2001 Volume 50 Issue 4 Pages 167-174,181
Roost site selection of ducks in relation to safety was investigated at 26 ponds from December 1992 to February 1993 and from December 1993 to January 1994 in northern Chiba Prefecture, central Japan. The degree of safety was evaluated as the size of the safe range. The escape distance (D) between the first individual escaping and the observer was estimated, based on five experiments on the Spot-billed Duck Anas poecilorhyncha the most common species in the study area. I subtracted an area of periphery (unsafe area: width D: 29.2 m from the accessible waterside) from the total area of each pond. The remaining water surface was defined as the safe range. Ten species, including eight dabbling ducks and two diving ducks, were observed at 17 ponds. The total areas and the safe ranges at the 17 ponds with dabbling ducks were significantly larger than at those nine ponds without them. A similar tendency was detected between the seven ponds with diving ducks and the 19 ponds without them. The number of dabbling ducks was positively correlated with the size of the safe range; however no such relationship was detected between the number of dabbling or diving ducks and the total area, or between the number of diving ducks and the size of the safe range. These differences seem to be related to the utilization patterns of the ponds between ducks with different foraging habits. The dabbling ducks utilized the ponds only for resting, while the diving ducks used them as resting and foraging sites. The safe range is a useful index for representing the safety of a roost.