A partnership between Oregon State University, Real Time Research, and the USGS - Oregon Cooperative Fish and Wildlife Research Unit


Predation on Salmonid Smolts by Piscivorous Waterbirds on East Sand IslandPredation on Salmonid Smolts by Piscivorous Waterbirds on East Sand Island

As a component of a comprehensive strategy for salmonid (Oncorhynchus spp.) recovery in the Columbia River Basin, management plans have been developed to reduce the impacts of double-crested cormorants (Phalacrocorax auritus) and Caspian terns (Hydroprogne caspia) nesting on East Sand Island on the survival of juvenile salmonids listed under the Endangered Species Act (ESA). Management initiatives are currently being implemented to reduce the size of the East Sand Island colonies through primarily lethal strategies (i.e., culling and egg oiling) for double-crested cormorants and dispersal of nesting birds to areas outside the Columbia River Basin for Caspian terns. The primary goal of this study was to estimate predation rates (percentage of available tagged fish consumed by birds) based on recoveries of passive integrated transponder (PIT) tags implanted in juvenile salmonids on the double-crested cormorant and Caspian tern colonies on East Sand Island. More specifically, the objective was to generate population-specific (salmonid evolutionary significant units [ESU] or distinct population segments [DPS]) predation rates on ESA-listed juvenile salmonids from the Columbia River Basin.

Predation rate estimates generated as part of this study integrated multiple factors of uncertainty in the tag recovery process, including imperfect detection of PIT tags on bird colonies, on-colony PIT-tag deposition probabilities, and temporal changes in fish availability to predators nesting on East Sand Island. Predation rates were used to compare and contrast impacts within a given fish population based on the fish’s rear-type (hatchery, wild) and migration history (in-river, transported). Predation rates were also used to compare and contrast smolt losses prior to and following bird management actions on East Sand Island; data critical in evaluating the effectiveness of management plans aimed at reducing predation rates. We also evaluated the relationship between predation rates and various biotic and abiotic conditions in the estuary that have potentially influenced these rates during 2000-2015. The goal of this analysis was to describe the influence of “management relevant” variables on predation rates, variables that resource managers may be able to control to some degree.

Predation rates indicated that impacts by double-crested cormorants on ESA-listed juvenile salmonids in 2015 were some of the highest ever recorded, with estimates of 14.5% (95% creditable interval [CI] = 10.5–22.4%) and 12.8% (95% CI = 9.3–19.6%) for Snake River spring/summer Chinook salmon (O. tshawytscha) and Snake River steelhead (O. mykiss), respectively. Impacts on salmon ESUs were comparable to those of steelhead DPSs in 2015, although the lowest rates were observed on two salmon populations (Snake River sockeye salmon O. nerka and upper Willamette River spring Chinook salmon, with < 2.5% of available fish consumed by cormorants in 2015). An investigation of temporal trends in predation rates indicated that double-crested cormorants consumed smolts in proportion to their relative availability, with the highest predation rates observed in May, when the largest numbers of PIT-tagged fish were available in the estuary. An investigation of predation rates dating back to 2000 indicated that smolt losses to double-crested cormorants were substantial in most years but highly variable over time. For example, annual predation rates by East Sand Island double-crested cormorants on Snake River steelhead have ranged from 1.9–16.6% for that particular DPS during 2000-2015. In comparison to smolt losses by double-crested cormorants, results indicated that Brandt’s cormorants (Phalacrocorax penicillatus) nesting on East Sand Island posed little risk to smolt survival in the estuary, with predation rates ≤ 0.6% for each salmonid ESU/DPS evaluated in 2015. Despite a steady increase in the size of the Brandt’s colony since it was first established on East Sand Island in 2006, predation rate estimates have remained below 1.0% for each salmonid ESU/DPS in all years.

Predation rates indicated that impacts by Caspian terns nesting on East Sand Island on ESA-listed juvenile salmonids were generally lower in 2015 compared with years past, with estimates ranging from 0.4% (95% CI = 0.1–1.5%) to 10.5% (95% CI = 8.2–15.0%) in Upper Willamette River spring Chinook and Upper Columbia River steelhead, respectively. Of those ESUs/DPSs evaluated, steelhead DPSs were predated at significantly higher rates than salmon ESUs. For instance, predation rates were 10.2% (95% CI = 8.2–14.6%) for Snake River steelhead but just 0.8% (95% CI = 0.2–1.5%) for Snake River fall Chinook in 2015. The finding that Caspian terns disproportionately consumed steelhead DPSs compared with salmon ESUs has been consistent since research was initiated on East Sand Island Caspian tern colony in 2000. Contrary to temporal results from double-crested cormorants, Caspian tern predation rates were the lowest when PIT-tagged smolt availability in the estuary was the highest; a finding consistent with predator-swamping, whereby an individual fish’s susceptibility to tern predation decreases when larger numbers of fish were available in the estuary.

If management is successful in significantly reducing the number of double-crested cormorants and Caspian terns that reside in the Columbia River Estuary during the spring/summer smolt outmigration, it is expected that there will be a commensurate reduction in smolt mortality. In 2015, the first year of double-crested cormorant management in the Columbia River Estuary, actions to reduce the number of double-crested cormorants nesting on East Sand Island were implemented after the peak of the smolt outmigration period. The percent of adult cormorants culled (ca. < 1% of adults) was also not large enough to have an effect in reducing cormorant predation. There was some evidence that management efforts to reduce the number of Caspian terns nesting on East Sand Island are associated with lower predation rates on ESA-listed salmonid populations. For instance, predation rates by East Sand Island terns on Upper Columbia River steelhead averaged 17.2% (95% CI = 15.7–19.3%) during 2000-2010, but were 9.9% (95% CI = 8.5–12.0%) during 2011-2015, associated in part with reductions in colony size due to tern nesting habitat restrictions on East Sand Island during the latter period. Despite the apparent reductions in Caspian tern predation associated with management, the target colony size for Caspian terns in the Columbia River Estuary has not been met, so the gains in survival of ESA-listed juvenile salmonids associated Caspian tern management have not been fully realized.

A relative comparison of predation impacts based on the fish’s rearing-type (hatchery, wild) indicated that hatchery and wild fish were equally susceptible to double-crested cormorant predation in the Columbia River estuary, with no consistent predation trend or preference by rear-type identified during 2006-2015. Unlike cormorants, there was evidence that Caspian terns disproportionately consumed hatchery spring/summer Chinook salmon compared with their wild counterparts during 2006-2015. Comparisons of predation impacts by a fish’s migration history (in-river, transported) also indicated differences in the relative susceptibility of smolts to bird predation, with transported Snake River fall Chinook and transported Snake River sockeye salmon disproportionately consumed by double-crested cormorants compared with in-river migrants during 2006-2015. There was also some evidence that in-river steelhead and in-river spring/summer Chinook were disproportionately consumed by Caspian terns and double-crested cormorants compared with transported fish but results were not consistent across all weeks and years. Data from this and other studies suggest the difference in fish susceptibility to bird predation is related to several behavioral and physical traits, including the size and condition of fish, the run-timing of fish, the abundance of fish, and predator-specific foraging techniques and behaviors. Additional research is needed to better understand these factors and how (or if) these factors can be managed to decrease predation rates on juvenile salmonids by double-crested cormorants and Caspian terns in the Columbia River estuary.

An evaluation of predation impacts from both double-crested cormorants and Caspian terns (i.e., cumulative predation rates) indicated that birds nesting on East Sand Island have annually consumed between 1.5–16.5% of available salmon per ESUs and between 12.0–40.6% of available steelhead per DPS during 2000-2015. Losses were comparable to or greater than those of other documented sources of smolt mortality (e.g., dam passage, predation by piscine predators) in the Columbia River Basin. Furthermore, impacts from bird predation in the estuary are on juvenile salmonids that have survived freshwater outmigration, including passage through the Federal Columbia River Power System, fish that have a higher probability of surviving than those that have yet to complete outmigration. Additionally, juvenile salmonids belonging to every ESA-listed ESU/DPS from the Columbia River Basin must pass through the Columbia River Estuary and are therefore susceptible to predation by birds nesting on East Sand Island. Despite recent reductions in the numbers of Caspian terns nesting on East Sand Island and predation rates by these birds on juvenile salmonids, the total number of birds (terns and cormorants combined) nesting on East Sand Island has remained fairly constant since 2000, as have average annual predation rates at the level of steelhead DPS and salmon ESU.

A multivariate analysis of factors influencing predation rates indicated that colony size was just one of several factors that explained variation in predation rates by Caspian terns and double-crested cormorants nesting on East Sand Island during 2000-2015. Fluctuations in large-scale climate indices (Multivariate El-Nino Index, North Pacific Gyre Oscillation), river operational strategies (spill, discharge), smolt abundance, and other factors were also important variables; demonstrating that a fish’s susceptibility to bird predation in the estuary is determined by a complex set of interacting factors. Results suggest that reductions in colony size will have to be both large and sustained before management goals to reduce avian predation can be fully realized in the Columbia River estuary.

-A. Evans, Q. Payton, B. Cramer, K. Collis, D. Lyons, and P. Loschl

Click here to see full report

Back to Feature Stories »