GAP Meetings 1999 Abstracts

Managing the Natural Resources of Illinois for the Future: Illinois Gap Analysis Project and Conservation 2000

Aycrigg, Jocelyn L., Mark Joselyn, Patrick W. Brown, Liane Suloway, and Linda Schwab

The National Gap Analysis Program (GAP) was initiated to address biodiversity conservation at the national level, while Illinoisš Conservation 2000 (C-2000) was established to address the natural resource needs of Illinois for the future. The Illinois Gap Analysis Project (IL-GAP) and C-2000 are cooperating to gather and maintain ecological data to monitor trends in Illinoisš ecosystems. C-2000 supports a variety of specific programs including the Critical Trends Assessment Program (CTAP), EcoWatch, and the Ecosystem Partnership Program. The data generated by these innovative programs will supplement IL-GAP. CTAP is a long-term program aimed at monitoring and preserving Illinoisš biological resources. CTAP has produced a land cover database of Illinois, conducted statewide scientific monitoring, defined the statešs Resource Rich Areas, and assisted Ecosystem Partnership Areas. Another major component of C-2000 is EcoWatch, which seeks to train łcitizen scientists˛ and students to monitor various ecosystems in Illinois. The Ecosystem Partnership Program is an additional component of C-2000 in which public/private partnerships attempt to combine natural resourdce stewardship with compatible economic and recreation development by bringing local stakeholders together. The existence of C-2000 in Illinois provides IL-GAP with valuable data and avenues in which to promote the goals and implementation of proactive biodiversity conservation at the community and landscape levels.

Keeping GAP Vegetation Mapping Current Over Time: Towards a Continuously Updated Multi-use Coverage

Division of Forestry, Minnesota Dept. of Natural Resources, 413 SE 13th Street, Grand Rapids, MN 55744

If GAP vegetation maps can be continuously updated in economical fashion without exhibiting major fluctuations in cover type acreages, they are likely to be in continuing demand for non-GAP use by state and federal agencies whose support may be critical to continuation of the program. Minnesotašs experiences in using regularly acquired satellite data for detection of forest change may point the way to an operational methodology for maintaining vegetation maps in current status while avoiding the expense and uncertainty associated with complete reclassification. Alternative approaches to GAP vegetation map updating are discussed and their costs compared.

Capen, David E.¹, Charles E. Ferree¹, Ernest W. Buford1, and Elizabeth H. Thompson²

¹Spatial Analysis Laboratory, School of Natural Resources, University of Vermont
²The Nature Conservancy

The objective of the Vermont Biodiversity Project is a conservation design that incorporates the full diversity of landscapes, communities, and species that occurs in Vermont. Recognizing the link between physical factors across a landscape and biological diversity, we describe a method for modeling landscape diversity as the basis for delineating areas for high biological diversity. We first used 30-meter digital elevation models to develop land-position indices and slope classes, which were combined to create 18 landform classes. We then incorporated 4 elevation zones, 8 classes of lithology, and 9 classes of surficial geology in an overlay analysis to create several hundred combinations of physical units in each of 7 biophysical regions in Vermont. Next, we used a grid of hexagonal cells, customized for each biophysical unit, as the basis for an analysis that selected up to 6 cells, but no more than 30% of the area of each region, that maximized physical diversity. We then employed indicators of biological diversity, including results of Gap Analysis, for an optimum design of priority conservation areas that maximized diversity, optimized representativeness, and took advantage of existing patterns of conserved land.

COMPARISON OF PROTECTED AREA PROPOSALS RESULTING FROM LIFE SCIENCE GAP ANALYSES WITH PROTECTED AREA ACHIEVEMENTS FOLLOWING A PLANNING PROCESS IN ONTARIO

Ontario Ministry of Natural Resources, 300 Water St., Peterborough, Ontario K9J 8M5

The life science component of Ontariošs system of protected areas focuses on representation of the naturally occurring landform/vegetation associations in each ecodistrict. The gap analysis algorithm used is iterative, selecting the site containing the most diverse assemblage of unrepresented landform/vegetation associations at each step. In a recently completed strategic planning project, gap analysis results for 39 ecodistricts in the planning region were provided to planning teams, including citizenšs panels (Round Tables) representing various stakeholders and interests (environmental, municipal, tourism, forestry, mining, hunting, etc.). Ecodistrict 3E-3 in the eastern boreal portion of the planning area is used to illustrate how gap analysis data were used, modified, and compared with alternatives, in an attempt to accommodate the needs of stakeholders, while achieving ecosystem representation. In this example, the level of representation (percent of total landform/vegetation features occurring within existing protected areas) prior to the project was 12%. The initial phase of planning resulted in no increase in the level of representation. However, subsequent generation of options (with the most optimistic alternative at 65% representation) and negotiations among stakeholders led to a proposal for protection of 54% of the landform/vegetation features. Given the large number of stakeholders and the contentious nature of resource allocation issues on Crown (public) lands, the final phase of this strategic planning process yielded results that are acceptable to most parties. Gap analysis results played a significant role in directing the process of selecting new protected areas to the most diverse sites.

¹National Gap Analysis Program, Moscow, ID
²University of Wyoming, Spatial Data Visualization Center, Laramie, WY
³Beartooth Mapping Inc., Red Lodge, MT

Decision support systems (DSS) represent a new paradigm in data serving and application by reducing or removing the need for users to gain expertise in GIS and allowing them to intelligently use information from fields other than their own by incorporating expert-systems tools. They are particularly relevant to GAP data which represents novel information for the decision-making processes in planning and management, yet the very novelty is an impediment to its application. The National Gap Analysis Program has invested in the development of two DSS pilot projects to date and intends to produce more for various clients to help solve this problem. These pilot systems, BEST and Refuge-GAP, were developed for very different users and scales of application; they represent the spectrum from small-area, short-term application to large-area, long-term planning. Developing DSS requires an iterative process to incorporate the data with the user group's decision process-maintaining the process in a recognizable form yet removing boundaries established by previous data limitations. The pilot DSS may be explored in the technical demo sessions; this paper will discuss issues in developing and implementing DSS including the need for a sound framework of data and scientific concepts necessary for establishing proper use and limitations.

GREAT PLAINS DATABASE/EXPERT SYSTEM FOR VERTEBRATE MODELING AND METADATA MANAGEMENT

The Kansas Gap Analysis Program (KS-GAP) developed an Access database/expert system to standardize vertebrate habitat and distribution model development. The database will also keep track of background information (metadata), maintain other species-specific habitat data (ancillary data), and provide output to the ArcInfo GIS for KS-GAP. With the database/expert system available to standardize choice criteria, the Midwest GAP Consortium (Kansas, Iowa, Nebraska, North Dakota, and South Dakota) realized it would be possible to model vertebrates regionally rather than on a state-by-state basis as has been done in the past. Regionalization will allow us to divide the burden of modeling and is expected to reduce edge-matching problems at state lines. Each state will identify species ranges within its boundaries, and expert review will still occur at the state level. We will describe the composition of the database and operation of the expert system. We will also present the results of a pilot project that is now under way.

AIRBORNE VIDEO SAMPLING EFFORT FOR ACCURACY ASSESSMENT OF THEMATIC MAPS: A SIMULATION EXPERIMENT

The number of land cover types intersected by video transects is positively related to total transect length, but the relationship is not linear. Three statewide land cover maps (Wyoming, Colorado, and Arkansas) from the Gap Analysis Program were used to generate sets of transects in three orientations (East-West, North South, random) and two configurations (systematic and random). For all three states a łlevel of diminishing returns˛ was reached, beyond which additional transect length added few new cover types to the sample. Transect orientation (E W vs. N-S vs. random) hade little effect on the number of cover types encountered, although vegetation zonation added a small advantage to systematic vs. random transects. Mean proportional area per cover type and mean number of polygons per type influenced the rate of capture of new types. Diminishing returns were achieved at 2495 km, 3355 km, and 1769 km for Wyoming, Colorado, and Arkansas, respectively.

EDUCATING THE PUBLIC THROUGH THE NATUREMAPPING PROGRAM AS A WIN-WIN SOLUTION FOR LOCAL DISSEMINATION OF GAP DATA SETS

Washington Cooperative Fish and Wildlife Research Unit, University of Washington

The Washington Gap Analysis Project (WA-GAP) began in 1991 and The NatureMapping Program the following year. NatureMappingšs goal is to facilitate the exchange of information between natural resource agencies, academia, land use planners, local communities, and schools through public education and participation in data acquisition. It began because there was a need for ground-truthing and data collection for WA-GAP. If public data are to be used, then the public must be trained. Training materials are primarily GAP oriented. Feedback to the public is through mailings and the web site which receives over 15,000 hits per month. Facilitating the exchange of information includes interactions between agency personnel and NatureMapping participants at workshops, through the exchange of materials, or working together on local projects. WA-GAP implementation efforts are to give local planners, elected officials, and the public tools they need to carefully consider biological conservation and biodiversity issues during their planning activities. NatureMapping is playing a role by using its network of particpants within each county as a resource for ground-truthing WA-GAP maps at a finer scale, and for individuals to become involved in their own countyšs long-term planning efforts by working with city and county land planners.

A GIS DECISIONFinlay, Ian SUPPORT SYSTEM FOR A COMBINED ENDURING FEATURES AND RTES APPROACH TO GAP ANALYSIS

Gap Analysis software developed by Linnet will be demonstrated using the Province of Ontario, Canada, as an example. The software allows both coarse- level analysis at an ecoregional level and fine filter analysis at an ecodistrict level. Coarse-level analysis involves overlaying various data sources (surficial geology, classified Landsat data, existing protected areas, anthropogenic disturbances [roads, forest cut-overs]) and providing reports of the representation status of each enduring feature type found within the ecoregion. Fine-filter analysis extends the coarse-level analysis to include vegetation data (Forest Resource Inventory) and automatically identify "candidate areas" based on representation, diversity, and size. The software allows the user to perform "what if" scenarios in a variety of different ways. For example, what if I change the minimum size considered for representation? What if I choose the second-best candidate instead of the best candidate? What if I change the boundary of this candidate? Using the software, you can preview what the representation will be if the candidate areas you select were to become protected areas. Rare, Threatened, and Endangered Species (RTES) data can also be incorporated into the analysis in the site selection process. Both coarse-level analysis and fine-filter analysis will be demonstrated. The technology that will be used in the demonstration will include custom software developed by Linnet, ARC/INFO for Windows NT, ArcView with Spatial Analyst, and Microsoft Access. Audience participation is encouraged, and there will be an opportunity to discuss incorporation of a variety of spatial "rules" into automated analytical techniques.

SELECTING PRIORITY AQUATIC CONSERVATION SITES: APPLICATION OF THE NATURE CONSERVANCYšS AQUATIC CLASSIFICATION

¹The Nature Conservancy, Freshwater Initiative, 8 S. Michigan Ave., Suite 2301, Chicago, IL 60603
²The Nature Conservancy, Freshwater Initiative, 4245 N. Fairfax Drive, Suite 100, Fairfax, VA 22203-1606

Through The Nature Conservancyšs Freshwater Initiative, we are applying a hierarchical classification system for freshwater communities to identify priority sites for conservation. With ecoregions across the United States and Latin America serving as broad-scale conservation planning areas, we are classifying and mapping streams, lakes, and nearshore areas using abiotic variables that structure biological communities. Using water quality, land use, biological data, and consultations with experts, we select areas across ecoregions thought to represent the most viable examples of all habitat types. Our goal in promoting aquatic classification is to ensure that conservation efforts use consistent approaches to identify aquatic targets and incorporate sites into portfolios that adequately represent the aquatic biodiversity of an ecoregion. Consistent methods among ecoregions will allow us to compare and analyze aquatic targets beyond ecoregional boundaries, assess each ecoregionšs stage of knowledge about aquatic targets, and develop meaningful measures of success. Drawing on examples from recently completed and ongoing efforts for ecoregions in the United States and Latin America, we present initial conservation priorities identified from application of the aquatic classification.

Florida Fish and Wildlife Conservation Commission, Office of Environmental Services, 620 S. Meridian Street, Tallahassee, FL 32301-1600

For over ten years, the Florida Fish and Wildlife Conservation Commission (FFWCC) has been actively involved in identifying significant areas of upland and wetland biodiversity. Maps depicting significant biological resource areas have been developed and are used in prioritizing land for acquisition by the state and in the development of regional and local land use planning. The FFWCC has begun a new project to comprehensively access the status of aquatic biodiversity throughout the state. The first phase of Floridašs aquatic resource project was an assessment of watersheds containing rare and imperiled fish species. Fish collection records were obtained to determine the Florida distribution of the selected species. River reaches and their surrounding watersheds that contained rare or imperiled fish species were identified. Watersheds were ranked based upon the number and relative rarity of the species present. Water quality and upland land use were analyzed to determine potential threats to the identified watersheds and river reaches. Watersheds identified as threatened were targeted for potential restoration, water quality improvements, or additional surveys. The current phase of the project involves assessing riverine and lake systems for all fish and mussel species. Florida is investigating how the GAP process might be applied in the estuarine and marine environment. Significant mapping and species-specific modeling has already been completed to define essential fish habitat in several estuarine areas. A project completed in Floridašs Big Bend Region integrated protection of upland, wetland, freshwater, and estuarine resources while accommodating secondary growth impacts from proposed major transportation improvements.

Rapidly expanding technology and access to more diverse sources of geographic imagery has led to promising new techniques for the use of image processing for Gap Analysis. With the recent successful launch of Landsat 7, remotely sensed multispectral imagery will be significantly less expensive and easier to acquire. Our workshop will expose the participant to the latest image processing methodologies, taking full advantage of satellite images, as well as aerial photography and RADAR data. The workshop will be conducted through a series of Gap Analysis scenarios, each designed to show how the latest software tools apply to a particular land analysis requirement. These scenarios and software tools include: * Autonomous Gap Analysis processing with a rule-based expert system, designed to automatically extract complex Gap Analysis terrain information from imagery.
* Making your GAP geographic database łortho-accurate˛ with high resolution aerial photography and the new ortho-rectification software for the PC.
* Land cover change detection using the new Image Analyst Extension for ArcView.
* Specific vegetation species identification with sub-pixel processing.
* Elevation data extraction with IFSAR RADAR imagery.
* 3D visualization of the environment with the VirtualGIS.

AN EVALUATION OF CONSERVATION THRESHOLDS USED TO DEFINE PROTECTED AND UNPROTECTED VERTEBRATES IN GAP ANALYSES

¹USGS Biological Resources Division, Maine Cooperative Fish and Wildlife Research Unit, University of Maine, Orono, ME 04469-5755
²Department of Wildlife Ecology, University of Maine, Orono, ME 04469-5755
³Natural Resource Ecology Laboratory, Colorado State University, Fort Collins, CO 80523-1499

Gap analysis is designed as a coarse-filter approach to geographically assess a statešs biodiversity in terms of plant communities and terrestrial vertebrates. One of the key analyses in GAP is an assessment of these two elements that are considered unprotected. Unprotected is defined in terms of 1) the amount of a speciesš habitat (or plant community) predicted to be in a state, and 2) the percentage of the statewide habitat (or plant community) that is on lands managed for natural processes and values (i.e., the Category 1 and 2 lands). In contrast, if an element is above these thresholds, it is considered secure from a conservation perspective. We evaluated these two threshold criteria for terrestrial vertebrates that regularly breed in Maine by plotting the frequency distributions of these criteria and determining where species are at risk versus not at risk, as defined by the Maine Department of Inland Fisheries and Wildlife (MDIFW). If the two criteria are correlated with conservation risk, then one would predict that species at risk should fall within the low side of both distributions, whereas species assessed to have little risk of extinction should occur mostly in the middle and high sides of the distributions. These predictions were tested with two sets of vertebrate species, high-risk species defined as those listed as endangered or threatened (ET) by the MDIFW, and a comprehensive listing of conservation risk covering all terrestrial vertebrates (i.e., Heritage Ranks). The quality of potential habitat a species has seems to be a good indicator of conservation status; the ET species had < 5,000 km2 of predicted habitat whereas 70% of the non-ET species were predicted to have > 10,000 km2 of habitat. In contrast, the percentage of a species potential habitat that is on protected land was not a good indicator of conservation status because variation in this threshold criteria was not associated with ET status nor with Heritage ranks. Given that Mainešs terrestrial vertebrate fauna has relatively few ET species and a low proportion of its total land area in conservation lands, the two key threshold criteria used in the Gap Analysis Program need to be evaluated under a greater range of conditions than those reported here (i.e., states with a faunal composition of > 4.8% ET species and > 5.3 % of land area in conservation lands).

¹Canadian Forest Service, Natural Resources Canada, P.O. Box 4000, Fredericton, New Brunswick, E3B 5P7, Canada
²3156 West 42nd Avenue, Vancouver, British Columbia, V6N 3H2,Canada (formerly Fundy Model Forest)

A fine-filter gap analysis was carried out in southern New Brunswick, Canada, within one of Canadašs ten Model Forests. The purpose, in keeping with the Model Forest objectives, was to develop and apply a systematic approach for filling the medium and fine-scale holes in a developing provincial protected areas system, which was based on coarse-scale enduring features of the landscape. Initial challenges ranged from the lack of a usable ecological land classification system to the many small land holdings in the area having absentee owners. Areas containing assemblages of rare species, remnants of once-more-common ecosystems, unusually high species or ecosystem diversity, and relatively untouched examples of late successional community types were identified using a series of steps. Methods employed for identifying sites ranged from seeking out the anecdotal gem and visiting herbaria to the traditional Gap Analysis GIS layering approach. All potential sites identified through these steps were visited and examined for their ecological significance. At present, a conservation stewardship program is under way to encourage landowners to protect chosen sites.

University of Idaho Cooperative Park Studies Unit and National Park Service Social Science Program

Gap analysis as a technique and the National Gap Analysis Program (GAP) are making significant progress. The scientific literature on gap analysis is growing. As gap analysis has developed, it has become increasingly clear that socioeconomic factors play a significant role in the distributions of biodiversity elements and their conservation. Research on the adoption, diffusion, and application of gap analysis is also important.

The National GAP Program has initiated a seed grant program to encourage and support the integration of socioeconomic research into gap analysis techniques and state projects. Four socioeconomic seed grants of up to $15,000 each will be awarded this year. A Request for Proposals (RFP) was distributed to all state GAP projects in June, and proposals are due 1 September 1999. The purpose of this presentation is to 1) describe the goals of the socioeconomic seed grant program, 2) provide examples of how socioeconomic research can be integrated into gap analysis, 3) review the RFP, 4) answer questions about the socioeconomic seed grant program, and 5) encourage state GAP projects to submit proposals.

MISSISSIPPI GAP META-PROJECTS: ENHANCING THE CAPABILITITES OF BIODIVERSITY PLANNING

Minnis, Richard B., Francisco J. Vilella, Christopher J. Reynolds, and Jacob L. Bowman

Mississippi Cooperative Fish and Wildlife Research Unit, Department of Wildlife and Fisheries, Mississippi State University, Mississippi State, MS 39762

The Mississippi Gap Analysis Program (MS-GAP) is a rapid, large-scale assessment of the habitat distribution and protective status of all of Mississippišs terrestrial vertebrate species. As such, we recognized the need to incorporate applications at scales finer than regional and/or continental. From the onset of MS-GAP we have initiated cooperation with Mississippišs natural resource agencies through the metaproject approach (sensu Indiana GAP). Our metaproject partners include the U.S. Forest Service, U.S. Fish and Wildlife Service, Mississippi Museum of Natural Sciences, and Mississippi Department of Wildlife, Fisheries, and Parks. These projects include 1) establishing a forest bird monitoring program for the National Forests of Mississippi, 2) developing the Mississippi Ornithological Atlas, 3) developing the Mississippi Herpetological Atlas, and 4) developing habitat models for selecting black bear reintroduction sites in Mississippi. Metaprojects have been instrumental in enhancing MS-GAP capabilities to customize needs, approaches, and applications that focus on local biodiversity issues, while retaining the original regional and continental goals of Gap Analysis.

A Habitat Affinity Index (HAI) was developed for fish and invertebrate species in estuarine waters of the Mid-Atlantic and Southeast regions. HAI defines habitat affinity based on the relative concentration of a species in a particular habitat, compared with the availability of that habitat in the study area. In the Mid Atlantic region, EPA's EMAP program provided trawl survey data for 120 sites, 482 samples, July through September, for four consecutive years (1990-1993). In the Southeast region, EPA/EMAP trawl survey data were for 153 stations, July through September, for two years (1994-1995). Dissolved oxygen, temperature, salinity, depth, substrate type, sediment contaminants and toxicity were measured at each site along with trawl catch by species and size. Although most estuarine species occur across a range of environmental gradients, many had discernible habitat affinities. In the Mid-Atlantic region, three-fourths of sciaenids had affinity for a particular salinity zone, whereas three-fourths of flatfish had affinity for a particular substrate type. In the Southeast region, the maximum catch of any given species was consistently greater at the nonpolluted sites, although overall catch was not significantly greater at nonpolluted sites. These analyses feature consistent sampling methods applied across an entire biogeographic province. In addition, a variance estimator allows statistical tests and comparisons among estuarine systems. A similar analysis is planned for Gulf of Mexico data.

THE CRITICAL ROLE OF PRIMARY POINT OCCURRENCE DATA IN MODELING SPECIES GEOGRAPHIC DISTRIBUTIONS

¹Natural History Museum, The University of Kansas, Lawrence, Kansas
²San Diego Supercomputer Center, San Diego, California

Knowledge of biodiversity and its distribution is almost universally incomplete: most species are known from relatively few records or localities, and most localities have either not been sampled or have been sampled only cursorily. For this reason, in interpreting biodiversity data, it is critical to include a step of inference to fill the holes in knowledge. In contrast to wildlife habitat modeling approaches, we strongly advocate basing such inferential efforts on primary point occurrence data, and using algorithms designed to produce ecological niche models to produce distributional models. The advantage of this approach lies in its link to biological processes: the niche is the key to population processes (e.g., source-sink dynamics), biogeographic processes (e.g., species invasions, distributional shifts in the face of changing climates), and other features of species biology. Moreover, basing models on primary point occurrence data allows models to be continually updated, even built in real-time for particular applications, allowing products based on distributional models to improve continually in quality and completeness. This focus on modeling ecological niches based on primary point occurrence data provides a greatly improved basis for products based on the distributions of elements of biodiversity.

ACCURACY ASSESSMENT OF PREDICTED VERTEBRATE OCCURRENCES: ARE DETAILED NATIONAL STANDARDS TECHNICALLY FEASIBLE?

¹Research Assistant, Department of Wildlife Ecology, Maine Cooperative Fish and Wildlife Research Unit, University of Maine, Orono, ME 04469-5755
²USGS Biological Resources Division, Maine Cooperative Fish and Wildlife Research Unit, University of Maine, Orono, ME 04469-5755

Gap Analyses are being done as state projects to determine the adequacy of the current conservation lands in protecting biodiversity. Ensuring that these statewide data can be integrated for regional and national level analyses, the Gap Analysis Program (GAP) has established data standards. Currently, the accuracy standard for predicting terrestrial vertebrate species richness on a site is about 80%. However, reliability may be influenced by state-specific factors such as test site size, survey length, geographical distribution, and faunal composition (e.g., % of community consisting of rare fauna). How species are defined as being present in the state for modeling may also influence the omission and commission errors. Results from Maine Gap Analysis were used to assess the need for a clear understanding of what a breeding species is in GAP, and how test site size and survey length influence the reasonableness of an overall accuracy standard of 80% for predicted vertebrate occurrences across all states. The conservative definition used in Maine-GAP to establish a list of regularly breeding species in inland Maine lead to low omission versus high commission errors, especially for birds. When a more liberal data set was used in the accuracy assessment, commission error significantly declined and omission error increased significantly (p > 0.03), as expected. Linear regression analysis also showed that commission error was significantly influenced by test site size and survey length (r2 = 0.85, p > 0.003). Because previous research has shown that wildlife habitat relationship models do a fair job of predicting vertebrate occurrences on sites, use of the statewide models should lead to reasonable results at that resolution. However, until there are nationwide field data that standardize which species to include (and exclude) in the modeling process as well as size and survey length of test sites, the adequacy of the predicted occurrences for regional and nationwide level analysis remains problematic.

USE OF SOUTH CAROLINA GAP PROJECT DATA TO DEVELOP A LANDSCAPE CONSERVATION STRATEGY FOR RED-COCKADED WOODPECKERS AND LONGLEAF PINE IN THE COASTAL PLAIN OF SOUTH CAROLINA

Gordon, David H.¹, Elise V. Schmidt, Cindy Aulbach-Smith², Frank X. Tian³, Bruce R. D. Richardson¹, James D. Scurry³, and David Otis⁴

¹U.S. Fish and Wildlife Service
²South Carolina GAP Analysis Program
³South Carolina Department of Natural Resources
⁴South Carolina Cooperative Fish and Wildlife Research Unit

The 525,000 acre Winyah Bay Focus Area located in the upper coastal plain of South Carolina is a biologically rich region including extensive longleaf pine (Pinus palustris) tracts and a significant red-cockaded woodpecker (Picoides borealis) population (>100 groups); both occur predominately on private lands. In 1993, a coalition of public and private organizations formed a Focus Area Task Force to seek landscape-scale conservation solutions to sustain the ecological integrity of the Winyah Bay Ecosystem. Immediate attention was given to protecting existing wetland and upland habitats from sprawling coastal development using perpetual conservation easements, public and private fee-title acquisitions, and cooperative, long-term land stewardship initiatives. Success in this effort led to the formation of a partnership in 1998 to develop and implement a landowner-based conservation strategy specifically for the endangered longleaf pine ecosystem and red-cockaded woodpecker in the Focus Area. The goal of the initiative is to secure landowner and community-based commitment to the long term stewardship of a viable red-cockaded woodpecker population and longleaf pine community within the Winyah Bay ecosystem. Land cover and generalized soil data developed from the SC-GAP Project are being used to identify potentially suitable sites to target for longleaf pine restoration. The goal is to increase the overall habitat base for red-cockaded woodpeckers and develop corridors of habitat that connect existing isolated tracts. This initiative will facilitate implementation of the U.S. Fish and Wildlife Servicešs red-cockaded woodpecker private lands conservation strategy using cooperative conservation partnerships incorporating Safe Harbor Agreements and Habitat Conservation Plans.

TAKING IT TO THE WEB: INTEGRATION OF COLORADOšS GAP INFORMATION INTO COLORADOšS NATURAL DIVERSITY INFORMATION SOURCE TO FACILITATE CONSERVATION PLANNING

The Colorado Division of Wildlife began using geographic information system (GIS) technologies in 1972 through its Wildlife Resource Information System (WRIS) efforts to provide wildlife habitat information for use in natural resource planning processes. The early 1980s saw expanded use of wildlife resource information at the county planning level through efforts of Division biologists using GIS to delineate significant wildlife habitats for consideration in the development of county master plans. Early information consisted of thematic maps of wildlife distributions and associated biological activity areas (winter range, lek sites, etc.) mapped by field biologists. Ancillary information on species use of habitats was collected in textual databases such as the Colorado Wildlife Species Database. The early 90s brought development of landscape-level habitat coverages that could be utilized with habitat association information to develop habitat models for species not previously mapped using thematic methods. Coloradošs participation through the National Gap Analysis Program provided opportunities for data integration, making habitat modeling feasible on a statewide basis. Through a collaborative design approach in the mid-90s, the System for Conservation Planning (SCoP) project developed methods to provide workstation based modeling outputs direct to county planners. Recently the Natural Diversity Information System (NDIS) was developed, leveraging SCoP programming code, Internet Map Server (IMS/ESRI) technologies, and Java programming, bringing old and new conservation planning data sets into broad public access. NDIS provides Internet access to information on wildlife habitats, vegetation patterns, conservation sites and occurrences, high-priority habitats, concerns if developed, and development patterns. NDIS can be accessed at http://ndis.nrel.colostate.edu.

U.S. Geological Survey, Biological Resources Division, Center for Biological Informatics

Trying to locate historical data or find data to augment your data set thus preserving your meager research budget? The FGDC and NBII Clearinghouse Gateways are a means to locate and evaluate data prior to obtaining and loading on your server. This workshop will utilize Internet connection to the Gateways and workshop participation to demonstrate the Clearinghouse as a search engine for data. You will leave the workshop with several metadata examples. These examples and a guide (a series of questions) will be used to illustrate metadata as a tool for data evaluation. The structure and concepts of metadata will be discussed as the framework describing data.

Materials provided:
Clearinghouse presentation
      NSDI Fact Sheets
      NBII Fact Sheets
Metadata examples
Metadata evaluation guide
Metadata presentation
      NSDI Fact Sheets
      NBII Fact Sheets

RECENT INNOVATIONS IN AERIAL VIDEOGRAPHY, AUTOMATIC RECTIFIED MOSAICS, CANOPY PROFILING, AND THE HOW-TO WEB SITE

Slaymaker, Dana, Zhigang Zhu, Edward Riseman, Howard Schultz, and Chris Holmes

The use of aerial videography to collect a distributed set of terrain point samples for Gap Analysis projects began as a fairly simple process. Small camcorders were mounted on light aircraft, logged with GPS, then viewed on TV monitors set up beside a computer to classify vegetation on Landsat images or to verify an existing classification. This was inexpensive but time-consuming because every video frame inspected had to be matched to the Landsat image by a combination of its GPS-logged photo center and pattern recognition. Combined with the unstable viewing image inherent in interlaced television, most interpreters found this process exhausting. Viewing conditions improved as it became practical to digitize the video and move it onto the same screen as the satellite data, but it was still necessary to deal with individual video frames and to use the GPS point files to locate the position of each.

In this paper, we present a different approach to interpreting aerial video data by mosaicking it into long strips that can be directly overlaid on base maps or satellite images. A simple program that generates non-georeferenced video mosaics can be downloaded from the web for around $60. The resulting strips cover a large enough area to be rectified to the Landsat image or 7 1/2 minute quads, making them much easier to interpret.

However, as part of a joint NSF sponsored project between the Department of Forestry and Wildlife Management and the Computer Science Vision Lab at the University of Massachusetts, we are developing a mosaic program that both georeferences and rectifies each frame during the assembly process. This work is part of a larger goal to develop automatic terrain reconstruction from the aerial video (using the massive redundancy of frame coverage to generate an accurate digital elevation model [DEM] in spite of the poor metric control of the camera lens), then ray tracing the mosaic to this surface. This requires a progressive frame video camera with far more precise coordinates for each exposure. We have switched to a digital video camera system and added a digital gyroscope to document the camera's orientation in flight. We are also using a pulse laser to measure distance to the ground and record a profile of its surface. This approach adds expense, size and weight to the camera system but, for some purposes, this will prove a reasonable tradeoff against the time and expense of manually handling individual video images. The use of the laser profile and DEM generation have opened up new survey applications for this instrumentation, measuring stand height, canopy structure and standing biomass as well as forest type in vegetation surveys.

Finally, we will use this opportunity to introduce our new aerial videography web site, an ongoing manual of everything we know and can collect on the use of video, digital and small format cameras in natural resource inventories.

ESTABLISHING INITIAL PRIORITIES FOR THE CONSERVATION OF BIODIVERSITY IN RIVERINE ENVIRONMENTS

Before you can establish conservation priorities you must define or develop your a) target element, b) assessment element(s), c) assessment unit(s), and d) assessment criteria and methods. From a biodiversity conservation standpoint, we believe that key ecological and evolutionary processes should be taken into consideration when defining these items and your assessment criteria and methods should provide truly relative assessments that correspond directly to your objectives. The first part of this presentation defines and describes the rationale behind the selection of these items for the Missouri Aquatic GAP Pilot Project, and also covers the associated assumptions. The second part of this presentation gives a general overview of how our initial conservation priorities will be established. Finally, it will briefly cover some alternative or additional approaches to establishing conservation priorities in riverine environments.

CHALLENGES AND OPPORTUNITIES FOR BIODIVERSITY ASSESSMENTS IN AQUATIC ENVIRONMENTS

Interest in the conservation of aquatic biodiversity has increased dramatically over the last five to ten years. However, rapid declines in the health of our nationšs aquatic ecosystems necessitate that we move rapidly from the łarm-waving stage˛ to a more concerted nationwide effort to conserve these important resources. Such an effort presents several major challenges. Some of the most pressing challenges include: a) establishing a central forum for the exchange of ideas and information among the numerous resource agencies and aquatic resource professionals across the nation, b) establishing a common strategy to inventory and assess our diverse aquatic resources, and c) establishing conservation strategies that effectively deal with the diffuse, diverse, and often distant threats to our aquatic resources. The National Gap Analysis Program is ideally situated to help us meet many of these challenges, and the initiation of Aquatic GAP pilot projects in New York and Missouri is an important and necessary first step though many challenges remain. Aquatic biodiversity conservation efforts also present some unique opportunities. For instance, our nation is mandated by law in the Clean Water Act to maintain the biological integrity of our nationšs waters. This legislative mandate has the potential to be a powerful tool in the conservation of not just our aquatic, but also our terrestrial, resources. However, implementation of the Clean Water Act has been hindered by a lack of accurate and detailed data; data that can be developed within an aquatic component of GAP. Another opportunity relates to the fact that the public places high priority on preserving the health of our nationšs waters because they see it as a human health issue. This is likely our greatest conservation opportunity, since private land management is absolutely essential for the conservation of most aquatic systems, especially riverine and coastal systems.

CRITERIA AND CONSEQUENCES TO GUIDE DECISION-MAKING WHEN DESIGNING A MAP ACCURACY ASSESSMENT

Stehman, Stephen V.¹, Sarah M. Nusser, F. Jay Breidt², and Erwin E. Klaas³

¹SUNY College of Environmental Science and Forestry, Syracuse, NY
²Iowa State University, Statistical Laboratory
³USGS/BRD, Iowa Cooperative Fish & Wildlife Research Unit, Ames, IA

Numerous decisions must be made when developing the protocol for assessing the accuracy of thematic maps. Designing an effective, practical accuracy assessment requires defining the criteria by which these decisions may be judged, and then making decisions based on the consequences of each choice. To further focus our thinking, it is useful to recognize three basic components of accuracy assessment: 1) the sampling design used to select the reference sample; 2) the response design used to obtain the reference land cover classification for each sampling unit; and 3) the estimation and analysis procedures. Decision-making criteria can be defined for each component. For example, criteria for evaluating sampling design options may include precision, cost, simplicity, and whether the design satisfies the definition of probability sampling. Response design criteria may include meaningful representation of land cover (related to hard versus fuzzy classes) and meaningful scale of the assessment unit (related to choice of pixel, polygon, or 2 ha block). Example criteria for the analysis component are consistency (a statistical property of estimation) and spatial representation of accuracy. Planning the accuracy assessment should then proceed focusing on the objectives outlined for the assessment and resolving difficult decisions on the basis of explicitly identified consequences (tradeoffs) of each choice. This presentation will focus primarily on statistical criteria and consequences for designing a practical, yet statistically valid accuracy assessment protocol.

Regional Conservation Biologist, Parks Canada-Cornwall Service Centre, 111 Water Street East, Cornwall, Ontario K6H 6S3

The Lake Superior Protected Area Managers (LSPAM) sponsored the World Wildlife Fund-Canada which leads the national Endangered Species Campaign with other conservation interests in the preparation of a protected area representation analysis (a gap analysis) for the entire Lake Superior Watershed. The project reconciled land classification differences across the US-Canadian border, accessed source data from both countries, developed and applied a new classification for large freshwater lakes, used its previously tested representation criteria, assessed the results, and provided maps of source data and results. The project demonstrates the capabilities of existing US and Canadian data and is an example of applied criteria for representation analyses. More detail is found in Representation Analysis of the Lake Superior Watershed by the World Wildlife Fund Canada (1998), provided to LSPAM as possible content for coordinated Lake Superior-wide protected area communication.

Thompson, Bruce C.^1,2, Robert A. Deitner^2,3, Julie S. Prior-Magee^1,2, and Kenneth G. Boykin^2,3

¹USGS Biological Resources Division,
²New Mexico Cooperative Fish and Wildlife Research Unit,
³Fishery and Wildlife Sciences Department, New Mexico State University, Las Cruces, NM 88003

During 1997-1998, we evaluated data products and outreach capability of the original New Mexico Gap Analysis Project (NM-GAP) by providing cross-project data sharing and staff availability, re-examining data layers and products, and assessing interest in and utility of NM-GAP data among counties statewide as conservation planning contacts. Substantive discrepancies were found in some deliverables (e.g., projection reference errors in metadata, poor spatial reference of some polygons in GIS coverages, inconsistencies in species distribution models). These discrepancies complicate focused work with portions of NM-GAP products, but were not sufficient to change any original conclusions. Minor changes were made to statewide coverages to rectify geospatial referencing needs and to revise 21 of 584 (3.6%) vertebrate distribution algorithms (1 mammal and 20 bird species). These revisions primarily involved simple adjustments in species associations with watersheds of occurrence. Related review of bird distributions by multiple experts in the state revealed varied concurrence on revisions needed for individual species. Our multiple contacts (mail, e-mail, and telephone) with planning program staff in all 33 New Mexico counties produced responses from 8 (24.2%), despite provision of substantial illustrated explanatory material. Generally, these 8 counties represented the least populous parts of New Mexico. These counties were superficially interested in NM-GAP data but largely did not have the hardware, software, or technical expertise to make extensive use of NM-GAP products. They were aware of likely advantage of GIS applications in planning but were preliminarily viewing GIS use for other alternatives like emergency 911 implementation. Significant challenges remain in assisting these counties to use available ecological data for conservation planning.

Umphrey, Gary J.¹, Geoff S. Ghitter², Daniel W. McKenney³, Donald H. Rivard⁴, and Alan D. Tomlin⁵

¹Dept. of Zoology, University of Western Ontario, London, Ont. N6A 5B7
²Mid America Remote Sensing Center, Murray State University, P.O. Box 9, Murray, KY 42071
³Canadian Forest Service, P.O. Box 490, Sault Ste. Marie, Ontario
⁴Parks Canada, 25 Eddy St., Hull, Quebec K1A 0M5
⁵Agriculture and Agri-Food Canada, Vineland Station, Ontario L0R 2E0

Substantial, and indeed bold, initiatives to expand Canadašs network of conservation lands have gone forward in recent years. The reserve selection process has been informed by several gap analysis approaches, including those developed by World Wildlife Fund Canada, individual provinces, Parks Canada, and the Canadian Forest Service. The various Canadian approaches tend to share a common conceptual root with U.S. GAP that recognizes the value of a łcoarse filter˛ conservation strategy. If the recommendations of these gap analysis approaches are actually implemented, it is likely that the reserves selected will contribute substantially to the goal of conserving Canadian biodiversity. But while łfine filter˛ components are employed to some degree in Canadian gap analysis approaches, the adequacy of the reserve network on an element-by element basis remains largely unknown. Not unexpectedly, this uncertainty contributes to the conflict between socioeconomic interests and conservationists over the use of land and natural resources.

We present examples that illustrate where shortcomings in geospatially explicit biodiversity information are leading to difficulties in developing Canadian policy on biodiversity and resource management, including new endangered species legislation. In particular, we suggest that a tendency to analyze Canadašs biodiversity as if the country was a geopolitical island (an łAustralia of the north˛), without a proper appreciation of how biodiversity elements and impact factors are distributed continentally, has led to distortions in setting conservation priorities. However, numerous components are already in place that could be integrated into a cooperative national program for developing biodiversity information conformable to that being developed by GAP projects in the U.S. and Mexico. Prospects and impediments for moving forward are discussed.

Noss, Reed F., James R. Strittholt, Kenneth Vance-Borland, Carlos Carroll, and Pamela A. Frost

An approach to conservation planning combining focal species, special elements, and habitat representation analyses was developed and applied to the Klamath Siskiyou bioregion. Important habitat for a viable population of the Pacific fisher was mapped. Natural Heritage Program data were used to map areas with high occurrences of rare and threatened species. GAP vegetation maps were combined with a physical habitat map developed for the region, and a gap analysis of the resulting representation classes was done. An analysis of road data identified roadless areas that became the foundation, along with existing wilderness areas, for our expanded protected areas proposal. Phase I of our reserve design would place 34% of the region under GAP 1 protection status, and another 16.5% under GAP 2 status. Phase II would expand the protected area to include approximately 60% of the region.

USGS ||Gap Analysis Program || NBII ||
Disclaimer || Privacy || Accessibility||
GAP Webmaster ||

GAP Annual Meetings - 1999 Abstracts