Patterns of habitat use in animals presumably have evolved in response to diverse selective processes, so we first examined whether the theory of natural selection formed the conceptual framework for published studies (N = 270) of nest-site selection by birds. Most (61%) studies of nest-site selection tested for pattern arising from natural selection (whether used nest habitat differed from available nesting habitat), many (54%) tested for evidence of the process of natural selection (whether unsuccessful nests differed from successful nests), some (10%) tested whether the process of natural selection caused subsequent adaptation, but remarkably few conceptually linked these elements or used the theory of natural selection as the rationale for their questions. We then tested for evidence of natural (phenotypic) selection with data for six species of ducks. At nests, we used six variables to describe vegetation structure/nest position and categorized patch types (pond edge, native grass, planted cover, shrubs, or trees) in which nests were found; nest fates (abandoned, depredated, or successful) were also determined. For Blue-winged Teal (Anas discors), Northern Shoveler (A. clypeata), and Mallard (A. platyrhynchos), there were significant patterns of nonrandom nest-site placement within a gradient of vegetation structure/nest position. For Blue-winged Teal and Gadwall (A. strepera), nest success varied within these gradients in a way that could exert directional selection. Several tests for adaptive nest-site choice were conducted. Nest fate did not influence fidelity of females to patch types. However, Mallards with previously unsuccessful nests dispersed farther than females with previously successful nests. Nonetheless, neither fidelity to patch type nor dispersal distance influenced subsequent nest success. In the long term (over 8 yr), there was a weak tendency within species for nest density to be higher among patch types where relative nest success was higher. In the short term (from year t to year t + 1), this pattern was not observed in a vegetation-structure/nest-position gradient for any species. The strongest evidence of adaptive response to nest fate was higher nest density on an island where nest success was relatively high.
Read full abstract