Ørjan Karlsen: "The trawl period covers what it is supposed to cover, but we probably do not cover the entire emigration period." Photo: Institute of Marine Research.

Traffic light trial: smolt trawls ‘carried out correctly’

A senior researcher at Norway’s Institute of Marine Research (IMR) has defended the smolt trawling methods used in connection with the state’s “traffic light” system that determines salmon production levels based on lice risk to wild fish.

Ørjan Karlsen is a member of the expert group that advises the state about which traffic light each of 13 production areas should be given. Areas with a red light must reduce production.

On Friday, Karlsen appeared as an expert witness in the court case between the state and the 25 salmonid farmers in production area 4 (PO4) who are contesting the red light given to their area last year. These include Mowi and Scottish Sea Farms co-owner Lerøy.

Karlsen rejected criticism that has come from the farmers in the case that scientists have trawled wrongly, at the wrong times, and that they have too few observations in their basis for a red light.

“The trawl period covers what it is supposed to cover, but we probably do not cover the entire emigration period,” he pointed out.

Hard to get fish

Karlsen added that there are capacity considerations and in some fjords the lice infection pressure is outside the fjords.

“The highest infection pressure is along the coast and not into the fjords. We have to go out to the coast to trawl, and there it is terribly difficult to get fish,” he told Sogn og Fjordane District Court.

When asked by government lawyer Hilde Lund about the criticism that has been directed at the number of fish and the condition of the fish, Karlsen said: “The people in the expert group are aware of representativeness in numbers, time and space. We are familiar with that. We trawl in six fjords, and we have done so for the past four years. These claims that we only catch weak and bad fish are simply not true. Whether we trawl inwards or outwards has no effect on the fish we catch.”

Underestimated infection pressure

Asked if there was a fixed way of trawling, he replied: “You drive crosswise and to and fro and really completely randomly, as long as it is within an area.”

Karlsen said that scientists’ conclusion is that they do not trawl too late in relation to the time of emigration of the wild smolt, which is an argument that has been presented by the farmers.

“The trawl period is adapted to emigration data from Lærdal, as well as knowledge of the rate of progression,” explained the researcher.

“The trawl period covers early migrating fish. Possible exceptions are from the two small middle rivers. The trawling period does not cover the end of the emigration period, which means that you will normally underestimate the number of salmon lice on the fish.”

Frank Nilsen giving evidence to the court online.

All available knowledge used

The court had earlier heard from Professor Frank Nilsen, former leader and now deputy leader of the expert group of ten who have made the professional assessments that form the basis for traffic lighting.

Nilsen explained that the expert panel comprised people with different backgrounds, chosen by the traffic light system’s steering group.

“We do not represent the institutions we come from, but ourselves and the competence we have,” Nilsen said, adding that the group’s way of working is to use all available knowledge.

“In addition to using all possible publications that we find, not only those published in international scientific journals, we also use the total knowledge that sits in the group by virtue of people being experts in their fields.”

Different models ‘are a strength’

The expert group had earlier in the case been criticised for the fact that data obtained from farming teams, according to the farmers, had not been taken into account. Nilsen denied this.

“One of the things that was discussed early on was whether a limit should be set in which knowledge should be assessed. We did not think it was right to limit ourselves to international scientific time shifts, so we have used everything we know when we have made our conclusion.”

Trond Hatland, the lawyer representing the farmers, asked if the expert group had gone deeper into infection models to validate and evaluate these. The validity of the models is a key plank in the farmers’ case.

“We have not gone into detail in the models more than the knowledge that the individual expert group members have,” said Nilsen.

“The advantage of having more than one model is to see if the data that comes out points in the same direction, and thus be able to see how great the uncertainty is. The models are built up a little differently, which we consider a strength, because then the sum will be able to give us a clear message.”

Water movement

He also pointed out that where variations between models were found, the group tried to find a cause. As an example, he said the model for PO4 in 2019 computed by the Veterinary Institute (VI) showed a lower lice impact than other models.

“Due to the coastal current bringing salt water into the fjord, more lice larvae were transported further in than normal. Since VI’s model is not a hydrodynamic model that does not take into account the movement of water masses, there is probably a very high probability of why there was a difference,” Nilsen said.

When asked what they had considered related to lice mortality in its different stages, Nilsen again answered that here they have used the available data.

“Here we have good knowledge, both based on lab experiments, but also from infection in the field. We have of course taken into account that there is a certain mortality in the louse when it grows in the various stages.”

‘Almost an insulting claim’

When government lawyer Lund asked a similar question, Nilsen replied: “It is almost an insulting claim that someone wonders if we know that the lice disappear from the salmon. We have good control over lice biology and how salinity and temperature affect this.”

Hatland asked about experiments from Canada, Scotland and Ireland showing high mortality of lice larvae in the sea right after hatching.

“From wild? How in the world have they measured it?” answered Nilsen, who did not know what the farmers’ lawyer was referring to.

Time of emigration

Hatland went on to ask about the time of emigration. Expert witnesses called by the farmers had previously questioned whether infection pressure was being calculated too late, after fish has migrated.

“We have written that it is one of the things we do not have the best overview of, but if we think we have a good enough overview, not least together with other conditions, then we have enough data to make assessments for the various areas,” said Nilsen.

Responding to the claim that assessment is based on a steady migration over 40 days, Nilsen said: “It’s not like that! It is based on an emigration profile, and not least the time the fish spend migrating out, which we have a pretty good idea about.”

Enough information for decison

State lawyer Lund asked Nilsen for his view on claims by others in court earlier that the knowledge base for the PO4 red light decision “was really bad”.

“As a scientist, I cannot say that it is bad. What lay people say is another matter. We have had enough to make our assessments,” Nilsen stated.

“Our job is to deliver a scientific report, conclude with lice-induced mortality in the production areas, explain data, uncertainty, concepts, etc. The underlying research has been done by others, we used it, and evaluate it in the best possible way to say as much as possible about lice-induced mortality,” he said.

He emphasised that he has so far not experienced any conflict within the group based on where the individual works.

“I have not experienced a power struggle. We are a well-composed group, which has good discussions, and we ask each other critical questions.”

Sea trout

Another aspect that has been a contentious issue in the trial is the use of data from sea trout to say something about the effects of lice on salmon.

“Where do you use sea trout in the data base,” Lund asked.

“The number of trout says something about the infection pressure. Based on the different lice stages, we can thus say something about when the salmon became infected, which can tell someone about high and low infection pressure during the period. Mortality and damage to sea trout are not used as a measure of mortality in salmon,” Nilsen emphasised.

He said this was because sea trout have a very different behaviour from salmon, and can, for example, migrate back into the rivers. The knowledge is not in place yet to be included, he said.

Snow melt included

Asked whether the expert group knew that when the snow melts in the spring, there will be a freshwater layer in the fjord, Nilsen said this had been taken into account.

“This is where the expert group comes into its own,” he added. “When the indicators point in different directions, the experts can go in to identify why.”

The judge asked Nilsen whether it is controversially scientific that farmed fish lice are transmitted to wild fish, or whether it is something that is widely agreed upon.

“In my opinion there is a clear agreement. I do not know of anyone serious who thinks otherwise,” said Nilsen.

“Do you have the impression that farmers also think this,” asked the judge.

“There may be some disagreement about the scope, but most people are probably aware that there is a connection,” Nilsen said.

Huge number of lice

Nilsen pointed out the difficulty in making understand how many lice come from a fish farm.

“Even though there are very few lice on the individual farmed fish, for example 0.1 female lice per fish, a facility with say, two million fish, with 0.1 sexually mature female lice per fish, will have 200,000 such lice, which produce minimum 50 copepods each per day. So, from one such farm alone, 10 million copepods are produced per day. When there are 3-400 million farmed fish in the sea, it will be a huge number,” he explained.

The court will hear from more witnesses this week.