How Many Mountains Can We Mine? Assessing the Regional Degradation of Central Appalachian Rivers by Surface Coal Mining

Abstract

Surface coal mining is the dominant form of land cover change in Central Appalachia, yet the extent to which surface coal mine runoff is polluting regional rivers is currently unknown. We mapped surface mining from 1976 to 2005 for a 19,581 km(2) area of southern West Virginia and linked these maps with water quality and biological data for 223 streams. The extent of surface mining within catchments is highly correlated with the ionic strength and sulfate concentrations of receiving streams. Generalized additive models were used to estimate the amount of watershed mining, stream ionic strength, or sulfate concentrations beyond which biological impairment (based on state biocriteria) is likely. We find this threshold is reached once surface coal mines occupy >5.4% of their contributing watershed area, ionic strength exceeds 308 μS cm(-1), or sulfate concentrations exceed 50 mg L(-1). Significant losses of many intolerant macroinvertebrate taxa occur when as little as 2.2% of contributing catchments are mined. As of 2005, 5% of the land area of southern WV was converted to surface mines, 6% of regional streams were buried in valley fills, and 22% of the regional stream network length drained watersheds with >5.4% of their surface area converted to mines.

Isotopic Imprints of Mountaintop Mining Contaminants

“…These results are consistent with earlier observations that the salinity, sulfate, and Se concentrations in MTM streams are linearly proportional to the extent of surface mining in the watershed. 14,15 A mass-balance calculation, using the Sr and 87 Sr/ 86 Sr in a mixing model, provides a sensitive evaluation of the fraction of the tributaries contributing to each of the downstream sampling points along the Mud River (Figure 1). We estimate that the tributaries provide between 5% and 33% of the dissolved constituents in the mainstem at different sampling locations along its flow (Figure 1 and Table S3, Supporting Information).…”

“…15 A survey quantifying the extent of surface mining in southern West Virginia revealed that the extent of surface mining disturbance within a catchment is highly correlated with the ionic strength and sulfate concentrations of the receiving streams. 14 The majority of coal production from surface mines in the southern Appalachian of West Virginia is derived from the Middle Pennsylvanian Kanawha and Allegheny formations (Figures S1, S2, Supporting Information). The Kanawha Formation is composed mostly of sandstone, gray and black shale (Carbolith), siltstone, and hosts several coal seams, including the high quality Coalburg and Stockton coals.…”

A critique of the use of indicator-species scores for identifying thresholds in species responses

“…These concerns are easily addressed via modification of TITAN's output (i.e., IndVal maxima are output as part of every TITAN object) or post hoc simulation (see appendix 1 in Bernhardt et al 2012) with minor effect on TITAN's core analysis. TITAN is relatively unique in its goal of deconstructing community responses into taxon-specific change as a complement to ecotoxicology, species distribution modeling, and community analysis.…”

“…We developed more realistic simulations using negative binomial distributions with overdispersion in R (version 2.15.1; R Development Core Team, Vienna, Austria; Appendix S1; available online from: http://dx.doi.org/10.1899/12-142.1.s1). We set mean abundance across observations to range from 2 to 10 individuals, which approximates the range of empirical values we have observed in our analyses of various survey data sets (e.g., King and Baker 2010, Bernhardt et al 2012. We set mean abundance to be a function of the response models suggested by fig.…”

“…Analyses of regression trees suggest that skewed samples can affect the pattern of variance when applied to idealized and smoothly varying functions (i.e., linear; Daily et al 2012). However, IndVal partitioning (as opposed to z-score partitioning) in TITAN appears relatively robust to strongly skewed samples when parameters from observed data are used with more realistic negative binomial simulations (see appendix 1 by Bernhardt et al 2012 for a detailed example). To aid users in better understanding the implications of skewed samples in specific data sets, forthcoming versions of TITAN will offer guidance for generating data-set-specific simulations and IndVal, rather than z-score, partitioning.…”