Cooperative Fish and Wildlife Research Units Program: Virginia
Education, Research and Technical Assistance for Managing Our Natural Resources

Virginia Project

Cerulean Warbler Response to Variable Density Thinning and Other Silvicultural Prescriptions – A Multi-state Adaptive Management Project

August 2012 - December 2014


Participating Agencies

  • Coop Unit HQ

Populations of cerulean warblers (Setophaga cerulea; CERW), a neotropical migratory songbird ranging throughout much of the Appalachian Mountains, have declined at one of the steepest rates relative to other North American warbler species based on estimates from the North American Breeding Bird Survey (BBS; Sauer et al. 2011). These birds breed in large blocks of mature forest containing large diameter trees, but have highly specific preferences for complex vegetation structure in the canopy that include a well-developed mid- and understory and a broken canopy with spreading crowns and long horizontal limbs (Jones and Robertson 2001, Jones e t al. 2001, Nicholson 2003, Perkins 2006). The amount of mature deciduous forest in eastern North America with these vegetation characteristics and requisite disturbance regimes has decreased or been substantially altered over the past century. A recently completed body of research (Bakermans and Rodewald 2009, Boves 2011, George 2009, Hartman et al. 2009, Perkins 2006, Wood and Perkins In Press) found that CERW benefit from forest management practices that promote a heterogeneous canopy structure. Over 6 years at study sites in Kentucky, Ohio, Tennessee, and West Virginia, CERW had greatest increases in territory density on 10 ha silvicultural shelterwood type harvests that retained 10-20 m2/ha (mean 12.6 m2/ha) of white oak (Quercus alba), chestnut oak (Q. prinus), hickory (Carya spp.), and sugar maple (Acer saccharum) residuals in Appalachian and Central Hardwood forests. Results from this work currently are being used to develop forest management guidelines for CERW, however, questions remain regarding minimum effective harvest size, timing and proximity of multiple harvests, longevity of positive CERW response, and need for subsequent stand re-entry or intermediate treatment, i.e., partial residual removal, prescribed burning or herbicide application to prevent stand capture by shade-tolerant mast-poor woody species. We are proposing a region-wide study during 2013-2016 to address cerulean warbler and associated species response to a variety of planned harvests and management activities. Partners will include American Bird Conservancy, U.S. Fish and Wildlife Service, U.S Army Corps of Engineers, West Virginia Division of Natural Resources, West Virginia Division of Forestry, Virginia Department of Game and Inland Fisheries, Ohio Division of Natural Resources, Maryland Division of Natural Resources, and the Kentucky Wildlife Resources Commission. The overarching management goal is to create CERW habitat via variable density thinning, shelterwood, and firewood harvests as well as non-commercial overstory herbicide treatments across the wide range of site conditions, age classes, and landscape configurations that are within the parameters of typical forest management at the state wildlife management area and state forest scale and purpose. Likely study areas include Elk River WMA, Stonewall Jackson WMA, and Coopers Rock State Forest in West Virginia, Gathright, Highland and Goshen Mountain WMAs in Virginia, Paintsville Lake WMA in Kentucky, Savage River State Forest in Maryland, and Tar Hollow WMA and Zaleski State Forest in Ohio. State partners have identified areas slated for management. The West Virginia Unit will take data collection responsibilities for the West Virginia and Ohio study areas and the Virginia Unit taking data collection responsibilities for the Virginia, Kentucky and Maryland study areas. The specific research objective is to quantify relative abundance and territory density of cerulean warblers and associated avian species to environmental parameters within the context of applied forest management activities. We propose at minimum a 4-year study at each site. We will sample 1 year pre-harvest (2013) to establish baseline conditions, 2 years during harvest (2014-15) as the various harvests and management activities are implemented across each site, and 1 year post-harvest (2016). Throughout each study area, we will establish avian point count stations with >250 m between points to quantify relative abundance of the overall avian community. Points will be placed across the managed area to ensure that each harvest type and management activity is represented. Each point will be surveyed twice in each breeding season with standard 10 min point count surveys by trained and experienced observers (Ralph et al. 1993). All individuals seen or heard will be recorded by species. Points will be sampled between a half hour after sunrise and 1030 on days with appropriate weather conditions. Observations will be categorized into five detection types (calling, displaying, flyovers, singing, visuals), and sex will be determined if possible. We will record detections by distance categories and one-minute time intervals to calculate detection probabilities (Farnsworth et al. 2005). We will use program DISTANCE to estimate density (Thomas et al. 2009) for species with a large enough sample size. For other species, we will calculate indices of relative abundance. For cerulean warblers and a set of focal species, we also will sample via territory mapping to better evaluate the potential for breeding within each of the different harvest types. Territory mapping will be conducted throughout the study area on 8 mornings in May and June each year by trained observers who record locations of singing males, pairs, and territorial activities (e.g., conspecific counter-singing and aggressive interactions). Surveys will be conducted between sunrise and 4 hrs after sunrise on mornings without significant precipitation or sustained high winds. We will then delineate territories of each species following Bibby et al. (2000), requiring clusters of ≥2 territory registrations and using territorial behaviors when possible to separate territories. Territory density will be calculated for each harvest unit. Additionally, sampling will be completed each year to characterize the structure and composition of vegetation in each harvest unit. Metrics will include such things as basal area by tree species, snag density, cover in vegetative height classes, shrub cover by species, and ground cover in several classes. Vegetation metrics will be related to relative abundance and territory density of avian species.