TWEL Kelly Atchison Thesis

Community structure and behavior of wintering birds in riparian forests: relationships with landscape matrix, microclimate, and local habitat
Kelly A. Atchison, MS
Advisor: Amanda Rodewald
Recent studies suggest that the landscape matrix mediates habitat suitability for breeding birds; however we have a poor understanding of matrix effects on wintering birds. Landscape context, specifically urban development within the matrix, should affect the suitability of a habitat patch for wintering birds by altering microclimate and food availability. I examined (1) the extent to which community structure of wintering birds in riparian forests was associated with the percent urban development within 1 km, and (2) if local habitat characteristics, microclimate, or supplemental food were possible underlying mechanisms of landscape-level associations. Since avian behavior may reflect habitat suitability, I observed behavior of Carolina Chickadees (Poecile carolinensis) and Downy Woodpeckers (Picoides pubescens). Based on previous theoretical and empirical studies, I predicted that birds in urban forests would (1) devote less time to foraging, (2) join flocks less often, and (3) increase vigilance compared with birds in rural forests. If found, these behaviors would collectively suggest better food resources in urban forests.
This study was conducted on 36 mature riparian forest sites along an urbanization gradient in central Ohio. Forest birds were surveyed along a 250 X 40 m transect adjacent to the river 3 times each year during December-February 2001-2002 and 2002-2003. Behavioral data were collected at 6 sites in 2002, while data were collected at 30 sites in 2003. Proportions of time spent foraging, flocking, and scanning for predators (i.e., vigilance) were recorded for Carolina Chickadee and Downy Woodpecker. Habitat characteristics, microclimate variables, and numbers of bird feeders within 300 m were recorded at each site. An information theoretic approach was used to test a priori candidate models at both the local and landscape scales, and Akaike’s Information Criterion values were used to determine which models best explained the variation in the data.
Results suggested that factors operating at both local and landscape-level scales are important to wintering forest birds. Species richness was best explained by the percent urban development in the landscape, and total abundance was best explained by urban development and forest width. Both species richness and total abundance were positively related to percent urban development, while total abundance was negatively related to forest width. Abundances of American Goldfinch (Carduelis tristis), American Robin (Turdus migratorius), Brown Creeper (Certhia americana), and Tufted Titmouse (Baeolophus bicolor) were best explained by percent urban development. In contrast, abundances of Carolina Chickadee, Carolina Wren (Thryothorus ludovicianus), Downy Woodpecker, and Northern Cardinal (Cardinalis cardinalis) were best explained by local variables, such as number of bird feeders within 300 m. Behavioral data suggest that (1) urban forests provide better food resources than more rural sites, and that (2) bird feeders are important local resources. Percent urban development was negatively related to proportion of time spent foraging for Carolina Chickadee in 2002 and Downy Woodpecker in 2003. Additionally, proportion of time spent foraging within riparian forests was negatively related to number of bird feeders for Carolina Chickadee. Flocking behavior of Downy Woodpeckers was negatively related to the number of bird feeders, and their vigilance was best explained by several local variables, including the average number of hard mast trees, average number of snags, number of bird feeders, and wind speed. Number of bird feeders within 300 m was positively correlated with urban development over the 6 sites used for behavioral data collection in 2002. Minimum, maximum, and ambient temperature were positively correlated with urban development across all sites.
Overall, the increased species richness and abundance of wintering bird communities, coupled with decreased foraging effort, suggests that urban development in the landscape matrix surrounding riparian forests may benefit forest birds in winter. Urban landscapes may provide favorable energetic conditions compared to rural landscapes by increasing temperatures and providing supplemental food resources. However, associations between local and landscape variables were inconsistent and sometimes weak, which prevents a clear understanding of the mechanisms by which urban development affects winter birds. Furthermore, habitat quality should be best reflected in differences in survival and nesting productivity during the breeding season, but these were not examined here. For example, urban development within the landscape may be detrimental to sensitive Neotropical migrant species during the breeding season. Ultimately, a careful examination of the year-round species assemblage is necessary to understand how urban development affects avian communities.