Scroll Top

reLAKSation no 1089

SEPA science: Before I start this commentary, I would remind readers that SEPA told the Scottish Parliamentary committee that sea lice from salmon farms were not responsible for the decline of wild fish on Scotland’s west coast. Against this background, SEPA have summarised the responses as being highly polarised. This was most apparent in relation to the science with the angling fraternity claiming that there is substantial evidence of harm being caused to wild salmonids, whilst the farming sector argue that there is insufficient evidence that sea lice from salmon farms have a significant impact on wild salmon populations.

Even though SEPA have said that sea lice from salmon farms are not responsible for wild fish declines, their policy appears to favour those who claim that salmon farms do damage wild fish stocks backed by Marine Scotland Science’s Summary of Science. However, despite saying that the proposed framework will not be implemented until the end of 2023, they have stated that they will not revisit the discussions on the scientific basis of the decision to introduce a regulatory control regime. As I have mentioned previously in relation to Marine Scotland Science, if SEPA, or Marine Scotland Science, who helped SEPA with the science responses to the consultation, are so sure of the science, why would they not be willing to spend just one day out of the next fifteen months to explain the scientific justification for the framework?

Whilst SEPA say they will not revisit the science; it is unclear whether they ever visited the subject at all and instead simply accepted the Marine Scotland Science view as expressed in their Summary of Science. I have been told that this summary is simply for information only and is not used to help form Government policy. When asked what science is used to form Government policy, I was referred to the Salmon Interactions Working Group (SIWG).

Regular readers may remember that I was critical of this group from the outset. Firstly, the salmon farming contingent was outnumbered by those who could be said to be opposing the industry and secondly, because the group refused to accept submissions, as they said they were not an enquiry seeking evidence. Yet, how could the group be expected to make the correct decisions if they were not in receipt of all the evidence? Judging from the limited meeting minutes available, one of the few scientific papers discussed was the Vøllestad paper that also appears in the summary of the science which states:

“Catches of wild salmon after the late 1980s declined on the Scottish Atlantic coast relative to elsewhere (Vøllestad et al. 2009). This area covers the majority of mainland Scotland’s salmon farms although the authors stressed that this did not prove a causative link with aquaculture.”

Surely, if the authors stressed that the observations made did not prove a causative link then why was it even discussed in SWIG and why is it included in the Summary of Science?  In addition, as I have discussed previously, the paper is flawed in other ways, simply because it is not about sea lice at all but if Marine Science Scotland wanted to discuss the paper with me, I would be more than happy to do so.

Finally with regard to SWIG and the science, I was informed more recently that the science was never discussed by the group because it was considered too contentious a subject for the group.

Irrespective of any discussion of the science, the group agreed a joint position during their meeting of 28th November 2018:

to acknowledge the potential hazard that farmed salmonid aquaculture presents to wild salmonids’

This statement is often bandied about by the wild fish sector as acknowledgement by the salmon farming industry that they do have an impact on wild fish.  Of course, there is an impact, just as anglers have an impact or any of the other identified pressures have an impact. The fundamental issue is the scale of the impact and I, and as I believe the salmon industry will argue too, that this impact is very small whereas the wild sector claim that it is extremely large. The only way of identifying the size of this impact is through the science, which is why discussion about the science is so important and necessary.

In their summary of the science, Marine Scotland Science state under experimental studies that:

“the reduction in numbers of returning salmon associated with lice in untreated smolts is in the range of 0-39%”

This is a big range and includes 0% or no impact to 39%, a figure often quoted by the wild fish sector. However, Marine Scotland Science has not managed to narrow down the impact or identify which is a reflection of the true impact. It would appear that they are sitting on the fence by offering a range of two extremes and they are supposed to be the scientific advisors!!!

What I do know is that a figure of 1% mortality has been derived from large scale experiments in Ireland, and yet this significant paper is not cited in the summary of science. By comparison. the top end figure of 39% has been derived from a statistical analysis of other research data and thus is not really an experimental study at all.

Surely, before any decision is made as to the appropriate action required, the actual scale of any impact must be identified. At the moment, it seems that measures are being imposed that bear no relation to the scale of the impact identified in the summary of science.

SEPA say in their response that they will consider the latest science before deciding on the appropriate risk threshold to use, so why won’t they consider all the ‘new’ science that isn’t included in Marine Scotland’s Summary of Science. One example of this follows this commentary.

Whatever anyone else might suggest, the science in relation to the risk-based framework is key.


One example: A recent bog by Tanglewest Douglas of Wild Fish Conservation (Salmon & Trout Conservation) considers the implications of warming temperatures on sea lice infestation. The blog cites four scientific references but none of these relate to the narrative that “juvenile stages of lice float in the water current to disperse so they move freely out in the sea lochs where Atlantic salmon and sea trout live.” Ignoring the fact that Atlantic salmon don’t actually live out in the sea lochs but pass quickly through them on their migration to their marine feeding grounds, there is little scientific evidence to support the claim that the larval lice disperse on the water currents. Other commentators from the wild fish sector talk about a ‘soup of lice’ or ‘clouds of lice’ yet cannot supply any proof of such observations.

The risk-based framework to be introduced by SEPA on advice from Marine Scotland Science is built around the assumption that larval sea lice spread through the ocean and infest wild fish along the way. The MSS summary of science states that:

“Infection of wild salmonid smolts by sea lice emanating from salmon farms depends on environmental dispersal patterns and behaviours of sea lice and fish. The processes involved and associated best available biological data are combined in the model presented by Murray & Moriarty (2021)”.

During their recent workshops, SEPA demonstrated their new model showing how plumes of larval sea lice might be expected to spread from a farm as illustrated by the purple area shown below:

However, the paper makes no mention of validating the model by conducting trawls of sea lochs to measure the numbers of sea lice larval present in the water column. An FOI request revealed that such validation was not part of the work so currently, we don’t actually know whether a soup of sea lice larvae can be found in this plume or whether it is just a model of how any inert particle might be expected to travel through the sea lochs, rather than living sea lice.

Interestingly, after I had written this commentary, I became aware that BBC Radio 4 had featured the SEPA sea lice risk framework in their programming. Peter Pollard from SEPA told Farming Today that “one of the things that we really need to do is develop the models to understand where sea lice are going.” However, that is a very blinkered approach. What we need to do is understand what is happening to sea lice in the marine environment and then develop the models accordingly. We could release hundreds of plastic ducks, for example,  around a salmon farm and then observe where they end up as a way of developing a model, but this does not mean the same patterns will happen to sea lice. Sea lice are biological organisms, not inert particles. It cannot be assumed that they will behave in exactly the same way.

The only actual experimental work referred to in this modelling paper dates to 2004 and was work carried out by the then Fisheries Research Services at the Shieldaig Field Station. In a nutshell, the researchers found sea lice nauplii close to the farm but ‘occurred less elsewhere’. It would seem that this description was a polite way of saying that they found hardly any other than at one specific location and this was where the River Shieldaig empties into the head of the loch. This is not surprising given that the natural infestation of lice on wild fish is believed to occur at river mouths when any migrating fish slow down to adjust from freshwater to salt and are more easily targeted. The area between the farm and the head of the loch was mostly free of lice.

These findings confirm the work of Costelloe et al. (1998) who only found larval sea lice consistently around a salmon farm in Killary Harbour in the west of Ireland. Elsewhere, larvae were recovered only sporadically. The paper also states that similar sporadic occurrences were recorded in two estuaries in which there were no salmon farms.

A more recent paper was presented at Sea Lice 2016 which was attended by representatives of Marine Scotland Science. This is the work of Nelson at. al. (2017). This paper also identified high larval sea lice concentrations around the farm, but this declined significantly within 100 metres of the pens.  Again, this paper is not cited in MSS’s Summary of Science.

There is a page on the Scottish Government website from Marine Scotland Information regarding a two-year project called Salmon Parasite Interactions in Linnhe, Lorn, and Shuna (SPILLS) that began in October 2020 and runs to December this year. One of the aims of the project is to calibrate the model to actual data from the water including historic sentinel cage and plankton tow data and new plankton data. There are rumours that this new data fails to match up with that which is modelled. Given the findings from the papers mentioned above, this would not be surprising. Unfortunately, Marine Scotland Science are unlikely to provide any indication of their findings before their report is published which could be months away but given the relevance of this work to the framework, any results should be made available now.

In the meantime, perhaps the Coastal Communities Network might provide an insight of the findings. In their submission to SEPA, they stated that they are liaising with the Marine Scotland Science SPILLS team, so clearly, they have access to this research that is denied to others.

Even though, research on proposed dispersal of larval sea lice stretches back twenty-five years, there is still no proof that such a sea lice soup even exists.

I would end this commentary with reference to two other pieces of work. The first is that Marine Scotland run the Scottish Coastal Observatory based in Loch Ewe, a sea loch well-known to those working on farmed- wild salmon interactions. The relevant page states that the sampling site at Loch ewe is located in a sheltered fish farm hosted by Jane and Willie Grant.  I have questioned this location and it may be incorrect, but the sampling site is near the location of the former Marine Harvest Loch Ewe site.

In total the observatory has 201,163 zooplankton records from 2002 to 2017. Of these 820 are for Caligidae larvae (The exact species is not identified in this study, but the sampling site is very close to what was an active salmon farm) and of these 67 were measured as less than one individual per cubic metre. Four were measured at between 1 and 2, five with between 3 and 4 and one with 6, one with 9 and one with 12 individuals. These were spread randomly over a period of sixteen years. Such levels of lice in the plankton are not really sufficient to generate any real concern, especially when the sampling site is so close to an active salmon farm.

Finally, in the absence of other data, one measure used to assess the impact of lice from salmon farms is the sea lice sampling of sea trout. This work has been used to suggest that sea lice larvae could travel over 30 km from the farm yet given that no significant levels of lice larvae have been found in the open water, I would like to suggest an alternative explanation. Rather than sea lice dispersing from the farm to areas of fish sampling, perhaps, the fish travelled into close proximity of the salmon farm and then continued to swim on to where they were eventually caught.

I’ll discuss this further in a future issue of reLAKSation together with the question whether sentinel cages offer an accurate method of assessing sea lice infestation.


Bang: I have previously written that the wild salmon sector has abdicated any interest in finding out what impacts on wild salmon at sea. They prefer to focus their attention on matters close to home and especially matters that are not connected to their own interests.

I have previously raised the subject of seismic testing as a possible explanation why salmon are failing to return to Scottish rivers. My attention has now been drawn to a Marine Scotland assessment website about impulsive noise ( It seems that between 2015 and 2017, 79% of marine noise is attributable to seismic activity and 17% to sub bottom geographical surveys. The site includes an image of the location of such noise cross all marine areas and not surprisingly, the west coast hardly features leaving the question as to why removal of ADDs had become such a focus except as a way of attacking the salmon farming industry. The noise generated elsewhere is much more significant.

I have been reminded of such seismic surveys because the Norwegian Institute of Marine Research has been conducting trials on the impact of seismic surveys on spawning cod. They say the background to this current work were large scale trials during the 1990s that cod and haddock were scared away from areas of seismic activity. In the case of salmon, such seismic activity could distort the fish’s ability to navigate due to disruption to the internal magnetic material.

Sadly, Government research has been directed towards control of salmon farming even though the impacts are negligible instead of doing even some basic assessment of the impacts of seismic guns. Is it a coincidence that salmon numbers began to decline at about the same time that oil and gas exploitation around Scotland was being first explored.