Biodiversity Guide - The Intertwine

appendix B Habitat Prioritization ModelingB. Habitat Prioritization ModelingIn November 2010, when The Intertwine Alliance launched theeffort to produce a regional conservation strategy and biodiversityguide for the greater Portland-Vancouver region, therewas no data-driven map of priority areas for conservation thatadequately covered both the urbanized and rural portions ofthe region. Previous efforts either worked on a larger regionalscale that for the most part discounted the habitat value ofurban areas (this was the case with the state conservationstrategies and Willamette Valley Synthesis Project), focusedon localized geographies and abruptly ended at jurisdictionalboundaries (e.g., Title 13), or covered most of the region butwere based solely on expert opinion (e.g., the Natural Featurescharrette process). The goal of the Regional ConservationStrategy for the Greater Portland-Vancouver Region and theaccompanying Biodiversity Guide was to add a unified regionalperspective to local efforts and to encourage a shared visionthat could facilitate cooperation to protect remaining valuablehabitat.We aimed to develop data-driven, science-based scalablemodels for determining the relative conservation valueof habitat in a way that would complement and support theRegional Conservation Strategy and accompanying BiodiversityGuide. We also wanted to (1) represent urban habitat in a waythat makes the best fine-scale habitat within or near urbanareas “competitive” with large, intact habitat blocks in theurban fringe, (2) cooperate with stakeholders to ensure theirbuy-in on the resulting product, and (3) create a foundationof work that others throughout the region could use for futureconservation efforts, such as wildlife connectivity mapping andconservation and restoration prioritization.Overall ApproachThe modeling effort was overseen by the GIS Subcommittee ofthe Regional Conservation Strategy (RCS) Steering Committee,which included representatives of federal, state, and localjurisdictions and nonprofit organizations. The Institute forNatural Resources (INR) conducted the primary data developmentand modeling with input from the GIS Subcommittee.INR provided multiple drafts and iterations for review by theGIS Subcommittee and RCS Steering Committee. In establishingthe criteria, methods, and threshold values for the models,the modeling team took into consideration the results ofextensive stakeholder consultation and basic conservationscience principles and incorporated scientific expertise.The modeling effort produced two regional map outputs(with accompanying GIS data): a high-value habitat map (FigureB-1) and a riparian habitat map (Figure B-2). Each mapwas based on a distinct set of criteria for relative habitat value.We used a raster-based analysis format to map and analyzethe region as square pixels in a rectangular grid. Each pixel wasscored uniquely based on the science-based criteria. A highresolution(5-meter) regional land cover map that INR createdfor The Intertwine Alliance (Figure A-4) served as a foundationaldata set for several criteria, but the models also requiredregional data on wetlands, bodies of water, floodplains, soiltypes, and roads. In several cases, we faced a tradeoff betweenusing the best available local data and creating or using aregionally consistent data set. In general, we used or createddata sets that provided consistent spatial information acrossthe region. In limited cases, such as with wetlands, we integratedlocal and regional data sets to produce a composite thatwe thought was more accurate while still reasonably consistent.Compiling data from numerous sources can cause variableresults. For example, the density of mapped wetlands in ClarkCounty is higher than in the rest of the region in part becauseof the mapping methods used to compile this data set.Two Habitat Models, Two Sets of CriteriaThe approach used to determine the conservation value ofhabitats consisted of developing two separate models—one forthe entire region (the high-value habitat model) and one forriparian areas (the riparian habitat model). For each model,the modeling team developed spatial data sets that representedcriteria for calculating the value of habitat.171