1. Changes in the coastal ecosystem of the Cape Verde Archipelago over the period 1981 to 2000: a simulation model using Ecosim. K.A. Stobberup 1, V. M. Ramos 2, and M.L. Coelho 1 1: IPIMAR - Instituto de Investigação das Pescas e do Mar, Portugal 2: INDP - Instituto Nacional de Desenvolvimento das Pescas, Cape Verde Presented at the Symposium on “Marine fisheries, ecosystems, and societies in West Africa: half a century of change”, 24-28 June 2002, Dakar, Senegal.

2. Part 1: Ecosystem characteristics and time series data for Ecopath models Part 2: Model comparison and simulation results

3. Geographic location

4. Oceanographic conditions Stronger influence from the Canary Current in the windy season from December to May. Northeasterly winds and colder sea surface water temperatures.

5. Revision of catch time series

6. Effort – Evolution over Time

7. Pelagic Catches – Evolution over Time

8. Other important catches

9. The first preliminary model for the period 1981-1985 (published in the context of SIAP) had several weak points that had to be dealt with: Statistics available for the 1981-1985 period were in need of revision in order to assess correctly the impact of fishing on the system. Difficulty in delimiting the system and problems in handling the oceanic, migratory groups (e.g. tuna). The model “fed heavily on” the Opitz (1993) model for the Caribbean coral reef ecosystem in terms of definition of functional groups and their corresponding diets. Developments

10. Functional groups defined specifically for Cape Verde (based on 99 more common fish and shark species and their corresponding diets)

11. Part 2: Model comparison and simulation results

12. 1.Construct Ecopath models for two time periods; 1981 - 1985 and 1994 - 1997. 2.Compare models and determine whether important changes have taken place. 3.Simulate (Ecosim) from the first period to the second and determine what are the possible causes of these changes. 4.Compare the model for the 1994 - 1998 period with simulation results. 5.Given the limited data available, consider whether the simulation model fits the observed trends. Steps undertaken

13. Model and Simulation Comparisons (higher trophic levels only!)

14. Fishing Mortalities estimated for the two time periods

15. Fitting to time series: biomass or CPUE The fitting process in Ecosim minimizes the SS in relation to absolute or relative biomass. But biomass estimates always consisted in one point only over a 15 year period. Thus, the fitting process included the catch time series (fitting by eye!) as well as the available information on absolute and relative biomass.

16. In order to simulate a general decrease in biomass a decreasing trend in primary productivity was necessary Forcing function applied to Primary Producers

17. Considering the limited data available, Ecosim proved to be a useful tool for time series fitting, but also to identify problems in the base model (e.g. parameters, diet). The coastal ecosystem of Cape Verde can not be regarded as a closed ecosystem. In order to sustain the system, it is necessary to include the import of food for tuna, in particular. Changes have occurred over the period under study, seen as a general decrease in biomass. This change can not be attributed to the effects of fishing and it appears to be a result of a “regime shift” in terms of productivity (consistent with Klyashtorin 1998 – ACI index) causing a consistent decrease overall. The Small Pelagics group plays a key role in the coastal ecosystem and our findings appear to support the theory of a tropical wasp-waist ecosystem, particularly for the pelagic component. Given the physical forcing on the system, Ecosim should be used with extreme care as a management tool in Cape Verde. Conclusions

18. Small Tuna Pelagic predators Demersal predators Sparids Large tuna Jacks Small pelagics Demersal fish An excercise on effects of increasing fishing effort Small tuna