Improving Abiotic Stress Mitigation via Predictive Modeling of Water Quality Parameters in Recirculated Aquaculture Systems

Abstract

A sustainable aquaculture solution can be provided by Recirculated Aquaculture Systems or RAS, nevertheless, Abiotic stress factors can negatively impact aquatic organisms' growth and well-being. This study’s purpose is to demonstrate how Random Forests machine learning method helps to develop a predicting model that can aid in forecasting and in the mitigation of abiotic stressors in Recirculating Aquaculture Systems by regulating water quality influences.
 The study used the historical data on water quality, such as temperature, dissolved oxygen, pH, ammonia, and TDS levels in constructing a Random Forest-based predictive model. Based from the results reveal, the developed prediction model using random forests machine learning method was 90% accurate in making prediction and improved abiotic stress in RAS.
 Understanding the complex relations between water quality indicators and abiotic stress variables in RAS is crucial for identifying major abiotic stress drivers and developing effective models for forecasting water quality parameters, which results in real-time insights and actionable information for making proactive decisions and employing adaptive management techniques.
 Furthermore, RAS improves aquaculture productivity while reducing environmental impacts, which results in increased productivity, resource utilization, and system performance. This study makes a vital contribution to the aquaculture sector by proposing a data-driven method to improve the control of water quality parameters in RAS and, eventually, raise the sustainability and effectiveness of Recirculating Aquaculture Systems