Evaluation of nitrogen cycling and fish production in seasonal ponds ('Fingerponds') in Lake Victoria wetlands, East Africa using a dynamic simulation model

A dynamic model was developed to simulate nitrogen (N) flows and fish production in seasonal wetland fish ponds (Fingerponds) based on organic manuring and natural food production. The model incorporates pond water depth, food availability, fish stocking densities, fish and fingerling weights at stocking, reproduction rate, manure type and application rates. The ponds were fertilized fortnightly with 1042 kg ha−1 chicken manure. The model captured the dynamics of hydrology, nutrients and fish and demonstrated that similar fundamental processes underlie fish production in these systems. The model predicted annual fish yields of up to 2800 kg ha−1. Simulated fish production, chlorophyll a and dissolved inorganic N concentrations were comparable with field measurements. Using the model, N budgets and estimates of all N flows were made. Most of the N input into the ponds (60–70%) accumulated in the bottom detritus of the pond and only 8–10% was converted into fish biomass, of which about half consisted of small fish. Fish production in Fingerponds was limited by turbidity induced light limitation and by nutrient limitation. Reduction of variability of fish production should come from reduced turbidity and sufficient nutrient input to minimize light limitation and maximize fish growth.