David E. Capen

EFFECTS OF FOREST ROADS ON HABITAT QUALITY FOR OVENBIRDS IN A FORESTED LANDSCAPE

Numerous studies have reported lower densities of breeding Ovenbirds (Seiurus aurocapillus) adjacent to forest edges. However, none of these studies has considered habitat use and reproductive success to address mechanisms underlying the observed pattern, and most were conducted in fragmented landscapes and ignored juxtapositions of forest with narrow openings such as roads. We studied the influence of forest roads on Ovenbird density in an extensively forested region of Vermont, evaluating habitat use and reproductive success relative to mechanisms proposed to explain the density-edge relationship. Territory densities on seven study plots were 40% lower within edge areas (0 to 150 m from unpaved roads) than within interior areas (150 to 300 m from roads). We simulated the distribution of Ovenbird territories and concluded that passive displacement, where birds perceive habitat interfaces as boundaries and limit their territories entirely to forest habitat, did not account for the observed density-edge pattern. Territory size was inversely related to distance from roads, providing an alternative explanation for reduced densities near edges and suggesting that habitat quality was higher away from roads. Pairing success was lower within edge areas than within interior zones, but the difference was not statistically significant. The proportion of males that produced fledglings did not differ between edge and interior areas. We conclude that habitat quality for Ovenbirds may be lower within 150 m of unpaved roads in extensive forested landscapes, affecting territory density and possibly reproductive success.

Effects of Small Forest Openings on the Breeding Bird Community in a Vermont Hardwood Forest

We examined the response of a forest bird community to the presence of small openings created by patch clear-cutting 0.4-ha plots within an extensive northern hardwood forest. We conducted bird censuses (June) and habitat measurements (July-August) in 102 census plots at four distances from openings in 1991 and 1992. Of 19 habitat variables measured, none differed significantly among forest plots at any distance from clear-cuts. Thirty-five (70%) of 50 bird species encountered during censusing were Neotropical migrants. Bird species richness did not differ as a function of distance from openings. However, species composition in plots within openings was least similar to that in plots farthest into forest, and most similar between the two distance categories farthest from openings. The movement of several forest-interior species away from openings, the addition of early-successional colonists in openings, and a high abundance of interior-edge species near openings contributed to the difference in species composition between openings and forest plots. As a group, Neotropical forest-interior migrants were significantly less abundant in openings than at any distance from them, and less abundant 50 m from openings than 200 m from them. Neotropical interior-edge migrants were significantly more abundant 50 m from openings than at any other distance. Nearctic migrants and nonmigrants did not respond to the presence of small openings. Of three locally common avian nest predators, none became more abundant in the openings. Brown-headed Cowbirds (Molothrus ater), also common locally, were never observed in or near the openings. Overall, bird species diversity increased in forested areas containing small openings due to the addition of edge and open-area nesters, but several forest-interior species were adversely affected by the presence of openings.

Landscape-level habitat use by brown-headed cowbirds in Vermont

We studied the distribution of brown-headed cowbirds (Molothrus ater) in a forested landscape in central Vermont to better understand effects of forest disturbance on the occurrence of this species in New England. We mapped 702 patches where chronic disturbance created permanent openings, such as pastures, lawns, and ski areas, and where recent logging had created temporary forest openings. We surveyed 67 randomly selected disturbance patches for cowbird presence during the 1992-93 breeding seasons; we detected cowbirds in 46% of the surveyed patches. We used logistic regression to develop predictive models from variables that described each disturbance patch in terms of size, shape, and position in the landscape relative to livestock areas and other disturbances. The model that best predicted patches used by cowbirds (75-80% correct classification) consisted of variables that described area of the patch, distance to closest chronic disturbance patch, and number of livestock areas within 7 km of the patch. From this model, we concluded that large patches (?= 9.8 ha) that were either chronically disturbed, or close to a chronic disturbance patch, and near a high concentration of livestock areas were most likely to contain cowbirds. We employed a grid-based geographic information system (GIS) procedure to map the probability of cowbird distribution in disturbance patches across the study area. The map illustrates that small (? = 4.0 ha), remote forest openings are unlikely to attract brown-headed cowbirds in Vermont.

Availability of nesting sites as a limit to woodpecker populations

Abstract We determined availability of potential cavity sites in a second-growth beech-birch-maple ( Fagus-Betula-Acer ) forest in central New York, USA and related these measures to estimates of wood-pecker abundance from line-transect surveys. Nine habitat variables, including seven indicators of decay, were used to develop a model to discriminate between cavity trees and non-cavity trees. Presence of broken tops and broken branch stubs > 10 cm diameter were the best predictors of bird use. Using the model to classify trees as potential nest sites, we estimated the density of such trees as 80.7 trees ha −1 . This supply of potential cavity trees should have been sufficient to support potential maximum woodpecker populations, but our estimates of density for five species of woodpecker were well below potential maximum woodpecker densities in northern hardwoods. Availability of suitable nest trees apparently was not limiting woodpecker on the study area, thus efforts to maintain or increase woodpecker populations should not focus solely on providing potential cavity trees because other factors may be limiting.

Management-Induced Reproductive Failure and Breeding Dispersal in Double-Crested Cormorants on Lake Champlain

Abstract We studied breeding dispersal of double-crested cormorants (Phalacrocorax auritus) associated with management practices that suppressed their reproduction on Lake Champlain in the northeastern United States. We implemented an experiment on one colony by spraying corn oil on cormorant eggs in portions of the colony and leaving other portions untreated. Gulls (Larus spp.) consumed cormorant eggs during the oiling process, but we reduced and then eliminated predation levels after the first year of the study. We used mark–recapture techniques within the experimental framework to measure rates of breeding dispersal for cormorants from the experimental colony and an unmanaged colony in Lake Champlain. Egg oiling increased the movement rate to the unmanaged colony by 3% during the year with no egg predation by gulls. When gulls depredated cormorant eggs at high rates during egg oiling, movement to the unmanaged colony increased by 20%. When cormorants are managed to reduce population sizes, methods that limit dispersal away from the managed colony may be most effective. Such methods would mitigate effects to nontarget populations and allow for a greater portion of the metapopulation to be managed.

Abundance and productivity of forest songbirds in a managed, unfragmented landscape in Vermont

Effective conservation of forest songbirds requires knowledge of responses to landscape pattern. We studied a breeding songbird community in a landscape characterized by scattered openings in a forest matrix in Vermont, 1993-94. We measured species composition abundance, and fledgling brood density in the forested areas of 10 500-ha sites: we compared a group of 5 disturbed-canopy sites (≤10% canopy disturbance) with a control group of 5 undisturbed-canopy sites. We conducted 456 point counts to determine species composition and abundance, and we conducted 60 line transect surveys to estimate fledgling brood density as a measure of productivity. Adult densities were not different between groups (P > 0.05) for 18 species listed as priority for conservation, although power was low. Combined brood densities of all species, Neotropical migrants, and ground nesters were not different among groups (P > 0.05). Brood densities of 3 Neotropical migrants alsa were not different (P > 0.05) However, combined brood density of area-sensitive, forest interior species was greater on control sites (P < 0.05). At the scale of this study, a minimal amount of canopy removal in the extensively forested landscap did not affect abundance and overall productivity of songbirds inhabiting the remaining forest, but canopy removal had begun to suppress productivity of some forest interior species.