A little bit of digging that I have done...4
http://onlinelibrary.wil....2000.tb00569.x/abstract"The timing of migration and breeding are key life-history traits; they are not only adaptations of populations to their environments, but can serve to increase reproductive isolation, facilitating further divergence among populations. As part of a study of divergence of chinook salmon, Oncorhynchus tshawytscha, populations, established in New Zealand from a common source in the early 1900s, we tested the hypotheses that the timing of migration and breeding are under genetic control and that the populations genetically differ in these traits despite phenotypic overlap in timing in the wild. Representatives of families from two populations were collected within a day or two of each other, reared in a common environment, and then released to sea from each of two different rivers, while other family representatives were retained in fresh water to maturity. The date of maturation of fish held in fresh water and the dates of return from the ocean and maturation of fish released to sea all showed significant differences between the two populations and among families within populations. The very high heritabilities and genetic correlations estimated for migration and maturation date indicated that these traits would respond rapidly to selection. Combined with the results of related studies on these chinook salmon populations, it appears that spawning time may not only evolve during the initial phases of divergence, but it may play an important role in accelerating divergence in other traits."
http://onlinelibrary.wil...serIsAuthenticated=falseInterannual variation in the timing of the return migration to fresh water of adult sockeye salmon, Oncorhynchus nerka, from 46 populations throughout the species North American range was examined in a broad analysis of how timing patterns are affected by marine and freshwater conditions. Migration timing data (measured at various points along the migration, including just prior to freshwater entry, just after freshwater entry, and near the spawning grounds) were examined for correlations with sea-surface temperatures (SST) prior to migration and to freshwater temperatures and flows during migration. Following a spring–summer period with warm SST, populations from southwestern Alaska tended to return early, Fraser River populations returned late, and populations from other regions showed no consistent patterns. Similarities between interannual timing of both nearby and distant populations indicated the presence of common or coincidental influences on timing. When riverine conditions related to timing, high flows and low temperatures were associated with late migrations, low flows and high temperatures were associated with early migrations. However, even counting stations at upriver locations showed correlations with SST. Notwithstanding some inconsistencies among the many populations examined and the indirect nature of the inferences, the results supported the hypotheses that (i) interannual variations in salmon distributions at sea reflect temperature conditions, and (ii) the date when salmon initiate homeward migration is a population-specific trait, largely unaffected by the fish's location at sea.
and lastly, from wiki - salmon run
Salmon spend their early life in rivers, and then swim out to sea where they live their adult lives and gain most of their body mass. When they have matured, they return to the rivers to spawn. Usually they return with uncanny precision to the natal river where they were born, and even to the very spawning ground of their birth. It is thought that, when they are in the ocean, they use magnetoception to locate the general position of their natal river, and once close to the river, that they use their sense of smell to home in on the river entrance and even their natal spawning ground.
Sometimes it's good for a change. Other times it isn't.