Survival of Mallard Broods in South-Central North Dakota

-Survival characteristics of 25 broods of Mallards (Anas platyrhynchos) were determined on a study area in the Missouri Coteau of south-central North Dakota in 1976-1977. Radio-equipped Mallard hens fledged at least one duckling in 7 of 16 (44%) broods produced in 1976, 5 of 9 (55%) in 1977, and 12 of 25 (48%) for both years combined. Of the 13 broods in which all young were lost, 11 (85%) were lost within the first two weeks after hatching. All losses of entire broods occurred in wetlands; few ducklings and no entire broods were lost during overland travel. Predation by mink (Must& vison) was apparently the principal cause of duckling mortality. The magnitude and pattern of brood mortality are key factors in the reproductive ecology of waterfowl. Measurement of brood survival is essential for calculating recruitment from nesting data (Cowardin and Johnson 1979). Yet, survival and mortality patterns during the prefledging period are poorly understood. Most duckling mortality in Mallards (Anus pfatyrhynchos) occurs before young are three weeks old (Dzubin and Gollop 1972, Ball et al. 1975) but the specific mortality factors of ducklings are largely unknown. Predation, accidents, and losses due to scattering or exhaustion are often cited as potential agents of mortality, but little direct evidence has been presented in the literature. Some workers have suggested that overland travel is particularly hazardous to survival of ducklings (Bellrose 1953, Keith 196 1, Odum 1970, Dzubin and Gollop 1972). However, Evans and Black (1956) found no evidence that mobile broods suffered more mortality than sedentary broods in prairie wetland habitat. This paper describes survival characteristics of Mallard broods on a study area in southcentral North Dakota as determined by radio telemetry and observation. To identify factors contributing to duckling mortality, we studied the effects of overland movement, predation, and hen-brood bonds on the survival of Mallard broods. STUDY AREA AND METHODS The study area in south-central North Dakota encompassed five tracts in Stutsman County. The largest tract (93.2 km’), referred to as the Medina Study Area, was 9.6 km south of Medina. This site was characteristic of the Missouri Coteau with moderately rolling glacial moraine, and from 2 to 14 wetland basins per square kilometer. The 41 .4-km2 central part of the area, where most radio tracking and observations took place, contained 10% ephemeral wetlands, 33% temporary wetlands, 35% seasonal wetlands, and 15% semipermanent wetlands (Stewart and Kantrud 197 1). The remaining wetlands were three permanent lakes of over 20 ha and several dugouts and fens. Land was principally used for livestock and small grain farming; about 40% of the uplands was in pasture and forage crops and 60% was cultivated. The second tract (2.61 km2) was a federally-owned Waterfowl Production Area (WPA), 6.4 km northwest of Medina. On this area, 16% of the wetland basins was seasonal and 84% was semipermanent. About 50% of the upland was cultivated or in forage crops and 50% was in native mixed-grass prairie or idle cover. During drought conditions in 1977, we also studied broods on three other WPA’s about 14 km east of Medina. The third and fourth tracts were primarily in federal ownership as WPA’s and contained single semipermanent wetlands of 18 1 and 22 ha, respectively. The fifth was a WPA of 180 ha containing 26 ha of wetland habitat. Uplands on those three tracts were in native prairie or a planted mixture of grasses and legumes. Mallard hens were captured on nests with remote-controlled traps (Shaiffer and Krapu 1978) during 1976 and 1977 at about 20 days into incubation. Each hen was fitted with a back-mounted radio pack weighing about 25 g and similar to that described by Dwyer (1972). The location of each radio-equipped brood hen was plotted one to eight times each day and night. Locations were determined by triangulation from known positions with hand-held