Response of Stream Biodiversity to Increasing Salinization

Abstract:

We used a large data set of macroinvertebrate samples collected from streams in both reference-quality (n = 68) and degraded (n = 401) watersheds in the state of Nevada, USA to assess relationships between stream biodiversity and salinity. We used specific electrical conductance (EC)(μS/cm) as a measure of salinity, and applied a previously developed EC model to estimate natural, baseflow salinity at each stream. We used the difference between observed and predicted salinity (EC-Diff) as a measure of salinization associated with watershed degradation. Observed levels of EC varied between 22 and 994 μS/cm across reference sites and 22 to 3,256 uS/cm across non-reference sites. EC-Diff was as high as 2,743 μS/cm. We used a measure of local biodiversity completeness (ratio of observed to expected number of taxa) to assess ecological response to salinity. This O/E index decreased nearly linearly up to about 25% biodiversity loss, which occurred at EC-Diff of about 300 μS/cm. Too few sites had EC-Diff greater than 300 μS/cm to draw reliable inferences regarding biodiversity response to greater levels of salinization. EC-Diff increased with % agricultural land use, mine density, and % urban land use in the watersheds implying that human activities have been largely responsible for increased salinization in Nevada streams and rivers. Comparison of biological responses to EC and other stressors indicates that increased salinization may be the primary stressor causing biodiversity loss in these streams and that more stringent salinity water quality standards may be needed to protect aquatic life.