Modelling the effects of spatio-temporal spawning variability on the transport and retention success of sardine (Sardinops sagax) eggs and larvae around South Africa
Master Thesis
2018
Permanent link to this Item
Authors
Journal Title
Link to Journal
Journal ISSN
Volume Title
Publisher
Publisher
University of Cape Town
Department
Faculty
License
Series
Abstract
The transport and retention of sardine (Sardinops sagax) eggs on the west and south coasts of South Africa were investigated using a Lagrangrian, particle tracking, individual-based model (IBM), coupled with two 3-D hydrodynamic Regional Ocean Modeling System (ROMS) model configurations of different generations. The same IBM is coupled with PLUME, an older ROMS model configuration used in a previous sardine IBM for the southern Benguela, as well as BENGSAFE, a more recent, better resolved ROMS model configuration. Results from the two runs were compared to see if and how the underlying ocean model affects modelled retention and transport. An updated IBM, extended spatially and temporally to better represent sardine spawning and recruitment, was then coupled with the BENGSAFE model configuration, and assessed. The effects of spawning area, and month, year and depth of particle release, were investigated for each coupled 3D-IBM simulation using linear models. The portion of particles transported from the south coast to the west coast was a focus throughout, as this is of great importance to the management of the sardine fishery. In all simulations, three main systems were identified depending on where particles were released and recruited. Two of these systems were retention-based: particles released on the west coast and retained in the west coast recruitment area (WC-WC), and particles released on the south coast and retained in the south coast recruitment area (SC-SC). The other system is transport-based and represents particles released on the south coast and transported by ocean dynamics to recruit in the west coast recruitment area. Results were similar for the west coast when either hydro- dynamic model configuration was used; depth of release proved to be important in this system with its effect varying depending where on the west coast particles were released. Differences occurred on the south coast, with more transport to the west coast and more offshore loss in the higher resolved BENGSAFE IBM than the PLUME IBM. The difference was attributed to the positioning of the PLUME model's boundary in close proximity to the Agulhas Current, and the conditions at this boundary not properly resolving the current. The initial BENGSAFE IBM was spatially restricted to match the PLUME IBM, and therefore south coast spawning and recruitment was not fully represented. Therefore, the BENGSAFE IBM was spatially extended on the south coast and run for a longer time period. Similar results were obtained in the extended BENGSAFE IBM and the initial BENGSAFE IBM. However south coast retention increased and the average percentage of particles transported to the west coast decreased slightly. Spawning area proved to be an important determinant of retention and transport success in all recruitment systems, and the further east a particle was spawned the less likely it was to be transported to the west coast, and the more likely it was to be retained on the south coast. Transport to the west coast was most successful in late winter; a significant monthly effect on transport success was present. Correlations between standardised anomalies for stock assessment model estimated recruitment and IBM modelled retention/transport for the west and south coast were not significant, highlighting that retention and transport are not the only determinants of recruitment success. The extended BENGSAFE IBM can act as a base level to which extra layers of biological complexities can be added, such as horizontal and vertical egg distribution matching observations, incorporating gonadosomatic index (GSI) values, diurnal vertical migration and buoyancy.
Description
Keywords
Reference:
McGrath, A. 2018. Modelling the effects of spatio-temporal spawning variability on the transport and retention success of sardine (Sardinops sagax) eggs and larvae around South Africa. University of Cape Town.