Testing the Role of Social Cues in Saltmarsh Sparrow Habitat Selection Decisions

Date of Completion

January 2010


Agriculture, Wildlife Conservation|Biology, Ecology|Psychology, Behavioral Sciences




The saltmarsh sparrow Ammodramus caudacutus is a species of national and global conservation concern that is threatened by its limited breeding range and vulnerability to sea-level rise. Despite detailed studies of its nesting habitat, our ability to predict this species' distribution remains deficient. Several lines of evidence suggest that these sparrows may combine social information with their assessment of the physical environment in order to select nesting habitat, yet the way in which birds integrate these disparate types of information is not understood. To resolve this uncertainty, I investigated how conspecific social cues may influence breeding habitat selection. My research indicates that nesting activity is more strongly related to cues describing total sparrow activity than to those specifically related to breeding activity, such as male song or female provisioning behavior. Although this result suggests that sparrow abundance could be a reliable cue of habitat quality, experimental manipulations of apparent sparrow densities indicate that saltmarsh sparrows do not use conspecific attraction either to select breeding sites within marshes, or to select which marshes to settle in. In addition, although previous accounts have described saltmarsh sparrows as semi-colonial, spatial tests of aggregation failed to detect any evidence for non-random patterns in nest placement, consistent with the hypothesis that nest placement is random with respect to other nests and that females are not attracted to settle near other nesting females. Finally, because nest flooding is a major cause of nest failure for this species and is not generally well understood, I studied the mechanics of nest flooding. During 2007-2009, all but 28 of 191 nests sampled were flooded at least once. Some nests, including those that produced young, were flooded up to 10 times within the nesting cycle. On average, the maximum tide height at which nests did not flood differed between successful nests and those that failed due to flooding by just 5 cm. When I modeled nest fate using variables related to tide height and nest timing, the top performing models all included variables related to tidal metrics. This suggests that avoidance of flooding is likely a major component of habitat selection decisions. ^