banner literature about projects tools meetings search links BBS GAP home GAP home
home

 


STATE PROJECT REPORTS
(Status as of fall 1999)

GapBulletin859-00.jpg 482x270

All completed products and reports will be available through the National GAP Web site at http://www.gap.uidaho.edu/Projects/Data.asp. Drafts and other products may be obtained from the state project PI as noted.

Alabama

Organizing

Anticipated completion date: December 2004

Contact: James B. Grand
Alabama Cooperative Fish and Wildlife Research Unit
Auburn University, Auburn
bgrand@acesag.auburn.edu, (334) 844-4796

A Gap Analysis Project for Alabama was initiated December 8,
1999, with a Cooperator’s Workshop. The anticipated start date is
March 2000.

Alaska

Not started

Arizona

Version 1 project near completion; Version 2 update under way

Anticipated completion date: June 2000; update: July 2004

Contacts: Kathryn A. Thomas, Project Leader
USGS Forest and Rangeland Ecosystem Science Center
Colorado Plateau Field Station, Flagstaff
Kathryn_A_Thomas@USGS.gov, (520) 556-7466 x235

Sarah R. Jacobs, AZ-GAP Update Coordinator
USGS Forest and Rangeland Ecosystem Science Center
Colorado Plateau Field Station, Flagstaff
Sarah.Jacobs@NAU.edu, (520) 556-7466 x240

Original AZ-GAP Project:
The Arizona GAP land cover map, with 54 land cover categories, is
finished and is available at http://srnr.arizona.edu/nbs/gap. The assessment of this map 
was completed through a partnership of Uni-versity of Arizona and Northern Arizona University. The land stew-
ardship map is completed and ready to be incorporated into the
final GAP report. Gap analysis for Arizona has been completed;
the tables and graphs are now being finalized. The AZ-GAP final
report is near completion. Our anticipated date of delivery is the
end of February 2000. We anticipate that all the data will be on the
Web site (listed above) by the end of March.

Southwest Regional GAP Project:
Land cover:
Arizona, Nevada, Utah, Colorado, and New Mexico
are embarking on a multiyear project to update GAP in the South-
west as a coordinated regional effort. The Southwest regional project
will develop new digital map databases, focusing first on the land
cover map. A consistent approach for mapping land cover is essen-
tial for success of a regional gap analysis. Consistency across state
boundaries can be accomplished by using a standardized classifi-
cation system (rather than each state having a unique system) and
mapping zones (rather than state boundaries).

A standardized classification system, the National Vegetation Clas-
sification System (NVCS), will facilitate use of the land cover map
throughout the Southwest. The NVCS is regarded as a major step
toward enhancing our ability to understand, protect, and manage
the natural resources of the United States. It provides a hierarchi-
cal framework for describing vegetation and a convention for iden-
tifying and naming additional vegetation types. A description of
the NVCS is available at http://consci.tnc.org/library/pubs/class/
index.html. A set of preliminary alliance names have been devel-
oped in Arizona, but it is expected that the project will expand and
further define alliances for Arizona.

Arizona will use satellite imagery from the latest earth satellite ob-
servation system, Landsat 7 Enhanced Thematic Mapper Plus, to
delineate preliminary land cover polygons. It is anticipated that the
improved spatial resolution (15 m) from the newly added panchro-
matic band will increase accuracy and definition in land cover map-
ping. Prior to classification, the imagery will undergo preprocess-
ing and stratification at a regional laboratory.

Mapping zones will create a seamless land cover map for the South-
west by dividing the area into ecological rather than administrative
units. The use of mapping zones will maximize information ex-
traction from the satellite imagery by separating the imagery into
smaller, more homogeneous areas prior to classification.

Animal modeling: Arizona will cooperate to produce a prelimi-
nary consolidated list of terrestrial vertebrate species and prelimi-
nary distribution maps for the Southwest region. The SW Gap
Analysis Update five-state team met in January 2000 to develop the
interstate strategy for vertebrate distribution modeling.

Arkansas

Complete (see http://www.cast.uark.edu/gap/)

California

Complete (see http://www.biogeog.ucsb.edu/projects/gap/
gap_home.html)

Colorado

Near completion

Anticipated completion date: June 2000

Contact: Donald L. Schrupp
Colorado Division of Wildlife, Denver
hqwris@lamar.colostate.edu, (303) 291-7277

Land cover: Land cover base coverage is complete and undergo-
ing accuracy assessment. Interpretation of air video for accuracy
assessment is also complete; preliminary traditional and fuzzy ac-
curacy estimates have been calculated for both overall map accu-
racy and accuracy of individual types. Interpretation of these sta-
tistics is under way.

Animal modeling: Animal models have been generated for 597
species, reviewed by cooperators, and applied to generate the spe-
cies/stewardship tables for CO-GAP’s final report.

Land stewardship: The land stewardship coverage for CO-GAP
was finalized with edits based on Bureau of Land Management’s
Areas of Critical Environmental Concern.

Analysis: Standard state report tables have been generated and ana-
lyzed for the CO-GAP final report. Table information was used to
prepare the analysis chapters for the final report, and the chapters
were distributed for team review. Copies were distributed for final
cooperator review in November 1999.

Reporting and data distribution: Development and delivery of
final CO-GAP deliverables to the National GAP Office is on target
for March 2000.

Other accomplishments and innovations: Materials for the CO-
GAP home page (http://ndis.nrel.colostate.edu/cogap/
cogaphome.html) are being developed concurrently with
deliverables for the National GAP Office. The CO-GAP home page
is targeted for public access in February 2000 (http://
ndis.nrel.colostate.edu/cogap/cogaphome.html).

Additionally, Colorado has joined with Arizona, Nevada, New
Mexico, and Utah in initiating the Southwest ReGAP project to
extend state GAP project work to the landscape level for use in
biodiversity planning efforts in the Southwest. Lee O’Brien will be
the Colorado state coordinator for the SW-ReGAP initiative. The
five states have already begun to develop a collective list of their
species models and plan to meet in January 2000 to develop an
integrated workplan.

Connecticut

(see Massachusetts, Connecticut, and Rhode Island)

Delaware

(see Maryland, Delaware, and New Jersey)

Florida

Near completion

Anticipated completion date: June 2000

Contact: Leonard Pearlstine
Florida Cooperative Fish and Wildlife Research Unit
University of Florida, Gainesville
pearlstinel@wec.ufl.edu, (352) 846-0630

Land cover: A 70-class (59 natural classes) land cover classifica-
tion from 1993/94 Landsat Thematic Mapper imagery is complete.
All of the classification has been reviewed by the Florida Fish and
Wildlife Conservation Commission except for two Landsat scenes
of the Florida panhandle that are currently under review. Accuracy
assessment is also being conducted by evaluating low altitude,
geocoded videography and digital photography.

Animal modeling: Habitat-affinity matrices have been completed
and matched to the final land cover classification. The ARC/INFO
AML programs to conduct spatial modeling of species distribu-
tions have been completed and tested. Wildlife habitat models will
be run for all species and should be completed by December 1999.

Land stewardship mapping: The Florida Natural Areas Inventory
(FNAI) of The Nature Conservancy Heritage Program has inde-
pendently compiled GIS coverages of conservation lands for Florida.
GAP ownership and management codes have been added as at-
tributes to the FNAI coverage. FNAI is working with FL-GAP to
translate their protection status rankings to the ranking scheme re-
quired for GAP.

Analysis: We are awaiting completion of the Florida land steward-
ship coverage to begin analysis of the land cover types. Analysis of
the wildlife habitat models will be conducted as the models are
completed. Expected date of completion for the analysis is late
December 1999.

Reporting and data distribution: The land cover and wildlife mod-
eling methodology sections of the final report are nearing comple-
tion. Draft products can be viewed at the FL-GAP Web site at
www.wec.ufl.edu/coop/gap. As products are completed, they will
be available on the FL-GAP Web site.

Publications:

Allen, C., L. Pearlstine, and W. Kitchens. (Accepted). Modeling
viable mammal populations in gap analyses.
Biological Con-
servation.

Georgia

Under way

Anticipated completion date: October 2001

Contact: Elizabeth Kramer
Natural Resource Spatial Analysis Laboratory
Institute of Ecology, University of Georgia, Athens
lkramer@arches.uga.edu, (706) 542-2968

Land cover: The Georgia Gap Analysis Project began in July 1998.
From January 1999 to January 2000 significant progress has been
made in all areas of the project. Specifically, we have:

1. Developed a protocol and progressed in creating a general land
cover map of the state on a county-by-county basis. To date, 36
counties (out of 159) have been completed. Figure 1 below shows
the progress of the state land cover mapping initiative.

Figure 1. Status of the general land cover map for Georgia.

2. Ground-checked two of the above counties with a reported over-
all accuracy of 87%.

3. Continued to refine the list of vegetation alliances to be mapped
in the second iteration of the land cover mapping effort. In some
instances this required aggregating several alliances into a single
mapping unit.

4. Began development of rules to be used in predicting aggregated
vegetation alliances.

5. Continued training seven image processors to implement the veg-
etation mapping protocol.

Over the next 12 months, we anticipate completing the general land
cover map of the state and will implement models of the distribu-
tion of vegetation alliances. Additionally, we will engage in a se-
ries of ground-truth surveys to assess the accuracy of these maps.
We are also working with Dana Slaymaker of the University of
Massachusetts to develop a new color infrared digital video sys-
tem, which we will use to gather high-resolution land cover data in
the spring/summer and fall of 2000.

Animal modeling: To date, a list of Georgia’s vertebrates has been
compiled, and a literature search is under way to determine habitat
requirements for vertebrate species known to breed or winter in
Georgia. These requirements are being entered into a relational
database and will be cross-walked to the general land cover classi-
fication and, when possible, to the more detailed vegetation alli-
ance classification. From this database, individual species models
will be built to determine spatial distributions of the vertebrates in
question. Locational information on rare species across all taxa
will be obtained from the Natural Heritage Program’s element oc-
currences database from the Georgia Department of Natural Re-
sources. Additionally, agreements are being set up with the Geor-
gia Museum of Natural History to obtain a copy of the database of
distributional records for all vertebrates in their collection, includ-
ing fish. GA-GAP has also implemented a pilot program focusing
on the development of methods for aquatic gap analysis.

Land stewardship mapping: With funds provided by the U.S. En-
vironmental Protection Agency, Region 4, land stewardship map-
ping has been completed, and a CD-ROM of the data has been pro-
vided to the National GAP office. See the section on data distribu-
tion for details on how to obtain a copy of this database. To date, 14
agencies and many individuals have requested and received copies
of the Georgia Conservation Lands database.

Analysis: We have begun a preliminary analysis of the distribution
and status of protected lands in the state, addressing such questions
as the proportion and spatial distribution of: protected lands in each
category of GAP status, protected lands in each ecoregion, pro-
tected lands that allow timber harvesting, and a breakdown of the
protected lands by managing authority. This analysis will continue
over the next 12 months as the land cover maps and vertebrate dis-
tribution maps are developed.

Reporting and data distribution: Maintenance of the Georgia con-
servation lands database has been turned over to Chris Canalos of
the Georgia Department of Natural Resources, Wildlife Resources
Division, phone: (706) 557-3032, fax: (706) 557-3033, e-mail:
chris_canalos@mail.dnr.state.ga.us. The database and its associ-
ated metadata can be downloaded from the Georgia GIS data clear-
inghouse (http://www.gis.state.ga.us/) or received via anonymous
FTP from the NARSAL Lab at the Institute of Ecology, University
of Georgia (UGA), following these instructions:

1. ftp greer.ecology.uga.edu, 2. cd pub, 3. bin, 4. prompt, 5. mget
conservation.*, 6. bye.

Other accomplishments and innovations: GA-GAP will provide
the base map for the Georgia Land Use Trends analysis program
(GLUT), a 1-million-dollar project funded by The Turner Founda-
tion. The GLUT project will also be run by Dr. Elizabeth A. Kramer
of UGA’s Institute of Ecology in cooperation with UGA’s Carl
Vinson Institute of Government. The project will retroactively cre-
ate a series of land cover maps for the state of Georgia representing
five increments for the past 25 years to show how Georgia’s rapid
growth has affected land use in urban and rural areas.

Hawaii

Organizing

Anticipated completion date: December 2004

Contact: Samuel M. Gon III
The Nature Conservancy of Hawaii, Honolulu
sgon@tnc.org, (808) 537-4508 x241

Land cover: Initial drafts of land cover have been completed for
native vegetation. The goal is to develop more detailed units for
both native and non-native vegetation. Masks for urban and major
agricultural lands are planned. 

Animal modeling: Compilations of Natural Heritage Program oc-
currence records for endangered birds are complete. We intend to
convene working groups to develop range maps for birds, bats, and
selected native invertebrates.

Land stewardship mapping: Major land ownership patterns for
the state are completed. We will develop management classes
through working group meetings and assign these attributes to ex-
isting managed land units.

Analysis: We intend to adapt analysis algorithms of other states to
apply to the higher resolution needed for small tropical insular sys-
tems. We will convene the analysis working group on this matter.

Idaho

Version 1 published in the peer-reviewed literature. Version 2 up-
date near completion.

Anticipated completion date: June 2000

Contact: Leona K. Bomar
Idaho Cooperative Fish and Wildlife Research Unit, Moscow
leonab@uidaho.edu, (208) 885-5788
http://www.wildlife.uidaho.edu/idgap.htm

Land cover: The ID-GAP land cover layer and final report chapter
are complete. The land cover classification recognizes 81 cover
types and is mapped at a resolution of 0.09 ha with a 2 ha MMU.

Animal modeling: Wildlife habitat relationship models have been
completed for 375 terrestrial vertebrates in Idaho. The models are
stored as georeferenced TIFF images with a native resolution of
0.09 ha.

Land stewardship mapping: The revised Idaho land stewardship
database is also complete. This data set represents a significant
improvement over the original Idaho land stewardship layer by in-
creasing spatial resolution to a 2 ha MMU and incorporating many
of the smaller managed areas in Idaho.

Analysis: Analysis of the protection status of Idaho’s land cover
types is complete. Analysis of wildlife habitat distributions will be
completed in early December 1999.

Reporting and data distribution: The final report for ID-GAP is
awaiting completion of the wildlife habitat distribution analysis.
All data, metadata, and documentation are currently available for
download from the URL above or by contacting the Idaho Coop-
erative Fish and Wildlife Research Unit.

Other accomplishments and innovations: We have conducted a
gap analysis of geomorphologic and climatic features in Idaho. We
will compare the results of this analysis with those from land cover
and wildlife habitat analyses.

Illinois

Under way

Anticipated completion date: December 2001

Contacts: Jocelyn L. Aycrigg, GAP Coordinator
Illinois Natural History Survey, Champaign
aycrigg@mail.inhs.uiuc.edu, (217) 244-2111

Linda Schwab, Assistant GAP Coordinator
Illinois Natural History Survey
schwab@mail.inhs.uiuc.edu, (217) 265-8425

Land cover: A general land cover classification for the state was
completed in October 1995. This classification identified 19 land
cover classes: four urban, three forest and woodland, three agricul-
ture, two grassland, five wetland, and two other categories (water
and barren areas). The alliance-level GAP vegetation classifica-
tion is being performed by stratifying along land cover classes.
Classification protocols are similar to protocols for UM-GAP (see
Bulletin No. 5, p. 35). Classification to the community/alliance
level has been completed for southern Illinois. We have concen-
trated our efforts on the western side of the state along the Missis-
sippi River near East St. Louis. We have obtained ancillary data
such has DEMs and forest inventory information, which have as-
sisted us in our classification. We plan to classify TM scenes north-

ward along the Mississippi River as well as in northeastern Illinois
in the coming year. Furthermore, we will explore methods for con-
ducting accuracy assessment in anticipation of completing the veg-
etation classification.

Animal modeling: We have created a list of amphibians, reptiles,
and mammals to be mapped. We are using specimens collected by
the Illinois Natural History Survey (INHS) and the University of
Illinois Museum of Natural History to obtain locational records for
each species. We have also obtained museum records from the
Smithsonian, American Museum of Natural History, Chicago Field
Museum, Kansas State Museum of Natural History, and Museum
of Vertebrate Zoology at Berkeley. Furthermore, we have obtained
input from wildlife habitat biologists throughout the state regard-
ing species known to occur in their district. We have completed
mapping the amphibian, reptile, and mammal collection of the Illi-
nois Natural History Survey and the University of Illinois Museum
of Natural History. We are currently mapping the mammal collec-
tions of Illinois from other museums. We plan to conduct an expert
review of the amphibian and reptile range maps as well as finish up
the habitat associations. We will also continue gathering habitat
association information for each mammal species. Information
gathered previously for the Illinois Fish and Wildlife Information
System will be helpful in developing habitat associations. We will
use the breeding bird survey for Illinois and the Illinois Breeding
Bird Atlas to create a list of bird species to be mapped and begin
delineating ranges for those species.

Land stewardship mapping: We have developed a land steward-
ship map for Illinois, attributed general ownership categories, and
assigned management status levels. The GAP coding scheme for
land units has been assigned to each property. The database needs
to be reviewed to determine if all properties have been included.

Analysis: We have completed some preliminary analyses using
amphibian, reptile, and mammal locational data to create species
richness maps using the EMAP hexagons. We have also started
some analyses of species that occur only in southern Illinois, for
which we have a completed alliance-level classification. We will
continue to do more analyses as our species and vegetation map-
ping progresses.

Reporting and data distribution: In the coming year, we will fin-
ish the statewide alliance-level classification of vegetation and con-
tinue our work on the species modeling. We will start writing seg-
ments of the report in the coming year.

Other accomplishments and innovations: A Web page for the
Illinois Gap Analysis Project has been created and can be viewed at
www.inhs.uiuc.edu//cwe/gap/gapintro.html. We were able to ob-
tain new cooperators with the assistance of the Illinois Department
of Natural Resources. The U.S. Department of Agriculture’s Na-
tional Agricultural Statistics Service and the Illinois Department of
Agriculture have agreed to work with us to obtain and classify state-
wide TM imagery.

Presentations on the Illinois Gap Analysis Project were given to the
senior agency managers of the Illinois Department of Natural Re-

sources, the Illinois Board of Natural Resources, the Illinois Chap-
ter of The Wildlife Society, and a University of Illinois class on
Ecosystem Management.

Listed below are projects that are starting up, ongoing, or have been
completed using the Land Cover Database of Illinois as well as
other data developed as part of GAP.

• Modeling wild turkey habitat in the Illinois landscape. T. Van
Deelen, P. Brown, M. Joselyn, D. Greer, T. Maples, and J. Garver.

• Dispersion of gray and fox squirrels. D. Rosenblatt and E. Heske.

• Fox and coyote ecology in central Illinois. T. Gosselink, T. Van
Deelen, R. Warner, and P. Mankin.

• Habitat use by raccoons and opossums. E. Heske and D.
Rosenblatt.

• Identification and classification of critical wildlife habitat. P.
Brown, M. Joselyn, J. Aycrigg, L. Suloway, and B. Zercher.

Indiana

Near completion

Anticipated completion date: August 2000

Contact: Forest Clark
U.S. Fish and Wildlife Service, Bloomington
forest_clark@fws.gov, (812) 334-4261 x206

Land cover: The Indiana land cover data layer is complete and is
being used in numerous projects, including as a foundation data
layer in the regional assessments under way by the Indiana
Biodiversity Initiative. These data have also been made available
on request to agencies, consultants, and NGOs. We hope to have
this and the other primary data layers on the Web early in 2000.

Animal modeling: The Indiana vertebrate models were completed
and integrated with the vegetation map to produce a draft product
for review. A panel reviewed the models and identified those for
which improvements were possible with the data available to the
project. These revised models have been drafted and are currently
being converted to digital form. We plan to run the entire set of
vertebrate models early in 2000.

Land stewardship mapping: The Indiana land stewardship data
layer is complete. Coding of stewardship types based on a matrix
developed for the project is ongoing. The basic protection status
coding is complete, and these data are ready for the analysis phase.

Analysis: The Indiana project will conduct the analysis over the
winter 2000.

Reporting and data distribution: The Indiana project has been
working on various sections of the final report and proposes to fin-
ish the first draft by late spring 2000.

Other accomplishments and innovations: The Indiana
Biodiversity Initiative, which is using GAP data extensively in its

assessment of biodiversity, recently received two major grants to-
taling over $60,000 to continue the Initiative’s work. The Indiana
Biodiversity Initiative functions as the implementation arm of Gap
Analysis in Indiana. The Grand Kankakee Marsh National Wild-
life Refuge, which used GAP data from both Indiana and Illinois in
its design, was approved by the Regional Director of Region 3 of
the U.S. Fish and Wildlife Service in August 1999.

Iowa

Under way

Anticipated completion date: December 2001

Contacts: Bruce Menzel, Co-Principal Investigator
Department of Animal Ecology
Iowa State University, Ames
bmenzel@iastate.edu, (515) 294-7419

Kevin Kane, Co-Principal Investigator
Manager, ISU GIS Support and Research Facility
Iowa State University, Ames
kkane@iastate.edu, (515) 294-0526

The Iowa Gap Analysis Project (IA-GAP) is in its third year. An
IA-GAP home page is accessible at http://www.ag.iastate.edu/cen-
ters/cfwru/iowagap/.

Land cover: Preparation of the land cover map has not progressed
as rapidly as we had hoped, thus we are revising our anticipated
completion date for the land cover to December 2000. Measures
have been taken to speed up the unsupervised classification process
by reducing the number of ambiguous cover classes that require
relatively longer amounts of time to differentiate. We reduced the
number of Iowa land cover classes that can reasonably be mapped
to 29. Three Landsat scenes, representing about 40% of the land
area of Iowa, have been mapped with these cover classes.

In 1998, we completed the aggregation of National Wetland Inven-
tory data into five major classes of wetlands (temporary, seasonal,
semipermanent, permanent, and open water). In 1999, this data
layer was integrated into the Phase 1 land cover map for the entire
state. Wetland vegetation is then classified using Wetland Inven-
tory codes and Landsat imagery with a recoding and overlay tech-
nique.

In summer 1999, we conducted field surveys of natural and
seminatural vegetation in 32 additional counties across Iowa with
the majority being in the western part of the state. These data were
digitized and will be used to assist in assigning map labels to classes
generated during the unsupervised classifications.

The IA-GAP staff continues to meet biannually with staff from
Kansas, Missouri, Nebraska, South Dakota, and North Dakota to
share data and information on polygon edge-matching, legend com-
patibility, accuracy assessment, and other problems that are com
mon 
to GAP in the Great Plains Region.

Vertebrate modeling: Species lists for mammals, birds, reptiles,
and amphibians are compiled and include state, federal, and global
status codes for each species. These lists, which are ready for ex-
pert review, appear on the IA-GAP Web site. Selected scientists
from across the state are being invited to serve as expert reviewers
of the species lists and habitat models. During 1999 we continued
to assemble historical data from museum and private collections
and to compile information from the literature on species occur-
rence, geographic ranges, and optimal habitat requirements.

Modeling of Iowa vertebrates has begun. Our goal is to complete
the modeling effort by December 2000 and begin the preparation
of distribution maps in January 2001, soon after the land cover map
is complete.

Land stewardship: We continued to acquire data on legal bound-
aries of federal, state, and county lands in Iowa. Acquisition of
county properties has consumed the most time; of 99 counties, 60
have been completely digitized, 20 are near completion, and data
for an additional 8 counties have been acquired but not digitized.
Data have been requested from the remaining 11 counties. State
land boundaries were initially digitized by the Iowa Department of
Natural Resources in 1991. This database has been recently up-
dated by the Department of Natural Resources (DNR) to include
new acquisitions and other changes up to mid-1999 and will be
made available to IA-GAP in January 2000. Boundary information
for federal lands has been obtained from the U.S. Fish and Wildlife
Service and the U.S. Army Corps of Engineers. We anticipate
completion of the land stewardship layer for IA-GAP in summer
2000.

Other accomplishments: Extension wildlife staff at Iowa State Uni-
versity, Department of Animal Ecology, with the help of an educa-
tional grant from the Iowa Resource Enhancement and Protection
program, conducted eight NatureMapping training workshops across
the state in 1999; 110 people were trained. The program has been
enthusiastically received by a diverse group of people, including
school teachers and environmental educators, landowners, nonprofit
conservation groups, businesses, and agencies. A database and Iowa
NatureMapping Web site are currently under development.
NatureMapping data will be sent for accuracy assessment to Iowa
State University via the NatureMapping Web site, then stored in a
publicly accessible database at the State of Iowa GIS Office server.
Additional funding is being sought to carry NatureMapping to an-
other level during the coming year.

In spring 1999, EPA Region 7 agreed to partially fund a pilot study
to develop field and analytical techniques for accuracy assessment
using ground surveys and the pixel (30 x 30 m) as the basic sam-
pling unit. The statistics laboratory at Iowa State University is de-
veloping a protocol for sampling design and data analysis to be
used in a regionwide approach for accuracy assessment of land cover
maps. In March 1999, coordinators from Missouri, Iowa, Kansas,
and Nebraska met with EPA administrator Marla Downing and stat-
isticians Sarah Nusser, Iowa State University, and Steve Stehman,

State University of New York-Syracuse, to discuss procedures for
the pilot field study conducted in 1999. We selected a four-county
area in NE Iowa to conduct our portion of the pilot study and col-
lected data from 145 sites in September and October 1999. Each
randomly selected pixel was located using a real-time GPS unit.
Each pixel in a nine-pixel matrix (target pixel in the center) was
assessed for vegetation and land cover and assigned a class corre-
sponding to one of the 29 map labels. The results of the pilot study
are expected to be available by March 2000, when a decision will
be made whether to proceed with a regionwide accuracy assess-
ment.

Kansas

Under way

Anticipated completion date: April 2001

Contact: Glennis Kaufman
Kansas State University, Manhattan
gkaufman@ksu.edu, (785) 532-6622

Land cover: Currently, mapping of the state has been completed to
the cropland/natural vegetation stage. All of the land cover has
been classed to the alliance level, but we continue to refine the cover
classes. The land cover layer will be completed in December 1999.
During the following nine months, accuracy assessment of the land
cover layer will be ongoing with a heavy investment of time occur-
ring during summer 2000. 

Animal modeling: Currently, species lists of terrestrial vertebrates
to be mapped are finalized following expert review. Range distri-
bution maps are being generated based on museum collections,
publications, and current observations. We estimate that 85% of
the literature has been collected for vertebrate models, and we are
in the process of building wildlife habitat relationship models both
for the state and Great Plains region. Ancillary databases also are
beginning to be discussed and built for use in the predicted habitat
relationship models. During the next nine months, all wildlife habitat
relationship models will be completed. Drafts of predicted species
distributions will be completed and sent out for expert review. Fol-
lowing expert review, final species distribution maps will be gener-
ated.

Land stewardship mapping: Currently, the land stewardship layer
is near completion. Polygon boundaries are done for 210 managed
areas throughout the state. As more land is purchased for conserva-
tion in the state, we will update the layer to keep it current. A
number of attributes (e.g., owner, county of location) already is in
the database that goes along with polygons. We are in the process
of querying land unit managers in cooperating agencies about the
level of protection in each area and will add this attribute informa-

tion as it is obtained. Our goal is to complete the stewardship layer
by June 2000.

Analysis: Analysis will be conducted by the end of 2000.

Reporting and data distribution: Reporting and data distribution
are expected to be initiated by fall 2000, and we anticipate asking
for an extension to complete the final report.

Other accomplishments and innovations: We have created a Great
Plains ACCESS database for modeling of vertebrates within the
Great Plains states (North Dakota, South Dakota, Nebraska, Kan-
sas, and Iowa). This database should allow us to produce seamless
maps of the predicted distributions of Great Plains vertebrates in
the future.

Kentucky

Under way

Anticipated completion date: June 2002

Contacts: Tom Kind (land cover)
Murray State University
tom.kind@murraystate.edu, (270) 762-3110

Terry L. Derting (animal modeling)
Murray State University
terry.derting@murraystate.edu, (270) 762-6327

Keith Wethington (land stewardship mapping)
Kentucky Department of Fish and Wildlife Resources, Frankfort
keith.wethington@mail.state.ky.us, (502) 564-7109

Land cover: All data layers for one physiographic province, the
Shawnee Hills of the Interior Low Plateaus, have been produced.
Layers include interpreted air video points, classified TM data, con-
vexity/concavity, slope, aspect, and National Wetlands Inventory
(NWI). A draft of a detailed vegetation map containing natural
vegetation map units for the Shawnee Hills has been produced, us-
ing a decision tree process, and is under review. Air video flight
lines in other portions of the state have been ground-truthed, and
air video has been recently collected over the eastern part of the
state. The short-term goal for the coming year includes incorporat-
ing urban (high and low density), agriculture (pastureland/grass,
cropland), and mined lands (bare ground, revegetated) into the draft
map. Completing a draft map of the entire state is the major goal
for the coming year.

Animal modeling: Considerable progress was made in 1999 to-
ward modeling vertebrate distributions in Kentucky. We finalized
the list of terrestrial vertebrate species to be modeled after its re-
view by our state experts. Our final species list includes 51 reptile,
52 amphibian, 63 mammal, and 198 bird species (364 species to-
tal). Range maps of current distributions were delineated. A Web
page for the review of the range maps was constructed with the aid

of the Kentucky Department of Fish and Wildlife Resources
(KDFWR), and the review process for the ranges has been com-
pleted. Final range maps will be completed in December 1999.
We are completing our database of the habitat association for each
species. We are also establishing a Web page, with the aid of
KDFWR, for the review of our habitat association information. The
habitat associations will be posted for review by the end of 1999.

In addition to the required tasks of Gap Analysis, we are using the
species range maps to map species richness in the ecoregions and
physiographic provinces of Kentucky. We have presented our analy-
ses of species richness at the 2nd Annual Biodiversity Conference
at Western Kentucky University and the 1999 Kentucky Academy
of Science meeting at Eastern Kentucky University. Through meet-
ings such as these we maintain open communication with biolo-
gists, educators, and other interested parties in the state.

The database of final habitat associations will be produced during
2000. Also, we will produce an initial wildlife/habitat relationship
model for each species. We plan on making draft maps of the pre-
dicted occurrence of species depending upon the status of the veg-
etation mapping for Kentucky.

Land stewardship mapping: Work on the stewardship layer has
focused on assessment of data available. Agencies at the federal,
state, and local government levels as well as nongovernment orga-
nizations were contacted. Available information regarding data for-
mat (digital or analog), management status, contact names, etc. were
obtained. Existing digital coverages will be acquired and converted
to a common projection during 2000. The large amount of nondigital
ownership data makes it improbable that all managed lands in Ken-
tucky will be included in the final analysis. A method for prioritiz-
ing the conversion of analog to digital data, focusing on conserva-
tion lands, will be developed.

Louisiana

Near completion

Anticipated completion date: June 2000

Contact: Jimmy Johnston (project leader)
USGS/National Wetlands Research Center, Lafayette
jimmy_johnston@usgs.gov, (318) 266-8556

Steve Hartley (land cover, analysis)
USGS/National Wetlands Research Center, Lafayette
steve_hartley@usgs.gov, (318) 266-8543

Land cover: The land cover section for the final report is 98%
complete, with an anticipated completion date of December 1999.
Data directory structure for final CD distribution was completed in
October 1999. We are working on the metadata for ancillary data
sets.

Animal modeling: Species distribution maps were reviewed and
corrected. Final digital data were delivered to the project office in
August 1999. Hyper-distribution maps were created for all four
major vertebrate groups (mammals, birds, reptiles, and amphibians)
in October 1999. Data directory structure for final CD distribution
was completed in October 1999. The animal modeling section for
the final report is 98% complete, with an anticipated completion
date of December 1999. We are still working on the metadata.

Land stewardship mapping: The data directory structure for final
CD distribution was completed in October 1999 for Louisiana’s
land stewardship and management. The Land Stewardship and
Management section for the final report is 98% complete, with an
anticipated completion date of March 2000.

Analysis: The analysis of land cover types and vertebrate species
by land stewardship and management status was completed in Oc-
tober 1999. Data directory structure for final CD distribution was
also completed in October 1999. The Analysis section for the final
report is currently under way, with an anticipated completion date
of December 1999.

Reporting and data distribution: Currently, the National Wetlands
Research Center is in the process of writing the final report, with an
anticipated completion date of April 2000.

Maine

Complete (see http://wlm13.umenfa.maine.edu/progs/unit/gap)

Maryland, Delaware, & New Jersey

Near completion

Anticipated completion date: October 2000

Contact: Ann Rasberry (land cover)
Maryland Department of Natural Resources, Annapolis
arasberry@dnr.state.md.us, (410) 260-8558

Rick McCorkle (animal modeling)
Delaware Bay Estuary Project
U.S. Fish & Wildlife Service
richard_mccorkle@fws.gov, (302) 653-9152 x17

Timothy A. Palmer (land stewardship)
Maryland Department of Natural Resources
tapalmer@dnr.state.md.us, (410) 260-8559

Land cover: During 1999 the land cover mapping made signifi-
cant progress. The Delmarva Peninsula, the
piedmont, southern New Jersey, and eastern mountains of Mary-

land have been completed as draft maps. The remainder of Mary-
land and New Jersey will be completed early in 2000, and accuracy
assessment will begin at that point.

Attempts to install the necessary hardware to import our video into
a GIS format are under way. This will enable additional video to be
used in the classification process where GPS capture was intermit-
tent during video flights (e.g., over the mountains in western Mary-
land).

A project Web page has been developed and will be available soon
via link to the national GAP Web page. Status of the project as well
as availability of products may be determined from information
found at that location.

Animal modeling: In 1999, most of the vertebrate modeling work
involved finalizing habitat layers, completing development of the
database, and fine-tuning the modeling software. In addition to
land cover-derived habitats, the habitat layers used in the modeling
include wetland/riparian buffers, elevation, forest area (measured
by thickness), riparian forest width, forest isolation, edge habitats,
road density, land use (e.g., pasture), aquatic habitats, special habi-
tat features (e.g., cliffs), and other special habitats that are gener-
ally smaller than the land cover MMU (e.g., vernal pools). The
feasibility of developing other potentially important habitat layers
(e.g., soils) is still being evaluated.

The database includes tables for species-unique codes, species
modeling status, species range, species/habitat associations by taxa,
forest metrics relationships (e.g., for forest interior dwelling birds),
species guilds, species taxonomy, habitat types, habitat cross-walks
(e.g., to alliance), species/landform relationships, species/wetland
buffer relationships, species-special habitat features relationships,
and several other tables to be used in the modeling or for various
queries. The database was developed in Personal Oracle8 and has
been successfully imported to MSAccess. The modeling software
was developed through ArcView Avenue scripting to enable as much
automation of the modeling as possible and includes some nice
custom query capabilities. Draft models and distribution maps will
be sent out for expert review during the winter of 1999/2000, and
final models will be run by April 2000.

Land stewardship mapping: During 1999 the stewardship layer
for the Maryland portion of the project was completed. The Dela-
ware portion is essentially complete; the work is under review as of
November 1999. For New Jersey, private conservation lands data
have been collected, but the public lands data are still outstanding.
The land stewardship layer is expected to be completed for Dela-
ware by the end of 1999 and for New Jersey in early 2000.

Analysis: GAP investigators expect to complete analysis of the pro-
tection and management status of biodiversity in Maryland and
Delaware by spring 2000 and expect the New Jersey analysis to be
completed by late summer 2000.

Reporting and data distribution: The final report writing for Mary-
land, Delaware, and New Jersey has begun and should be com-
pleted by October 2000. The GIS layers for Maryland and Dela-
ware will be distributed in ARC/INFO/ArcView format, with asso-
ciated data tables in Microsoft Access format, on CD-ROM. A
similar approach will probably be used in New Jersey. The status
and availability of these products will be reported via the GAP Web
page.

Other accomplishments and innovations: By developing our ver-
tebrate modeling scripts in ArcView’s Avenue, we have provided
both a user-friendly interface for accessing data tables and custom
control over the modeling process. Users can query for species/
habitat relationships or maps of species distributions, or survey the
base GIS layers used in the modeling; lists of species occupancy
can also be generated from input of habitat information. In addi-
tion, users can query the community alliance and wildlife habitat
tables, generating a list of potential alliances/habitat by input of
plant species, location, or physical characteristics of the site. Ef-
forts are under way to make this software package portable for even-
tual distribution of the GAP data sets.

Data sets are being used for considerable analysis even in their draft
form. Examples include identifying areas with high potential for
supporting forest interior neotropical migrant songbirds on the
Delmarva Peninsula, priority sites for forest restoration and/or ac-
quisition, and a multiagency federal, state, and private collabora-
tion on the Delmarva Peninsula (the Delmarva Conservation Corri-
dor Initiative).

Massachusetts, Connecticut, &
Rhode Island


Complete; update under way

Anticipated completion date: June 2000

Contact: Curtice Griffin and John Finn
University of Massachusetts, Amherst
cgriffin@forwild.umass.edu (413) 545-2640
finn@forwild.umass.edu (413) 545-1819

Land cover: An accuracy assessment of the land cover map com-
pleted in 1997 is now under way. Preliminary results indicated that
there were significant classification errors, especially in the Cape
Cod region. An error model is being developed from the accuracy
assessment project. A revised land cover map will be developed
from this error model. Additionally, development of a new land
cover map is planned within the next 18 months as part of an NSF-
funded project with the Department of Computer Science at the
University of Massachusetts.

Animal modeling: With completion of the expert review of mam-
mal range maps during summer 1998, all vertebrate models are
complete. Predicted habitats for all 273 vertebrates modeled in the
Southern New England region were identified and coarse species
richness maps developed for each taxonomic group. Additionally,

habitats were identified throughout the Connecticut River water-
shed for priority species of neotropical migratory birds. We plan to
redo the vertebrate habitat maps once the revised land cover map is
available.

Land stewardship mapping: All conservation lands in the region
are mapped and classified according to conservation status. The
database for Connecticut was not as well developed as for Massa-
chusetts and Rhode Island. Over 15% of Southern New England is
classified as conservation lands, and about 7% of the land area was
classed in the categories 1 & 2.

Analysis: Species richness analyses have been completed. The fi-
nal gap analysis will be complete by June 2000.

Reporting and data distribution: All data layers are currently avail-
able on the Southern New England Gap Analysis home page (http:/
/outsider.fnr.umass.edu/gaphome.html). We plan to distribute the
vertebrate models and predicted habitat data layers as an ArcView
project. Revision of the land cover map will begin in April 2000.
The final vegetation map will not be distributed on CD until revi-
sions are complete. A manual for incorporating GPS-logged aerial
videography into land cover mapping efforts is under development
and will be distributed on CD-ROM in May 2000. The final report
and a report on accuracy assessment will be available in June 2000.

Other accomplishments and innovations: Conducted two one-
day workshops for regional planners on “Gap Analysis in the Con-
necticut River Watershed: Landscape-based Approaches for Con-
serving Biodiversity.“ The workshops were funded in part by the
Silvio O. Conte National Fish and Wildlife Refuge and the Univer-
sity of Massachusetts.

Michigan

Under way

Anticipated completion date: September 2002

Contact: Mike Donovan
Michigan Department of Natural Resources
Michigan Resource Information System, Lansing
donovanm@state.mi.us, (517) 335-3445

Land cover: Classification of the Northern Lower Peninsula is ex-
pected to be completed in the second quarter of FY 2000. Classifi-
cation of the Southern Lower Peninsula will begin in the first quar-
ter of FY 2000. Classification of the Eastern Upper Peninsula is
expected in the second or third quarter of FY 2000. The Western
Upper Peninsula is completed in a preliminary version that upon
review may require revision.

Animal modeling: The animal modeling effort will begin in FY
2000 with an expected collaboration with Michigan State Univer-
sity.

Land stewardship mapping: The stewardship layer was begun in
1996 and continues. Completion is expected in summer of 2000.

Analysis: Gap analysis is scheduled to begin in FY 2001.

Reporting and data distribution: Land cover and stewardship data
will begin to be served by UMESC in the second and third quarter
of FY 2000.

Minnesota

Under way

Anticipated completion date: September 2002

Contact: David Heinzen
Minnesota Department of Natural Resources
Resource Assessment Unit, Grand Rapids
david.heinzen@dnr.state.mn.us, (218) 327-4449 x222

Land cover: The state has been divided into 29 spectrally consis-
tent classification units (SCCUs) based on a procedure described in
the Upper Midwest GAP Image Processing Protocol. Ten classi-
fied SCCUs, complete with accuracy assessment, are scheduled for
delivery to UMESC in the first quarter of FY 2000.

Animal modeling: A working group of species experts has been
formed by the Wildlife Section of the Minnesota DNR. Work to-
wards the development of the vertebrate species distribution mod-
els will begin in FY 2000.

Land stewardship mapping: Preliminary public ownership/stew-
ardship mapping is completed. The map is based on an ARC/INFO
vector coverage of the Public Land Survey. Section corners are
georeferenced. From those, a polygon grid of 40-acre tracts has
been generated. Each 40-acre polygon is then attributed for owner,
manager, stewardship category, etc. In FY 2000 UMESC will cross-
walk the MN DNR public owner categories to the official GAP
categories. Independent verification and cross-referencing of the
Minnesota DNR lands stewardship classification will begin in FY
2000.

Analysis: Gap analysis is scheduled to begin in FY 2001.

Reporting and data distribution: Land cover and stewardship data
will begin to be served by UMESC in the second and third quarter
of FY 2000.

Mississippi

Near completion

Anticipated completion date: January 2001

Contact: Francisco J. Vilella (principal investigator)
MS Cooperative Fish and Wildlife Research Unit
Mississippi State University
fvilella@cfr.msstate.edu, (662) 325-0784

Richard B. Minnis (coordinator)
MS Cooperative Fish and Wildlife Research Unit
Mississippi State University
rminnis@cfr.msstate.edu, (662) 325-3158

Land cover: Land cover mapping has been completed. The final
map product has been aggregated to 43 classes. Among them are a
number of classes discerning structure or age of the overstory veg-
etation. Particularly, four classes have been obtained that relate to
the age of Mississippi’s pine ecosystems. Although this informa-
tion is not required for the current mapping standards, the person-
nel from the Spatial Information Technologies Laboratory at Mis-
sissippi State University believed they could extract this informa-
tion from the TM images. These data have proven very useful in
refining vertebrate distributions for species such as Red-cockaded
Woodpecker (
Picoides borealis ) and Bobwhite Quail ( Colinus
virginianus
). Accuracy assessment is nearly complete and should
be finalized by December 1999. Map classes have been associated
with NVCS alliances for each of the five ecoregions in the state.
Metadata construction is in progress.

Animal modeling: Wildlife habitat relationship models are being
applied to the final vegetation layer. Finished models are under
review by vertebrate committees. Amphibian and mammalian spe-
cies are finished pending final review. Bird species are still under
scrutiny by the review team. Ancillary data from 900 bird point
counts conducted in Mississippi’s national forests and a large col-
lection of museum records have aided in the refinement of habitat
relationships and distributions. Models are expected to be final-
ized and reviewed by early spring 2000.

Land stewardship mapping: The land stewardship layer for the
state is complete. The Mississippi Department of Wildlife, Fisher-
ies, and Parks provided MS-GAP with GPS-located boundaries for
all state and federal lands. Additional properties, such as TNC and
tribal lands, have also been incorporated. Metadata construction is
near completion for these data.

Analysis: Macros have been written to perform the final analyses
for the project. Testing of the macros on preliminary distributions
has proven effective. All analyses should be complete by late spring
2000.

Reporting and data distribution: The final report is being written
as phases of the project near completion. The land cover section is
nearly complete and will be finalized with the accuracy assessment.
Sections on stewardship and vertebrate distributions are currently
in progress with completion expected in late spring 2000.

Other accomplishments and innovations: Through the process
of data acquisition and sharing and development of strong working
relationships with MS-GAP cooperators, a movement has started
in the state to provide “life after GAP.” Cooperative research ef-
forts to apply knowledge and expertise in the field of spatial tech-
nology to natural resource management (such as applying spatial
technologies to conservation law enforcement, aiding Partners-in-
Flight with establishing priority lands, and optimizing locations for

aquaculture facilities) have prompted an initiative to be placed be-
fore the Mississippi legislature. The initiative, called the “Natural
Resource Decision Support System,” would provide recurring funds
to support three new employees at Mississippi State University.
These persons would provide biannual updates and refinements to
the MS-GAP land cover and provide support to agencies in terms
of applying spatial technologies and existing spatial data to help
solve current natural resource problems. This initiative would pro-
vide needed information to policy makers and resource managers
to make scientifically based and ecologically sound management
decisions.

Missouri

Complete. Publication products under development.

Contact: Timothy L. Haithcoat
Geographic Resources Center and Missouri Spatial Data
Information Service, University of Missouri, Columbia
haithcoatt@missouri.edu, (573) 882-1404

Land cover: Complete. Phase I land cover was completed by the
Missouri Resource Assessment Partnership. This was then gener-
alized from the original 30 m cell size to the 2 ha minimum map-
ping unit used for vertebrate modeling. Also, 5, 10, 20, 40, and 100
ha land cover bases were created using an in-house generalization
AML to try to maintain some of the linear features and connectiv-
ity present in Missouri’s file. Phase II land cover contains 46 classes
and will be provided in summer 2000 as an update on the National
GAP home page.

Animal modeling: Complete. Among the 348 vertebrates mod-
eled were 66 mammals, 164 birds, 74 reptiles, and 44 amphibians.
These models were created using the base land cover map of 2 ha
and derivative products such as ecotones, core interiors, regional-
ized areas, and riparian measures. Other data layers were devel-
oped to aid in the modeling, such as wetland components, prairies,
soils, precipitation, temperature, population density, road density,
as well as many other feature types such as caves, springs, etc. An
additive weighting model was developed for each species, which
calculated an index to occurrence. It is somewhat similar to a Bayes
Theorem approach in that it uses subjective weights in an additive
fashion. These weights represent what the literature, biologists,
and the modeler believe to be the likelihood of a species occurring,
given that a specific habitat or landscape component exists.

Land stewardship mapping: Complete. The stewardship layer
was created by the Missouri Resource Assessment Partnership.
Public lands comprise only 6.7% of Missouri with 4.7% under fed-
eral and 2% under state jurisdiction. The average size of these hold-
ings is very small. Less than 1% of Missouri falls within areas
designated as management status 1 or 2. All areas greater than 16
ha were analyzed for biodiversity components.

Analysis: Complete. Species biodiversity as measured by species
richness was calculated on five spatial extents. These included
county, EMAP hexagons, 7.5-minute quadrangle, quadlet (1/6 of
quadrangle), and public land survey section. These results were
comprehensively reviewed against the Missouri Fish and Wildlife
Information Systems county lists for all terrestrial species to pro-
vide a measure of reliability and consistency. A more refined re-
view was conducted against the results of the Breeding Bird Atlas
project at the quadlet level for all bird species. This was followed
by the analysis of these richness indices and individual species pre-
dicted occurrences within the context of stewardship. Species tied
to grassland complexes are the most at risk, followed by those spe-
cies that require larger contiguous areas. Wetlands have high rich-
ness, but there are few areas outside preserves where these habitats
exist in Missouri. Focus for Missouri’s future biodiversity efforts
must be on private land holders as they control over 90% of the
state.

Reporting and data distribution: Completed. The draft final re-
port was submitted in January 2000. Data distribution from these
analyses will be posted on the Missouri Spatial Data Information
Service (MSDIS) at http://msdis.missouri.edu. A link will be cre-
ated on this site to the AMLs and programs written in support of
this effort.

Other accomplishments and innovations: The methods used in
the generalization of the land cover are fairly unique as they were
developed to try to capture the linear “travelways” present across
many of Missouri’s landscapes. The program was developed in
AML and utilizes ARC/INFO processing in both grid and vector
environments. Programs were developed to take advantage of the
matrix (i.e., agricultural, grassland, or forested) within which the
process was being run. In addition, regional variables were created
to allow for the maintenance of polygons within which only diago-
nal connectivity exists.

The methods developed for vertebrate modeling expand the usabil-
ity of the data sets by acknowledging and incorporating the fact
that these data are fuzzy and that “habitat” is not always associated
with a single cover type but rather a suite of cover types in juxtapo-
sition that can provide for all the species’ needs. The modeling
approach of mapping this as a continuum of response (occurrence)
and then not clipping these results to a range map but rather cali-
brating these measures to a range greatly expands the users’ flex-
ibility with these data as well as not perpetuating sampling errors
or biases that can permeate the range information.

Montana

Complete (see http://www.wru.umt.edu/reports/gap/)

Nebraska

Under way (http://www.calmit.unl.edu/gap/)

Anticipated completion date: October 2001

Contacts: Geoffrey M. Henebry
CALMIT, University of Nebraska-Lincoln
ghenebry@calmit.unl.edu, (402) 472-6158

James W. Merchant
CALMIT, University of Nebraska-Lincoln
jmerchant1@unl.edu, (402) 472-7531

Land cover: We have completed a preliminary statewide classifi-
cation and are in the process of refining the product. We have been
working with Iowa, South Dakota, and Kansas on accuracy assess-
ment strategies. We conducted a pilot study of accuracy assess-
ment during summer 1999 and are currently planning for the sum-
mer 2000 accuracy assessment field campaign. In addition, we are
working with the USDA/NRCS to get county-level land cover draft
maps out to NRCS field offices across the state to solicit comments
from local experts. Plans for next 12 months: finalize land cover
mapping, including integration of NWI data, and conduct land cover
accuracy assessment.

Animal modeling: We have completed assembly of species lists
for the state. The herpetiles list has undergone expert review and
has been revised. The bird and mammal lists are currently in expert
review. We have initiated model development for reptiles and am-
phibians; based on preliminary results we are incorporating climate
and soils data as complementary environmental variables. We are
cooperating with the Great Plains Regional GAP modeling group
to develop common models and modeling methodology. We con-
tinue to work with the Nebraska State Museum on integrating their
species occurrence records into a format usable in GAP. As the
land cover products from the Rainwater Basin and Cooperative
Hydrology Study (COHYST) projects become available, we will
apply the animal models to these data, which are based on more
recent TM images (1997-98), to assess the effects of multi-date
thematic data on predicted species occurrences. Plans for next 12
months: Finalize bird and mammal species lists, ingest relevant
museum data, complete herpetile and mammal models, initiate bird
models, select and apply accuracy assessment methods to each ani-
mal model, Web-publish animal models as they are completed to
solicit comments and review.

Land stewardship mapping: A preliminary product is currently
undergoing review and refinement. We continue to work closely
with the Nebraska Games and Parks Commission in this aspect of
the project. Plans for next 12 months: finalize stewardship data-
base.

Analysis: Analysis is pending completion of animal models. Plans
for next 12 months: Initiate analyses as animal models are com-
pleted.

Reporting and data distribution: Metadata assembly, data lin-
eage, and methods documentation are ongoing.

Other accomplishments and innovations:

1. Ongoing assessment of elk habitat in northwest Nebraska in co-
operation with the NE Game and Parks Commission and the School
of Natural Resource Sciences.

2. Cooperating with Rainwater Basin Joint Venture (U.S. Bureau of
Reclamation – $180,000; NE Game and Parks Commission) in land
cover mapping using 1997-98 TM data.

3. Cooperating with the COHYST of the Central Platte, which is
sponsored by five natural resource districts, two public power dis-
tricts, NE Game and Parks Commission, NE Natural Resources
Commission, and the NE Department of Water Resources ($87,000).
The project will produce improved land cover mapping including
crop-level discrimination across a region covering one-third of
Nebraska using 1997-1998 TM data and digital orthophotos.

Nevada

First-generation GAP data will be provided through the National
GAP Web site by summer 2000; this preliminary project will not
have a standard report.

Version 2 update under way

Contact: Bruce Jones
U.S. EPA, Las Vegas
jones.bruce@epamail.epa.gov, (702) 798-2671

Land cover: Nevada is participating in the regional GAP update
for the Southwest. During the year 2000, organizational tasks in-
clude listing potential cooperators and their contributions, compil-
ing relevant literature, collecting masking data, interacting with re-
gional plant ecologist on alliance descriptions, and interacting with
the regional office (Utah State University Remote Sensing/GIS
Laboratories, Logan) on initial mapping procedure and concurrent
field season.

Animal modeling, land stewardship mapping, and analysis: Or-
ganizing during the year 2000.

Reporting and data distribution: Anticipated for spring 2004.

New Hampshire

(see Vermont and New Hampshire)

New Jersey

(see Maryland, Delaware, and New Jersey)

New Mexico

Version 1 complete (see http://rgis.unm.edu/price.htm). Version 2
update under way

New Mexico is included in a five-state (AZ, CO, NV, NM, UT)
regional effort to update and expand information from GAP projects
conducted for the individual states in the 1990s. Mapping of biotic
elements and land stewardship will follow established GAP meth-
ods but will be conducted cooperatively for the region, without re-
gard to state boundaries. The current New Mexico project is in the
start-up phase for this multiyear Southwestern Regional Gap Analy-
sis Project (SW-ReGAP). Project responsibilities in New Mexico
focus on reinitiating the cooperative network, assisting a regional
lab with imagery analysis, and acquiring selected vegetation and
animal occurrence data. Specifically we will 1) develop the part-
nership infrastructure for land cover mapping and other data set
preparation, 2) identify training site data for land cover mapping,
3) assist a regional lab with selecting ancillary data for the full
project, 4) prepare a rule base for making land cover mapping deci-
sions, 5) consolidate existing animal distribution models for
ecoregional task assignment, 6) evaluate recent changes to land stew-
ardship data, and 7) create and maintain regionwide Web pages,
FTP site, and list server to facilitate regional cooperation and data
exchange.

Results of this project will be used to develop New Mexico’s con-
tribution to the full multiyear SW-ReGAP. This step is critical to
ensure that appropriate research staff and facilities are brought to
bear on the ecoregional conservation assessment. Further, this
project will provide the starting point for the first-ever conserva-
tion evaluation of animals and plant communities throughout the
Southwest that crosses state jurisdictional boundaries.

To develop the infrastructure necessary to complete the long-term
project, we are incorporating personnel from several departments
within New Mexico State University (animal and range sciences,
biology, fishery and wildlife sciences, geography, entomology, plant
pathology, and weed science) as well as the agricultural experiment
station, the physical sciences laboratory, and the Army Research
Laboratory at White Sands Missile Range. These experts are pro-
viding input for this portion of the project by identifying training
site data, ancillary data, scene selection, and creation of the rule
base. The SW-ReGAP home page (http://leopold.nmsu.edu/
fwscoop/swregap) will provide information for both project per-
sonnel and cooperators and the general public.

New York

Near completion

Anticipated completion date: June 2000

Contact: Charles R. Smith
New York Cooperative Fish and Wildlife Research Unit
Cornell University, Ithaca
crs6@cornell.edu, (607) 255-3219

Land cover: Accuracy assessment, as described in last year’s sta-
tus report, continued in 1999. We evaluated map accuracy at three
different levels in the nested hierarchy of the National Vegetation
Classification System (NVCS). Using conventional accuracy as-
sessment methods, our land cover map accuracy at an NVCS level
approximately equivalent to Anderson Level I is 74.4%. At the
NVCS subclass level, our map accuracy is 56.5%, and at the level
of NVCS superalliance, our accuracy is 42.0%. Values for map
accuracies using the more computationally intensive “fuzzy” accu-
racy methods were generally a few percentage points better than
values derived by conventional accuracy assessment methods.

Animal modeling: We are continuing a successful and productive
cooperation with the ongoing NY Amphibian and Reptile Atlas,
sponsored by the New York State Department of Environmental
Conservation (NYSDEC), with field work ending in 1999 (http://
www.dec.state.ny.us/website/dfwmr/wildlife/herp/index.html). We
have received updated information on current distributions of am-
phibians and reptiles from the herpetological atlas project through
the 1998 field season. Association matrices, relating vertebrate
species occurrences to each of the 45 land cover types we have
identified were completed, reviewed by teams of experts, and as-
sessed for accuracy during the third quarter of 1999. Predicted
occurrences of terrestrial vertebrates were assessed for accuracy
using known occurrences of species from recent museum data (mam-
mals only), recent herpetological atlas field data, checklists of birds
from state parks, breeding bird occurrence data from the NY Breed-
ing Bird Atlas, and checklists of birds and other vertebrates from
federal refuges. The NY-GAP Project also is cooperating with
NYSDEC and the Federation of New York State Bird Clubs to as-
sure that new information collected as part of the Second NY Breed-
ing Bird Atlas Project, scheduled to begin in 2000, can be incorpo-
rated fully into the NY-GAP database for future gap analysis appli-
cations.

Land stewardship mapping: During 1999 we substantially refined,
expanded, and updated our existing land stewardship coverage to
reflect significant acquisitions of new public lands and increased
availability of accurate statewide land stewardship information in
digital form. Included in the land stewardship coverage at this time
are boundaries for all state wildlife management areas, state for-
ests, state parks, NewYork City reservoir watersheds with restricted
access, Adirondack Park Preserve, Catskill Park Preserve, Depart-
ment of Defense lands, national parks and historic sites, federal
wildlife refuges, and large preserves managed by The Nature Con-

servancy, represented by approximately 8,000 polygons. Assign-
ment of management status categories to these areas was completed
during the last half of 1999, and stewardship codes were assigned
in consultation with staff from NYSDEC and the New York State
Office of Parks, Recreation, and Historic Preservation.

Analysis: Accuracy assessments for maps of predicted vertebrate
distributions were completed during the last quarter of 1999. Over-
all map accuracy for predicted vertebrate distributions, at the
ecoregional scale, is 84.2%, with 86.2% for amphibians, 83.6% for
reptiles, 89.2% for birds, and 77.9% for mammals. Gap analysis
for the statewide database will be initiated and completed during
the last quarter of 1999 and the first quarter of 2000.

Reporting and data distribution: A draft final report will be sub-
mitted to the National Gap Analysis Program Office for review in
March 2000. After review, a final report will be completed and
submitted by late June 1999. Data distribution is expected to be
primarily on CD-ROM, with limited hard copy distribution, and
from Web pages. Data distribution by NYSDEC is expected to be
accomplished on a regional basis using a central server linked to
GIS workstations in each of the nine regional offices of NYSDEC.
The creation of the Cornell University Geospatial Information Re-
pository (CUGIR; http//cugir.mannlib.cornell.edu), a major National
Spatial Data Infrastructure node, also offers opportunities for addi-
tional modes of data display and distribution not available previ-
ously. Once produced by USGS, CD-ROM products will be of-
fered for sale through the NY Cooperative Fish and Wildlife Re-
search Unit.

Other accomplishments and innovations: See note on page 57
of this issue.

New York Aquatic Gap Analysis

Complete

Contacts: Marcia Meixler, Project Leader
NY Aquatic Gap Analysis, Cornell University
msm10@cornell.edu, (607) 255-2038

Mark Bain, PI, NY Aquatic Gap Analysis
Cornell University
mbb1@cornell.edu, (607) 255-2840

Habitat characterization for Aquatic GAP: Habitat was charac-
terized using the parameters stream size, habitat quality, water qual-
ity, gradient, and riparian forest cover. The first three parameters
were combined to form a habitat characterization from which fish
diversity was predicted. The latter three parameters were used for
macroinvertebrate diversity predictions. The first round of habitat
characterization involved static, manually intensive classifications
from topographic and Mylar land use overlay maps. In an effort to
deviate from such limiting classification, the NY Aquatic Gap Analy-

sis group developed computerized macros to automate classifica-
tion from digital elevation models, land use, and road and railroad
coverages. This provided equal or better accuracy, increased flex-
ibility, and enabled us to calibrate the model using previously col-
lected data. The calibrated habitat characterization incorporated
five additional GIS layers (surficial geology, bedrock geology, depth
to bedrock, point-source pollution, priority waters) and involved
optimization using discriminant analysis procedures.

Analysis: Field data were collected in the summer of 1998 on fish
species diversity, macroinvertebrate family diversity, stream width
and depth, substrate, general habitat assessment, water chemistry,
and gradient at 39 sites. This information was used to test the five
parameters in habitat characterization and overall diversity of fish
and macroinvertebrates.

Reporting and data distribution: The NY Aquatic Gap Analysis
group maintains a Web site for dissemination of up-to-date infor-
mation on model methodology and results. This site can be found
at http://www.dnr.cornell.edu/hydro2/aquagap.htm. In addition, a
comprehensive report was compiled in October 1999 for the U.S.
Geological Survey discussing the methods, results, products, analy-
ses, and conclusions to date.

North Carolina

Under way

Anticipated completion date: December 2000

Contact: Alexa McKerrow
North Carolina Cooperative Fish and Wildlife Research Unit
North Carolina State University, Raleigh
mckerrow@unity.ncsu.edu, (919) 513-2853

Land cover: In 1999 we expanded the vegetation mapping efforts
inland with two separate efforts: one in the southern mountains and
the other in the sandhill region. In addition, field data collection
and video interpretation were completed for the piedmont. We have
continued to utilize the decision-rule process for mapping, with the
addition of variables derived from the digital elevation models for
the mountains. By spring 2000 we will be completing the state-
wide land cover map by following the protocols developed in each
of the smaller-scale mapping efforts.

Animal modeling: Draft county and hexagon range maps are cur-
rently available for review by experts in the state. With the help of
the North Carolina Natural Heritage Program and the North Caro-
lina State Museum we will be finalizing those data layers by early
2000. In addition to soliciting one-on-one reviews and holding joint
meetings for review of the range and habitat association data lay-
ers, we are developing a Web site to support that process as well as
final data delivery needs. By February 2000 final ranges and habi-
tat associations will be available for incorporation into the final
analysis.

Land stewardship mapping: Several separate state efforts have
been undertaken to develop the data layers required to develop the
land stewardship layer. A reasonable base data layer with respect
to boundaries exists for state-owned lands and wildlife refuges. Fed-
eral lands will require a considerable effort to bring them up to
date. We have been in communication with the North Carolina
Natural Heritage Program, North Carolina Center for Geographic
Information and Analysis, and the Fish and Wildlife Ecological
Services Raleigh Field Office and are hoping to develop a coopera-
tive agreement to develop a set of data layers that will meet each of
the individual programs’ needs in addition to those of the North
Carolina GAP Project. We are hoping to have a fulltime staff mem-
ber assigned to this task by January 2000.

Analysis: Final analysis will be initiated in June 2000.

Reporting and data distribution: Reporting and data distribution
will begin in March 2000, starting with the basic data layer devel-
opment and concurrent with the final analysis.

North Dakota

Under way

Anticipated completion date: October 2003

Contact: Larry Strong
Northern Prairie Wildlife Research Center, Jamestown
larry_strong@usgs.gov, (701) 253-5524

Land cover: Three dates of Thematic Mapper (TM) imagery
(spring, summer, and fall) for each of 14 path-row orbit combina-
tions were acquired to provide complete coverage for North Da-
kota (ND). A seven-class land cover classification for ND was pro-
duced by building upon the recent efforts of Ducks Unlimited (DU)
and U.S. Fish and Wildlife Service (FWS) in the Prairie Pothole
Region of ND. Guided K-means clustering and maximum likeli-
hood classification of spring and fall TM images were performed
to produce a land cover classification compatible with the DU/FWS
land cover map for areas south and west of the Missouri River.
Vegetation surveys were conducted on thirty 28.6 x 28.6 km study
areas distributed among eight ecoregions in the summer of 1999.
Within ecoregions, study areas were chosen to maximize: 1) access
to public lands, 2) the amount of natural and seminatural vegeta-
tion, and 3) variability of surface geology and soil types. Global
Positioning Systems were used to delineate vegetation polygons on
public lands. A large number of vegetation polygons on private
land were delineated on 1:31,680 scale color prints of TM imagery
from roadside surveys conducted after ground surveys of public
lands in the study areas. Vegetation community element of occur-
rence records from ND Natural Heritage Program’s Biological Con-
servation Database were obtained and converted to a georeferenced
vector object. Access to the ND State Lands Range Inventory Da-
tabase for school lands was granted, and efforts are under way to
convert the data to a georeferenced vector object. Several digital

vegetation and land cover classifications for selected areas in ND
were obtained from ND-GAP cooperators. Training data sets will
be constructed for a more detailed vegetation and land cover classi-
fication using the three dates of TM imagery and elevation, cli-
mate, soils, and geology data in a classification tree analysis.

Animal modeling: Checklists of all vertebrate species in North
Dakota have been completed and reviewed by expert reviewers.
For herpetiles and birds, locational data have been assembled into
GIS coverages, and range maps have been completed and subjected
to expert review. Literature reviews have started, and wildlife-habitat
relationship (WHR) models have been completed and reviewed for
two species, respectively. For mammals, most locational data have
been assembled into GIS coverages, and large mammal range maps
have been completed and reviewed. Small mammal range maps
have been sent to expert reviewers. Literature reviews have started,
and WHR models have been completed and reviewed for three spe-
cies. Efforts in the next 12 months will include the creation of
mammal range maps, literature reviews for most of the vertebrate
species, generation of WHR models and requests for expert review,
and the development of environmental data sets for modeling spe-
cies distributions.

Land stewardship mapping: Land stewardship data were obtained
from the Army Corps of Engineers, Bureau of Land Management,
Bureau of Reclamation, and National Park Service. Land steward-
ship data have been requested from the FWS, the Forest Service,
and the ND Game and Fish Department. The FWS has an aggres-
sive effort to construct land stewardship data for North Dakota and
will make the data available for GAP. We obtained a CD-ROM of
the ND Department of Transportation base map data, which includes
coverages for federal and state public lands in ND. During 2000
we will continue to acquire existing digital data from appropriate
agencies.

Ohio

Under way

Anticipated completion date: May 2003

Contact: Donna N. Myers
U.S. Geological Survey, Columbus
dnmyers@usgs.gov, (614) 430-7768

Thomas Waite
The Ohio State University, Columbus
Waite.1@osu.edu, (614) 430-7768

Land cover: The Ohio GAP project has undergone a change in
personnel. Completion of the land cover map is being negotiated
with The Ohio State University’s Center for Mapping. The goal for
the 12 months from January 2000 will be to begin production of the
land cover map using 1999 MRLC data obtained from the OhioLink
program.

Animal modeling: A new PI for animal modeling, Dr. Thomas
Waite, of The Ohio State University, Department of Evolutionary
Ecology and Organismal Biology, has been selected. Graduate stu-
dent Troy Wilson has compiled a species list of terrestrial verte-
brate taxa in Ohio from existing information sources. The goal for
the next 12 months will be to develop a database of information on
species distributions, location information, references, and habitat
affinities. Data sources on species occurrence and distribution have
been identified for the Ohio Aquatic GAP project. Ohio Aquatic
GAP will use the valley segment classification system to identify
stream valley segment types in Ohio. The classification is planned
to be completed by January 2000.

Land stewardship mapping: The Ohio Department of Natural Re-
sources (DNR) is compiling a land ownership map for Ohio. OH-
GAP will coordinate efforts with the Ohio DNR to produce a land
stewardship map by December 2000.

Reporting and data distribution: A fact sheet will be completed
by March 2000.

Oklahoma

Near completion

Anticipated completion date: June 2000

Contact: William L. Fisher, Assistant Unit Leader
Oklahoma Cooperative Fish and Wildlife Research Unit,
Stillwater
wfisher@okstate.edu, (405) 744-6342

Land cover: The final land cover map is complete and under re-
view.

Animal modeling: Modeling of the distribution of the 427 terres-
trial vertebrate species is under way.

Land stewardship mapping: The final land stewardship map is
complete.

Analysis: Overlay analysis to identify potential gaps in biodiversity
conservation will begin following completion of animal distribu-
tion modeling.

Reporting and data distribution: Planning is under way.

Oregon

Near completion

Anticipated completion date: June 2000

Contact: Jimmy Kagan, Director/Ecologist and GAP PI
Oregon Natural Heritage Program
jkagan@tnc.org, (503) 731-3070 x332

Current (1993) land cover: An original land cover map (Version
1) was completed for Oregon in 1992. A second-generation land
cover map (Version 2) has been recently completed (August 1998,
using 1993 imagery) and is currently available. Metadata are avail-
able for this second-generation cover, but the accuracy assessment
is incomplete, and the classification is not tied to the NVCS. An
associated vegetation manual has been completed for the new map.
A description of the mapping process is included in chapter 2 of the
OR-GAP final report.

Historic (~ 1850) land cover: A statewide historic land coverage
was created by OR-GAP, modified from a coverage developed by
the Oregon Biodiversity Project. The coverage approximates a
1:100,000 scale, and metadata are complete. The historic land cover
categories are tied to the NVCS and have been cross-walked to the
wildlife habitats developed for the current land coverage (Version
2). Descriptions of the mapping process and classification are in-
cluded in chapter 3 of the OR-GAP final report.

Animal modeling: Animal modeling has been completed for Or-
egon three times, the first time using the Version 1 GAP vegetation
map. The result of this project is the book
Atlas of Oregon Wildlife
by Csuti et al. The initial models relied on hexagon distributions
prepared for all native wildlife species. Since these were completed
over eight years ago, we updated the hexagon distribution covers
for all wildlife species in March 1999. Version 2 models (ARC/
INFO programs - AMLs) were developed for all vertebrate species
using the updated hexagon distributions and the Version 2 land cover,
with an updated wildlife habitat relationship matrix to reflect the
differences between the first- and second-generation land cover-
ages. The updated modeled distributions received limited peer re-
view through local experts. As the Version 2 maps are fairly similar
to the thoroughly reviewed Version 1, OR-GAP feels the review
has been sufficient. OR-GAP also developed historic models using
similar techniques with modified hexagon distributions (based on
historic distribution, including extirpated species) and the historic
land coverage. The details of the current and historic modeling
process are included in chapter 4 of the final state report.

Land stewardship mapping: The land stewardship cover is com-
plete and was used to develop the gap analysis. The coverage and
metadata for 1999 (used in the analysis) are posted at the OR State
GIS Service Center. An updated version of the cover will be pro-
duced in 2000. A summary of the stewardship mapping and cat-
egorization effort was included as chapter 5 in the draft final Or-
egon report.

Analysis: Analysis using the Version 2 data is completed. A sum-
mary of the analysis is included in chapter 6 of the OR-GAP draft
state report.

Reporting and data distribution: The draft final state report is
being peer-reviewed by the national GAP staff and by Oregon peer
reviewers. A limited number of hard copies of the final report will
be produced. The final report and the basic coverages will be posted
at or linked to the Oregon GAP home page, which will be main-
tained at either the Oregon Natural Heritage Program or the Or-

egon Department of Fish and Wildlife office in Portland. We in-
tend to develop links to allow downloading these coverages from
many sources. The basic coverages that are currently posted in-
clude the stewardship cover, the first- and second-generation GAP
vegetation covers, and a presettlement vegetation coverage. All
final GAP products, databases, and coverages will also be included
on the CD-ROM, with an ArcView application we developed that
creates historic and existing species distributions.

During the next 10 months of the project, using funds from the U.S.
Fish and Wildlife Service and the state of Oregon, we will distrib-
ute GAP data to watershed councils, local governments, and other
decision makers in Oregon. We will also train others in the use of
the data and in assessment methods and tools. We intend to pro-
vide GAP data, Oregon Natural Heritage Program threatened and
endangered species data, and analysis to assist in watershed, basin,
or ecoregional planning efforts. We are working with the North-
west Office of the Defenders of Wildlife to distribute GAP data and
also to assure the updated GAP coverages are included on their
CD-ROM product,
Oregon’s Living Landscape, an Interactive In-
troduction to Oregon’s Biodiversity
. This CD-ROM was produced
as part of the Oregon Biodiversity Project, in which both the Or-
egon Natural Heritage Program and the Oregon Gap Analysis Pro-
gram were partners.

Other accomplishments and innovations: As part of our associa-
tion with the Oregon Biodiversity Project, an application was de-
veloped that allows both GAP and Natural Heritage data to be sum-
marized on a watershed basis. This application was developed by
the Defenders of Wildlife staff and is included on their CD-ROM.
We are working on developing a similar, Web-based application.
We are hoping to adapt the Colorado State University Internet site
for distributing OR-GAP and Oregon Natural Heritage Program
data.

Pennsylvania

Near completion

Anticipated completion date: June 2000

Contact: Wayne L. Myers
Penn State University, University Park
wlm@psu.edu, (814) 863-0002

Land cover: Land cover maps have been completed with general-
ized categories. Accuracy assessment is pending.

Animal modeling: Models are completed and have been reviewed.
Fish, bird, and mammal models were run for mapping; herp mod-
els are running.

Land stewardship mapping: Stewardship mapping has been com-
pleted.

Analysis: is in progress for reporting.

Reporting and data distribution: Expect draft ready for review
by national GAP by February 29, 2000. Web service will be pro-
vided by Pennsylvania Spatial Data Access (PASDA) at Penn State
University.

Rhode Island

(see Massachusetts, Connecticut, and Rhode Island)

South Carolina

Under way

Anticipated completion date: March 2001

Contact: Elise V. Schmidt
South Carolina Cooperative Research Unit
Clemson University, Clemson
schmidt@water.dnr.state.sc.us, (803) 734-9097

Land cover: We have completed the initial land cover map for the
state and are now in the process of incorporating ancillary data. We
are using data from the National Wetlands Inventory and NRCS
soil surveys to further refine our initial classification of 28 land
cover types. There are other digital databases available that will be
used to add additional land cover types. The land cover is expected
to be complete by August 2000.

Animal modeling: The database for animal distribution and habi-
tat affinities is complete and has been reviewed by state experts.
We are now compiling the expert reviewer comments to determine
where more research needs to be conducted. SC-GAP has data
from state and university museums as well, which are ready to be
incorporated. A user interface is being developed for limited ac-
cess to the database through the Internet, so that the experts can
review the final maps and database. We will complete sampling for
ant diversity in all physiographic regions and most land cover types
in the state within the year 2000. Ant diversity will be a part of our
animal modeling. The animal database is expected to be complete
by August 2000.

Land stewardship mapping: The land stewardship database is
complete except for a small portion of properties. We are working
with involved parties to ensure correct classification of GAP status
on public lands in South Carolina.

Analysis: The gap analysis is expected to begin in August 2000
and be completed by December 2000.

South Dakota

Under way

Anticipated completion date: May 2001

Contact: Jonathan A. Jenks, GAP PI
South Dakota State University, Brookings
jonathan_jenks@sdstate.edu, (605) 688-4783

Vickie J. Smith, GAP Coordinator
South Dakota State University, Brookings
vickie_smith@sdstate.edu, (605) 688-5124

Land cover: Land cover classification was completed for eastern
South Dakota in June 1999. Twelve categories were separated, in-
cluding two alliance classifications, which were delineated using
on-screen digitizing, and four wetland categories from the National
Wetland Inventory (NWI). An image of land cover for eastern South
Dakota can be viewed at http://wfs.sdstate.edu/sdgap/eastland.htm.
Currently, two of nine western South Dakota scenes have been clas-
sified using alliance training data from the USGS/TNC Vegetation
Mapping Program for Wind Cave National Park. Within the two
scenes, 19 categories are present, including 13 association-level
categories. Assessment is under way to determine the accuracy of
this method of classification. Completion of the South Dakota land
cover map is expected by May 2000. Accuracy assessment using a
stratified random sample for the state will begin during the summer
of 2000.

Animal modeling: Distribution maps are completed, reviewed, and
revised for 88 mammal species. They can be viewed at http://
wfs.sdstate.edu/sdgap/mammals.html. Small mammal/vegetation
associations are being determined at Wind Cave National Park.
These data will be used in accuracy assessments of small-mammal
distributions. Maps are nearly completed for 43 herp species, and
the avian species list is under review. South Dakota is working
cooperatively with four surrounding states (North Dakota, Kansas,
Iowa, and Nebraska) to create regional models for all vertebrate
species. Currently, 16 species have been modeled for the region.
Literature review has been conducted for nearly 80% of the species
to be modeled by South Dakota.

Land stewardship mapping: We have added state parks and rec-
reation areas, The Nature Conservancy holdings, and state wildlife
refuges to our stewardship map in the past year. We are attempting
to acquire information for a three-mile-wide section along the bor-
ders of Montana and Wyoming. Permission has been granted to
include only two of the seven Indian reservation boundaries in our
stewardship map. Our current stewardship map can be viewed at
http://wfs.sdstate.edu/sdgap/stewardmap.htm.

Aquatic GAP: Plans are under way to hire a graduate student in
early 2000. River reach files have been acquired, and watershed
boundaries have been formulated for three major river systems in
eastern South Dakota, through a project researching presence of

Topeka shiners in eastern South Dakota streams (see article on page
35 of this issue). Habitat information has been collected on his-
toric locations of the Topeka shiner. These data will be used as a
pilot project for Aquatic GAP to formulate predictions of the cur-
rent Topeka shiner distribution in these stream systems.

Tennessee

Completion date unknown, please contact coordinator below.

Contact: Jeanette Jones, Project Coordinator
Tennessee Wildlife Resources Agency, Nashville
jjones2@mail.state.tn.us, (615) 781-6534

Sue Marden, Vertebrate Ecologist
Tennessee Wildlife Resources Agency
smarden@mail.state.tn.us, (615) 781-6637

Land cover: The detailed vegetation map is completed. The veg-
etation map was produced using classification techniques applied
to Landsat TM imagery and aerial videography. Accuracy assess-
ment was performed using a subset of points set aside from the
aerial videography interpretation. Final figures for accuracy as-
sessment are being tabulated. Completion of metadata and prepa-
ration of data for final delivery to National GAP remain to be done.

Animal modeling: Predicted species distributions and species rich-
ness data have been produced for Tennessee’s 364 terrestrial verte-
brate species. The species distribution data are in the process of
being written onto CDs to send to National GAP.

Land stewardship mapping: The land stewardship layer is com-
pleted. Lands mapped are current through December 1997. The
public lands coverage has been updated, while land management
status needs to be assigned before any further analysis.

Analysis: Gap analysis has been completed.

Reporting and data distribution: The final report is near comple-
tion. Data are being written to CD-ROMs for delivery to National
GAP. Plans are to present TN-GAP data as part of the Tennessee
Wildlife Resources Agency (TWRA) Web page.

Other accomplishments and innovations: The Tennessee
Biodiversity Program (established by the Tennessee Conservation
League [TCL]) and TWRA’s GIS division are continuing to work
together to provide planners and community leaders, landowners,
natural resource professionals, and educators with information on
Tennessee’s natural resources. TWRA provides TN-GAP data and
related GIS data layers as ArcView files to county planners and
community leaders. Managing Natural Resources -
A Planning
Guide for the Elk River Watershed of South Central Tennessee and
Northern Alabama
was published in 1999 by TWRA, TCL, Ten-
nessee Valley Authority, and National GAP as a planning guide for
developing and carrying out natural resource conservation and man-
agement programs.

Texas

Under way

Anticipated completion date: December 2000

Contact: Nick C. Parker
Texas Cooperative Fish and Wildlife Research Unit
Texas Tech University, Lubbock
nparker@ttu.edu, (806) 742-2851

Land cover: The land cover map for Texas has been completed; all
scenes have been stitched together. Data from field work, com-
pleted in all 254 counties in Texas, were used to classify the scenes.
Data from videography were used for accuracy assessment. Over-
all accuracy of the classified product was above 80%. Approxi-
mately 6,000 photographs with UTM coordinates were taken in the
field. These photographs are in a database and are being prepared
for distribution through the Web.

Vertebrate modeling: We have identified 637 terrestrial vertebrate
species as being native to and breeding in Texas. GIS layers repre-
senting the range extents for each of these species have been devel-
oped from existing range maps. In addition, a database consisting
of 34,441 location records for mammals and birds has been devel-
oped. Habitat profiles have also been prepared for all 637 species
being modeled, and statewide GIS layers have been created from
the following profile variables: precipitation, temperature, soils, hy-
drology, ecoregions, and elevation. Preliminary distribution mod-
els for herpetofauna and mammals have been completed. Fifty per-
cent of the preliminary distribution models for birds have been com-
pleted.

Analysis: Under way.

Reporting and data distribution: Draft maps were provided to 89
landowners in West Texas to solicit their evaluations for use in ac-
curacy assessment. Draft maps have also been prepared for Texas
State Parks, the National Park Service, the U.S. Border Patrol,
USDA, and cooperators in Texas and Mexico.

Other accomplishments: Data prepared for West Texas are being
used to prepare selected species-specific maps (e.g., prairie dog
towns and Scaled Quail distribution).

Utah

Update under way

Anticipated completion date: June 2004

Contact: Doug R. Ramsey
Remote Sensing/GIS Laboratories
Utah State University, Logan
doug@nr.usu.edu, (435) 797-4484

The first-generation Utah Gap Analysis project has been completed.
An update is under way. We are part of Southwest Regional GAP,
which includes Nevada, Colorado, New Mexico, and Arizona. To
date we have been updating the ancillary data layers and research-
ing the feasibility of adding a soil component to our ancillary data
set. We have also been reviewing the vegetation reference data
collected in our first gap analysis for use in the update. As part of a
regional effort we have been helping each state define mapping
zones. The purpose of the mapping zones is to divide the landscape
into similar units of landform features, soils, and biotic elements.

Older Utah GAP products are still readily available for both UNIX
and PC computer systems. We also have a CD-ROM on Intermoun-
tain Region Land Cover Characterization that incorporates GAP
data from Nevada, Southern Idaho, Western Wyoming, and Utah.

Land cover: Modeling land cover characteristics will not begin
until all vegetation reference data are collected for a mapping zone.
The collection of data is scheduled to begin in spring 2000.

Animal modeling: Dr. Thomas C. Edwards of the Utah Coopera-
tive Fish and Wildlife Research Unit will undertake the wildlife
habitat modeling for Utah. Dr. Edwards can be reached by phone
at (435) 707-2529, by fax at (435) 797-4025, and by e-mail at
tce@nr.usu.edu.

Land stewardship mapping: Individual management units for the
state of Utah will be updated for public and private lands.

Analysis: All data layers are scheduled for analysis completion by
2004.

Reporting and data distribution: All products derived from Utah’s
Gap Analysis as well as Southwest GAP are scheduled for comple-
tion by 2004.

Vermont and New Hampshire

Near completion

Anticipated completion date: June 2000

Contact: David Capen
School of Natural Resources, University of Vermont, Burlington
dcapen@nature.snr.uvm.edu, (802)656-2684

Land cover: Land cover mapping for both Vermont and New Hamp-
shire is complete. Edge-matching between New Hampshire and
Maine gave acceptable results.

Animal modeling: Refining models and processing additional an-
cillary data for about 25% of vertebrate species; modeling is com-
plete for other species.

Land stewardship mapping: Complete for both states, including
private conservation parcels and hundreds of ownerships protected
by easements.

Analysis: In progress; macros for most steps have been developed
and tested. Final analysis awaits refinement of selected habitat
models.

Reporting and data distribution: Reports will be submitted for
review in early 2000; data distribution should follow later in the
year.

Other accomplishments and innovations: Both Vermont and New
Hampshire have undertaken statewide reserve selection projects that
have complemented the Gap Analysis effort. Each of these projects
has incorporated physical factors of the landscape into their analy-
ses. We believe that this approach, coupled with the land cover
maps of GAP, offers a more sensitive and more comprehensive
means of identifying hotspots of diversity than does the modeling
of vertebrate distributions.

Virginia

Near completion

Anticipated completion date: June 2000

Contact: Scott D. Klopfer
GIS and Remote Sensing Project Coordinator
Fish and Wildlife Information Exchange, Blacksburg
fwiexchg@vt.edu, (540) 231-7348

Land cover: The final VA-GAP land cover map was completed in
spring 1999. This map is currently available through anonymous
FTP from the Fish and Wildlife Information Exchange (FWIE)
server at ftp.fwie.fw.vt.edu. The final map contains 26 land cover
classes and two forest complex classes. For more information on
VA-GAP’s land cover mapping efforts, please contact Scott Klopfer
(sklopfer@vt.edu).

Animal modeling: Vertebrate modeling is near completion.

Land stewardship mapping: The stewardship map for VA-GAP
is complete and available for download at ftp.fwie.fw.vt.edu. This
coverage contains all federal, state, and some privately owned lands
in Virginia. For more information regarding these efforts, please
contact Scott Klopfer (sklopfer@vt.edu).

Analysis: The quantitative accuracy assessment of the vegetation
map is completed. The accuracy of the final land cover map was
found to be between 67% (conservative estimate) and 87%. Analy-
sis of species distributions and protection gaps is near completion.

Reporting and data distribution: Many of the basic data sets used
by VA-GAP are available to the public via FTP. Our Web-based
data distribution node is expected to be complete by spring 2000.
The final VA-GAP report is in progress, and plans for a statewide
VA-GAP data use workshop are under way.

Other accomplishments and innovations: VA-GAP continues to
serve as a means of communication for federal and state agencies

in Virginia. The results of our accuracy assessment indicate that
our abiotic factor (a.k.a. “enduring feature”) modeling/remote sens-
ing hybrid classification method was able to provide us with accu-
rate land cover classifications.

Washington

Complete (see http://www.wa.gov/wdfw/wlm/gap/dataprod.htm)

West Virginia

Near completion

Anticipated completion date: September 2000

Contact: Charles Yuill
West Virginia University
Natural Resource Analysis Center, Morgantown
cyuill@wvu.edu, (304) 293-4832 x4492

Land cover: Complete; additional verification continuing.

Animal modeling: Near completion. Models and range maps are
complete. Modeling under way using revised final land cover.
Herpetiles are complete. Range maps are on project map server.

Land stewardship mapping: Complete; updated to 9/1/99.

Analysis: Near completion.

Reporting and data distribution: Near completion. Final report
for project is under way. A report for wide distribution has been
initiated and will be published from state funding sources. All map-
ping will be placed on map server, and CDs will be produced. Data
distribution is under way for available products; EPA, USGS, and
TNC have obtained products to date.

Other accomplishments and innovations: A grant was obtained
to fund the West Virginia Land Status and Trends Project, a five-
year project that will update WV-GAP and conduct detailed assess-
ments of forests, agricultural/open lands, landscapes of special con-
cern, and land use conversions in West Virginia. The project will
also refine and apply data using landscape ecology metrics within
the Watershed Characterization and Modeling System (WCMS) in
a landscape atlas project for The Canaan Valley Institute (a multistate
NGO focusing on watersheds).

Wisconsin

Under way

Anticipated completion date: September 2002

Contact: Tim Weiss
Wisconsin Department of Natural Resources
Bureau of Wildlife Management, Madison
weissn@dnr.state.wi.us, (608) 267-9428

Land cover: Classification of the state followed the Upper Mid-
west Gap Image Processing Protocol and is now completed. Cross-
walking the remote sensing land cover classes to the National Veg-
etation Classification System will begin in FY 2000.

Animal modeling: Vertebrate modeling will be undertaken by the
Upper Midwest Environmental Sciences Center (UMESC) in co-
operation with the state DNR beginning in FY 2000.

Land stewardship mapping: Public land ownership/stewardship
mapping is being undertaken by the Wisconsin DNR and is par-
tially completed. Federal lands and state lands are mapped to date;
county lands remain to be mapped.

Analysis: Gap analysis will be undertaken by UMESC and is sched-
uled to begin in FY 2001.

Reporting and data distribution: Land cover data will begin to
be served by UMESC in the second and third quarter of FY 2000.

Wyoming

Complete (see http://sdvc.uwyo.edu/wbn/gap.html)