Final Report Summary:
Bruce C. Thompson
This research included all of New Mexico, a 314,920 km2 landscape that reflects a varied geologic and natural history. New Mexicos diverse array of species is attributable to complex connections of regional biogeographic components from the Great Plains, Rocky Mountains, Great Basin, and Chihuahuan and Sonoran Deserts.
Land Cover Classification and Mapping
We developed our land cover classification scheme in cooperation with the New Mexico Natural Heritage Program and consultation with experts on New Mexico vegetation. The final land cover map has approximately 24,260 polygons representing 42 mapped classes that include 33 terrestrial and riparian vegetation community classes, 2 hydrologic feature classes, 2 aquatic classes, 2 urban classes, and 3 classes of barren, rocky, or mined ground. We assessed accuracy of the final land cover map during February-July 1995 by ground-truthing 1,763 polygons with cooperation from state and federal agencies and a variety of other knowledgeable and interested people. Conservative accuracies among mapped classes ranged from zero to 80% at grouped cover-type level. Highest accuracy was associated with agricultural land cover, high-elevation conifer forest, urban vegetation, desert scrub, and natural surface waters. Accuracy among classes generally improved dramatically by accounting for ecotones and inclusions.
Predicted Animal Distributions and Species Richness
We modeled 584 species (26 amphibians, 96 reptiles, 324 birds, and 138 mammals) relative to species-specific data on associations with land cover types, mountain ranges, watersheds, elevation, slope, water, soils, and known general range. We consulted experts to review first-draft maps of species distribution predictions. To assess distribution predictions, we obtained species occurrence data for birds in a county in the northwest corner of New Mexico and amphibians, reptiles, and birds of a military reservation in southern New Mexico. Comparison of predicted animal presence to records of occurrence ranged from 53.8% to 88.6% accuracy among three taxonomic groups for two locations. Omission errors were more prevalent for the county data, whereas commission errors were more prevalent for all taxonomic groups compared for the military area. These patterns related to degree of recent specific surveys of test areas.
Considering all 584 animal species included in our project, we predicted the richest areas in the state to contain 327 species, 56% of the total. Richest areas among taxonomic groups contained 53.8% of 26 amphibian species, 59.4% of 96 reptile species, 65.7% of 324 bird species, and 47.8% of 138 mammal species. Assessment of data for breeding distribution of birds relative to year-round distribution of birds indicated distinctions between those data sets for drawing conclusions about bird richness.
Land Stewardship and Management Status
We used a public domain map of land ownership categories in New Mexico at 1/4-1/4 section (40 acre or 16 ha) resolution. With additional data about specific stewardship boundaries incorporated from federal and state agencies, several tribes, land trusts, and private landholders, we added 670 polygons to the ownership. Before assigning management status categories, we assessed views about management classification from a work group of various federal and state government agencies, tribal representatives, environmental organizations, and private landholders statewide. From variability in the responses, we concluded that the land management categories are not interpreted and applied in the same way by all individuals. Thus, we developed a dichotomous key to consistently assign status to the stewardship boundaries (Crist et al. 1996).
Private lands (45%) were the dominant category of stewardship; federal stewardship was dominated by Bureau of Land Management, U.S. Forest Service, and military lands. We identified 18 general categories of land tracts represented in management status 1 and 2 lands. These categories included an array of federal, state, and private managing entities. Distribution of management status in New Mexico was estimated as 2,418 km2 of Status 1 (1%), 19,354 km2 of Status 2 (6%), 89,833 km2 (29%) of Status 3, and 203,320 km2 (65%) of Status 4.
Analysis Based on Stewardship and Management
Management status 1 and 2 represent about 7% of the New Mexico landscape. We identified 11 natural land cover classes each represented by less than an estimated 100,000 hectares. Six of these restricted classes (Madrean Lower Montane Conifer Forest, Madrean Closed Conifer Woodland, Broadleaf Evergreen Interior Chaparral, Graminoid Wetlands, Riverine/Lacustrine, and Basin/Playa) each had less than 10% of their estimated area in Status 1 and 2. Statewide, 20 natural land cover classes each had less than 10% of area in Status 1 and 2. Of these classes, nine (primarily Madrean Forest and Woodland, Interior Chaparral, Broadleaf Sand-Scrub, and various Wetlands) each had less than 10,000 hectares in Status 1 and 2 areas. Management Status 1 and 2 lands were nearly all distributed among a variety of federal agencies and functions. Private and tribal stewardship is significant in the overall distribution of many land cover classes; 5 of the 11 most restricted classes have at least 45% of area on private and tribal lands.
We identified 35 species with no more than 1% of their predicted distribution on Status 1 and 2 lands. Nearly 45% of these species were reptiles and amphibians, despite those taxonomic groups representing 21% of all species included in analyses. Six of the nine species with no predicted distribution on Status 1 and 2 lands were amphibians and reptiles which have restricted distributions in southern New Mexico. Overall, 465 species (79.6%) each had less than 10% of their distribution on Status 1 and 2 lands. Importantly, all users of these data should recognize that some species primarily distributed on Status 3 and 4 lands adequately meet their biological needs there. Judicious evaluation will be needed to determine which species represent biological gaps.
Data Use and Availability
NM-GAP data are presented in a format that will operate on a PC configured to run ARC/INFO and ArcView current to November 1996. However, all possible combinations of data queries were not tested. A workstation may be necessary for some operations. NM-GAP data products and documentation may be acquired from the Resource Geographic Information System (RGIS) of New Mexico at (505)277-3622; Internet at http://rgis.unm.edu:8080, or from the national GAP Home Page on the Internet at http://www.gap.uidaho.edu/gap.
Crist, P.J., J.S. Prior-Magee, and B.C. Thompson. 1996. Land management status categorization in Gap Analysis: A potential enhancement. Gap Analysis Bulletin 5:20-22.