Results

Another problem to be addressed before an AusRivAS approach could be used in estuarine health monitoring is to understand seasonal patterns in estuarine benthos. In southern Australia estuarine macrobenthos shows strong temporal but not seasonal variability. In tropical and subtropical estuaries there are strong seasonal patterns in numbers and diversity, though numerically abundant taxa change from year to year. When using AusRivAS models to assess river health it is assumed similar seasonal patterns exist from year to year. Consequently, for AusRivAS, providing samples are all collected in the same seasons, the reference samples and test samples do not have to be collected in the same years. At least in southern Australia this may not be true for estuaries, so that a new set of reference samples would need to be collected whenever a test was required. This would be prohibitively expensive.

Considering that the AusRivAS approach could not be used in all estuaries and that the sampling effort required to give reliable data would be relatively expensive, it was decided to review other options for looking at estuarine health. Environmental health can be indicated by the macrobenthos community structure; the number of species present, the distribution of individuals among the taxa, the proportion of rare taxa, etc. Such community descriptors can be depicted graphically in k-dominance curves, plots of cumulative percent abundance versus species ranked from most to least common. The problems with using k-dominance curves for assessing national or regional environmental health is that only a limited number of curves can be compared on each plot.

Our field work generated a large database that could be used to construct k-dominance curves for Victorian and southern New South Wales estuarine communities. We investigated multivariate analysis to compare large numbers of curves with each other and test sites. Community indices, many of which described the k-dominance curve provided a multivariate ordination analysis of the samples. The ordination plots could be divided into sections containing communities indicative of good, questionable and poor environmental health.

We demonstrated that this ordination method could be used to rate samples from

Power Stations discharge thermal effluent into estuaries

estuaries elsewhere in Australia. Following improved treatment of effluent being discharged from a Tasmanian pulp mill over time there was a clear movement from communities indicative of poor and questionable environmental health to good environmental health. When analysed with the south eastern Australian data, samples from Western Australian and Central Queensland estuaries were also found to fit into the same pattern with samples indicative of stressed communities falling in the same part of the ordination. Thus it was shown that the database of estuarine community structure we had developed could be used as a starting point to assess Australian estuarine health.