No abstract

Recent studies have documented adverse effects to biological communities downstream of mountaintop coal mining and valley fills (VF), but few data exist on the longevity of these impacts. We sampled 15 headwater streams with VFs reclaimed 11-33 years prior to 2011 and sampled seven local reference sites that had no VFs. We collected chemical, habitat, and benthic macroinvertebrate data in April 2011; additional chemical samples were collected in September 2011. To assess ecological condition, we compared VF and reference abiotic and biotic data using: (1) ordination to detect multivariate differences, (2) benthic indices (a multimetric index and an observed/expected predictive model) calibrated to state reference conditions to detect impairment, and (3) correlation and regression analysis to detect relationships between biotic and abiotic data. Although VF sites had good instream habitat, nearly 90 % of these streams exhibited biological impairment. VF sites with higher index scores were co-located near unaffected tributaries; we suggest that these tributaries were sources of sensitive taxa as drifting colonists. There were clear losses of expected taxa across most VF sites and two functional feeding groups (% scrapers and %shredders) were significantly altered. Percent VF and forested area were related to biological quality but varied more than individual ions and specific conductance. Within the subset of VF sites, other descriptors (e.g., VF age, site distance from VF, the presence of impoundments, % forest) had no detectable relationships with biological condition. Although these VFs were constructed pursuant to permits and regulatory programs that have as their stated goals that (1) mined land be reclaimed and restored to its original use or a use of higher value, and (2) mining does not cause or contribute to violations of water quality standards, we found sustained ecological damage in headwaters streams draining VFs long after reclamation was completed.

In 2003, we compared two benthic macroinvertebrate sampling methods that are used for rapid biological assessment of wadeable streams. A single habitat method using kick sampling in riffles and runs was compared to a multiple habitat method that sampled all available habitats in proportion of occurrence. Both methods were performed side‐by‐side at 41 sites in lower gradient streams of the Piedmont and Northern Piedmont ecoregions of the United States, where riffle habitat is less abundant. Differences in sampling methods were examined using similarity indices, two multimetric indices [the family‐level Virginia Stream Condition Index (VSCI) and the species‐level Macroinvertebrate Biotic Integrity Index (MBII)], their component metrics, and bioassessment endpoints based on each index. Index scores were highly correlated between single and multiple habitat field methods, and sampling method comparability, based on comparison of similarities between and within sampling methods, was particularly high for species level data. The VSCI scores and values of most of its component metrics were not significantly higher for one particular method, but relationships between single and multiple habitat values were highly variable for percent Ephemeroptera, percent chironomids, and percent Plecoptera and Trichoptera (Hydropsychidae excluded). A similar level of variability in the relationship was observed for the MBII and most of its metrics, but Ephemeroptera richness, percent individuals in the dominant five taxa, and Hilsenhoff Biotic Index scores all exhibited differences in values between single and multiple habitat field methods. When applied to multiple habitat samples, the MBII exhibited greater precision, higher index scores, and higher assessment categories than when applied to single habitat samples at the same sites. In streams with limited or no riffle habitats, the multiple habitat method should provide an adequate sample for biological assessment, and at sites with abundant riffle habitat, little difference would be expected between the single and multiple habitat field methods. Thus, in geographic areas with a wide variety of stream types, the multiple habitat method may be more desirable. Even so, the variability in the relationship between single and multiple habitat methods indicates that the data are not interchangeable, and we suggest that any change in sampling method should be accompanied by a recalibration of any existing assessment tool (e.g., multimetric index) with data collected using the new method, regardless of taxonomic level.