Date of this Version
Stahlecker, Dale W. "Using National Wetlands Inventory Maps to Quantify Whooping Crane Stopover Habitat in Oklahoma.", In: Stahlecker D. W., ed. 1992. Proceedings of the Sixth North American Crane Workshop, Oct. 3-5, 1991, Regina, Sask. (Grand Island, NE.: North American Crane Working Group, 1992), 62-68.
Three stratified random samples of the 416 National Wetland Inventory (NWI) maps within the western Oklahoma portion of the Wood Buffalo-Aransas whooping crane (Grus americana) migration corridor were used to evaluate the availability of wetland roost sites. Wetlands were eliminated as potential roosts if visibility was obscured by vegetation or slope, or if certain human activities occurred within 100-800 m. Thirty percent of all wetlands >0.04 ha passed map review, but only 7% passed when ground truthed. NWI map review was a poor predictor of suitability (33 % correct) but a good predictor of unsuitability (97 % correct). Most (>75%) wetlands in western Oklahoma are man-made impoundments dug to water livestock. They are generally small « 1 ha), dammed, in steep drainages, and tree-lined. Most stock ponds do not meet the horizontal visibility requirements of roosting whooping cranes. Map review of wetlands > I ha (no randomized field review) suggests that 2-4 suitable crane roosts are available per 100 km2 in the migration corridor in Oklahoma. Three major rivers and approximately 20 large reservoirs (> 16 ha) provide the best roost sites. Without adequate ground truthing, review of NWI maps, even when supplemented by Soil Conservation Service photo maps, overestimate suitable roost sites, especially in areas where rainfall is sufficient to support woody vegetation > 1 m in height.