Development of an Integrated Water Quality Prediction Model for Recirculating Aquaculture Systems under Tropical Conditions in South West Nigeria

The reliable prediction of water quality dynamics in recirculating aquaculture systems is essential for preventing catastrophic dissolved oxygen and ammonia events that cause mass fish mortality, yet mechanistic models calibrated to the specific biofilter microbial communities, water chemistry, and ambient temperature regimes of South West Nigeria are absent from the literature. This study develops and validates an original integrated water quality prediction model for recirculating aquaculture systems operating under tropical conditions, calibrated using data from four commercial tilapia and catfish recirculating aquaculture system farms in Lagos and Ogun States. The model couples a dynamic fish bioenergetics submodel with a nitrification kinetics submodel, a carbon dioxide production and stripping submodel, and a dissolved oxygen mass balance submodel, implemented in a custom Python simulation environment. Biofilter nitrification kinetics parameters were determined experimentally from ammonia oxidation rate measurements at representative temperatures (25 to 34 degrees Celsius) and substrate concentrations using batch respirometric assays with biofilter media. Model calibration was performed against 24-hour time-series water quality profiles measured at 15-minute intervals across all four farms. Validated model predictions of dissolved oxygen (RMSE = 0.31 mg/L) and total ammonia nitrogen (RMSE = 0.08 mg/L) were within acceptable engineering design tolerances. The model was used to simulate the consequences of feeding interruption, power outage, and biofilter backwash events, generating scenario-specific early warning thresholds for operators. The study presents an original predictive tool for recirculating aquaculture systems management in tropical Nigeria.

Keywords: recirculating aquaculture systems, water quality modelling, nitrification, Lagos, tropical aquaculture