April 1st: The Scottish Government published their finalised river conservation gradings for the 2025 season on Tuesday 1 April. I thought that this might have been some form of April Fool’s joke, but it seems that it actually is a joke.
Firstly, the proposed river gradings were published on 7t August 2024. It has taken to 1 April 2025 to issue the finalised gradings, even though the fishing season has been underway since the 11 January.
It might be understandable that it has taken so much time to finalise the gradings if there were many changes, but the April publication simply states:
The River Leven (Dunbartonshire) is awarded a Grade 3 due to the presence of the Grade 3 Endrick Water SAC within the system.
I appreciate that it is important to get the conservation status right to safeguard the future of wild salmon but the bad news for the Scottish Government is that if the list they provided is correct then the River Leven has not changed its status at all. It is the River Clyde, not the River Leven that has been downgraded from grade 2 to grade 3. The River Clyde was one of the rivers that had been highlighted in August 2024 as having improved and no longer requiring mandatory catch and releases. It seems that this was not the case at all.
I am not sure if that statement is right because for the provisional gradings last Augst the Scottish Government changed the way they graded the rivers etc from grade 1, 2, and 3 to Good, Moderate and Poor. However, for the finalised gradings, the gradings have returned to a numbered system.
Unfortunately, the problems with these finalised gradings do not stop there. In August 2024, the Snizort Fishery District was divided into ten different assessment areas. One of the changes highlighted back in August was that Hinnisdal to Haultin had also improved and no longer required mandatory catch and release. The Hinnisdal to Haultin area was divided into three assessment areas – The River Haultin, The River Hinnisdal and the River Romesdal all of which were classified as Moderate (Grade 2).
In the finalised list, Snizort has eight, not ten assessment areas and this is because Hinnidsal to Haulton is now just listed as the River Hinnisdal. The River Haultin and Romesdal have disappeared from the list. Critically, the River Hinnisdal is no longer classified as a Grade 2 river but has improved to become a Grade 1, something the Scottish Government failed to mention.
It is interesting that there also happens to be a 2,000-tonne salmon farm located within 2-3 km of the river mouth,
However, the absurdity of the conservation grading assessments is illustrated by the Clyburn Fishery District in the Outer Hebrides. This is divided into nine assessment areas eight of which are classified as Poor (Grade 3) and one is Moderate (Grade 2). This is Laxadale Lochs, which also is very close to another salmon farm producing 2000 tonnes.
The point about the Clayburn Fishery District whose sea trout fishery collapsed long before the arrival of salmon farming is that in 2023, the last year for which catch data has been published, the total catch was 1 MSW large salmon, 2 ISW grilse, 19 sea trout and 13 finnock. The size of this catch hardly merits the district to be carved up into nine assessment areas. It also brings into question how Laxadale Lochs can be considered Moderate even if all the Clayburn fish were caught from that one fishery.
The problem of this system is that with west coast fishery districts being divided into so many assessment areas (143) and 114 being assessed as poor, the wild fish sector then claim that this is evidence of the negative impact of salmon farming as was used to instigate the 2018 REC inquiry.
The whole system is flawed. If wild salmon are so threatened, just make catch and release mandatory across all rivers and then there is no need for any such assessment. Killing fish is just no longer acceptable. It is no longer a joke
Appealing? The headline of a news story from the salmon angler’s representative organisation ‘Wild Fish’ states that the salmon farming industry has launched an attempt to undermine SEPA’s Sea Lice Risk Framework (SLRF). However, SEPA do not need any help in undermining the SLRF as they have already done the job themselves. They just are unable to consider that they have effectively done so by ignoring some basic evidence. The SLRF centres on the SEPA’s sea lice dispersal model, which predicts when and where infective sea lice larvae can be found in Scottish marine waters. Unfortunately, SEPA, backed by Marine Directorate scientists have such an implicit belief in their model, that they have never actually tested it in real life. The reason why they have not done this is simple, no-one has actually found any infective sea lice larvae in open water in anything but the most miniscule numbers. In 2021, the SPILLS project identified just 20 larvae over a period of seven months. The reason why they didn’t find any more is also simple – they are not there. As I have written previously, SEPA are stuck in Model Land, which is a far cry from the reality in which Scotland’s salmon farmers operate. The lack of any such evidence clearly contradicts SEPA’s claims that the SLRF provides an evidence-led approach. However, the main reason why SEPA are undermining their own regulation is their blanket refusal to discuss the science and the evidence. SEPA might have held two consultations, but it is clear they ignored any submissions that challenged their approach.
Whilst I have argued that the salmon farming industry should have challenged the SLRF regulation long ago, my understanding is that they have tried to work with SEPA but have now reached the point that SEPA have become so intractable that the salmon farming industry have been left with no choice but to take a stance. The Ferret has described salmon farmers as trying to sabotage the new lice controls, but this is far from the case. SEPA have varied the licence conditions without any basis and consequently, the industry has submitted appeals for all 210 farms. This could have been avoided had SEPA been willing to have open and meaningful discussions, but they have such inherent belief in their framework that they are not interested in doing so. Whilst others might claim otherwise, the reality is that there has never been a proper discussion over the science of sea lice and despite this challenge, SEPA and Marine Directorate scientists still refuse to engage in any proper scientific discussion. As I have pointed out previously, if they are confident in the scientific basis for this regulation, they should have no problem in defending it to others with a different view. Yet, it seems that irrespective of the appeals, SEPA appear intent on continuing down the same road. According to the Ferret, a spokesman for SEPA said that “Having been identified by Scottish Ministers as the lead body responsible for managing the risk to wild salmon and sea trout from sea lice from fish farms, SEPA will continue to implement the sea lice regulatory framework to ensure environmental protection from existing and new development”. However, Fish Farming Expert was told by the Scottish Government that the licence changes are frozen, and any changes will not take effect until the appeals are finally determined or withdrawn.
Despite what the critics such as Wild Fish say, my understanding is that the industry is still cooperating with some aspects of the Framework such as sentinel cage work. This is SEPA’s attempt to validate their model by measuring any infection pressure on fish held in small cages around the Scottish coast. Unfortunately, as SEPA continue to avoid discussion of the science of sea lice, they still will not accept that lice infection of sentinel fish is more likely to come from passing fish than infective sea lice larvae. Certainly, no scientist over the last thirty years has found these larvae in any sort of number that could cause wild fish any problems.
The most important point about the Sea Lice Risk Framework is that it results from the cumulation of many years of complaints from the wild fish sector that sea lice associated with salmon farms are responsible for causing harm to wild fish. In a blog on their website, Wild Fish point to the fact that the current regulation began with their 2015 petition to the Scottish Parliament in which they called for controls against sea lice to be strengthened. The problem with the petition is that none of the MSP’s in the Petition Committee or the REC Committee to whom it was referred questioned the ‘evidence’ that Salmon & Trout Conservation, as Wild Fish were then called, had included in their petition. It was just accepted that it must be correct. In fact, the evidence was actually extremely weak. Firstly, they claimed that the new conservation assessments showed that all fisheries on the west coast were in a poor state so that no fish could be exploited. In fact, this was a new assessment, and the Government had taken a cautious approach initially. Last year, 30 fisheries or rivers in and around salmon farms were judged to be of conservation status that could be exploited.
The scientific aspect of their claim came from a 2015 paper by Thorstad et. al. which estimated that between 12 and 29% fewer unprotected salmon as determined by smolt release trials were recaptured. Wild Fish translate this into ‘Mortality attributable to salmon lice can lead to an average of 12-29% fewer salmon spawners.’ The 12-29% figure is calculated from nine different papers. However, when the papers are investigated, they include Jackson’s 2013 paper that claims a mortality of about 1%, well below the figures used by Thorstad and Wild Fish. The papers also include a short communication, which was not peer reviewed from Martin Krkosek which criticised Jackson’s paper claiming that they had got the statistics wrong and then a subsequent paper from Jackson highlighting that it was Krkosek who was mistaken. This is Martin Krkosek who is a colleague of anti-salmon farm campaigner Alexandra Morton. I am not convinced that this figure holds water, but I am not a statistician. The way that many papers are written often make it difficult to conduct exactly the same analysis of the data.
Interestingly, I have been looking at a different paper, which is cited in the Scottish Government’s Summary of Sea Lice Science. This is by four Scottish Government scientists including Dr John Armstrong who took part in the Salmon Interactions Working Group. The paper is Middlemas et al. 2013 and is titled ‘Relationship between sea lice levels of sea trout and fish farm activity in Western Scotland.’
As a side issue, it is worth mentioning that the Sea Lice summary lists 41 references, five of which are not cited in the body of the document, and one is simply incorrect. I have informed the Marine Directorate more than once, but as I write this the reference as posted on the website remains incorrect. Seemingly the Marine Directorate don’t appear interested in ensuring the science related documents they post are accurate.
The Middlemas et. al. (2013) paper shows the relationship between sea lice infestations on wild fish. They found that sea lice infestation was influenced by the distance to the nearest farm up to a distance of 31 km, although the range using 95% confidence limits was 13 to 149 km. The authors concluded that there was link between salmon farms and sea lice burdens on sea trout in the west of Scotland.
I am not so convinced about their findings for the simple reason that sea trout do not necessarily remain in the location where they were sampled. They move around and thus a fish sampled at 2 km from a farm might end up 20 km away and vice versa.
The paper describes that sea trout were caught at a variety of locations during May to July of each year from 2003 to 2009. The results section states that a total of 6590 sea trout were sampled from 48 different sites. They add that 3462 (52.5%) were infested with sea lice, but they don’t specifically say that 3128 were free of any lice.
They continue that of the sampled sea trout 4867 were below 198 mm in length which they say means that they are of a size relevant to the Wells threshold (Wells et al. 2006). However, the accuracy of this threshold is a totally different discussion. Of these smaller fish 2273 (46.7%) were infested with lice. Again, the authors omit to highlight how many fish were free of any lice which was 2594 (53.3%) fish. Of these smaller fish 612 (12.6%) had lice levels above Wells threshold of 13 lice and thus are potentially at risk of damage. The authors also fail to mention that the risk is to 9.3% of all the fish caught.
As I have the full dataset of sea lice monitoring from 1997 to 2019, which I obtained under FOI, I am able to following the same process as the authors of the paper. I have checked and rechecked the numbers because there are so many different parameters. The full dataset contains the results of monitoring 21628 fish. Of these 8832 fish were sampled from 2003 to 2009 inclusive. However, the authors only used fish caught during May, June and July of these seven years and when fish caught between August one year and April the next are excluded, the remaining fish total 7359 in number. This is 769 fish more than appear in the Middlemas paper even though I have used the exact same dataset.
When the locations of the of the sites which were sampled, May-July 2003 to 2009 are totalled, the final number is 58, which is ten more than the 48 quoted in the Middlemas paper. The obvious question is what has happened to these 769 fish and ten sites? Why were they excluded from the Middlemas analysis?
I asked the Scottish Government under FOI for the list of 48 sites used in their study so am able to now identify which sites have been excluded. If I just look at just these sites, the number of fish sampled range from 1 fish to 160. When totalled, they add up to 661 fish which when subtracted from the 7359 I identified as meeting the requirement for inclusion in the study means that there are still 108 fish more than the 6590 used by Middlemas. The numbers simply do not add up. I have asked the Scottish Government for clarification. I am still waiting.
One of the excluded sites can be probably accounted for by a spelling error with Kannaird Wester Ross being also logged as Kanaird unknown. They are probably the same site, and it is a puzzle why Scottish Government scientists never bothered to check the list before allowing it to be released. There are other potential errors in site identification in the main list which have never been clarified but probably don’t affect this analysis.
Whilst Middlemas says 612 fish meet the Wells threshold; I make the number to be 835 fish out of 5193 smaller fish. This may be more than the number highlighted by Middlemas, but the point is that the analysis has not used all the available data, and it is unclear why. For example, I could see that some sites may be excluded because they sampled only one fish. but the Middlemas analysis sampled three sites with a sample size of one fish, whilst two of the sites that were not included also sampled just one fish. Equally, some sites were sampled over just one year, others two, others three and so on.
This is not the first time I have questioned the accuracy of the science coming out of the Marine Directorate. The Ives paper from last year also did not match the number of sites with the number of fish. Last week, I highlighted that provisional salmon catch numbers appear to exceed the final count. And the summary of science reference list is inaccurate. From my perspective, if the data is not accurate, how can anyone be sure that the analysis is too. The Marine Directorate say that the science of sea lice is settled. How can we be sure, when their output is riddled with inaccuracies.
It is no wonder that the salmon farming industry has had to appeal the conditions of the Sea Lice Risk Framework. The science on which it is built is not only inherently flawed, but also built on inaccurate data.