Spatial structure of co-occurring anchovy and sardine populations from acoustic data: implications for survey design

A geostatistical analysis has been undertaken on the spatial structure of co-occurring adult and recruit populations of anchovy, Engraulis capensis, and sardine, Sardinops sagax, in the southern Benguela upwelling region, using information from two acoustic surveys. The study was prompted by the need for a more efficient design for surveying sardine abundance, which is increasing in relation to that of anchovy; the current acoustic survey design is based on the distribution of anchovy. Variograms of fish density and density indicator variables were computed, as well as cross-variograms between the indicator variables. The sardine variograms were less structured than the anchovy variograms, with slightly greater nugget effects, indicating greater randomness in space at small scales. The indicator variograms showed progressive loss of structure with increasing density. Anchovy formed high-density schools during the day, breaking into larger, low-density aggregations at night. Sardine, on the other hand, remained in relatively high-density schools throughout the day. The cross variograms revealed some spatial continuity between low- and high-density areas for anchovy, but no such transitional structures were evident for sardine. It was concluded that sardine are more patchily distributed than anchovy and, unlike anchovy, they may not have a single common way of occupying space at the population level, perhaps partly because of the broader age structure of the population. It is concluded that, while the current acoustic survey design is well suited to the spatial distribution of anchovy, it is not as well suited to that of sardine. Future survey designs should he more tailored to the spacial distribution of sardine, possibly by using sampling methods to cope with the highly patchy distributions expected.