Improving assessment and risk-based management of freshwater salinisation

Salinisation is a common and increasingly threatening environmental stressor to freshwater ecosystems globally. Despite this, there is a lack of environmentally relevant data on the effects of salinisation for freshwater macroinvertebrates. This is a problem because in order to manage and mitigate effects caused by freshwater salinisation, we first need to more fully understand how salinity affects freshwater biota. This thesis comprises a series of papers that aim to improve salinity risk assessment and water quality guideline derivation for lotic freshwater ecosystems by contributing empirical data to knowledge gaps and suggesting future research directions. Trade-offs between data quality and quantity in salinity risk assessment using species sensitivity distributions (SSDs) were investigated. The dominant taxonomic group was the prime determinant of protective concentrations (PCs), and decreases in data quantity decreased community representativeness. There is a need to compile SSDs that reflect natural biotic assemblages; however, there are insufficient ‘high quality’ data to do this. Until more data become available, the use of right-censored data in SSDs is recommended. The use of such data can provide a more representative sample of species from which to estimate ecologically relevant PCs, and cannot result in under protective concentrations. Salinity is composed of multiple major ions, yet typically studies have used representative single salts (e.g. NaCl) or artificial sea water to determine its toxicity. This thesis researched the use of representative salts, finding that site specific and whole effluent toxicity testing produces more environmentally realistic salinity risk assessment of saline wastewaters. The thesis investigates mechanisms of salinity toxicity in two model species, Paratya australiensis (Decopoda) and Austrophlebiodes pusillus (Ephemeroptera), and their experience of temporally varying salinity. Mortality in P. australiensis co-occurred with osmoregulartory breakdown. However, latent (or delayed) mortality effects occurred when P. australiensis are returned to ambient waters, despite haemolymph salinity returning to a normal baseline level. This indicates that a rise in haemolymph salinity is not the sole cause of mortality from increased salinity in this species. Latent effects are not currently considered in salinity risk assessment, and are not incorporated into current water quality guideline derivation protocols; thus the effects of salinity may be underestimated. In A. pusillus osmoregulartory breakdown did not precede death irrespective of the temporal course of the salinity increase. Furthermore, significant mortality occurred well below the isotonic point. This finding contrasts with common osmoregulatory theory which implies no mortality below the isotonic point, and has not been previously reported. Thus, this research challenges the extent of our current understanding of the relationship between osmoregulation and mortality. Further studies are urgently needed on the osmoregulation of other salt sensitive species. Current salinity risk assessment and water quality guideline derivation protocols likely underestimate the effects of salinity on freshwater macroinvertebrates. Salinity risk assessment and water quality guideline derivation protocols need to compile SSDs that are more representative of the communities they aim to protect, and use ionic proportions of the salt source. Further research is required on the implications of temporally varying salinity and latent mortality effects, and into the physiology of osmoregulation for salt sensitive species