Large-scale declines in marine survival of many Pacific salmon populations occurred over the last three decades. We examined when in the life history this mortality was expressed using the POST prototype array.
We first compared the size at tagging of released smolts and of the survivors at distant marine array locations, hundreds of
kilometers away from the release site and requiring ≥1 month travel time. The mean, variance, and higher order moments of the
size-frequency distribution was equivalent for each species examined (sockeye, steelhead, chinook, & coho). This indicates that mortality processes did not substantially distort the size distribution of survivors and that larger smolts did not survive better than smaller smolts above our pre-specified minimum size thresholds for surgical implantation (130 mm & 140 mm for Vemco V7 & V9 tags). We also found that survival to adult return of acoustically tagged Fraser River sockeye and Snake River spring Chinook smolts matched that of the overall (untagged) run in at least some years of tagging. As surgical implantation of acoustic tags does not therefore substantially reduce smolt survival after release, it is possible to calculate how much of the overall mortality is expressed in the first 1~2 months after release and how much afterwards; the results indicate that total mortality still to occur after the first 1 month of life in the sea equals or exceeds that occurring to that time. Thus events later in the life history still have the potential to determine much of the declining marine survival of the salmon stocks we have examined.