Northern Prairie Wildlife Research Center
Another characteristic of grassland bird assemblages in North America is the prevalence of short-distance migrants (MacArthur 1959, Willson 1976). In this study, short-distance migrants composed more than one-half of the observed species and indicated pairs in each year. These patterns are comparable to those of other grasslands in North America but contrast greatly with northeastern deciduous forests, where long-distance migrants and permanent residents dominate the breeding avifauna (MacArthur 1959, Willson 1976).
One species, Northern Waterthrush (Seiurus noveboracensis), was recorded only in 1967, compared with 32 species detected only in 1992 and/or 1993. Although differences in observer abilities between 1967 and 1992-93 may account for the increase in the number of species between the two periods (Table 1), many other recent or irregular events may have contributed to this increase. These include food-based nomadism (Pine Siskin [Carduelis pinus], DuBowy 1983; Short-eared Owl [Asio flammeus], Stewart 1975; Long-eared Owl [Asio otus], Marks et al. 1994), increased nest box availability (Hooded Merganser [Lophodytes cucullatus], Doty et al. 1984; Eastern Bluebird [Sialia sialis], Sauer and Droege 1990), successful reintroduction and management programs (Wild Turkey [Meleagris gallopavo], Johnson and Knue 1989; Canada Goose [Branta canadensis], Lee et al. 1984), and increased habitat availability (species associated with woody vegetation, see below). Nonetheless, most of the 32 species that were not observed in 1967 are considered rare, uncommon, localized, or irregular breeders in North Dakota (Faanes and Stewart 1982, DeSante and Pyle 1986).
Our results (Table 1 and Table 2; Appendix 1 and Appendix 2) are consistent with earlier reports showing that breeding bird populations in grassland ecosystems are tremendously dynamic, exhibiting considerable annual variation in abundance and frequency of occurrence (Graber and Graber 1963, George et al. 1992, Zimmerman 1992). Within breeding habitat associations, our observed patterns of population change were remarkably similar and consistent among species with different migratory behaviors. First, many species associated with woody vegetation (e.g. House Wren [Troglodytes aedon]) and human-made structures (e.g. Barn Swallow [Hirundo rustica]) generally increased between 1967 and 1992-93 (Appendix 1 and Appendix 2). Second, many grassland (e.g. Savannah Sparrow) and wetland (e.g. Wilson's Phalarope) species declined between 1967 and 1992-93, though many exhibited slight (e.g. American Coot), moderate (e.g. Western Meadowlark), or dramatic (e.g. Common Yellowthroat) increases between 1992 and 1993.
The two patterns of long-term population change evident in our data were consistent with trends from the BBS. For both surveys, most species with significantly increasing population changes were associated with human-made structures or woody vegetation, and most species with significantly decreasing population changes were associated with wetlands or grasslands. This concordance illustrates that both independently derived measures of population change likely were recording the same phenomena. Moreover, of the 46 and 64 species that had significant population changes in our survey and the BBS, respectively, 26 species had significant trends in both data sets; all 26 species were considered common species in North Dakota (Appendix 1 and Appendix 2). Only one of the 26 species, Grasshopper Sparrow, had a population change that was opposite in sign. This discrepancy could be attributed to many factors, including differences in survey methodology or trend analysis. However, the long-term trend from the BBS may have masked the recent population increase in Grasshopper Sparrows in the Great Plains; these increases have been associated with the increased availability of perennial grassland habitat established by the Conservation Reserve Program (Reynolds et al. 1994, Johnson and Igl 1995).
The causes of the above patterns of population change may involve conditions in the breeding, migration, or wintering seasons (Sherry and Holmes 1992). The prevalence of parallel patterns of population change among habitat associations, but not among migratory behaviors, suggests that these population changes may be caused, at least in part, by conditions on the breeding grounds (Finch 1991). A common feature of breeding birds observed in this study is their dependence on habitats on the breeding grounds; most of these species breed in the northern Great Plains but winter elsewhere. For example, if breeding ground conditions are primary influences on bird population dynamics, then populations of long-distance migrants should show similar patterns of change as short-distance migrants. The examples of Blue-winged Teal (Anas discors) and American Coot for wetlands, Baird's Sparrow (Ammodramus bairdii) and Savannah Sparrow for grasslands, and House Wren and Blue Jay for open woodlands are illustrative (Appendix 1 and Appendix 2).
What conditions or factors might limit populations on the breeding grounds in the northern Great Plains? Habitat loss and changes in land use have been viewed as major potential causes of population change of many breeding birds in the Great Plains (Knopf 1988, 1994; McNicholl 1988). Within the last century, the landscape of the Great Plains has been greatly modified by a number of human-induced changes. The once abundant grasslands and wetlands of the Great Plains have been drastically reduced, altered, and fragmented by intensive agriculture, roads, tree plantings, encroachment by woody vegetation, invasion of exotic plants, and other human activities. Knopf (1994) described several historic, contemporary, and continuing influences on breeding birds and their habitats in the Great Plains, including: (1) removal of native grazers and transformation to intensive, domestic livestock grazing; (2) cultivation of native grasslands; (3) loss of wetlands; and (4) woody development in the form of tree plantings and ecological invasions. The first three influences were implicated in the recent declines of grassland bird populations, as well as wetland associates. The fourth provided increased nesting opportunities for species associated with woody vegetation that generally were lacking or were limited in presettlement times. Likewise, Herkert (1994), Vickery et al. (1994), and Warner (1994) implicated fragmentation of grasslands in the recent declines of breeding birds in eastern and midwestern grasslands.
Before European settlement, the landscape of North Dakota included vast expanses of grassland, consisting primarily of northern mixed-grass prairie with some tallgrass prairie on the extreme eastern edge of the state (Risser et al. 1981). Moreover, the grasslands of the prairie pothole region (i.e. Drift Plains and the Missouri Coteau; Fig. 1) were dotted by millions of shallow wetland basins. Since settlement, North Dakota has lost about 49% of its wetlands (Dahl 1990) and 75% of its native grasslands (Samson and Knopf 1994), much of this before 1967. We compared the statewide changes in bird populations (Table 1 and Table 2, Appendix 1 and Appendix 2) with the overall changes in land use area (Table 3) between 1967 and 1992-93. Long-term population changes in our study and BBS were consistent with the notion that these changes were, in part, caused by changes in land use on the breeding grounds (e.g. Knopf 1994). Declines in wetland species were commensurate with declines in wetland area. The increases in species associated with human-made structures and woody vegetation were consistent with the overall increases in area of those habitats. The declines in grassland birds paralleled the decrease in the combined total area of grassland and secondary grasslands (hayland and planted cover). Bernstein et al. (1990) repeated the historic survey of Kendeigh (1941) in northwestern Iowa and found similar declines in grassland birds and increases in species associated with woody vegetation.
Land use area, however, changed very little between 1992 and 1993, but our results suggest that many grassland and wetland species, as well as several other species, shifted similarly by increases in frequency and abundance between the two recent years (Table 1 and Table 2, Appendix 1 and Appendix 2). Droege and Sauer (1989) and Peterjohn and Sauer (1993) suggested that breeding bird population declines on BBS routes in the late 1980s and increases in the early 1990s reflect extreme drought conditions and amelioration of drought conditions, respectively. Extended periods of below-average rainfall are known to affect populations of grassland and wetland birds (e.g. Johnson and Grier 1988). In this study, the increases between 1992 and 1993 corresponded with the amelioration of the long-term drought conditions and are consistent with observations of behavioral flexibility and opportunism in habitat selection (Cody 1985) and weather-related shifts in distribution and abundance by grassland (Wiens 1974) and wetland (Johnson and Grier 1988) species. George et al. (1992) also noted rapid recoveries of grassland bird densities after drought. Reproductive success of grassland birds varies annually, with very poor productivity during some drought years (George et al. 1992).
In conclusion, our results indicate that BBS data reflect real population changes for many common species in North Dakota. The increase in species associated with human-made structures and woody vegetation and decreases in wetland and grassland species in agricultural landscapes have been reported previously, both in North America (McNicholl 1988, Knopf 1994, Herkert 1995) and elsewhere (Böhning-Gause and Bauer 1996). Our data suggest that land use changes, and probably drought conditions, influenced some breeding bird population changes in North Dakota between 1967 and 1992-93. Although these factors would not necessarily affect different species to the same extent, the fact that our surveys detected significant population changes for several common species (e.g. Northern Shoveler [Anas clypeata]) that the BBS did not detect and vice versa (e.g. Sora [Porzana carolina]) deserves further investigation. Furthermore, we cannot attribute all population changes to conditions on the breeding grounds. However, because many birds in the Great Plains are short-distance migrants, many of the factors driving population changes on the Great Plains are likely North American processes (Knopf 1994, Herkert 1995). Nonetheless, our data indicate that factors on the breeding grounds are having a major effect on breeding bird populations in the northern Great Plains.