Continuous response of benthic macroinvertebrate assemblages to a discrete disturbance gradient: consequences for diagnosing stressors

In a previous study, we found that underlying geology, acidic precipitation, and acidic mine drainage (AMD) associated with coal extraction in the Monongahela River basin, West Virginia, interact to produce discrete water-quality types among streams, including: high-quality reference, soft, hard, transitional, and severely impaired AMD types. In this study, we tested the prediction that this discrete impairment template would produce correspondingly discrete benthic macroinvertebrate assemblages. Furthermore, we posited that unique macroinvertebrate genera would serve as strong indicators of each water-quality type. Macroinvertebrate assemblages were strongly influenced by stream water chemistry, and tests of compositional similarity identified statistical links with water-quality types. However, assemblage composition was highly variable within and among water-quality types, a finding that provides greater support for continuous than for discrete structure. High variability in assemblage composition among streams of the same water-quality type could not be explained by differences in physical habitat or by variation in water chemistry. Benthic invertebrate assemblages also exhibited significant nestedness, but only AMD assemblages were clear subsets of reference-type assemblages. Furthermore, indicator species analysis found relatively few genera that were indicators for specific water-quality types, but they were strong for reference streams (e.g., Epeorus, Dolophilodes), soft-water streams (e.g., Simulium, Leuctra), and hard-water streams (e.g., Ectopria). Therefore, little evidence was found for a direct correspondence between benthic macroinvertebrate assemblages and the discrete water-quality template. We suggest that metacommunity dynamics might prevent a strong correspondence despite the localized effects of water chemistry. Consequently, use of benthic macroinvertebrate assemblage data to diagnose specific water-quality stressors might continue to be difficult until we factor in the effects of invertebrate dispersal and assemblage dynamics at a watershed scale.