PNNL

Abstract

Ship wakes produced by deep-draft vessels transiting the lower Columbia River have been observed to cause stranding of juvenile salmon. Proposed deepening of the Columbia River navigation channel has raised concerns about the potential impact of the deepening project on juvenile salmon stranding. The Portland District of the U.S. Army Corps of Engineers requested that the Pacific Northwest National Laboratory design and conduct a study to assess stranding impacts that may be associated with channel deepening. The basic study design was a multivariate analysis of covariance of field observations and measurements under a statistical design for a before and after impact comparison. We have summarized field activities and statistical analyses for the “before” component of the study here. Stranding occurred at all three sampling sites and during all three sampling seasons (Summer 2004, Winter 2005, and Spring 2005), for a total of 46 stranding events during 126 observed vessel passages. The highest occurrence of stranding occurred at Barlow Point, WA, where 53% of the observed events resulted in stranding. Other sites included Sauvie Island, OR (37%) and County Line Park, WA (15%). To develop an appropriate impact assessment model that accounted for relevant covariates, regression analyses were conducted to determine the relationships between stranding probability and other factors. Nineteen independent variables were considered as potential factors affecting the incidence of juvenile salmon stranding, including tidal stage, tidal height, river flow, current velocity, ship type, ship direction, ship condition (loaded/unloaded), ship speed, ship size, and a proxy variable for ship kinetic energy. In addition to the ambient and ship characteristics listed above, site, season, and fish density were also considered. Although no single factor appears as the primary factor for stranding, statistical analyses of the covariates resulted in the following equations: •Stranding Probability ~ Location + Kinetic Energy Proxy + Tidal Height + Salmonid Density + Kinetic energy proxy × Tidal Height + Tidal Height × Salmonid Density. •Stranding Probability ~ Location + Total Wave Distance + Salmonid Density Index. •Log(Total Wave Height) ~ Ship Block + Tidal Height + Location + Ship Speed. •Log(Total Wave Excursion Across the Beach) ~ Location + Kinetic Energy Proxy + Tidal Height The above equations form the basis for a conceptual model of the factors leading to salmon stranding. The equations also form the basis for an approach for assessing impacts of dredging under the before/after study design.