Oceanic, riverine, and genetic influences on spring chinook salmon migration timing

Migrating salmonids often return to their spawning habitats in overlapping timing patterns of multiple stocks (populations) collectively called a run that varies in its genetic makeup across and within years. Managers, tasked with developing harvest strategies on these runs, may have preseason estimates of total run size but little information on run timing. Without both it is difficult to assess a run's status in real time. Consequently, to avoid overharvest, managers tend to control the timing of harvest. However, this strategy may inadvertently affect the component stocks disproportionately and therefore the run's diversity. Thus, accurate estimates of run timing are needed to improve management. We developed a model that includes genetic and environmental factors to predict the mean run timing of chinook salmon (Oncorhynchus tshawytscha) at Bonneville Dam on the Columbia River, Oregon, USA. The model predicted mean run timing (P < 0.00001, r(2) = 0.78) by characterizing genetic run timing components from the arrival timing of precocious males returning one year prior to the remainder of the adults and environmental influences of oceanic and riverine flows that impede or advance the run timing. Variations in the relative abundances of the populations in the run explain 62% of the interannual variation in mean run timing while the oceanic and riverine factors combined account for 15.5%. We suggest that when genetic run timing characteristics are preserved in species with multiple maturation strategies the information can be used to improve run time predictions and maintain genetic diversity of harvested species.