Northern Prairie Wildlife Research Center
Strip cover provides habitat for most of the NTMBs of agricultural landscapes because it is used by birds requiring either herbaceous or woody vegetation (Table 1). These areas, which include grassed waterways, terraces, fencerows, roadsides, and windbreaks/shelterbelts, usually provide habitat that is more long-term than that provided by areas enrolled in cropland-retirement programs. Food (e.g., arthropods, weed seeds) is often more abundant in strip cover than in crop fields, and complex vegetation structure provides nesting sites, song perches, and cover (Rodenhouse et al. 1993). Many of these habitats are associated with conservation practices used to control soil erosion, such as crop rotation, terraces, contour planting, strip intercropping, grassed waterways, and windbreaks. Some kinds of strip cover are subject to disturbances such as mowing. During the past few decades, the average size of cropped fields has increased, reducing the amount of strip-cover habitat (Rodenhouse et al. 1995).
Grassed waterways are heavily used by birds. These waterways are natural drainage systems or channels constructed to transport water off crop fields at a non-erosive velocity. Various grass species are planted in these waterways to slow the flow of water. In Iowa, 48 bird species used grassed waterways, compared with only 14 in surrounding rowcrop fields (Bryan and Best 1991). The most abundant bird species using waterways were red-winged blackbirds, dickcissels, barn swallows, grasshopper sparrows, brown-headed cowbirds, song sparrows, and western meadowlarks. Total bird abundance in waterways averaged 2,198 birds/census count/l00 ha, compared with 682 for crop fields. Eleven bird species nested in the grassed waterways; red-winged blackbird and dickcissel nests were most common (Bryan and Best 1994). Considering nests of all species, more than twice as many nests were found in forbs as in grass. Sweetclover (Melilotus spp.) and curly dock (Rumex crispus) were predominant nest substrates. Orientation of crop rows relative to the waterway influenced agriculture-related disturbance. Nest densities were greater when the rows paralleled the waterways. Nest success (Mayfield estimate) of red-winged blackbirds and dickcissels in waterways was 8.4 and 22.0%, respectively. Predation was the most common cause of nest loss (57% of all nest losses), followed by mowing (16%).
The practice of planting and maintaining grassed waterways clearly provides habitat for a variety of NTMBs. Mowing, however, affects the quality of this habitat. Mowing alters the structure of the vegetation, which in turn may affect the bird community. Dickcissels, common yellowthroats, and red-winged blackbirds preferred to nest in waterways with tall (>60 cm) grass cover, whereas nest densities of vesper sparrows were greater in mowed waterways (Bryan and Best 1994). Grasshopper sparrows nested only in waterways that had been mowed the previous year; sedge wrens nested only in waterways that had not been mowed the previous year. The timing of mowing may also affect populations. Birds that have been displaced from mowed hayfields may move into grassed waterways with have suitable vegetation structure. Mowing waterways at the peak or late in the nesting season may interfere with some birds' last nesting attempt of the season.
Terraces are another kind of strip cover that results from conservation-oriented land use in cropland. Wildlife use of terraces is poorly known but has been examined by D.W. Beck (USDA Natural Resources Conservation Service, Des Moines, Iowa unpubl. data). Beck observed 13 bird species using grassed backslope terraces and found evidence of nesting by mallards, dickcissels, vesper sparrows, red-winged blackbirds, and ring-necked pheasants.
Fencerows are linear habitats that separate agricultural fields with a fenceline and associated vegetation. Best (1983) examined bird use of fencerows with only herbaceous vegetation, herbaceous vegetation with scattered woody plants, and continuous woody vegetation. The number of species was greatest in the continuous woody type and least in the herbaceous type, regardless of season.
In Michigan, 16 species nested in herbaceous and woody fencerows; nest density in fencerows was 43.5 nests/ha (Shalaway 1985). Fencerow width, adjacent field type, and area of open shrubs (i.e., <50% shrub cover 1.5-2.0 m above the ground) most influenced nest density. Wider fencerows, which were more heterogeneous and had greater shrub coverage than narrow fencerows, supported greater nest densities. Fencerows bordered by old fields had more nests than those bordered by crop fields. Nest density and abundance increased with shrub abundance. Song sparrows, American robins, northern cardinals, red-winged blackbirds, gray catbirds, brown thrashers, northern flickers, and ring-necked pheasants were the most frequent nesters in fencerows. Apparent nest success (successful nests/active nests) in fencerows was 58% overall. Raccoons (Procyon lotor), red foxes (Vulpes vulpes), striped skunks (Mephitis mephitis), and long-tailed weasels (Mustela frenata) were responsible for most nest losses. Nest success was lower for larger, ground-nesting game birds than for passerines nesting on or above the ground.
Herbaceous cover along roadsides is used by many species. In Iowa, 35 bird species were observed in roadsides compared with 26 species in adjacent rowcrop fields (Camp and Best 1993). Grass was the dominant vegetation in roadsides and forbs were uncommon. Increased diversity of roadside vegetation can increase bird-species richness (Paruk 1990, Warner 1992). Abundance of some bird species in roadsides was related to vegetation height and vertical density (Camp and Best 1993). Both of these vegetation characteristics are influenced by burning and the composition of grasses (exotic vs. native). The number of bird species observed was inversely related to grass coverage and directly related to the amount of bare ground (Camp and Best 1993). Red-winged blackbird nests were the most common and nest success (Mayfield estimate) was 26%. Predation was the main cause of nest failure, accounting for the fate of 52% of all active nests. Mowing appeared to benefit some species; vesper sparrows and meadowlarks nested on mowed roadside shoulders.
In Illinois roadsides, 92% of all passerine nests were of red-winged blackbirds; other nesting species included dickcissels, brown thrashers, eastern meadowlarks, grasshopper sparrows, vesper sparrows, song sparrows, and sedge wrens (Warner 1992). Nest densities in roadsides were affected by the kind of habitat in the vicinity of the roadside. The number of passerine nests in roadsides was higher in years in which a smaller proportion of the study area was planted to small grains, presumably because less of the non-roadside habitat was suitable for nesting. The 4-mi2 study plot with the most grassland area had the highest density of passerine nests in linear, grassland habitats. Similarly, roadside study plots with the highest density of pheasant nests were in proximity to other prime nesting habitats such as hay (Warner and Joselyn 1986, Warner et al. 1987). In Iowa, the highest densities of red-winged blackbirds were found in roadsides adjacent to idle grasslands and hayfields (M. K. Koob and L. B. Best, Iowa State Univ., Ames, unpubl. data).
Densities of passerine nests in Illinois roadsides also were affected by characteristics of the roadsides (Warner 1992). Nest densities were greater in interstate roadsides than in secondary roadsides and increased with width of the roadside. Nest densities were 4-5 times greater in interstate roadsides dominated by brome-alfalfa (Bromus inermis-Medicago sativa) than in those dominated by fescue (Festuca spp.). There were over six times more passerine nests on managed (mowing deferred until 1 August) than on unmanaged (frequently mowed) secondary roadsides. In this Illinois study, as in the Iowa study (Camp and Best 1993), mowing was confined to the roadside shoulders; the sloped sections were left unmowed.
Higher nest densities in a particular landscape do not necessarily indicate that a species is doing better there; high densities could lead to lower nest success. Ring-necked pheasants in Illinois had low annual nest success in linear habitats (farmland corridors and managed roadsides) in years with low amounts of grassland (strip-cover habitats, forage crops, small grains) per hen in the spring (Warner 1994). Also, nest success in roadsides was lower in years with high nest densities than in years with low densities (Warner et al. 1987). Similar findings have not been reported for passerines.
Shelterbelts are a common feature of midwestern landscapes (Johnson and Beck 1988). For many decades, federal agencies have encouraged farmers to plant shelterbelts or windbreaks to decrease wind erosion and to protect crops, livestock, and farmsteads. Establishment of shelterbelts can provide positive economic returns (Brandle et al. 1992). Shelterbelts usually consist of one to five rows of trees and shrubs (Capel 1988) but can be much wider, especially around farmsteads. Because shelterbelts usually contain trees, they provide habitat for forest and forest-edge species.
In North Dakota, at least 64 species of birds are known to have bred in shelterbelts or tree claims (trees planted by homesteaders) (Cassel and Wiehe 1980). The most common species were the brown thrasher, mourning dove, vesper sparrow, least flycatcher, eastern kingbird, black-billed cuckoo, yellow warbler, American goldfinch, gray catbird, clay-colored sparrow, and American robin. The vesper sparrows and American goldfinches were presumably not nesting in the shelterbelts (see Yahner 1982). Yahner (1982, 1983a) recorded 87 bird species using farmstead shelterbelts in Minnesota and documented nesting in 17 of these. In South Dakota, the number of bird species increased with shelterbelt area both during spring migration and during the breeding season (Martin 1980, 1981). The density and diversity of breeding birds was significantly correlated with both the age and size of shelterbelts in North Dakota (Cassel and Wiehe 1980).
The configuration of shelterbelts can affect the species composition of the avifauna. Shelterbelts with only a few rows of woody plants attracted more birds associated with open habitats, whereas those with many rows attracted more birds associated with forested habitats (Cassel and Wiehe 1980). Yahner (1983a), however, provided correlations between variables describing vegetation structure and avian community or population variables in shelterbelts and found that the perimeter and length of shelterbelts were associated more often with measures of community structure than were area and width of shelterbelts.
The avian species composition of shelterbelts also can be affected by vegetation structure and landscape context. Martin and Vohs (1978) found the highest bird species diversity in South Dakota shelterbelts with developed tree canopies and lush grass layers; dense shrub growth under the trees was not preferred. In Minnesota, vegetation structure was a major factor determining bird community structure in shelterbelts, with older belts having lower densities of shrubs, higher densities of trees, and greater bird-species richness (Yahner 1983a). Distance to wooded and old-field habitats, and the amount of cropland and pastureland surrounding shelterbelts, influenced which birds were found in shelterbelts. A concentration effect occurred in isolated shelterbelts, which were more likely to have certain species, probably because of the paucity of similar habitat in the vicinity. In addition to species occurrence, it is important to examine nest success (Yahner 1982, 1983b). Apparent nest success (successful nests/active nests) of mourning doves and American robins was 32% and 56%, respectively.
Additional perspectives on shelterbelts can be found in several papers. Johnson and Beck (1988) reviewed shelterbelt management and wildlife. They classified bird species on the basis of how much they benefit from shelterbelts and discussed characteristics of shelterbelts that relate to wildlife use. Schroeder (1986) presented a Habitat Suitability Index model for wildlife species richness in shelterbelts, which reliably predicts species richness over much of the Great Plains (Schroeder et al. 1992). Podoll (1979) distinguished between shelterbelts as a habitat that is essential for some species but merely used by other species. Johnson (1996) points out that trees may have detrimental effects on prairie birds.