COS 29-5
Avian response to forest disturbance associated with Marcellus shale gas development: A long-term case study of impacts on area-sensitive species

Tuesday, August 11, 2015: 9:20 AM
337, Baltimore Convention Center
Laura S. Farwell, WV Cooperative Fish and Wildlife Research Unit, West Virginia University, Morgantown, WV
Petra B. Wood, U.S. Geological Survey, WV Cooperative Fish & Wildlife Research Unit, West Virginia University, Morgantown, WV
James Sheehan, WV Cooperative Fish & Wildlife Research Unit, West Virginia University, Morgantown, WV
Greg George, WV Cooperative Fish and Wildlife Research Unit, West Virginia University, Morgantown, WV
Background/Question/Methods

In the last decade, extraction of natural gas from the Marcellus shale has increased exponentially in the central Appalachians, a region that contains some of the largest expanses of deciduous forest remaining in the eastern US.  These forests comprise important breeding habitat for many neotropical migrants, including several species of conservation concern.  To better understand the impacts of shale gas development on breeding forest songbirds, we conducted annual point count surveys during 2008-2014, at 142 survey stations in a 4,500 ha study area of primarily mature hardwood forest in northwestern West Virginia.  We also mapped, categorized, and analyzed land cover disturbance throughout the study area on an annual basis using a combination of leaf-off and leaf-on aerial imagery (WV SAMB 2003, NAIP 2007 and 2011), satellite imagery (Landsat 7 and 8, Quickbird 2009), and personal knowledge of disturbance events.  We assessed patterns of avian response to annual changes in land cover at two spatial extents: a broad ‘landscape’ scale (site-level, 4,541 ha), and a local scale within a 300-m radius of each survey station.  We also assessed avian response to general disturbance levels from human activity, at each survey station on an annual basis.

Results/Conclusions

We found that shale gas well pad and infrastructure development resulted in a 4% loss in forest cover at the site (decreasing from 95% to 91% forest), and a 48% increase in forest edge density (increasing from 36 m/ha to 53 m/ha).  Although these changes in forest cover seem relatively minor, we also observed evidence of changes in bird abundance and distribution in response to this disturbance. Generalized linear mixed models revealed significant relationships between species abundances and forest cover, gas disturbance, edge density, and levels of human disturbance activity.  For example, abundance of cerulean warblers (Setopha cerulea) and worm-eating warblers (Helmitheros vermivorum) – both area-sensitive species of conservation concern – declined by 70% and 63% during the seven-year study period, respectively. Cerulean warblers were positively related with forest cover at a landscape scale, while worm-eating warblers were negatively related with disturbance activity at a local scale. Conversely, brown-headed cowbirds (Molothrus ater) – an edge-associated, brood parasitic species – increased 800% over the seven-year period, and were negatively related with forest cover at both a landscape and local scale.  Our results suggest that Marcellus shale gas development has the potential to affect regional forests and area-sensitive avian species, at multiple spatial scales.