae to influence their fate has been a recurrent theme of marine ecology. Consequently, ecologists have focused on post-settlement processes to explain the distribution and abundance of benthic marine organisms. Coral larvae, in particular, are typically viewed as simple organisms with little capacity to control their position in the water column or sense their environment. In this thesis, I demonstrate that coral larval ecology can have a profound affect on adult distributions. Firstly, I examined the depth patterns of coral settlement over a two-year period at four locations around Lizard Island on the Great Barrier Reef. Many taxa showed a pronounced and consistent decline with depth. In particular, settlement of Isopora and Pocillopora was largely restricted to the reef crest. A similar pattern was evident in the adults of these taxa, suggesting that larvae can recognize and respond to cues from the parental habitat. To test this hypothesis the larvae of six common coral species, with contrasting depth distributions, were introduced into aquaria containing tiles conditioned at a depth of 2m, 12m and unconditioned tiles. Larval substratum preferences generally corresponded with those predicted on the basis of the depth distribution of the adults. Zone-specific species showed a clear and pronounced preference for tiles conditioned in the parental habitat. For example, Goniastrea aspera a reef-flat species, settled on shallow tiles in densities 4 times greater than on deep tiles. Similarly, Fungia horrida, a species locally restricted to deeper water, was 6 times more abundant on deep tiles. These results confirm that the depth distribution of these species is influenced, in part, by patterns established at settlement. Another area where coral larval ecology has been neglected is in investigating the scale of dispersal. To test the likelihood of localised recruitment and the potential of coral larvae for long distance dispersal, I compared the frequency distribution of settlement and the longevity of larvae in four Acropora and two faviid corals. Some settlement was record within 4 d of gamete release in all species, indicating a shorter pre-competent period in these species than has been generally accepted. Competence peaked at over 50% of the surviving cohort within 7-10 d after which the proportion competent to settle dropped rapidly in all species except A. valida and A. millepora, where 50% of the surviving cohort remained competent for over 30 d. Maximum competence periods were 110 d for A. valida, 60 d for A. millepora, 36 d for G. retiformis, 34 d for A. gemmifera and P. daedalea. These competence periods should produce sufficient gene flow to prevent populations diverging over a wide geographical area. However, larval survivorship was low in all species with less than 50% of larvae alive after 14 d and less than 10% alive after 30 d. Low survivorship combined with a rapid drop in the proportion of larvae competent to settle after two weeks suggests that numbers of migrants are unlikely to be sufficient to sustain population abundance on distant reefs.