Home - Table of Contents

State Reports - Maine

Maine Gap Analysis (ME-GAP) has entered its final year. Land cover mapping is progressing well, much of the vertebrate information needed is in hand and compiled, and land ownership and management status databases are complete. We have begun preparation of the final report, have made ME-GAP products available to cooperators and on the Web (http://wlm13.umenfa.maine.edu/progs/unit/gap), and have presented preliminary results to a variety of state and private conservation organizations.

Land Cover Map

Landsat Thematic Mapper (TM) imagery for the state is being used, along with ancillary GIS data, to delineate approximately 45 habitat types. A classification scheme was developed that represents a compromise between habitats required as input into wildlife-habitat models and the classes we felt we could discern from the TM imagery and available ancillary GIS layers.

Supervised, unsupervised, and hybrid approaches to image classification are being used. Training samples for the supervised portions of our image classification are derived from statewide aerial videography flown in June and October 1994. Approximately 11 million frames of wide angle (210 ha coverage) or zoom (0.09 ha coverage) are available from the videography for use in training TM data and testing of the resulting map. Spectral confusion occurs among some habitat types, requiring ancillary GIS data to differentiate these classes. Ancillary data sources include aerial videography, wetland polygons from the U.S. Geological Survey’s (USGS) Land Use/Land Cover Digital Analysis (LUDA) database, transportation network derived from the USGS Digital Line Graph database, and field-checked locations for blueberry fields and hay fields in eastern and central Maine.

Once the statewide classification is complete and all TM scenes are mosaicked together (anticipated in August 1997), the map will be tested using polygons defined from the aerial videography. Eighty percent accuracy is desired for all major habitat types. To date, three TM scenes in northern Maine have been classified with all three techniques, and classes are being labeled by habitat type.

Vertebrate Data

We had previously compiled range and ecological information for each of the terrestrial vertebrate species that breeds in Maine into synopses. We continued to receive feedback on the synopses, including a review of the information with our neighboring states to improve edge-matching. The information compiled has been useful to many public and private land managers in the state. We have used the information and other sources in a publication on Maine herpetofauna, and provided data to an author summarizing ranges for all vertebrates in New England. Species ranges, conservation ranks, and general habitats used are provided on our Web site.

Land Stewardship

ME-GAP personnel cooperated with the Maine State Planning Office to complete "A Conservation and Public Lands Database" (CAPLD) for the state. The 1:100,000 scale CAPLD includes many revisions and updates to previously released 1:250,000 scale land ownership maps and was constructed to meet GAP specifications. The final coverage contains about 3,130 polygons for 120 owners and includes state and federal lands, tribal lands, private nonprofit conservation lands (e.g., those owned by The Nature Conservancy), and selected public areas not managed for biodiversity (e.g., playgrounds, cemeteries, campuses). Each ownership block is coded as to consideration given to biological diversity in management plans. CAPLD has been provided to cooperators in digital and paper form and has been passed on to our state GIS office for distribution to the public. More details of how CAPLD was developed are presented in our home page cited above.

Other Progress

Randy Boone completed a thesis that forms a biogeographical foundation for ME-GAP, described elsewhere in this bulletin. Randy and Jeff Hepinstall, a remote sensing specialist, have joined ME-GAP as Research Associates, and two graduate students will be joining the team. One of the new students will assist in completing ME-GAP and test the predicted vertebrate distributions; the second student will assess the representativeness of Maine’s conservation lands. As products are generated from ME-GAP, conveying results has become paramount. We constructed a Web site which contains extensive information on ME-GAP. Finally, we have formally presented preliminary results to professional organizations, personnel from state and private nonprofit organizations, and, most exciting, to a Land Acquisition Priorities Committee convened by the Governor of Maine.

Project Information

Home - Table of Contents