The current special habitats classification project started in 1994 as a non-forested habitat mapping and classification project. Fish and Wildlife staff developed a non-forest habitat classification and coding system as an adjunct to a forest classification and growth model system developed by Jim Arnie (at a time when commercial logging was expected to continue in the CRMW). The Arnie classification system used a nine-digit code, with the first four digits used to classify forested areas (called Species I codes) and the next five used for non-forested areas (Species II codes). Of the five non-forested classification digits, four described the area within the non-forested polygon, while the final digit identified the source of the species II code (photo interpretation, field checked, etc.).
The goal was to estimate the total amount of acres of all non-forested habitat types in the CRMW by making the shape of non-forested polygons correspond to boundaries seen on aerial photographs, and conducting a limited amount of field verification. In addition, a one-mile buffer around the CRMW was also classified, but not field verified. From 1994-2000 the GIS tbs99_a layer (a product of the Mason, Bruce and Girard classification of Landsat data) was modified by several CRMW Fish and Wildlife staff, as they interpreted aerial photographs and classified non-forested polygons using the species II codes. Results from this initial project were used in the development of the HCP.
The current project was renamed Special Habitats because it included some forested areas (e.g., forested wetlands, riparian areas). Because the Forest Ecology group was no longer using the Arnie classification system, a new classification and coding system, no longer limited to four digits, was developed (see excel file ‘special habitat codes’ described in data description section). It utilized a “key letter” approach, in which the habitat association was identified by the starting letter, with the second letter identifying the subtype (e.g., the key letter for meadows was M, with the upland meadow code = MU, alpine meadow = MA, and wet meadow = MW). All codes had two digits, and included association types plus individual species codes. A combination of the two digit codes was used to describe all special habitat polygons, with association type listed first and individual species codes (if any) listed in order of dominance. So the number of digits varied between polygons, depending on the number of dominant vegetation species present. A separate column in the attribute table identified whether the polygon was classified using photo interpretation or had been field verified (in which case the person was identified). Most polygons within the CRMW were field verified, while those within one mile of the watershed were classified using photo interpretation based on the field verification done within the CRMW.
The 1998 orthophoto of the CRMW was overlain with the GIS layer developed by earlier Fish and Wildlife staff, and existing GIS layers of roads, rivers, contours, and previously mapped wetlands were added and used for reference. Special habitat polygon boundaries were modified to match boundaries seen on the 1998 orthophoto, many polygons were split and more detailed boundaries drawn (e.g., areas of bare talus were separated from areas of shrub-covered talus), some polygons were deleted, and other added. During field verification, many small patches of special habitat (e.g., rock outcrops) were seen that had not been included from the photo interpretation. These were subsequently added and classified. All classification codes were changed from the Species II 4-digit system to the new coding system.
