An individual-based modelling approach to examine life history strategies of sardine Sardinops sagax in the southern Benguela ecosystem

Abstract

Hypotheses regarding the spawning strategy and recruitment of sardine (Sardinops sagax) in the southern Benguela ecosystem are tested using an individual-based Lagrangian particle tracking model linked with a 3-D hydrodynamic model of the region. Experiments focus on the dispersion of eggs and larvae among possible spawning and nursery areas. The two main areas of interest were the west coast upwelling region and the south coast shelf region (Agulhas Bank). A stage-based temperature-dependent development model is incorporated and vertical positioning schemes are tested. The spatial distribution and size structure of the sardine spawning stock for the period 1991-1999 are presented and a simple size-based fecundity model, combined with modelled recruitment, is used to determine the relative importance of each spawning and nursery area. The area of spawning plays a fundamental role in determining the destination of spawned eggs, and recruitment of sardine in the southern Benguela ecosystem appears to be divided into three recruitment systems by the circulation of the region: eggs spawned west of Cape Agulhas recruiting on the west coast (the WAB/WC-WC system), eggs spawned east of Cape Agulhas recruiting on the west coast (the CAB-WC system), and eggs spawned east of Cape Agulhas recruiting to the south coast (the SC-SC system). There is a slight increase in retention in the two nursery areas during winter, but the transport of eggs and larvae from the Agulhas Bank to the west coast is optimal during spring to early summer. Slow development arising from cold temperatures on the west coast could negatively impact recruitment by increasing offshore loss of individuals before they develop to a stage when they are able to actively avoid offshore currents and through its effect on mortality rate. This could explain the spatial separation of spawning and nursery areas in this system. The vertical position of individuals has an effect on the level of modelled recruitment and mortality rate, but observed vertical distributions of sardine egg and larvae do not significantly increase the level of modelled recruitment to optimal nursery areas. This suggests that efficient transport and retention are traded-off against other factors such as predator avoidance or prey abundance. Observed size structure and spatial distribution of the sardine spawning stock for 1991-1999 fluctuated greatly with most spawning centred on the western Agulhas Bank. When spawning was centred east of Cape Agulhas, recruitment was poor. No significant relationship could be established between potential reproductive output reaching the west coast and estimated recruitment, but positive recruitment anomalies required good transport to, and retention on, the west coast. A conceptual model of the early life history of sardine is proposed in light of limitations imposed by transport and retention of individuals. Lower primary production and the possibility of higher predation on the Agulhas Bank suggest that the south coast supports less recruitment than the west coast. The hypotheses tested using available data and model results could improve the understanding of recruitment of sardine in this complex ecosystem. These need to be validated by field observations. Additionally, further avenues for research that could help in developing a better understanding of the sardine life history in the southern Benguela ecosystem are suggested.