USGS - science for a changing world

Northern Prairie Wildlife Research Center

  Home About NPWRC Our Science Staff Employment Contacts Common Questions About the Site

Wildlife Monographs

Red Fox Predation on Breeding Ducks
in Midcontinent North America

Results and Discussion

Duck Mortality Estimates

Expansion Factors for Estimating Total Ducks Taken by Foxes.—Our predation rate indices can be expanded into meaningful estimates of total numbers of ducks taken by foxes during the denning season by accounting for (1) ducks brought to dens, but not detected in surface food remains, (2) ducks taken by foxes, but not brought to or detected at dens, (3) average number of dens used by individual fox families to rear pups, and (4) numbers of fox families. The expansion formula is:

Total Ducks Taken by Foxes

=PRI x 1/PSD x 1/PDD x NFD x NFF,


   PRI = predation rate index,
   PSD = percentage of total ducks at dens accounted for by surface food remains only,
   PDD = percentage of ducks taken by foxes that are brought to dens,
   NED = average number of dens used by foxes to rear a litter, and
   NFF = number of fox families.

Johnson and Sargeant (1977) used this method to determine red fox impact on mallard sex ratios in North Dakota, and they provide most of the detail needed for expansion of PRI.

Johnson and Sargeant (1977:9) found that 49% of mallards found at completely excavated dens were detectable in surface food remains. We found that 53% of all ducks from 235 completely excavated dens were detected in surface remains. However, the proportion detected in surface remains increased as the denning season progressed because older pups fed more above ground (Fig. 9). Thus, PSD = 53% and we used an expansion factor of 1/0.53 = 1.9 to account for below ground ducks.

gif -- graph(percent of ducks above ground vs. time of the fox denning season)

Fig. 9. Relation between percentage of ducks at red fox rearing dens that were detected in surface food remains and week of the fox denning season (r=0.873, P < 0.0002). Correlation based on 235 completely excavated fox dens and weighted for numbers of ducks (in parentheses) in each week category.

Johnson and Sargeant (1977:9-10) provided unpublished data by G. L. Rohde concerning the proportion of mallards taken by red foxes that were recovered at excavated dens. His findings were based on 285 radio-equipped mallards taken by foxes. Monthly proportions recovered at dens were 18% (Apr), 47% (May), and 16% (Jun). We visited very few dens in April; thus an expansion factor reflecting the May and June rates is most applicable to our data. Because of the time lapse between a fox taking a duck and our finding remains of the duck at a den, we believe our early (22 Apr-11 Jun) and late (12 Jun-20 Jul) sample periods best reflect G. L. Rhode's May and June findings, respectively. About one-half of the ducks we found at dens were from each of these periods (Table 4); therefore we assume that 32% (average of G. L. Rohde's May and Jun results) of the ducks taken by foxes in our study were recoverable from dens. With PDD = 32% we used an expansion factor of 1/0.323.1 to account for the additional ducks taken by foxes that were unaccounted for in remains found at completely excavated dens.

Data reported by Johnson and Sargeant (1977:11) on numbers of dens used by fox families apply equally well to our study. Their data for 16 fox families indicated that red foxes in agricultural areas use an average of 4.8 dens to rear each litter. However, they reasoned 1 den should be subtracted to account for the 4-5 weeks before pups emerge from dens and few prey are brought to dens. We found few dens during the preemergence period. Thus, we used NED = 3.8 to account for the number of rearing dens used by the average fox family during the fox denning season.

The above 3 factors result in an overall expansion factor of 1.9 x 3.1 x 3.8 = 22.4 to convert fox predation rate index values into estimates of total numbers of ducks taken by the average fox family during the entire fox denning season. The expansion factor is large because a very small proportion of overall fox predation is revealed by examinations of food remains found above ground at the average fox den.

Fox Impact on a North Dakota Duck Population.—Numbers of red fox families (NFF) were determined for each of the 6 detailed study townships in the 3-county intensive study area for each year, 1969-73 (Allen et al. 1974, Sargeant et al. 1975). We used numbers of families per unit area as our measurement of population density. Average densities were lower on the Drift Plain than on the Missouri Coteau, but densities increased greatly in both regions from 1969 to 1973 (Table 10).

By applying the expansion factor of 22.4, we estimated the average number of dabbling ducks taken per fox family during the denning season each year ranged from 16.1 on the Drift Plain in 1971 (a drought spring) to 65.9, also on the Drift Plain, in 1969 (a very wet spring) (Table 11). The average estimate of numbers of dabbling ducks taken per fox family per year for the 5-year period was 25% higher on the Missouri Coteau than on the Drift Plain. Moreover, the estimated average annual take of dabbling ducks by foxes per square kilometer on the Missouri Coteau was more than double the average annual take per square kilometer on the Drift Plain.

The total number of dabbling ducks taken by foxes on the 13,528-km2 3-county intensive study area was estimated to range from about 34,000 to 56,000 annually (Table 11). Annual variations were caused primarily by differences in fox and duck densities. The fox population increased substantially during the 5-year period and, except for 1972, the dabbling duck population declined (Table 8). The result was relative constancy in the total number of dabbling ducks taken by foxes each year except 1972. In 1972 the combined effects of a moderately dense fox population and a relatively high dabbling duck population with good nesting conditions resulted in greatly increased predation.

The percentage of the dabbling duck population taken by foxes in the intensive study area was calculated from duck density data provided by Stewart and Kantrud (1974) and from our estimate of total annual mortality per square kilometer (Table 11). Stewart and Kantrud (1974) found that dabbling duck densities on the Drift Plain and Missouri Coteau were similar and averaged 16.5 pairs/km2. We estimated the foxes took an average of 2.97 dabbling ducks/km2 annually from the intensive study area (Table 11), and we found that 75% of the ducks taken by foxes in that area were hens. This mortality represented an average annual take of 13.5% of the hens and 4.5% of the drakes.

Total Number of Ducks Taken.—We expanded our estimate of numbers of dabbling ducks taken by foxes in the 3-county 11 intensive study area during each denning season to the full pothole region of North Dakota (Table 11). We believe the predation rate indices from the intensive study area (Table 8) apply to all of the North Dakota pothole region though they average 1.66 ducks/den, which is lower 14.8 than the average of 1.80 ducks/den obtained for all dens visited in east river North Dakota (Table 7). The latter figure is higher than the average for the intensive study area primarily because it includes 39 dens from USFWS refuges where the predation rate averaged 4.4 ducks/den. Evidence indicates that the fox density data (Table 10) also apply to the entire North Dakota pothole region (Allen and Sargeant 1975, Johnson and Sargeant 1977:7). Stewart and Kantrud (1974) found there were no great differences in dabbling duck densities throughout this region except for about 26% fewer dabbling ducks (primarily blue-winged teals) in the northwestern portion of the Drift Plain. Expansion of our estimates of average annual mortality for the Drift Plain and Missouri Coteau portions of the intensive study area (Table 8) to North Dakota's 94,245km2 pothole region (68,661 km2 of Drift Plain and 25,584 km2 of Missouri Coteau) resulted in an estimated average annual take of about 242,000 dabbling ducks (2.56/km2) during the approximate 3-month fox denning season.

Assessments of losses of dabbling ducks to foxes in other parts of the midcontinent area are more speculative than for east river North Dakota; nevertheless, they are useful for placing the magnitude of fox predation on breeding ducks in the midcontinent area in perspective. Data were sufficient to estimate average annual mortality for all parts of the Prairie Pothole Region (except northern Montana), for Iowa, and for most of the Great Plains west of the Missouri River in North Dakota and South Dakota.

Based on differences in predation rate indices (Table 7), breeding duck densities, and fox population densities, we estimated the average number of adult dabbling ducks taken annually by foxes during the denning season in each of the other breeding ground areas sampled was 74,000 in east river South Dakota, 2,000 in Iowa, 13,000 in Minnesota, 13,000 in west river North Dakota, and 31,000 in west river South Dakota. For the Prairie Pothole Region of Canada, we estimated that an average of about 386,000 adult dabbling ducks was taken annually by foxes during the denning season. In deriving these estimates we assumed fox densities in east river South Dakota, Iowa, and Minnesota were the same as in east river North Dakota, but that in the 2 west river areas and Canada, fox densities were 50 and 27% of that in east river North Dakota, respectively. Our combined average annual mortality estimate for all areas was 761,000 adult dabbling ducks. This estimate, however, does not include diving ducks, ducks taken by foxes after the denning season, and ducks taken in portions of the midcontinent area not sampled (i.e., Prairie Pothole Region of northern Montana). We estimated an average of about 6,000 diving ducks were taken annually by foxes in east river North Dakota during the denning season based on our finding that 2.4% of the ducks found at dens in that area were diving ducks. We believe that considerably more diving ducks were taken by foxes in the Prairie Pothole Region of Canada than in North Dakota (even though fox densities were much lower) because of the much larger area involved and the increased proportion of diving ducks in the population.

Fox predation on adult ducks after the denning season was probably far less severe and less hen-selective than during the denning season. In much of the midcontinent area, especially during the dry years, the nesting season was nearing completion by the end of the fox denning season. Once nesting is completed, most ducks remain on or near water where they are relatively secure from foxes. Also, maturing fox pups forage on their own after the denning season (Allison 1971) thereby reducing their dependence on adults. Because pups are inexperienced, they are more apt to feed on readily available small prey than on large prey such as adult ducks. With hunting demands decreased, adult foxes may prey less on ducks and more on seasonally available small items such as insects and berries (Scott 1943:467468).

The average annual take of ducks by foxes in areas not included in our estimates of total ducks taken was probably small compared to that for the rest of the region because of relatively low nesting duck populations, low fox populations, or both. A realistic estimate of the average annual take of adult ducks by red foxes during spring and summer in the total midcontinent area (includes the Prairie Pothole Region of Canada) during 1969-73 was about 900,000 birds, annually.

Previous Section -- Results and Discussion - Fox Predation Rate Indices
Return to Contents
Next Section -- Results and Discussion -- Duck Contribution to Fox Diet

Accessibility FOIA Privacy Policies and Notices

Take Pride in America logo logo U.S. Department of the Interior | U.S. Geological Survey
Page Contact Information: Webmaster
Page Last Modified: Saturday, 02-Feb-2013 05:54:10 EST
Sioux Falls, SD [sdww55]