Two numerical models of the circulation and upwelling in the South Australian region were constructed. The first, a coarse resolution model (CRM), was designed to capture the equatorward Sverdrup transport that leads to the deep upwelling favourable Flinders Current. The average transports for the model period, December 1998-March 1999 (obtained using daily atmospheric forcing), compare favourably with estimates obtained from a summer-time climatology of a global ocean model. Within the CRM, a fine resolution model (FRM) was developed. This model has a very fine mesh and was able to resolve scales down to 2 km in the primary region of study: the mid-Bight to Robe. Along the open boundaries of the FRM, the velocities are specified using daily output from the CRM. In order to represent wind forced energy that is generated on the West Australian shelves, a coastal-trapped wave (CTW) “paddle” was added at the western boundary of the FRM (Esperance). This paddle is modulated using sea level observations from Esperance. A similar CTW paddle was introduced at the eastern boundary of the FRM, using sea level observations from Portland. Good agreement with coastal sea level data was obtained for signals in the 5-20 day (weather) band. These results were obtained using daily averaged atmospheric forcing fields. Preliminary results obtained using 6 hourly atmospheric fields, which include the sea breeze, also indicated that some of the 1-3 day variability could be accounted for and that more realistic, deeper surface mixed layers would result.

 

During a period of upwelling favourable winds, the model results indicated that water is drawn shorewards across the 200 m isobath in the SA Gulfs region. The coldest water (13oC) is upwelled from depths of 200 m and at sites to the south of Kangaroo Is and off Robe. Through bottom advection this water eventually spreads to the north-east and north-west of Kangaroo Is. The upwelling off the Eyre peninsula is identified to result from water upwelled to the south of Kangaroo Is. The surface signature of the upwelling compares favourably with satellite derived sea surface temperature (SST): plumes of cold 16oC water were advected to the north-west. For other regions the model SST was also in general agreement with the data except near the mid-Bight coastal region, where the model temperatures were too cool. The latter result supports the analysis above, that the coastal heating here is under-represented by the forcing fields.

 

A review of the literature has shown that population, breeding success and foraging patterns of land breeding marine predators such as seals and seabirds, are sensitive to changes in their marine environments and can be used as appropriate monitoring tools of ecosystem change, and fishery impacts. To undertake such studies, however, is not a simple exercise. It requires an understanding of the spatial and temporal variability in the ecosystem, trophic relationships between predators and prey and the distribution of foraging effort of predators so that discontinuities between the areas of potential anthropogenic impact and other areas can be understood and assessed. A suite of appropriate species, parameters and sites were identified to focus future research in this area.

 

We examined the potential suitability of a range and seal and seabird species to provide data on the spatial and temporal variability of their prey resources within and outside the main pilchard-fishing zone (PFZ) within the eastern GAB ecosystem. For each of these species we also assessed a range of population status, reproduction success and foraging parameters that could potentially be monitored at sites within and outside the PFZ. Each of these was discussed with respect to previous studies that had demonstrated their use, and the practical/logistic feasibility of undertaking such monitoring in a systematic fashion at a range of suitable sites.

 

A preliminary conceptual and trophodynamic model for the pelagic waters in the eastern GAB ecosystem was developed to describe and facilitate an understanding of the structure and dynamics of the food web in the region. An Ecopath model was developed that incorporated all available data. A preliminary conceptual food-web model was also developed, based on trophic levels and interactions. There was considerable uncertainty in many of the parameters and dietary profiles entered into the Ecopath model and as a consequence considerable adjustments were required in order to balance it. Most notable were adjustments to the dietary composition of fur seals (highest trophic level), and the biomass of little penguins and shearwaters. Additional modifications had to be made to the biomass of the tuna and kingfish/samsonfish groups, and it is likely that the modified parameters used were unrealistic. In addition, due to paucity of data, some species were amalgamated into groups that in future should be separated. Groups such as benthic and pelagic sharks have not been included. 

 

As this model is only in the developmental stage, it is not yet capable of providing quantitative predictions. It does, however, provide a useful framework within which the complex nature of trophic interactions between species groups can be conceptualised and importantly help to identify data-poor components. Already, this approach has revealed major gaps in data on the biomass of species groups and their trophic interactions, which will form the focus of studies to be undertaken as part of the long-term project, detailed in an FRDC proposal for funding 2005.