Surviving the journey: The migration of smolts from river to sea

Migration allows animals to exploit spatiotemporally variable resources to maximise lifetime fitness, but poses substantial mortality risk. Mortality is rarely constant throughout the migration, but occurs at survival bottlenecks where risks are disproportionately high, such as narrow corridors that limit individuals’ predator escape options. One widely studied animal migration is the freshwater-marine transitions of anadromous salmonids. The smolt migration from river to sea can be particularly hazardous, and is often associated with high mortality. Because the number of smolts entering the sea is associated with the number of returning adults, increasing smolt survival can enhance adult returns and recruitment. As such, the smolt stage is often considered a critical life stage for the management and conservation of salmonids. This thesis aimed to uncover bottlenecks to smolt survival, with the focus on two iconic species of salmonids inhabiting rivers across the North Atlantic: the Atlantic salmon (Salmo salar) and the brown trout (Salmo trutta).
Because the research in this thesis originates from rivers in Denmark, Ireland, and Northern Ireland, where capture methods varied among countries, the first manuscript (MS I) compared the post-release impacts of two widely used smolt capture methods: pre-migration capture (electrofishing, not delaying migration) and during-migration capture (screw trap, potentially delaying migration). I found that pre-migration and during-migration capture yielded similar results in terms of behaviour and survival, supporting during-migration capture as a low-impact method and suggesting that this thesis’s other findings are comparable and representative of wild populations.
The remaining manuscripts investigated smolt survival across various river systems. MS II compared the migration behaviour and survival of Atlantic salmon and brown trout smolts within the same river, fjord, and year in Denmark, and revealed that both species shared similar behaviour and survival in the river, but displayed different migration strategies in the fjord: Atlantic salmon (F1 hatchery-reared) maintained similar speed as in the river, likely prioritising migration to reach oceanic feeding grounds, while brown trout (wild and F1 hatchery-reared) slowed down, likely to forage in the fjord. The fjord was also found to be a survival bottleneck for hatchery-reared brown trout, possibly due to poor anti-predator behaviour. These findings highlight that differences in behaviour and survival may arise only during parts of animals’ migration, likely impacting how and where stressors affect species along their migrations. MS III investigated Atlantic salmon smolt survival in the River Erriff, one of Ireland’s national index rivers, used to monitor marine survival and population trends based on smolt and adult return counts. I identified a predation hotspot below the smolt counting facility (Wolf trap) at a narrow, constricted bedrock cascade, which likely limited predator escape options, and combined with smolts’ recent handling in the trap, increased their predation risk. Because interrupting smolts during-migration did not impact survival (MS I), the predation pressure downstream Wolf traps may indicate predator habituation caused by the recurrent release of smolts each morning. Consequently, parts of the marine mortality estimated from index rivers may actually be freshwater mortality caused by a river survival bottleneck. In MS IV, I investigated if wetland restoration in the River Gudenaa, Denmark, created an ecological trap for migrating brown trout smolts by reducing water flows and/or providing an attractive area with elevated food resources compared to the river, thus encouraging smolts to spend more time in high risk areas, ultimately increasing passage time and mortality rates. Although smolt progression rates were slower, the restoration did not impact survival, likely due to the retention of a defined river channel with consistent flow. These findings show that electronic tracking can help evaluate efficiency of management interventions designed to reduce negative impacts on migratory animals. Collectively, the research in this thesis emphasises the importance of maintaining adequate natural habitats and environmental cues both along animal’s migration routes and at their destinations.
This thesis shows that high smolt mortalities can occur in rivers and coastal areas, with the location and magnitude of survival bottleneck shaped by local physical, ecological, and anthropogenic factors. Identifying where and why smolt losses occur is essential to design strategic management plans that increase smolt output and subsequent adult returns, ultimately boosting population recruitment. As anthropogenic stressors continue to change the ecosystems that migratory animals rely on, it is increasingly important to pinpoint and disentangle the numerous factors impacting survival across their complex life cycle.