An ecological study was undertaken on the Powell River system in Virginia from 1988 - 1990 to evaluate the effects of land uses on hydrology, water quality, sedimentation (particle size < 2 mm), and fish assemblages and to derive ecological indicators for monitoring. The hydrologic (disturbance) regime of the sixth order reach was Classified. Although surface coal mining appears to have reduced flow variance, the changes were not sufficient to alter the hydrologic classification.

Tributaries draining coal-mined lands had elevated specific conductivity, iron, manganese, sulphate, and turbidity relative to tributaries in unmined watersheds or in the sixth order mainstem (p ≤ 0.05). The fifth order reach, which had a greater proportion of watershed area surface-mined, exhibited higher specific conductivity, iron, sulphate, turbidity, and total solids than the sixth order (p ≤ 0.05). Despite the upstream-to-downstream, lower-to-higher water quality gradient, there was no compelling evidence from the use of state water quality standards and a nine-variable water quality index that water quality differentially affected fish assemblages.

The primary physical habitat gradient in the lower river was sedimentation. Upstream, embeddedness increased in shallow-water habitats (p ≤ 0.001) and sediment depth increased in pools (p ≤ 0.001). Stepwise multiple regression analysis indicated that variation in sediment level was explained primarily by the proportion of the watershed surface-mined (R² = 0.75).

Index of Biotic Integrity scores for fish assemblages were not correlated with sedimentation in shallow-water or in pool habitats (p > 0.066). Functional metrics, as opposed to taxonomic metrics, however, varied with sedimentation (p ≤ 0.002). In shallow-water habitats, omnivore relative abundance increased and specialized insectivore abundance decreased in higher sedimented sites. Top carnivore abundance decreased in pools with higher sedimentation levels. Abundance of lithophilous benthic spawners, postulated as the most sediment-sensitive reproductive guild, was not consistently correlated with sedimentation.

Nine fish species were classified as sediment-intolerant and eight were classified as sediment-tolerant. The sediment-intolerant group was composed mostly of benthic insectivores. Habitat analyses indicated that sediment-intolerant species, as a category, utilized microhabitats with low sedimentation levels relative to the sediment-tolerant species group (p ≤ 0.0001). Species not classified as either sediment-intolerant or sediment-tolerant utilized microhabitats intermediate in sedimentation level (p ≤ 0.0001).