97th ESA Annual Meeting (August 5 -- 10, 2012)

PS 111-250 - Modeling vertebrate species in the US: Species modeling efforts of the Gap Analysis Program

Friday, August 10, 2012
Exhibit Hall, Oregon Convention Center
Jocelyn L. Aycrigg1, Jeff Lonneker1, Matt Rubino2, Nathan Tarr2, Kenneth G. Boykin3, Alexa McKerrow4, Gary P. Beauvais5, Tracey Gotthardt6 and William A. Gould7, (1)National Gap Analysis Program, University of Idaho, Moscow, ID, (2)Biology Department, NCSU, Biodiversity and Spatial Information Center, Raleigh, NC, (3)Fish, Wildlife and Conservation Ecology, New Mexico State University, Las Cruces, NM, (4)Core Science Systems, United States Geological Survey, Raleigh, NC, (5)Wyoming Natural Diversity Database, University of Wyoming, Laramie, WY, (6)University of Alaska, Alaska Natural Heritage Program, Anchorage, AK, (7)International Institute of Tropical Forestry, San Juan, PR
Background/Question/Methods

The goal of the Gap Analysis Program (GAP) is to build species range maps and distribution models with the best data available for assessing conservation status of vertebrate species within the entire US. These data describe patterns of species geographic location and basic ecological characteristics. Spatial patterns of species occurrence across the landscape provide insights into biodiversity conservation within the US. To achieve this goal, we started by compiling a species list of the US, which includes about 2000 amphibians, birds, mammals, and reptiles. We use a national database of standardized 12-digit hydrological units (HUCs) to represent species ranges. A species range provides the biological context within which to build each species’ distribution model. A species’ distribution, at 30 meter resolution, is created using a model to predict areas suitable for occupation within its range. It can be considered a spatial subset of a species’ range.  We build a deductive model for each species distribution using a suite of spatial variables including habitat types, elevation, hydrological characteristics, distance to/from forest edge, and habitat patch size. We base each modeled distribution on habitat associations described in peer-reviewed literature and expert input.

Results/Conclusions

We are currently working to complete our deductive models while also looking to improve some models through inductive modeling. This is our first attempt to build species models across species ranges rather than stopping at state boundaries.  And presently, we have created > 1200 species ranges and >700 species distribution models. All of these data are available to view and download from GAP’s Species Viewer (gapanalysis.usgs.gov). Our species ranges have been used to identify patterns of species occurrences in the US and our distribution models are informing the regional Crucial Habitat Assessment Tool (CHAT) being developed by the Western Governors’ Association. These data will provide a base from which we can iteratively improve the species’ range and/or model when new data become available. We will also use these data for a national biodiversity assessment of vertebrate species.