Northern Prairie Wildlife Research Center
We attribute the differences in nest success between coyote areas and fox areas to coyote suppression of fox abundance. We found evidence of fox predation at 4% of the depredated large-clutch nests in coyote areas compared with 27% in fox areas, indicating that our assignments of study areas to canid species were accurate. We believe all areas were suitable for occupancy by coyotes, foxes, or both. Although there was more grassland and less cropland at coyote areas than in fox areas, the difference was small (12 percentage points). Previous studies have shown intensively farmed areas are suitable for coyotes to occupy (Andelt and Gipson 1979, Sargeant et al. 1987, Sargeant et al. 1993). Our methods of selecting study areas, which resulted in much interspersion of coyote and fox areas, reduced the likelihood that some local factor (e.g., disease) differentially affected predator abundance. We found no difference in species composition of predator communities, with exception of canids, on coyote areas compared with fox areas. Although the possibility that coyotes affected abundance of predators other than foxes sufficiently to have a noticeable effect on duck nest success cannot be ruled out, there is little evidence in the literature indicating that interspecific relationships between coyotes and other predators regulate abundance of other predators, except with foxes. Sargeant et al. (1993) examined relationships between coyotes and the 4 other large carnivores present in our study areas. Except for a strong inverse relationship between coyotes and red foxes, the only other negative relationship they detected was a weak inverse relationship with raccoons.
The finding that average nest success in coyote areas was 15 percentage points greater than in fox areas is important because this level of nest success was above the 15-20% suggested thresholds for population stability of several dabbling duck species (Cowardin et al. 1985, Klett et al. 1988). With 17% nest success as the threshold level for population stability, our data indicated that all 17 coyote areas, but only 7 of 13 fox areas, had nest success above that level.
The variability in nest success we found among coyote and fox areas shows that presence of coyotes does not ensure high nest success. Conversely, the presence of foxes does not necessarily result in low nest success. Some fox areas had higher nest success than some coyote areas. Many factors probably affect nest success in individual areas, such as location of canid rearing dens, number of individuals in the canid family, availability of alternative prey, differences in individual foraging habits, and abundance of other predators. The greater variability of nest success in our mixed-canid areas compared with coyote or fox areas is consistent with probable variation in use of such areas by coyotes and foxes.
Two factors may explain why our estimates of nest success rates in coyote and fox areas are higher than rates generally reported for the PPR (Greenwood et al. 1987, Johnson et al. 1987, Klett et al. 1988). First, our study was conducted when substantial cultivated land was enrolled in the Conservation Reserve Program and seeded to grassland habitat (Bjerke 1991); many of our study areas had large enrolled fields of grassland nearby. The additional grassland may have resulted in greater dispersion of duck nests and reduced their risk to predation (Ricklefs 1969). Second, drought persisted throughout the study and probably reduced duck abundance (Stoudt 1971, Krapu et al. 1983) and nesting effort (Smith 1969, Krapu et al. 1983, Greenwood et al. 1987). Low nest density may have a positive influence on nest success by reducing predator efficiency in finding nests (Marshall 1967, Weller 1979, Hill 1984).
Although our results show that coyotes benefit ducks by reducing nest depredation by foxes, coyotes also can prey extensively on duck nests, especially where coyotes are considered abundant (Glup and McDaniel 1988). Our study was conducted in a region where coyote populations were relatively low. Sargeant et al. (1987) showed that family territories of coyotes in North Dakota averaged 61 km², which was probably comparable with family territory size of coyotes in our study areas.
The magnitude of difference in nest success that we found between coyote and fox areas probably does not apply to all of the PPR, to all habitat types, or to all years. For example, in our areas there were few American crows, which are believed to be important predators of duck nests in other parts of the PPR (Dzubin and Gollop 1972, Johnson et al. 1989, Sargeant et al. 1993). Depredation by these or other predator species that were uncommon or absent on our study areas could negate the benefits of coyote suppression of fox predation. Also, higher populations of predator species that occurred in our study areas could negate the benefits of coyote suppression of fox predation. Alternatively, the benefits to nest success also could be greater than we documented if other predator species populations were lower than in our study areas.