English

Abstract

Elevated levels of total dissolved solids (TDS) have been identified as stressors to aquatic life in Central Appalachian coalfield streams. At present there are no aquatic life water quality criteria for TDS in the primary coal-producing Central Appalachian states (KY, VA, WV). In all three states, mining-related impacts on aquatic life have been characterized using measures of benthic macroinvertebrate community structure. Impacts of mining on aquatic life have been documented in the literature, but our understanding of impacts from TDS is confounded because elevated TDS rarely occurs independent of other stressors in coalfield streams. Potential TDS covariates in coalfield streams include acidic pH, toxic metals, sedimentation, in-stream and riparian habitat degradation, trophic structure alteration, and hydrologic modification. As a means of isolating TDS effects, we identified 17 headwater streams in Virginia's coalfield region that represent a gradient of TDS concentrations, where influence from non-TDS stressors was minimized (i.e., pH between 6.0 and 9.0, low metal concentrations, reference-quality habitat, primarily forested land-use). Benthic macroinvertebrate communities were sampled from these streams in Spring 2009. Organisms were enumerated and identified to the family/lowest practicable level. These data were then used to calculate common benthic macroinvertebrate community metrics. In addition, TDS and component ions were measured for water collected concurrently with biological samples. Data were analyzed for significant associations between TDS and biological metrics. We identified stream sites with elevated TDS where influence from non-TDS stressors was minimal. Dominant components of TDS were sulfate (46% by weight, mean), bicarbonate (26%), and calcium (13%). Several benthic macroinvertebrate richness measures were correlated negatively with increasing TDS (p 0.05) within the range of TDS that we measured (28-792 mg/L). Our results suggest sulfate is a good candidate for single-parameter prediction of biological condition.