Do magnesium and chloride ameliorate high sodium bicarbonate concentrations? A comparison between laboratory and mesocosm toxicity experiments.

Increasing salinity is a concern for biodiversity in many freshwater ecosystems globally. Single species laboratory toxicity tests show major differences in freshwater organism survival depending on the specific ions that comprise salinity types and/or their ion ratios. Toxicity has been shown to be reduced by altering ionic composition, despite increasing (total) salinity. For insistence, single species tests show the toxicity of sodium bicarbonate (NaHCO3, which commonly is a large proportion of the salts from coalbeds) to freshwater invertebrates is reduced by adding magnesium (Mg2+) or chloride (Cl-). However, it is uncertain whether reductions in mortality observed in single-species laboratory tests predict effects within populations, communities and to ecosystem processes in more complex multi-species systems both natural and semi-natural. Here we report the results of an outdoor multi-species mesocosm experiment to determine if the effects of NaHCO3 are reduced by increasing the concentrations of Mg2+ or Cl- on: a) stream macroinvertebrate populations and communities; b) benthic chlorophyll-a and; c) the ecosystem process of leaf litter decomposition. We found a large effect of a high NaHCO3 concentration (≈4.45 mS/cm) with reduced abundances of multiple taxa, reduced emergence of adult insects and reduced species richness, altered community structure and increased leaf litter breakdown rates but no effect on benthic chlorophyll-a. However, despite predictions based on laboratory findings, we found no evidence that the addition of either Mg2+ or Cl- altered the effect of NaHCO3. In semi-natural environments such as mesocosms, and natural environments, organisms are subject to varying temperature and habitat factors, while also interacting with other species and trophic levels (e.g. predation, competition, facilitation), which are absent in single species laboratory tests. Thus, it should not be assumed single-species tests are good predictors of the effects of changing ionic compositions on stream biota in more natural environments.