Analysis of disease outbreaks on salmon farms supplying Marks and Spencer

Independent experts investigate publicly available Scottish Sea Farm data

Ecohustler’s salmon petition and campaign is just a small part of a network of communities, scientists, business people and activists deeply concerned about the state of salmon farming in Scotland and the impacts it is having on marine ecosystems who are now taking action to regulate this industry.

In response to the 100,000 petition signers, Steve McLean (Head of Agriculture & Fisheries Sourcing at M&S Food Group) sent an email to Ecohustler essentially maintaining that there is no problem with salmon farming and that the company - Scottish Sea Farms - supplying Marks and Spencer “delivers high welfare and environmental standards of which we are proud.”

Ecohustler invited experts within the wider network to contribute in responding to this claim. Using publicly available data and building on analysis carried out by Inside Scottish Salmon Feedlots a completely different story emerges. It is clear that just like other salmon factory farms the farms supplying M&S are destroying Scotland's marine ecosystems in three main ways -

  • Parasites from feedlots are killing wild salmon
  • Salmon farm sewage is dumped directly into sea lochs harming other marine life
  • Plastic and chemicals are polluting the coastline

Image courtesy of IISF

Using the data from the Scottish Government’s Fish Health Inspectorate (FHI) website, reporting all mortality events above certain threshold levels between Feb 2017 and March 2021.

The acronyms mean:

The most striking things are:

The losses from SSF’s freshwater and marine sites that are reported here are a minimum of 1.46 million fish.

This is a minimum figure because of the 280 events at SSF farms that exceeded the FHI’s reporting threshold (therefore serious events), 169 report the number of fish killed as ‘not provided’. That is 60% of the total.

1.46 million fish died in the 111 events for which figures are provided. If we estimate the total number of fish killed by scaling that figure up in proportion to the number of reported incidents, the real total might be as high as 3.69 million

Image courtesy of IISF - a diseased animal in a scottish salmon farm_

Several farms including Inner Mangaster, Kempie Bay, Kerrera B, Fishnish B, Dunstaffnage, Bring Head, Sian Bay, Tanera, Walters (Lismore) and Wyre reported a large number of events without giving any figures at all.

Under these circumstances, how can M&S claim to know what is happening on these farms and be happy that this is responsible behaviour?

Feb 2017 and March 2021, the farms with the worst reported mortality (number of animals dying in the cages) according to the figures given were:

  • Bloody Bay - 68,463
  • Drury Voe - 50,003
  • Kisholm A (South) - 62,642
  • Kisholm B (North) - 109,204
  • Kisholm West - 97,500
  • (most deaths at the last two farms attributed to hydrogen peroxide treatments)
  • Lismore N - 30,187
  • Loch Creran B - 29,894
  • Loch Creran D - 29,894
  • (odd isn’t it?)
  • Nevis C -110,854
  • Snaraness - 163,523 (gill issues - fish to be incinerated - smacks of some nasty infectious disease)
  • South Sound - 33,190
  • Loch Damph (freshwater) - 176,594 young fish died when transferred to sea

Image courtesy of IISF

How can this be consistent with M&S’s position?

Many of these events mention ’post-treatment’ as at least a partial cause. Many of the treatments named are physical sea lice treatments which maim many fish.

This is poor husbandry, not an environmental cause or death caused directly by disease.

Norwegian reports show that salmon feel pain when warmed up in thermolicer machines, a physical treatment commonly used by SSF.

SSF has recently been promoting its newly built thermolicer vessel the Kallista Helen

The 68,463 fish that died at Bloody Bay were in a series of events caused by hydrogen peroxide treatment. Hydrogen peroxide is bleach. It is used to treat sea lice and gill disease.

Imagine the suffering of fish whose sensitive gills were exposed to bleach.

Some events reported to the FH show the cause of death as due to strong tides. Tides are entirely predictable. These farms are not correctly sited if that can happen.

The official data on sea lice, chemical use and percentage mortality at all SSF farms during the years 2018-2019

As you would expect there is a mixed picture.

The farms marked in grey in this list were empty (fallow) during the 2018-19 period. Orkney farms marked pale yellow - they often have the best sea lice results. No one knows why.

Sea lice

I have highlighted all lice figures that were above the Scottish Government’s reporting threshold of 2 adult female (AF) lice per fish, and above 0.5 AF lice per fish between February and the end of June.

The industry’s voluntary Code of Good Practice says that farms should have no more than 0.5 AF lice per fish between 1st February and 30th June inclusively as that is when wild salmon smolts are running.

Sea trout are present all year in coastal waters and therefore at risk from high lice levels at all times of year.

The Aquaculture Stewardship Council has set a much lower level (0.1 AF lice per fish) for the farms it certifies, during wild smolt migration time. This is more like a responsible level. No SSF farms are ASC certified.

The North Atlantic Salmon Conservation Organisation (NASCO), to which the UK belongs, says that fish farms should not increase the natural level of sea lice in the water.

SSF’s farms are clearly very far from protecting wild salmon from the effects of the sea lice released from their fish farms.

Marine Scotland Science, in its statutory advice letters to planning authorities, is clear that even when farms are managed within the industry’s CoGP lice levels, they still release significant numbers of lice larvae into the sea. Its usual wording is that ‘adherence to the suggested criteria for treatment of sea lice stipulated in the industry CoGP may not necessarily prevent release of substantial numbers of lice from aquaculture installations.'

MSS now accepts that sea lice from fish farms can affect populations of wild salmon and sea trout in Scotland: ‘The body of scientific information indicates that there is a risk that sea lice from aquaculture facilities negatively affect populations of salmon and sea trout on the west coast of Scotland.'

Based on the two years 2018-2019, the SSF farms with the fewest sea lice are mostly in Orkney. They had some really high lice numbers on some of their other farms in this period.

The following farms exceeded 0.5 adult female lice per farmed fish in April - May:

Bellister (Shetland) - M&S/SSF may say there are no breeding salmon in Shetland. There are breeding sea trout.

(This farm and many others show a surge in mortality tonnages and monthly percentage of biomass dying towards the end of each farm’s 20 month production cycle. I have marked these events in red as well. Presumably many of the mortality events reported to FHI happen during this period, when sea lice are getting out of control and pesticide and physical treatments are killing the fish, many of which have already been weakened by disease, according to the FHI reports) - and Dunstaffnage, Fishnish B, Fiunary, Grunna Voe (Shetland), Holms Geo (Shetland), Camas Doun Point, Allta a Chois, Kishorn Outer, Walters (Lismore East), Dubh Sgeir (Lismore N), Port nan Ledaig (Lismore B), Creran A, Loura Voe (Shetland), South Sound Mangaster Voe (Shetland), Ardlintigh (Nevis C), Scallastle, Slocka (Shetland), Tanera 1, Tanera 2, Teisit Geo (Shetland) and Vidlin N.

There are some really shocking monthly mortality percentages in this document, also highlighted in red

The same document shows emamectin benzoate usage which is regular on many farms. I have marked them in red.

SEPA imposed a much lower interim level of use for this substance several years ago because SEPA found its use to be correlated with declines in the number and diversity of crustaceans, but this only applied to new and expanded farms. Farms with existing licences have continued to use it because they are allowed to, despite knowing that it causes harm. The new EQS levels for emamectin benzoate have been delayed for years.

Azamethiphos and deltamethrin pesticides were hardly used by SSF during this time, although the mortality reports submitted to FHI do show deaths due to azamethiphos treatments later than this. Hydrogen peroxide treatments are not included in the 2018-2019 data.

Using the latest weekly sea lice counts for all active Scottish salmon farms.

The red highlighting between 1st February to 30th June shows all the farms that have exceeded the 0.5 industry Code of Good Practice level for adult female lice per farmed fish during that period.

From 1st July the CoGP maximum acceptable level is 1.0 Af lice per fish. After 1st July ≥0.5 AF lice is shown in yellow, and ≥1.0 in red.

As an example, of the active farms during week beginning 7th June, 12 farms reported figures exceeded the CoGP lice level, while 15 did not exceed that level. That’s 44% of SFF farms failing to protect wild fish during that week.

Below are the farm survey classifications (coloured shapes) superimposed on each SSF’s farm’s biomass graphs. Some farms seem to be missing from this database. Possibly they are know by more than one name.

The data comes from Scotland’s Environment website (This does not work in Chrome or Safari but does with Firefox). These classifications are based on the results of the seabed samples conducted every two years by the farm’s operators.

This document shows the farms whose results are not good or are unclear because the last satisfactory or borderline rating (also a pass) was several years ago.

Some farms show reductions in peak biomass in production cycles following poor environment performance, ideally then followed by a ’satisfactory’ survey result. This is what ought to happen, to protect life on the seabed, but it does not happen at all farms.

In many case the survey data has not been analysed by SEPA, often several years after it has been collected. These are marked in black as ‘to be evaluated’.

Examples of this and failing farms include **Bellister, Bring Head, Fishnish A, Fishnish B, Puldrite Bay, Lippie Geo and Vidlin North. **

Why are these farms supplying M&S?

Bloody Bay was borderline in 2018 then ’to be evaluated’ next time. Shuna failed in 2018 and the latest survey (2019) is still ’to be evaluated’ two years on.

Some farms such as Dunstaffnage, Charlottes Bay, Achnacroish (Walters), Scallastle and Tanera 1 and 2 might be OK, but their surveys have not been assessed for three or four years.

Kempie Bay may be OK, but its abrupt ’satisfactory’ grade after two borderlines seems a little unlikely. It has not reduced its biomass. Teisti Geo is similar - what changed?

How can M&S claim to know that the SSF farms that supply them are complying to SEPA’s seabed environmental standards, when SEPA has not classified many of these farms for such a long time?

This is the same type of data for better performing SSF farms. Many of these have not been assessed for several years.

Separate analysis of sea lice data for SSF farms completing their production cycle in 2020, based on published figures, show that SSF’s worst farm in this category was Vidlin. It had sea lice figures >0.5 for 11 months of its 18 month cycle, >1.0 for 10 months and >2.0 for 8 months. This was the third worst performance of all the farms in Scotland that completed their cycle in 2020.

Next worse SSF farms were Mangaster (7 months >0.5, 6 months >2.0), Fishnish A (13 months >0.5, 7mo >1.0), Scallastle (12mo >0.5, 3 >1.0), Fishnish B (6mo >0.5, 5mo >1.0), Fiunary (6mo >0.5), and Slocka Ronas Voe (8mo >0.5, 4 mo>1.0, 2mo >2.0).

Overall mortality

Lismore East had 36.6% overall mortality during that production cycle. Mangaster had 27.2%, Vidlin had 17.3%.

Some SSF farms had lower mortality, which lowers the average to 13.6% over the two years, in the 14 SSF farms completing their production cycle in 2020.

Tanera had the lowest mortality, over the two years, totalling 3.6% overall.