Recovery of Endangered Species Act—listed salmonids in the Columbia River basin has relied upon the efficacy of the U.S. Army Corps of Engineer's juvenile salmon transportation program to move fish past Snake and Columbia River hydropower dams. The effectiveness of this program has been assessed by the indirect method of comparing smolt‐to‐adult returns. We present some of the first data and mortality estimates of barged and run‐of‐river (ROR) radio‐tagged juvenile spring–summer Chinook salmon Oncorhynchus tshawytscha after release in the lower Columbia River, representing years of study. Our data suggest that smolt mortality (1) is very low for ROR and barged fish between Bonneville Dam and the estuary proper, a migratory distance of 180 river kilometers (rkm); (2) occurs in the lower estuary (rkm 0–46); (3) varies more across dates within a year than between years or between passage types (barged or ROR); (4) increases with time within a season and increasing numbers of avian predators, including Caspian terns Sterna caspia and double‐crested cormorants Phalacrocorax auritus; and (5) is estimated to be 11–17% of all smolts annually. Preliminary evidence suggests that at least some smolt mortality is influenced by differential predation by avian predators on Chinook salmon infected with Renibacterium salmoninarum and possessing low smoltification levels (relatively low gill Na+,K+‐ATPase activity). Fish type (barged or ROR) did not appear to influence mortality because of avian predation. This project was also the first to identify avian predators as a major source of mortality for out‐migrant Columbia River basin salmonids.

No abstract

Migrating wild (W) and hatchery‐reared (H) chinook salmon Oncorhynchus tshawytscha and steelhead Oncorhynchus mykiss juveniles were sampled after loading into fish‐transport barges at Lower Granite Dam on the Snake River, Washington, and after barge transportation downstream to Bonneville Dam on the Columbia River. Stress indices (increased plasma cortisol and glucose concentrations and decreased plasma chloride concentrations) were higher (P < 0.001) for chinook salmon sampled during midseason (early to mid‐May), when fish loading densities in barges were at seasonal maximums, than were stress indices for those sampled earlier or later. Cortisol concentrations in chinook salmon were correlated with steelhead densities after loading of barges (P < 0.0001, R2 = 0.41) and after arrival of barges at Bonneville Dam (P < 0.0001, R2 = 0.65). Cortisol concentrations were not correlated with gill Na+, K+–adenosine triphosphatase activities, which were higher in W than in H fish of both species. Cortisol concentrations were higher (P < 0.0001 in 1994, P = 0.02 in 1995) in W than in H chinook salmon, and concentrations declined in both groups during barge transportation early and late in the migration season but not during midseason. In contrast, cortisol concentrations were lower (P < 0.001) in W than in H steelhead, were not correlated with steelhead loading densities, and declined in both W and H fish during barge transportation on all sampling dates. Electrolyte disturbances were greater in chinook salmon than in steelhead, but disturbances were similar for W and H fish of both species. Stress‐related water gain was, however, greater (or was compensated more slowly) in W than in H fish. These results indicate that chinook salmon are more stressed by barge transportation than are steelhead. If the viability of juvenile chinook salmon is reduced by adverse physiological, immunological, or behavioral responses to transportation stress, reductions in survival rates should be largest for fish transported during midseason, when densities of juvenile steelhead in the fish‐transport barges are highest.