A Trophic Flow Model of the Caeté Mangrove Estuary (North Brazil) with Considerations for the Sustainable Use of its Resources

The Caeté Estuary lies within the world's second largest mangrove region, 200 km south-east of the Amazon delta. It has an extension of about 220 km²and is subjected to a considerable human impact through intensive harvest of mangrove crabs (Ucides cordatus) and logging of mangroves. In order to integrate available information on biomass, catches, food spectrum and dynamics of the main species populations of the system, a trophic steady state model of 19 compartments was constructed using the ECOPATH II software (Christensen & Pauly, 1992). Ninety-nine percent of total system biomass is made up by mangroves (Rhizophora mangle, Avicennia germinans andLaguncularia racemosa ), which are assumed to cover about 45% of the total area and contribute about 60% to the system's primary production. The remaining biomass (132 g m⁻²) is distributed between the pelagic and benthic domains in proportions of 10% and 90% respectively. Through litter fall, mangroves inject the main primary food source into the system, which is either consumed directly by herbivores (principally land crabs, Ucides cordatus) or, when already metabolized by bacteria, by detritivors (principally fiddler crabs, Uca spp.). These two groups are prominent in terms of biomass (80 g and 14·5 g m⁻²), and food intake (1120 g m⁻² yr⁻¹and 1378 g m⁻² yr⁻¹respectively). According to the model estimates, energy flow through the fish and shrimp compartments is of relatively low importance for the energy cycling within the system, a finding which is contrary to the situation in other mangrove estuaries reported in the literature. The dominance of mangrove epibenthos is attributed to the fact that a large part of the system's production remains within the mangrove forest as material export to the estuary is restricted to spring tides, when the forest is completely indundated. This is also the reason for the low abundance of suspension feeders, which are restricted to a small belt along the Caeté River and the small creeks which are watered daily. Phytoplankton, temporarily refloating benthic diatoms, neritic zooplankton and small pelagic fish dominate the (low) pelagic biomass. Total system throughput (10 559 g m⁻² yr⁻¹) and mean transfer efficiency between trophic levels (9·8%) calculated by the model fit well into the range reported for other tropical coastal ecosystems. The very high gross efficiency of the fishery (catch/net primary production) of 8·6% and its low trophic level (2·1) is explained by a high harvesting rate of mangroves and the fact that the main animal resource in the system are the mangrove crabs (Ucides cordatus), which feed at the first trophic level. The model was balanced asuming a turnover rate for the land crabs of P/B=0·25 (P/B: production per unit of biomass) which is possibly too high. If this value was replaced by a (possibly more realistic) lower value, the model would not balance, suggesting a situation in which more biomass is being harvested than produced, which hints to an overexploitation of this resource A ranking of the various system components in terms of their contribution to the system function (ascendency sensu Ulanowicz, 1997) revealed that detritus and associated bacteria contribute 34%, mangroves 19%, fiddler crabs 13%, phytoplankton and microphytobenthos 10%, mangrove crabs 10%, and the remaining 14 groups 14% to the total ascendency. Summary statistics of the model are given and compared with those of other coastal ecosystems.