1. Migration and survival of Atlantic salmon smolts in a river and a fjord: the effects of different release strategies from a hatchery
Eva B. Thorstad1, Ingebrigt Uglem1, Bengt Finstad1, Cedar M. Chittenden2, Rune Nilsen3, Finn Økland1 & Pål Arne Bjørn3
1Norwegian Institute for Nature Research (NINA), 2Norwegian College of Fishery Science, 3Nofima Marine, Norway
Aim and methods
Declines in wild Atlantic salmon populations have led to the widespread release of hatchery-reared salmon smolts as a mitigative strategy.
Aim of study:
To compare the success of three smolt release strategies.
Hatchery-reared acoustically tagged smolts were released in the river mouth (one group, n = 33) and 9 km upriver (two groups, n = 66) (▲). The smolts were recorded by receivers (●) at three sites in the fjord. Depth tags were used to distinguish between live smolts and tags being in the stomach of predators that had eaten tagged smolts.
Results
A large proportion of the smolts were lost in the river (64%), likely due to predation or reduced motivation to migrate.
The group exposed to a longer transport time and direct release in the river did not differ in survival or migration from the group allowed to acclimate for 48 h in a river net pen.
Mortality in the marine environment did not differ among groups. Marine mortality was 37% during the first 2 km, with at least 25% due to predation by marine fishes. The total marine mortality over 37 km was 68%.
Conclusions
The large immediate loss of smolts after release in the river, emphasises the need for improving smolt production and riverine release strategies.
Hatchery-reared smolts (pictured left, top) are often much larger and have a higher fat content than wild smolts from the same stock (pictured left, bottom). Hatchery regimes producing a more ”natural or ecological” smolt may improve the success of hatchery-releases.

2. How can we recognise live smolts from smolts that have been eaten by
fish predators in telemetry studies?
Eva B. Thorstad1, Ingebrigt Uglem1, Bengt Finstad1, Cedar M. Chittenden2, Rune Nilsen3, Finn Økland1 & Pål Arne Bjørn3
1Norwegian Institute for Nature Research (NINA), 2Norwegian College of Fishery Science, 3Nofima Marine, Norway
Atlantic salmon post-smolts (n = 57) and fish predators (Atlantic cod and saithe, n = 10) were tagged with depth tags. Their behaviour was recorded by 25 automatic receivers in the Romsdalsfjord, Norway.
Atlantic cod
Figure 1. Depth recordings of four individual fish predators.
Figure 2. Depth recordings of four individual post-smolt tags. The two post-smolts in the upper graphs show a typical vertical behaviour of post-smolts, whereas the post-smolts in the lower graphs show the typical vertical behaviour of Atlantic cod and saithe. They have likely been eaten by such a predator and the tag remained within the stomach of the predator. The tag ID 53 seems to have been expelled by the predator after 14 days and remained on the sea floor after that. Note the different scales on the y and x axes.
One Atlantic cod and one saithe showed a typical post-smolt behaviour, i.e. directed movement out of the fjord. Hence, if a tagged post-smolt is eaten by an untagged predator, the latter might be interpreted as a post-smolt.
However, vertical behaviour differed between predators and post-smolts, with predators swimming much deeper (Figure 1 and 2).
Vertical behaviour was used to identify whether the tag was still in a live post-smolt, or whether it had been eaten by a predator.
If depth tags had not been used, mortality of post-smolts would have been underestimated (26% instead of 37% during the first 2 km of the marine migration).
Tags from eaten post-smolts remained in the body of fish predators for up to 47 days (average 29 days).