Process Description:
CHANJ Version 1.0 identifies terrestrial wildlife habitat cores, corridors, and road mitigation opportunities. It was developed by first creating a raster base habitat layer statewide with values from 0 (non-habitat) to 1 (habitat). The components of the base layer were: A) LULC classes coded as Habitat (1), Marginal Habitat (0.5), and Non-Habitat (0), and B) Wetlands/Riparian Habitat (1) composed of the Landscape Project Riparian Corridor, Flood Prone areas, Hydric soils, and LULC Wetlands, with LULC Urban removed. The two layers were combined with the maximum value of the cell kept. Next, Railroads (0), and Roads (0) were added in, taking the minimum cell value. Core Mapper (in the Gnarly Landscape Utilities ArcGIS toolbox (http://www.circuitscape.org/gnarly-landscape-utilities) was used to model core areas. The inputs for the modeling included: 1) the base habitat layer, 2) a moving window radius of 500m, which represents the 75th percentile of average home range sizes of all 126 target CHANJ species, 3) a minimum average habitat value of 0.69, which was informed by analyses of species location data and their average habitat values, 4) county roads, highways, and interstates as core barriers, and 5) a minimum core threshold size of 78.5 ha, which represents the 75th percentile of average home range sizes of all 126 target CHANJ species. Further refinement included removing core areas that did not meet the minimum threshold size after subtracting out area of Marginal Habitat and open water. To model the corridors, or continuous swaths of habitat between cores, we used the Linkage Mapper ArcGIS toolbox (http://www.circuitscape.org/linkagemapper) and coarsened the grid size to 20m to ease computation time. We first created a resistance layer by using the base habitat layer from the core mapping to calculate an average habitat value grid using a 100m radius moving window size. We then used the following equation (99 * (1 – [ave_hab_grid])) + 1) to reverse the values and convert average habitat values to resistance values such that the resultant raster had values from 1 (low resistance; easier to move through) to 100 (high resistance; harder to move through). We made additional modifications to account for landscape features that represent high resistance to movement, but were not yet reflected in the resistance grid. LULC classes were coded to extract water bodies (See Appendix __) and were given a high resistance value (100), and differing levels of roads were given the following resistance values: Local (25), County (50), Highways (50), High Traffic Volume (>10,000 vehicles/day) (75). The inputs for the cost-weighted distance and least-cost path modeling included: 1) the resistance layer and 2) the final core areas. We selected a cost distance cut-off threshold of 16.76km for the corridor mapping, and erased the core areas from the resultant corridor mapping to remove overlap. We classified the corridor mapping into 5 equal interval bins. Finally, we coded roads into 3 categories: 1) Local, 2) County roads and Highways with <10,000 vehicles/day and 3) County roads and Highways with >10,000 vehicles/day and selected just those road segments that intersected with the corridor and corridor mapping and from those deleted any road segments adjacent to LULC Urban areas on one or both sides. The resultant road segments were combined with the core and corridor mapping and all three components were converted to polygons for the final product.