Measuring and modelling the dispersal of salmon farm organic waste over sandy sediments

Abstract

Fish farm waste dispersal models are widely used but have only been directly validated to a limited extent. Two shallow (<20 m) Atlantic salmon farms (Bay of Meil and Quanterness) in Orkney, Scotland were studied. Bay of Meil has peak near-bed currents of 9.7 cm s-1 whereas Quanterness has flows up to 31.6 cm s-1. Sediment tray traps which allow resuspension to occur were deployed at each site. The patterns of particulate organic carbon (POC) deposition into the traps were in broad agreement with the observed water current directions and results from infaunal benthic monitoring. Despite the markedly different flow regimes at the 2 sites, most of the deposition occurred within 210 m of the cage perimeters. POC footprints were then modelled using the particle tracking model NewDEPOMOD. For Bay of Meil, a footprint was obtained using the recommended parameter defaults, but the spatial extent was too constrained compared to the sediment tray results. For Quanterness, all simulated particles were lost from the model domain and the critical erosion shear stress had to be increased to unrealistic levels to obtain a footprint. The failure to find a common set of parameter values applicable to both sites, despite their similar depths and sandy seabed, suggests that there remain unresolved issues, likely in how NewDEPOMOD handles waste resuspension. The sediment trays provided a direct method for quantifying the organic carbon deposition, facilitating direct validation of the dispersal model and demonstrating that further research is needed on fish farm waste dispersal at coarser sediment sites.