Redesigning Modern Portfolio Theory to Improve Spatial Recovery Planning for Oregon Coast (OC) Coho Salmon

Oregon Coast (OC) coho salmon (Oncorhynchus kisutch) are a federally listed threatened species under the Endangered Species Act. It is integral to conserve this species due to their ecological importance in nutrient cycling and cultural significance to Indigenous peoples. The combination of their threatened status and significance creates a sense of urgency for conservation organizations, like the Wild Salmon Center, to efficiently allocate their budgets. In this project we redesigned Modern Portfolio Theory (MPT) to optimize habitat restoration spending. MPT is traditionally used in finance to inform portfolio managers what the risks and returns are of investing in different portfolios of assets. In our redesigned application, the 21 populations of OC coho salmon are treated as assets, with the increase of salmon abundance and variance directly relating to the amount of money allocated to conserve each population. More specifically, we applied our new approach to mitigating barriers that inhibit salmon from traveling back to their natal streams. To do this, we gathered and collected data on fish passage barriers, average project costs to remove barriers, and estimated how removing barriers would affect the abundance of coho salmon. We analyzed portfolios under multiple budgets and scenarios that prioritize conservation spending in watersheds important to Indigenous peoples. This endogenous application is the first of its kind in the conservation field and can be applied to a multitude of species or restoration actions beyond OC coho salmon and barrier mitigation. 

Methods

Adult Coho Salmon Spawner data: This dataset was given to us by our client, the Wild Salmon Center, but was sourced through the Oregon Adult Salmonid Inventory & Sampling Project run by the Oregon Department of Fish & Widlife (ODFW).  
National Hydrography Reaches: All files were given to us by our client the Wild Salmon Center. Each file were given and downloaded as .shp files for each individual population in the OC coho Salmon Evolutionary Significant Unit (ESU). These files were created specifically for the Wild Salmon Center and NOAA by Terrainworks using USGS' National Hydrography Dataset. Many variables are outputs of different climate models ran by Terrainworks using the variables within USGS' orginal hydrography dataset. These .shp files were used to better understand stream habitat within the ESU.
Population Boundaries: All files were given to us by our client the Wild Salmon Center as a .shp file with all populations included. Each individual population was then exported to its own .shp file. All population boundaries were used to clip the reach data and barrier data in ArcGIS Pro to only look at ESU specific attributes.
Oregon Fish Passage Barriers: This dataset was sourced from ODFW's Fish Barrier Data basin. The barrer data for the state of Oregon was downloaded as a layer package. For our project, the layer package was clipped by each individual population boundary .shp files. This allowed us to isolate the barriers that are within the ESU.
Cost Data: The cost data was collected by downloading a .xlsx file from the Oregon Watershed Restoration Inventory and it managed by the the Oregon Watershed Enhancement Board. The inventory contains information on self reported restoration projects that have taken place in Oregon from 1995 - 2022. The file was then filtered by project type. All project types that involved bridges, culverts, dams, waterfalls, tide gates, fords, and cascades were used in our analysis. Each project type was given its own tab in the .xlsx file to calculate an average cost for each barrier type. The average cost was then applied to our barrier data where each barrier type received it's estimated cost to mitgate. This data informed our R model to estimate the impacts of barrier mitgation on adult salmon population returns.