Avian Occurrence in Woodlands and Grasslands on Public Areas Throughout Eastern South Dakota

The purpose of this study was to inventory nongame bird species on public lands throughout eastern South Dakota in woodland and grassland habitats of different patch sizes and landscape matrices in various physiographic regions. Nongame birds were surveyed in 524 (99 native and 425 planted) woodland patches on 307 public areas in eastern South Dakota using belt transects. Woodlands were surveyed once during either the 1997 or 1998 breeding season. Woodlands were surveyed completely with the exception of extremely large (\u3e 7 5 ha) tracts ( \u3c 1 % of the sample). A total of 85 different avian species were detected in 524 woodland patches on public areas in eastern South Dakota Native woodlands attracted significantly more woodland obligate species than planted woodlands across all size categories. Woodland edge species were significantly higher in planted woodlands. The largest planted woodlands in this study attracted an average of only 2 ( 14%) of the 13 woodland obligate avian species that commonly occur in eastern South Dakota, while the largest native woodlands attracted over 5 (40%). Differences in woodland obligate species richness between native and planted woodlands were due in part to structural differences between the patches. Native woodland patches in this study had taller trees, more snags, fewer and shorter shrubs and were larger than planted woodlands. Individual planted woodlands were generally homogeneous, linear stands of trees and shrubs and exhibited little variation within the patch. Thus, native woodlands provided a greater diversity of nesting and foraging habitat required by woodland obligate bird species. The influence of local and landscape factors affecting avian use of grasslands (n=380) were studied on public areas in eastern South Dakota to determine the effect of habitat fragmentation on grassland bird communities. I quantified land cover using a GIS at 3 spatial scales (buffers at 400-m, 800-m and 1600-m radii from the center of the transect within the grassland patch) to determine whether proportions of land cover types were related to the occurrence of grassland birds. Landsat Thematic Mapper imagery from 8 scenes covering eastern South Dakota was used to classify uplands into tilled and untilled vegetation classes. Ground-trothing was used to further differentiate CRP, idled grassland, pastureland, hayland and trees. Data on avian species occurrence were collected using fixed-width belt transects (I per patch) during the breeding seasons of 1997 and 1998. Grassland obligate bird species occurrence in this study was influenced by region, the surrounding landscape and within patch habitat variables. Differences in bird occurrences between grassland patches in the tall and mixed grass prairie regions of eastern South Dakota for 7 of 9 grassland obligate species may be accounted for by regional distribution and abundance, differences in landscape structure, e.g., area and attributes of the grasslands, or vegetation structure. Three of 9 grassland obligate species, the Upland Sandpiper and Dickcissel and Grasshopper Sparrow in the mixed grass prairie region. were positively associated with area of the grassland patch. The Sedge Wren. Clay-colored Sparrow and Savannah Sparrow were positively associated with the proportion of grassland habitat within either the 800-m or 1600-m buffers. Abundance of grassland habitat in the landscape explained more variation in occurrence for these 3 species than did area indicating that the landscape in which patches are imbedded is more important to these species than size of the grassland patch. Two grassland obligate species, the Savannah and Grasshopper Sparrow, exhibited negative relationships with measures of wooded habitat at the patch and/or landscape level. To manage for the entire community of grassland bird species, grasslands need to be large enough (\u3e 125 ha) or be part of landscapes with enough grassland habitat (2:40%) to support the most area sensitive species. Grasslands should have little (\u3c10%) or no edge occupied by wooded habitat and woodland habitat within a 400-m buffer should be \u3c1% when the primary management objective is to enhance grassland obligate avian species