palmer lab - palmer lab @ university of maryland - the … · 2019. 9. 26. · ann. n.y. acad. sci....

Ann. N.Y. Acad. Sci. ISSN 0077-8923

ANNALS OF THE NEW YORK ACADEMY OF SCIENCESIssue: The Year in Ecology and Conservation Biology

The environmental costs of mountaintop mining valley filloperations for aquatic ecosystems of the CentralAppalachians

Emily S. Bernhardt1 and Margaret A. Palmer21Department of Biology, Duke University, Durham, North Carolina. 2University of Maryland Center for Environmental Science,Solomons, Maryland

Address for correspondence: Emily S. Bernhardt, Department of Biology, Box 90338, Duke University, Durham, NC [email protected]

Southern Appalachian forests are recognized as a biodiversity hot spot of global significance, particularly for endemicaquatic salamanders and mussels. The dominant driver of land-cover and land-use change in this region is surfacemining, with an ever-increasing proportion occurring as mountaintop mining with valley fill operations (MTVF).In MTVF, seams of coal are exposed using explosives, and the resulting noncoal overburden is pushed into adjacentvalleys to facilitate coal extraction. To date, MTVF throughout the Appalachians have converted 1.1 million hectaresof forest to surface mines and buried more than 2,000 km of stream channel beneath mining overburden. The impactsof these lost forests and buried streams are propagated throughout the river networks of the region as the resultingsediment and chemical pollutants are transmitted downstream. There is, to date, no evidence to suggest that theextensive chemical and hydrologic alterations of streams by MTVF can be offset or reversed by currently requiredreclamation and mitigation practices.

Keywords: stream ecosystems; mountaintop mining; appalachians; water quality; macroinvertebrates

Introduction

“Streams are the gutters down which flow theruins of continents.”

Leopold, Wolman, and Miller (1964)

Most of the world’s great rivers are born in vastnetworks of tiny headwater mountain streams. Eachcontributing stream carries eroded sediments andweathered solutes from its watershed toward thesea. When human activities alter the vegetation, thesoil, and the contours of watersheds for resourceextraction, agriculture, or settlement, we changethe relationship between the stream and its valley.Most types of land-use change result in the increasedmovement of water, sediments, and dissolved chem-icals from the uplands to downstream ecosystems.1

Of all the many varieties of land-use change, noneis as marked and as irreversible as surface mining,which strips away vegetation and surface soils andmechanically alters the underlying geology, leavingbehind a landscape that is fundamentally recon-

toured, requires centuries for the development ofnew soil horizons, and where water is routed alongnew flowpaths, either overland or through uncon-solidated rock spoil toward the draining stream. Un-consolidated rock has vastly increased surface areasexposed to the elements that are susceptible to in-creased rates of physical and chemical weathering;this leads to substantial losses of rock-derived chem-icals from the watershed to downstream ecosys-tems. Mountaintop mining with valley fill oper-ations (MTVF) is a relatively recent approach tosurface mining techniques in which surface miningactually lowers watershed ridges to extract coal andfills watershed streams with the quantities of leftoverrock. Throughout the southern Appalachians of theUnited States, small and steep forested watershedsare being converted to flattened grasslands in or-der to acquire coal. Streams within these watershedsare lost, buried beneath overburden. The impacts ofthese landscape alterations are not contained withinmine boundaries—mined watersheds export largervolumes of water, masses of sediment, and higher

doi: 10.1111/j.1749-6632.2011.05986.xAnn. N.Y. Acad. Sci. 1223 (2011) 39–57 c© 2011 New York Academy of Sciences. 39

Mountaintop mining impacts on aquatic ecosystems Bernhardt & Palmer

Figure 1. (A) Hotspots for at-risk fish and mussel species (from Master et al.7). (B) A Blue Ridge salamander (photo by J.S.Pippen). (C) Hotspots of rarity and richness. (D) Regional concentrations of aquatic biodiversity; both reprinted with permissionfrom Precious Heritage, c©TNC and NatureServe, 2000.

concentrations of a variety of rock-derived chem-icals to downstream rivers. Despite laws designedto regulate environmental effects from coal mining,such as the Surface Mining Control and ReclamationAct (SMCRA) of 1977 (Refs. 2 and 3) and sections402 and 404 of the Clean Water Act, examination ofthe known influences of MTVF on the environmenthas led some to designate the Central Appalachiansas an “environmental sacrifice zone.”4 While legaldebates rage and court cases focused on the permit-ting and regulation of surface mining, particularlyMTVF, continue to grow in number and complexity,our goal is to focus on the environmental impacts.In this review, we will describe how mountaintopremoval mining affects aquatic ecosystems through-out the Central Appalachians.

Aquatic ecosystems of the CentralAppalachians

The forests of the Central Appalachian ecoregionare rich in natural resources, including the produc-

tive temperate forests that characterize the regionas well as rich reserves of bituminous coal. The Ap-palachian mountain region forms the headwatersfor many major U.S. rivers including the Potomac,Susquehanna, and Ohio rivers, which collectivelyprovide water resources to tens of millions of U.S.residents. Coal mining in this region has supportedthe growth of the U.S. energy economy since theearly 19th century, and billions of tons of recov-erable coal deposits are still contained within theregion. In contrast to much of the rest of the east-ern United States, much of this ecoregion remainsforested, and the dominant land-cover change in theregion results from surface mining.5,6

Because the southern Appalachians escapedglaciation, these are among the oldest mountain-ous ecosystems on Earth. As a result of their antiq-uity and their topographic diversity, the landscapeof the southern Appalachians supports among thehighest levels of biodiversity and endemism in thetemperate zone7,8 (Fig. 1). More than 2,000 species

40 Ann. N.Y. Acad. Sci. 1223 (2011) 39–57 c© 2011 New York Academy of Sciences.

Bernhardt & Palmer Mountaintop mining impacts on aquatic ecosystems

Uniquechemistry Refuge from

high flows

Therman refuge

Refuge frompredators

Refuge fromcompetitors

Refuge frominvasive species

Spawning sites

Nursery habitats

Rich feeding areas

Drifting insects

Benthicorganisms

Emerginginsects

Dissolved andparticulate

organic matter

Organic matterstorage

Nutrienttransformations

Exportdownstream

Export

Export

Migrants

Migrants

Migrantsfromdownstream

Figure 2. Factors that contribute to the biological importance of headwater streams in river networks. Attributes on the rightbenefit species unique to headwaters and also make headwaters essential seasonal habitats for migrants from downstream. On theleft are biological contributions of headwater ecosystems to riparian and downstream ecosystems. From Meyer et al.12

of vascular plants have been recorded in the re-gion, and the rivers and streams of the southernand Central Appalachians have been identified asthe most biologically diverse freshwater systems inNorth America.8 The ecoregion is home to 10% ofglobal salamander diversity and freshwater musseldiversity.7,9

The burial of 2,000 km of Central Appalachianstreams thus represents a significant loss of freshwa-ter ecosystem habitat in the region. The role of head-water streams in supporting high levels of biodiver-sity has been emphasized in a great deal of scientificresearch.10–12 These small streams provide habitatsfor a rich array of species, which enhances the bi-ological diversity of the entire river system. Someof the species are unique—that is, the only place ina river network these species occur is in the head-waters.13,14 They provide a refuge from predatorsand changes in temperature for some species15 andcan be important spawning and nursery grounds forothers.16,17 Meyer et al.12 provide a succinct sum-

mary of why the protection of headwater streamsis critical for preserving regional biodiversity, sup-porting downstream food webs, and reducing sed-iment and nutrient loading to downstream waters(Fig. 2).

Invertebrate diversity is particularly high in head-water streams.18 For example, Stout and Wallace14

sampled 23 intermittent Appalachian streams andrecorded 86 insect genera from more than 47 fam-ilies. Amphipods, isopods, copepods, cladocerans,and ostracods are particularly common, and the lat-ter three may reach abundances of >10,000/m.2,19

Meyer et al.12 documented that many of the smallerinterstitial taxa such as rotifers, gastrotrichs, andoligochaetes are both diverse and abundant in head-water streams. Fish can also be abundant in head-waters, and while fish diversity generally increaseswith stream size, many headwater species are notfound elsewhere in the river network—headwatersmay disproportionately contribute to network-widediversity and play a critical role in the genetics of

Ann. N.Y. Acad. Sci. 1223 (2011) 39–57 c© 2011 New York Academy of Sciences. 41


fish populations.12,20 Generally, the species foundin these small streams are small in body size—suchas minnows, darters, and sculpins—but also may in-clude salmonids such as those found in cold NorthCarolina headwaters.21

Where fish are absent in Appalachian streams,amphibians are common and are typically the topaquatic predators. Headwater seeps and ephemeraland intermittent streams provide vital habitat foramphibians, many of which are state and/or fed-erally threatened and endangered.22–25 Salaman-ders are the most common vertebrate in headwa-ters of the region25,26 and appear to be particularlyvulnerable to land-use change. Ford et al.27 docu-mented that diversity and abundance of salaman-ders in the southern Appalachian mountains wasreduced when forests are clear-cut, even after morethan 75 years of regrowth. Loss of salamander popu-lations from headwater streams can have ecosystem-wide consequences since they influence insect pop-ulation dynamics, regulate detritus food webs, andlink stream and terrestrial food webs.23

Stream ecosystem habitat loss through burial anddegradation resulting from mining runoff occursin the headwater streams “where rivers are born”(sensu Meyer et al.).12 Headwater streams are crit-ically important components of river networks be-cause their flow and associated biota, sediment, anddissolved constituents feed all downstream waters.Any major changes affecting headwater tributariesor any activity that isolates or cuts off these trib-utaries from the lower part of the watershed willhave important consequences for hydrologic pro-cesses, sediment delivery, channel morphology, bio-geochemistry, and stream ecology further down-stream in the watershed.28–30 Surface mines andfilled valleys deliver enormous quantities of mining-derived sediments and solutes to downstream rivers,which fundamentally alters the biological commu-nities and the biogeochemical cycling throughoutthe river network.

The coal mining process

Coal and coal impuritiesCoal supplies nearly half of U.S. electricity needsannually.31 Coal minerals are a fossil fuel derivedfrom the lithification of peats produced in wet-lands millions of years ago and that were subse-quently buried under sediments. Coal minerals pri-marily consist of densely packed organic carbon

that, when burned, produces energy. Coal miner-als that have undergone more extensive lithificationare denser and thus have higher energy productionpotential and greater economic value. Coal miningin the Central Appalachians produces high-gradebituminous coal with a high heat value.32 Alongwith the vegetation-derived carbon, coal mineralsare also highly enriched in pyrite minerals producedas a byproduct of sulfate reduction in the anoxic sed-iments of ancient wetlands33 as well as a variety oftrace metals that accumulated in the original plantand microbial tissues. When coal minerals previ-ously buried in bedrock are exposed to air and wa-ter, these pyrite minerals are oxidized, generatingsulfuric acid and iron hydroxides. It is this reactionthat generates the acid mine drainage problems as-sociated with some coal mines. Throughout muchof the Central Appalachians, carbonate parent ma-terial provides extensive internal buffering capacityso that many surface mines generate alkaline minedrainage, with mine runoff having elevated pH thatis rich in the acid anion sulfate (SO42−) and the basecations calcium and magnesium (Ca2+ and Mg2+).The sulfuric acid generated by pyrite oxidation facil-itates rapid weathering of the surrounding rocks andleads to the leaching and export of trace elementsto draining streams. When coal is burned, thesetrace elements can produce significant air qualityproblems—particularly through the production ofthe volatile SOx and NOx gases that generate acidrain34 and the emission of mercury as an airbornepollutant.35

Mountaintop mining with valley fill operationsThe advent of MTVF enabled surface mines to ex-pand considerably in size.36 MTVF practices involvethe removal of the top 50–200 vertical meters ofmountain summits and ridges to extract embed-ded coal strata.37 To access the shallow coal seamsembedded within the ridges, mining companies be-gin by clearing existing forests and then remov-ing the topsoil. A combination of explosives andheavy machinery is then used to loosen and removethe dirt and rock material that covers coal seams(Fig. 3A). Mechanized draglines are then used toremove rubble to clear access to the now exposedcoal seams on the rock face (Fig. 3C). MTVFproduces large volumes of unconsolidated rockwaste and coal debris, often referred to as “overbur-den.” In the steep terrain of Appalachia, the cheapest



Figure 3. Photos of mountaintop mining activity. Panel A shows active and reclaimed portions of the Hobet mine in southernWest Virginia; note the active explosion on the long wall at center left and the protection of the Berry Branch cemetery in the smallforested ridge in the bottom left. Panel B is a photograph of a valley fill draining a portion of the Hobet mine. Panel C shows adragline excavator at work on a long wall. These are the largest mobile machines built today, and they are capable of excavating30–60 metric tons of material with every pass of the dragline bucket. Panel D shows iron precipitates coating the bed of a streamdraining a mountaintop mining operation in Kentucky. Photos A&C are by Vivian Stockman, with the help of SouthWings.org.Photo B was taken by Ty Lindbergh, and Photo D was taken by Ken Fritz.

and most common disposal option for overburdenis to push it into adjacent valleys, creating what arecalled valley fills. Individual valley fills can be hun-dreds of feet wide and more than a mile in length,and each fill buries the headwater streams of theformer valley under tens to hundreds of meters ofoverburden (Fig. 3B). Most of the coal producedin the Central Appalachians is washed prior to ex-traction to reduce its sulfur content, a process thatleaves behind large coal slurry ponds enriched inpyrite minerals and other trace elements leached

from coal.37 Slurry ponds and coal washing facili-ties may be built on site in some mining operations,while other mines transport coal to centralized fa-cilities. Once a surface mine permit area has beendepleted of coal, mining companies are required toperform reclamation activities, including regradingand recontouring the land surface, adding a top-soil substitute, and seeding the area. Regulatory per-mitting requires that the burial of streams beneathvalley fills and coal slurry ponds be mitigated by theconstruction of natural channels or near reclaimed



mines or the restoration of degraded streams withinthe watershed.

Mountaintop mining is widespread across theCentral Appalachian Mountains and is particularlyintense in southern West Virginia, eastern Ken-tucky and Tennessee, and southwestern Virginia.The U.S. Environmental Protection Agency (EPA)predicts that by 2012, mountaintop mining will haveremoved ∼5700 km2 of Appalachian forests andburied ∼3200 km of streams.36 In fact, surface min-ing and mine reclamation activities are currently thelargest driver of land-use change in the Appalachianregion.38

The Central Appalachian region supplies ∼19%of U.S. coal production annually.31 Mountaintopmining is practiced by numerous companies in theregion because it can be done at large scales, allowsaccess to shallow coal seams, and generates greaterprofits than underground mining. The large size ofMTVF mines (they can be many square kilometersin size) and the use of large machinery and explo-sives both reduce the number of workers neededand create economies of scale. Economic analysisshows that MTVF require far fewer workers perton of coal produced than does underground min-ing;31,36 this allows Central Appalachian states tocompete more effectively with western states in theextraction and sale of coal. Surface mining can beeffectively accomplished with heavy machinery runby only a few workers. Indeed, the size of the coalmining workforce has dropped precipitously in theAppalachian region, falling from 171,000 workersin 1980 to 71,000 workers in 1993 without declinesin production.39 The extent of MTVF in the CentralAppalachians increased markedly as a result of the1977 and 1990 amendments to the Clean Air Act tocurtail sulfur emissions,40 because these regulationsincreased demand for the low sulfur coal character-istic of much of the Central Appalachians. MTVFtarget shallow coal reserves, often only a few feetthick, which are typically inaccessible to traditionalunderground mining practices.

The legal and regulatory framework governingMTVFIn the United States, debates over the social and envi-ronmental impacts of MTVF are conspicuous due toseveral high profile federal court cases, widely pub-licized exchanges between the EPA and the ArmyCorps of Engineers (ACOE) over permitting deci-

sions, advocacy by local and national nongovern-mental organizations, and protests by miners relatedto Congressional hearings and pending legislation.Several federal laws and multiple government agen-cies regulate the practice of mountaintop mining.The federal laws that most directly relate to MTVFare the Surface Mining Control and ReclamationAct (SMCRA; 25 U.S.C. 1201) and the Clean Wa-ter Act (CWA; 33 U.S.C. 1252).41 The governmentagencies responsible for administering these lawsare the EPA, the ACOE, and the environmental andnatural resource departments of state governments.Enacted in 1977, the SMCRA requires the regulationand inspection of surface mines and the restorationor reclamation of abandoned mines. The CWA of1972 is the central environmental statute protect-ing U.S. surface waters. The CWA aims to restoreand maintain the physical, chemical, and biologicalintegrity of streams, rivers, and water bodies. TheCWA and its implementing regulations require thatburying streams with discharged materials shouldbe avoided and, if unavoidable, mitigation must ren-der nonsignificant the impacts that mining activi-ties have on the structure and function of aquaticecosystems. Section 404 of the CWA regulates dis-charge of fill material into waterways by establish-ing a permit process administered by the ACOE andoverseen by the EPA. Regulators and environmentaladvocacy groups have interpreted the language ofthe CWA differently, with numerous lawsuits chal-lenging agency permitting.41,42

Environmental impacts

MTVFs have both local and regional effects onaquatic ecosystems. The most obvious impact ofMTVF on aquatic ecosystems is the local destructionof stream segments that are buried beneath valleyfills, mined through or converted to waste treat-ment systems in the form of ponds at the base offills. Streams that have been filled no longer exist—thus, MTVF lead to a net loss of stream habitat andstream and riparian ecosystem function in the wa-tersheds in which it occurs. In addition, MTVF havefar-reaching downstream impacts through the ex-port of sediments and dissolved substances (solutes)from mined watersheds. Furthermore, the removalof vegetation from mined watersheds, the flatten-ing of valley contours, and the compaction of soilon mined sites fundamentally alters the patterns ofwater flow through impacted valleys and changes



the delivery of water and the composition of dis-solved and particulate materials provided to largerreceiving streams.

Effects at the local scaleLost streams. Individual surface mines in the Cen-tral Appalachians can range in size from a few squarekilometers to the 40 km2 Hobet mine in southernWest Virginia, the largest individual surface minein Appalachia. Typically, mining companies applyfor a single permit that allows them to mine coal inmultiple adjacent headwater valleys. The first step ininitiating a new MTVF mine is to convert a portionof the draining stream into a sediment pond thatwill collect sediments generated during the miningoperation. Ultimately, the sediment pond will sit atthe base of a newly generated valley fill. The orig-inal draining stream will continue to receive all ofthe surface runoff generated from the watershed;yet the headwaters will be eliminated; instead, waterwill flow over and through blasted rock and throughthe increasingly thick valley fill and into a settlingpond en route to the stream network. Settling pondsare critical because new mines immediately begin togenerate sediment pollution as mine access roads arebuilt and forests are cleared. The rate of sedimentexport only increases as a result of blasting along theridges and the disposal of rock waste directly intothe stream drainage network.

Lost soils. One of the most severe and long-lastingconsequences of surface mining is the permanentloss and homogenization of the forest soils. Somemining operations attempt to preserve soil and storeit to replace on the mine surface following mining,but even with careful management, a large volume ofsoil will inevitably be lost via erosion, burial withinoverburden, and oxidation by soil microbes. Whatsoil remains will be homogenized so that organicand mineral soils are mixed, significantly alteringsoil horizons that developed over prior centuries.In addition to the significant losses of soil carbonassociated with soil removal,43 soil macro- and mi-cronutrients are also lost. Thus, the surface of areclaimed mine will at best be covered with a thinlayer of very altered soils overlaying crushed andcompacted rock. This change in the vertical profileof soils is coupled with massive changes in topog-raphy at the scale of the catchment, both of whichhave important implications for the hydrology andvegetation structure of the recovering watershed

and its draining stream; these changes persist in-definitely. The landscape left behind after surfacemining operations cease must undergo a primarysuccession sequence similar to that following glacialretreat.44–46 Many decades are required for foreststo reestablish in the thin or nonexistent soils overnewly exposed rock (Fig. 4B). In addition to thechallenge of growing in thin soils, vegetation mustestablish roots in alkaline soils, which can inhibittree growth.47,48 Without thick soils and sufficientnutrients, there is little evidence to suggest that na-tive trees characteristic of the region can be culti-vated or maintained on reclaimed mines.36,43 Thereare few peer-reviewed reports of forest growth onreclaimed mine sites in the Central Appalachians,and their conclusions are divergent. Several stud-ies49,50 have reported limited woody tree recruit-ment and growth in a series of MTVF sites reclaimedunder the SMCRA, while Rodrigue et al.51 andAmichev et al.52 estimated similar levels of for-est productivity between 14 surface mine sitesreclaimed prior to the 1977 enactment of theSMCRA and adjacent unmined stands. Revegetationis required under the SMCRA, but this has typicallyinvolved hydroseeding the recontoured landscapewith a mixture of grasses44 so that forest regrowthdepends upon dispersal and recruitment. Recentstudies demonstrate that the forests and grasslandsgrowing on reclaimed surface mines sequester andstore far less carbon than unmined sites, both invegetation biomass and in the soil.43,48,53

Altered hydrology. As a result of reduced vegeta-tion and evapotranspiration, surface-mined water-sheds typically have higher annual water yields thantheir forested counterparts.54 Mining also leads tosignificant changes in watershed hydrographs.55–57

Recent work analyzing hydrological changes as afunction of the amount of a watershed that hasbeen surface mined showed that peak flows increaselinearly with the percent of the watershed mined,even if the land has been reclaimed.58 These hy-drologic changes persist because reclamation usingearth-moving machinery compacts soil layers thusdecreasing porosity and water infiltration.59 Com-pared to regions that have not been mined, infil-tration may be as much as an order of magnitudelower60,61 so that mined sites respond to rainfallmore like urban watersheds, where impervious sur-faces lead to high surface runoff during storms.58



Figure 4. At top a photo of the (A) unmined landscape of the central Appalachians juxtaposed against (B) a photo of a reclamationsite on the Wind River mine after more than 20 years of reclamation. Both photos by Vivian Stockman with the help of South-Wings.org. At bottom (C) a photo of a typical central Appalachian stream juxtaposed against (D) a photo of a created stream builtto mitigate for stream burial by mining overburden. Photo C by Jack Webster. Photo D excerpted from a Technical memorandum:Ecological Functions in Created Stream Channels Prepared by Ecology and Environment, Inc. Source caption reads: “Example of astream creation project 15 years post-creation (NPDES Outlet). Nicholas County, WV.”

Longitudinal impactsThree fundamental scientific principles are criticalto understanding why cumulative impacts of MTVFon downstream aquatic resources are so impor-tant. First, changes at the watershed scale influencestream hydrology throughout the catchment.62 Thetiming and magnitude of stream flows result fromcomplex interactions between rainfall, plants, topo-graphic relief, and soil properties of all land above adrainage point.63 Once vegetation is lost and soil iscompacted, as occurs on mined and even reclaimedmine land, hydrological changes negatively affectstream biota and water quality.53,58 Second, streamwater chemistry is shaped by processes that occur asrainwater infiltrates the ground and moves throughpore spaces and soil on its way to streams.64,65 Wateremerging from valley fills carries with it dissolvedconstituents that are toxic or damaging to biota (dis-cussed extensively later in this review). Third, thedownhill movement of water and one-way flow in

stream networks means that whatever happens onland or in first- and second-order streams (headwa-ters) not only determines sediment and water flowin the immediately impacted streams and rivers,but also determines ecological structure and func-tions of larger waterways into which these tributariesflow.66

Therefore, individual mines not only profoundlyaffect stream water quality, community structure,and ecosystem functions immediately downstreamof valley fills (Fig. 3D), but multiple mining opera-tions within larger watersheds have cumulative ef-fects on larger downstream rivers through increasedloading of dissolved substances derived from alka-line mine drainage.

Sulfate. Just as coal burning for power generationproduces sulfur aerosols (SOx), the exposure of coalseams during coal mining provides many opportu-nities for the leaching of sulfate (SO42−) from coal



wastes into surface waters. Unlike SOx emissionsthat distribute sulfur aerosols regionally, miningactivities lead to a localized point source of SO42−

to the drainage network. As a consequence of re-gional SOx emissions, freshwater systems through-out North American and Europe have had tenfoldor greater increases in SO4− concentrations. As a re-sult of mining activities, impacted streams in WestVirginia often have 30- to 40-fold increases in SO42−

concentrations,67,68 with some streams reported tohave SO4− concentrations higher than found in sea-water.69 The relationship between mining activitiesand high sulfate concentrations is so well establishedthat the 2008 WVDEP West Virginia Integrated Wa-ter Quality Monitoring and Assessment Report sug-gested that SO42− concentrations >50 mg L−1 couldbe used as an indicator of mining activity.70

The headwater mountain streams of the CentralAppalachians that are being impacted by mountain-top mining were historically dilute and oligotrophic.Earlier studies in major watersheds of West Virginiadirectly linked increases in river sulfate load to in-creasing coal production in the watershed2 and usedtime-series analysis to show that sulfate concentra-tions in streams continue to increase after miningactivities end.2 Likewise, a U.S. Geological SurveyNational Water-Quality Assessment Program studyfound that in the Kanawha–New River Basin, totaliron (Fe) and manganese (Mn) decreased in streamsdraining watersheds with ongoing coal productionbetween 1991 and 1998, while sulfate concentra-tions continued to increase.57 Both of these studiesdocumented an increase in SO42− loading to majorriver systems that corresponded to increases in coalextraction within their watersheds.

This fundamental change in the chemistry ofheadwater streams can have important local- andwatershed-scale impacts on aquatic organisms andecosystem functions. Elevated sulfate concentra-tions will stimulate microbial sulfate reduction instream and wetland sediments. In this reaction,microbes use SO4− in place of oxygen in their con-sumption of organic matter. The sulfur (S) in sul-fates is converted to sulfide (HS−)—this is the re-action that gives salt marshes and wetlands theircharacteristic rotten egg (sulfurous) odor:

SO2−4 + 2CH2O → HS− + HCO−3 + CO2 + H2O.As sulfate concentrations increase, the produc-

tion of sulfide also increases, and this has important

implications for the receiving ecosystems. Sulfideis directly phytotoxic to many aquatic plants.71–74

Elevated sulfide also has important biogeochemi-cal impacts. Sulfide binds strongly with Fe in sedi-ments, converting it to pyrite minerals. While thishas positive benefits in terms of reducing Fe con-centrations in sulfate-rich mine drainage, it alsohas implications for nutrient pollution. Much ofthe phosphorus (P) present in freshwater ecosys-tems at any given time is bound to iron minerals.In environments with high sulfide concentrations,sulfide interferes with the Fe–P bonds and P is re-leased to the water column.75,76 By this mechanism,sulfate additions can lead to the eutrophication offreshwater streams, wetlands, and lakes without anyadditional nutrient loading (so-called indirect orinternal eutrophication).74 High sulfate loading canalso make freshwater ecosystems more sensitive tonutrient pollution by preventing abiotic reactionsfrom sequestering P in inaccessible forms in thesediments. High sulfide can also inhibit nitrification(the process by which ammonium is converted to ni-trate) in sediments and thereby dramatically reducedenitrification rates—again contributing to a re-duced nitrogen removal efficiency within S-pollutedsediments and promoting or enhancing nitrogeneutrophication.77

Co-occurring contaminants. While an increase insulfate loading is the most predictable and well-documented water quality consequence of moun-taintop mining in the Appalachians, many othersubstances are released to surface waters as a resultof mining activity. Throughout much of the CentralAppalachians, the presence of significant carbon-ate and base cations in parent material neutralizesthe acidity of sulfuric acid released from weatheredcoal deposits, but the neutralization leads to dra-matic increases in Ca2+, Mg2+, and HCO3− ions.This natural acid buffering potential can lead toan increase in the pH of receiving streams (ratherthan the more well understood acidification as-sociated with acid mine drainage). The release ofthese ions contributes to dramatic increases in theelectrical conductivity and total suspended solidswithin the water column of receiving streams.67 Ananalysis of all small streams in West Virginia sam-pled by the West Virginia Department of Environ-mental Protection found that sulfate concentrationswere highly correlated with conductivity, Ca, Cl,Fe, Mg, and Hardness—all of which contribute to



heightened ionic stress in these impacted streams.69

The abundance of Al, Mn, and Se was also found toincrease with SO42− concentration.69

Elevated conductivity. Recent studies67,68,78 foundthat specific conductivities in streams below in-dividual valley fills were 2–10 times higher thanfound in nearby unmined, reference streams.Typical specific conductance levels in low-orderstreams in West Virginia ranged from 13 to253 �S/cm,68,79 while valley fill-impacted streamswere found to far exceed these values (502–2540 �S/cm).67,68

For many streams, it is the cumulative or additiveimpact of elevated concentrations of multiple stres-sors that leads to biological impairment, and this isundoubtedly a part of the reason that conductivity(a cumulative measure of ionic strength) is such aneffective predictor of biological impairment.80 Theionic stress associated with high conductivity canhave direct toxicity as well as providing an indica-tion of the additive impacts of a variety of solutes.High conductivity can be directly toxic to aquaticorganisms by disrupting osmoregulation.81,82 Thisis particularly important for aquatic insects withhigh cuticular permeability. Mayflies in particularare highly sensitive to ionic stress as they regulatetheir ion uptake and release using specialized struc-tures within their gills, their integument, and inter-nally via Malphigian tubules.83 For these sensitivetaxa, large increases in certain ions can disrupt wa-ter balance and ion exchange processes and causestress or death. Tests for conductivity toxicity formayflies have often proved inconclusive,82,84–87 yetthese studies are typically based on lab toxicologicaltests with hardy organisms that are easy to rear inlab settings (i.e., Hexagenia, Centroptilum, Cloeon,Isonychia) and are likely to be less sensitive than themayfly genera that appear especially susceptible toionic stress (e.g., ephemerellids, heptageniids).68,88

Rather than being directly lethal, high conductivitymay encourage sensitive taxa to drift downstreamfrom polluted reaches89—an effect that would notbe measured in the closed vessels of laboratory trials,but that could strongly alter community composi-tion in the field.

The clear patterns linking high conductiv-ity to a loss of mayfly (Ephemeroptera) taxahas ecosystem-scale importance since mayflies of-ten account for 25–50% of total macroinverte-

brate abundance in the least-disturbed CentralAppalachian streams.68,80 The finding that en-tire orders of benthic organisms are nearly elim-inated in MTVF-affected streams suggests thatalkaline mine drainage is fundamentally chang-ing the structure of aquatic macroinvertebratecommunities.68,80,90

It is widely recognized that individual contami-nants rarely exist alone, and although many ecotoxi-cological studies examine the impacts of single con-taminants on laboratory organisms, it is the actualcombined toxicity of constituents in field settingsthat is of interest.91,92 In cases where an associationof contaminants is well characterized (e.g., the tracemetals and cations associated with alkaline or acidmine drainage or the road runoff associated withhigh-traffic volume corridors), a concentration-addition method should be applied to assess theircumulative impact.92–95 Classic laboratory toxico-logical tests have largely failed to find toxic effectsof any single component of alkaline mine drainagepollution. In contrast, the weight of field evidencesuggests that mining activities in watersheds oftendegrade downstream water quality and lead to dra-matic alterations in macroinvertebrate communitystructure.36,68,69,80,90 Mine sites may vary consider-ably in the extent to which they affect legally regu-lated solutes in downstream waters, yet the valley filloperations studied to date are clearly causing height-ened conductivity and high SO42− concentrations.These increases in conductivity and sulfate are as-sociated with a loss of sensitive macroinvertebratetaxa from affected stream reaches.68,69,80,90 Cur-rent statistical and empirical work suggests that forCentral Appalachian streams, conductivities above300 �S cm−1 represent a threshold for sensitive in-vertebrate taxa.68,80

All available data show that it becomes increas-ingly unlikely to find an unimpaired aquatic benthiccommunity as conductivity increases.36,68,69,80,96

As conductivity (and the associated SO42−, Ca2+,Mg2+, HCO3−, and trace metals) increases in Cen-tral Appalachian streams, the biological communitybecomes less diverse and sensitive species (especiallyEphemerellidae and Heptageniidae mayflies) arelost. High conductivities and high sulfates can per-sist long after mining activities cease,2,57,68 and thereis no empirical evidence documenting recovery ofmacroinvertebrate communities in the streams im-pacted by alkaline mine drainage.



Effects at the landscape scaleThe EPA predicts that by 2012, an area over5600 km2 of Appalachian forests will have been im-pacted or destroyed by mountaintop removal min-ing (see Ref. 36 EIS, Appendix I). Surface miningis converting the large expanses of interior for-est found in Central Appalachia into a fragmentedpatchwork of smaller forest patches with a higherproportion of forest edge habitat. In numerous stud-ies examining the impacts of urban development oragricultural intensification on streams, authors havesuggested that land-cover alteration affecting morethan 10% of the watershed area leads to significantlosses in stream biodiversity. To date, there have beenno studies that explicitly link the cumulative extentof mountaintop mining to the water quality and bi-ological community structure of impacted streams.Such information is critically important for guid-ing regulatory decision making. New research byPetty et al.90 suggests that mining severity (prox-imity to stream and extent of mining) is tightlylinked to degradation of steam biological com-munities providing strong evidence of cumulativeimpacts.

Reclamation, remediation, and mitigationof MTVF

The CWA 404 guidelines mandate permit applicantsto avoid, minimize, and mitigate impacts on thewaters of the United States to prevent significantdegradation of the nation’s waters. When impactsare unavoidable, stream habitat and functions lostthrough mining and filling are subject to ameliora-tion through mitigation. Because mitigation actionsmust replace lost stream resources and ecologicalfunctions, the value of those natural resources andfunctions must be assessed prior to their loss. Thisvalue is then used to determine how much mitiga-tion is required.

Compensatory mitigation for MTVF generallyincludes efforts to create streams on the mining siteand sometimes to also restore or enhance degradedstreams in nearby watersheds. The SMCRA regula-tions require that watersheds impacted by surfacemines be returned to a semblance of their formershape at the conclusion of mining activities. Yetthe resulting manmade valleys are flattened, hydro-logically altered grasslands with rock-lined drainsflowing downslope. It is often these former drainsthat are restored to mitigate for newly permitted

stream burial even though the hydrology and thevegetation of the reclaimed mines are very differentfrom their premining state. Stream creation propos-als typically involve regrading mined land and re-configuring mine drains or building new channelsthat are similar in physical form (e.g., width, depth,sinuosity) to the streams that were buried (Fig. 4C& D). Structural complexity of these channels maythen be increased by adding wood or arrangingrocks to help create physical habitat. Assessmentsdeveloped by the regulatory authority are then re-quired; these usually involve a focus on the physicalstructure of the stream even though structural andfunctional assessments are legally required. Streammitigation efforts that meet a series of simple re-quirements for stream channel shape, streambedcharacteristics, and riparian vegetation survival aredeemed adequate for mitigation. Mitigation creditcan be earned for restored streams regardless ofwhether the constructed channels hold water forany time during the year, support any aquaticorganisms, or perform any aquatic ecosystem func-tions.97,98 The compensatory mitigation assess-ments98 required by the ACOE have been success-fully challenged in district court, but these rulingshave been overturned upon appeal (see summary inRef. 97).

Perhaps the most fundamental problem with thecurrent regulatory framework is that permittingagencies fail to consider the cumulative impacts ofissuing multiple mining permits in the same basin,and the impacts of mining outside of the permitboundaries are considered to be minimal, or to betaken care of through the National Pollutant Dis-charge Elimination System (NPDES) permitting.NPDES permit violations in the region are frequent,and when fines are levied, the costs are minimal rel-ative to mining profits.99

Deficiencies of current mitigation practices

Failure to address ecosystem functionAs described above, healthy streams are living, func-tional systems that support a number of criticalecological processes including the processing ofnutrients; the decomposition of organic matter;and microbial, primary, and secondary produc-tion.100–102 To date, mitigation plans associated withmountaintop mining have not used readily availablemethods for directly measuring ecological func-tions; however, these processes must be measured



in order to determine how and whether they maybe brought back to the right levels and directionthrough mitigation. There are now abundant sci-entific studies outlining how to make and interpretsuch measurements103,104 and how such measure-ments should be used to evaluate the success of arestoration project.105–109

The use of well-accepted methods for measur-ing ecological functions110 is important becauseecological functions evaluate dynamic propertiesof ecosystems that underlie an ecosystem’s abilityto provide vital goods and services.102,103,108,111,112

Functions reflect system performance, and theirmeasurement requires quantification of ecologicalprocesses such as primary production or nutrientuptake.110 This should be reflected in the mitiga-tion plan if the plan is to mitigate functions that arelost due to the mining through or filling streams.Functional measures have been used to compare de-graded versus restored versus reference streams inmany settings,105–107,113,114 but research on ecologi-cal functions of created or restored MTVF streamsis in its infancy.115

Stream creation versus stream restorationMTVF projects destroy healthy streams that are of-ten in minimally disturbed, forested watersheds.Many mitigation plans propose to regrade the landand then construct a channel that has similar di-mensions (width, depth, slope, sinuosity, etc.) tothe one destroyed. Thus, the goal is to create astream; however, all the natural flow paths andlandscape topography have been destroyed. As “ev-idence” that stream creation is a routine practice,mitigation plans often cite projects that are actu-ally channel reconfigurations or projects that havespatially shifted a section or meander of a channel—these are not the same as stream creation because,for most restoration projects, the natural flow pathsare still intact. In practice, ecological stream restora-tion varies along a continuum from removing on-going impacts to a stream (e.g., preventing toxicinputs) and letting the system recover on its own;to enhancing habitat in the stream or its surround-ing riparian zone (e.g., adding coarse woody de-bris to streams and planting vegetation); to full-scale restoration that involves manipulations ofan existing stream channel (e.g., regrading banksand planting trees along a stream with erodingbanks).116–118

Some mitigation plans refer to channel creationprojects as “restoration” or “reconstruction,” dra-matically broadening the definition of restorationwith no evidence that historic ecological conditionscan actually be restored,119 but while a constructedchannel may occupy the same map coordinates as apreviously filled stream, virtually every aspect of thisnew channel and its watershed will be dramaticallyaltered by surface mining operations. In particu-lar, hydrologic connections between the watershedand its stream will be profoundly disrupted. As aresult, built channels may not receive, contain, orexchange surface waters for any substantial periodof time, and what water makes it to the channelwill likely carry with it high sediment loads andsubstantially increased solute concentrations. Evenwell-constructed channels are thus highly unlikelyto provide habitat that mimics or matches unminedstreams in the region.

Between 2002 and 2005, we developed thefirst comprehensive database on stream and riverrestoration for the United States.120,121 From our ex-tensive work with scientists and restoration practi-tioners,122 we do not know of a single case (of 38,000projects in our database) in which stream creation,as proposed by many MTVF compensatory miti-gation plans outlined, has been shown to recreatestream hydrology, much less to support the eco-logical functions lost when streams are filled. Con-trary to suggestions made in many mitigation plans,the very concept of creating streams with levels ofecological functioning comparable to natural chan-nels on sites that have been mined through remainsuntested.123 No peer-reviewed scientific studies doc-ument a case where the critical features of streamecosystems (flow, biota, and nutrient and materialprocessing) have been created in a surface-minedlandscape. Even with far less damage to their water-shed, stream restoration projects that involve chan-nel modification have a low probability of promot-ing recovery of geomorphic processes or biologicalcommunities.124–126

Structure and function are not equivalentMost mine permit compensatory mitigation plansthat include stream enhancement or restoration arebased on the Natural Channel Design (NCD) ap-proach,127 but the NCD approach to stream restora-tion is not an ecological restoration approach andit has never been shown to promote ecological



recovery. In fact, results from recent studies pointin the opposite direction.126 Evidence to date sug-gests that extensive channel engineering, which istypical of the NCD approach, may in fact causedamage to streams in need of restoration; for ex-ample, species diversity may actually decrease fol-lowing restoration and may decrease over time.128

The NCD approach is fundamentally focused onchannel form (structure), not ecological function.This approach was designed by David Rosgen127 toaddress channel stability based only on building achannel structure (shape, slope, etc.) that is able totransport the sediment and water inputs that areexpected to be delivered to the stream. There is noscientific evidence supporting the assumption thatrestoration of channel form will lead to full restora-tion of function.111,129,130 How a stream looks (itsform) is simply not the same as how it processesmaterials and supports life (its function).

Most MTVF stream mitigation plans assume thatselection of a channel type from a channel clas-sification scheme such as the NCD approach willnecessarily result in full ecological restoration, butthey also assume that use of the NCD approachguarantees successful creation of a channel from ageomorphic and hydrologic perspective. However,channel designs based on a classification system thathas not been fully evaluated at the site can lead toserious failures,125,131 because the specific processesand history of the river under study were not ade-quately understood. If mitigation projects fail andchannels are unstable, this could cause new envi-ronmental degradation. Even if created channels aregeomorphically stable, this cannot be taken as evi-dence that ecological functions have been restored.Indeed, the NCD scheme of classification does notdeal with ecological functions at all.

While use of the NCD scheme for stream restora-tion has been very common in the past, current sci-ence has documented numerous reasons that use ofthis scheme for restoration can be extremely prob-lematic.125,131–138 The NCD method in no way takesinto account a whole array of biophysical factors thatdetermine the ability of the channel to support all ofthe living resources in pristine streams in the area.Such factors include the intensity and duration ofsunlight reaching the stream, which is determinedin part by the vegetative structure; inputs of organicmatter upon which the food web depends; and onnitrogen and carbon levels in the soil and streambed.

In fact, an analysis of > 75 channel reconfigu-ration projects overwhelmingly showed that theseprojects failed to restore or even markedly changestream biodiversity.128 The fundamental problemwith classification-based restoration approaches isthat they assume fixed end points and rigid classifi-cation schemes in which the type of stream desiredcan be achieved by constructing a specific channelform.

Existing mitigation approaches fail to include anymechanisms that will reduce the export of SO42−,HCO3−, Ca2+, Mg2+, Fe, and trace metals frommined sites, or that will remediate these impacts forthe water columns of constructed channels. Regard-less of the structural similarity between constructedchannels and reference sites, if the water flowingthrough these mitigated channels comes into con-tact with overburden, it will contain the charac-teristic signature of alkaline mine drainage. Thus,the capacity for even a channel that is structurallyand hydrologically similar to reference streams tosupport a diverse aquatic fauna and an ecosystemfunctional capacity similar to those lost when un-mined streams are buried will be limited due to thesevere water quality degradation associated with wa-ter flowing through mined landscapes. The mitiga-tion projects associated with MTVF are not designedto actually mitigate for the water quality impactsgenerated, and these long-term, long-distance im-pacts represent unmitigated stressors to the streamreaches below valley fills and to the full river networkextending downstream.

Importance of ephemeral and intermittentstreamsOften, mitigation plans propose to mitigate forthe burial of ephemeral or intermittent headwa-ter streams through the restoration or creation ofperennial streams. While it may be tempting toview intermittent streams as harsh environmentsbecause they are dry during parts of the year,diverse assemblages are adapted to such condi-tions and may recover from flood and droughtdisturbances more quickly than biota only foundin perennial reaches.139 Headwater streams con-tribute to the aquatic ecosystem in important waysthat make them different from perennial streams.Intermittent and ephemeral streams have uniquecharacteristics that distinguish them from peren-nial streams.139–141 The most obvious difference is



hydrological—surface water is only present part ofthe year, and this attribute leads to the supportof unique species and characteristic communities oforganisms that would not exist if flow were peren-nial. In particular, intermittent and some ephemeralstreams provide unique habitat for a diverse popu-lation of insects and other animals, from macroin-vertebrates to salamanders.142 Because intermittentand ephemeral streams have a seasonal mosaic ofhabitat types, they typically support fauna that maybe poor competitors with the taxa found in peren-nial reaches.143

The interaction of groundwater and surface wa-ter that takes place in the uppermost stream seg-ments within river networks helps regulate down-stream water quality, affecting both aquatic life andwater quality below. Intermittent and ephemeralstreams are critical to biogeochemical processes thathave watershed-scale impacts, and streams that gothrough wet and dry cycles may support high ratesof denitrification.144 Furthermore, intermittent andephemeral streams supply water and sediments thatare important to downstream perennial reaches.143

Finally, these smallest of streams act as a link be-tween terrestrial ecosystems and perennial reaches,and when they are rewet following dry periods,the inundation of dry organic matter (especially inforested region) may release large amounts of dis-solved organic matter to fuel downstream ecosys-tems.143

Altered energetics in constructed channelsThe energetic basis of the stream food web of moun-tainous Appalachian streams is leaf litter from thesurrounding trees.30,145 For most of the year, bac-teria, fungi, and aquatic insects consume the leavesand wood that fall or are washed into the streamfrom the surrounding forest.146 There may be briefperiods of the year (between snowmelt and leafout and between autumn litterfall and first snow)when aquatic plants (algae) are important food re-sources. Constructed channels on or below valleyfills are in high-light environments with early vege-tation consisting primarily of short-stature grasses.Thus, while the forested streams in the region aretypically fueled by leaf litter from the surroundingforest, created streams will likely be fueled by algalproduction.147 Without a forest canopy, water tem-peratures in the constructed channels will be signif-icantly hotter in summer and significantly colder inwinter than in the forested streams.

Furthermore, there is no evidence provided thatdiversion of water flow to ditches or low-lying pointscreates a stream. Subsurface and surface flow pathsto natural streams may be complex, and the resi-dence time of the water in the groundwater variesbefore it reaches streams.148–150 Without a thor-ough scientific study including hydrological anal-ysis of groundwater, surface water, and hyporheicinteractions (rates of flow and flow paths), thereis no evidence that the water resources left af-ter the mining and mitigation will compensate forwhat was lost. Yet there is abundant scientific evi-dence that these hydrologic interactions determineecosystem functions including rates of whole streammetabolism, nutrient processing, organic matter de-composition, productivity, and reproduction of in-vertebrates and fish.151–153 Successful restoration re-quires that key processes and linkages beyond thechannel reach (upstream/downstream connectivity,hillslope, floodplain, and hyporheic/groundwaterconnectivity) also be considered.154–159 The impor-tance of these linkages is without question; water,sediment, organic matter, nutrients, and chemicalsmove from uplands, through tributaries, and acrossfloodplains at rates and concentrations that down-stream organisms have evolved to depend upon.

Conclusions

In summary, the environmental impacts of MTVFsin the Central Appalachians are severe, large scale,and long lasting. In addition to the permanent burialand loss of headwater streams directly impacted bymining, many additional river miles are being de-graded by the cumulative impacts of altered flowsand increased pollutant from both past and presentmining activities in the region. Whether or notindividual component ions within mining-derivedrunoff reach streamwater concentrations that areindividually lethal or toxic to aquatic life, the cu-mulative effect of elevated concentrations of mul-tiple contaminants is clearly associated with a sub-stantial reduction in water quality and biologicalintegrity in streams and rivers below mine sites.All research to date indicates that conductivity is arobust measure of the cumulative or additive im-pacts of the elevated concentrations of multiplechemical stressors from mine sites that lead to bi-ological impairment of streams. Each constituentpollutant increases conductivity and they may haveadditive or multiplicative ecological impacts. To



date, mitigation practices and restoration effortshave not been effective in ameliorating water pol-lution from MTVF sites. Furthermore, efforts toreclaim vegetation and restore the full diversity ofplant species in mined watersheds have not provedsuccessful to date.

Conflicts of interest

The authors declare no conflicts of interest.

References

1. Allan, J.D. 2004. Landscapes and riverscapes: The influenceof land use on stream ecosystems. Annu. Rev. Ecol. Evol.Syst. 35: 257–284.

2. Sams, III, J.I. & K.M. Beer. 1999. Effects of Coal-MineDrainage on Stream Water Quality in the Allegheny andMonongahela River Basins—Sulfate Transport and Trends.(Water-Resources Investigations Report 99–4208) Re-trieved September 15, 2010, from http://pa.water.usgs.gov/reports/wrir 99-4208.pdf.

3. Surface Mining Control and Reclamation Act, 30 U.S.C.1201–1328, 91 Stat. 445. (1977).

4. Fox, J. 1999. Mountaintop removal in West Virginia: anenvironmental sacrifice zone. Organ. Environ. 12: 163–183.

5. Hooke, R.L. 1999. Spatial distribution of human geomor-phic activity in the United States: comparison with rivers.Earth Surface Processes and Landforms 24: 687–692.

6. Saylor, K.L. 2008. Land Cover Trends Project: Cen-tral Applachians. U.S. Department of the Interior, U.S.Geological Survey. Washington, DC. Retrieved fromhttp://landcovertrends.usgs.gov/east/eco69Report.html

7. Master, L.L., S.R. Flack & B.A. Stein. 1998. Rivers of Life:Critical Watersheds for Protecting Freshwater Biodiversity.The Nature Conservancy. Arlington, VA.

8. Stein, B.A., L.S. Kutner & J.S. Adams. 2000. Precious Her-itage: The Status of Biodiversity in the United States. OxfordUniversity Press. New York.

9. Green, N.B. & T.K. Pauley. 1987. Amphibians and Reptilesin West Virginia. University of Pittsburgh Press. Pittsburgh,PA.

10. Lowe, W.H. & G.E. Likens 2005. Moving headwater streamsto the head of the class. Bioscience 55: 196–197.

11. Meyer, J.L. 2003. Where Rivers Are Born: The ScientificImperative for Defending Small Streams and Wetlands.American Rivers and the Sierra Club, Washington,D.C. Retrieved September 1, 2010, from http://www.americanrivers.org/library/reports-publications/where-rivers-are-born.html.

12. Meyer, J.L., D.L. Strayer, J.B. Wallace, et al. 2007. The con-tribution of headwater streams to biodiversity in river net-works. J. Am. Water Resour. Assoc. 43: 86–103.

13. Erman, N.A. & D.C. Erman. 1995. Spring permanence,Trichoptera species richness, and the role of drought.J. Kans. Entomol. Soc. 68: 50–64.

14. Stout, B. & J.B. Wallace. 2005. A Survey of Eight AquaticInsect Orders Associated With Small Headwater StreamsSubject to Valley Fills from Mountaintop Mining. Mountain-

top Mining/Valley Fills in Appalachia: Final ProgrammaticEnvironmental Impact Statement. US EnvironmentalProtection Agency. Washington, DC. Retrieved fromwww.epa.gov/region3/mtntop/pdf/Appendices/Appendix%20D%20Aquatic/StoutWallaceMacroinvertebrate.pdf.

15. Franssen, N.R., K.B. Gido, C.S. Guy, et al. 2006. Effects offloods on fish assemblages in an intermittent prairie stream.Freshwater Biol. 51: 2072–2086.

16. Wigington, P.J., J.L. Ebersole, M.E. Colvin, et al. 2006. Cohosalmon dependence on intermittent streams. Front. Ecol.Environ. 4: 513–518.

17. Erman, D.C. & V.M. Hawthorne. 1976. Quantitative im-portance of an intermitten stream in spawning of raimbowtrout. Trans. Am. Fish. Soc. 105: 675–681.

18. Clarke, A., R. Mac Nally, N. Bond, et al. 2008. Macroinver-tebrate diversity in headwater streams: a review. FreshwaterBiol. 53: 1707–1721.

19. Galassi, D., P. Mermonier, M.J. Dole-Olivier, et al. 2002. Mi-crocrustaces. In Freshwater Meiofauna: Biology and Ecology.A.L. Rundle, S.D. Roberston & J.M. Schmid-Araya, Eds.:135–175. Backhuys Publishers. Leiden.

20. Burridge, C.P., D. Craw, D.C. Jack, et al. 2008. Does fishecology predict dispersal across a river drainage divide?Evolution 62: 1484–1499.

21. Moyle, P.B. & B. Herbold. 1987. Life history parameters andcommunity structure. In Stream Fishes of Western NorthAmerica: Comparisons with Eastern North America. W. J.Matthews & D. J. Heins, Eds.: 25–32. University of Okla-homa Press. Norman, OK.

22. Beachy, C.K. & R.C. Bruce. 1992. Lunglessness in Plethon-dontid Salamanders is Consistent with the Hypothesisof a Mountain Stream Origin–a response to Ruben andBoucout. Am. Nat. 139: 839–847.

23. Petranka, J.W. 1998. Salamanders of the United States andCanada. Smithsonian Institution Press. Washington, DC.

24. Reid, L.M. & R.R. Ziemer. 1994. Evaluating the Bio-logical Significance of Intermittent Streams. USDA For-est Service, Pacific Southwest Research Station. RetrievedSeptember 1, 2010, from http://www.rsl.psw.fs.fed.us/projects/water/2IntermitStr.htm.

25. Davic, R.D. & H.H. Welsh. 2004. On the ecological roles ofsalamanders. Annu. Rev. Ecol. Evol. Syst. 35: 405–434.

26. Johnson, B.R., J.B. Wallace, A.D. Rosemond, et al. 2006.Larval salamander growth responds to enrichment of anutrient poor headwater stream. Hydrobiologia 573: 227–232.

27. Ford, W.M., B.R. Chapman, M.A. Menzel, et al. 2002. Standage and habitat influences on salamanders in Appalachiancove hardwood forests. For. Ecol. Manage. 155: 131–141.

28. Wipfli, M.S. & D.P. Gregovich. 2002. Export of invertebratesand detritus from fishless headwater streams in southeast-ern Alaska: implications for downstream salmonid produc-tion. Freshwater Biol. 47: 957–969.

29. Wipfli, M.S., J.S. Richardson & R.J. Naiman. 2007. Ecolog-ical linkages between headwaters and downstream ecosys-tems: Transport of organic matter, invertebrates, and wooddown headwater channels. J. Am. Water Resour. Assoc. 43:72–85.

30. J.R. Webster, E.F. Benfield, T.P. Ehrman, et al. 1999. Whathappens to allochthonous material that falls into streams? A



synthesis of new and published information from Coweeta.Freshwater Biol 41: 687–705.

31. USDOE. 2010. Annual Coal Report 2009 (DOE/EIA 0584)U.S. Energy Information Administration. Coal Productionin the United States – An Historical Overview. http://www.eia.doe.gov/cneaf/coal/page/acr/acr sum.html.

32. USDOE. 1996. U.S. Coal Reserves: A Review and Up-date. (DOE/EIA-0529(95)). Energy Information Ad-ministration, U.S. Department of Energy. RetrievedSeptember 1, 2010, from http://tonto.eia.doe.gov/ftproot/coal/052995.pdf.

33. Howarth R.W. & J.M. Teal. 1980. Energy flow in a salt marshecosystem: the role of reduced inorganic sulfur compounds.Am. Nat. 116: 867–872.

34. Likens, G.E., R.F. Wright, J.N. Galloway, et al. 1979. Acidrain. Sci. Am. 241: 43–51.

35. Pacyna, E.G., J.M. Pacyna, F. Steenhuisen, et al. 2006. Globalanthropogenic mercury emission inventory for 2000. At-mos. Environ. 40: 4048–4063.

36. USEPA. 2005. Final Programmatic Environmental Im-pact Statement (PEIS) on Mountaintop Mining/Valley Fillsin Appalachia (EPA 9-03-R-05002). U.S. EnvironmentalProtection Agency. Retrieved September 1, 2010, fromhttp://www.epa.gov/region3/mtntop/eis2005.htm.

37. USDOE. 2006. Coal Production in the United States –An Historical Overview. Energy Information Ad-ministration, U.S. Department of Energy. RetrievedSeptember 1, 2010, from http://www.eia.doe.gov/cneaf/coal/page/coal production review.pdf.

38. Townsend P.A., D.P. Helmers, C.C. Kingdon, et al. 2009.Changes in the extent of surface mining and reclamationin the Central Appalachians detected using a 1976-2006Landsat time series. Remote Sens. Environ. 113: 62–72.

39. USDOE. 1995. Coal Data: A Reference (DOE/EIA-0064(93)). Energy Information Administration, U.S. De-partment of Energy. Retrieved September 1, 2010, fromftp://ftp.eia.doe.gov/coal/006493.pdf.

40. Ackerman, B. & W. Hassler. 1981. Clean Coal/Dirty Air.Yale University Press. New Haven, CT.

41. Davis, C.E. & R.J. Duffy. 2009. King coal vs. reclamationfederal regulation of mountaintop removal mining in Ap-palachia. Admin. Soc. 41: 674–692.

42. Hasselman, M. 2002. Bragg v. W-Va. Coal Ass’n and theunfortunate limitation of citizen suits against the state incooperative federalism regimes. Ecol. Law Quart. 29: 205–229.

43. Fox, J.F. & J.E. Campbell. 2009. Terrestrial carbon dis-turbance from mountaintop mining increases lifecycleemissions for clean coal. Environ. Sci.Technol. 44: 2144–2149.

44. Angel, P., V. Davis, J. Burger, et al. 2005. The AppalachianRegional Reforestation Initiative. (Advisory Number 1). Re-trieved September 1, 2010, from http://arri.osmre.gov/.

45. Chapin, F.S., L.R. Walker, C.L. Fastie, et al. 1994. Mech-anisms of primary succession following deglaciation atGlacier Bay, Alaska. Ecol. Monogr. 64: 149–175.

46. Delmoral, R. & L.C. Bliss. 1993. Mechanisms of primarysuccession–insights resulting from the eruption of MountSt-Helens. Adv. Ecol. Res. 24: 1–66.

47. Burger, J., D. Graves, P. Angel, et al. 2005. The ForestryReclamation Approach (Forest Reclamation Advisory Num-ber 2). U.S. Office of Surface Mining. Retrieved September1, 2010, from http://arri.osmre.gov/.

48. Chatterjee, A., R. Lal, R.K. Shrestha, et al. 2009. Soil carbonpools of reclaimed mine soils under grass and forest landuses. Land Degrad. Devel. 20: 300–307.

49. USEPA. 2005. National Management Measures to Con-trol Nonpoint Source Pollution from Urban Areas(EPA 841-BN-05-004). U.S. EPA Office of Water.Washington DC. Retrieved September 1, 2010, fromhttp://water.epa.gov/polwaste/nps/urban/index.cfn#08.

50. Groninger, J.W., S.D. Fillmore & R.A. Rathfon. 2006. Standcharacteristics and productivity potential of Indiana sur-face mines reclaimed under SMCRA. North. J. Appl. For.23: 94–99.

51. Rodrigue, J.A., J.A. Burger & R.G. Oderwald. 2002. Forestproductivity and commercial value of pre-law reclaimedmined land in the eastern United States. North. J. Appl. For.19: 106–114.

52. Amichev, B.Y., J.A. Burger & J.A. Rodrigue. 2008. Carbonsequestration by forests and soils on mined land in theMidwestern and Appalachian coalfields of the U.S. For. Ecol.Manage. 256: 1949–1959.

53. Simmons, J.A., W.S. Currie, K.N. Eshleman, et al. 2008.Forest to reclaimed mine land use change leads to al-tered ecosystem structure and function. Ecol. Appl. 18: 104–118.

54. Bosch, J.M. & J.D. Hewlett. 1982. A review of catchmentexperiments to determine the effect of vegetation changeson water yield and evapo-transpiration. J. Hydrol. 55: 3–23.

55. Messinger, T. 2003. Comparison of Storm Response ofStreams in Small, Unmined and Valley-Filled Water-sheds, 1999–2001, Ballard Fork, West Virginia (U.S. Ge-ological Survey Water-Resource Investigations Report02-4303). Retrieved September 15, 2010, from http://pubs.usgs.gov/wri/wri024303/pdf/wrir024303.pdf.

56. Messinger, T. & K.S. Paybins. 2003. Relations Between Pre-cipitation and Daily and Monthly Mean Flows in Gaged, Un-mined and Valley-Filled Watersheds, Ballard Fork, West Vir-ginia, 1999–2001 (U.S. Geological Survey Water-ResourceInvestigations Report 03-4113). Retrieved September 15,2010, from http://pubs.usgs.gov/wri/wri034113/.

57. Paybins, K.S., T. Messinger, J.H. Eychaner, et al. 2000. WaterQuality in the Kanawha -New River Basin West Virginia, Vir-ginia, and North Carolina, 1996–98 (U.S. Geological Sur-vey Circular 1204). Retrieved September 15, 2010, fromhttp://pubs.water.usgs.gov/circ1204/.

58. Ferrari, J.R., T.R. Lookingbill, B. McCormick, et al. 2009.Surface mining and reclamation effects on flood responseof watersheds in the central Appalachian Plateau region.Water Resour. Res. 45: W04407.

59. Chong, S.K., M.A. Becker, S.M. Moore, et al. 1986. Char-acterization of reclaimed mined land with and withouttopsoil. J. Environ. Qual. 15: 157–160.

60. Negley, T.L. & K.N. Eshleman. 2006. Comparison of storm-flow responses of surface-mined and forested watershedsin the Appalachian Mountains, U.S.A. Hydrol. Process. 20:3467–3483.



61. Potter, K.N., F.S. Carter & E.C. Doll. 1988. Physical-properties of constructed and undisturbed soils. Soil Sci.Soc. Am. J. 52: 1435–1438.

62. Brooks, K.N., P.F. Fofolliott, H.M. Gregersen, et al. 2003.Hydrology and the Management ofWatersheds, 3rd ed. IowaState Press. Ames, IA.

63. Tong, S.T.Y. & W.L. Chen. 2002. Modeling the relation-ship between land use and surface water quality. J. Environ.Manage. 66: 377–393.

64. Bormann, F.H. & G.E. Likens. 1974. Nutrient cycling. Sci-ence 155: 424–429.

65. Likens, G.E. & F.H. Bormann. 1974. Linkages be-tween terrestrial and aquatic ecosystems. Bioscience 24:447–456.

66. Allan, J.D. & M.M. Castillo. 2007. Stream Ecology, 2nd ed.Springer. Dordrecht, The Netherlands.

67. Hartman, K.J., M.D. Kaller, J.W. Howell, et al. 2005. Howmuch do valley fills influence headwater streams? Hydrobi-ologia 532: 91–102.

68. Pond, G.J., M.E. Passmore, F.A. Borsuk, et al. 2008. Down-stream effects of mountaintop coal mining: comparing bi-ological conditions using family- and genus-level macroin-vertebrate bioassessment tools. J. N. Am. Benthological Soc.27: 717–737.

69. Palmer, M.A., E.S. Bernhardt, W.H. Schlesinger, et al. 2010.Mountaintop mining consequences. Science 327: 148–149.

70. WVDEP. 2008. West Virginia Integrated Water QualityMonitoring and Assessment Report. West Virginia De-partment of Environmental Protection, Charleston, WestVirginia. Retrieved September 1, 2010, from http://www.wvdep.org/Docs/16495 WV 2008 IR SupplementsComplete Version EPA Approved.pdf [Id. at 21.].

71. Lamers, L.P.M., S.J. Falla, E.M. Samborska, et al. 2002.Factors controlling the extent of eutrophication and toxicityin sulfate-polluted freshwater wetlands. Limnol. Oceanogr.47: 585–593.

72. van der Welle, M.E.W., J.G.M. Roelofs & L.P.M. Lamers.2008. Sulfate-induced entrophication and phytotoxicity infreshwater wetlands. Sci. Total Environ. 406: 426–429.

73. Wang, F.Y. & P.M. Chapman. 1999. Multi-level effects ofsulphur-iron interactions in freshwater wetlands in TheNetherlands. Environ. Toxicol. Chem. 18: 2526–2532.

74. Lamers, L.P.M., H.B.M. Tomassen & J.G.M. Roelofs. 1998.Biological implications of sulfide in sediment - A reviewfocusing on sediment toxicity. Environ. Sci. Technol. 32:199–205.

75. Caraco, N.F. 1993. Disturbance of the phosphorus cycle–acase of indirect effects of human activity. Trends. Ecol. Evol.8: 51–54.

76. Caraco, N.F., J.J. Cole & G.E. Likens. 1989. Evidence forsulfate-controlled phosphorus release from sediments ofaquatic ecosystems. Nature 341: 316–318.

77. Joye, S.B. & J.T. Hollibaugh.1995. Influence of sulfide in-hibition of nitrification on nitrogen regeneration in sedi-ments. Science 270: 623–625.

78. Johnson, B.R., A. Haas & K.M. Fritz. 2010. Use of spa-tially explicit physicochemical data to measure downstreamimpacts of headwater stream disturbance. Water Resour.Res. 46: W09526.

79. Bryant, G., S. Mcphilliamy & H. Childers. 2002. A Surveyof the Water Quality of Streams in the Primary Region ofMountaintop/Valley Fill Coal Mining. US EnvironmentalProtection Agency, Philadelphia, PA.

80. USEPA. 2010. A Field-based Aquatic Life Bench-mark for Conductivity in Central Appalachian Streams(External Review Draft). Retrieved September 1,2010, from http://cfpub.epa.gov/ncea/cfm/recordisplay.cfm?deid=220171#Download.

81. Hart, B.T., P. Bailey, R. Edwards, et al. 1991. A review of thesalt sensitivity of the Australian freshwater biota. Hydrobi-ologia 210: 105–144.

82. Hassell, K.L., B.J. Kefford & D. Nugegoda. 2006. Sub-lethaland chronic salinity tolerances of three freshwater insects:Cloeon sp and Centroptilum sp (Ephemeroptera : Baetidae)and Chironomus sp (Diptera : Chironomidae). J. Exp. Biol.209: 4024–4032.

83. Gaino, E. & M. Rebora. 2000. The duct connectingMalpighian tubules and gut: an ultrastructural and com-parative analysis in various Ephemeroptera nymphs (Ptery-gota). Zoomorphology 120: 99–106.

84. Chadwick, M.A., H. Hunter, J.W. Feminella, et al. 2002. Saltand water balance in Hexagenia limbata (Ephemeroptera :Ephemeridae) when exposed to brackish water. Fla. Ento-mol. 85: 650–651.

85. Goetsch, P.A. & C.G. Palmer. 1997. Salinity tolerances ofselected macroinvertebrates of the Sabie River, Kruger Na-tional Park, South Africa. Arch. Environ. Contam. Toxicol.32: 32–41.

86. Kefford, B.J., P.J. Papas & D. Nugegoda. 2003. Relativesalinity tolerance of macroinvertebrates from the BarwonRiver, Victoria, Australia. Mar. Freshwater Res. 54: 755–765.

87. Kennedy, A.J., D.S. Cherry & R.J. Currie. 2003. Relativesalinity tolerance of macroinvertebrates from the BarwonRiver, Victoria, Australia. Arch. Environ. Contam. Toxicol.44: 324–765.

88. Echols, B.S., R.J. Currie & D.S. Cherry. 2010. Preliminaryresults of laboratory toxicity tests with the mayfly, Isonychiabicolor (Ephemeroptera: Isonychiidae) for development asa standard test organism for evaluating streams in the Ap-palachian coalfields of Virginia and West Virginia. Environ.Monit. Assess. 169: 487–500.

89. Wood, P.J. & A.P. Dykes. 2002. The use of salt dilutiongauging techniques: ecological considerations and insights.Water Res. 36: 3054–3062.

90. Petty, J.T., J.B. Fulton, M.P. Strager, et al. 2010. Landscapeindicators and thresholds of stream ecological impairmentin an intensively mined Appalachian watershed. J. N. Am.Benthological Soc. 29: 1292–1329.

91. Reeves, M.K., P. Jensen, C.L. Dolph, et al. 2010. Multiplestressors and the cause of amphibian abnormalities. Ecol.Monogr. 80: 423–440.

92. Burton, G.A., R. Pitt & S. Clark. 2000. The role of traditionaland novel toxicity test methods in assessing stormwater andsediment contamination. Crit. Rev. Environ. Sci. Technol.30: 413–447.

93. Drescher, K. & W. Boedeker. 1995. Assessment of thecombined effects of substances–the relationship between



concentration addition and independent action. Biomet-rics 51: 716–730.

94. Berenbaum, M.C. 1985. The expected effect of a combi-nation of agents–the general solution. J. Theor. Biol. 114:413–431.

95. Konemann, H. 1981. Fish toxicity tests with mixtures ofmore than 2 chemicals–a proposal for a quantitative ap-proach and experimental results. Toxicology 19: 229–230.

96. Pond, G.J. 2010. Patterns of Ephemeroptera taxa loss inAppalachian headwater streams (Kentucky, U.S.A.). Hy-drobiologia 641: 185–201.

97. Sangi, E. 2010. Equating stream structure with function:The Fourth Circuit’s misstep in Ohio Valley EnvironmentalCoalition v. Aracoma Coal Co. Ecol. Law Quart. 37: 701–709.

98. USACOE. 2007. Functional Assessment for High GradientStreams in WV. U.S. Army Corps of Engineers HuntingtonDistrict. Retrieved September 1, 2010, from http://www.lrh.usace.army.mil/Documents/index.cfm?id=12018&pge id=1101.

99. Duhigg, C. 2009, Sept. 12. Clean water laws are ne-glected at a cost in suffering. The New York Times Re-trieved September 1, 2010, from http://www.nytimes.com/2009/09/13/us/13water.html.

100. Naiman, R.J., N. Décamps & M.E. McClain. 2005. Riparia:Ecology, Conservation, and Management of Streamside Com-munities. Elsevier. Burlington, MA.

101. Palmer, M., A.P. Covich, B.J. Finlay, et al. 1997. Biodiversityand ecosystem processes in freshwater sediments. Ambio 26:571–577.

102. Palmer, M.A. & D.R. Richardson. 2008. Provisioning ser-vices: a focus on freshwater. In Princeton Guide to Ecology:625–633. Princeton University Press. Princeton, NJ.

103. Gessner, M.O. & E. Chauvet. 2002. A case for using litterbreakdown to assess functional stream integrity. Ecol. Appl.12: 498–510.

104. Peterson, B.J., W.M. Wollheim, P.J. Mulholland, et al. 2001.Control of nitrogen export from watersheds by headwaterstreams. Science 292: 86–98.

105. Bukaveckas, P.A. 2007. Effects of channel restoration onwater velocity, transient storage, and nutrient uptake in achannelized stream. Environ. Sci. Technol. 41: 1570–1576.

106. Kaushal, S.S., P.M. Groffman, P.M. Mayer, et al. 2008. Ef-fects of stream restoration on denitrification in an urban-izing watershed. Ecol. Appl. 18: 789–804.

107. Roberts, B.J., P.J. Mulholland & A.N. Houser. 2007. Effectsof upland disturbance and instream restoration on hydro-dynamics and ammonium uptake in headwater streams.J. N. Am. Benthological Soc. 26: 38–53.

108. Fellows, C.S., J.E. Clapcott, J.W. Udy, et al. 2006. Benthicmetabolism as an indicator of stream ecosystem health.Hydrobiologia 572: 71–87.

109. Udy, J.W., C.S. Fellows, M.E. Bartkow, et al. 2006. Mea-sures of nutrient processes as indicators of stream ecosys-tem health. Hydrobiologia 572: 89–102.

110. Hauer, R.H. & G. Lamberti. 2006. Methods in Stream Ecol-ogy, 2nd ed. Academic Press. London.

111. Falk, D.A., M.A. Palmer & J.B. Zedler. 2006. Foundations ofRestoration Ecology. Island Press. Washington, DC.

112. Fischenich, J.C. 2006. Functional objectives for streamrestoration (ERDC TN-EMRRP-SR-52). U.S. Army En-gineer Research and Development Center Vicksburg,MS. Retrieved September 1, 2010, from www.wes.army.mil/el/emrrp.

113. Filoso, S. & M.A. Palmer. 2011. Assessing stream restora-tion effectiveness at reducing nitrogen loads to downstreamwaters. Ecol. Appl. In press.

114. Sudduth, E.B. & E.S. Bernhardt. 2011. Testing the field ofdreams hypothesis: functional responses to urbanizationand restoration in stream ecosystems. Ecol. Appl. In press.

115. Fritz, K.M., S. Fulton, B.R. Johnson, et al. 2010. Structuraland functional characteristics of natural and constructedchannels draining a reclaimed mountaintop removal andvalley fill coal mine. J. N. Am. Benthological Soc. 29: 673–689.

116. FISRWG. 1998. Federal Interagency Stream RestorationWorking Group: Stream corridor restoration: principles, pro-cesses and practices (United States National EngineeringHandbook, Part 653). Washington, D.C.

117. Karr, J.R. & E.W. Chu. 1999. Restoring Life in Running Wa-ters: Better Biological Monitoring . Island Press. Washington,DC.

118. Williams, D.D., N.E. Williams & Y. Cao. 1997. Spatial differ-ences in macroinvertebrate community structure in springsin southeastern Ontario in relation to their chemical andphysical environments. Can. J. Zool.-Revue Canadienne DeZoologie 75: 1404–1414.

119. Palmer, M.A. & S. Filoso. 2009. Restoration of ecosys-tem services for environmental markets. Science 325: 575–576.

120. Bernhardt, E.S., M.A. Palmer, J.D. Allan, et al. 2005.Ecology–Synthesizing U.S. river restoration efforts. Science308: 636–637.

121. Palmer, M.A., E.S. Bernhardt, J.D. Allan, et al. 2005. Stan-dards for ecologically successful river restoration. J. Appl.Ecol. 42: 208–217.

122. Bernhardt, E.S., E.B. Sudduth, M.A. Palmer, et al. 2007.Restoring rivers one reach at a time: Results from a surveyof U.S. river restoration practitioners. Restor. Ecol. 15: 482–493.

123. Palmer. M.A. 2009. Reforming restoration: science in needof application and applications in need of Science. EstuariesCoasts 32: 1–17.

124. Palmer, M.A., H.L. Menninger & E. Bernhardt. 2010. Riverrestoration, habitat heterogeneity and biodiversity: a failureof theory or practice? Freshwater Biol. 55: 205–222.

125. Smith, S.M. & K.L. Prestegaard. 2005. Hydraulic perfor-mance of a morphology-based stream channel design.Water Resour. Res. 41: W11413.

126. Tullos, D.D., D.L. Penrose, G.D. Jennings, et al. 2009. Anal-ysis of functional traits in reconfigured channels: implica-tions for the bioassessment and disturbance of river restora-tion. J. N. Am. Benthological Soc. 28: 80–92.

127. Rosgen, D.L. 1996. Applied River Morphology. WildlandHydrology. Pagosa Springs, CO.

128. Palmer, M.A., H.L. Menninger & E. Bernhardt. 2009. Riverrestoration, habitat heterogeneity and biodiversity: a failureof theory or practice? Freshwater Biol 55: 205–222.



129. Hilderbrand, R.H., A.C. Watts & A.M. Randle. 2005. Themyths of restoration ecology. Ecol. Soc. 10: 17.

130. Palmer, M.A., R.F. Ambrose & N.L. Poff. 1997. Ecologicaltheory and community restoration ecology. Restor. Ecol. 5:291–300.

131. Kondolf, G.M., M.W. Smeltzer & S.F. Railsback. 2001. De-sign and performance of a channel reconstruction projectin a coastal California gravel-bed stream. Environ. Manage.28: 761–776.

132. Gillilan, S. 1996. Use and misuse of channel classifi-cation schemes. Stream Notes Newsletter. USDA For-est Service Stream Systems Technology Center. Re-trieved September 1, 2010, from http://www.stream.fs.fed.us/news/streamnt/oct96/oct96a2.htm.

133. K.E. Juracek & F.A. Fitzpatrick. 2003. Limitations and im-plications of stream classification. J. Am. Water Resour. As-soc. 39: 659–670.

134. Niezgoda, S.L. & P.A. Johnson. 2005. Improving the ur-ban stream restoration effort: Identifying critical form andprocesses relationships. Environ. Manage. 35: 579–592.

135. Roper, B.B., J.M. Buffington, E. Archer, et al. 2008. The roleof observer variation in determining Rosgen stream typesin northeastern Oregon mountain streams. J. Am. WaterResour. Assoc. 44: 417–427.

136. Shields, F.D., A. Brookes & J.P. Haltiner. 1999. Geomorpho-logical approaches to incised stream channel restoration inthe United States and Europe. In Incised River Channels. S.Darby & A. Simon, Eds.: 371–394. John Wiley and Sons.New York.

137. Simon, A., M. Doyle, M. Kondolf, et al. 2007. Critical evalu-ation of how the Rosgen classification and associated “nat-ural channel design” methods fail to integrate and quantifyfluvial processes and channel response. J. Am. Water Resour.Assoc. 43: 1117–1131.

138. Slate, L.O., F.D. Shields, J.S. Schwartz, et al. 2007. Engi-neering design standards and liability for stream channelrestoration. J. Hydraul. Eng.-Asce 133: 1099–1102.

139. Fritz, K.M. & W.K. Dodds. 2005. Harshness: characteriza-tion of intermittent stream habitat over space and time.Mar. Freshwater Res. 56: 13–23.

140. Bonada, N., M. Rieradevall & N. Prat. 2007. Macroinver-tebrate community structure and biological traits relatedto flow permanence in a Mediterranean river network. Hy-drobiologia 589: 91–106.

141. Doyle, M.W. & E.S. Bernhardt. 2011. What is a stream?Environ. Sci. Technol. 45: 354–359.

142. Collins, B.M., W.V. Sobczak & E.A. Colburn. 2007. Sub-surface flowpaths in a forested headwater stream harbor adiverse macroinvertebrate community. Wetlands 27: 319–325.

143. Bond, N.R. & P. Cottingham. 2007. Ecology and hy-

drology of temporary streams: implications for sustain-able water management (eWater Technical Report). eWa-ter Cooperative Research Centre. Canberra. RetrievedSeptember 1, 2010, from http://ewatercrc.com.au/reports/Bond Cottingham-2008-Temporary Streams.pdf.

144. Butturini, A., S. Bernal, E. Nin, et al. 2003. Influences of thestream groundwater hydrology on nitrate concentrationin unsaturated riparian area bounded by an intermittentMediterranean stream. Water Resour. Res. 39(4): W1110.

145. Wallace, B.J. et al. 1995. Wallace. Long term dynamicsof coarse particulate organic matter in three Appalachianmountain streams. J. N. Am. Benthological Soc. 14: 217–232.

146. Wallace, J.B., D.H. Ross & J.L. Meyer. 1982. Seston and dis-solved organic carbon dynamics in a southern Appalachianstream. Ecology 63: 824–838.

147. Hill, W.R., M.G. Ryon & E.M. Schilling. 1995. Light limita-tion in a stream ecosystem: responses by primary producersand consumers. Ecology 76: 1297–1309.

148. Fisher, S.G., R.A. Sponseller & J.B. Heffernan. 2004. Hori-zons in stream biogeochemistry: Flowpaths to progress.Ecology 85: 2369–2379.

149. Jones, Jr., J.G. & P.J. Mulholland. 2000. Streams and GroundWaters. Academic Press. San Diego.

150. Poole, G.C., S.J. O’Daniel, K.L. Jones. et al. 2008. Hydro-logic spiralling: The role of multiple interactive flow pathsin stream ecosystems. River Res. Appl. 24: 1018–1031.

151. Doyle, M.W., E.H. Stanley, D.L. Strayer, et al. 2005. Effectivedischarge analysis of ecological processes in streams. WaterResour. Res. 41: W11411.

152. Poff, N.L., J.D. Allan, M.B. Bain, et al. 1997. The naturalflow regime. Bioscience 47: 769–784.

153. Poff, N.L. & J.V. Ward. 1989. Implications of stream-flow variability and predictability for lotic communitystructure–a regional analysis of streamflow patterns. Can.J. Fish. Aquat. Sci. 46: 1805–1818.

154. Wohl, E., P.L. Angermeier, B. Bledsoe, et al. 2005. Riverrestoration. Water Resour. Res. 41:W10301.

155. Graf, W.L. 2001. Damage control: Restoring the physicalintegrity of America’s rivers. Ann. Assoc. Am. Geographers91: 1–27.

156. Gregory, S.V., F.J. Swanson, W.A. McKee & K.W. Cum-mins. 1991. An ecosystem perspective on riparian zones.Bioscience 41: 540–551.

157. Palmer, M.A., E.S. Bernhardt, J.D. Allan, et al. 2005. Stan-dards for ec