Nest Site Selection in Eastern and Western Kingbirds: A Multivariate Approach

Abstract

-Nest site selection by Eastern and Western kingbirds (Tyrannus tyrannus and T. verticalis) was studied at Delta Marsh, Manitoba, and results were analyzed using multivariate techniques. Principal component analysis of overstory vegetation characteristics for randomly selected plots defined the habitat space of the study area in terms of the major components of variation in the vegetation. Discriminant function analysis of the random plots identified the major trend in the vegetation (corresponding roughly to a north-south topographical gradient), and characterized the differences between northand south-zone habitat types. Principal component analysis of the vegetation characteristics of kingbird nest sites determined the relative position of each species in the habitat space. Virtually all of the Western Kingbird sites and most of the Eastern Kingbird sites appeared to be situated in that portion of the habitat space considered to represent north-zone habitat. Projection of the nest site data onto the random plot discriminant axis indicated that Eastern Kingbird nest sites were more evenly distributed than those of Western Kingbirds along the major trend of variation in the forest vegetation. Discriminant function analysis of the nest sites identified the variables most important in characterizing the difference between the nest sites of the two species. Western Kingbird sites were characterized by fewer but larger trees than Eastern Kingbird sites. Western Kingbird nest trees were larger and taller, and Western Kingbird nests were placed higher than those of Eastern Kingbirds. Eastern Kingbirds behaved as habitat generalists, selecting nest sites from both available habitat types, and selecting nest trees from a wide range of tree sizes. Western Kingbirds behaved as specialists, selecting sites in only the north-zone habitat type, and nest trees from the larger sizes. The Eastern Kingbird (Tyrannus tyrannus) and the Western Kingbird (T. verticalis) are sympatric over much of western North America (Fig. 1). The area of sympatry constitutes 35% of the breeding range of the Eastern Kingbird, and 60% of that of the Western Kingbird. Within their common range, where one species is abundant the other is usually much less so, and there is also regional variation as to which species exploits the wider range of habitats (Hiatt 1942, Hamilton 1962, Smith 1966, Tatschl 1973). Over much of its range the Eastern Kingbird nests along the edges of woodlots, in orchards, near marshes, and frequently in riparian communities (Bent 1942, Smith 1966). In the parts of its range where woods are scarce, the Eastern Kingbird inhabits open country, often nesting in low shrubs. The Western Kingbird is typical of dry grassland areas, but over much of its range its distribution appears to depend on the presence of at least a few trees for nesting. It is, therefore, a common inhabitant of prairie riparian woodlands, where it often reaches high nesting densities (e.g., see Carothers et al. 1974). This preference for trees or other tall structures (e.g., utility poles) for nesting suggests that the recent expansion of the breeding range of the West rn Kingbird (since 1900) may have been influenced by the planting of trees and erection of man-made structures which accompanied settlement of the plains at the turn of the century (see Nice 1924). The association of these species in Manitoba is the result of this recent breeding range expansion of the Western Kingbird (Taverner 1927). Analysis of census data from the Breeding Bird Survey (Robbins and Van Velzen 1967) for the period 19701975 revealed that the Eastern Kingbird was four times as abundant as the Western Kingbird in Manitoba. Both species breed in high densities on the forested dune ridge