Nitrogen removal by denitrification in the sediments of a shallow lake

Most surface waters in the Netherlands are highly eutrophicated due to high loadings with the nutrients nitrogen (N) and phosphorus (P). To improve the water quality of lakes often the phosphorus loading is reduced. Due to phosphorus release from the sediments the success of the recovery of these lakes, however, is limited. Therefore renewed interest is directed to the fate of nitrogen in surface waters: perhaps a reduction of the N loading may result in a better water quality. This interest is furthermore achieved by the international programs for protection and restoration of the North Sea and the river Rhine.In the nitrogen cycle bacterial processes in the sediment are very important. When the external nitrogen loadings are reduced, these benthic processes (resulting in the internal loading) dominate more and more the nitrogen loading and the concentration in the overlying water. Little is known, however, about the magnitude of the various nitrogen process rates or about the effects of reduction of the nitrogen loading.The aim of this study was therefore to investigate and quantify the important microbial N processes (ammonification, nitrification and denitrification) and the related fluxes to the overlying water, in order to obtain a better insight in the possibilities to diminish eutrophication by reduction of the N loading. Special attention was paid to the denitrification as by this process nitrogen can be removed from the sediment-water system as N 2 to the atmosphere. The experiments in this study were performed with sediments sampled from the shallow lake Wolderwijd/Nuldemauw, The Netherlands.In the literature various methods to measure denitrification rates are described. In Chapter 2 the results of three of these methods, the N 2 flux Method, the 15N isotope pairing technique and the mass balance approach, are compared. Results obtained with the N 2 flux method were in agreement with mass balance data, but were higher than the results obtained with the 15N isotope pairing technique. Various checks demonstrated that after a pre incubation period of about 10 days the coupled denitrification can be estimated very well with the N 2 flux method and that the N 2 flux is not due to leakage of atmospheric nitrogen. An important assumption for the calculation of the coupled denitrification by the 15N isotope pairing is the uniform mixing of the nitrate species ( 14NO 3-and 15NO 3-)When micro-sites exist in the sediment, uniform mixing of the nitrate species probably cannot be assumed because of the very tight coupling between nitrification and denitrification at these sites. It is therefore suggested that the difference in the results is caused by an underestimation of the coupled denitrification by the 15N isotope pairing technique, due to the occurrence of micro-sites. In the further research the N 2 flux method was used.In order to estimate the ammonification rates beside the nitrification and the denitrification rates also anoxic cores were incubated. From sediments incubated under anoxic conditions and without nitrate and/or nitrite in the overlying water unexpected but remarkable N 2 fluxes were estimated (Chapter 3). Thorough investigations demonstrated that these fluxes were not due to N 2 leakage or oxygen leakage followed by nitrification and denitrification. Therefore it was concluded that the anoxic N 2 fluxes were real. No explanation, however, for the source and/or the mechanism leading to these N 2 fluxes could be given, although ammonium oxidation by Fe and Mn might be possible.The influence of temperature and organic matter content on the various N fluxes was investigated by incubating muddy and sandy sediments sampled throughout the year at 2, 12 and 23 °C (Chapter 4). As expected N tot (NH4 ++ NO x-+ N 2 ) fluxes increased with increasing temperature. A temperature factor of 1.9 ± 0.3 for a 10 °C increase was estimated for both sediment types. At the same temperature the N tot fluxes from muddy sediments were generally higher than from the sandy sediments and the species composition of the N tot flux differed with the sediment type. This was due to the higher organic matter content of the muddy sediments. In general especially the denitrification was influenced by the organic matter content: 75-90% of the N tot flux was denitrified in the muddy and only 45-65% in the sandy sediments. However, in cores from both sediment types collected just after a spring bloom of phytoplankton, the contribution of the various N species to the N tot flux was changed: much higher NH 4+and much lower NO x-and N 2 fluxes were calculated, especially at the high incubation temperature. It was assumed that this shift in species composition was mainly due to the freshly settled and easily degradable organic matter. At higher concentrations of easily degradable organic matter more oxygen is consumed by the oxic mineralization and CH 4 oxidation and no or less oxygen remains for the oxidation of nitrogen. Enhanced temperatures stimulate this further. A direct proof and quantitative data regarding the influence of easily degradable organic matter on the various nitrogen processes, however, are not given in this study.Several factors like the O 2 availability and the (freshly settled) organic matter content may vary throughout the lake and can cause spatial variability in the nitrogen fluxes. Spatial variability might influence therefore the total N loss of a lake as well. The spatial variability of the N fluxes was investigated by collecting sandy and muddy sediment cores throughout lake Wolderwijd/Nuldernauw and measuring the sediment characteristics and nutrient fluxes (Chapter 5). Within the same sediment type the sediment characteristics and nutrient fluxes showed little variation, whereas between sediment types significant differences were found. In comparison with the spatial variability of terrestrial soils however, this difference is small. It was hypothesized that this was due to the much smaller influence of micro-sites in lake sediments than in terrestrial soils.By principal component analysis (PCA) performed on the sediment characteristics and/or the nutrient fluxes the main sediment types were readily distinguished. It was therefore concluded that the N fluxes depended on the sediment type, although no obvious and direct relationship between the sediment characteristics and the N fluxes measured was found.Once enough light can penetrate to the sediment-water interface, benthic algae can develop. Their presence can influence the release of nutrients (Chapter 6). In laboratory release experiments, performed under dark and light conditions to simulate the situation without and with benthic algae, it was demonstrated that benthic algae were able to grow on nutrients released from the sediments. The flux of inorganic nitrogen to the overlying water thus decreased due to benthic algae. Due to an indirect effect furthermore the loss of nitrogen to the atmosphere increased. The photosynthesis of the benthic algae caused an increase of the 02 penetration and therefore the coupled nitrification-denitrification was stimulated. In this study the coupled denitrification was stimulated by the benthic algae with about 50%. Hence, benthic algae can reduce the nutrient release from the sediment and may accelerate the rate of recovery from eutrophication.In Chapter 7 the results of the laboratory experiments are interpreted for the field situation by comparing these results with data from mass balances of 1985-1992. Based upon the results from the experiments a general concept, showing the combined influence of temperature and (easily degradable) organic matter on the N loss due to coupled denitrification, is postulated. The concept implies that initially the coupled denitrification increases with increasing contents of organic matter and/or increasing temperature. A further increase however, changes the character of the sediment into more anoxic conditions and the coupled denitrification decreases. Therefore the N 2 fluxes measured in the laboratory at low temperatures may be a correct estimation for N 2 flux in the colder months in the field situation. If however the field temperature and/or the availability of easily degradable organic matter increase, the laboratory N loss may no longer be representative for the N loss in the field.From the comparison of the two estimates it was concluded that in the field situation on average about 60% of the N tot input is removed on an annual base, of which about 35% by the uncoupled and about 65% by the coupled denitrification. This clearly demonstrates the importance of the coupled denitrification in the annual N removal from a lake. In individual months however, storage in the sediment may be an important process in the total N loss. Although in each month also about 60% of the N tot input was removed, the contribution of the various N loss terms is highly variable. Based upon the available data, it is in general impossible to discriminate between the coupled denitrification and the storage of nitrogen in the sediments in spring and summer.It was concluded that reduction of the N loading, along with a reduced P loading may result in an improvement of the water quality. For the perspectives for lake restoration by reduction of the N loading it was however hypothesized that in situations with a high algal biomass a significant improvement cannot be expected immediately: the N loss by denitrification initially will remain limited and no sufficient reduction of the stock of nitrogen is obtained. However, the stock of nitrogen will reduce and once the organic matter production decreases, the improvement of the water quality will accelerate.In the last chapter the results obtained are compared with data from literature and the occurrence of micro-sites and the dependence of the coupled denitrification from the nitrate concentration in the overlying water are discussed. Furthermore the requirements of a model that is suitable to analyze the various processes occurring in a sediment column and of a model capable to predict the whole lake N cycle are discussed. Finally the main conclusions and recommendations are presented: several nitrogen transforming processes can be quantified and the influence of temperature, sediment type and benthic algae is clearly demonstrated. Furthermore several indications were found for the influence of fresh and easily degradable organic matter in the various nitrogen processes and for the importance of the coupled nitrification- denitrification in the total N loss in the field situation. For a good interpretation of the laboratory results to field conditions, however, more quantitative data are needed concerning the effect of the availability of easily degradable organic matter on the nitrogen processes.