Foraging and Dispersal of Managed Caspian Terns on the Columbia Plateau
The primary objective of this study in 2015 was to monitor and evaluate management initiatives implemented to reduce predation on U.S. Endangered Species Act (ESA)-listed populations of salmonids (Oncorhynchus spp.) by Caspian terns (Hydroprogne caspia) nesting on Goose Island in Potholes Reservoir and on Crescent Island in the mid-Columbia River. Specifically, this study was designed to evaluate dispersal of Caspian terns dissuaded from nesting on Goose and Crescent islands and changes in Caspian tern predation rates (number consumed/number available) on juvenile salmonids as a consequence of management.
In January 2014, the U.S. Army Corps of Engineers – Walla Walla District (Corps) completed the Inland Avian Predation Management Plan (IAPMP). The goal of the IAPMP is to reduce Caspian tern predation rates on ESA-listed salmonids from the Columbia River Basin to less than 2% (per colony and per ESA-listed population) by redistributing Caspian terns from the two largest colony sites in the Columbia Plateau region (i.e. colonies on Goose Island in Potholes Reservoir and on Crescent Island in the mid-Columbia River) to sites outside the Columbia River Basin. In 2015, the Corps and the U.S. Bureau of Reclamation (BOR) implemented Phase II of the IAPMP by installing a variety of “passive nest dissuasion” materials prior to the 2015 nesting season that were designed to prevent Caspian terns from nesting on either island. An effort was also made to prevent nesting by the two species of gulls that nest abundantly on both islands (California gulls [L. californicus] and ring-billed gulls [L. delawarensis]), on the theory that nesting gulls would attract prospecting Caspian terns and could limit the efficacy of efforts to dissuade Caspian terns from nesting on the two islands. Once Caspian terns and gulls arrived on Goose and Crescent islands to initiate nesting, active nest dissuasion (i.e. human hazing) was used in an attempt to dissuade both Caspian terns and gulls from nesting anywhere on either island.
Despite the lack of suitable Caspian tern nesting habitat on Goose Island in 2015, some Caspian terns displayed persistent fidelity to the Pothole Reservoir area throughout the nesting season, likely due to the history of Caspian tern nesting on Goose Island since 2004 and the presence of a large breeding colony of gulls on the island that continued to attract prospecting Caspian terns to the site. Another factor that might explain the strong site fidelity of Caspian terns to the Potholes Reservoir area is the paucity of alternative Caspian tern colony sites in the vicinity. Caspian tern use of Goose Island for roosting and nesting during 2015 was largely limited to areas near the shoreline where passive nest dissuasion had not been installed. Active nest dissuasion (hazing), collection of any Caspian tern eggs that were discovered, and high rates of gull predation on newly-laid Caspian tern eggs were successful in preventing the formation of a Caspian tern colony on Goose Island in 2015. A total of 43 Caspian tern eggs were found on Goose Island and nearby islets in 2015, and these eggs were laid in 39 different nest scrapes. Seventeen Caspian tern eggs were collected under permit, 23 tern eggs were depredated by gulls soon after they were laid, and three tern eggs ultimately produced chicks. In the end, only two pairs of Caspian terns nesting in separate areas of the island succeeded in hatching eggs and raising young on Goose Island and nearby islets in 2015.
Passive and active nest dissuasion techniques were successful in preventing nesting and roosting by both Caspian terns and gulls on Crescent Island in 2015. This result was somewhat unexpected because it was the first year that nest dissuasion was implemented at Crescent Island and because Caspian terns and gulls have nested consistently on Crescent Island for nearly three decades. One factor that likely contributed to the absence of nesting Caspian terns on Crescent Island was the use of closely-spaced fence rows of privacy fabric as passive dissuasion over much of the suitable Caspian tern nesting habitat on Crescent Island; similar fencing was not deployed at Goose Island due to shallow, rocky soils. Another factor was the successful dissuasion of all gulls from nesting on Crescent Island in 2015; gulls are breeding associates of Caspian terns and attract prospecting Caspian terns to nest near their colonies. At Goose Island, gull nesting could not be prevented using the passive and active dissuasion techniques at our disposal, whereas at Crescent Island gulls never habituated to our hazing techniques and abandoned Crescent Island to establish a new colony on Badger Island (located on the mid-Columbia River just one kilometer upstream from Crescent Island) in 2015. Similarly, Caspian terns displaced from Crescent Island were able to relocate to an alternative colony site on the mid-Columbia River, the Blalock Islands (70 river kilometers downriver from Crescent Island), where Caspian terns have nested in small numbers over the last decade.
The total estimated breeding population of Caspian terns in the Columbia Plateau region during 2015 was 769 breeding pairs at five separate colonies (i.e. the Blalock Islands on the mid-Columbia River [677 breeding pairs], Twinning Island in Banks Lake [64 breeding pairs], Harper Island in Sprague Lake [10 breeding pairs], an unnamed island in Lenore Lake [16 breeding pairs], and Goose Island in Potholes Reservoir [2 breeding pairs]). The estimated total size of the breeding population of Caspian terns in the Columbia Plateau region in 2015 was similar to the estimated population size in 2014 (755 breeding pairs), but still generally lower than the numbers observed during 2000-2013. These results suggest that although nest dissuasion actions implemented on Goose and Crescent islands in 2015 were highly effective in reducing the numbers of Caspian terns nesting at these two colonies, formerly the largest Caspian tern colonies in the region, they did not result in a significant reduction in the total number of Caspian terns breeding in the region to date. This was due to the more than 10-fold increase in the number of Caspian tern nesting in the Blalock Islands in 2015 compared to 2014. The Blalock Islands colony in 2015 was similar in size to the largest Caspian tern colony recorded anywhere in the Columbia Plateau region since intensive monitoring began in 2000.
Juvenile salmonids made up 67.3% of the diet of Caspian terns nesting on the Blalock Islands in 2015; this is consistent with results from previous years for Caspian terns nesting at Crescent Island. However, a larger proportion of the salmonids in the Blalock Islands tern diet were steelhead (34%) compared to the tern diet at Crescent Island. We estimated that Caspian terns nesting on the Blalock Islands consumed ca. 550,000 juvenile salmonids in 2015 (95% CI = 310,000 – 800,000), including ca. 240,000 steelhead (95% CI = 130,000 – 350,000). Steelhead consumption by terns nesting at the Blalock Islands colony in 2015 was likely greater than at either Crescent Island or Goose Island in any previous year for which results are available.
After two years of implementation of the IAPMP, satellite-tracking of tagged Caspian terns has indicated several broad categories of response to management: (1) stay and search or compete for nest sites in reduced habitat, (2) move to a nearby colony and attempt to nest there, returning to the colony of origin if nesting fails, (3) engage in a long distance dispersal to a more favorable colony, or (4) wander nomadically across the Columbia Plateau region or a much larger area. Terns tagged at Potholes Reservoir in 2014 or 2015 have generally stayed nearby and searched for nesting habitat, moved to nearby colonies (Banks Lake, Sprague Lake) and returned to Potholes Reservoir when those colonies failed, or wandered nomadically, often across large portions of Washington, Oregon, and northeastern California. Terns tagged at Crescent Island in 2015 primarily moved to a nearby colony (the Blalock Islands), but a few individuals exhibited long distance dispersal to the Columbia River estuary. In 2015, several tagged terns visited constructed islands in interior Oregon and northeastern California, and a few of these visits were sustained and suggested possible nesting attempts. We did not detect any movement of satellite-tagged terns to the newly constructed tern islands at Don Edwards NWR in the San Francisco Bay area during 2015.
Drought across the Pacific Northwest likely limited dispersal of Caspian terns from the Columbia Plateau region to the islands constructed in interior Oregon and northeastern California. Low flows in the Columbia River led to greater availability of nesting habitat at the Blalock Islands due to exposure of low-lying islands that terns nested on there. Low winter snowpack in interior Oregon and northeastern California resulted in low water levels and limited nesting and foraging habitat, particularly at Crump Lake in the Warner Valley and Malheur Lake in the Harney Basin. Expected relaxation of regional drought conditions in 2016 may make colony locations outside of the Columbia River Basin more attractive as nesting locations to prospecting terns displaced from the Columbia Plateau region by management.
Resightings of Caspian terns that were previously color-banded indicated that some terns exhibited site fidelity to the Potholes Reservoir area, while the Blalock Islands experienced a large influx of terns from the Crescent Island colony in 2015. Evaluation of inter-regional movements of Caspian terns in 2015 revealed net movements to the Columbia Plateau region from the managed colony at East Sand Island in the Columbia River estuary, as well as from the Corps-constructed alternative colony sites in interior Oregon and northeastern California; the latter two regions experienced severe drought in 2015.
Recoveries of passive integrated transponder (PIT) tags on Caspian tern colonies were used to estimate predation rates (percentage of tagged fish consumed by terns) to evaluate the efficacy of tern management initiatives to increase smolt survival in the region. PIT tag data were also used to evaluate smolt consumption rates by other piscivorous colonial waterbirds (e.g., California gulls, ring-billed gulls, American white pelicans [Pelecanus erythrorhynchos], double-crested cormorants [Phalacrocorax auritus]) on juvenile salmonids. Results indicate that management efforts to reduce the size of the Goose Island and Crescent Island Caspian tern colonies were successful in nearly eliminating predation by terns from these two colonies in 2015, with predicted Caspian tern predation rates ranging from < 0.1% to 1.5% (depending on the ESA-listed salmonid population) at Goose Island and < 0.1% (for all salmonid populations) at Crescent Island. This likely was the first time since the Crescent Island colony of Caspian terns formed in 1986 when no salmonid smolts were consumed by Caspian terns nesting on Crescent Island. Predation rates on juvenile salmonids by Caspian terns nesting on Goose Island were lower in 2015 than in 2014 (< 1% to 2.9%), the first year of tern management, and significantly lower than predation rates observed during 2007-2013, prior to tern management. For instance, predation rates by Goose Island terns on ESA-listed Upper Columbia River spring Chinook salmon and steelhead averaged 2.5% and 15.7%, respectively, during 2007-2013.
Despite a dramatic reduction in predation rates on smolts by Caspian terns nesting on Goose and Crescent islands in 2015, a significant increase in predation rates was observed for Caspian terns nesting at the Blalock Islands (John Day Reservoir) and at Twinning Island (Banks Lake), colonies where terns dissuaded from Goose and Crescent islands relocated following the implementation of management actions in 2015. Predation rates were highest on steelhead, with an estimated 8.2% (95% CI = 5.9–12.4%) of Upper Columbia River steelhead consumed by Blalock Islands terns and 2.6% (95% CI = 1.8–3.9%) of Upper Columbia River steelhead consumed by Twinning Island terns in 2015. These predation rates exceeded the IAPMP target goal of < 2% per ESA-listed salmonid population for these two tern colony sites. Predation rates by Caspian terns nesting at the Blalock Islands were also substantial for Snake River steelhead, with an estimated 8.0% (95% CI = 6.0–11.8%) consumed by terns in 2015, the highest predation rate on Snake River steelhead recorded for any Caspian tern colony in the Columbia Plateau region since 2007. Predation rates on salmon populations were significantly lower than those on steelhead populations, with predation rates of < 2.0% for all ESA-listed salmon populations evaluated in 2015. Predation rates by Caspian terns nesting at all colonies in the Columbia Plateau region combined during 2015 were similar to or higher than those observed in previous years due to the large and unprecedented number of Caspian terns (677 breeding pairs) that nested at the Blalock Islands in 2015.
Estimates of consumption rates of juvenile salmonids by gulls nesting at certain colonies in the Columbia Plateau region were also substantial in 2015, particularly consumption by California gulls nesting on Miller Rocks (The Dalles Reservoir), Island 20 (McNary Reservoir), and the Blalock Islands. Similar to predation on smolts by Caspian terns, consumption rates by gulls were generally highest on steelhead populations relative to salmon populations, with the highest consumption rates observed on Upper Columbia River steelhead by gulls nesting on Miller Rocks (13.2%; 95% CI = 8.3–21.1%), Island 20 (7.9%; 95% CI = 5.2–12.0%), and the Blalock Islands (6.1%; 95% CI = 3.4–10.5%). Gull consumption rates of Snake River steelhead were also substantial (9.7% by gulls nesting on Miller Rocks), but lower than rates observed for Upper Columbia River steelhead in 2015. Consumption rates of salmon populations by gulls nesting at colonies in the Columbia Plateau region were generally < 2.0%, with the exception of consumption of Upper Columbia River spring Chinook salmon (3.5%; 95% CI = 2.1–6.0%) and Snake River sockeye salmon (7.4%; 95% CI = 4.1–13.1%) by gulls nesting on Miller Rocks. Consumption rates by gulls from colonies in the Columbia Plateau region during 2015 were significantly higher than those observed at the same gull colonies in previous years, with a roughly 2- to 5-fold increase observed at some gull colonies in 2015. Further research is needed to better understand the mechanisms that influence fish susceptibility to consumption by gulls and why consumption rates were significantly higher in 2015, a year of reduced river flows and increased water temperatures, compared with previous years. Regardless of the reasons, smolt consumption rates associated with certain gull colonies were comparable to or higher than predation rates associated with Caspian tern colonies in 2015, and were some of the highest consumption rates associated with any piscivorous waterbird colony in the Columbia Plateau region since 2007. Predation by American white pelicans nesting at the Badger Island colony was, however, low (< 1.0% per ESA-listed salmonid population), indicating that pelicans nesting at this colony posed little risk to smolt survival in 2015.
To better understand the spatial and temporal distribution of smolt consumption by piscivorous colonial waterbirds in the middle Columbia River (i.e. from the tailrace of Rock Island Dam to just upstream of the confluence with the Snake River), detections/recoveries of acoustic tags and PIT tags implanted in juvenile steelhead and sockeye salmon (i.e. double-tagged smolts) were used to evaluate predation rates, specifically within the Priest Rapids Project. Results were based on last known detections of live fish passing telemetry arrays, coupled with the recovery of PIT tags from these fish on nearby bird colonies. Results indicate that an estimated 2.7% (95% CI = 1.7–4.4%) and 2.3% (95% CI = 0.2–4.7%) of tagged steelhead were consumed by Caspian terns within the Wanapum and Priest Rapids developments, respectively. Predation rates in 2015 were significantly lower than those observed prior to implementation of the IAPMP (4.0–10.1% per development, depending on the year). Avian predation rates on double-tagged sockeye salmon in 2015 (the only year analyzed) were estimated at 1.2% (95% CI = 0.5–2.3%) and 0.7% (95 CI = 0.1–1.6%) in the Wanapum and Priest Rapids developments, respectively, indicating that avian predators posed little risk to tagged sockeye smolts traveling through the middle Columbia River in 2015. Reductions in the number of Caspian terns nesting within foraging distance of the Priest Rapids Project is likely a contributing factor to recent improvements in survival of juvenile salmonids, particularly survival of steelhead, in the middle Columbia River.
-Bird Research Northwest
Click here to see full report