Northern Prairie Wildlife Research Center
Robert J. Kawula and R. J. Gates, Cooperative Wildlife Research Laboratory and Department of Zoology, Southern Illinois University at Carbondale, Carbondale, IL 62901 USA
We used radiotelemetry to investigate home range and habitat use patterns in a population of wood ducks that nests entirely in natural tree cavities in southern Illinois during 1994 and 1995. Female wood ducks were marked with necklace-type transmitters and located daily throughout the breeding season. Individual hens also were monitored every 2 hr throughout diurnal periods (30 min before/after sunrise/sunset) once/week. Fixed kernel (95% volume contours) home ranges were determined using program KERNELHR. Least-squares cross validation (LSCV) was used to select smoothing parameters to calculate utilization distributions (UD) within home ranges. Three-dimensional UD's were plotted for each hen and home ranges constructed by projecting the 95% volume contour onto a 2-dimensional surface. Polygons containing 95% of UD's were used to extract habitat composition within home ranges from a geographical information system (GIS) coverage. The GIS coverage included 7 land cover classes that were interpreted from color infrared aerial photography. We used compositional analysis to investigate habitat preferences at 2 scales: home range composition in relation to the entire study area, and habitat composition at use sites in relation to home range composition of individual hens. Home range size and habitat preferences were compared between 1994 (n = 26) and 1995 (n = 35) when seasonal flooding patterns differed. We also compared home range and habitat preferences of nesting versus non-nesting hens, and among prenesting and post-nesting periods. Nesting effort differed between 1994 (31% of hens) and 1995 (54%). Habitat composition of home ranges differed between these years (λ = 0.720, F = 3.493, P = 0.005), although there were no differences in habitat use within home ranges (λ = 0.813, F = 2.068, P = 0.072). Hens used more seasonal flooded forest in 1995 when there was sudden flooding during pre-nesting and nesting, compared to 1995 when water levels were low but stable throughout the nesting season. Habitat preferences did not differ (P > 0.05) between nesting and non-nesting hens at either scale. Habitat composition within home-ranges differed before and after hatching (λ = 0.426, F = 4.948, P = 0.002), although habitat composition of use sites did not change (λ = 0.699, F = 1.580, P = 0.200). Hens showed a high degree of fidelity to specific use sites throughout the nesting cycle, despite changes in habitat conditions. This may constrain foraging opportunities of some hens if they are unable to adapt movement or habitat use patterns to sudden flood events. Our study demonstrates the interface of radiotelemetry with GIS to determine home range and habitat use patterns of a wild vertebrate. We discuss applications of recent developments in home range and habitat use analyses to answer basic questions about breeding biology of wood ducks.