Assessment of biogeochemical transformation of nitrate in shallow groundwater in the agricultural landscape : a thesis presented in partial fulfilment of the requirements for the degree of Master of Science in Soil Science at Massey University, Manawatu, New Zealand

Abstract

In groundwaters, denitrification or subsurface denitrification (SD) has been identified as a key attenuation process. Where leached nitrate (NO₃⁻) can be reduced to dinitrogen (N₂ — a harmless gas), offering an ecosystem service in terms of reducing water pollution. However, partial denitrification (PD) can release nitrous oxide (N₂ — a greenhouse gas), resulting in a pollution swap from liquid to gaseous pollution and adding to greenhouse gas emission. There is very limited information available about occurrence, characteristics and dynamics of subsurface denitrification in shallow groundwaters across New Zealand agricultural catchments.

We studied 6 pastoral farms (DF 1, 2, 3; SC 1, 2, 3; ARM 1, 2, 3; CAM 1, 3; SR 1, 2, 3; BUR 1, 2 and 3) located in various hydrogeological settings in the Manawatu and Rangitikei Rivers catchments, in the lower North Island of New Zealand. We collected 7 sets of monthly groundwater observations at 17 piezometers from March to September 2018 to characterize the groundwater monthly chemical variations. The collected groundwater samples were analyzed for groundwater redox status, including dissolved oxygen (DO), oxidation-reduction potential, pH, NO₃⁻-N, iron (Fe²⁺), manganese (Mn²⁺) and sulphate (SO₄²⁻). We also conducted a set of push-and-pull tests to gain insights into dynamics of subsurface denitrification occurring in the groundwater samples at the study sites. We quantified changes in concentration of NO₃⁻-N, Br⁻ (tracer), dissolved N₂O-N and excess N₂ during the push-and-pull tests.

Our results suggested a spatially variable groundwater redox conditions and SD occurring across the study sites. The piezometers DF 2, 3; SC 1, 2; CAM 3; ARM 1, 2 and 3 showed anoxic redox status. Only the piezometers SC 3 and CAM 1 presented mixed redox condition. While the piezometers DF 1; SR 1, 2, 3; BUR 1, 2 and 3 indicated oxic conditions with some variability over the study.

Nitrate is being reduced in the anoxic piezometers DF 2, 3; SC 1, 3; ARM 1, 2, 3 and CAM 3, showing no NO₃⁻-N accumulation (< 0.5 mg L⁻¹). One of the piezometers with mixed redox condition (CAM 1) showed NO₃⁻-N accumulation (> 6 mg L⁻¹) while the piezometer SC 3 showed variability in NO₃⁻-N accumulation ranging from 0.02 mg L⁻¹ to 22.56 mg L⁻¹. The oxic piezometers SR 1, 2, 3; BUR 1, 2 and 3 showed NO₃⁻-N accumulation (> 3 mg L⁻¹) except for piezometer DF 1 that showed variability in NO₃⁻-N concentrations ranging from 0.01 mg L⁻¹ to 3.75 mg L⁻¹ over the study.

The concentrations of the electron donors Fe²⁺ and Mn²⁺ were found to be suitable for SD on anoxic piezometers DF 2, 3; SC 1, 2; CAM 3 and ARM 1, 2, 3 (> 1 mg L⁻¹ and > 0.05 mg L⁻¹ respectively). The piezometers with mixed redox status SC 3 and CAM 1 ranged just over the redox threshold for identifying redox processes (0.1 – 1.0 mg L⁻¹ and > 0.05 mg L⁻¹ respectively). In general, the piezometer with oxic redox status (DF 1, SR 1, 2, 3 and BUR 1, 2, 3) showed [Fe²⁺] and [Mn²⁺] below the threshold for identifying redox processes (< 0.1 mg L⁻¹ and < 0.05 mg L⁻¹ respectively) and not suitable to support SD.

The dominant terminal product of SD, whether was complete denitrification (N₂ — as end product) or partial SD (N₂O — as end product) spatially varied according to the redox status of the groundwater. Push-pull test results showed an increase in excess N₂ and N₂O-N concentrations at DF 3, ARM 3, CAM 3, BUR 3. The push-pull test conducted at SR 3 and SC 3 showed inconclusive results. Piezometers CAM 3 and ARM 3 showed the highest suitable conditions for SD followed by DF 3. Piezometer BUR 3 showed the highest partial SD rate. Therefore BUR 3 is considered in general, the less suitable piezometer for SD.

Our observations highlight the influence of different hydrogeological settings on spatial variability of partial (pollution swamp) or complete (ecosystem service) SD in shallow groundwaters. A better understanding and quantification of spatial and temporal variability of SD process will support information, design and formulation of targeted and effective management measures for sustainable agricultural production while protecting soil, water and air quality.