• Stockwell, Sarah S.
  The Graduate School, University of Maine
  

This study relates the abundance and distribution of birds in eight Maine peatlands to peatland vegetation, size, and geomorphology, and uses data on peatland birds to test predictions of the density assessment hypothesis of habitat selection. I censused birds in eight Maine peatlands of varying types and sizes during May and June of 1983 and 1984. Eight distinct vegetation types occurred in the peatlands. I used a variable-width transect method for censusing birds and estimated bird densities from a modified Emlen method and the Fourier Series Estimator in computer program TRANSECT.

Bird species richness (BSR, number of species censused) of the peatlands l sampled (48-81 per site, 101 in total) was greater than that reported for other habitats in Maine. Among the eight vegetation types, bird species richness was not correlated with foliage height diversity (vertical heterogeneity of vegetation), but density of birds (BD, number of birds per 40 ha) and bird species diversity (BSD, Shannon Diversity lndex) were correlated with foliage height diversity.

I conclude that horizontal heterogeneity of vegetation is more important than vertical structure of vegetation in influencing bird species richness and bird density in these peatlands.

I tested three predictions of the density assessment hypothesis of Fretwell and Lucas on habitat selection in birds. First, linear regression revealed that those species with higher population densities used a wider variety of habitats than those species with lower population densities, both on intraspecific and interspecific levels. Second, most species whose population level changed showed increases or decreases in density simultaneously across several habitats. Third, most species showed changes in dominance consistent with those predicted by the density assessment hypothesis, especially ¡f modified to account for philopatry. These findings generally support the density assessment hypothesis of density-dependent habitat selection.