Scottish Model Protection Agency: Fisheries Management Scotland (FMS) held their annual conference at the end of March but given some of the content it might have been better scheduled for 1st of April!
One of the speakers was SEPA’s Head of Ecology, Peter Pollard who spoke about SEPA’s sea lice risk framework, a new form of regulation to protect wild fish. Peter had previously told the Scottish Parliament’s REC committee that sea lice associated with salmon farms were not responsible for the decline of wild fish numbers along the west coast. Clearly, if sea lice associated with salmon farms are not responsible for the decline of wild fish, then SEPA’s risk framework is totally unnecessary. There is planet of evidence to support Peter’s statement that sea lice and salmon farming is not the reason why wild salmon stocks are in such a dire state, but SEPA persist with the implementation of this framework because they say they have been told to by Scottish Ministers. SEPA are not interested in hearing the evidence but more about this later.
Peter told the conference that the framework is focussed on the pest – the infective stage sea louse – and what they are trying to do is manage concentrations of this louse in the environment so they don’t reach levels that will lead to harmful numbers of lice on the salmon as they leave the rivers and head off to the high seas.
He continues: ‘The framework is based on assessing the risks to post smolts as they navigate their way through the wild salmon protection zones. How that works is we use modelling to model the concentration of infective stage sea lice in these zones and then we model the movement of ‘virtual’ salmon smolts from the rivers leading to these protection zones. We might run several thousand ‘virtual’ smolts and then we explore the ‘potential’ exposure they experience to assess the risk. It is a risk framework based on modelling – quite sophisticated modelling – which factors in the concentration of lice and the movement of the fish through the zone’.
What Peter is saying is that they have developed a hypothetical model of the interaction between sea lice and wild fish but what Peter cannot say is whether the model bears any resemblance to what actually happens to sea lice and wild salmon in Scotland’s sea lochs. SEPA do not appear to have made any attempt to validate the model which would demonstrate that the model’s predictions are borne out in the sea.
SEPA’s risk assessment model is simply a computer model that has not been validated by real data. It is nothing more than a very large and expensive computer game, which SEPA aim to play, but which they will do at the expense of the salmon farming industry.
What Peter and his colleagues at SEPA, as well as scientists at the Marine Directorate, are reluctant to admit is that whilst they say that they will be managing concentrations of sea lice in the environment, they haven’t a clue as to what the actual concentrations of sea lice in the sea lochs actually are. They therefore have no idea as to whether the regulation they will impose on the salmon farming industry will have any impact on protecting wild fish or not. Crucially, Peter doesn’t appear willing to discuss this possibility, let alone that the assumptions on which SEPA have based their model are wrong.
I was rather surprised to see that Peter had made his presentation to the FMS conference in person, especially as the Cabinet Secretary had not. I had previously been told by SEPA management that I should meet with Peter to express my concerns, and I requested a face-to-face meeting at a location of his choice. He refused saying that we could meet remotely. However, I am not convinced that this is the best method to talk about important and complex issues. Peter told me that he was busy and speaking remotely would be the best use of his time, although I am unclear as to how an hour of speaking remotely is any different to a hour of face-to-face unless one of the participants is busy doing something else off screen at the same time, something which happens all too often and is probably why SEPA have previously failed to understand the major shortfalls of their approach.
I have recently contacted Peter again to ask if he had reconsidered my request. His answer was that
For efficiency reasons, our normal practice is to host such discussions via MS Teams.
Yet it doesn’t seem a very efficient way to spend a day travelling to Aviemore to give a 15-minute presentation to the angling fraternity, which has done nothing to allay any of my concerns about the SEPA risk framework.
The biggest flaw in the modelling is that from some of the earliest research into sea lice, no-one has yet found sea lice larvae in the concentrations predicted by various sea lice dispersal models. Researchers from the 2021 SPILLS project wrote that just because they didn’t find any sea lice larvae (20 in 372 samples) doesn’t mean they are not there.
A 2017 paper from the Marine Directorate and the University of Aberdeen reported sampling plankton at Loch Sheldaig every week for over ten years (500 plus samples) and identified 434 sea lice copepodids. Ten years seems to be a long period of time to be sampling and still not to be persuaded that larval sea lice are not present in the numbers predicted by the models.
The fact that no-one has actually found the concentrations of larval sea lice that SEPA aim to manage does not appear to have deterred SEPA from progressing on with the implementation of this framework.
Rather than instigate some research to find these lice to validate their model, SEPA have adopted a different approach, the mainstay of which is to use sentinel cages as a method of validation. Peter explained to the conference about sentinel cages saying ‘that is putting out farmed smolts into pens in the sea to make sure the models we’ve done are reflecting what the environment is experiencing. It allows us to test the skill of the model against real data in the environment from the lice that are picked up on smolts held in these fixed cages for short periods. It gives the confidence in the models and the confidence is really important to enable us to take decisions.’
As the first sentinel cages are planned to be deployed in October, he said that ‘It doesn’t matter it is outside the migration window, it’s still going to tell us if the model reflects lice in these areas.’
Sentinel cages are such a good way of measuring sea lice infestation pressure that none have been deployed in Scotland since 2013. The SPILLS project team were clearly so impressed with this method that this 2021 project simply took the data from 2011 to 2013 to input into their models. Interestingly, whilst Peter said that using sentinel cages outside the migration window will still offer an idea of lice infestation, not one of the 2011-2013 study sentinel cages recorded an impactful lice infestation during the critical migration window in any of the three years.
By comparison, higher infestation was recorded during the autumn which coincided with the return of adult fish with the naturally occurring lice infestation so highly prized by anglers.
SEPA assume that any lice infestation on fish contained in sentinel pens must have come from a salmon farm. Yet, sea lice have been naturally infesting fish for millennia and it is more than possible that sentinel fish could be infested from passing fish.
From 1987, the farm in Loch Ewe was blamed on spreading sea lice that caused the collapse of the Loch Maree sea trout fishery. In their paper of 2006, Butler and Walker state that the nearest active farms were 40km by sea to the north in Little Loch Broom and 55 km to the south the Loch Torridon. It is unlikely that clouds of infective sea lice larvae travelled such distances to Loch Ewe to infest the new uninfected farm so the lice must have come from somewhere else and that somewhere else is from naturally occurring lice already infesting fish. At the same, long-term data from the Scottish Coastal Observatory in Loch Ewe shows very low larval lice concentrations even during the time the farm was operational.
In his presentation, Peter said that SEPA also were working internationally, as they have got an unspecified grant funding arrangement with the Institute of Marine Research in Norway, and they will be working alongside SEPA on the sentinel cage work. It would be interesting to hear from Peter Pollard why he has refused a face-to-face meeting to discuss issues such as sentinel cages whilst at the same time SEPA are more than happy to pay for advice and collaboration from overseas. It is unclear what IMR scientists can bring to the table that can’t already be found in Scotland. What is clear is that IMR has taken a very different approach to sentinel cages than in Scotland. Sentinel cages in Norway are often deployed for a three-week period. How a three-week deployment can represent an infestation pressure on migrating salmon smolts that will typically clear the farming areas in one or two or days is a mystery. At the same time, the young smolts will be increasingly stressed due to the lack of food and succumb to secondary infestation, that they might normally withstand. The Scottish sentinel cage trial lost significant numbers of fish during the deployment, something which was rarely discussed.
It is clear from the latest Traffic Light System assessment that sentinel cages are no longer considered an important part of the assessment process, which begs the question why SEPA are placing so much emphasis on their use.
Interestingly, Peter relates how juvenile stocks from river systems with predicted high lice loads will be compared with stocks from systems with low predicted lice loads, a process which he calls pairing. Surely, if sentinel cages are to be deployed, they should also be deployed in areas with no farms as a control. Loch Ewe might be a candidate location.
Finally, Peter talks about some of the monitoring work and said that he is hoping, very much hoping that Fisheries Management Scotland and some of their members will be really keen to be involved in the steering and delivery groups. When this was first announced, I wrote to SEPA to request involvement in these groups but did not receive even the courtesy of a reply. Clearly, they are more interested in people from the wild fish sector who will be delighted to support SEPA’s efforts to protect their sport rather than have a qualified scientist involved who will question everything they try to do, and rightly so.
The real question about protecting wild salmon is whether this should be done from a theoretical, so-called sophisticated model world or from one where real action is taken. Unfortunately, the Salmon Interaction Working Group which decided that wild salmon needed protection from sea lice even though there is no evidence that sea lice are responsible for the current state of wild salmon stocks, were more interested in penalising the salmon farming industry rather than protecting wild salmon. SEPA appear more than happy to act on their behalf.
Dr John Armstrong from the Marine Directorate also spoke at the Fisheries Management Scotland conference about the state of wild salmon stocks. As you will read next time, the Wild Salmon Strategy’s Science and Evidence Group’s action plan as explained by Dr Armstrong is unlikely to make a real difference. One of their priorities is to issue guidance about placing large woody structures in Scottish rivers…..
The presentation can be viewed at: