Determination of the denitrification capacity of unconsolidated rock aquifers using 15N tracer experiments at groundwater monitoring wells - development of a new method to assess actual and future denitrification in aquifers

The nitrogen cycle is one of the most important nutrient cycles in terrestrial ecosystems and has been dramatically altered by increasing anthropogenic inputs of reactive nitrogen (Nr) to terrestrial ecosystems. Nitrate (NO3−) emissions from the agricultural sector are the dominant source of Nr fluxes to aquatic systems like aquifers and NO3− pollution of groundwater is a problem due to eutrophication of surface waters receiving polluted groundwater, possible indirect emissions of nitrous oxide (N2O) and increasing costs for keeping the standard for NO3− in drinking water. Denitrification is the most important process of NO3− attenuation in groundwater and is accompanied by an irreversible loss of reduced compounds, which leads to an inevitable reduction of the denitrification capacity of aquifers. Denitrification is difficult to measure and to predict on aquifer or river catchment scale. Therefore, knowledge about the denitrification capacity of aquifers is important for the designation of nitrate vulnerable zones. Against this background, the main objective of this thesis is to improve methods to measure denitrification in situ as well as in the laboratory and to develop an approach to estimate the denitrification capacity of aquifers. All investigations were conducted within two Pleistocene sandy aquifers in Lower Saxony, Germany, the Fuhrberger Feld aquifer, situated NE of the city of Hannover, and the Großenkneten aquifer SW of the city of Bremen. Both aquifers receive considerable NO3− inputs via seepage waters from agricultural fields and intense denitrification is known to take place within both aquifers. To improve laboratory measurements of denitrification, an automatic sampling and calibration unit coupled to a membrane inlet mass spectrometer (ASCU-MIMS) suitable for online measurement of denitrification (chapter 2) was developed and tested during a 15N tracer experiment with incubation of aquifer material from the Fuhrberger Feld aquifer. It was shown that online analysis of denitrification rates measured with ASCU-MIMS was in good agreement with the well established offline isotope analysis by GC-IRMS. From 3 investigated 15N aided mathematical approaches the approach given by Spott and Stange (2007) was found to be most suitable for the determination of denitrification from ASCU-MIMS raw data. The approach given by Mulvaney (1984) can be used under certain circumstances (chapter 2). The latter approach has the advantage that it is not necessary to analyse molecular ion mass 30, which is often biased. To estimate denitrification capacity of aquifers it was tested if the stocks of reduced compounds (SRC) of the investigated aquifer samples from both aquifers could be estimated from the measured cumulative denitrification during one year of anaerobic incubation (Dcum(365)) (chapter 3). Dcum(365) showed good linear regressions with the SRC of aquifer material from the reduced zone of both aquifers. From this finding it is concluded that Dcum(365) is a useful indicator for the denitrification capacity of aquifer material. Overall, median SRC (1.3 g N kg−1) and Dcum(365) (15.6 mg N kg−1 yr−1) of sulphidic aquifer samples was 5 respectively 10 times higher than the one of non-sulphidic samples. Initial denitrification rates measured at the beginning of incubations were poorly related to predict Dcum(365), indicating that short-term incubations are not suitable to predict the SRC. Among the tested sediment parameters total organic carbon (Corg) and KMnO4-labile organic carbon (Cl) yielded the best predictions of Dcum(365) for the whole data-set of aquifer material from both aquifers. Regression analysis revealed that for non-sulphidic and sulphidic aquifer material different sediment parameters yielded the best regressions with Dcum(365) and the SRC. The kinetics of denitrification during the conducted incubations could be described with zero-order kinetics, suggesting that the NO3ˉ concentration during the experiments was not limiting the measured denitrification activity (chapter 3). Denitrification in the non-sulphidic samples, where only organotrophic denitrification occurred, was kinetically much slower than in the sulphidic samples. Push-pull 15N tracer tests for the measurement of in situ denitrification rates (Dr(in situ)) were tested in the aquifers as an approach to estimate SRC without the need to collect and analyse aquifer material. These tests were carried out in groundwater monitoring wells at the same position and with filter screens in the same depths as the origin of the incubated aquifer samples. Dr(in situ) ranged from 0.0 to 51.5 µg N kg−1 d−1 and were lower than average laboratory rates (Dr(365)), especially for in situ measuring points in the NO3−-free groundwater zone of both aquifers where sulphides are mostly present. After pre-conditioning of one multilevel well by the repeated injection of NO3ˉ containing groundwater in the zone of NO3ˉ free groundwater to stimulate denitrifying bacteria, Dr(in situ) increased strongly which resulted in much better agreement of Dr(in situ) and laboratory Dr(365). From these results it is concluded that the deeper NO3‾-free groundwater zones of aquifers require pre-conditioning prior to Dr(in situ) measurements to obtain potential in situ denitrification rates reflecting SRC. It is assumed that this is due to the slow response of the microbial community to the initial input of NO3-.