While never a dominant feature in the landscape, natural grassland systems once formed an ecologically cohesive and important “archipelago” or “shifting mosaic” of patches of open grasses and broad-leaved herbaceous flowering plants called forbs, usually on calcareous soils in belted formations, across much of the East and West Gulf Coastal Plains and, more rarely, in parts of the Ozark Highlands and Gulf Coast. About 99% of these landscapes have been converted to other land uses or have transitioned into forested conditions due to fire suppression and ecological succession. As these landscapes disappear, so to do the species that rely on their characteristically open structure for habitat. Perhaps the most imperiled of the terrestrial ecological system types, Grassland-Prairie-Savanna was identified as one of the nine “Broadly Defined Habitats” developed by NatureServe and the U. S. Fish and Wildlife Service as a habitat framework for conservation as described in the Draft Integrated Science Agenda of the GCPO LCC.
Prairie patch polygons in the Tombigbee National Forest, Chickasaw County, digitized by the Mississippi Natural Heritage Program, now part of our Draft Dataset of Known Prairie Patches in the GCPO. World Imagery Source: Esri, DigitalGlobe, GeoEye, i-cubed, Earthstar Geographics, CNES/Airbus DS, USDA, USGS, AEX, Getmapping, Aerogrid, IGN, IGP, swisstopo, and the GIS User Community.
In order to support a Draft Rapid Ecological Assessment of Grassland-Prairie-Savanna, GCPO staff created a spatial data layer of all known examples of natural prairie within our geography, using spatial datasets derived from a variety of sources. The spatial data layer shows observed and noted prairies in Alabama, Arkansas, Mississippi, and Missouri. Currently we have no records of prairie patches for parts of the GCPO outside of these states. The dataset is available for viewing or exporting at our Conservation Planning Atlas (CPA).
When we began the process of writing the ecological assessment of the Grassland-Prairie-Savanna priority system last summer, we realized early on that spatial data products describing land cover, land use, vegetation class, or ecological systems, such as those produced by the Gap Analysis Project, LANDFIRE, NatureServe, or the National Land Cover Database (NLCD), could make a valuable contribution but could not be relied upon exclusively to answer our primary questions of where, how much, and in what condition. These spatial data products do a very good job with broad and differentiated classes such as developed land, open water, forest, and herbaceous cover. Within the forest class, the distinction between evergreen, hardwood and mixed forest classes is widely acknowledged to be fairly accurate at large landscape scales. Herbaceous classes are much more difficult to parse out using remote sensing and the various classification and feature extraction algorithms, and the fact that prairie patches are relatively small and scattered (compared to forests) further complicates things. Native prairies and other grasslands must be burned, mowed, or otherwise disturbed to suppress woody encroachment and to enhance desired habitat conditions for wildlife. Consequently these natural habitats experience a phenological cycle – a regular succession of structural stages – similar but not identical to those exhibited by other herbaceous land cover classes such as pasture/hay and cultivated crops. Consequently, a recently-burned prairie may be difficult to distinguish from a recently hayed pasture or a recently harvested cropland in an aerial image. In order to offset the uncertainty in the available geospatial data, we enhanced our analysis of the amount and condition of prairie-grassland-savanna by incorporating a spatial data layer of all observed natural prairie patches in our region. We appealed to partners in each of the eleven states in our geography, asking them to share information about where prairie conservation is taking place in their areas.
We obtained geospatial data layers describing prairies in Alabama, Mississippi, Arkansas, and Missouri. We also obtained data from Illinois and Tennessee but were unable to use it for a variety of reasons. Since the Alabama layer was the largest file (14,595 polygons totaling 15,533 acres), we appended the Mississippi, Arkansas, and Missouri spatial datasets to it, producing a dataset of 14,713 polygons totaling 30,475 acres. We eliminated attributes unique to the source layers, and added attribute fields for state, county, and acres. You can read more about how the layer was produced in the dataset description page on our Conservation Planning Atlas, or in the Draft Ecological Assessment of Grassland-Prairie-Savanna on the GCPO LCC website.
The Draft Dataset of Known Prairie Patches in the GCPO describes the location, size, and configuration of prairies that we know about. Where are the prairies that we don’t know about? There are certainly more to be found. We encourage our partners to provide feedback in the comments section of the map on the CPA. The comments tab allows the user to draw polygons, lines and points in addition to written comments. To learn more about how to leave comments, watch this video tutorial.
We look forward to continuing engagement with partners as we work to understand and conserve this important ecological system.