Spatial and temporal variability in lipid content and gonadosomatic index of sardine (sardinops sagax) in the Southern Benguela ecosystem

Abstract

Determining lipid content for sardine (Sardinops sagax) in the Southern Benguela Ecosystem (SBE), through applying a method of fat staging, is a relatively easy method to use as a biological time-series to assess the condition of sardine. The condition of fish is an important indicator for fluctuations in the population size of post-recruit fish. Depending on the amount of fat associated with the stomach, one of seven fat stages was assigned to each sardine. Visually assigned fat stages were then converted to a lipid content value, given as a percentage of wet body mass, by applying a conversion equation to these fat stages. Time-series of lipid content and gonadosomatic index (GSI) were generated using general linear models (GLM). Results from these GLM's explained 34% and 39% of the observed variance in lipid content and GSI respectively. Monthly and annual least square means (LS mean) were derived from these GLM's to show seasonal variability in lipid content and GSI, for the period 1996 to 2003. Clear seasonal patterns in both lipid content and GSI were observed. Lipid content showed a decrease, but an interannual trend in GSI could not be observed, for the time series. The study area, which ranged from the Orange River mouth (28°40'S and 16°30'E) on the west coast to Port Elizabeth (25°37'E and 33°57'S) on the east coast was divided into four smaller zones to allow for spatial tests. Zone 1 being on the west coast and Zone 4 on the east coast. Both lipid content and GSI showed strong spatial differences, with lipid content being at highest levels in Zone 4 throughout the time series, and the highest GSI levels being observed in Zone 1.