Settlement vs. environmental dynamics in a pelagic-spawning reef fish at Caribbean Panama

How are the dynamics of the settlement of pelagic larvae of marine shore or ganisms into benthic habitats affected by larval production and mortality d ynamics, and by environmental (wind and tidal) control of settler delivery? We examined the dynamics of reproduction of the pelagic-spawning fish Thal assoma bifasciatum from that perspective at San Bias, Caribbean Panama. During April-June 1993, similar to 75% of females spawned each day, with me an output per female per day at about one-third the potential maximum. Outp ut peaked several days before the new and full moon, when tides were high d uring the afternoon spawning period. Daily fluctuations in output were weak ly correlated with light (+), swell (-), and onshore wind (-) on preceding days. Settler production (determined from otolith back calculations of settlers' fertilization dates) peaked during the wet season (May-November), when wind s were light and variable, spawning-period tidal heights were moderate to l ow, and tidal flows were off-reef at night. These conditions correspond to when mast settlement occurs. Settler production was minimal during the dry season (January-March), when onshore trade winds were strong and consistent , and tidal conditions were the reverse of those in the wet season. Settlement usually occurred in one variably timed pulse per lunar cycle, ar ound the new moon. Variation in monthly settlement strength over 11 yr (198 6-1996) was unrelated to onshore wind stress during either settler arrival or settler production. The daily dynamics of settlement were unrelated to t he concurrent dynamics of wind or tides, or wind dynamics up to a week befo re settlement. Settler production was much less continuous than spawning. Wet season settl er-production dynamics were unrelated to tidal dynamics, but (weakly) negat ively related to onshore wind stress. Although the dynamics of settler prod uction and settlement were similar, bouts of settler production were about twice as long as bouts of settlement. The average lunar patterns of settler production and observed spawning were similar (weakly bimodal). Settlers p roduced during different lunar phases differed in the length and variabilit y of their pelagic larval durations (PLDs). Fish with different PLDs all te nded to settle around the new moon. Weakly semilunar spawning may reflect effects of weak tidal influences on e gg survivorship, or of environmental constraints on adults' spawning capaci ty that also limit spawning to well below the potential maximum. Dry season trade winds may produce the seasonal low of settlement by depressing spawn ing or forcing the onshore transport of eggs and larvae. If settlement is t ransport-dependent, weak variable winds may have little effect on currents during the wet season, or local geography may limit such effects; tidal dyn amics may not affect wet-season settlement dynamics because small (<0.6 m) tides produce weak flows. Alternatively, settlement may not be dependent on transport mechanisms. Short-term settlement dynamics result from both the dynamics of larval surv ival (as most larvae have near-mean PLDs) and larvae delaying or advancing settlement to settle around the new moon. Larvae spawned at different lunar phases have different patterns of variation in their PLDs but achieve simi lar relative settlement success. Larvae with extended PLDs may not suffer l ow settlement success because the inevitable age-dependent decline in survi val has smaller effects on relative settlement success than does large mont hly variation in the survivorship of fish spawned during different lunar ph ases.