Northern Prairie Wildlife Research Center
Although the species richness of grasslands is relatively low, the ecosystem is essential for the survival of several bird species, such as burrowing owl (Athene cunicularia), mountain plover (Charadrius montanus), Sprague's pipit (Anthus spragueii), Henslow's sparrow (Ammodramus henslowii), and Baird's sparrow (Ammodramus bairdii).
Loss, degradation, and fragmentation of grassland habitats have contributed to the apparent decline of several grassland bird species (Herkert et al. 1993). Prescribed burning, grazing, and haying are used to maintain existing prairies or restore prairie habitats. Different grassland bird species react differently to these treatments. The habitat requirements and responses to management have been the subjects of numerous studies (although many of them suffer from lack of replication, controls, and random assignment of treatments). Syntheses of the literature on many grassland bird species are available on the Web (see USGS-BRD 1999). This article emphasizes landscape issues, especially those related to habitat fragmentation.
Habitat fragmentation involves the separation of large, contiguous areas of habitat into smaller patches isolated from one another. Three types of effects of fragmentation can be distinguished: patch size effects, edge effects, and isolation effects. Patch size effects are those that result from differential use or reproductive success associated with habitat patches of different sizes. Some of these effects may be induced by edge effectsphenomena such as avoidance, predation, competition, or brood parasitism that differ near a habitat edge compared with the interior of a habitat patch. And isolation from similar habitat can influence use of a particular habitat patch.
Each of these factors can affect (1) the occurrence or density of birds using a habitat patch; (2) reproductive success, either through predation rates on eggs or young or through brood parasitism rates; or (3) competition with other species. Effects due to competition are not well-known, but studies on the other features are summarized next.
Patch size effects: occurrence and density. Species that have lower densities or are absent from small habitat patches are referred to as "area-sensitive" (Robbins 1979). Species with large home ranges, such as greater prairie-chicken (Tympanuchus cupido), upland sandpiper (Bartramia longicauda), and northern harrier (Circus cyaneus), are typical area-sensitive species that are rarely present in small habitat patches. But even species with territories small enough to easily fit in small habitat patches can be area-sensitive, for reasons which are as yet poorly understood. Several grassland passerines have been identified as area-sensitive in different geographic areas: Baird's sparrow and grasshopper sparrow (Ammodramus savannarum) in Prairie Canada (McMaster and Davis 1998); Baird's sparrow, among others, in the northern Great Plains of the USA (Johnson et al. in preparation); grasshopper sparrow and western meadowlark (Sturnella neglecta) in Minnesota (Johnson and Temple 1986); Henslow's sparrow in Missouri (Winter 1998); grasshopper sparrow, Henslow's sparrow, bobolink (Dolichonyx oryzivorus), savannah sparrow (Passerculus sandwichensis), and eastern meadowlark (Sturnella magna) in Illinois (Herkert 1994); bobolink, savannah sparrow, Henslow's sparrow, and grasshopper sparrow in New York (Bollinger 1995); and vesper sparrow (Pooecetes gramineus), savannah sparrow, grasshopper sparrow, bobolink, and eastern meadowlark in Maine (Vickery et al. 1994). The extent of area sensitivity, however, varies geographically, probably because it depends on the structure of the landscape surrounding a prairie patch.
Consistent with their area sensitivity, few grassland-dependent birds were observed by Powell (1998) in surveys of three tallgrass prairie national parks, all of which were small. In contrast, she found that the 4,449-ha Tallgrass Prairie National Preserve supported large numbers of area-sensitive species such as grasshopper sparrow, greater prairie-chicken, eastern meadowlark, and upland sandpiper (A.N. Powell, personal communication).
Patch size effects: nest success. Burger et al. (1994) indicated that artificial nests survived at a higher rate in larger than in smaller prairie patches in Missouri, but artificial nests may not accurately reflect real nests. Winter (1998) recently showed that dickcissels (Spiza americana) were more successful in larger than in smaller prairies in Missouri.
Edge effects: occurrence and density. Wiens (1969) suggested that savannah sparrows, grasshopper sparrows, and vesper sparrows tended to avoid forest edges in Wisconsin. Delisle and Savidge (1996) indicated that few grasshopper sparrow nests were located within 60 m of a habitat edge in Nebraska. Helzer (1996) found that, in wet meadows along the Platte River in Nebraska, grasshopper sparrows avoided woody and cornfield edges, and bobolinks avoided woody edges.
Edge effects: predation. Nests of bobolinks and western meadowlarks in Minnesota that were >45 m from a habitat edge were more successful than those that were closer (Johnson and Temple 1990). Burger et al. (1994) reported that artificial nests <50 m from woody edge in Missouri were depredated at greater rate than those farther away. Delisle and Savidge (1996) found no edge effect for grasshopper sparrows in Nebraska. Winter, Johnson, and Faaborg (in review) reported that, in Missouri, predation rates on Henslow sparrow and dickcissel nests were higher near shrubby edges, and that there was more activity by mid-sized carnivores near edges.
Edge effects: brood parasitism. Brown-headed cowbirds (Molothrus ater) lay their eggs in the nests of other species, thereby reducing the productivity of the hosts. Johnson and Temple (1990) detected higher parasitism rates of nests within 45 m of forest edge for the clay-colored sparrow (Spizella pallida) and western meadowlark in Minnesota. In Missouri prairies, Winter, Johnson, and Faaborg (in review) found higher parasitism rates for nests within 50 m of shrubby edge.
Isolation effects: occurrence and density. Winter (1998) reported that for southwestern Missouri, Henslow's sparrow and dickcissel densities were greater in patches with more grassland within 5 km, and were lower in patches with more forest within 5 km. J. R. Herkert (personal communication) indicated that grasshopper sparrows and bobolinks in Illinois were encountered more frequently in large patches with greater amounts of grassland within 5 km.
Isolation effects: nest success. Winter (1998) found that dickcissel nest success was higher in Missouri prairies with greater areas of grassland within 5 km.
Interactions and regional variation. Some features mentioned above may interact, or operate differently in different areas. For example, Winter (1998) detected an interaction between patch size and landscape in Missouri, where Henslow's sparrow densities were independent of patch size in landscapes with a lot of grassland, whereas in forested landscapes densities were lower in small patches than in large patches. She also noted that Henslow's sparrows occurred in 30-ha patches in Missouri, whereas Herkert et al. (1993) indicated that they required patches exceeding 100 ha in Illinois.