Avian Predation on Juvenile Salmonids at John Day and the Dalles Dams on the Columbia River in 2010

Background: Avian predators are one highly visible cause of smolt mortality at hydropower dams. In the Columbia basin, if sever enough this predation may prevent a dam from meeting its Biologic Opinion required survival goals for ESA listed salmonid passage. In 2010 our task was determining the impact of avian predators on fish passing John Day and The Dalles Dams. Our objectives were: 1) Determine species composition and numbers of piscivorous birds; 2) Estimate smolt consumption by gulls; 3) Determine the effectiveness of intense boat hazing and a new synthetic avian line array at John Day Dam.

Methods: To quantify avian consumption observers used binoculars to count gulls, the rate of attacks (dives), and determine if an attack was successful (fish in bill) during the smolt outmigration between 8 April to 28 July 2010.  We then estimated salmonid consumption using those variables and diet information from weekly gull stomach collections.  Additionally, counts were collected on a much smaller population of other fish eating birds.

Results: California gulls were the dominant piscivore at both dams followed by western grebes, American white pelicans,  double crested cormorants, Caspian terns, osprey, and great blue heron. The daily abundance of gulls at John Day ranged from zero on 8 April to a brief high of 118 on 14 June, declining rapidly to two on 17 June, yielding a seasonal mean of 17.  At The Dalles gull daily abundance ranged from zero on 8 April to a high of 133 on 19 May, slowly decreasing to four on 27 July, yielding a seasonal mean of 34.

Overall, 349 California Gull stomachs were collected. Stomachs from John Day Dam contained 93 salmonids, seven lamprey, three other fish, eight unidentified fish, and insects.  We also recovered 14 PIT tags, eight of which were broken, thus un-readable tags.  Gull stomachs from The Dalles Dam contained 72 salmonids, 95 lamprey, and three unidentified fish as well as five readable PIT tags and landfill starches.

At John Day Dam our estimate of smolt consumption, which includes additive and compensatory sources of mortality, was 18,000 (11,000 - 26,000 95% CI). This is a reduction of 61,000 (76%) from 2009 when 80,000 smolt were consumed and is attributed to intensive boat hazing and a large avian deterrent line array. At The Dalles Dam our estimate of smolt consumption, which includes additive and compensatory sources of mortality, was 98,000 (58,000 - 145,000 95% CI).