Model Land again: Following my account last week of finding myself trapped in Model Land whilst I attended a SEPA online workshop, I then had a second encounter with Model Land but this time in person. The experience was equally frustrating because those in Model Land are not interested in what exists outside their closely contained world.
Model Land does exist, and it was to be found a couple of weeks ago in a special session of the latest Marine Alliance for Science and Technology for Scotland (MASTS) conference that took place in Glasgow at the beginning of November. The special session was on Sea Lice Surveillance and Modelling and was led by scientists from the Marine Directorate and the Scottish Association for Marine Science (SAMS). Apparently, they had decided that as Sea Lice 2024 was no longer to take place, then a session at the MASTS conference was a good substitute. Unfortunately, they didn’t really tell anyone about it except those signed up to receive MASTS newsletters or were in the Model Land club. Since Marine Directorate scientists have led the narrative about sea lice dispersal through the Salmon Interactions Working Group and now the Sea Lice Risk Framework, it is not surprising that rather than a wider programme of sea lice research and discussion, the programme was restricted to modelling and the continued attempts to validate the models through surveillance.
The session consisted of seven presentations all from Marine Directorate, their former employees and SAMS. All the presenters were connected in some way or another, so the programme has a rather incestuous feel with the session becoming something of a Modellers Love-In.
I am only going to discuss one of the presentations which was delivered by Dr Alexander Murray from the Marine Directorate and was titled –
Optimizing policy advice from sea lice dispersal models: the Knowledge Strengths approach.
This single presentation encompassed all that is wrong with the sea lice narrative and why the salmon farming industry is now caught up in the monster that is the Sea Lice Risk Framework. This is because the policy advice given to Government comes from what is effectively a single blinkered source that is too reliant on theory without the backing of any conclusive evidence. Surely, policy advice should be based on a range of factors and not just the model. It’s not surprising that the Marine Directorate refuse to discuss real evidence because clearly it doesn’t fit in with their narrative, or the advice they give.
I was once asked who I thought represented the greatest threat to the salmon farming industry in Scotland from someone who expected me to say anglers or activists, but my reply was Marine Directorate Science, and I still believe this to be true. They are the ones providing advice behind the scenes to Scottish Ministers, yet they refuse to discuss anything other than their models and their belief, encouraged by the angling fraternity, that salmon farming is a threat to wild fish. As I highlighted in the last issue of reLAKSation, a letter sent on behalf of the Cabinet Secretary began ‘Sea lice are internationally recognised as posing a risk to wild Atlantic salmon and Sea trout.’ This is clearly based on advice the Cabinet Secretary received. However, such international recognition only comes from a small clique of scientists whose research is at the centre of such recognition but more importantly, refuse to consider any evidence that contradicts their own research. As I keep saying, if they are so confident about their narrative, then they should have no problem in explaining why the counter evidence is invalid. Instead, there is just a wall of silence. Being in the same room as some of these scientists didn’t help improve the dialogue.
Returning to the presentation about optimizing policy, it was explained that a working group was formed consisting of research groups, Marine Directorate, the Institute of Marine Research and Firum of the Faroes and most surprising SEPA. This gives the impression that these groups have colluded to influence governments to act against the salmon farming industry based on their models, none of which have been properly validated or shown to actually reflect what is happening in the sea in real life. This is the world of Model Land, which these organisations appear to inhabit.
The presentation ended with a conclusion that models and outputs must be transparent to external scientists. There must be no black holes. Well, I am one external scientist for whom neither the models or outputs can be described as transparent but then I don’t live and breathe in Model Land. Sadly, Dr Murray failed to give any examples of such transparency, probably because in my opinion, the models are only transparent to those external scientists who are willing to be part of Model Land.
Surveillance: The last day of the MASTS conference featured a number of workshops including one on sea lice surveillance organised by SAMS. Regular readers may remember that as part of the 2021 SPILLS project, SAMS researchers attempted to find sea lice larvae in the water column but only managed to identify 20 larvae from 372 samplings. The SPILLS reports stated that just because they couldn’t find any, doesn’t mean that they are not there. I disagree, which is why they can’t find the numbers as predicted by the various models.
Of course, and just for clarification, when I say that the sea lice are not there, I mean that they are not there as predicted by the models. Sea water is full of plankton and there will be some sea lice larvae present, but as the 1996 Irish study showed, they will be in very low numbers.
As you can see from the image of the model taken from the SPILLS website, copepodid sea lice concentrations of 10 lice larvae /m3 are predicted (in yellow). Even at these concentrations, they are not present as a soup, cloud, or curtain, as the anti-salmon farming critics claim. Yet the first presentation of workshop began stating that sea lice are very rare with concentrations of 0.1-1.0/m3. A concentration of 0.1-1.0/3 is well short of the 10 lice/m3 concentration shown in the model.
This summer SAMS undertook a new survey of lice in Loch Linnhe over three days. They pumped water from 1m and 12m and sieved them through 150 microns to collect any lice larvae. The findings were that 17 out of the samples were found to contain larvae with numbers of lice in a sample ranging from 1-12 giving a density of 0.09-1.05 lice/m3.
SAMS were surprised to find that 3 nauplii and 10 copepods were retrieved from the 1m depth sample whist 32 nauplii and 11 copepods were found at 12m depth. Yet, this isn’t really surprising as Loch Linnhe has different flows at different depths, and it is likely that the lice at depth were simply being swept along and not searching out a host as the established narrative would suggest.
A later presentation provided some further references to sea lice concentrations in open water. Under the title ’The Needle and the Haystack’, examples from published data indicated concentrations of 0.075 copepodids/m3 (2013) 0.29/m3 (2011) and 0.014/m3 unpublished.
Not one of these findings would validate any model, yet after thirty years of searching for sea lice larvae in the sea, the idea that the models are correct remains the basis of current sea lice thinking. How long will it be before there is some acceptance that the particle dispersal models are purely theoretical and that the sea lice are not where they are expected to be. Perhaps, if the modellers had begun with a real-life organism rather than using an arbitrary particle, we might be in a very different place today?
Validation: The final presentation of the workshop was from SEPA. This came as a surprise to me because regular readers will know the difficulty, I have encountered trying to speak to SEPA face to face. Anyway, more about that later.
The workshop was told that SEPA had selected sentinel cages as a way of validating the model because the SPILLS project had recommended using this method. Yet, if I remember correctly, the SPILLS project did not even deploy one sentinel cage themselves. They relied on data generated between 2011 and 2013. If sentinel cage is considered so valuable, then why did SPILLS not deploy them as part of the project? The Loch Linnhe data is now at least 12 years old. I presume that the age of the data was considered irrelevant as the narrative it supposedly supports is even older.
Work Package 4 of the final SPILLS report states:
Sentinel fish lice count data provided data which could be compared with confidence to modelled infective pressure (lice days m–2) over the deployment period. Data collection or model validation should be focused in late summer and Autumn when lice numbers tend to be highest
Over the three-year deployment period, the lice count during the spring period were very low and it was only during the autumn that lice counts were noticeable peaking in 2013. This is why SPILLS says data collection for model validation should be carried out in the autumn period.
This is where I have a major problem. The whole point of the Framework is to essentially to protect wild salmon as they migrate from river to their marine feeding grounds. This happens in the spring. This is when the model should be validated not the autumn. The rest of the year is totally irrelevant for this purpose. This is such a clear flaw in the Framework and the associated narrative it is surprising that those who endorsed the final report did not express a similar concern. However, it is only necessary to look at the makeup of the steering group to see that the SPILLS project lacks impartiality. The Steering group consists of Alan Wells of Fisheries Management Scotland, Peter Pollard of SEPA, John Armstrong of Marine Directorate and Alex Adrian of Crown Estate. Where is the representation from the independent scientific community and/or the industry? But then the coordinating institution was Marine Scotland Science, an organisation is advising policy makers based on their own research. Like SEPA, FMS, the Marine Directorate continue to avoid any discussion on the possibility that they may be wrong.
I discussed sentinel cages in the last issue of reLAKSation but a couple more points are worth mentioning. I would like to use the analogy of crossing a road. It is possible to cross a road with light traffic repeatedly without getting hit but if someone was to stand in the middle of the roadway and remain there for a week, the chances of being hit increase significantly. Wild salmon do not remain in one place for weeks on end without moving so keeping a fish in one cubic metre for a long period of time is not representative in any way. It now turns out that deploying a sentinel cage is not as straightforward as might be implied. Seemingly, sentinel cages deployed in the open water require a licence for a variety of reasons. However, salmon farms can deploy sentinel cages in their immediate vicinity because they are already licenced or have the relevant permissions to hold fish in cages. I may be wrong because I am not included in the discussion, but farms might be encouraged to deploy the cages around the farm as part of SEPAs validation rather than deploying cages in areas well away from farms. This would be a waste of time because we know that if the farm has lice, then lice larvae can be found in concentration in the surrounding waters. It is elsewhere that is the issue.
There really does need to be a proper discussion about the Sea Lice Risk Framework which is not about the model or modelling but about whether the evidence shows sea lice do or do not represent a risk to wild salmon.
I have seen with my own eyes that representatives of the Marine Directorate and SEPA can attend a sea lice workshop. I did ask the representative of SEPA whether SEPA would be willing to attend a similar workshop to discuss the wider issues. The answer was yes in principle. SAIC organised this sea lice surveillance workshop, which was part of an event promoted as an alternative to Sea Lice 2024. Surely SAIC should be able to organise a similar workshop for scientists from industry, SEPA and the Marine Directorate to help understand the whys and wherefores of the Sea Lice Risk Framework that exist outside that blinkered world of Model Land. Surely, these organisations can spare one day to ensure that what they are planning actually ‘does what it says on the tin.’ and that everybody is on board.