The role of the feeding migration and diet of Atlantic salmon (Salmo salar L.) in yolk-sack-fry mortality (M74) in the Baltic Sea

Atlantic salmon (Salmo salar L.) populations exist in about 80 rivers draining to the Baltic Sea. Salmon reproduction has ceased in many rivers due to anthropological changes in river conditions and in some cases salmon stocks have been lost due to overfishing. To compensate for losses in smolt production, large-scale smolt releases into rivers have been carried out since the 1950s in the Baltic Sea area. In 2004, 7.1 million salmon smolts initiated their migration from the river mouths to the sea, of which 75% originated from hatcheries and the remainder were of wild origin. The rearing of salmon in most of the hatcheries around the Baltic Sea has been based on eggs obtained from ascending, feral spawners. However, in Finland captive spawners have been used almost exclusively to supply the eggs, while feral spawners have only been used to renew brood fish in hatcheries. In 1974, in a Swedish salmon hatchery, exceptionally high yolk-sac fry mortality was observed. The mortality was recorded only within a few family groups, while the others were viable. As no causative factors were found in the hatchery or in rearing practices, it was concluded that the syndrome must be related to nutritional factors in the sea environment from where the spawners ascended the river. Therefore, it was named as M74 syndrome. The letter M came from the Swedish word miljörelaterad, meaning environmentally caused, and 74 was the year (1974) when the syndrome was first detected. The syndrome has since been detected in the salmon populations ascending the rivers draining into the Gulf of Bothnia and Gulf of Finland, but not in the Baltic Proper rivers, except in the River Mörrumsån in southern Sweden. The M74 syndrome has always only been detected in a proportion of the females in different spawning populations. The syndrome strengthened rapidly in early 1990s, but almost disappeared during the early 2000s. These findings gave rise to the hypothesis that the M74-positive salmon had fed on different food compared to healthy ones. Therefore, I investigated the distribution pattern of salmon in the feeding areas and their diet in these areas. I also sought to determine causative factors in the food. Moreover, I have studied the migration patterns of salmon postsmolts with tag recovery data and the possible biotic and physical factors affecting the selection of feeding areas. Those areas where salmon perform their feeding migrations have been determined with the aid of salmon smolt tagging data by mapping the tag recoveries of feeding fish. I have also used the distribution of offshore catches to determine the feeding area selection. Stomach analyses of feeding salmon caught in different areas have used to reveal differences in food composition. The selection of the feeding area might take place during postsmolt migration, which is strongly guided by thermal zones in the inshore areas. It appears that during postsmolt migration, an abundant availability of fish of an edible size, such as 0+ herring, causes postsmolts to stop their further migration, and the actual feeding migration is then mainly guided by movements of prey fish shoals. Salmon populations in the rivers draining into the Gulf of Bothnia migrate to the southern Baltic The role of the feeding migration and diet of Atlantic salmon in yolk-sack-fry mortality (M74) 7 Proper to feed, but during some years more feeding takes place in the southern basin of the Gulf of Bothnia (Bothnian Sea) and the northern Baltic Proper. Stomach sampling suggests that salmon prey species in the central and southern Baltic Proper are sprat (Sprattus sprattus L.), herring (Clupea harengus L.) three-spined stickleback (Gasterosteus aculeatus L.) and sandeel (Ammodytes sp.) in order of importance. In the northern Baltic Proper, Bothnian Sea and Gulf of Finland, salmon prey dominantly on herring followed by stickleback and to a minor degree on sprat. The analysed data suggest that when the feeding migration of salmon is emphasized in the northern Baltic Sea, an increased proportion of M74 spawners in the rivers follows, and vice versa. It seems that a diet dominated by herring is a causative factor for M74 syndrome in the northern rivers. It might also be so that the causative factor is not species related but the reason could be found in the food web of salmon. In the northern Baltic Sea the composition of plankton species available for herring differs from that occurring in the central and southern Baltic Proper. A deficiency in thiamine, which has been shown to be the main reason for the syndrome, may be a result of a herring-orientated diet. Thiaminase activity is probably higher in herring than in sprat, and thus the thiamine stores of ingested herring are destroyed faster in the salmon stomach. In addition to a herring-dominated diet, a further reason may also lie in the food of salmon prey species. Small limnic plankton species dominate in the northern feeding areas, while in the southern areas larger neritic species are common. Poor nutritional conditions in the northern Baltic Proper were reflected in a retardation of herring growth beginning in the early 1980s and even more profoundly after the late 1980s, when the annual growth of sprat also decreased. There are indications that the neritic plankton species are better food for herring and sprat, containing more carotenoids than limnic ones and probably also a greater thiamine content. The M74 syndrome detected in the Gulf of Finland salmon populations seems to fit the same possible explanations. Those salmon feeding in the Gulf of Finland and northern Baltic Proper have suffered from the syndrome, while salmon migrating to more southern areas have produced viable offspring. Latvian and Polish salmon have not suffered from the M74 syndrome, even though at least a certain proportion of postsmolts migrate to the northern Baltic Sea. However, it is probable that later migration to the more southern areas and the different diet there prevents the emergence of the syndrome. The syndrome detected in the River Mörrumsån might also be related to the northern feeding areas of its salmon population. However, no detailed data on the feeding migration of this salmon strain are available. In the future, the factors leading to a strengthening of M74 mortality are still likely to exist. An increased proportion of feeding salmon in the northern Baltic Sea might result in a growing number of M74-positive females in the spawning populations. However, possible improvement in the salmon food web in the northern Baltic Sea might diminish the negative effect of feeding in the north.