Predictive Mapping of Seabirds, Pinnipeds and Cetaceans off the Pacific Coast of Washington

This report presents long-term seasonal distribution maps of selected seabird, pinniped and cetacean species off the Pacific Coast of Washington. The maps were created to support state-led marine spatial planning and responsible stewardship of natural resources by the Olympic Coast National Marine Sanctuary. They are intended to distinguish persistent areas of high relative density from areas of low relative density, and are useful for identifying ecologically important areas, recognizing and mitigating impacts from human uses and coastal hazards, and improving our understanding of marine environments. 

Predicted relative density distribution maps were constructed using associative models linking at-sea species observations with environmental covariates. Associative models relied on species observations compiled from federal, state, and non-governmental monitoring programs with data between 1995 and 2014. Environmental variables, such as depth, sea surface temperature, and indices of primary productivity, were processed from long-term archival satellite, oceanographic, and hydrographic databases.

The compilation of at-sea species observations represents the first attempt to combine eleven selected survey programs, and is a substantial combination of nearshore and offshore survey effort. As far as we are aware, the compilation prepared for this report is the largest synthesis of recent seabird, pinniped, and cetacean observations in the study area, in terms of both number of observations and number of programs combined. 

A boosted generalized additive modeling framework was applied to associate seabird and environmental covariate data sets and develop contiguous, accurate predictions of relative density. To improve model performance, the modeling framework allowed for flexible relationships and multi-way interactions between environmental variables while accounting for sampling heterogeneity between and within datasets.

Model performance was assessed using cross validation and a range of model fit and bias diagnostics. All models showed good performance based on model performance diagnostics, and expert reviewers agreed all maps were valuable representations of species distributions. Reviewers included ecologists, coastal resource managers, and modelers from multiple agencies and organizations. 

These maps represent an important step towards improving our understanding of the long-term spatial distributions of selected seabirds, pinnipeds, and cetaceans, identifying persistent hotspots of relative densities, and more effectively planning offshore human activities. The seabird, pinniped, and cetacean predictions are already being used by the Washington Department of Fish and Wildlife to identify ecologically important areas off the Pacific Coast of Washington, and they intend to utilize this information in planning for offshore renewable energy development.


Menza, M., J.B. Leirness, T. White, A. Winship, B.P. Kinlan, L. Kracker, J.E. Zamon, L. Ballance, E. Becker, K.A. Forney, J. Barlow, J. Adams, D. Pereksta, S. Pearson, J. Pierce, S.J. Jeffries, J. Calambokidis, A. Douglas, B.C. Hanson, S.R. Benson, and L. Antrim. 2016. Predictive Mapping of Seabirds, Pinnipeds and Cetaceans off the Pacific Coast of Washington. NOAA Technical Memorandum NOS NCCOS 210. Silver Spring, MD. 96pp. doi:10.7289/V5NV9G7Z

Download PDF