MEPS 296:93-106 (2005)  -  doi:10.3354/meps296093

ABSTRACT: The potential linkages among Cuba and geographically associated northwestern Caribbean locations were examined through simulated transport of snapper larvae for 5 harvested snapper species. The analyses are based on a coupled biophysical model incorporating realistic, intra-annual varying currents from a single year (1984), a Lagrangian stochastic scheme, and larval behaviors to find settlement habitat. Sequential runs centered on peak spawning months and lunar phases estimated the degree to which each spawning event contributes larvae to distant populations or to neighboring populations on the complex Cuban shelf. Results suggest that considerable levels of self-recruitment (ca. 37 to 80% total recruitment) structure Cuban snapper populations, in particular, those from the southern and north-central regions. The northern snapper populations exported larvae to the southern Bahamas, specifically to Cay Sal Bank (ca. 11 to 28% total recruitment), while, for more distant locations, the import of larvae from Cuba was negligible depending on the species. Regional oceanographic regimes for cubera, dog and gray snappers and site utilization (e.g. shelf geomorphology) for mutton snapper caused most within-species recruitment variability. However, a small lag in peak spawning times contributed significantly to high recruitment variability among species. Active virtual larvae stand a better chance of reaching settlement habitat, whereas spatial distribution of recruitment is similar but less structured (i.e. homogeneous low abundance) for passive larvae. This modeling approach produces spatio-temporal predictions of larval pathways with explicit measures of variance. Further, it allows for the quantification of relative levels of connectivity, a component needed in the design of marine reserve networks.

KEY WORDS: Spawning aggregations · Larval transport · Recruitment · Biophysical modeling · Connectivity · Marine reserve network · Lutjanidae · Cuba