USGS - science for a changing world

Northern Prairie Wildlife Research Center

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

An Assessment of Predator Barriers and Predator Control to Enhance Duck Nest Success on Peninsulas

John T. Lokemoen and Robert O. Woodward

Table of Contents



Annual recruitment in duck populations in the northern plains is less than its potential because predators take nesting females, eggs, and ducklings (Sargeant and Arnold 1984). Most predation is caused by Franklin's ground squirrels (Spermophilus franklinii) and 6 carnivores (raccoon [Procyon lotor], mink [Mustela vison], striped skunk [Mephitis mephitis], badger [Taxidea taxus], red fox [Vulpes vulpes], and coyote [Canis latrans]) (Sargeant and Arnold 1984). Although breeding waterfowl are impacted by predation in upland habitats, nest success is usually high on islands where access by mammalian predators is impeded (Giroux 1981, Willms and Crawford 1989). Moreover, duck nest success has been enhanced where nesting habitat is surrounded with electric fencing and predators are removed from the enclosed areas (Lokemoen et al. 1982, Greenwood et al. 1990).

Although duck nest success can be high on islands, island construction is expensive (Lokemoen 1984). We added predator barriers (i.e., fences or moats) and removed predators from peninsulas in wetlands to determine if this management tactic would increase duck nest success over that on peninsulas without predator barriers and predator control.

Study Area and Methods

Peninsula Selection

In 1984, 10 pairs of peninsulas in semipermanent and permanent wetlands (Stewart and Kantrud 1971) were selected from a sample of approximately 70 sites in central North Dakota. Pairs were chosen because they were close neighbors (Symbol - Mean x separation ≤15 km), contained undisturbed, grassy, or low shrub cover, and were adjacent to wetlands suitable for brood rearing. One peninsula of each pair was randomly selected for a fence or a moat and predator control; the other peninsula served as a control. Eight of the predator barriers were electric fences and 2 barriers were water-filled moats. Managed peninsulas ranged from 1.6-31.6 ha (Symbol - Mean x = 14.6 ha). Controls varied from 1.6-38.9 ha (Symbol - Mean x = 9.7 ha).

Our study was initiated in 1985 on 5 pairs of fenced peninsulas and on 1 pair of moated peninsulas. Two pairs of fenced peninsulas and 1 pair of moated peninsulas were added in 1986, and 1 additional pair of fenced peninsulas was added in 1987. All peninsular pairs were studied through 1989. Data from 1 moated pair were not used in 1987 and 1988 because high water flooded the control site.

Electric Fences and Moats

Electric fence barriers extended across the base of the peninsulas and into the water on each side. Fences averaged 333 m in length (range = 62-788 m), with an average of 289 m of fence on dry land and 44 m in wetland. Fences were 18-gauge, 2.5-cm-mesh poultry netting that extended from 30 cm below ground level to 1.7 m above ground. The lower 60-90 cm of the poultry netting was vinyl-clad; the upper portion was galvanized. All poultry netting was attached to 2.4-m-long wooden posts set in the ground 0.6 m. Two 12.5-gauge energized wires were placed on the side of the fence facing the headland, 1.2 m above ground, and 6-13 cm out from the poultry netting. Wires were held in place by fiberglass rods driven into the wooden posts and insulators attached to the wire. Another energized wire was placed 6 cm above the poultry netting. The top 30 cm of the fence extended toward the peninsula headland at a 45 angle. Fences were powered by a solar-charged battery and a high-voltage electrical energizer.

Moats were excavated at the base of both peninsulas to create a water barrier. Each moat was 91 m wide, 0.9 m deep, and averaged 630 m in length.

Predator Removal

Predators were controlled on all managed peninsulas in our study to make the system effective. Constructing barriers without removing predators would be futile, because enclosed animals likely would depredate nests. Also, several aquatic carnivores are endemic to the study area, and they must be removed if they gain access to protected areas. We assumed that trapping alone would not increase nest success, because predators disperse and rapidly occupy vacant habitat (Sargeant and Arnold 1984:164).

Mammalian predators were captured using an average of 4 quick-kill body traps and 1 leg-hold trap/managed peninsula. Predator trapping extended from early April (when fences were energized) until mid-July 1985-1989. Traps were placed only on the protected portion of the peninsulas. Records of predators removed from managed peninsulas were collected during 1986-1989.

Nest Searches

During 1985-1989, we searched all upland cover on peninsulas for waterfowl nests, twice in both May and June. Nest searches were conducted with all-terrain vehicles (83%, n = 71 searches) that pulled a 38-m chain with 0.8-cm-diameter links, or with 2 people that pulled a 30-m weighted rope (17%, n = 15 searches) (Klett et al. 1986). We placed a wire marker 4 m north of each duck nest and recorded the species, clutch size, and incubation stage (Klett et al. 1986). Nest success was determined using the modified Mayfield method (Johnson 1979). Nest density was calculated by dividing the total number of nests by the area of upland on each peninsula.

Construction Costs

Fence construction costs were estimated by multiplying total length of all fences by mean cost/m incurred in building 4 fences. Ducks Unlimited, Inc. (Great Plains Reg. Off., Bismarck, N.D.) funded moat construction. Mean annual maintenance costs of fences and moats included fencing supplies and predator traps, labor for repairs and trapping, and travel. Labor costs were based on a federal wage scale of $8.91/hour. Travel costs were estimated at $0.25/km. We used the Water Resources Council (Washington, D.C.) standard amortization rates of 0.109 for a projected 20-year fence and 0.090 for an assumed 50-year moat. We calculated the net number of ducklings hatched on managed peninsulas by subtracting the number of ducklings hatched on the control peninsulas. To determine the number of fledged birds, we multiplied the number of hatchlings times an annual survival rate of 0.54 (Lokemoen 1984). Finally, annual fence and moat costs were divided by the number of fledged young to determine cost/fledgling from either fenced or moated peninsulas.

Statistical Methods

Significance was deemed a priori at alpha = 0.05. We used a chi-square test of homogeneity (Dowdy and Weardon 1983:113) to determine if frequencies of captured predator species differed between moated and fenced peninsulas. Because we captured small numbers of several species, we only tested for differences among raccoons, striped skunks, and others combined. We also combined data across years. We used this same approach to test the null hypothesis of no difference in frequency of occurrence of duck species nesting on managed peninsulas and control peninsulas. A chi-square goodness-of-fit test (Dowdy and Weardon 1983:108) was used to compare species composition of ducks nesting on managed and control peninsulas with that observed in the pothole region of North Dakota during 1985-1989 (U.S. Fish and Wildl. Serv. 1990). Significant overall results were followed by pairwise Z-tests to identify specific differences with respect to predator and duck species composition (Neter et al. 1982:438-441).

We used a repeated-measures analysis of variance (ANOVA) to test for differences in nest success and number of hatched young/ha between fenced and control peninsulas (Milliken and Johnson 1984:322-350). Repeated-measures ANOVA was used because the same peninsulas were measured each year. The number of hatched young/ha were natural-log transformed for the analysis to correct for unequal (P < 0.001) variances (Levene's test [Milliken and Johnson 1984:22]). We did not make statistical comparisons for moated peninsulas because of small sample size (n = 2). We used simple correlation analysis (Steel and Torrie 1980:272) to determine possible relationships between peninsula area and nest density and between peninsula area and the number of hatched young/ha. All statistical analyses were done with the SAS PC System (SAS Inst. Inc. 1987).


Species Composition

Mallards (Anas platyrhynchos), gadwalls (A. strepera), and blue-winged teal (A. discors) were the most numerous nesting ducks on peninsulas (Table 1). They accounted for 75 and 86% of all nests on managed and control peninsulas, respectively. Species composition of ducks nesting on managed peninsulas differed (χ² = 1,169.3, 6 df, P < 0.001) from the overall species composition of breeding ducks in North Dakota. Disproportionately more gadwalls and lesser scaup (Aythya affinis) occurred on managed peninsulas than statewide (Table 1). Mallards, blue-winged teal, and northern shovelers (Anas clypeata) were less common. On control peninsulas, gadwalls and blue-winged teal composed more of the population than expected (χ² = 125.9, 6 df, P < 0.001), whereas other species occurred less than expected. We also detected a difference (χ² = 87.15, 6 df, P < 0.001) in species of nesting ducks between managed and control peninsulas, with gadwalls and lesser scaup more common and blue-winged teal less common on managed peninsulas (Table 1).

Nest Success and Nest Densities

Mean nest success on fenced peninsulas was greater (F = 26.81; 1,7 df; P = 0.001) than that on control peninsulas (Table 2). Mean numbers of young/ha hatched on fenced peninsulas were 10 times greater (F = 8.46; 1,7 df; P = 0.023) than numbers of hatched young/ha on control peninsulas. On fenced peninsula pairs, there were neither yearly fluctuations in nest success (F = 1.19; 4,41 df; P = 0.329) nor in number of hatched young/ha (F = 2.10; 4,48 df; P = 0.096).

Although only 2 pairs of moated and control peninsulas were available for assessment, moats and predator control seemed to increase duck nest success and production. Overall, nest success was 75 and 14% on moated and control areas, respectively (Table 3). Annually, <1 hatchling/ha was produced on control peninsulas compared to >20 ducklings/ha on peninsulas with moats and predator control.

Of 6 primary nesting species, gadwalls were most numerous during the 5 years, with a mean density of 1.19 nests/ha. Mean densities of nests/ha were 0.56, 0.48, 0.46, 0.22, and 0.11 for mallard, blue-winged teal, lesser scaup, northern pintail (Anas acuta), and northern shoveler, respectively. Averaged over all species and years, there were 3.0 nests/ha (SE = 0.536) on fenced peninsulas (n = 36), 3.2 nests/ha (SE = 0.826) on moated peninsulas (n = 7), and 0.7 nests/ha (SE = 0.114) on control peninsulas (n = 43).

Fences in this study were not an obstacle to broods moving from nests to wetlands, because fences were relatively short (Symbol - Mean x = 333 m) and extended straight across the peninsula. However, broods have had problems exiting peninsulas with fences >800 m long and fences with corners.


During 1986-1989, 88 predators (Symbol - Mean x = 2.8 animals/yr/peninsula) were removed from fenced peninsulas and 14 (Symbol - Mean x = 2.3 animals/yr/ peninsula) from moated peninsulas. Five carnivore species and 1 Franklin's ground squirrel were captured on the peninsulas. There were differences (χ² = 9.34, 2 df, P = 0.009) in the proportion of predator species captured at moated and fenced peninsulas. Of the total predators, more raccoons (86%) were removed from moated peninsulas (Z = 2.96, P = 0.003) than from fenced peninsulas (43%). Conversely, more (Z = 2.93, P = 0.003) striped skunks were captured on fenced peninsulas (49%) compared to moated peninsulas (7%). There was no difference (Z = 0.10, P = 0.917) in the proportion of "other predators" (red fox, badger, and mink) captured at either type of peninsula.

Peninsula Cost and Size Considerations

Average cost of constructing 8 fences ($23/m) was less than that of constructing 2 moats ($328/m) (Table 4). Maintenance costs (including trapping and removing predators and making repairs) were similar for both barriers because each was visited regularly during the nesting season. Mean annual cost of fenced peninsulas divided by the mean annual production (net) of fledglings (190 hatched young × 0.54 = 103) resulted in a cost of $12/fledgling. The mean annual cost of moated peninsulas divided by the mean annual production (net) of fledglings (564 hatched young × 0.54 = 305) yielded a cost of $62/fledgling. We found no relationship between either nest density (r = -0.07, 18 df, P = 0.76) or the number of young hatched/ha (r = -0.05, 18 df, P = 0.83) and peninsula area.


Species Composition

The high proportion of gadwalls on managed peninsulas probably reflected the response of this species to successful breeding in habitats with tall nesting cover and numerous nearby wetlands (Lokemoen et al. 1990). In the Dakotas, gadwalls were an important nesting species on islands (34% [Duebbert et al. 1983] 49% [Willms and Crawford 1989]) and in tall cover with predator control (27% [Duebbert and Lokemoen 1980]) and without predator control (24% [Duebbert and Lokemoen 1976]). Lesser scaup may be an important breeding species on peninsulas and islands because they prefer semipermanent and permanent wetlands, where most peninsulas and islands were located, and they are philopatric (Johnson and Grier 1988). Also, lesser scaup may nest more on peninsulas and islands because of the increased opportunity to nest near shorelines (Keith 1961).

For unknown reasons, mallards were not as common as expected on managed peninsulas despite their propensity to respond to successful breeding habitats (Lokemoen et al. 1990). We suggest that blue-winged teal and northern pintail were not numerous on peninsulas because they are not highly philopatric (Johnson and Grier 1988) and usually do not seek tall nesting cover or insular habitats (Lokemoen end Woodward 1992).

Nesting on Managed Peninsulas and Control Peninsulas

Mayfield nest success rates on fenced (Symbol - Mean x = 54%) and moated (Symbol - Mean x = 75%) peninsulas with predator removal were similar to those on nesting islands. Apparent nest success rates in island studies were 49% in prairie Canada (Giroux 1981), 62% in the Dakotas and Montana (Lokemoen and Woodward 1992), and 35% in North Dakota (Willms and Crawford 1989). In comparison, duck nest success rates of 15 and 16% have been reported for North Dakota uplands (Klett et al. 1988, Higgins et al. 1992). Although the nest success rates observed on managed peninsulas were similar to those on islands, nest densities were less than the 6.6-13.8 nests/ha observed by Giroux (1981), Willms and Crawford (1989), and Lokemoen and Woodward (1992). Our peninsulas were nearly 9 times larger than the referenced islands, which probably affected dispersion and density of nesting ducks.


In this experiment, we reduced predation of waterfowl nests on small parcels of land by combining the use of barriers and predator control. The system was effective probably because mammalian predator access to peninsulas was impeded and trapping removed most animals that gained entry. Willms and Crawford (1989) recorded females being killed on islands by mink. Fewer adult females may be killed on managed peninsulas and islands as opposed to mainlands, because mammalian predator numbers are kept low. In this study, more raccoons were captured on moated than on fenced peninsulas probably because raccoons can readily cross water. Striped skunks were more commonly captured on fenced peninsulas because they probably initiated residency at these sites when the fences were not energized.

Cost and Size Considerations

Fenced peninsulas produced ducklings more economically than moated peninsulas primarily because of reduced capital costs. Most electric-fence maintenance was needed on the segment that extended into the water. Managers reduced fence damage in wetlands by using cattle panels. These panels of steel rod (4.9 m long, 1.3 m high) were used to extend the fence into the wetland in spring. The surface of each panel was covered with 2.5-cm poultry netting and an energized wire was placed 6 cm below the top of the panel. Panels were removed each fall before freeze-up.

The cost of each fledged duck from fenced peninsulas was less than the cost of fledged ducks from man-made islands and small rock islands built in the Dakotas (Lokemoen 1984). However, the cost of fledged ducks from fenced peninsulas was similar to that of ducks fledged from nest baskets, but greater than the cost of adding electrified fences to existing nesting cover (Lokemoen 1984).

Small units of habitat have a higher proportion of edge than large units, often resulting in increased productivity and higher densities of vertebrates. In this study there were no differences in nest densities among peninsulas of various sizes, indicating that there is no benefit in managing small peninsulas.

Management Recommendations

Adding predator barriers and predator control to peninsulas is a cost-effective method compared to island construction for increasing duck production in the prairie pothole region. Predator barriers on peninsulas provide primary benefits to gadwalls and lesser scaup and secondary benefits for mallards, blue-winged teal, and other upland-nesting ducks. Electric fences are superior to moats because they are most cost effective.

The dryland portion of peninsula fences should be solidly built to reduce maintenance, and the wetland portion should be removed each fall to reduce damage from ice. All managed peninsulas should be visited regularly to maintain traps, remove predators, and repair structural damage or electrical circuit interruptions at fences.

Fences on peninsulas can be built completely of small wire mesh. These components give peninsular fences at least 3 advantages over 4-sided fences (Greenwood et al. 1990) in that ducklings need not pass through fences to access the wetland, the mesh denies entry to small mammalian predators, and the electric wires are too high to be short-circuited by standing vegetation. At peninsulas with long fences or those with corners, brood departure may be delayed and exits should be added (P. J. Pietz, North. Prairie Wildl. Res. Cent., Jamestown, N.D., pers. commun., 1992).

Managed peninsulas should contain suitable nesting cover for ducks in the form of tall, dense grasses and forbs or low shrubs (< 1.25 m) (Dwernychuk and Boag 1973, Duebbert and Lokemoen 1976). Low shrubs and grass-legume mixtures can be established to provide nesting cover (Duebbert et al. 1981, Snyder 1982).

Predator barriers probably would not be economical on small areas (< 1 ha) because few nests may occur yet management costs would approach those of larger peninsulas. Also, peninsulas in wetlands that are often dry are poor candidates for this type of management.


To enhance duck nesting success in the prairie pothole region of North America, we added predator barriers and predator control to peninsulas in lakes. The study included 8 peninsulas with electrified wire fences and 2 peninsulas with water-filled moats. Fences and moats were constructed at the base of the peninsulas. Success and densities of duck nests found on the peninsulas with predator barriers were compared to the success and densities of duck nests found on similar nearby controls. Average duck nest success of 54% found on fenced peninsulas was higher (P < 0.01) than the 17% on unfenced controls. Moated peninsulas also had higher nest success (75%) than unmoated control peninsulas (14%). Nesting gadwalls and lesser scaup preferred peninsulas with barriers over control sites. Mallards, blue-winged teal, and northern pintail also were common nesters on peninsulas with barriers but did not select them. The cost/fledgling on fenced peninsulas was $12, compared to $62/fledgling on moated peninsulas. The cost of young fledged on fenced peninsulas was intermediate compared to the costs of birds produced by other waterfowl management practices.

Acknowledgments. We thank D. B. Henry, Chief, Division of Refuges and Wildlife for North Dakota, U.S. Fish and Wildlife Service, and many helpful biologists in this Division, including M. W. Goos, H. C. Hultberg, R. F. Johnson, S. A. Knode, M. R. McEnroe, M. J. Rabenberg, R. D. Shupe, and L. D. West. We also appreciate generous help of Ducks Unlimited biologists M. O. Bennett, R. D. Hoffmann, R. L. Meeks, J. R. Singleton, and R. A. Warhurst, and North Dakota Fish and Game Department biologists M. A. Johnson, J. D. Hare, R. W. Renner, and L. J. Schoonover. Statistical treatments and insights by W. E. Newton, T. L. Shaffer, and D. J. Twedt were invaluable. The manuscript benefitted from review by R. E. Kirby, J. E. Austin, and A. B. Sargeant.

Literature Cited

DOWDY, S., and S. WEARDON. 1983. Statistics for research. John Wiley and Sons, Inc., New York, N.Y. 537pp.

DUEBBERT, H. F., and J. T. LOKEMOEN. 1976. Duck nesting in fields of undisturbed grass-legume cover. J. Wildl. Manage. 40:39-49.

_____, and _____. 1980. High duck nesting success in a predator-reduced environment. J. Wildl. Manage. 44:428-437.

_____, E. T. JACOBSON, K. F. HIGGINS, and E. B. PODOLL. 1981. Establishment of seeded grasslands for wildlife habitat in the prairie pothole region. U.S. Fish and Wildl. Serv. Spec. Sci. Rep.-Wildl. 234. 21pp.

_____, J. T. LOKEMOEN, and D. E. SHARP. 1983. Concentrated nesting of mallards and gadwalls on Miller Lake Island, North Dakota. J. Wildl. Manage. 47:729-740.

DWERNYCHUK, L. W., and D. A. BOAG. 1973. Effect of herbicide-induced changes in vegetation on nesting ducks. Can. Field-Nat. 87:155-165.

GIROUX, J.-F. 1981. Use of artificial islands by nesting waterfowl in southeastern Alberta. J. Wildl. Manage. 45:669-679.

GREENWOOD, R. J., P. M. ARNOLD, and B. G. MCGUIRE, 1990. Protecting duck nests from mammalian predators with fences, traps, and a toxicant. Wildl. Soc. Bull. 18:75-82.

HIGGINS, K. F., L. M. KIRSCH, A. T. KLETT, and H. W. MILLER. 1992. Waterfowl production on the Woodworth Station in south-central North Dakota, 1965-1981. U.S. Fish and Wildl. Serv. Resour. Publ. 181. 88pp.

JOHNSON, D. H. 1979. Estimating nest success: the Mayfield method and an alternative. Auk 96:651-661.

_____, and J. W. GRIER. 1988. Determinants of breeding distributions of ducks. Wildl. Monogr. 100. 37pp.

KEITH, L. B. 1961. A study of waterfowl ecology on small impoundments in southeastern Alberta. Wildl. Monogr. 6. 88pp.

KLETT, A. T., H. F. DUEBBERT, C. A. FAANES, and K. F HIGGINS. 1986. Techniques for studying nest success of ducks in upland habitats in the prairie pothole region. U.S. Fish and Wildl. Serv. Resour. Publ. 158. 24pp.

_____, T. L. SHAFFER, and D. H. JOHNSON. 1988. Duck nest success in the prairie pothole region. J. Wildl. Manage. 52:431-440.

LOKEMOEN, J. T. 1984. Examining economic efficiency of management practices that enhance waterfowl production. Trans. North Am. Wildl. and Nat. Resour. Conf. 49:584-607.

_____, and R. O. WOODWARD. 1992. Nesting waterfowl and water birds on natural islands in the Dakotas and Montana. Wildl. Soc. Bull. 20:163-171.

_____, H. A. DOTY, D. E. SHARP, and J. E. NEAVILLE. 1982 Electric fences to reduce mammalian predation on waterfowl nests. Wildl. Soc. Bull. 10:318-323.

_____, H. F. DUEBBERT, and D. E. SHARP. 1990. Homing and reproductive habits of mallards, gadwalls, and blue-winged teal. Wildl. Monogr. 106. 28pp.

MILLIKEN, G. A., and D. E. JOHNSON. 1984. Analysis of messy data. Vol. 1: designed experiments. Van Nostrand Reinhold, New York, N.Y. 473pp.

NETER, J., W. WASSERMAN, and G. A. WHITMORE. 1982. Applied statistics, second ed. Allyn and Bacon, Inc., Boston, Mass. 773pp.

SARGEANT, A. B., and P. M. ARNOLD. 1984. Predator management for ducks on waterfowl production areas in the northern plains. Proc. Vertebr. Pest Conf. 11:161-167.

SAS INSTITUTE INC. 1987. SAS/STAT guide for personal computers. Version 6 ed. SAS Inst. Inc., Cary, N.C. 1,028pp.

SNYDER, W. B. 1982. Minimum tillage techniques for establishing shrubs in clump plantings. Colo. Div. Wildl. Spec. Rep. 53., Denver. 17pp.

STEEL, R. G. D., and J. H. TORRIE. 1980. Principles and procedures of statistics: a biometrical approach. Second ed. McGraw-Hill Book Co., New York, N.Y. 633pp.

STEWART, R. E., and H. A. KANTRUD. 1971. Classification of natural ponds and lakes in the glaciated prairie region. U.S. Fish and Wildl. Serv. Resour. Publ. 92. 57pp.

U.S. FISH and WILDLIFE SERVICE. 1990. Waterfowl breeding population survey for South Dakota and North Dakota. Migr. Bird Manage. Off., Laurel, Md. 8pp.

WILLMS, M. A., and R. D. CRAWFORD. 1989. Use of earthen islands by nesting ducks in North Dakota. J. Wildl. Manage. 53:411-417.

This resource is based on the following source (Northern Prairie Publication 0877):

Lokemoen, John T., and Robert O. Woodward. 1993. An assessment of predator barriers and predator control to enhance duck nest success on peninsulas. Wildlife Society Bulletin 21:275-282.

This resource should be cited as:

Lokemoen, John T., and Robert O. Woodward. 1993. An assessment of predator barriers and predator control to enhance duck nest success on peninsulas. Wildlife Society Bulletin 21:275-282. Jamestown, ND: Northern Prairie Wildlife Research Center Online. (Version 13JUN2001).

John T. Lokemoen and Robert O. Woodward, Northern Prairie Wildlife Research Center, U.S. Fish and Wildlife Service, Jamestown, ND 58401

Downloading Instructions -- Instructions on downloading and extracting files from this site.
(Download) ( 17K ) -- An Assessment of Predator Barriers and Predator Control to Enhance Duck Nest Success on Peninsulas
Installation: Extract all files and open index.htm in a web browser.

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: Friday, 01-Feb-2013 19:49:55 EST
Menlo Park, CA [caww54]