ABSTRACT.-Predation patterns on artificial waterfowl and passerine nests were studied in upland habitats adjacent to freshwater marshes located in urban, agricultural and natural habitats, near Ottawa, Ontario in 1989 and 1990. Nest predation on waterfowl nests did not change during the breeding season in urban and in agricultural habitats but it increased significantly in natural habitats. There were no significant differences over time of year in predation on passerine nests in any of the three types of habitats. Nest camouflage was not a factor in preventing nest destruction. Nest predation was higher on passerine nests than on waterfowl nests in urban and natural habitats but was similar in agricultural habitats. Losses of waterfowl nests were similar among habitats in both years. Passerine nests were more frequently destroyed in natural habitats than in urban and agricultural habitats. Small predators such as the blue jay (Cyanocitta cristata) and red squirrel (Tamiasciurus hudsonicus), more commonly observed in urban and natured squirrel habitats compared to agricultural haft hats, would most like commonly explain differences in urban and nest predation among habitats compared to agricultural has difitats, would most likely explain differences between nest predation among habitats as well as differences between nest types.
Identifying causes of nest predation has been the subject of numerous studies in recent years because nest loss to predation is a major cause of mortality in birds. Researchers have used artificial nests to overcome difficulties associated with studying real bird populations such as insufficient number of nests to examine predation patterns (Major and Kendal, 1996; Bayne and Hobson, 1997). Most studies were designed to simulate passerine or waterfowl nests in a large array of situations ranging from single-species studies (O'Reilly and Hannon, 1989) to large-scale studies of waterfowl breeding ecology (Butler and Rotella, 1998). Although nest predation observed on simulated nests may not accurately predict success of real nests (Butler and Rotella, 1998; Wilson et al., 1998), studies using artificial nests can be useful in detecting patterns of nest predation in different experimental settings (Wilson et al., 1998; Vander Lee et al., 1999).
Site-specific factors such as nest concealment (Bowman and Harris, 1980; Sugden and Beyersbergen, 1987) or nest height (Yahner and Scott, 1988; Rudnicky and Hunter, 1993), and area-specific factors such as nest density (Goransson et al., 1975; Esler and Grand, 1993) or habitat (Janzen, 1978; Rudnicky and Hunter, 1993) can influence nest success, sometimes with conflicting results (Major and Kendal, 1996). The predator community inhabiting a site is an important factor affecting predation on artificial nests because predation patterns differ among habitats with respect to their specific predator community (Martin, 1987; Picman, 1988; Miller and Knight, 1993). Time within the breeding season may also be important because nest predation may change during the breeding season (Crabtree et al., 1989; Wilson et al., 1998). Martin (1987) showed that predation pressure remained constant throughout the season in evergreen forests in Arizona, whereas Best and Stauffer (1980) observed increasing nest predation as the breeding season progressed in a variety of riparian habitats in Iowa.
Jobin and Picman (1997) studied predation patterns in eight freshwater marshes located in urban, agricultural and natural habitats near Ottawa, Ontario. They showed that increasing nest predation on artificial waterfowl and passerine nests in the summer was related to decreasing water level in all three habitats because mammalian predators could invade the marshes' shallow waters later in the season. However, the continuous presence of the marsh wren (Cistothorus palustris), an important passerine nest predator, was related to constant passerine …