Planning and managing freshwater cage aquaculture in Scotland with combined catchment and lake models of phosphorus loading

Abstract

Phosphorus (P) is often the main limiting nutrient in freshwater lakes and reservoirs, so it is important to monitor and manage total P (TP) loading to minimise negative impacts on water quality and reduce the risk of eutrophication. In cage aquaculture, P enters the surrounding environment via waste feed, faeces, and excretory products. Static models for predicting P have been used for several decades, yet their capacity to represent the dynamic nature of lake systems remains unclear. There is also a need to consider TP loading from the wider catchment to better represent the range of P inputs to lake systems. This study developed a novel modelling framework by integrating the InVEST Nutrient Delivery Ratio (NDR) to simulate P loading from the catchment with 3 lake models to simulate TP: the static Dillon & Rigler and OECD models, and the dynamic LakeMab model. The approaches were demonstrated using four freshwater lakes (lochs) in Scotland where cages are used to grow Atlantic salmon (Salmo salar) smolts: Arkaig, Lochy, Shiel, and Garry. The mean relative errors for the modelled mean TP levels compared to the measured mean values were between: 1.54–2.57 μg L−1 for Dillon & Rigler, 1.86–2.74 μg L−1 for OECD, and 1.45–3.91 μg L−1 for LakeMab. The findings indicate that the three models have similar error rates; despite a slightly poorer fit, the dynamic model successfully captured seasonal fluctuations in TP. This study demonstrates an approach that can strengthen the monitoring, management, and regulation of freshwater aquaculture.