COWBIRDS BREEDING IN THE CENTRAL APPALACHIANS: SPATIAL AND TEMPORAL PATTERNS AND HABITAT SELECTION

Abstract

To help minimize the impact of brood parasitism by female Brown-headed Cowbirds (Molothrus ater) on Nearctic-Neotropical migrants, we studied their spatial and temporal patterns and habitat selection throughout the breeding season. The study was conducted on an anthropogenic forest landscape in the central Appalachian Mountains of western Maryland. This landscape was characterized by different forest types and stages of succession resulting from logging, agriculture, and low-density housing development, as well as by natural and human-made corridors. Using radiotelemetry, cowbird movements were monitored in 3-h time periods during the day, beginning at 0500 hours Eastern Standard Time. Females were generally alone or with one or two other individuals from 0500 to 1100 hours. They were located in deciduous forest and brush near edges formed by road, power line, or stream corridors and other clearings, presumably searching for nests to parasitize. Breeding core ranges of individual females tended to be spatially separated from one another. Females would travel, on average, 2.3 km from breeding to feeding areas. From 1100 hours until dusk, feeding aggregations of ≥12 individuals could be found in grazed pasture or other areas of short grass. Feeding core ranges overlapped considerably. At dusk, cowbirds would fly an average of 3.0 km to roosting areas in deciduous forest near (X¯ = 1.3 km) where they would be found the next morning, presumably breeding. Cowbirds roosted singly or in small aggregations in trees near forest edges. Although the mean home range was 1592 ha, only 3.5% or 31.6 ha, on average, was actually used for different activities. Most individuals had at least two feeding core ranges (X¯ = 16.5 ha total), one breeding core range (X¯ = 9.5 ha), and one roosting core range (X¯ = 1.8 ha) or combination roost/breed core range (X¯ = 9.8 ha). Multiple-use core ranges were generally larger than single-use core ranges. Breeding and roosting core ranges were often elongated, with the long axis paralleling a linear canopy opening, e.g., a road, power line, or stream corridor. Multiple-use core ranges were more circular in shape. Breeding and roosting core-range centroids in the forest were often <50 m from a forest-road, brush, power line, or stream edge. Feeding core ranges were nearly always associated with agricultural land. In selecting home ranges, cowbirds incorporated brush and deciduous forest in greater proportions than those found within the study area. At the next level of habitat selection, agricultural land and deciduous forest were the most important habitat types within core ranges. Breeding core ranges had higher numbers of seedlings and saplings and large (>22.5 cm dbh) snags than were found within forest interior >250 m from an edge. Conservation measures designed to reduce use of large forest tracts by cowbirds in this region would include eliminating feeding areas, minimizing canopy openings or edges, and perhaps managing existing edges in order to disperse nests of potential host species.