Electric fences to reduce mammalian predation on waterfowl nests

We evaluated electric fences as predator barriers to reduce high losses of waterfowl nests to mammalian predation at Waterfowl Production Areas (WPAs). The work was done in 1978-81 on 3 paired sites in central North Dakota and western Minnesota. Resident mammalian predators were trapped from inside the exclosures. All 3 fences operated during the study period with few major maintenance problems. Nest success in the enclosures was 65% in North Dakota and 55% in Minnesota vs. 45 and 12% in the respective controls. Cover inside the electric fence produced 7.8 more young/ ha than cover in control plots in North Dakota during the 3 years. Cover inside the 2 electric fences in Minnesota yielded 9.5 and 4.3 more young/ha than cover in control plots during the 3 years. Using construction costs only we estimated that each additional duckling produced in cover protected by electric fencing cost $0.65 in North Dakota and $0.87 in Minnesota. WILDL. SOC. BULL. 10:318-323 Recently, waterfowl production in the Prairie Pothole Region of eastern North Dakota and western Minnesota has been limited by severe predation on nests. Cowardin and Johnson (1979) summarized nesting studies accomplished on unmanaged habitats in North Dakota and estimated that the adult female mallard (Anas platyrhynchos) population was declining by about 2% annually. D. Johnson (U.S. Fish and Wildl. Serv., Jamestown, ND, unpubl. data) reported waterfowl nest success of 16.2% in eastern North Dakota on 13 WPAs. Success of 113 nests located on 8 WPAs in western Minnesota in 1980 was 8% (H. Doty, U.S. Fish and Wildl. Serv., Fergus Falls, MN, unpubl. data). With a nest success of 16-17% needed to maintain population stability (Cowardin and Johnson 1979) it was obvious that waterfowl recruitment in these eastern prairie localities was insufficient. Egg losses in the above studies were attributed mainly to small carnivores including the striped skunk (Mephitis mephitis), red fox (Vulpes vulpes), raccoon (Procyon lotor), and badger (Taxidea taxus). Also, red fox on the 318 average annually kill an estimated 18% of the incubating mallard hens in the Prairie Pothole Region of North Dakota (Johnson and Sargeant 1977). Where suitable wetlands and nesting cover occur in conjunction with low predator populations, high waterfowl production has been reported. In northwestern Minnesota Balser et al. (1968) concluded that 60% more Class I ducklings (Gollop and Marshall 1954) were produced on areas where predators were controlled. In North Dakota on islands free from mammalian predation, Hammond and Mann (1956) and Duebbert (1966) found high nest densities and success. In pothole habitat in South Dakota Duebbert and Lokemoen (1980) noted that eggs hatched in 90% of 842 nests in dense nesting cover when predator control was in effect. Breeding mallards responded positively, and nest densities increased from 0.7 to 3.5/ha. Pairs increased 496% when predator reduction was most effective. Waterfowl would benefit notably by reduced nest predation, but predator removal programs have not been practical because of This content downloaded from 207.46.13.85 on Fri, 18 Nov 2016 04:26:58 UTC All use subject to http://about.jstor.org/terms ELECTRIC FENCES PROTECT WATERFOWL NESTS * Lokemoen et al. 319 biological, social, and economic concerns. As a possible solution to the predation dilemma we searched for techniques that would be primarily non-lethal and could be applied to managed situations at reasonable costs. Specifically, we sought to exclude predators from planted nesting coverts by some type of barrier. The most worthy possibility for trial was a newly designed electric fence system which uses high powered, low impedance energizers developed in Australia and New Zealand (Nesbitt 1978). These new energizers allow the charging of long fence lines effectively with electrical power sources. The use of electric fencing in wildlife management is not new. McAtee (1939) described various electric fencing applications > 40 years ago. Recently, electric fences were used with some success to exclude red foxes from a nesting colony of sandwich terns (Sterna sandvicensis) (Forster 1975). Sargeant et al. (1974) used small fences to keep carnivores from individual sharp-tailed grouse (Pedioecetes phasianellus) and waterfowl nests in North Dakota. The objectives of this study were to determine (1) if mammalian-caused nest losses would be reduced in nesting cover encircled by electric fences, and (2) if electric fences would operate effectively under field conditions with minimum maintenance. We thank R. J. Greenwood who reviewed the manuscript and appreciate the early encouragement of the project by H. F. Duebbert and J. R. Serie. We are grateful to W. R. Goforth, former director of Northern Prairie Wildlife Research Center, who supported the North Dakota study phase and C. R. Madsen, project leader of the Mid-Continent Waterfowl Management Unit Project, who supported to the Minnesota trials. We are grateful for the help provided by the staff at the Audubon National Wildlife Refuge, particularly B. Gastineau. We appreciate the work of V. Gums, A. Rondeau, and R. Naze who built the initial fences. STUDY AREAS AND METHODS Electric fencing tests were conducted at 3 paired study sites in the eastern Prairie Pothole Region of the U.S. in 1978-81. One study site was on the Gaub WPA in McLean County, North Dakota, near Mercer. The other 2 sites were located on the Stowe Lake WPA in Douglas County, Minnesota, near Alexandria. Land use around the Gaub WPA was 55% in small grains, 30% pasture, 10% wetland, and 5% miscellaneous habitats. Land use around the Stowe Lake study site was 55% in row crops, 12% hay and pasture, 8% woodlands, 20% lakes and wetlands, and 5% miscellaneous habitats. The Gaub WPA included a 33-ha field of seeded nesting cover adjacent to a 20-ha semipermanent marsh (Stewart and Kantrud 1971). The vegetation was primarily intermediate wheatgrass (Agropyron intermedium) and alfalfa (Medicago sativa). One 8.6-ha portion of cover was enclosed by an electric fence (NDF), and an adjacent 8.6-ha portion of cover served as the control (ND-C). The Minnesota study sites were placed in a block of cover just south of the 219-ha Stowe Lake. One pair of study plots contained the same species of vegetation found in the North Dakota site with the addition of tall fescue (Festuca arundinacea). The exclosure (MN-Fl) and the adjacent control (MN-Cl) in this cover type each contained 10 ha. The 3rd pair of study sites were in old field vegetation comprised primarily of smooth bromegrass (Bromus inermis) and quackgrass (A. repens). Both the exclosure (MN-F2) and the nearby control (MN-C2) were 7 ha. Fences were constructed with 7 strands of 12.5 gauge smooth wire. The wire spacing on all fences was approximately 10, 20, 36, 56, 76, 102, and 127 cm above ground level (Fig. 1). Wires 1, 2, 3, 5, and 7 above ground level were charged, and wires 4 and 6 were grounded. All corners were anchored wood posts. At ND-F a fiber glass post was placed every 16 m. Stronger fences were considered necessary in Minnesota where wood posts and fiber glass poles were driven alternately about every 11 m. A soil sterilant "Pramitol 25E" was sprayed on a 2 m wide strip of ground centered on the fence to prevent vegetative growth. Gallagher E-12' energizers (Gallagher Electronics Ltd., Hamilton, New Zealand), each powered by a 12-V battery, were used as the power source for the fences. The fences were usually electrified in April or early May. Nest searches were usually started a week or more after fences were activated; they were terminated in July. In North Dakota nest searches were conducted every 7 days using the cable-chain drag (Higgins et al. 1969). In Minnesota, nest searches were conducted about every 17 days using a weighted rope in 1979 and a heavy chain pulled between 2 vehicles I Reference to trade names does not imply government endorsement of commercial products. This content downloaded from 207.46.13.85 on Fri, 18 Nov 2016 04:26:58 UTC All use subject to http://about.jstor.org/terms 320 Wildl. Soc. Bull. 10(4) 1982