Influence of Size-Selective Mortality on Growth of Gulf Menhaden and King Mackerel Larvae

Abstract

Gulf menhaden Brevoortia patronus and king mackerel Scomberomorus cavalla represent two widely different larval life history strategies: feeding on large and small prey, respectively. We back-calculated lengths at age for wild and laboratory-reared larvae of gulf menhaden and wild king mackerel using direct proportion procedures then constructed matrices of observed age (rows) by increment number (columns) for mean back-calculated lengths at age. The coefficient of variation (100±SD/mean) in length at age was greater for observed than for back-calculated length at age for both wild and laboratory-reared gulf menhaden and for king mackerel. Columns in the length-at-age matrix of wild gulf menhaden showed significant trends of increasing back-calculated length at age for older larvae, but the matrix for laboratory-reared fish did not. We suggest that size-selective mortality—the culling of slower-growing larvae—was the cause of the different error structures of observed and back-calculated lengths at age as well as of the increasing back-calculated lengths at age for older larvae in the matrix of wild gulf menhaden. Predation may have been the cause of size-selective mortality because wild larvae were exposed to predation and laboratory-reared larvae were not. Slopes of regressions of back-calculated length on observed age for columns of the matrices indicate the time trend and intensity of size-selective mortality; in wild gulf menhaden larvae, size-selective mortality began after hatching, reached a plateau at 5–8 d, then declined markedly after 14 d, which suggests that the influence of predation was mainly expressed during this period. Size-selective mortality caused average growth (mean back-calculated or observed length at age) to appear higher for both species, but especially for gulf menhaden, because the smallest larvae of a given age were removed. We adjusted back-calculated growth by removing the effect of size-selective mortality with analysis of covariance and estimated that the observed growth rate was 25% higher than the adjusted rate for wild gulf menhaden and 7% higher for wild king mackerel.