The seasonal pattern of dissolved oxygen, and hypolimnetic deoxygenation, in two tropical Australian reservoirs

Abstract

The dissolved oxygen concentrations in Darwin River Reservoir (DRR) and Manton River Reservoir (MRR), both located in the wet/dry tropics of Australia, were investigated over an 8‐year period. Average oxygen concentrations were lower in MRR than in DRR, indicating the dominance of consumptive processes not compensated by photosynthetic production of oxygen, despite MRR’s higher chlorophyll a concentration. With the onset of thermal stratification, dissolved oxygen in the hypolimnion of each reservoir was depleted at average rates of 3.4–7.1 mg L−1 month−1. These are higher than rates reported for temperate water‐bodies (0.10–3.0 mg L−1 month−1). This was ascribed primarily to the effect of temperature on microbial metabolism, while the influence of the reservoirs’ high epilimnion to hypolimnion volumetric ratios (approximately 10:1) and reservoir trophic state were considered secondary. Due to the temperature dependence of hypolimnetic oxygen depletion, the trophic classification of lakes based on hypolimnetic deoxygenation and anoxia is not globally applicable, but is applicable to water‐bodies of similar hypolimnetic temperatures, especially when morphometric influences are also taken into account. Both reservoirs experienced long periods of anoxia (average: 9 months in MRR, 5 months in DRR). In DRR, the extent of hypolimnetic anoxia was related to hypolimnion volume. Hypolimnetic anoxia in the reservoirs, when quantified by the anoxic factor (AF; number of days that a sediment area, equal to the whole‐lake surface area, is overlain by anoxic waters), was high compared with similar water‐bodies in North America, providing quantitative evidence that anoxia in more prevalent in tropical reservoirs. Ammonia, iron and manganese accumulated in the hypolimnion of both DRR and MRR, with concentrations decreasing exponentially from the sediments to the oxycline. Sediment phosphorus release was detected in MRR, but not detected in DRR, despite the reservoir’s long periods of hypolimnetic anoxia. The high water temperatures and lengthy periods of hypolimnetic anoxia of the reservoirs provided a favourable environment for anoxygenic phototrophic bacteria.