Bias in hatchery-origin coho salmon survival estimates due to underestimated prerelease mortality
The widely accepted belief that hatchery-origin salmon survive less well than natural-origin or wild salmon can be, at least in part, an artifact of the way hatchery salmon survival is estimated. Hatchery salmon are often marked several months before release, while natural salmon are marked during their seaward migration. Underestimated prerelease mortalities result in biased low survival estimates. In British Columbia, although hatchery rearing practices have been modified to reduce unrecorded mortalities, experimental evidence indicates that coho salmon (Oncorhynchus kisutch) smolt survivals continue to be underestimated by ∼13%. Better reporting and incorporation of survival bias in data sets and analyses as well as additional work to evaluate bias for other regions and species over time is needed.