Breeding season movements and resource selection of female wood ducks in southern Illinois

Most studies of habitat use and movements of wood ducks have been conducted during the brood rearing and non-breeding periods. However, little information is available on habitat use and movements of hen wood ducks during prenesting, egg laying and incubation. Availability of resources during this period are critical for breeding success of hens. I examined home ranges, movements, and habitat use of natural cavity nesting wood duck hens during the 1994 ( n = 29) and 1995 (n = 41) breeding seasons using radiotelemetry. A higher proportion of hens attempted nesting in 1995 (21 of 35; 60%) than in 1994 (8 of 19; 42%) when habitat conditions were drier and the spring flood pulse occurred later compared to 1994. Fixed kernel home ranges also were larger in 1995 (Z = -3.13, P = 0.002) compared to 1994. Home ranges also were larger ( Z = 2.09, P = 0.037) and mean distance moved between consecutive locations were greater for nesting than non-nesting hens. Habitat preference was examined at 2 spatial scales using compositional analysis. Habitat composition within the home range differed (P < 0.001) from the entire study area. The proportion of habitat types at radiolocations also differed (P < 0.001) from the proportion of habitat types within home ranges. Flooded live forest and scrub-shrub habitats were most preferred and open water and cropland habitats were least preferred at both spatial scales. Habitat composition of home ranges differed between years (Lambda = 0.619, P = 0.005). Home ranges in 1994 contained more seasonally flooded forest while 1995 home ranges contained more scrub-shrub, cropland and open water habitats, likely a result of drier habitat conditions in 1995. Although I found no differences (Lambda = 0.699, P = 0.200) in proportions of radiolocations within each habitat type before and after hatching, habitat composition within pre- and post-hatching home ranges differed between years (Lambda = 0.426, P = 0.002), indicating a shift in home ranges between the 2 periods. Similar to other studies, open water was avoided by brooding hens. I could not discriminate between used and unused forest patches in 1994, however, univariate comparisons found used patches contained more mast-producing trees than unused forest patches. Size and shape discriminated used and unused flooded forest patches in 1995. Larger, more irregularly shaped patches were used due to dry conditions in early spring. Flood pulses caused a shift in areas used by hens in 1994 (73%) and 1995 (77%). However, mean pairwise distances of nesting hens were smaller, indicating a greater tendency to return to the same habitat patches. Wood ducks exploit unstable habitats that are susceptible to widely fluctuating conditions. However, timing and extent of the spring flood pulse may have affected nesting effort during my study. High fidelity to habitats within home ranges may have a negative impact on nesting effort, and ultimately recruitment, if fluctuations occur during critical times during the breeding period. Nevertheless, hen wood ducks appear to be resilient to habitat fluctuations, especially if they occur later in the breeding season.