A comparison of GIS- and hydrodynamic-based depositional modelling for the purpose of predicting freshwater finfish cage footprints

Description

Aquaculture regulators frequently require the use of a depositional model to accompany new site licence applications and support requests for changes to feed quotas or production. DEPOMOD is the most widely used model to predict the extent and intensity of deposition of solid waste from fish farms and associated environmental impacts. Within freshwater systems, DEPOMOD is not able to predict certain parameters with accuracy, and the production of a DEPOMOD prediction is time consuming. It also requires extensive data collection and analysis using specialized software that must be purchased and licensed. Although other modelling processes exist, such as FVCOM, they too tend to be data intensive and expensive resulting in a need to explore alternative and more practical modelling process for freshwater systems. In particular, a modelling approach that incorporates cage movement, while focusing less on the details of hydrodynamic processes, may produce footprint predictions with similar accuracy and less costs. Additionally, if this approach proved promising, then it may be possible in the future to consider how mooring standards could be designed as a way to minimize impacts of deposition on the benthic environment.

This project aimed to use existing data sets to compare the statistical error of depositional modelling estimated using DEPOMOD to that of a GIS-based model (MSDM) that incorporates cage movement. This information can then be used to develop a new basic computational program that will more accurately estimate the depositional footprint of a cage farm based on feed usage and quality, production size, average current speeds, and bathymetry.

Findings

Both DEPOMOD and MSDM predictions were compared with sediment trap measures of carbon depositions at two fish farms in Lake Huron. Average model error was smaller for the MSDM compared to DEPOMOD suggesting further development of this approach is warranted. In the future, the model must be adjusted so that it spreads the deposited waste across the total cage water surface area. Current model input coding has all waste enter at the center of the cages and it produces unrealistic low predictions in some locations under cage artifacts that adversely impacted our model diagnostic measures.

An accepted functioning depositional model should make it possible for regulators to make objective science based decisions on the suitability of new farm sites and to defend licencing decisions to the public and industry.

Program name

Aquaculture Collaborative Research and Development Program (ACRDP)

Years

One year: 2018-2019

Principal investigator

Cheryl Podemski, research scientist, Fisheries and Oceans Canada, Freshwater Institute, Central and Arctic Region
Email: cheryl.podemski@dfo-mpo.gc.ca

Team members

Jamie Raper, physical scientist, Fisheries and Oceans Canada, Freshwater Institute, Central and Arctic Region
Gord Cole, co-owner, Aqua-cage Fisheries

Collaborators

Jamie Hooft, senior scientist, Aqua-cage Fisheries Ltd.