Prairie Duck Populations and Predation Management

Raymond J. Greenwood and Marsha A. Sovada

Table of Contents

• Introduction
• Recovery of Duck Populations
• Strategies for Managing Predation
• Future Predation Management in the Prairie Pothole Region
• Acknowledgments
• References

Introduction

The purpose of our paper is to discuss strategies for managing mammalian predation on prairie-nesting ducks and issues likely to affect predation management in coming years. Predation management is a complex, often emotional subject for which there are no easy solutions or pat answers (Berryman 1972). Because knowledge of the functioning of natural systems can help direct management efforts, we first want to briefly examine some events that we think may have influenced the recent recovery of duck populations. Near the end of the last decade breeding populations of many North American prairie-nesting ducks were greatly depressed; estimates for mallards (Anas platyrhynchos), blue-winged teal (A. discors), and northern pintails (A. acuta) were at or near record low levels. By 1995, however, the estimated total breeding population of ducks in North America was the highest since 1980 and 11% above the long-term (1955-94) average (U.S. Fish and Wildl. Serv. and Can. Wildl. Serv. 1995).

A primary factor affecting populations of prairie-nesting ducks is low recruitment due to predation on nesting females (Johnson and Sargeant 1977, Sargeant et al. 1984), eggs (Cowardin et al. 1985, Klett et al. 1988, Greenwood et al. 1995), and ducklings (Talent et al. 1983, Cowardin et al. 1985, Orthmeyer and Ball 1990). Factors of predation on the breeding grounds are related to habitat conditions (Cowardin et al. 1985, Greenwood et al. 1995) and make-up of the predator community (Greenwood 1986, Johnson et al. 1989, Sovada et al. 1995). Species that nest early in the spring are especially vulnerable to predation (Johnson and Sargeant 1977, Blohm et al. 1987, Greenwood et al. 1995). Risk to breeding females from predation is greatest during years of good water conditions when the nesting period is prolonged.

The Prairie Pothole Region (PPR) of the northern Great Plains is the primary breeding ground for ducks in North America (Batt et al. 1989). Most upland areas suitable for farming in the PPR have been converted to cropland (Lynch et al. 1963, Boyd 1985, Sugden and Beyersbergen 1984, Higgins 1977). In many areas, the only remaining upland habitat for nesting ducks is along roadsides, fence rows, and other isolated small areas of undisturbed cover (Greenwood et al. 1995, Higgins 1977, Sugden and Beyersbergen 1984).

Recovery of Duck Populations

Four events seem to have influenced the recent recovery of duck populations.

Reduced predator control. In the mid-1970's, coyote (Canis latrans) populations in the PPR began to rebound following years of suppression (Johnson and Sargeant 1977). Rebound followed the 1972 ban of toxicants to control predator-caused damages on federal lands or through federally-funded programs (Nixon 1972), restrictions on aerial hunting, and easing of other control programs. Although the coyote and red fox (Vulpes vulpes) have always occupied the PPR (Hall and Kelson 1959), in sympatric populations, coyotes exclude red foxes from large areas (Sargeant et al. 1987, Voigt and Earle 1983). Both species prey on nesting ducks, but because foxes are particularly severe predators of upland-nesting ducks and eggs (Sargeant et al. 1984, Johnson et al. 1989), a coyote-dominated predator community is more favorable to nesting ducks than one dominated by red foxes (Sovada et al. 1995). Although coyotes were scarce throughout much of the PPR in the late 1970's (Johnson and Sargeant 1977), they occupied much of the PPR in North Dakota and South Dakota by the early 1990's (Sovada et al. 1995). Security provided to coyotes by large expanses of land idled through the Conservation Reserve Program (CRP) enacted by the U.S. Department of Agriculture (Young and Osborn 1990) and reduced harvest resulting from low pelt prices probably enhanced recent expansion of coyote populations.

Drought of 1980's. A prolonged period of drought occurred throughout much of the PPR from the mid-1980's into the early 1990's. Drought affected the Canadian portion of the PPR first (Greenwood et al. 1995) and later the U.S. portion (U.S. Fish and Wildl. Serv. and Can. Wildl. Serv. 1995, Todhunter 1995). An obvious effect of drought was the precipitous decline in breeding ducks (U.S. Fish and Wildl. Serv. and Can. Wildl. Serv. 1995). Less obvious was the decline in mink (Mustela vison) populations under drought conditions (Sargeant et al. 1993). Mink are serious predators of over-water nesting ducks and ducklings (Eberhardt 1973, Eberhardt and Sargeant 1977, Arnold and Fritzell 1987). Drought reduces reproduction of mink (Eberhardt 1974) and has catastrophic effects on mink populations (Sargeant et al. 1993). In the early 1990's, mink populations were extremely low in much of the PPR.

CRP enacted. Beginning in 1985, millions of hectares of cropland were restored to perennial grassland under the CRP (Young and Osborn 1990). In North Dakota, South Dakota, Minnesota, and Montana, this amounted to >3 million hectares (Jones and Kruse 1995), of which about 2.5 million were in the PPR (Kantrud 1993). CRP fields are attractive to ducks and can be highly productive (Kantrud 1993, Reynolds et al. 1994). However, during the mid-1980's, potential of CRP fields for increasing recruitment of ducks was greatly diminished due to drought (Kantrud 1993). CRP fields also are attractive to and provide productive nesting habitat for a variety of nongame bird species (Johnson and Schwartz 1993a, 1993b).

Drought ended. During 1993-95, conditions for breeding ducks may have been the best ever observed in some areas of the PPR (Krapu 1994, U.S. Fish and Wildl. Serv. and Can. Wildl. Serv. 1995). Nesting cover was abundant in the U.S. portion of the PPR due to CRP. In North Dakota, beginning in June 1993, precipitation was above long-term averages for 14 of 19 consecutive months (Nat. Oceanic and Atmos. Admin. 1993, 1994). After years of drought, numerous ponds contained abundant detritus that provides optimum conditions for production of benthic macro-invertebrates (Neckles et al. 1990) needed by laying females and ducklings (Swanson and Duebbert 1989). Red fox numbers were low, favoring high nest success, and mink numbers were low, favoring high rates of duckling survival. These are conditions that lead to production of "Super-Crops" of ducks (Lynch 1984); however, such conditions seldom are synchronized for more than a few years at a time. Lynch et al. (1963) suggested that where habitats are altered by agriculture, high duck production is likely in only 2-3 years out of 10.

Strategies for Managing Predation

The CRP replaced cropland with grassland in much of the United States portion of the PPR, but when enrollment periods begin to expire in 1996, prairie landscapes again will change markedly. Change is also inevitable in predator populations. So the question arises again, "What can be done to reduce effects of predation on duck production in the PPR and how should management be applied for maximum effect?"

There are nine mammals and seven birds that commonly prey on ducks in the PPR (Sargeant et al. 1993). A variety of methods have been tried or suggested for reducing effects of some of these predators on ducks (Sargeant and Arnold 1984). Most methods protect only the nesting hen and eggs from mammalian predators; few target avian predators or are designed to enhance survival of ducklings. Most methods affect duck production on a local scale; few methods have potential to affect duck populations at the landscape scale. Methods can be grouped into those that (1) physically separate or conceal prey from predator; (2) alter predator foraging behavior or food supply; or (3) affect predator distribution or abundance, including both lethal and non-lethal methods.

Separate or conceal prey

Management of grassland habitat to provide dense nesting cover (Duebbert et al. 1981) is practiced under the assumption that heavy cover will repel predators and prevent nests from being detected. Although dense cover is attractive to ducks and several studies have reported high nest success in this habitat (Schranck 1972; Duebbert and Lokemoen 1976; Kirsch et al. 1978; Livezey 1981; Cowardin et al. 1985; Sugden and Beyersbergen 1986, 1987), a review by Clark and Nudds (1991) was inconclusive on reported benefits of dense nesting cover. Clark and Nudds (1991) suggested that concealment of nests was important to survival of duck nests in the PPR only when predation was predominantly by birds. Use of dense nesting cover is appealing for various, especially economic reasons (Lokemoen 1984), but use of this practice alone in local areas probably is of limited value in reducing predation on nesting ducks where mammals are primary predators. Planting of dense vegetation actually may be counterproductive where Franklin's ground squirrel (Spermophilus franklinii) occur because this predator prefers dense cover (Choromanski-Norris et al. 1989).

Physical barriers, including moats constructed across peninsulas (Lokemoen and Messmer 1994, Lokemoen and Woodward 1993) and nesting islands (Jones 1975, Giroux 1981, Higgins 1986, Lokemoen and Messmer 1993) also are used to protect nests. Fences, usually electrified, are used as barriers across peninsulas (Lokemoen and Woodward 1993), around tracts of nesting habitat in uplands (Lokemoen et al. 1982, Greenwood et al. 1990, Gatti et al. 1992), and around individual nests (Sargeant et al. 1974). Barriers can be effective in preventing destruction of nests by mid-sized and larger mammalian predators, especially when used in conjunction with trapping or a toxicant to keep predators from the protected area during the nesting season (Greenwood et al. 1990).

Barriers are not effective with all mammalian predators. Semi-aquatic species like mink and raccoons (Procyon lotor) are not deterred by water barriers, and Franklin's ground squirrels, mink, and weasels (Mustela erminea and M. frenata) can easily penetrate fences. In addition, mortality sometimes occurs among ducklings that cannot exit through mesh fences (Sargeant et al. 1974, Pietz and Krapu 1994, Trottier et al. 1994). Electric fences are one of the most cost-effective methods of increasing nest success of ducks in local areas; creation of water barriers by excavating dikes or constructing islands is one of the most costly methods (Lokemoen 1984).

Another method of separating nests from mammalian predators is with elevated structures (Zenner et al. 1992), including nest baskets (Bishop and Barratt 1970, Doty et al. 1975) and earth-filled culverts (Higgins et al. 1986, Ball and Ball 1991). The mallard is the only duck that readily uses nest baskets (Bishop and Barratt 1970, Doty et al. 1975). Although elevated structures are relatively cost-effective (Lokemoen 1984), each nest basket generally protects only one nest. Single earth-filled culverts may simultaneously contain nests of both mallards and Canada geese (Branta canadensis) (Ball and Ball 1991). Raccoons sometimes climb support poles, and avian predators may destroy eggs in nest baskets unless overhead covering is provided (Doty 1979). Methods that separate or conceal prey from predators function mostly at the local scale, although individual applications can be aggregated regionally to reduce administrative costs, and possibly increase effectiveness.

Alter foraging or foods

Conditioned taste aversion (CTA) employs a natural biological process to alter the foraging behavior of predators (Nicolaus et al. 1982, 1983; Nicolaus 1987; Conover 1989, 1990; Dimmick and Nicolaus 1990). To protect eggs from predation, individual predators must first develop an aversion to chemically treated eggs placed where predators will encounter and consume them. Aversion then must be generalized to untreated eggs in naturally-occurring nests. The aversion must translate to visual, olfactory, or other externally provided cues, because if consumption or disturbance of eggs is required to elicit the aversion, the method is of little value in protecting eggs in naturally-occurring nests.

Several compounds have potential for eliciting a CTA, but some are toxic and the response to compounds differs among species (Conover 1989, 1990). Thus, it is unlikely that a single compound will be effective for any major group of egg-eating predator in the PPR. CTA may be especially difficult to establish with red foxes. Foxes seldom eat eggs immediately upon encounter, but rather cache eggs for future consumption (Kruuk 1964, Tinbergen 1965). Tests with American crows (Corvus brachyrhynchos) have shown encouraging results and suggest that individual crows may avoid consumption of eggs for a considerable time after being conditioned to avoid them (Dimmick and Nicolaus 1990). Search for potentially usable compounds and high cost of evaluating each compound with individual predator species is likely to be a deterrent to screening and testing of a variety aversive agents. For compounds that may be consumed by non-target species, registration may be difficult because of food and drug regulations. Unless it is possible to develop a CTA simultaneously in several species of predators inhabiting an area, compensatory predation by animals that have no aversion to consuming eggs may reduce success of this method.

Manipulating availability of food has been suggested to reduce consumption of eggs (Crabtree and Wolfe 1988). Theory behind this approach is based mostly on findings from Arctic and sub-Arctic ecosystems, where there are fewer species of predators and prey than in the PPR. In those ecosystems, abundance of alternate food for predators has been found to be inversely related to predation rates on birds and eggs (Larson 1960, McInvaille and Keith 1974, Pehrsson 1986, Summers 1986, Beintema and Müskens 1987). Byers (1974) suggested that nest success of ducks in the PPR was higher during years when indices to small rodent numbers were highest. Crabtree and Wolfe (1988) suggested that provision of supplemental foods for striped skunks reduced predation on duck nests. However, a study similar to Crabtree and Wolfe (1988) conducted in central North Dakota during 1993-94 failed to confirm the benefits to nesting ducks of providing supplemental foods for striped skunks during the nesting season (Unpublished data, D. G. Pietruszewski and R. J. Greenwood: 1994).

Lethal methods

Lethal control of predators to increase nest success or productivity of ducks has shown variable results; efficacy depends on method used and size of area treated (Balser et al. 1968, Duebbert and Kantrud 1974, Duebbert and Lokemoen 1980, Greenwood 1986, Doty and Rondeau 1987, Sargeant et al. 1995). Removal of individual species of predators (Greenwood 1986) or several species of predators (Sargeant et al. 1995) during the duck nesting season from small (<500 ha) areas managed for nesting ducks may result in moderate increases in nest success; however, variability in nest success among treated areas can be high due to compensatory predation by unaffected predators. Restrictions on methods permitted for use and other factors (e.g., inflexible work schedule of field personnel) may impact success of removal efforts (Sargeant et al. 1995). Removal of several species of predators with unlimited techniques from relatively large (<26,000 ha) areas, either during the duck nesting season (Balser et al. 1968, Schranck 1972, Garrettson et al. 1996) or year round (Duebbert and Kantrud 1974, Duebbert and Lokemoen 1980), can result in high duck nest success and production of broods. Lethal control can be quite cost effective (Lokemoen 1984), especially when toxicants that require no daily maintenance are used. Depending on the methods used, lethal control can affect duck production over large areas of landscape.

Although effective, lethal control is unpopular today and often is opposed by the public, even to protect endangered species. Issues raised involve humane treatment of animals, moral issues, and anti-hunting and anti-trapping sentiment (Kirkpatrick and Turner 1985). Other issues also affect use of lethal control. For toxicants, costs of registration are extremely high and likely will detract from development of new compounds. Furbearer managers and fur harvesters may object to reduction of valued furbearers. To be effective in managing predation on nesting birds, predators must be removed just before and during the nesting season; because pelts have little value in spring, there is little incentive for fur harvesters to participate in control efforts at that time.

Non-lethal methods

Inhibiting reproduction is an appealing non-lethal approach to population management of wild animals that has been little evaluated (Allen 1982, Kirkpatrick and Turner 1985, Warren 1995). This method was proposed as early as the 1960's for managing predation by carnivores (Balser 1964, Linhart and Enders 1964), but its use in North America today is primarily with ungulates (Garrott 1995). With carnivores, inhibition of reproduction is suggested to reduce predation by reducing food demand of females that have no young to support and by reducing recruitment, thereby decreasing population size. Perhaps the most compelling reasons for using reproductive inhibitors to control predator populations are social; the approach is humane and likely to be accepted by the public (Kirkpatrick and Turner 1985). However, those authors questioned whether the cure is worse (i.e., unforeseen side effects) than the problem and suggested that extensive preliminary studies are necessary, species by species, to assess treatment effects before field application is attempted. In the PPR, this could include evaluation of effects on at least eight carnivores with variable breeding times.

An important consideration in use of reproductive inhibitors is the proportion of a population that remains reproductively active after treatment (Garrott 1995). This is influenced by the proportion of animals detected, the proportion detected that is treated, and the efficacy of the agent. Ideally a population to be managed should be isolated geographically to limit immigration and emigration (Garrott 1995). There are few isolating mechanisms among predator populations in the PPR. In addition, reproductive and/or dispersal characteristics of some carnivores in the PPR (e.g., red foxes [Allen 1983], raccoons [Fritzell 1978]; striped skunks [Greenwood and Sargeant 1994]) permit rapid recovery from population depletion.

Among canids, compounds tested or suggested for inhibiting reproduction range from steroids to, more recently, immunocontraceptive vaccines (Kirkpatrick and Turner 1985). Vaccines offer short-term reversible contraception suitable for rare or highly-valued species, but may not be adequate where long-term population control is required or where the goal is dramatic reduction of the population (Garrott 1995). Where long-term control is desired, a technique that results in permanent sterility is more appropriate. However, widespread use of treated baits that may result in permanent sterility is unlikely to be popular with the public, especially where domestic animals may be exposed. Another non-lethal approach to managing canid populations utilizes interspecific interactions among species. Red foxes avoid areas occupied by coyotes (Voigt and Earle 1983, Sargeant et al. 1987) and this interaction can influence productivity of nesting ducks (Johnson et al. 1989, Greenwood et al. 1995, Sovada et al. 1995). At low population levels, coyotes are thought to be less detrimental to nesting ducks than are red foxes (Sovada et al. 1995). Coyotes also may influence presence of raccoons, and badgers may influence presence of striped skunks (Johnson et al. 1989). Management of carnivore communities to favor coyotes may benefit nesting ducks over large areas of landscape, but also can create other conflicts. Coyotes prey on livestock and wild ungulates, especially young. Conflicts in objectives may thus ensue even within a game management agency, because management goals may differ among individuals or groups.

Future Predation Management in the Prairie Pothole Region

Wildlife professionals in rural areas continue to be faced with the challenge of reducing overabundant populations of wild animals; with increasing frequency, managers in urban and suburban settings face a similar problem (Warren 1995). Traditional methods of managing predator populations (e.g., hunting, trapping, poisoning) may not be appropriate or permitted in many areas, but few new approaches are available. Low fur prices due to instability in world markets have reduced, at least temporarily, incentives for public harvest of furbearing species. Changing human demographics have reduced numbers of fur harvesters in rural areas. Thus, managers in the PPR likely will get little help from the public in managing populations of mammalian predators, at least in the near future. This fact and fiscal austerity will require managers to focus efforts to enhance duck production where returns will be greatest.

The PPR is a dynamic area where factors that affect duck production, especially condition of wetland habitats, change continually. In this unstable environment, ducks have become opportunists in settling breeding habitat (Johnson and Grier 1988). Historically, ducks in the PPR had numerous options for breeding areas. Presently, however, in much of the PPR, options for breeding ducks are limited due to conversion of wetland and upland habitats to cropland.

Reynolds et al. (1996) discussed methods for identifying landscapes with greatest management potential for duck production, based on abundance of wetlands. We believe identification efforts also should include abundance of grassland and should include the entire PPR, not just the U.S. portion. Nest success in the PPR is correlated with amount of perennial nesting cover in the landscape (Greenwood 1987). Knowledge of distribution of both wetland and grassland habitats at the landscape scale would enable managers to better target areas for treatment. In addition, this information would provide a much needed framework for prioritizing areas for protection of threatened habitats.

We believe managers should conduct some type of periodic assessment at the landscape scale of predator populations, at least canids. Make-up of the predator community affects duck production (Johnson and Sargeant 1977, Greenwood et al. 1995, Sovada 1995); knowledge about the predator community is important for decisions about management to enhance duck production. Although changes in predator populations usually occur gradually (Sargeant et al. 1993), factors such as epizootics may result in rapid changes. Allen and Sargeant (1975) provided guidelines for an inexpensive and reliable method for assessing distribution and abundance of red foxes in North Dakota by rural mail-carrier surveys. Queries of farm operators and rural residents could provide additional information on canid distribution.

We believe that plans to enhance duck production, including predation management, should be based on long-term goals and be tailored to specific local and regional needs; a variety of approaches probably will be necessary. Management applied at the landscape scale to benefit duck populations, especially habitat enhancement (e.g., planting of nesting cover), is likely to be most effective when conducted in areas where nest success is highest (Greenwood et al. 1995). In areas where predation is severe and management to enhance duck production is mandated because of ownership or other reasons, more intensive management (e.g., use of barriers) of local areas may be the best choice. However, some types of intensive management (e.g., predator removal from small tracts) applied at the local scale may not be cost-effective when used where predation rates are high (Sargeant et al. 1995).

Evaluation is an important component of management. If the management goal is to enhance duck production, evaluation should focus on changes in duck recruitment. Other factors (e.g., number of predators removed, amount of nesting cover planted, number of wetlands restored) are easy and appealing to document, but may have little relation to actual number of ducks recruited (Sargeant et al. 1995).

Besides management based on long-term goals, we believe managers should be prepared to respond to events (e.g., excess precipitation, epizootics in predators) that temporarily may favor high duck production in certain areas. Such unpredictable events are characteristic of the PPR and offer unique opportunities for management, especially when they occur in the grassland portion of the PPR that is known for its high potential to produce ducks (Lynch 1984). As conditions change and become less favorable for duck production, emphasis can be redirected. In this manner, the dynamic nature of the PPR can be exploited as conditions dictate.

For example, an outbreak of sarcoptic mange (Sarcoptes scabiei) presently is reducing coyote and red fox populations in North Dakota and Minnesota (S. H. Allen, personal communication: 1996). This epizootic has potential to affect canid populations for several years (Pence and Windberg 1994). Where areas with depleted canid populations overlap with identified areas of high potential for duck production, temporary action (e.g., short-term contracts to restore wetlands; planting of nesting cover) might be used to attract breeding ducks. Simultaneous effort to reduce impacts of non-canid species in these areas might provide additional benefits to ducks.

We believe there is little rationale to expend resources for management beyond normal operations in areas where wetland conditions are poor for breeding ducks. Rather, we believe it makes more sense to consolidate resources from areas with poor wetland conditions for use in areas that have the greatest immediate potential (including favorable predator community) for breeding ducks. Consolidating resources will require administrators to be flexible, respond quickly, and focus on the entire PPR instead of local or regional aspects. The North American Waterfowl Management Plan (Can. Wildl. Serv. and U.S. Fish and Wildl. Serv. 1986) provides a mechanism for cooperation and pooling of resources among management agencies. The challenge for future managers will be to focus their efforts to enhance duck production where benefits are greatest while embracing the economic, ecological, and ethical dimensions of predation management.