Seawater intrusion: an appraisal of taxa at most risk and safe salinity levels

ABSTRACT Seawater intrusion into low‐lying coastal ecosystems carries environmental risks. Salinity levels at these coastal ecosystems may vary substantially, causing ecological effects from mortality to several sublethal endpoints, such as depression of rates of feeding, somatic growth, or reproduction. This review attempts to establish safe salinity levels for both terrestrial and freshwater temperate ecosystems by integrating data available in the literature. We have four specific objectives: ( i ) to identify the most sensitive ecological taxa to seawater intrusion; ( ii ) to establish maximum acceptable concentrations–environmental quality standards (MAC‐EQSs) for sea water (SW) from species sensitivity distributions (SSDs); ( iii ) to compile from the literature examples of saline intrusion to be used as predicted environmental concentrations (PECs) and to compute risk quotients for the temperate zone; and ( iv ) to assess whether sodium chloride (NaCl) is an appropriate surrogate for SW in ecological risk assessments by comparing SSD‐derived values for NaCl and SW and by comparing these with field data. Zooplankton, early life stages of amphibians and freshwater mussels were the most sensitive ecological receptors for the freshwater compartment, while soil invertebrates were the most sensitive ecological receptors for the terrestrial compartment. Hazard concentration 5% (HC 5 ) values, defined as the concentration (herein measured as conductivity) that affects (causes lethal or sublethal effects) 5% of the species in a distribution, computed for SW were over 22 and 40 times lower than the conductivity of natural SW (≈ 52 mS/cm) for the freshwater and soil compartment, respectively. This sensitivity of both compartments means that small increments in salinity levels or small SW intrusions might represent severe risks for low‐lying coastal ecosystems. Furthermore, the proximity between HC 5 values for the soil and freshwater compartments suggests that salinized soils might represent an additional risk for nearby freshwater systems. This sensitivity was corroborated by the derivation of risk quotients using real saline intrusion examples (PECs) collected from the literature: risk was >1 in 34 out of 37 examples. By contrast, comparisons of HC 5 values obtained from SSDs in field surveys or mesocosm studies suggest that natural communities are more resilient to salinization than expected. Finally, NaCl was found to be slightly more toxic than SW, at both lethal and sublethal levels, and, thus, is suggested to be an acceptable surrogate for use in risk assessment.