Northern Prairie Wildlife Research Center
Declines in grassland bird populations have been attributed to a number of causes, including the loss of suitable nesting habitat on the breeding grounds. The vast grasslands that once dominated the landscape in the northern Great Plains are largely gone or degraded (Samson and Knopf 1994). Recent analysis of BBS data indicates that some species that were declining in North Dakota before the CRP began (i.e., 1967-1986), such as the Grasshopper Sparrow and the Lark Bunting, have shown population increases coincident with the establishment of perennial grasslands after the CRP began (i.e., 1987-1992) (Reynolds et al. 1994). These results, in combination with our findings, suggest that the CRP provides not only important breeding habitat for some grassland birds but also a possible vehicle for restoring abundant populations of these species.
Most of the species more common in cropland than in CRP fields select the sparser cover that cropland affords (e.g., Graber and Graber  and Stewart  for Killdeer; Owens and Myres  and Skinner  for Horned Lark; Owens and Myres  for Chestnut-collared Longspur; Wiens , Owens and Myres , and Rodenhouse and Best  for Vesper Sparrow). Conversely, the Upland Sandpiper frequently forages in sparse cover (Graber and Graber 1963) but prefers to nest in somewhat more robust vegetation (Kantrud and Higgins 1992, Bowen and Kruse 1993). Higher densities of Upland Sandpipers found in cropland may reflect use primarily for foraging, and CRP fields may offer nesting habitat that is otherwise limited in highly cultivated or heavily grazed areas of the northern Great Plains.
Although the comparisons we made apply to North Dakota, the implications extend to other areas in the northern Great Plains. Currently, there are nearly 4 million ha enrolled in CRP in the northern Great Plains (Montana, Minnesota, North Dakota, and South Dakota), which accounts for nearly one- quarter of the total land in the CRP nationwide (Johnson et al. 1994).
Recent evidence suggests that CRP also benefits grassland bird populations in other areas of intensive agriculture such as the Midwest. In Nebraska, King (1991) found 16 species of birds in CRP fields and 13 species in native prairie during the breeding season, compared with only two species, Horned Lark and Killdeer, in cropland. Likewise, in Iowa, Patterson (1994) found 16 species of birds nesting in CRP fields compared with only two species, Horned Lark and Vesper Sparrow, in similar areas of cropland. Our data indicate that the addition of CRP habitat in the North Dakota landscape reduced numbers of a few species, such as Horned Larks and Vesper Sparrows, species whose populations had not declined in the last quarter-century.
Some breeding birds were undoubtedly missed in our single survey (e.g., Järvinen and Lokki 1978). The same techniques were used in both surveys, however, so any bias should be consistent. Because estimates of population size are minimal, projections of population change anticipated from termination of CRP are conservative. Projections also assume habitat-specific densities of species and that densities in one habitat will not change dramatically in response to changes in the availability of another. This simplifying assumption is necessary and probably adequate as long-term averages. The year-to-year variability evident in the densities of birds (Table 2) also points to the need to consider averages over long periods of time.
Our projections are based on the assumption that all CRP fields would be returned to cultivation. Some small fraction would likely be retained in perennial vegetation and used for grazing or forage production (Mortensen et al. 1989). Such habitat would support different bird populations from cropland, but probably lower numbers than CRP habitat maintains.
Although we did not examine reproductive success, it seems reasonable that birds are more successful in robust and undisturbed cover than they would be in crop fields, most of which are cultivated, sprayed, or otherwise disturbed frequently during the breeding season (e.g., Best 1986). For several waterfowl species, Kantrud (1993) and Reynolds et al. (1994) found that nest success in CRP fields was as high or higher than in habitats specifically purchased and managed for waterfowl production.
The Conservation Reserve Program is an agricultural commodities program and is inherently expensive. As such, it does not compete with programs to conserve habitat for wildlife. The primary benefits of CRP include major reductions in soil erosion in the Great Plains; from 1987 to 1992 North Dakota croplands experienced a 68% reduction in erosion from wind and a 22% reduction in erosion from water (USDA 1994). CRP does, however, offer far greater benefits to breeding birds and other wildlife populations than do other agricultural programs such as annual set-asides and summer fallow.