top of page
Letter to ECCLR 26 Mar 2018 top
Home >Info > Letter to ECCLR 26 March 2018

Letter to the Environment, Climate Change and Land Reform Committee

26 March 2018

Dear Mr Dey


I hope this finds you well.


The Friends of the Sound of Jura were delighted to read the ECCLR’s thorough and uncompromising report on the environmental impact of salmon farming. Thank you for listening so carefully to the evidence.


We were interested to read the ECCLR committee's letter to SEPA, posing some outstanding questions to which the agency has not given satisfactory answers. We would be very interested in these answers and hope they be published on the ECCLR inquiry webpage.


The Marine Scotland letter is also interesting and asks some excellent questions. Marine Scotland has a very broad remit, including encouraging aquaculture, which may generate political pressures inside the organisation. This may explain why its advice to planning authorities on wild fish impacts has been criticised by them as being ambiguous and unhelpful. This conflict of interest could compromise Marine Scotland’s ability to be a fair broker on whether wild fish would be harmed by fish farms. We note that at present Scottish Natural Heritage seems to play no role in making these decisions, referring all enquiries about fish farms’ impacts on these two PMF species to Marine Scotland.


We are interested in SEPA’s answers to your questions, in part, because there have been a number of recent applications for salmon farms larger than the previous maximum permitted size of 2500 tonnes. Marine Harvest’s current proposal for Cheesebay, North Uist is one example. This is at the EIA scoping option stage.  (Screen grab below -


Cheesebay has a proposed new maximum size of 5950 tonnes. The same company has applied for planning permission for a 3500 tonne farm at Pol na Gille (Argyll) and is holding pre-application consultations about expanding its North and South Carradale farms to 3750 tonnes each.


Regarding the Cheesebay proposal, SEPA’s response to the screening and scoping request ) is to note that 

they ‘consider it to be at the applicant’s commercial risk to submit a planning application at this time for a fish farm to accommodate a maximum biomass in excess of that which we can licence under the current licensing regime and modelling.’ but add that ‘NewDepomod will be used to predict potential deposition resulting from the farm and to predict a sustainable maximum standing biomass to operate at the farm …'

'…SEPA's current licensing regime and the use of only AutoDepomod modelling, allows a maximum biomass of up to 2500 tonnes. However, we may allow a higher biomass if supported by more detailed hydrodynamic modelling; this would need to be submitted and approved by SEPA’s modellers.'


The site is close to existing farms in Grey Horse Channel, Groatay and Vacassay, so, to their credit, SEPA specify that, ‘an assessment of any potential effects from combined impacts (due to overlap of benthic impacts and sea-lice medicine treatments, both in-feeds and baths) will be required’ and a similar cumulative impact assessment for nutrient loading.


Regarding their Carradale proposals, Marine Harvest say that there has never been a ban on farms above 2500 tonnes but that this size is the upper confidence limit of AutoDepomod. They add that they have had initial discussions with SEPA on using NewDepomod to model pollution from a larger farm at N Carradale and that they will produce a full modelling statement in due course to address SEPA’s requirements.


The industry is clearly pressing on with its plans for expansion without waiting to see what the REC committee’s inquiry concludes.


What concerns us most is that these examples of expansion are based on using SEPA’s unproven NewDepomod computer model, which will be the mainstay of SEPA’s DZR system, if it is approved. 


Computer models can be set up to yield the results that best suit their users, and they are only as good as the site validation data they are given. NewDepomod is no doubt better than the current model (AutoDepomod) but at least that was peer-reviewed, unlike NewDepomod, and given that its developers list that it ’supports industry expansion’ as one of its three aims (SEPA FOI 188192 Nov 2017  ‘20160804_AB_REFINING SEA-BED PROCESS MODELS FOR AQUACULTURE Final Report v4’ pg 1), we feel that its opaque assumptions and its site validation data collection regime should be publicly scrutinised before it is used to predict the dispersion of pollution for the purpose of issuing any licences for expanded fish farms.


One of NewDepomod’s developers, who works for SEPA, agrees that the model fails to be able to predict where most of the waste ends up (because much of it will travel the 1km to the edge of its modelled box and then be ignored) and that a better model would probably be too computationally expensive. 

A lot more data is needed about the flow around the Scottish coast for these models to be of much value. The model cannot allow for less frequent but important resuspension events, such as storms. 

We have asked a professional user of similar computer models about NewDepomod. He said, ‘I would also like to have 3d turbulence spectra for many sites and depths. Even then I’m not sure the model would tell you much more than that the pollution is very widely and patchily dispersed.’


Surely the central questions that need to be answered before deciding how much more to pollute the sea are:

What weight of nutrient is it safe to dump into the coastal environment, and what effect will the patchy wide-spread deposition of sea louse treatments, nutrients and other fish farm waste have on the ecosystem?


Best wishes


Friends of the Sound of Jura

bottom of page