USGS Home
Contact USGS
Search USGS

A. Ducks in the Prairie-parkland Region

The Mallard is the most-studied duck species, and the prairie-parkland region is where it has been studied most intensively. Yet, only recently have sufficient data become available to begin to assess the effects of mortality during the breeding season on recruitment and population levels of this important game species. Work on this subject was stimulated by: (1) Anderson and Burnham's (1976) examination of the compensatory relationships between hunting and nonhunting mortality, (2) application of the Mayfield method to studies of nest success that showed success was much lower than previously thought (Miller and Johnson 1978), (3) discovery of the high rate of mortality of hens from predation (Johnson and Sargeant 1977), and (4) the decline of the Mallard population (Trauger and Stoudt 1978, Canadian Wildlife Service and U.S. Fish and Wildlife Service 1986, Johnson and Shaffer 1987).

Johnson and Sargeant (1977) incorporated data on the incidence of Mallards found at red fox dens in North Dakota into a simulation model to evaluate the effects of fox predation (including scavenging) on the sex ratio of breeding Mallards. They estimated that about 45% of adult female Mallards died each year from all causes combined, including hunting, and that 68% of the annual mortality occurred during spring and summer, mostly during the breeding season. They concluded that predation by red foxes was the major source of female mortality during spring-summer and that the foxes took about 18% of the adult female and 5% of the adult male Mallards residing in North Dakota each year. This level of predation was sufficient to cause a disparate sex ratio of 118 males per 100 females, even though hunters killed a greater proportion of males than females during fall and winter. Subsequent studies have confirmed that the spring-summer mortality rate of adult female Mallards in the prairie-parkland region is high (Table 12-1).

Sargeant et al. (1984) showed that, relative to mortality rates of Mallards, Northern Pintails were more vulnerable to predation by foxes, Northern Shovelers were about as vulnerable, and both Blue-winged Teal and Gadwalls were less vulnerable. The differences related largely to differences in nesting chronologically of those species. These authors estimated that predation by foxes accounted for an average annual loss of over 800,000 adult ducks (nearly all dabbling ducks, of which about 75% were females) from the prairie-parkland region.

Nest mortality is the most important factor affecting duck recruitment rates in the prairie-parkland region (Cowardin and Johnson 1979). Numerous studies have reported very low nest success for dabbling ducks in the region (Table 12-3). Recent analysis of nest success records from numerous studies in Minnesota, North Dakota, and South Dakota (corrected for known biases including unequal searching effort in different habitats) suggested that annual nest success of Mallards averaged 9% (Klett et al. 1988). A recent study of duck nest success on 17 study areas scattered throughout the prairie-parkland region of Alberta, Manitoba, and Saskatchewan (also corrected for known biases) indicated that Mallard nest success averaged 12% (Canadian Wildlife Service and U.S. Fish and Wildlife Service 1987). Both Canadian Wildlife Service and U.S. Fish and Wildlife Service (1987) and Klett et al. (1988) found that nest success of Northern Pintails was similar to that of Mallards, whereas nest success of later-nesting Blue-winged Teal, Northern Shovelers, and Gadwalls ranged from 12% to 18%. Although average nest success rates were low in both studies, other investigators showed that dabbling ducks at certain sites, such as islands and other areas where there were few predators, generally had much higher than average nest success, up to 93% (e.g., Duebbert 1966; Vermeer 1968; Duebbert 1982; Duebbert et al. 1983; Cowardin et al. 1985, for data for Lostwood National Wildlife Refuge; Table 12-3). Also, there is evidence that current agricultural trends toward less frequent tillage, such as that resulting from increased planting of no-till winter wheat, will result in increased nest success (Cowan 1982, Duebbert and Kantrud 1987). Neither Canadian Wildlife Service and US. Fish and Wildlife Service (1987) nor Klett et al. (1988) evaluated nest success of diving ducks, but other studies indicate it tends to be higher than for dabbling ducks (Bouffard et al. 1987, Table 12-3).

The meager information on survival of dabbling duck broods in the prairie-parkland region indicates that more than 50% of ducklings die before fledging (Table 12-3).

Cowardin et al. (1985) studied mortality of breeding female Mallards and their eggs and young in a 10,041 km² area in North Dakota and used a model developed by Cowardin and Johnson (1979) to assess Mallard recruitment rates. Cowardin et al. (1985) estimated breeding season mortality rates were 0.19 for adult females (suggested to be an underestimate), 0.92 for nests, and 0.57 for ducklings (not given, but calculated by us from data presented and an average brood size of 8.4 at time of hatch [obtained from A. T. Klett, personal communication]). Cowardin et al. (1985) concluded that, on average, each adult female in the spring population fledged 0.27 young females. They predicted that this recruitment rate would result in a 20% annual decline in the spring population unless it was bolstered by an influx of immigrants. Nest success was the most important variable affecting the annual recruitment rate, with 15% success (about twice as high as the observed rate) needed to achieve population stability.

The estimated 15% nest success needed to achieve population stability for Mallards is a useful benchmark against which Mallard reproductive performance can be judged. Klett et al. (1988) suggested that the 15% threshold level applied equally well to Northern Pintails, but that a threshold level of about 20% nest success was more appropriate for the Northern Shoveler, Bluewinged Teal, and Gadwall because of the lower renesting potentials of these species. An even higher threshold level would be expected for most diving ducks and other species that exhibit delayed maturity and seldom renest.

For Mallards and other dabbling ducks in the prairie-parkland region, available data indicate that in much of the region the collective mortality during the breeding season is close to or exceeds that needed for population stability. Hence, mortality during the breeding season has a pronounced effect on numbers of birds available for harvest and has likely contributed to lowered breeding populations.

B. Geese in the Artic Tundra Region

Attention, of course, tends to focus on populations that have declined to low levels. Concern about the effects of subsistence harvests and other causes of mortality of geese on the Yukon-Kuskokwim Delta increased steadily during the 1970s as the goose populations plummeted (see reviews by Raveling 1984, Cook 1986 , King and Derksen 1986, Mitchell 1986, Pamplin 1986). Increased predation during the early 1980s, primarily by arctic foxes, has further depressed goose populations on the Yukon-Kuskokwim Delta. The rates of goose nest losses documented during the late 1960s and 1970s (about 0.11-0.39 [Table 12-2]) have been much higher in the 1980s, when the goose populations have been very low (Stehn 1986 in Pamplin 1986). For example, nest mortality in 1984 and 1985 was 0.66 and 0.68 for Cackling Canada Geese and 0.86 and 0.73 for Brant, for each year, respectively. Similarly, nest mortality rates for Emperor Geese averaged 0.63 (N = 70) in 1982-83 on a study site where losses averaged 0.34 (N = 134) during 1977-79 (D. G. Raveling, unpublished data). Reasons for increased predation by arctic foxes appear to be associated with a combination of low human harvests of foxes during winter, adequate winter food for foxes from sea mammal carcasses unretrieved by native hunters, and low rodent prey populations during spring (native hunters, personal communication; D. G. Raveling, personal observation and unpublished data).

Arctic fox predation poses a serious threat to continued existence of the diminished Brant colonies on the Yukon-Kuskokwim Delta (Raveling 1989). Arctic fox families tend to occupy relatively exclusive areas during spring and summer (Eberhardt et al. 1982), and, therefore, individual Brant colonies are generally hunted by only a few foxes. One fox family can almost eliminate the production of a small Brant colony (a few hundred pairs or less). For example, MacInnes and Misra (1972) reported that one arctic fox destroyed over 200 (all) Snow Goose nests in a 3 km² area of the McConnell River delta, Northwest Territories within 5 days. In contrast, the predator swamping effect (Wittenberger and Hunt 1985) of a large colony (many thousand pairs) would likely dampen the effect of this predation by overwhelming the capability of the foxes to take a large proportion of available eggs. Perhaps a crash in the fox population on the Yukon-Kuskokwim Delta will solve this problem of severe predation by itself, but in the meantime, the ability of these geese to expand their numbers is being slowed or prevented by predation. Regardless of one's views of the relative effect of factors leading to the massive declines of geese on the Yukon-Kuskokwim Delta, the balance between the number of geese and their predators has recently changed.

The history of Snow Geese on Wrangel Island also illustrates how subsistence hunting, human activities favoring arctic foxes, and predation by foxes on geese can interact to depress a goose population and then prevent or slow its recovery once numbers of geese are diminished (see review by Bousfield and Syroechkovskiy 1985). A human settlement established on the island in 1926 for purposes of hunting sea mammals resulted in: (1) the annual gathering of hundreds of thousands of Snow Goose eggs, (2) an increased and nonoscillating arctic fox population because of an abundant and stable food supply provided by wastes from the harvest of sea mammals (and later by wastes from harvest of an introduced reindeer herd), and (3) arctic foxes taking up to 40%-80% of Snow Goose eggs and goslings each year. By the mid-1970s, when subsistence egg collecting was stopped, the Snow Goose population had declined to perhaps 10% of its former size and was restricted to one major colony that could not expand because of predation by the foxes.