This study examined the correlations between watershed and reservoir characteristics, and water quality parameters related to algal blooms in seven subtropical reservoirs. Analysis of the dissimilarity of physico-chemical parameters resulted in separation of the reservoirs into three main groups: four reservoirs with the highest proportion of agriculture and/or urban land use in their watersheds; two reservoirs with a high proportion of forest cover; and one small reservoir with a relatively pristine watershed intermediate between the other two groups. All reservoirs were dominated by cyanobacteria, and at times, had species capable of producing toxins. However, the three reservoirs with the lowest percentage forest cover ( approximately 50%) had the highest frequency and magnitude of toxic species, principally Cylindrospermopsis raciborskii. Analysis of dissimilarity of algal species composition resulted in three reservoir groups similar to that for the physico-chemical parameters, with the exception of the reservoir with the highest percentage urban land use being an outlier. Across all reservoirs, percentage forest cover in the watershed, watershed area and reservoir volume were all significantly correlated with algal cell concentrations and total nitrogen (TN), but not with chlorophyll a concentrations. Total phosphorus (TP) was only correlated with the proportion forest cover in the watershed, suggesting that reservoir volume and depth were of less importance for TP than for algal cell concentrations or TN. These results suggest that watershed pattern and reservoir characteristics, such as water volume and depth, have a measurable effect on the type of algal blooms in reservoirs.

This study examined the sources and fate of nutrient inputs from two principal tributaries to the eutrophic subtropical Wivenhoe reservoir: an unregulated river and a dammed river with regular releases, during a period of declining reservoir water levels. Nutrient budgets were constructed over a period of 6 years, and combined with short-term data on nutrient concentrations and forms, and d 15 N stable isotope data. Our study found that over a 6 year period, there was net retention of phosphorus (P) in the reservoir, with 60% of inputs retained. Most of the P input load came from the unregulated river, with an agricultural catchment, during periods of high flow. During one event half of the total TP load from the unregulated river in the study period was delivered in only 12 days. Much of the P was dissolved inorganic P (DIP) and was derived from high P concentrations in soils and sediments. This highlights the importance of appropriate catchment management practices to reduce P losses from terrestrial systems because retention of P in reservoir sediments reduces the availability of this nutrient for agricultural production. In contrast, there was negligible retention of nitrogen (N). The unregulated river was an important source of N derived from N fixation in the river and adjacent soils, while the source from the dammed river was mostly reprocessed N. The high retention of P relative to N is consistent with relatively higher accumulation of P in sediments.