|Abstract or Summary
- Decision support systems (DSS) have been used to a very limited extent in pond aquaculture. This study documents the development of a DSS (POND) which allows representation of an entire pond aquaculture facility
- Decision support systems (DSS) have been used to a very limited extent in pond aquaculture. This study documents the development of a DSS (POND) which allows representation of an entire pond aquaculture facility, and provides analysis capabilities in the form of simulation models and an economics package. Simulation tools in POND include temperature, water budget, fertilization, and fish bioenergetics models. Verification of the water temperature model at sites in Thailand, Honduras and Rwanda indicated that it would accurately predict daily temperatures over entire seasons or diurnal temperatures over one day intervals if complete input weather datasets are available. Similarly, adequate estimates of water requirements can be obtained from the water budget model. Sensitivity analysis with the former model, and results obtained from the latter, indicate that input weather datasets should include air temperature, relative humidity, short-wave solar radiation, precipitation and wind speed measurements. The fertilization model estimates fertilizer application rates on the basis of nutrient concentrations, gross primary productivity and nutrient recycling processes. Model output was more conservative compared to rates used in Honduras, Thailand and the Philippines, suggesting that responsive fertilization strategies which account for ambient pond water conditions are more efficient than fixed input strategies. The bioenergetics model accounts for the effects of size, water temperature, photoperiod, dissolved oxygen and unionized ammonia on fish growth. The model was calibrated and validated for Nile tilapia (Oreochromis niloticus), tambaqui (Colossoma macropomum), pacu (Piaractus mesopotamicus), common carp (Cyprinus carpio), and channel catfish (Ictalurus punctatus). Model experiments generated useful information regarding supplemental feed initiation and fish feeding rates. A resource substitution function was also used in this model to analyze the consumption of endogenous and exogenous food resources by Nile tilapia. This function suggests that adding supplemental feed to tilapia ponds may increase phytoplankton biomass because feed is preferentially consumed. A genetic algorithm-based technique was developed to automatically calibrate the bioenergetics model. This technique generates best-fit parameters by comparing results of multiple model runs to observed data. In general, results obtained from all the models suggest that POND should be a useful tool for managers, planners and researchers involved with pond aquaculture.