Benthic diatoms as bioindicators of a point source thermal discharge to an estuary

The work presented here describes a novel approach for determining past and present effects of a power station thermal plume discharge on an estuary. The technique utilised benthic diatoms, combining palaeoecological and contemporaneous multi-taxa approaches, to demonstrate that diatoms were not only responding to the plume in the modern day environment, but also indicated a changing temperature regime within the receiving water bay over time. The approach provided new information and demonstrated a range of techniques that may be employed for improved management of thermal discharges to enclosed coastal water bodies. To determine the spatial distribution of benthic diatoms in relation to a power station cooling field, the discharge at Vales Point, Lake Macquarie, was selected as the study site and first sampled in 2003. The initial fieldwork sampled sediments along a thermal gradient with increased distance from the discharge point. This spatial pattern was then replicated within two other southern lake control embayments. Multi-variate analyses determined that temperature and a number of other variables associated with the plume explained gradients in the diatom flora of the receiving bay. Satellite imagery and high resolution logger data then provided detailed plume and lake temperature information and indicated that thermal loading of lake water and thus the potential for ecological effects were expected to be greater in autumn/winter compared to summer. To determine changes in receiving bay water quality over time and to establish prepower station baseline conditions, a sediment core was obtained from Wyee Bay, sub sectioned, 210Pb dated and analysed for fossil benthic diatoms. Diatom-inference models were then developed based upon a localised reference dataset obtained from sample sites across Lake Macquarie and NSW central coast estuaries. The assemblage profile displayed periods (several years to decades) of relative homogeneity and heterogeneity (years to a decade), coincident with the major phases of power station operation. A change in assemblages was also observed around 1925-30 and prior to power station commissioning, indicating a change to receiving bay ecology at an earlier time. Although the salinity and temperature data were both adequate for modelling, reconstruction errors meant that only longer-term trends were discernable for temperature. When compared to real-time data (monitoring) variability in salinity within the core was relatively well represented by the model and appeared to respond to climatic factors such as SOI and rainfall. Longer-term salinity trends, however, were not as reliable as those for temperature. In an effort to improve the temperature-inference model for the Wyee Bay core an alternative reference dataset was developed by sampling across a natural temperature gradient. Triplicate samples from 13 estuaries from Noosa (southern Queensland) to Eden (southern NSW) (26 °S – 37 °S) were analysed for benthic diatoms and combined with environmental variables to establish a Temperature-Latitude (TL) dataset. Sensitivity testing was also conducted to examine the roles of eveness/unevenness (structure) within the sample design, numbers of species and taxonomic resolution on the multi-variate analyses and model output. While latitude, salinity and phosphate were the variables that consistently explained the greatest proportion of variability across the different datasets, salinity, temperature and nutrients were dominant when latitude was removed. Most models using TL datasets improved reconstruction errors; however, the nested (nTL1) dataset provided the best inferred-temperature history for Wyee Bay when validated by long-term monitoring data. Generally, the environmental variables attributable to gradients between lakes were the same regardless of the number of species, sample design or taxonomic resolution. To increase our understanding of the changes in diatoms and lake ecology over a greater pre-power station and pre-industrial (heavy metals) period two additional cores were obtained for the final phase of the study. At both a plume-effected and a control site cored sediments were analysed for heavy metals and 210Pb to establish and cross validate geochemical chronology and identify pre-industrial boundaries. While only sediments of the Crangan Bay core were preserved adequately for dating it provided key information relative to the original Wyee Bay core. Prior to heavy metal enrichment in the southern lake (1925- 1942), Crangan Bay (control) (20-42 cm) and Wyee Bay (pre-1935) supported a similar diatom flora. Thus, it was inferred that the environmental condition of the lake’s southern embayments were likely to be similar at that time and relatively stable for Crangan Bay to at least ~1790. The work described here demonstrated the applicability of benthic diatoms as tools for understanding spatial and temporal changes in a coastal lake associated with a power station thermal plume. Diatoms and the multi-taxa approach could be utilised as bioindicatortype tools for regular broadscale assessments of south-east coast estuarine health