Breeding Dispersal by Birds in a Dynamic Urban Ecosystem

Abstract

Changes in land cover during urbanization profoundly affect the diversity of bird communities, but the demographic mechanisms affecting diversity are poorly known. We advance such understanding by documenting how urbanization influences breeding dispersal-the annual movement of territorial adults-of six songbird species in the Seattle, WA, USA metropolitan area. We color-banded adults and mapped the centers of their annual breeding activities from 2000-2010 to obtain 504 consecutive movements by 337 adults. By comparing movements, annual reproduction, and mate fidelity among 10 developed, 5 reserved, and 11 changing (areas cleared and developed during our study) landscapes, we determined that adaptive breeding dispersal of sensitive forest species (Swainson's Thrush and Pacific wren), which involves shifting territory and mate after reproductive failure, was constrained by development. In changing lands, sensitive forest specialists dispersed from active development to nearby forested areas, but in so doing suffered low annual reproduction. Species tolerant of suburban lands (song sparrow, spotted towhee, dark-eyed junco, and Bewick's wren) dispersed adaptively in changing landscapes. Site fidelity ranged from 0% (Pacific wren in changing landscape) to 83% (Bewick's wren in forest reserve). Mate fidelity ranged from 25% (dark-eyed junco) to 100% (Bewick's wren). Variation in fidelity to mate and territory was consistent with theories positing an influence of territory quality, asynchronous return from migration, prior productivity, and reproductive benefits of retaining a familiar territory. Costly breeding dispersal, as well as reduced reproductive success and lowered survival cause some birds to decline in the face of urbanization. In contrast, the ability of species that utilize edges and early successional habitats to breed successfully, disperse to improve reproductive success after failure, and survive throughout the urban ecosystem enables them to maintain or increase population size.

Fig 1. Classification of satellite images (top row) into suburban vegetation classes (bottom row) at original breeding site during year of occupancy (Territory A, Year 1), original breeding site during year it was abandoned (Territory A, Year 2), and new breeding site during year of occupancy (Territory B, Year 2).

In this example of a Pacific wren, the breeding male moved 647m from Territory A to Territory B in association with clearing of a portion of his former territory for construction of a new subdivision.

Fig 2. Breeding dispersal by songbirds in an urbanizing environment.

The frequency of movements by distance category is separated by guild in this stacked bar figure (gray portion of bars is 47 distances moved by avoiders and black portion is 457 distances moved by adapters and exploiters).

Fig 3. Average (+1SE) breeding dispersal distance by guilds in each of three landscapes.

Sample size above error bars.

Fig 4. Breeding dispersal distance of adapters / exploiters (A) and avoiders (B) in response to vegetation changes associated with construction of new subdivisions.

The most influential vegetation variable for each guild is used as the x-axis. For adapters / exploiters this was the total amount of forest within 100m of the new territory center. For avoiders it was the difference in total forest within 100m of the new territory center minus the old territory center (positive values indicate increases in forest cover at the territory each bird dispersed to. Shading of points indicates mate fidelity (A) or annual productivity (B), which were found to be important to each guild. Least-squares regression line ± 95% CI are fitted to data.