Passive integrated transponder (PIT) tags implanted in Columbia River basin juvenile salmonids Oncorhyncus spp. were recovered from breeding colonies of Caspian terns Sterna caspia and double‐crested cormorants Phalacrocorax auritus on Rice Island, a dredge spoil island in the Columbia River estuary. Tags were recovered to assess the relative vulnerability of different salmonid species, stocks, and rearing types to avian predators. We detected 50,221 PIT tags at the two bird colonies, mostly from juvenile chinook salmon O. tshawytscha and steelhead O. mykiss raised in hatcheries; 72% of the total tags were from the tern colony and 28% from the cormorant colony. Tagged steelhead smolts were more vulnerable to predation by both bird species than were yearling chinook salmon. More than 15% of PIT tags from steelhead smolts that were available in the estuary in 1998 were detected at the bird colonies compared with 2% of PIT tags from yearling chinook salmon. The greater vulnerability of steelhead may reflect size‐dependent selection by avian predators. Salmonids listed under the Endangered Species Act and unlisted salmonids were equally vulnerable to predation by both terns and cormorants. Hatchery‐raised yearling chinook salmon were more vulnerable than their wild counterparts to predation by terns, a surface‐feeding species; however, hatchery‐raised and wild yearling chinook salmon were equally vulnerable to predation by cormorants, a diving species. These results suggest that hatchery‐raised yearling chinook salmon, and hatchery‐raised steelhead in some years, are more vulnerable to tern predation than wild fish because they have a greater tendency to reside near the water surface where terns forage.

We estimated the consumption of juvenile salmonids (Oncorhynchus spp.) and other forage fishes by Caspian terns (Sterna caspia) nesting on Rice Island in the Columbia River estuary in 1997 and 1998 using a bioenergetics modeling approach. The study was prompted by concern that Caspian tern predation might be a substantial source of mortality to out-migrating juvenile salmonids from throughout the Columbia River basin, many populations of which are listed as threatened or endangered under the U.S. Endangered Species Act. The bioenergetics model used estimates of the energy requirements of the tern population and the proportion of tern energy requirements met by various prey types. The resulting estimate of the number of juvenile salmonids consumed by Rice Island Caspian terns was 8.1 million (5.9–10.4 million) in 1997 and 12.4 million (9.1–15.7 million) in 1998. Tern predation rates on juvenile salmonids were substantial, representing up to 15% of the juveniles to reach the estuary from some listed populations. Nevertheless, based on simple age-structured models of salmonid populations, it appears unlikely that management of Caspian tern predation alone would reverse salmonid declines. Management to reduce tern predation could, however, contribute to a comprehensive strategy to recover imperiled salmonid populations in the Columbia River basin.

We recovered passive integrated transponder (PIT) tags from nine piscivorous waterbird colonies in the Columbia River basin to evaluate avian predation on Endangered Species Act (ESA)‐listed salmonid Oncorhynchus spp. populations during 2007–2010. Avian predation rates were calculated based on the percentage of PIT‐tagged juvenile salmonids that were detected as passing hydroelectric dams and subsequently were consumed and deposited by birds on their nesting colonies. Caspian terns Hydroprogne caspia (hereafter, “terns”) and double‐crested cormorants Phalacrocorax auritus (hereafter, “cormorants”) nesting on East Sand Island in the Columbia River estuary consumed the highest proportions of available PIT‐tagged salmonids, with minimum predation rates ranging from 2.5% for Willamette River spring Chinook salmon O. tshawytscha to 16.0% for Snake River steelhead O. mykiss. Estimated predation rates by terns, cormorants, gulls of two species (California gull Larus californicus and ring‐billed gull L. delawarensis), and American white pelicans Pelecanus erythrorhynchos nesting near the confluence of the Snake and Columbia rivers were also substantial; minimum predation rates ranged from 1.4% for Snake River fall Chinook salmon to 13.2% for upper Columbia River steelhead. Predation on ESA‐listed salmonids by gulls and American white pelicans were minor (<2.0% per ESA‐listed salmonid population) relative to predation by terns and cormorants. Cumulative impacts were greater for Snake River and upper Columbia River salmonids than for salmonids originating closer to the estuary because upriver salmonids must migrate past more bird colonies to reach the ocean. Predation rates adjusted for colony size (per capita rates) were significantly higher for terns and cormorants nesting at inland colonies (upstream of Bonneville Dam) than for those nesting in the estuary, suggesting that inland colonies have a greater reliance on salmonids as a food source. Management actions to increase salmonid survival by reducing avian predation in the estuary could be offset if birds that disperse from the estuary relocate to inland nesting sites on or near the Columbia River.

We investigated colony size and diet composition of piscivorous waterbirds (gulls, terns, and cormorants) nesting on the lower Columbia River from the mouth (river km 0) to the head of McNary Pool (river km 553) in 1997 and 1998. The study was prompted by concern that avian predation might constitute a significant source of mortality to juvenile salmonids Oncorhynchus spp. during out‐migration. The diet of California gulls Larus californicus and ring‐billed gulls L. delawarensis nesting in colonies above The Dalles Dam (river km 308) included few fish and very few juvenile salmonids. The sole exception was a small colony of California gulls in which salmonids accounted for 15% (by mass) of the diet. Juvenile salmonids were, however, an important component of the diet of colonial waterbirds nesting in the Columbia River estuary. On Rice Island (river km 34), salmonids accounted for 74% (by mass) of the diet of Caspian terns Sterna caspia, 46% for double‐crested cormorants Phalacrocorax auritus, and 11% for glaucous‐winged−western gulls L. glaucescens × L. occidentalis. Juvenile salmonids were especially prevalent in the diets of colonial waterbirds on Rice Island during April and May. By comparison, juvenile salmonids were significantly less prevalent in the diet of cormorants and gulls nesting lower in the estuary on East Sand Island (river km 8), presumably due to the greater availability of marine forage fishes. Our results indicate that avian predation on juvenile salmonids in the lower Columbia River is more prevalent in the estuary than near the large upriver gull colonies. Furthermore, the high incidence of salmonids in the diets of Caspian terns, cormorants, and gulls nesting on Rice Island suggests that the impact of avian predation on survival of smolts may be reduced by discouraging piscivorous birds from nesting there, while encouraging nesting on East Sand Island and other sites nearer to marine foraging areas.