The marine life of sea trout (Salmo trutta L.)

Abstract

The research presented in this thesis revolves around the migration behaviour and survival of wild sea trout (Salmo trutta L.) in the marine environment. This area of research has seen increasing academic and managerial interest recently, as several aspects of sea trout survival and behaviour in the marine environment remain unknown or scarcely described. Sea trout individuals returning to spawn in their natal stream can constitute a high proportion of the overall spawning biomass in a population and contribute a disproportionately large production of juveniles, thus making their success and survival in the marine environment important for population sustainability. We therefore used a combination of biotelemetry, morphological measurements and physical samples to study different aspects of the marine life of sea trout entering the marine environment for the first time (smolts) and veteran sea trout re-entering the marine environment after spawning (kelts). By using acoustic telemetry on smolts (MS I) and kelts (MS II) from Karup and Simested rivers, which enter the Danish Limfjord, we were able to document the lowest minimum survival observed in any fjord system for sea trout (20 % for smolts and 48 % for kelts). We also found specific behavioural traits suggesting that the fish were adapted to the unique conditions of the fjord. Both smolts and kelts left the fjord in the eastern direction within a short window of time. The kelts entered the fjord gradually during winter and spring, and resided in two southern compartments of the fjord, with 41 % of the fish re-entering the river several times before making their final movement into the fjord and subsequently into the Kattegat. In April, sudden increases in temperature appeared to spark the onset of the migration out of the fjord. Most tagged fish left the fjord within a period of two weeks with mean progression rates through the fjord being 14.7 km-1 in the first part and 31.8 km-1 in the narrower, eastern part of the fjord. Fish from both Karup and Simested rivers behaved similarly in terms of migration timing, movement rates and mortality in the fjord. This suggests that the observed behaviour of both smolts and kelts represents an adaptation, and is thus important for survival and success in the Limfjord. By using data storage tags on sea trout kelts from seven Danish rivers, we were able to provide the first detailed reports on survival and behaviour of sea trout at sea (MS III, IV and V). The tags measured temperature and depth of the fish every five minutes and revealed a characteristic diel behavioural pattern where the fish performed several dives deeper than 5.0 metres during daytime and resided almost constantly at shallow depths during nighttime. The fish appeared to reside in areas with temperatures warmer than the mean (for the seas surrounding Denmark) during spring when mean surface (0 – 2 m) temperatures were below 11o C, and in areas with temperatures colder than the mean during summer when surface temperatures started exceeding 15o C. The fish showed a marked response to sea surface temperatures exceeding 17o C and disrupted their characteristic diving behaviour to remain submerged to depths of 5-15 metres almost permanently during such conditions. The data suggested the fish were close to land during spring and had entered open sea during summer. We employed a new type of DST where a string of floats enabled the recovery of the tag. This approach provided us with the possibility, for the first time, to study the detailed behaviour of fish that did not survive. The data revealed mean survival times at sea of 14.3 days (excluding angled fish) for the fish that died at sea, suggesting that the early part of the marine period is particularly 4 dangerous for sea trout kelts. The fish that died at sea were also in significantly poorer condition (P = 0.005) when tagged than those that survived the marine period. They also appeared to behave differently by performing fewer dives and visiting more shallow depths than the survivors in the initial period at sea. This was particularly true for the fish that were caught by recreational anglers, which raises the possibility of harvest-induced selection in sea trout. By applying a Hidden Markov Model on temperature and depth data recorded by the DSTs, we were able to reconstruct kelt migration tracks at sea. These tracks corroborated the expectation that sea trout are more coastally orientated than e.g. Atlantic salmon Salmo salar that have been studied with similar methods in other studies. All estimated positions of tagged individuals were within 100 km from the nearest coastline, but individuals migrated up to 580 km away from their natal river along the coast. All reconstructed tracks suggested a movement into areas with less saline waters had occurred in the first 15 days at sea, which would increase the metabolic scope available for growth for the fish. In sum, the results from MS III, IV and V therefore suggest that sea trout may seek out different areas favourable for growth at different parts of their marine period. The results of the present PhD-project sheds new light on the behaviour and survival of sea trout in the marine environment, provides input for managers in the region and identifies new areas of interest for future investigations.