文摘
We measured the daily abundance of larvae of eight species of ocean-spawned, estuarine-dependent fishes to determine the effect of sampling frequency on the mean and variance estimates during larval immigration past a permanent sampling station inside Beaufort Inlet, North Carolina, mid-November 1991 to mid-April 1992. Species of interest were Brevoortia tyrannus, Lagodon rhomboides, Leiostomus xanthurus, Micropogonias undulatus, Mugil cephalus, Paralichthys albigutta, P. dentatus, and P. lethostigma. Our data suggest that sampling at in tervals > 7 days can lead to excessive variance in abundance estimates. For all species, abundance varied as much as an order of magnitude from night to night. Proportional residuals from polynomial models of the seasonal recruitment pattern for a given species were used to assess the potential influence of nine environmental variables on daily densities. Twenty-seven of 72 correlations of proportional residuals with environmental variables were significant (P < 0.05). Proportional residuals were positively correlated with time after dusk for six of eight species and were negatively correlated with turbidity for five of eight species. However, interpretation of correlations must be done cautiously because a species recruitment pattern may coincide with normal seasonal change in one or more environmental variables. Variability in transport of larvae, from offshore to near the inlet and then through the inlet to the station, probably influences species abundance at the sampling station more than locally acting environmental variables. Daily collections or B. tyrannus larvae provided otoliths (n=1,341) showing that a large number of younger larvae, averaging 55 days posthatch, arrived at the station in mid-March on the date of maximum observed daily density (160 larvae per 100 m3).