Nitrate leaching under arable land : monitoring, mitigation measures & memory effects

The loss of nitrogen (N) from agroecosystems is one of the biggest unsolved environmental problems of our time: on the one hand, nitrogen is a critical yield-limiting factor and therefore, N supply in agriculture in the form of fertiliser is crucial to ensure food security. On the other hand, part of this fertiliser is lost to other environmental compartments, e.g. by leaching through the soil to the aquifer in the form of mobile nitrate. Due to this contamination, drinking water limits for nitrate are nowadays exceeded in many regions. In this study, we investigated the N transport and cycling processes during three cropping seasons (2017-21) under 11 arable fields on silty loams in the Gäu Valley on the Swiss Central Plateau. The crop rotations included silage maize after ploughing the grass-clover ley, winter cereals and canola. The goal was to compare monitoring techniques for nitrate leaching, to determine the main influencing factors for nitrate leaching from the root zone, and to increase understanding of nitrate transport dynamics across the vadose zone. Therefore, we used ion-exchange resin-based Self-Integrating Accumulators (SIA), soil coring for extraction of mineral N (Nmin), Suction Cups (SCs) complemented by a HYDRUS 1D model, and a Vadose Zone Monitoring System (VMS) to assess nitrate leaching in the soil and in the unsaturated zone down to 6 m depth. We also tested if a reduction of the N fertiliser level or a change of fertiliser type reduces nitrate leaching, and calculated surface N balances, including atmospheric deposition, fertilisation, biological N fixation, and N output via yield. All four monitoring techniques were suited to measure N leaching, but represented different N transport and cycling processes, and varied in spatio-temporal resolution. The average annual leaching measured with SIA devices was moderate for grass-clover leys, canola and maize (38, 42, and 44 kg N ha-1 a-1), and high for cereals (116 kg N ha-1 a-1). Averaged over all crop rotations (71 kg N ha-1 a-1, without strips with mitigation measures), this is triple the amount that is compatible with the national legal target concentration in groundwater (25 mg NO3 - L-1). This quality target was also surpassed in 55 % of SC samples. The N balance cumulated over three seasons depended highly on the share of grass-clover ley in the crop rotation that compensated for the negative balances of the the other crops (-50, -29, and -45 kg Ntot ha-1 a-1 for canola, cereals and maize versus +126 kg Ntot ha-1 a-1 for grass-clover ley). Thus, the main drivers for nitrate leaching were the long-term inputs of organic fertilisers with a high share of unavailable N especially to grass-clover leys, the related accumulation in the soil organic N pool, a high mineralisation potential, and finally the N release that is unsynchronised with the plants’ needs. Monthly SC measurements showed that nitrate is mainly leached from the root zone during autumn and winter, caused by elevated soil pore water mobilisation in this period. However, this seasonal leaching pattern was not transferred to deeper soil layers in the unsaturated zone, where further N transport was heavily affected by denitrification in clay-rich zones, and by preferential flow including bypassing in desiccation cracks and fractures in the consolidated layers. Due to the high mineralisation supply from the soil organic N pool, estimated to be 59 kg N ha-1, the fertiliser reduction was only partially visible in the nitrate leaching data, and a change of fertiliser type had no significant effect on nitrate leaching during the research period. To conclude, the biogeochemical legacy of N dynamics estimated to be in the range of decades, increases the time lag between an intervention on the surface and a visible effect in the aquifer far beyond the one that is explicable with the hydrologic legacy alone, that is estimated to be of several years. We recommend long-term monitoring to further elucidate the impact of these memory effects and establish realistic groundwater quality goals. The autumn Nmin content in a large number of fields can be used as an indicator for regional nitrate losses to the aquifer, as it was shown to significantly correlate with the subsequent winter leaching. The N fertiliser recommendation should be regionally adapted and account for the long history of manure application with high amounts of unavailable N, and the high mineralisation potential of the local soils.