Diatom communities in lakes and streams of varying salinity from south-west Western Australia : distribution and predictability

Abstract

The distribution pattern of diatoms from lakes and streams of varying salinity in the south-west of Western Australia was investigated. A total of 95 water bodies were sampled and separated into freshwater (<3 ppt), hyposaline (3-20 ppt), mesosaline (20-50 ppt) and hypersaline (>50 ppt). The south-west and specifically the inland wheatbelt region has been severely influenced by secondary salinisation, due to clearing of native vegetation for agriculture. There has been little research on diatom communities from salt-affected systems, with this data providing the basis for the development of an inference model based on species optima and tolerance limits to salinity.Physico-chemical variables measured from the study sites were collated and assessed. Salinity ranged from freshwater (0.04) to hypersaline (156.80 ppt), and pH ranged from acidic (2.90) to alkaline (10.51). Dissolved oxygen levels were recorded from 1.11 to 18.67 mgL[superscript]-1, water temperature from 6.30 to 28.10 ºC and peripheral vegetation scores from 1 (little or no cover) to 5 (dense cover). Analysis of variance (ANOVA) showed that salinities were significantly higher in standing waters located further inland, compared to flowing waters in high rainfall areas. Hypersaline wetlands displayed significantly lower dissolved oxygen levels, higher water temperatures and reduced peripheral vegetation, compared to freshwaters. The pH of hypersaline sites was also significantly lower, associated with surrounding land use or underlying geology. The data collected provides important baseline information, with implications for aquatic biota.The community structure of diatoms in relation to varying salinity concentration was explored. An artificial substrate collector (JJ periphytometer) was used to standardise sampling and ensure diatom assemblages were representative of ambient water quality. A total of 217 taxa were identified with the highest diversity observed in freshwater sites (up to 33 species), and limited to less than 15 in hypersaline waters. According to BIOENV, salinity was the key factor influencing diatom community structure. SIMPER analysis found a number of discriminating taxa between salinity ranges, specifically between assemblages from freshwater and hypersaline sites. Taxa such as Achnanthidium minutissimum and Gomphonema parvulum were indicative of freshwaters. In comparison, Amphora coffeaeformis and Nitzschia ovalis were associated with hypersaline water bodies.Diatom community structure was also examined from 20 hypersaline wetlands in the wheatbelt region with varying pH. Characteristic taxa including Amphora coffeaeformis, Hantzschia sp. aff. baltica and Nitzschia ovalis showed a wide tolerance to salinity and pH, or hypersaline acidic conditions. BIOENV analysis found there were no observable differences between diatom assemblages in relation to salinities above 50 ppt and that pH was highly correlated to species composition. The increasing occurrence of acid saline lakes is of concern and is most likely attributable to deep drainage practices and continued use of fertilisers in agricultural areas.A diatom-based transfer function was developed from the south-west dataset, to document species optima and tolerance limits to salinity. CCA analysis showed that salinity accounted for a significant and independent amount of variation in the diatom data enabling an inference model to be derived. The most successful model was generated using tolerance-downweighted weighted averaging, with a high coefficient of determination and low prediction errors that remained high after jackknifing. The optima of freshwater diatoms were similar to those reported from other regions of the world, although the optima of hypersaline species tended to be higher. Comparatively, the model performed very well, with the potential to be applied in future paleolimnological studies.In conclusion, this study has shown diatoms to be effective biomonitoring tools, providing the basis for future sampling strategies that assess the biodiversity of salt-affected water bodies in Western Australia. Potential indicator species from different salinity ranges were identified and the sensitivity of diatoms allowed for the development of a statistically robust inference model for the south-west. The reported optima and tolerance limits of important taxa may be further explored to evaluate the success of remediation measures implemented for secondary salinised systems in this region.