WP4

To develop a modelling tool for predicting the environmental response to waste inputs from marine cage farms production of sea bass (Dicentrarchus labrax) and gilthead sea bream (Sparus auratus) in the eastern Mediterranean

Objectives

To review the models used in the prediction of environmental impacts of marine cage farming, which have been developed in Norway, Scotland and elsewhere
To develop a model to predict the environmental response to waste inputs from marine cage farms production of sea bass and gilthead sea bream in the eastern Mediterranean
To utilise the data obtained on the physical characterisations of a wide range of farm sites to allow model parameterisation using both real and synthetic current data sets
To utilise the detailed environmental and faunal data obtained from the selected farm sites to validate model predictions with real data

Existing models and their use

A modelling tool now exists for predicting the response of sediments and their benthic macrofaunal assemblages to waste inputs from salmon cage farms. This model (DEPOMOD) has established correlations between benthic indices and predicted waste deposition and accumulation and is now being used in the environmental control of salmonid farms in Scotland

How will the existing model be developed?

A similar model will be developed for use in predicting fish farm effects in the oligotrophic waters of the Eastern Mediterranean. Since fish farming in these areas causes qualitatively and quantitatively different impacts both on the water column and benthos compared to Atlantic Europe, the model requires specific customisation of existing management practices and major re-parameterisation of predictive modules. The existing model needs to be developed in the following areas:

a validated component which models losses of waste particles to wild fish feeding needs to be added
measurements of settling velocities of feed and faecal particles common to fish farmed in the Mediterranean for use in the model
inclusion of temperature in the model where necessary to model fish growth and other factors such as degradation of solids discharged
establish correlations between solids accumulation and benthic response for several sites with sensitivity analyses

What are the deliverables?

A predictive model to assess the environmental impact of marine fish cage farms in the Eastern Mediterranean.

The model output would provide predictions of:

the area of sediment around the cages affected by solids deposition
the expected response of the benthic macrofauna to varying levels of solids deposition
the probable response of wild fish populations to varying captive fish production levels

Each component of the model will be validated independently so that the overall model performs reliably. The final model software will be a WindowsTM user friendly menu driven application with model documentation and user manual.

Background information on the existing model - input data

A flow diagram detailing the existing model DEPOMOD is shown below:

current velocity measurements for an area close to the fish farm with measurements at different depths in the water column
horizontal and vertical dispersion coefficients for the area
bathymetry, numbers, dimensions and positions of cages

For the stage of the growing cycle to be modelled:

feed input data (kg feed d-1 for the whole farm) and mean fish size
water content and digestibility of the feed
food and faecal settling velocity data

Background information on the existing model - predictions

This plot to the right shows the predicted solids deposition (g solids m-2 bed yr-1) for a fish farm (cages indicated by squares).

Although these results are for a particular stage of the growing cycle, different scenarios can easily be tested with the model

Established relationships between predicted solids accumulation (x-axis) and a benthic response (y-axis) have been established and an example of such a relationship is shown below

Circles demonstrate the variation in the benthic composition of duplicate grabs and the envelope of acceptable precision is defined to take account of this natural variation (88% in EAP, n = 42 stations)

The next plot shows the predicted benthic index for the site Different scenarios can be undertaken in the model to predict benthic impacts at different stages of the growing cycle or for different biomass consent levels.

Further details are available from: chjc@dml.ac.uk