A multimetric approach: Utilizing diatom communities to assess acid mine drainage impacts on South African wetlands

Acid mine drainage (AMD) is a major concern for wetlands located in coal-bearing regions, particularly where mining activities and associated spoil material are present. In this study, a diatom-based index is proposed to detect acidification related to AMD in wetlands surrounding coal mines, with a specific focus on the principal environmental stressors associated with AMD: low pH, low alkalinity, elevated electrical conductivity (EC), and elevated sulphate concentrations. Chloride was included as a secondary indicator reflecting mine-related ionic enrichment rather than direct AMD-generating processes. A total of 87 samples were collected from 45 sites across two seasons (dry and wet) to capture spatial and temporal variation in diatom community structures. Epiphytic diatoms were sampled using standard protocols, and species composition was derived alongside environmental variables. This dataset formed the basis for multivariate analyses linking diatom community structure to environmental gradients associated with acidification. Canonical correspondence analysis (CCA) demonstrated that diatom species distributions were primarily structured along an acidification gradient driven by pH (−0.93), alkalinity (−0.79), and sulphate (0.65), with a secondary ionic gradient associated with chloride and EC. Weighted averaging (WA) was used to calculate environmental optima for species under key stressors, enabling diatom taxa to be positioned along ecological gradients. These optima were used to calculate community-weighted metrics (CWMs), which were integrated into two multimetric indices (MMIs). Redundancy analysis (RDA) of the derived indices revealed strong and consistent associations with the same environmental gradients identified in the CCA, pH (−0.91), sulphate (0.67), and alkalinity (−0.78), and clearly distinguished between acidified and non-acidified wetlands. The correspondence between CCA and RDA results indicates a hierarchical compression of ecological information from species assemblages to indices while retaining dominant environmental gradients, making the proposed MMI a practical tool for distinguishing wetland acidification under AMD influence where routine water quality monitoring is constrained.