Description:This layer is the land-cover/land use data compiled for the California Gap Analysis Project. It contains vegetation attributes for landscape scale map units, including canopy dominant species, canopy density, presence of regional endemic species, and inclusion of wetland habitats. Although polygons are classified into several schema (California Natural Diversity Data Base or "Holland" system, California Wildlife Habitat Relationships habitat types), data on presence of dominant canopy species are provided to allow customized classifications to meet a user's needs."
Attribute Accuracy: unknown
Attribute Accuracy Explanation:
Because source information ranged widely in date and reliability,
the
current database is uneven in both level of detail and accuracy. We
did not have the resources to assess the statistical accuracy of the
land-cover/vegetation map and associated database. However, we have
appraised the product using less formal methods that have guided our
use of the product.
The distribution of each dominant plant species in the coverage was
compared to the documented distribution recorded in the CalFlora database
which was derived from the Munz flora and revised with some more current
data. Outlying locations of species in the GAP database were re-examined
to either confirm that the location was documented in the data source
or to change the species code if it appeared to be an incorrect interpretation
or a data input error. A similar comparison was made for each community
type with the written description in Holland, 1986."
At the landscape scale of mapping land-cover, it is not possible to include all vegetation and habitat features. The following guidelines were used in interpretating and generalizing the source data.
The minimum mapping unit is 100 ha for upland community types and 40 ha for wetland communities. To account for mosaics of communities below this resolution, each map unit was attributed with up to three community types, each of which had to be >10% of the map unit area. The spatial location of individual stands of vegetation therefore are not provided.
The classification was based on dominant overstory species. The minimum canopy closure criterion is 25%. Thus forest and woodland formations must have >25% tree cover, shrublands must have <25% tree cover and >25% shrub cover, and herbaceous types must have <25% trees, <25% shrubs, and >2% herbs.
Because the map is of actual, not potential, vegetation, the existing types were recorded even where there is evidence of recent disturbance. One exception is recently burned chaparral, which was typed as chaparral unless there was some strong evidence of type conversion.
Species must represent >20% of the canopy of the cover type in the map unit to be listed as a dominant species.
Some attributes are not completed for all map units. Wetland data and crown closure are only recorded if determined from aerial photos or other high resolution source. Data were not readily available for all map units and thus these fields may be blank (i.e., value of "0"). In the Mojave Desert region, species data were often indeterminate from TM imagery, and high resolution sources were unavailable. In many instances, therefore, only a community type is assigned to such map units, and the species data are blank. The project guidelines in effect when the Sonoran Desert and Southwestern California regions were being mapped only required primary and secondary types to be recorded. The tertiary type and species data in these two regions are always blank, except where modified in later revisions.
The attributes for developed and agricultural lands contain variable depth of detail, depending on the source information used. For example, in some areas, the community type was assigned to "Agricultural Land" of unknown uses. Where such lands were known to be either cropland or orchard/vineyard, they were so indicated. In a few sites, these may be further subdivided such as Irrigated Hayfield or Evergreen Orchard. Such distinctions are useful for modeling wildlife species distributions because the WHR database distinguishes the habitat suitability among these for different species. However, it should be noted that our classification is inconsistently applied, with greater detail in some places than others. It was decided that it was worth maintaining the most information possible at the expense of a consistent, but less detailed, classification."
This section begins with general methods used for the CA-GAP mapping of land cover, followed by details on regional sources. Polygon boundaries were derived from photointerpretation of 1990 Landsat Thematic Mapper digital images, supplemented by 1990 HAP photography and large scale vegetation maps. Attribute data came from the 1990 HAP photography, the 1930's VTM Survey Maps (i.e., Wieslander), field visits, and large scale vegetation maps.
Southwestern California sources: The sources described above were all used for this region, as well as the SANDAG MSCP land cover map, the Metropolitan Water District map of the San Diego Pipeline No. 6 study area, and a land cover map of Camp Pendleton Marine Corps Base.
Davis, F. W., P. A. Stine, D. M. Stoms, M. I. Borchert, and A. D.
Hollander, 1995. Gap analysis of the actual vegetation of California:
1.
The Southwestern Region. Madroño, 42: 40-78.
Sonoran Desert sources: The basic approach in delineating areas was to divide the desert region into rocky slopes, bajadas, and valley floors as viewed on 1990 TM images. An extensive field survey of the dominant perennial vegetation of the entire region was undertaken for determining the floristic information. Because of the extent of the area to be inventoried, the greatest coverage of the desert could be affected by using the major highways and other roads and tracks of the desert as transects. Approximately 10,000 miles of road were traveled and 86% of the map units were viewed.
Subsequent to the draft GAP land-cover map, the Bureau of Land Management
and the California Department of Fish and Game conducted a joint effort
to enhance the map with added detail for critical wildlife habitats
in their Northern and Eastern Colorado Desert planning area (Dorweiler
1997). Sand dunes and dry lake beds were delineated from helicopter
overflights using a global postioning system to determine geographic
coordinates. Desert dry wash woodland was mapped independently in
the southern portion through supervised classification of July, 1994,
Landsat TM imagery after masking known areas of creosote scrub and
mountains to reduce confusion of spectral signatures. Due to difficulties
separating dry wash woodland in the northern part of the planning
area, BLM used on-screen digitizing from orthophoto quads and hand
drawn maps instead. Based on field work and examining aerial photos,
some additional corrections in classification were made to the draft
GAP map. The California GAP obtained a June, 1997, version of the
BLM map and incorporated most of the modifications. Polygon
boundaries were generalized and attributes adjusted as needed to be
consistent with the database design philosophy and structure of the
GAP
land-cover map.
Gray, M. V., 1994. A Digital Multipurpose Vegetation Map for the Colorado Desert of California. Masters thesis, Department of Geography, University of California, Santa Barbara.
Mojave Desert sources: The Mojave Region vegetation database was developed using existing maps as baseline source data, including the East Mojave Resource Inventory from Southern California Edison, the BLM West Mojave map (Clark, personal communication), Edwards Air Force Base, and a BLM/USGS map derived from MSS imagery. Maps were updated and refined using Thematic Mapper satellite imagery, a current 1:150,000 road atlas (to identify areas of residential or industrial development), and extensive reconnaissance level field survey. The revisions made by the Bureau of Land Management for their Northern and Eastern Colorado Desert planning area (Dorweiler 1997) included a small part of the southern Mojave Desert region. The revisions were incorporated into the final GAP land-cover as described above for the Sonoran Desert region.
An accuracy check of a portion of the coverage was conducted using data obtained by the BLM during vegetation characterization studies in the east Mojave in 1975. Four hundred sixty-two transects were measured in this study and a listing of one to four visually dominant species for the stand in which each transect was placed was made. Each transect was not specifically georeferenced; however, its location was recorded to the resolution of a section (typically one square mile). A GIS coverage of all sections containing transects was overlain on the vegetation coverage and a listing of the cover types occurring within the sections was obtained. Visual dominant species were compared with Holland cover types for each section and if the dominant species could be cross referenced to any of the Holland cover types occurring within the section, the mapping was considered to be accurate. The accuracy check in the east Mojave subsampled 1.6% of the total regional land area. Overall accuracy was 84.0% for the subsample; however, accuracy among the eight cover types occurring on the transects varied between 65% and 100%.
Thomas, K. A., 1996. Vegetation and Floristic Diversity in the Mojave
Desert of California: A Regional Conservation Evaluation. Ph.D.
dissertation, University of California, Santa Barbara.
East of the Sierra Nevada sources: Landscapes on the eastern flank of the Sierra Nevada range south of Mono Lake were generalized from the 1980 Vegetation Resource Inventory (U. S. Forest Service, unpublished maps). The White-Inyo Mountains were generalized and labeled from the Landsat TM-based Inyo National Forest vegetation map. The floor of the Owens Valley was derived from the 1:24,000 scale BLM-SCS soil-vegetation mapping. The remaining areas were delineated subjectively by photointerpretation of patterns in the satellite imagery in conjunction with the CALVEG map (Parker and Matyas 1981), another BLM soil-vegetation map published in the Bodie/Coleville Grazing EIS for the Bodie and Antelope Valley areas, and the VTM maps.
Sierra Nevada sources: VTM maps and miscellaneous recent vegetation maps (notably the vegetation databases from the national forests and parks) were used. Where such maps were lacking we relied on USFS soil and vegetation survey notes (alpine and subalpine areas surveyed by R. Taskey), our own 1994/95 field reconnaissance surveys, forest patch type descriptions from the SNEP late seral old growth database, CALVEG, the Sequoia Mediated Settlement Agreement map of stands of Big-tree forest, and the map of foothill woodland types prepared by Pillsbury et al. (1991). Our draft map was extensively updated in timber-producing areas using USFS maps of timber plantations and shrub-dominated timberlands.
Davis, F. W., and D. M. Stoms, 1996. Sierran vegetation: A gap analysis. Pages 671-689 in Sierra Nevada Ecosystem Project: Final Report to Congress, vol. II, Assessments and scientific basis for management options. Davis: University of California, Centers for Water and Wildlands Resources.
Central Western California sources: Vegetation polygons were attributed using VTM data, maps of hardwood forests and woodlands (Pillsbury et al. 1991) and redwoods (Fox 1988), and field surveys. Air photo interpretation techniques were used to confirm, enhance, and in some cases where no other data were available, supply polygon attributes. National High Altitude Photography (NHAP) and NASA-JPL color infrared transparencies were viewed stereoscopically to identify vegetation types, percent coverage, canopy closure, and disturbance. The NHAP photos are at a scale of 1:58,000 and dated from 1980-1984, while the NASA-JPL photos are at a scale of 1:65,000 and dated late 1980s to early 1990s.
Northwestern California sources: Vegetation polygons were delineated in a first iteration from the CALVEG map (Parker and Matyas 1981) and Fox's (1988) map of redwoods. The polygons were attributed using a combination of VTM data, a map of hardwoods (Pillsbury et al. 1991), the map of redwoods (Fox 1988), field surveys by UCSB staff (562 polygons were checked in the field), and aerial photography. In addition, the GIS Potential Natural Vegetation coverage for the Six Rivers National Forest and GIS coverages of timber types for the Klamath and Shasta Trinity National Forests were used. Expert opinion was solicited from Forest Service botanists and ecologists.
Thorne, J. H., 1997. Gap Analysis: The vegetation of Northwestern
California. Masters thesis, Department of Geography, University of
California, Santa Barbara.
Great Valley sources: Initial landscape boundaries were derived from the USGS Digital Land Use and Land Cover maps (DLULC, USGS 1986) mapping. This base map was then edited subjectively by photointerpretation of patterns in the satellite imagery to improve registration of distinct edges and to account for recent and use changes. Wetlands were added from the 1:24,000 scale digital National Wetlands Inventory (NWI) maps from the U. S. Fish and Wildlife Service. The detailed map units of NWI, which were available in digital form only from Fresno northward at the time, were generalized into larger landscapes for GAP purposes. Other source maps for delineating landscapes include relic patches of native perennial bunchgrass prairie (Barry 1972 and redrawn by Dremann 1988); Carrizo Plain from the Bureau of Land Management, Bakersfield District; Southern San Joaquin Valley from the California Energy Commission (Anderson et al. 1991); and riparian forests (Nelson and Nelson 1984).
Cascade Ranges sources: Floristic information was derived mainly
from our own field surveys plus digital maps from the Lassen, Modoc,
Klamath, and Shasta National Forests. Some species information was
also obtained from a recent hardwoods inventory (Pacific Meridian
Resources, unpublished 1996), the Sierra Nevada Ecosystem Project's
mapping of Late Successional-Old Growth Forests (Franklin et al. 1996),
the land cover map of the Eagle Lake Resource Area of the BLM (version
dated January, 1996), the 1996 land cover
map jointly produced from 1990 TM imagery by the USFS and California
Department of Forestry and Fire Protection, Soil-Vegetation maps from
the USFS and State of California, and unpublished maps from the VTM
survey (Wieslander 1946). Site visits to 232 landscape units were
also used to resolve conflicts between sources or to verify older
information.
Modoc Plateau sources: Sources of floristic information were the
following: Soil Conservation Service 1:24,000 soil vegetation maps,
which cover most of the private land. Vegetation Type Map (VTM survey
maps; Wieslander 1946), which cover the southeast corner of the region.
Detailed vegetation maps for Lava Beds National Monument contained
in (Erhard 1979). Modoc National Forest vegetation map, which was
produced in 1978 from aerial photographs. Bureau of Land Management
range survey maps, covering the arid Surprise Valley area. BLM integrated
management plan database, derived from a August 13, 1989, TM image
centered on the Madeleine Plains. A preliminary
version of the map was used, but then updates were made from the January,
1996 revision. Field reconnaissance in areas for which no existing
vegetation map was available. Accessible roads through polygons were
driven with numerous stops made to check vegetation, using binoculars
and a 200x stereoscope from vantage points to survey areas distant
from the road, and to identify conifers on nearby slopes and ridges."
Landscape size map units were delineated on-screen over the TM imagery on the basis of relatively homogeneous color, tone, and texture. (For information on processing of the TM data, see TM metadata). In many cases, boundaries were added or moved based on higher resolution data from aerial photography or larger scale vegetation maps. Dominant species data were obtained from existing sources, primarily the VTM survey maps. Percent cover of the primary, secondary, and tertiary types were estimated from aerial photography. Crown cover was similarly estimated. Wetlands were recorded from aerial photography or direct field observation, or in some cases from detailed maps. Special species were noted from existing map sources or from field observation. Assignment of polygons to classes such as CNDDB communities was made by interpretation of the dominant plant species list associated with each map unit in a look-up table. These combinations of dominant species were also assigned to Wildlife-Habitat Relationships habitat types in collaboration with Barry Garrison, manager of the California Wildlife-Habitat Relationships Program. For further details, see:
Davis, F. W., P. A. Stine, D. M. Stoms, M. I. Borchert, and A. D.
Hollander, 1995. Gap analysis of the actual vegetation of California:
1.
The Southwestern Region. Madroño, 42: 40-78.
Davis, F. W., P. A. Stine, and D. M. Stoms, 1994. Distribution and
conservation status of coastal sage scrub in southwestern California.
Journal of Vegetation Science, 5: 743-756.
Stoms, D. M., F. W. Davis, P. A. Stine, and M. Borchert, 1992. Beyond the traditional vegetation map towards a biodiversity database, in Proceedings of GIS/LIS'92, San Jose, CA, November 10-12, 1992, pp. 718-726."
Source Distance Resolution (meters): 50
Attribute Attribute description
REGION Jepson biophysical region of California
WHR1 Primary WHR habitat type
CROWN1 WHR canopy closure class of primary habitat type
WHR2 Secondary WHR habitat type
CROWN2 WHR canopy closure class of secondary habitat type
WHR3 Tertiary WHR habitat type
CROWN3 WHR canopy closure class of tertiary habitat type
WHRWET Presence or absence of 9 WHR wetland habitats
CNDDB1 California Natural Diversity Data Base ("Holland")
natural
communities classification code of primary type
PCT1 Proportion of polygon covered by primary cover type
CNDDB2 California Natural Diversity Data Base ("Holland")
natural
communities classification code of secondary type
PCT2 Proportion of polygon covered by secondary cover type
CNDDB3 California Natural Diversity Data Base ("Holland")
natural
communities classification code of tertiary type
PCT3 Proportion of polygon covered by tertiary cover type
SP1A Code for co-dominant species in cover type covering the
largest fraction of the polygon (primary type)
SP1B Code for co-dominant species in primary cover type
SP1C Code for co-dominant species in primary cover type
SP2A Code for co-dominant species in cover type covering the
second largest fraction of the polygon (secondary type)
SP2B Code for co-dominant species in secondary cover type
SP2C Code for co-dominant species in secondary cover type
SP3A Code for co-dominant species in cover type covering the
third largest fraction of the polygon (tertiary type)
SP3B Code for co-dominant species in tertiary cover type
SP3C Code for co-dominant species in tertiary cover type
SPWA Code for most widespread canopy species in the wetland
portion of the polygon
SPWB Code for second most widespread canopy species in the
wetland portion of the polygon
SPWC Code for third most widespread canopy species in the
wetland portion of the polygon
SPSA Code for a species of special concern either because it is
narrowly endemic, threatened or endangered that is present
in the polygon
SPSB Code for a species of special concern either because it is
narrowly endemic, threatened or endangered that is present
in the polygon
*** Begin DFG Appended Attributes ***
LABEL Polygon map label code used in the DFG Land Cover maps.
Label numbers were derived by concatenating 9 species codes
(SP1A through SP3C), assigning a unique ID to each unique
group of concatenated species codes, then joining these
unique IDs back to the primary attribute table. Therefore,
the LABEL code represents unique groupings ONLY for these 9
species codes.
SEARCH LABEL attribute with appended "M" character to
provide for
easier database or text file searches (to avoid confusion
with other number groupings).
FREQUENCY Frequency of LABEL attribute within the attribute table.
COLOR Color code used with the associated landcov.avl file to
portray an identical color scheme as depicted on the DFG
Land Cover Maps.
ACRES Calculated from ARC/INFO default AREA polygon attribute."