Understanding the ecological basis for health, treatment and aesthetic risk: Studies of a small unregulated ephemeral subtropical Australian catchment

Globally, drinking water supplies are unique as a result of differing water storage capacities, climate, regional geology, treatment types and anthropogenic effects. Traditionally, middle to large sized waterbodies have dominated ecological studies about water supplies. However, small water storages are used to supply a significant portion of human water demands, particularly in geographically isolated locales. Despite this, there is a dearth of information on small waterbodies that are more susceptible to environmental change during natural rainfall events or anthropogenic impacts such as contamination. This study investigated Baffle Creek, a small relatively unmodified, unregulated subtropical ephemeral catchment located on Australia’s eastern seaboard, and the southernmost catchment to drain into the Great Barrier Reef Marine Park. The aim was to identify potential health, treatment and aesthetic risks, with particular reference to metals and phytoplankton risk in the drinking water supply. Data collection was performed over 28 months including two wet and two dry seasons. The study consisted of five surface water sites and two supply chain infrastructure sites. Water quality variables and phytoplankton concentrations were assessed monthly. Metal and nutrient concentrations were quantified from water and soil samples collected quarterly. Analysis of Similarity (ANOSIM) results indicated that water quality variables within the upstream, midstream and downstream sections of the catchment were statistically different and thus likely to pose different risks to the water supply. In the upstream waters, risks included pH reduction combined with an increase in colour (organic matter), whereas the dominant risk in the downstream sections was elevated conductivity. A conceptual model was created to facilitate visualisation of differences between the sections of the catchment and in turn, aid with management decisions. The hydrogeochemistry of Baffle Creek was predominantly driven by rain events. Following a flushing event, concentrations of dissolved aluminium and nitrate increased in the surface and drinking water, whereas manganese and ammonia were undetectable in the drinking water and occurred only in small concentrations in the surface waters. In contrast, when rainfall events occurred without an associated flush, manganese, iron and ammonia concentrations increased in the drinking water. The hydrochemistry of manganese and iron in the supply chain infrastructure was positively correlated with ammonia concentrations, and aluminium concentrations were correlated with nitrate concentrations. Currently the drinking water supply does not pose a threat in regards to chronic manganese exposure in humans; however, elevated concentrations occur periodically; and may become more problematic with climate change scenarios. Hierarchical agglomerative clustering with the addition of a similarity profile test (SIMPROF) of phytoplankton data and comparison with a principal component analysis of corresponding environmental data were used to validate the relationships between environmental gradients and variations in the phytoplankton assemblages. A biological-environmental stepwise test (BEST) was performed to assess correlations among environmental variables with the biological data, describing which changes in environmental variables were associated with the ecological composition of phytoplankton communities. The Spearman’s rank correlation between environmental conditions and phytoplankton communities was significant p < 0.05, however, weak ρ = 0.329. Variable rainfall - and therefore streamflow - led to the proliferation of genera adapted to the dynamic instream conditions, with picoplanctonic phytoplankton (Aphanocapsa and Nannochloris) being predominant in abundance and flagellated phytoplankton predominating from a bio-volume perspective (Trachelomonas, Peridinium, Cryptomonas and Euglena). There were 66 phytoplankton genera identified and from this, 10 coherent “functional groups” were statistically identified with certain environmental tolerances and sensitivities, as well as potential health, treatment and aesthetic considerations. There were 13 genera of cyanobacteria identified which may be potentially toxic, 11 phytoplankton genera that may cause treatment risk and 20 genera that may be an aesthetic risk. A novel quantitative approach was utilized for grouping of environmental variables, phytoplankton assemblages, tolerances, sensitivities and multiple-risk information to assist water managers to have a more precise understanding of the suite of risks posed by phytoplankton populations in ephemeral subtropical systems. Based on field data, a semi-quantitative risk matrix methodology was developed for use as a supplement to existing Water Safety Plans. For example, the primary risk identified for the Baffle Creek catchment was the presence of Cyanobacteria in the pre-chlorinated supply chain, specifically Limnothrix redekei. To respond to this, a preliminary proposal was suggested involving use of clean sand on the sub-surface layer of the bank filtration, complemented with biologically active sand as a surface cap. Furthermore, culturing techniques reported in this study can potentially be used to optimize assessment for L. redekei populations surrounding water extraction points. The present study provides insights as to some of the most important factors affecting the safety and palatability of drinking water from small water supplies. Furthermore, models were developed to provide managers of water treatment plants with additional information to help inform decision management and management of small water supplies, particularly in subtropical environments globally.