We deployed three replicate larval light traps off the upper Florida Keys from June-September 2001 to measure the delivery of settlement-stage fish larvae to the coral reefs. Nightly measures of larval abundance were compared to water temperatures measured across the outer reef, nearby wind records, and the alongshore and cross-shelf components of the currents measured at the seaward edge of the reef. These time series, together with satellitederived sea surface temperature and color fields indicate that a very large multi-taxa larval pulse on 20 July was directly associated with the passage of a Florida Current (FC) sub-mesoscale frontal eddy embedded within the elongated remnant of a mesoscale eddy. A second large pulse of larvae occurred when a similar mesoscale eddy passed the upper Keys in mid-June. Periods of increased tidal bore activity occurred with the passage of these eddies. Semidiurnal internal tides caused near-bottom onshore intrusions of cooler slope waters during periods of onshore meanders of the FC front when the downstream baroclinic flow and stratification increased at the reef margin. The high abundance and similarity in larval ages within taxa on 20 July indicate a Keys shelf origin, although the temporal and spatial scales of entrainment cannot be resolved. The passage of a third mesoscale eddy in September did not result in a larval pulse, possibly because of a mismatch between biological and physical criteria, several of which must be met for larval transport by mesoscale eddies to be successful.

Thirteen cohorts of the Caribbean reef fish Thalassoma bifasciatum were collected over 4 yr (2000 to 2003) in the upper Florida Keys, USA. Juvenile fish were censused and collected from replicate reefs shortly after settlement. The otoliths were examined to obtain early life history information such as timing of spawning, larval growth, pelagic larval duration, size-at-age, timing of settlement, and juvenile age and growth. Mean water temperature over the reef during the larval period explained 78% of the variation in larval growth among cohorts. Faster-growing warm-water fish had shorter pelagic larval durations (PLD), and larval growth explained 85% of the variation in PLD. Relative (otolith) size-at-settlement was a function of larval growth as well as PLD: settling larvae were largest at intermediate water temperatures. Early juvenile growth was also directly related to water temperature, which enabled smaller warm-water settlers to grow rapidly and eventually exceed juvenile size-atage of the cooler-water cohorts. Cohorts encountering intermediate water temperatures remained the largest throughout early juvenile life on the reef. The relative size of recruitment events could not be explained by any larval or juvenile trait, nor by water temperature. Recruitment was generally low for cool-water cohorts but quite variable among warm-water cohorts. This may be due to increased difficulty in sustaining high growth rates in warm water (i.e. beyond a threshold temperature of 28.5°C) or the interference of mesoscale advection processes. When 4 cohorts that settled during the passage of mesoscale eddies were omitted from the analysis, 61% of the variation in recruitment magnitude could be explained by water temperature alone. The dynamic oceanographic setting of the Florida Keys may obscure the relationship between seasonal water temperature, early life history traits, and magnitude of recruitment events. KEY WORDS: Reef fish settlement and recruitment • Larval and juvenile growth • Pelagic larval duration • Otoliths • Water temperature Resale or republication not permitted without written consent of the publisher Thalassoma bifasciatum. New recruit, older juveniles, and terminal phase male (counterclockwise from upper right). Water temperature largely determined growth rates and duration of the pelagic phase in bluehead wrasse larvae. The relationship between water temperature and fish recruitment was decoupled by variable feeding success in warm oligotrophic water or by transport due to mesoscale oceanographic circulation.

New recruits of the bluehead wrasse Thalassoma bifasciatum were censused and collected from nearshore reefs of Barbados, West Indies, every 2 weeks for 20 months. Their temporal coincidence with low salinity (<34Á5) water during their pelagic larval stage was determined by comparing the otolith records of new recruits with conductivity and temperature records from a current meter moored 2 km off the west coast of the island. Larval residence in a low salinity North Brazil Current (NBC) ring appeared to have a negative impact on growth. Larvae that encountered a NBC ring for at least 7 days during either the first half of the larval period exhibited slower larval growth than those that did not encounter a ring for 7 days during any part of their larval period. As a result of this slower growth, larvae that encountered low salinity waters had a longer pelagic larval duration and were larger at the time of settlement. The magnitude of settlement was not distinctly related to the presence or absence of a NBC ring, but the largest settlement event occurred at the end of the longest ring event. Early juvenile growth did not vary between fish that had encountered a ring and those that did not, so size differences at settlement were propagated through the first week of life on the reef. The potentially opposing attributes of fast and slow-growing larvae (e.g. fast growing larvae with shorter larval stage duration but smaller size at settlement and higher susceptibility to reef predation), and the resulting differential mortality on the reef may promote the persistence of individuals in the population with contrasting life history traits, and contribute to the lack of a relationship between larval growth and recruitment success. Positive transport related effects of rings (i.e. enhanced retention during some ring events) may further complicate matters by outweighing the negative impact of rings on larval growth. # 2004 The Fisheries Society of the British Isles

We examined the relationship between the pelagic environment and the lunar timing of reproduction and recruitment of a coral reef fish by comparing patterns exhibited by fish under normal oceanographic conditions with patterns exhibited by fish that experienced significantly different pelagic conditions (i.e. encounter with meso-scale, low-salinity North Brazil Current [NBC] rings passing by the island of Barbados). We used a 20 mo time series of bluehead wrasse Thalassoma bifasciatum recruitment, and compared larval growth and the timing of settlement and (successful) spawning recorded in the otoliths of individual recruits that encountered an NBC ring (RING fish) with those that did not (NO RING fish). Spawning occurred during all times of the lunar cycle, but during NO RING conditions, only those larvae spawned during the first-quarter moon were retained nearshore. Successful NO RING recruits were spawned during the first-quarter moon, grew rapidly as larvae, and settled during the third-quarter moon and neap tides. In contrast, during RING events, larvae spawned during all lunar phases were retained nearshore. Successful RING recruits were spawned over the entire lunar cycle, grew more slowly as larvae, and settled during both quarter moons. Fish settling during the first-quarter moon were of higher condition than third-quarter-moon settlers, which is consistent with the concept of higher predation losses and selective mortality of settlers during less-than-optimum periods. Synchronized settlement patterns can be decoupled from spawning patterns by pelagic processes and flexible larval growth schedules. Variable pelagic conditions may contribute to the maintenance of daily reproduction and flexible growth histories in marine species.

No abstract