Th-109-13
Bayesian Analysis of Size-Selective Pressure Exerted on in-Migrating Atlantic Salmon By Fish Passage Facilities on the Penobscot River, Maine

George Maynard , Wildlife, Fisheries, and Conservation Biology, University of Maine, Orono, ME
Joseph Zydlewski , U.S. Geological Survey, Maine Cooperative Fish and Wildlife Research Unit, Orono, ME
William Halteman , Math and Statistics, University of Maine, Orono, ME
Erik Blomberg , Wildlife, Fisheries, and Conservation Biology, University of Maine, Orono, ME
Successful upstream passage of Atlantic salmon through dams is critical to the long-term stability of sea-run populations. The influence of size on passage success of adult Atlantic Salmon was investigated at six fishways at hydropower dams in the Penobscot River. From 2002-2004 and 2010-2012, in-migrating adults (n = 4,610) were tagged with Passive Integrated Transponders (PIT) at Veazie Dam, the most seaward dam on the Penobscot River (rkm 47). Passage was monitored using PIT antenna arrays. For each dam, Bayesian logistic regression was used to analyze relationships between passage success and factors hypothesized to influence passage, including fish size, flow, water temperature, and river entry date. At passage facilities in the lower river, success was strongly negatively related to fork length; fish passing through Great Works Dam (rkm 59) were ≈ 5.5% less likely to pass for every 10cm increase in length. Thus, at that dam, a 75cm salmon was ≈ 18% less likely to pass than a 45cm salmon. The additive effects of navigating through multiple fishways may have effectively culled the largest fish from each spawning cohort. Such persistent reduction in access to spawning habitat of the largest adults likely impairs the probability of population persistence.