Phosphorus dynamics in the sediment of a eutrophic lake

Since the sixties eutrophication has been recognized to affect the quality of surface waters. The prolonged loading with nutrients has led to high algal concentrations in the water and to concommitant environmental problems such as the depletion of oxygen, the production of toxins and the development of tedious animal populations e.g. bream. A series of management measures have been carried out to combat eutrophication in the Netherlands. However, despite all efforts, the majority of the dutch surface waters is still considered to be eutrophic. The disappointing results are often attributed to processes in the sediment that can delay the improvement of the water quality. This thesis deals with the phosphorus dynamics in the sediment of an eutrophic lake. In contrast to most studies which are chemically orientated, attention here is mainly given to the role of microbial processes in the phosphorus cycle. The research was performed in eutrophic Lake Loosdrecht where the external phosphorus loading was recently decreased from 35-40 to 10-15 mmol.m ^-2.y ^-1.

The importance of microbial processes for the release of phosphate by the sediment was investigated by comparing sterilized and non-sterilized columns (chapter 2). Columns were sterilized by γ-irradiation (25 kGy). The combination of temperature and γ-irradiation experiments made it possible to distinguish between microbial and physico-chemical processes. The release of dissolved phosphate from the sediment is controlled by microbial processes on a short-term and a long-term basis. Microbial mediated release responds directly on an increase in temperature. This is probably due to the induction of changes in the chemical environment such as a decrease in oxygen content of the surface layer. Mineralization of organic matter results in a mobilization of phosphate and is prerequisite to sustain the phosphate release on a long-term basis.

Phosphate fluxes over the sediment-water interface were calculated using measured concentration gradients in the pore water and were compared to fluxes measured under laboratory conditions (chapter 3). Results were analysed with a statistical method (Redundancy Analysis) to detect patterns of variation in pore water chemistry and in measured and calculated fluxes, that could be ascribed to environmental variables. Initial fluxes of phosphate measured in sediment columns, which varied between -7.7 and 1330 μmol.m ^-2.d ^-1, correlated significantly with the calculated fluxes over the sediment-water interface. The high correlation between calculated fluxes of ammonia, phosphate and methane and measured initial flux of phosphate, conclusively pointed to mineralization of organic matter as driving force for phosphate release from the sediment. Redundancy Analysis demonstrated that the rates of mineralization and of phosphate release are high in autumn. This was ascribed to an increased sedimentation at the end of the growing season.

The importance of anaerobic mineralization processes fluctuated seasonally (chapter 4). At high anaerobic mineralization rates (>600 μmol organic carbon M ^-2h ^-1), sulfate reduction was limited by sulfate and methanogenesis accounted for over
80% of the total. At low anaerobic mineralization rates, measured in winter and spring, sulfate reduction was predominant. There was little methanogenesis below 5 cm depth in the sediment which indicated a rapid decrease of degradable organic
matter with depth.

A new method was develloped to quantify the contribution of bacterial processes to the phosphate uptake of aerobic freshwater sediment (chapter 5). The method was tested on iron hydroxyphosphate that was either synthesized or formed under in situ conditions, and a pure culture of Acinetobacter 210 A. Using a mild acid extraction we could distinguish between chemical and biological phosphate uptake. This method allowed the solubilization of the entire iron hydroxyphosphate fraction but did not extract bacterial phosphate.

The method was applied to determine the contribution of bacterial processes to the phosphate uptake of the surface sediment. Phosphate uptake of randomly sampled surface layers of the sediment was considerable and ranged from 12 to 138 μmol.g ^-1on a dry weight basis. Phosphate uptake was correlated positively with the amount of extractable iron and phosphate and negatively with dry weight. The contribution of bacterial processes ranged from 12 to 32 %. Addition of an easily degradable substrate, such as acetate, to the sediment stimulated the uptake of phosphate and augmented the biologically bound phosphate fraction.

A diffusion chamber was designed to investigate the effect of an enhanced oxygen consumption of the surface sediment on the phosphate flux (chapter 6). The diffusion chamber consisted of two compartments separated by a Teflon membrane. In the upper part a thin sediment layer was present and the lower part was continuously flushed with gas. The hydrophobic membrane allowed for diffusion of gasses from the lower part through the sediment layer towards the headspace of the upper part. In the diffusion chambers the methane oxidation was artifically increased to 9.8 mmol.m ^-2.d ^-1. This resulted in an increase of the oxygen consumption rate by a factor two compared to controls without methane oxidation (8.6 vs 17.7 mmol.m ^-2.d ^-1). The methane oxidation significantly decreased the oxygen penetration depth (2.5-4.0 vs 1.0-2.0 mm). However, despite the shrinkage of the oxidized microlayer, no differences were found in phosphate flux over the sediment water interface. Batch experiments with standard additions of methane revealed that the growth of methanotrophic bacteria contributes to the phosphate uptake of aerobic sediment. Results indicated that a decrease in chemical phosphate adsorption caused by a decease in the oxygen penetration depth, could be compensated for entirely by the growth of methanotrophic bacteria.

Finally it was concluded that the eutrophic conditions in Lake Loosdrecht are maintained by a relatively small but dynamic pool of phosphate (chapter 7).
Calculations with a simple model indicated that restoration measures such as dredging or addition of iron(III) compounds will not result in a long-term improvement of the water quality. The construction of isles and dikes might induce a decrease in turbulence and thus contribute to an improvement of the water quality. A further reduction of the external phosphorus loading is prerequisite to reduce the amount of phosphorus compounds in the water. However, the changes in the structure of the ecosystem that have been induced by the eutrophication might be irreversible. Whether Lake Loosdrecht can be restorated into a clear water system dominated by macrophyts depends on ecological interactions between bacteria, algae, zooplankton and fish in the ecosystem but is impossible to predict.