59% of young salmon failed to complete run to coast

by EWAN LAMB

The first phase of a research project by the Tweed Foundation to discover survival rates among the river's sea-bound smolt (young salmon) population has produced results suggesting virtually all of the losses occur between Galashiels and the village of Sprouston, near Kelso.

The so-called smolt run by tens of thousands of juvenile fish involves a 50-mile journey from the Gala area to Berwick-on-Tweed is a vital element in the overall well-being of Tweed's salmon numbers. So a survival rate of just 41% as shown by the initial stage of the research would seem to set alarm bells ringing.

But the authors of the report on the 2019 pilot study involving the tagging of just 60 smolts from the Gala Water with transmitters warn the sample was too small to draw any firm conclusions. Much more work will be needed before recommendations to protect the smolts can be put forward. It has been suggested many of the smolts fall prey to predatory birds - cormorants and goosanders - as they head downstream.

The newly published report explains: "The salmon smolt is the final product of the river. Losses at this stage of the life cycle cannot be replaced. With numbers of returning adult salmon continuing to decline due to increasing marine mortality, a Tweed smolt tracking study has been setup by The Tweed Foundation to investigate in-river smolt survival and assess the management options available. 

"The 2019 tracking work is a pilot study that will guide more comprehensive work in 2020 and 2021. Sixty smolts from the Gala Water were tagged with acoustic transmitters and tracked down the main river (a distance of 80 km) with fixed and mobile receivers to determine smolt survival to sea and where losses took place.

"Thirty-nine of the 60 tagged fish came out of the Gala Water. The remaining fish either did not leave the Gala Water or were removed from the analysis due to a receiver not working in Berwick harbour.

Sixteen out of the 39 tagged fish that were detected at Galafoot were recorded at the receivers located in Berwick (41% survival).  With such a small sample relative to the overall population, it cannot be assumed that these results accurately reflect the wider population of smolts.

"Twenty-one out of the 23 tag losses were between Galafoot and Sprouston (91%), even though most of this stretch of the river has licensed mitigation in place for goosanders and cormorants (scaring and control). The remaining two smolt losses were between Sprouston and Cornhill, meaning that there were no recorded losses between Cornhill and Berwick."

According to the research team if the losses concentrated in the Middle Tweed accurately reflect the Tweed smolt population, then the implication may be that the current licensed management for goosanders needs to be reviewed, with planned tracking work in 2020 and 2021 testing alternative strategies for improving smolt survival.

It is hoped to expand the study for future years if sufficient funding can be secured. Smolt tag sponsorship will also be on offer for those who wish to support the important work.

But the report repeats its warning: "It is important to remember that a sample size of 39 tracked fish represents a fraction of a percent of the total smolt population. To provide some context, the annual Tweed salmon smolt production could be between one and two million smolts per year. For the Gala Water we have a more accurate estimate of salmon smolt output of 15 to 20,000 fish each year."

The report then goes on to say: "With the continuing caution of making strong conclusions about the results due to the relatively small sample size, some consideration can be given to the number of goosanders and cormorants in the river during the study and their potential impact upon the smolt run.

"Regular counts of goosanders and cormorants are carried out by the Tweed Foundation and River Tweed Commission staff with assistance from volunteers in January, April, May and October each year.

"On 4th April this year we counted 250 goosanders and 44 cormorants between Ettrickmouth and Berwick which covers the extent of the (smolt) tracking study area. The count of 250 goosanders was the second highest on record (compared to an average of 174). Through April and into May, male goosanders leave the river for their annual migration to Northern Scandinavia and female birds move into the tributaries to nest. May counts are therefore normally much lower: this year’s count, on the 22nd May, 2019, recorded 118 goosanders and 14 cormorants. 

"Cormorant numbers are much lower than goosanders, particularly in May, because they migrate back to the coast to breed. As their potential damage to the smolt run is therefore minimal, goosanders present the larger threat to the smolt run."

Previous research in the 1990s indicated Tweed goosanders were consuming five smolt-sized salmon per day during the smolt run, which is low relative to Northern rivers (often between 15 and 20). 

"To provide some context for potential damage to the smolt run, using a figure of 200 goosanders for April and May (higher than the average) and a consumption of five smolts per day produces a smolt loss of 61,000 fish. A loss of 61,000 smolts out of 1,000,000 (probably an underestimate) produces a ballpark figure of a 6% loss, which is much smaller than the figure of 59% produced from the 2019 tracking work."

Future smolt tracking work as part of the study may focus on the comparison of survival rates between areas of high bird predator management (scaring and removal) in the Middle Tweed with areas that have no management. 

"Detecting a significantly higher survival of smolts where bird numbers are managed would provide the evidence that more intensive, targeted, mitigation is required for a measurable improvement in smolt survival. Ultimately, the aim has to be the ability to strike a sensible balance between the management of the abundance of smolts, that are successfully able to negotiate their way to the sea and the number of avian predators.

"A pilot study to track goosanders is also being carried out in Autumn 2019, which will potentially allow birds to be tracked during the smolt run in 2020. This will allow us to gain a better understanding of their feeding behaviour and, in particular, where they feed, and to investigate the effect of scaring techniques."