UW scientist Julia Parrish briefs Clallam County MRC on 26‑year COAST seabird monitoring program

Professor Julia Parrish of the University of Washington told the Clallam County Marine Resources Committee that COAST (Coastal Observation and Seabird Survey Team) is a 26‑year volunteer‑based seabird monitoring program that combines standardized beach surveys, paper data sheets and an online verification system to produce long‑term baselines for coastal bird mortality.

"We designed an early version of AI with actually accurate feedback," Parrish said, describing COAST’s verification workflow and training model, and adding that volunteers are the program’s ‘‘heart and soul.’’ She said COAST trains volunteers in the field, gives participants a protocol and field guide, and verifies volunteer identifications centrally.

Why it matters: Parrish said maintaining a long‑term baseline is crucial to detecting anomalies such as disease events, heat‑related die‑offs, or oil‑spill impacts. Using COAST data, her team compares year‑to‑year encounter‑rate graphs to identify unusually large mortality events and to estimate their scale and potential causes.

Local scale and notable records: Parrish reported that COAST’s database now contains hundreds of survey sites with active participation; she said 73 sites in Clallam County have been surveyed historically and that the county has contributed thousands of surveys to the regional dataset. She highlighted a local find: measurements from a photographed wing recovered on Hobuck Beach matched a purple gallinule, which Parrish described as the first recorded Washington State COAST record for that species.

On disease and large die‑offs, Parrish used two examples: a 2015–16 rhinoceros auklet die‑off around Protection Island tied to septicemia that depressed fledging success and colony occupancy for multiple years, and a larger 2014–16 common murre mortality event that, when modeled from colony and beach data, suggested multi‑million losses in some regions. "Those colonies in Alaska, they have not recovered," she said, noting regional differences in recovery and model uncertainty.

Parrish also flagged more recent mortality in Southern California linked to highly pathogenic avian influenza and said COAST is monitoring for similar signals on the outer Washington coast. She emphasized modeling uses beyond mortality detection—hindcasting oil‑spill paths, harmful‑algal‑bloom (HAB) work, bycatch and predator interaction studies—and said those applications make COAST data useful to managers and responders.

Questions and collaboration: During Q&A, Brian Tilbermann asked whether published murre mortality estimates matched the numbers Parrish discussed. Parrish acknowledged uncertainty in beach‑based recovery models and described how colony‑based techniques and follow‑up surveys informed larger Alaska estimates. She also described how COAST negotiates private‑property access and uses memoranda of understanding with tribes and landowners when sites are not publicly accessible.

Parrish suggested expanding a tribal intern program run from La Push and strengthening MRC‑COAST coordination for training, refresher courses and community events. "We should be celebrating all of those" forms of public engagement, she said, contrasting community science with large‑scale citizen‑science monitoring.

What’s next: Parrish said she would share the presentation slides with the MRC and welcomed follow‑up on partnerships, internships and additional refresher trainings for Clallam County volunteers.

Sources and limitations: The article draws exclusively on the meeting transcript. Several large mortality figures discussed by Parrish were presented as model estimates; Parrish acknowledged model uncertainty and sometimes treated numbers as approximate. Where the transcript contained inconsistent spellings or numbers, this article standardizes place names (Clallam County) and species names (e.g., purple gallinule) to conventional usage while noting when figures were described as estimates.