.
EFFECTS OF NATURAL GAS DEVELOPMENT ON FOREST ECOSYSTEMS
Mary Beth Adams, W. Mark Ford, Thomas M. Schuler, and Melissa Thomas-Van Gundy1
Abstract.—In 2004, an energy company leased the privately owned minerals that underlie the Fernow Experimental Forest in West Virginia. The Fernow, established in 1934, is dedicated to long-term research. In 2008, a natural gas well was drilled on the Fernow and a pipeline and supporting infrastructure constructed. We describe the impacts of natural gas development on the natural resources of the Fernow, and develop recommendations for landowners and land managers based on our experiences. Some of the effects (forest clearing, erosion, road damage) were expected and predictable, and some were unexpected (vegetation death from land application of fluids, an apparent increase in white-tailed deer presence). Although this is a case study, and therefore the results and conclusions are not applicable to all hardwood forests, information about gas development impacts is sufficiently rare that forest managers, research scientists, and the concerned public can learn from our experience.
INTRODUCTION
The increased demand for natural gas during recent decades, coupled with advances in extraction technology, has led to an increase in exploration and development in many previously unexplored areas in the United States, most notably areas underlain by Marcellus shales in the eastern part of the nation. The impacts of increased development pressure on other natural resources, including soil water, wildlife, and vegetation, are relatively undocumented for forest ecosystems in the eastern United States. Landowners, land managers, and policy makers require such information about these effects to help them make decisions about mineral resource development.
In 2008, drilling for natural gas and subsequent pipeline construction were implemented to extract the privately owned minerals underneath the Fernow Experimental Forest in West Virginia. The Fernow is well known for long-term silviculture, watershed, and ecological research (Kochenderfer 2006). In this paper, we describe some of the impacts of this natural gas development on the natural and scientific resources of the Fernow, and identify opportunities to mitigate possible impacts based on our experiences. This report describes results from a single case study, and the data described herein generally come from post-hoc monitoring, not designed experimentation. In addition, many of the potentially sensitive components of the ecosystem were not monitored; information on fauna, in particular, is lacking. Nonetheless, conclusions can be drawn which may be useful to other land managers.
…
DRILLING AND HYDROFRACING FLUIDS
Fluids from the drill pit were land-applied at two locations on the Fernow in June 2008, with nearly immediate impacts on vegetation. The fluid from the drill pit was sprayed into the air and onto the vegetation, although this application method may not be the standard protocol. The assumption was that if the drill pit fluids met the standards specified in the permit, there would be no damage to the vegetation or soil. Permit levels, as established by the West Virginia Division of Environmental Protection, dictate land-applied drill pit fluids must have chloride concentrations less than 12,500 mg/L and a pH between 6 and 10. State regulations require only one sample of the effluent be taken from the distribution hose during land application. The drill pit fluids met the permit concentration levels, but total vegetation mortality was observed on the first fluid application site almost immediately after land application. Because of this unexpected effect, a second fluid application site was negotiated, and a smaller amount of fluids applied to this site.
At the first fluid application site, many trees, shrubs, and understory plants showed immediate responses to the fluid application, with leaves turning brown, wilting, and subsequent leaf and bud mortality. We also observed that taller trees, whose leaves were not contacted by the fluids, also began showing decline symptoms about 10 days after the ground vegetation; these symptoms included leaf browning, leaf curling, and premature leaf drop. Premature leaf fall ranged from 227 to 1,395 kg ha-1, or about 10 to 45 percent of annual autumn leaf fall biomass (Adams 2008). We inferred from these observations that some of the vegetation was damaged immediately by contact with the drill pit fluids, but most likely many of the larger trees were killed as a result of uptake of the fluid through roots from the soil.
All herbaceous and shrub vegetation within the perimeter of the application area showed damage symptoms in 2008; 115 trees (>2.54 cm d.b.h.) exhibited decline symptoms. In 2009, that number had increased to 147 trees (basal area 3.8 m2). Some recovery of understory vegetation occurred in 2009. Note, however, that more than 50 percent of the trees had no foliage in 2009, and 65 percent of the trees had less than one-third full crown. Mortality was most evident in American beech, with bark sloughing from the bole and branches on 38 percent of the beech trees within the application area perimeter.
Damage symptoms on the second fluid application site were less dramatic (browning of leaves, particularly of northern red oak seedlings), most likely because a much lower volume of fluids was applied, and the operator took care to apply the fluids onto the ground, rather than spraying it onto the vegetation.
We hypothesized that the vegetation on fluid application site 1 was killed as a result of very high salt loading to the site. Although concentrations of chlorides met the permit criteria, an estimated 302,800 L of fluid were applied to this site, resulting in a load of 11,355 kg chlorides per ha. Such a loading far exceeds load limits established elsewhere, such as Oklahoma (450 kg ha-1), Wisconsin (275 kg ha-1 on a 2-year basis;
http://www.dnr.state.wi.us/org/water/wm/ww/gpindex/57665_permit.pdf ), and in Saskatchewan (400 kg ha-1;
http://eps.mcgill.ca/%7Ecourses/c550/Environmental-impact-of-drilling/Sask_Drilling_Waste_Guidelines.pdf ).
Monitoring of soil chemistry over time has confirmed that high levels of chlorides, sodium, and calcium were found in the application areas immediately after the fluids were applied (Fig. 2). Although concentrations have decreased over time, soil chloride and sodium levels were still significantly elevated in May 2009 in the application area relative to the control area.
…