Growth and mortality of fish larvae in the northcentral Gulf of Mexico and implications to recruitment

Extensive plankton collections were taken during September for several years along the northcentral Gulf of Mexico's inner continental shelf. Significant differences in growth rates were found for both vermilion snapper and Atlantic bumper larvae collected at stations where hydrographic conditions were similar, indicating the probable existence of small-scale patchily-distributed feeding conditions. Such variability in growth rates is probably inevitable given the high variability in the observed spatial and temporal distribution of surface chlorophyll a levels, macrozooplankton biomass, copepod abundance, and dispersion of fish larvae.;Cruise estimates of mortality coefficients (Z) for larval vermilion snapper, red drum, and Atlantic bumper were derived from the descending limbs of age-frequency distributions (adjusted for stage duration). Estimates of Z for larval vermilion snapper (n = 2,581, 4 cruises) and Atlantic bumper (n = 39,247, 7 cruises) ranged from 0.18 to 0.29 and 0.20 to 0.39, respectively. Mortality coefficients for red drum larvae (n = 13,658, 15 cruises) were highly variable (Z = 0.17-1.62), which is most likely confounded by the aggregate, synchronous, batch-spawning behavior of red drum which would modulate the relative abundance of age classes. The overall estimate of Z for red drum larvae (0.31), obtained by pooling catch data, was very similar to values found for vermilion snapper and Atlantic bumper.;Projections of juvenile recruitment based on these data show that even in a sub-tropical climate, where stage durations of larvae are short, changes in growth rates, particularly when combined with small changes in mortality rates, can cause large differences in cumulative larval survival. Even if the mortality rate is constant, twice as many vermilion snapper and Atlantic bumper larvae potentially survived to a length of 6 mm at stations with the fastest observed growth rates than at stations with the slowest observed growth rates. Accurately predicting recruitment variability may prove to be intractable because of the massive logistical task of adequately quantifying the spatial and temporal variability of the trophic relationships and physical dynamics involved.