An experimental test of information use by North American wood ducks (Aix sponsa): external habitat cues, not social visual cues, influence initial nest-site selection

Birds may use a variety of cues to select a nest site, including
external information on habitat structure and nest site
characteristics, or they may rely instead on social information
obtained directly or indirectly from the actions of
conspecifics. We used an experimental manipulation to determine
the extent to which a California population of the wood duck (Aix sponsa)
used social information gleaned from visual cues inside nest boxes that
might indicate the quality or occupancy of that site. Over two nesting
seasons, we manipulated the contents of newly installed boxes to simulate
one of three states: (1) presence of wood duck eggs, indicating current
use of a nest site; (2) presence of down and shell membranes, indicating a
previously successful nest; and (3) control nests with fresh shavings
indicating an unused box. In addition, we measured habitat characteristics
of the area surrounding each box to assess the use of external, non-social
information about each nest site. We found no evidence that females laid
eggs preferentially, or that conspecific brood parasitism was more likely
to occur, in any of the treatments. In contrast, nest site use and
reproductive traits of wood ducks did vary with vegetation cover, and
orientation and distance of the box from water. Our results suggest that
personal information, not social information, influence initial nest site
selection decisions when females are unfamiliar with a site. Social cues
likely become increasingly important once nest sites develop their own
history, and a population becomes well established.