Wetland density is believed to be an important determinant of home range size variation in mallards (Anas platyrhynchos), but hypothesized effects of upland habitat and female size and age have not been adequately evaluated. Thus, I investigated correlates and consequences of home range size variation using radio-tracking data for 131 female mallards studied on 12 Canadian prairie parkland sites, 1995-1998. Home range size and habitat composition varied within and among study areas; overall, home range size variation was best modeled to include effects of seasonal and semi-permanent wetlands (β = -0.06 ± 0.01 SE) and wood-shrub habitat (β = -0.03 ± 0.01 SE). Contrary to predictions, I obtained no support for a positive association between home range size and female body size or a negative relationship between home range size and female age. After controlling effects of wetland density, mean home range sizes were larger on study areas with lower mallard breeding pair densities. I suspect that individual home ranges were smaller in areas of high pair density because of increased intraspecific competition for breeding space. A higher proportion of wood-shrub habitat may have contributed to smaller individual home range sizes because of greater relative availability of preferred nesting habitat. Likewise, a high proportion of wetlands in home ranges could enhance access to important resources such as food, leading to smaller home range sizes.<p> Reproductive and survival consequences were investigated using 8 variables to distinguish between three reproductive categories (females that either did not nest, nested but failed, or nested successfully) and two survival categories (dead versus alive) with discriminant function analysis. Successful females were clearly separated from non-nesting females by having smaller home ranges (95% kernel estimate) with higher percentages of wood-shrub and habitat treatment but lower percentages of seasonal and semi-permanent wetlands. Females that did not nest were further distinguished from nesting females by being younger, structurally smaller and having larger home ranges composed of higher percentages of seasonal and semi-permanent wetlands. Date of first nesting (standardized by study area) was not associated with home range composition. Survival was also unrelated to either home range composition or female attributes. Overall, breeding performance was better described by variation in landscape characteristics than by female attributes, a finding that is consistent with other recent evidence from breeding ducks.

Wetland density is believed to be an important determinant of home range size variation in mallards (Anas platyrhynchos), but hypothesized effects of upland habitat and female size and age have not been adequately evaluated. Thus, I investigated correlates and consequences of home range size variation using radio-tracking data for 131 female mallards studied on 12 Canadian prairie parkland sites, 1995-1998. Home range size and habitat composition varied within and among study areas; overall, home range size variation was best modeled to include effects of seasonal and semi-permanent wetlands (β = -0.06 ± 0.01 SE) and wood-shrub habitat (β = -0.03 ± 0.01 SE). Contrary to predictions, I obtained no support for a positive association between home range size and female body size or a negative relationship between home range size and female age. After controlling effects of wetland density, mean home range sizes were larger on study areas with lower mallard breeding pair densities. I suspect that individual home ranges were smaller in areas of high pair density because of increased intraspecific competition for breeding space. A higher proportion of wood-shrub habitat may have contributed to smaller individual home range sizes because of greater relative availability of preferred nesting habitat. Likewise, a high proportion of wetlands in home ranges could enhance access to important resources such as food, leading to smaller home range sizes.<p> Reproductive and survival consequences were investigated using 8 variables to distinguish between three reproductive categories (females that either did not nest, nested but failed, or nested successfully) and two survival categories (dead versus alive) with discriminant function analysis. Successful females were clearly separated from non-nesting females by having smaller home ranges (95% kernel estimate) with higher percentages of wood-shrub and habitat treatment but lower percentages of seasonal and semi-permanent wetlands. Females that did not nest were further distinguished from nesting females by being younger, structurally smaller and having larger home ranges composed of higher percentages of seasonal and semi-permanent wetlands. Date of first nesting (standardized by study area) was not associated with home range composition. Survival was also unrelated to either home range composition or female attributes. Overall, breeding performance was better described by variation in landscape characteristics than by female attributes, a finding that is consistent with other recent evidence from breeding ducks.