Preface: reLAKSation was originally conceived as being views on the news and especially in relation to aquaculture and the market. As my involvement in wild fish/farmed fish interaction has increased, this has been reflected in the content of this commentary. The past couple of weeks has seen the focus on one news story that has dominated the mainstream media leaving little else to comment on.
The most important long-term story in Scotland has been the decision to impose a risk-based framework on the salmon farming industry to protect wild salmon and sea trout. Yet by comparison, the wild fish sector is being asked to show voluntary restraint in killing the same fish for sport. It does seem that salmon farming is being unfairly targeted especially when the regulator tasked with imposing the framework agrees that salmon farming is not responsible for the declines in wild fish. This new measure comes at a time when other salmon farming nations are also being subjected to possible controls, leaving the impression of a wider crusade against the salmon farming industry. This is why I continue to highlight the inconsistencies in the science being used to justify the new measures.
Some might argue that my views are anti-angling, but I do not have a problem with angling per se. What I cannot accept is that the wild fish sector persistently use aquaculture as a scapegoat for issues which they have failed to address themselves.
I also am puzzled by the reluctance of Scottish Government and Scottish Government scientists to discuss the science. I suspect that their science has never been challenged previously and thus are simply unsure of how to respond and so have just shut the doors to any discussion. A recent article in the Guardian newspaper about the UK badger cull very much resonated with me. The UK Government is persisting with the cull despite new scientific evidence to suggest that it serves little purpose. The Guardian reported that instead of discussing the science, the UK Government have simply refused to entertain the new findings.
Whilst wild fish interactions take up some of my time, I still follow the markets as this is Callander McDowell’s core business. I hope to discuss some of the latest market issues in a forthcoming issue of reLAKSation.
Uncertainty: One of the most important aspects of the forthcoming SEPA risk-based framework will be the monitoring. This is expected to be a key feature of the second consultation on the new framework. Interestingly, the science of monitoring is not considered as part of the Scottish Government’s Summary of Science relating to sea lice. The nearest to any reference to monitoring is hidden in the statement that: data collected throughout the west coast of Scotland showed that the proportion of individual sea trout with stress-inducing sea louse burdens decreased with the distance from the nearest salmon farm (Middlemas et al. 2013).
I have previously discussed this paper and reanalysed some of the data and have not reached the same conclusion as the authors. In addition, I repeated the exercise with data from 2011 to 2015 which was published by the then Rivers and Fisheries Trusts of Scotland (RAFTS) and did not find that sea lice infestation on fish declined as the distance increased to the nearest salmon farm. As I have pointed out fish do not stay in the same place.
However, even the results in Middlemas et al (2013) are not conclusive. The paper states that there was considerable uncertainty around the cut off distance of 31km (range 13-149km). I’m sure that I would be uncertain too about claiming a defined cut off distance when the results produced such a expansive range covering 136 km.
Although the paper was published in 2013, the data was collected during May to July by the five local fishery trusts between 2003 and 2009. That means that this data is now 13 years old but as we know the fishery trusts have continued collecting data ever since and are actually paid to do so by Marine Scotland Science. A FOI request revealed that since 2010, MSS has provided funding of £624,325.60 to the trusts to fund a programme of work in relation to aquaculture/wild fish interactions of which sample netting has been a component.
As this work is carried out on behalf of MSS, it would be expected that there would be a strict scientific protocol in place. One was put in place through the Scottish Fisheries Coordination Centre in 2008/9 but even this seem to miss out one very important aspect of the scientific aspect of this monitoring programme and that is adherence to specific sites so that long term trends can be studied.
The Middlemas paper does not mention which sites were used. Instead, they provide a map indicating the location of each sample site and more interestingly, the size of the point indicates the number of years that data from that site was available. The different sizes of the points suggest that data was available for different sites ranging from one year to all seven years. In all, there appears to be 41 different sampling sites with most being in Argyll (24). It is unclear why just those sites with longer runs of data were the only ones to be used. Perhaps the decision not to use those sites with most data is why the eventual conclusion drawn from the paper was one of such uncertainty.
More critically, given the uncertainly and the continued provision of data, it is surprising that MMS has not repeated the work to clarify this uncertainty. Perhaps, one reason would be the difficulty of analysing data from the same site, year after year.
The RAFTS report of 2011 (available on FMS website) states that twenty-eight core sites had been selected with the aim of achieving good coverage of sites across a range of varying distances from active salmon farms. In 2012, the number of core group of sites had fallen to 22, then in subsequent years to 24, 24, and in 2015 there were 27. In 2017 (no report in 2016) there were just 22. It seems that this core group was not a core group at all but one that continually changed. This is not the way science is conducted.
By 2019, according to the sea lice database, just nineteen sites were sampled. It’s not clear what happened to the other nine core sites. Of the nineteen, just four sites have been tracked since the late 1990s, although none run consistently from 1997 to 2019. Another two were monitored from the early 2000s, with the rest from the late 2000s onwards, except there which are relatively recent additions to the list. In 1997, there were just six sites sampled. Why they were never tracked continually remains a mystery.
Apparently, some of the reason why sites were later rejected include difficulty in netting and not catching enough fish (and maybe because enough sampled fish did not have any lice) but since 1997, a total of 97 sites have been sampled for sea trout with lice. Of these twenty-eight were sampled in just one year and another twenty sites were sampled for just two or three years.
Over the twenty-three years recorded in this database, 21,628 sea trout have been sampled, yet there does not seem to have been any real analysis of the data by MSS. Some of the fishery trusts highlight the few examples of fish infested with high numbers of lice as proof of the damaging effect of salmon farms, even though these fish are very much the minority. There seems to be no interest in why, if 13 lice can kill a smolt, fish can be caught with hundreds of lice that are still alive and are not considered near enough death that they need to be euthanised. Equally, there is no mention of recaptured fish that have fewer lice on the second occasion.
Some of the fishery trusts do try to apply their sampling to Taranger’s risk analysis despite its obvious limitations. These are described in a video presentation posted on the Fisheries Management Scotland website (https://fms.scot/wp-content/uploads/2021/12/210321-Aqua-Monitoring-workshop-report-.pdf). The last limitation described is that there is no estimate of uncertainty. This is an interesting statement because whilst I suspect that those promoting such risk analysis have never actually spoken with Geir Lass Taranger, I have been fortunate to have had the opportunity to do so. The one clear message that I took away from this meeting can be summed up in one word – uncertainty.
2017: Although the rivers and Fisheries Trusts of Scotland had been disbanded by 2018, someone form the West Coast Fisheries Trusts put together a report of the 2017 sweep net post smolt monitoring results in a similar way to the past reports from RAFTS from 2011 to 2015 (all available on the FMS website). The summary stated that L. salmonis sea lice were present at all of the monitoring sites with over 50% of the sea trout infested at 16 sites and with every fish infested at two of the sites. An analysis of the data indicated that 9 sites were at high risk of population changing effects, whilst only 7 were of low risk. There was no mention of what risk existed for the remaining 8 sites.
The body of the report mentions that two new sites were introduced to the monitoring programme for 2017 due to low trout numbers and difficulty in sampling at other sites. There is also mention that the programme follows the protocol from SFCC that there should be a target of more than 30 fish in each sample and that each site should be sampled at least twice. The report also confirms that the sampling programme is funded by the Scottish Government.
The report continues that of the 22 sites surveyed only 13 returned samples of 30 fish or more whilst two sites had sample size of just 2 fish. Three sites were only sampled on one occasion whilst a further site only recorded fish on the second sampling date. However, the sampling records listed in the MSS database suggest a further two sites were sampled on just one occasion and another site was not sampled at all. The last record for this site was in 2015 even though the current report lists prevalence and abundance figures for the site.
The report states that 900 fish were sampled although the database totals 880 fish for the dates and sites covered. The analysis of the data says that one site had a prevalence of 100% whereas the table in the report shows that four sites show a 100% prevalence. The other three are ignored because the sample size was below the 30 fish threshold. What the report fails to mention is that the fourth site only had a sample size of 39 fish which is close to the threshold. The protocol does not reveal how the figure of 30 fish was chosen but what is clear is that in Norway, Taranger recommends a sample size of more than 100 fish. Given that the fishery trusts are keen to relate their findings to Taranger, then perhaps they should aspire to a much higher sample size. Sample size is an important factor in measurement of parasite risk, but it has not been given sufficient consideration in the running of this sampling programme.
The report says that prevalence is only an indication of the percentage of infested fish versus uninfected fish and that abundance – the average number of lice in a sample gives a better indication of the distribution of sea lice amongst the population yet the initial focus remains on prevalence.
The report found that abundance ranged from 0.11 per fish to 110.44 lice per fish even though small sample sizes and one fish with a high lice count can dramatically alter the profile. Using abundance, the report suggest that 10 sites had an abundance greater than 13 lice, the threshold estimated by Wells et al (2006). It is no coincidence that Dr Wells is now head of FMS, the organisation now running the sea lice programme. It also brings into question the impartiality of the whole monitoring programme given the effort that FMS take to try to impose greater regulation on the industry.
One part of the report especially caught my eye. It begins that in 2012(13), Middlemas et al. analysed the West Coast fishery’s trust’s sweep netting data from 2003 to 2009 and concluded that the potentially damaging levels of sea lice infestations was related to distance frm the nearest farm. The report continues that the maximum range of effect of sea lice from farms was 31km but that there remains an inherent uncertainty with this estimation of distance due to the previous study being focused on localised investigations. With refinement of the programme since 2011, the report suggests that the new data is available to Marine Scotland Science, and it is envisaged that the uncertainties identified in the previous work can be further explored and definitive conclusions drawn.
Now where have I heard that before?
Finally, as MSS have not conducted any analysis of the 2017 data, I would like to offer a different perspective to that of the West Coast Fishery Trusts. Understanding a parasite population requires large samples, primarily because parasites are not distributed normally. Thus, the largest sample available for 2017 is the full dataset of 880 sea trout and the distribution of their parasite loading is shown in the following graph.
In this case, the majority of fish are not lice free, but the largest percentage certainly is. The reason why there is such a large spread of lice infestation (340 lice on fish and 520 on another) is the small size of many samples. Two sites sampled just two fish and the likelihood is that small samples tend to be the weaker easier to catch fish rather than a more typical range. The reality is that the majority of fish are either lice free or carry a very small number of lice. How this represents a significant risk to the wild population is unclear.
It’s not our fault: Recently, the Scottish Government issued a consultation on the encouragement of voluntary catch and release with the aim of achieving 100% across all of Scotland. The publication of the October edition of Fly Fishing and Fly-Tying magazine (FFFT) provides a first glimpse of the angler’s viewpoint of these proposals.
Whilst I say it is a first glimpse, the magazine repeats views expressed by the Angling Trust following a decision to introduce mandatory catch and release in Welsh rivers. The Trust said at that time that the imposition of mandatory catch and release was ‘illogical, disproportionate and impractical’. They said that for many game anglers, it is very important that they have the right to take a fish, even if they choose not to exercise it for conservation reasons. They conclude that voluntary catch and release is the best way to proceed leaving FFT to suggest that the underlying message being: the demise of salmon in the UK is not the anglers’ fault, so why regulate them?
That is a very good question indeed. SEPA have said that salmon farming is not to blame for the demise of wild fish on Scotland’s west coast, yet not only are they planning new regulation to protect wild salmon and sea trout, but it is also some of the most draconian, but unnecessary regulation possible.
In the case of salmon farming, there is still not a shred of evidence that the industry has had an impact on wild fish, despite claims stretching back over thirty years that it has. By comparison, the official Scottish Government catch data shows that anglers have killed over 5.9 million wild fish since 1952, most of which were breeding fish and they say that the demise of wild fish is not their fault.