Ecology of the Cerulean Warbler in the Cumberland Mountains of East Tennessee

The objective of this study was to describe the distribution, local breeding densities and population trends, breeding biology, and habitat selection of the Cerulean Warbler (Dendroica cerulea) in the Cumberland Mountains of Morgan, Campbell, Scott, and Anderson Counties, Tennessee. Point counts were conducted during three years at six study sites. The Cerulean Warbler was detected at 50% of the point counts and was the sixth most frequently encountered bird. Spot-map censuses were conducted on five plots at three study sites. Cerulean Warblers occurred on all census plots at an average density of 84.9 pairs/100 ha. Average density on individual plots ranged from 21.3 to 137.6 pairs/100 ha. Although population trends on the census plots were not uniform, there was an overall decline during the late 1990s.

The breeding biology of the Cerulean Warbler was studied by observing courtship behavior and nest-building, and monitoring a total of 52 nests. Female warblers built nests, incubated eggs, and brooded nestlings. Females fed nestlings more often than males; the rate at which males and females removed fecal sacs from the nest did not differ. Mean clutch size was 3.7 ± 0.13 SE. For the years 1996-1999, nest success, measured by the Mayfield method, averaged 0.3708; the daily nest success rate averaged 0.9685 ± 0.0057 SE. The major cause of nest failure appeared to be weather events, including presumed nestling starvation during unusually cool and wet weather.

Nests were built in 12 taxa of deciduous trees. Although there were some plot-specific preferences, most tree taxa were selected in proportion to their availability. Nests averaged 15.9 m ± 0.81 SE (range 7.0 - 36.3m) above ground. Nest trees averaged 23.6 m ± 0.90 SE (range 10.2 - 42.2 m) tall and 39.6 cm± 2.53 SE (range 12.3 - 76.2 cm) in diameter at breast height.

Habitat characteristics were sampled at nest sites, territory sites, and random sites on census plots, and at point count locations. Basal area was consistently greater, and both shrub cover and canopy cover consistently less, at nest sites than at random sites, at territory sites than at random sites, and at nest sites than at territory sites. Point count locations where Cerulean Warblers were present had larger diameter live trees and more sapling cover than locations where the warbler was absent. Logistic regression models successfully classified over 70% of sites on census plots but were less successful in predicting Cerulean Warbler presence at point count locations.