Three brooding corals, Seriatopora hystrix, Stylophora pistillata, and Pocillopora damicornis in Nanwan Bay, southern Taiwan, revealed a diel rhythm of larval release. Planulation by S. hystrix and S. pistillata was highly synchronized with one peak of planula release occurring close to sunrise. Planulae of P. damicornis were released throughout the day with two peaks occurred in the early morning and at night. We maintain corals in laboratory with varied light-dark cycle and temperature to determine the mechanism of their release rhythm. S. hystrix did not release larvae under constant light and constant dark, thus the release of larvae in S. hystrix is not controlled by an endogenous rhythm. Peak of larval release occurs after 23hrs of light cue and under the dark. Temperature treatments, in 1¢J intervals from 23.5-28.5¢J, did not change the timing of larval release. We infer that the diel rhythm of larval release of S. hystrix controlled by sunrise. S. pistillata did not release larvae under constant light and constant dark. The larval release pattern of S. pistillata is similar to S. hystrix. Thus, we infer that sunrise may be the cues for the diel rhythm of larval release of S. pistillata. P. damicornis releases larvae under constant light and constant dark, and exhibit a rhythm of larval release with 37 h periodicity under constant dark. The regulation of its larval release is complex. These results suggest that the interspecific mechanism controlling diel rhythm of larval release may be different.

The seasonality, lunar and diurnal periodicity of planulation of five scleractinian corals, Seriatopora hystrix, Stylophora pistillata, Pocillopora damicornis, Euphyllia glabrescens and Tubastraea aurea, were compared. Corals were collected monthly and maintained in outdoor, flow-through systems to quantify the number of larvae released. Planulation of the five species occurred throughout the year. The long period of reproduction may be related to the warm mean seawater temperature in Nanwan Bay. S. pistillata released few larvae during late summer and early fall, suggesting the seasonality of reproduction. The pattern of lunar cycle can be classified into three strategies: (1) Planulation of S. hystrix and P. damicornis showed clear lunar periodicity and the peak of larval release shifted with different months. The peak of planulation occurred around the full moon in winter then shifted gradually to the first quarter moon phase in summer. The phase shift of larval release may be related to the seasonal variation of seawater temperature and the sudden temperature drops caused by upwelling. (2) Planulation of S. pistillata showed clear lunar periodicity and the peak of larval release occurred around the full moon. This may promote the larval dispersal. (3) E. glabrescens and T. aurea showed clearly lunar periodicity and the peak of larval release changed with different months. The peak of planulation shifted from full moon to new moon during winter and spring while it shifted from new moon to full moon during summer and fall. The pattern of diurnal cycle also can be classified into three strategies: (1) Planulation of S. hystrix and S. pistillata revealed a well-defined diurnal pattern with most planulae being released close to sunrise. (2) P. damicornis and E. glabrescens released larvae throughout the day, with two peaks occurring in the early morning and in the night. Corals of all three pocilloporids and E. glabrescens timed larvae release relative to the light-dark cycle suggesting that diurnal cycle determined when larvae were released. Larvae released in the dark may have lower risk of predation effect. Furthermore, the peak planulation occurred close to sunrise may be advantageous for some mature larvae to have light cues for quick settlement within the natal reef. (3) The azooxanthellate coral T. aurea released larvae throughout the day with no apparent cycle. These results suggest that zooxanthellae may play a role in regulating the diurnal cycle.