The paradox of "premature migration" by adult anadromous salmonid fishes: patterns and hypotheses

In several groups of anadromous fishes, but especially the salmonids, some populations migrate from the ocean to fresh water many months prior to spawning. This "premature migration" reduces growth opportunities at sea, compels them to occupy much less productive freshwater habitats, and exposes them to extremes of flow and temperature, disease, and predation. We first review migration in salmonids and find great variation in timing patterns among and within species, relative to the timing of reproduction. Premature migration is widely distributed among species but not in all populations, and we propose two hypotheses to explain it. First, the fish may be making "the best of a bad situation" by entering early because access to suitable breeding sites is constrained seasonally by flow or temperature regimes, so they sacrifice growing opportunities at sea. Alternatively or additionally, some populations may be "balancing risks and benefits" as they trade off the benefits of growth at sea against the risk of mortality there. In this model, the reduced risk of mortality at sea must be balanced against the risk of mortality in freshwater habitats from thermal stress, disease, and predators. Premature migration may be favored where temperatures and flows are moderate or where lakes provide safety from predators and reduce energetic expenditure. Consistent with this hypothesis, early return is characteristic of larger, older salmonids (that would benefit less from additional time at sea to grow than would smaller fish). Finally, we consider the vulnerability of premature migrants to climate change and selective fisheries. Migration timing is an important part of the portfolio of phenotypic diversity that conveys resilience to species, population complexes, and the fisheries that depend on them. The premature migrants are often especially valued in fisheries and also often of particular conservation concern, and the phenomenon merits further research.