Characterisation and potential optimisation of seepage wetlands for nitrate mitigation in New Zealand hill country : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Soil Science, Massey University, School of Agriculture and Environment, Palmerston North, New Zealand

Thumbnail Image
Open Access Location
Journal Title
Journal ISSN
Volume Title
Massey University
The Author
Diffuse nitrate (NO₃⁻) loss to pastoral waterways in hill country headwater catchments is a water quality concern in many countries with pasture-dependent economies, including New Zealand (NZ). Sheep and beef farming is the dominant land use in NZ hill country which are often located in headwater catchments. As these primary industries strive toward production growth to meet global demand for meat exports, this agricultural intensification will introduce more NO₃⁻ to its waterways. This contrasts with the recently enacted National Policy Statement for Freshwater Management 2020 (NPS-FM) which recognises the significance and calls for the protection of small wetlands in recognition of their ecosystem services including nutrient regulation, water quality improvement as well as associated social well-being. Nitrate mitigation in low-order streams in pastoral headwater catchments are important due to their proportionally large catchment coverage and major contribution to the national NO₃⁻ load to NZ rivers. Seepage wetlands in hill country landscapes can be a N-sink and, therefore, is a potentially cost-effective and natural NO₃⁻-mitigation tool for improved water quality from the pastoral headwater catchments. Seepage wetlands are features that occur along low-order streams in the low gradient of hill country landscapes. Their organic matter-rich sediment, saturated conditions and locations at the convergence of surface and subsurface NO₃⁻ rich flow pathways make seepage wetlands a unique landscape feature in terms of NO₃⁻ reduction via denitrification processes. However, denitrification is spatially and temporally variable as the process is influenced by the wetland sediment and hydrological properties. Several studies have demonstrated that seepage wetlands can be a potential NO₃⁻ sink and have quantified high sediment denitrification capacities in individual wetlands. However, variations in sediment and denitrification properties across a range of wetlands and a comprehensive study of seepage wetland hydrological characteristics that influence NO₃⁻ attenuation have not been undertaken, particularly in pastoral hill country landscapes in NZ. This thesis has examined the spatial variabilities of seepage wetland denitrification and the denitrification-influencing sediment properties across four hill country seepage wetlands within the Horizons Regional Council administrative boundary in NZ. The spatial gradients of sediment properties were examined vertically (at 15 cm depth intervals) and horizontally (within- and between- wetlands) in seepage wetland sites. Sediment physicochemical (water content (WC), pH, Eh) and chemical properties (dissolved organic carbon (DOC), NO₃⁻, NH⁴⁺, %total carbon or %TC, %total nitrogen or %TN, C:N, dissolved Fe²⁺ and dissolved Mn²⁺) and sediment denitrification enzyme activity (DEA), that represents sediment denitrification capacity, were quantified. The DEA values were highest at the surface depths across all wetland sites. Based on the wide range (560-5371 µg N₂O-N kg⁻¹ DS h⁻¹) and distinctive surface DEA values, the seepage wetland study sites were categorised into high-performing H-DEA (>3000 µg N₂O-N kg⁻¹ DS h⁻¹) and comparatively low-performing L-DEA (<1000 µg N₂O-N kg⁻¹ DS h⁻¹) sites. The H-DEA sites measured 7 to 10 times higher surface DEA values compared to the L-DEA sites. Spatial variability of denitrification in seepage wetlands was mainly driven by sediment WC, NO₃⁻, %TC, %TN, C:N, dissolved Fe²⁺ and dissolved Mn²⁺ (p≤0.05). The H-DEA site measured high WC (78%) which was above the threshold for denitrification and high sediment NO₃⁻ (15.9-18.5 mg NO₃⁻N/kg DS), in contrast to the L-DEA sites (WC 39.8-37.4%, 2.5-3.97 mg NO₃⁻N/kg DS). The heterogeneity of WC explained the heterogeneous distribution of DEA within the individual L-DEA sites. The sediment properties accounted for only 58-73% of the overall spatial variability in DEA, suggesting that additional wetland characteristics such as wetland hydrology, could have an important influence on denitrification in seepage wetlands. The seepage wetland hydrology and associated NO₃⁻ removal were characterised in detail at one of the L-DEA sites located on Tuapaka farm. During the hydrological characterisation, streamflow discharge and water quality were monitored at inflow and outflow for a 2-year period (June 2019-May 2021). Shallow groundwater quality was monitored at the 0.5, 1 and 1.5 m depths at the inflow, midflow and outflow positions in the wetland for a 1.5-year period (November 2019-May 2021). The seepage wetlands site demonstrated a stream inflow-dominated hydrology (83-87%) with small seepage contributions (8-14%) to the seepage wetland hydrology. Precipitation was found to be the major hydrological and associated NO₃⁻ removal (means attenuation) driver in the seepage wetland site. The seepage wetland was found an overall NO₃⁻ sink that on an average removed 23% of the annual NO₃⁻ inflow. Compared to the stream inflow (<0.03 mg NO₃⁻N/L), higher shallow groundwater NO₃⁻ concentrations (<0.11 mg NO₃⁻N/L) suggests that seepage is potentially an important NO₃⁻ source in these wetlands. High flow conditions, high winter precipitation and direct grazing during low flow periods are potentially major NO₃⁻ loss hot moments. In contrast, initial rapid infiltration at the onset of high precipitation events in early winter and spring and dissipated flow conditions highlighted opportunities for NO₃⁻ attenuation in the wetland and were identified as major NO₃⁻ removal hot moments. An overall dissipated flow condition driven by seasonally equivalent precipitation (22% of annual precipitation in winter) facilitated considerably higher annual NO₃⁻ removal of 40.8% (2.78 kg NO₃⁻N) in the wetland in year 2, in contrast to very low NO₃⁻ removal (0.3%, ~0.02 kg NO₃⁻N) under an erratic annual precipitation distribution (38% of annual precipitation in winter) in year 1. These findings suggest there is scope to improve NO₃⁻ removal by optimising flow conditions to slow flow in seepage wetlands to minimise NO₃⁻ loss during NO₃⁻ loss hot moments. In a follow-up laboratory-scale seepage wetland intact sediment column experiment, the effectiveness of diffuse flow, via subsurface outflow, was investigated for the optimisation of the wetland NO₃⁻ removal. During the experiment, the flow intervention altered the NO₃⁻ reduction-constraints observed in the preceding hydrological study and facilitated anaerobic conditions conducive to denitrification to capitalise on the sediment denitrification capacity, which was quantified during the preceding seepage wetland sediment characterisation study. The flow intervention involved vertical downwelling of NO₃⁻ rich (5 mg NO₃⁻N/L) pastoral surface runoff and subsequent horizontal discharge through a subsurface sediment column depth of 15 cm depth, collected from the Tuapaka seepage wetland site. The effectiveness of the subsurface drainage intervention for NO₃⁻ removal was assessed by monitoring the subsurface outflow water quality. The study showed that flow intervention achieved 50-96% NO₃⁻ removal from NO₃⁻ rich surface runoff. Based on the observations from the column study, two separate optimal operational HRTs of 2 and 13 hr are recommended to achieve large NO₃⁻ removal (50% from NO₃⁻ input of 5 mg NO₃⁻N/L) in a short period of time and large reduction in NO₃⁻ concentration at the outflow (<0.15 mg NO₃⁻N/L), respectively. The reasonably short period of HRT for such high NO₃⁻ removal efficiency (50-96%) supports the potential for the application of subsurface outflow intervention as a practical in-situ NO₃⁻ mitigation strategy, which warrants further research. This study also acknowledges the associated technical limitations of translating the laboratory-based findings to the field scale and recommends future studies to overcome these research limitations including high sediment compressions during intact sediment column samplings from the field, for example. The thesis not only demonstrates a flow intervention strategy to improve NO₃⁻ mitigation via flow regulation in seepage wetlands, but also guides future management by identifying the potential seepage wetland hot spots in the landscape (chapter 3) and the NO₃⁻ removal hot moments in the wetlands (chapter 4) and also by recommending necessary HRTs for flow intervention (chapter 5). In summary, this thesis has generated a robust dataset that improves our understanding of seepage wetland characteristics and their influences on NO₃⁻ removal at spatial and temporal scales. From an application perspective, this research provides new knowledge as to ‘where’, ‘when’ and ‘how’ seepage wetlands can be targeted to enhance their role in NO₃⁻ mitigations in hill country landscapes. This information has the potential to accelerate the integration of seepage wetlands into the toolbox of NO₃⁻ management strategies that could be used at a farm scale to improve water quality leaving NZ pastoral headwater catchments.
Figure 2-5 (= Liu et al., 2021 Fig 8) was removed for copyright reasons.
Agricultural pollution, Nitrates, Environmental aspects, Hill farming, Wetlands, New Zealand