Reduction of nitrogen losses by manipulating carbon inputs and pasture composition : A thesis submitted in partial fulfilment of the requirements for the Degree of Doctor of Philosophy at Lincoln University

The nitrogen (N) cycle is very important to New Zealand agriculture, with N being a key nutrient needed for plant growth. However, there are major environmental concerns relating to the loss of N from agricultural systems. These include nitrate (NO₃⁻) leaching and the emission of nitrous oxide (N₂O). Nitrate leaching is of concern due to its potential human health risks and eutrophication of fresh water ways; while N₂O is a potent greenhouse gas and an ozone depleting substance. There is also a significant economic loss to farmers, as lost N needs to be replaced through fertilisers. Pastoral grazing systems have a high risk of NO₃⁻ leaching and N₂O emissions due to the high N loading rates of urine patches (approximately 613 kg N ha⁻¹ for cattle). Carbon (C) has many important interactions with the N cycle, these include immobilisation/mineralisation, nitrification, and denitrification. These interactions are of importance due to their potential to be manipulated in a way that reduces N losses from agricultural systems. Carbon inputs can be manipulated through growing different pasture/crop types with contrasting traits (rooting depth, carbon allocation, root exudates, winter growth), adding urine from animals fed on different diets and applying artificial C inputs. There is however, a lack of detailed knowledge on how C affects the N cycle in New Zealand grazed pastoral systems, on shallow stony soils. The aim of this research was to improve our understanding of these interactions and help develop techniques where C inputs can be manipulated to reduce N losses. To do this three lysimeter trials were conducted at Lincoln University’s Ashley Dene Research Development Station (ADRDS). These lysimeters contained Balmoral stony silt loam and were treated with cattle urine. Lysimeter experiment 1 was split into two trials; the objective of Trial 1 was to determine the effect of artificial inputs of readily available C on the N cycle; the objective of Trial 2 was to determine the effect of different urine compositions and crop types on the N cycle. Trial 1 discovered that applying readily available C to soil significantly reduced N leaching losses, without causing an increase in N₂O emissions. This was attributed to the added C increasing immobilisation of N in the soil. Trial 1 also revealed that the perennial ryegrass (Lolium perenne)/white clover (Trifolium repens) (PRG/WC) pasture leached 58% less N than the lucerne (Medicago sativa) crop. This reduction was attributed to the higher winter plant growth and N uptake of the PRG/WC pasture, reducing the amount of mineral N available to be leached. Trial 2 discovered that under PRG/WC pasture, urine from cows feed a diet of fodder beet (FB) (Beta vulgaris) leached 64% less NO₃⁻-N than urine from cows feed a diet of PRG/WC; even at the same urine N loading rate. Soil under the FB urine had significantly lower ammonia oxidising bacteria (AOB) amoA gene abundance (P = 0.005) and amount of soil NO₃⁻-N (P = 0.026), suggesting that the FB urine has a biological nitrification inhibitor (BNI) effect. Trial 2 also discovered that the PRG/WC pasture leached 65-84% less N than the bare fallow FB soil. This reduction was attributed to the higher winter plant growth and N uptake of the PRG/WC pasture, reducing the amount of mineral N available to be leached. Lysimeter experiment 2 discovered that by lowering the urine-N loading rate by 28%, from 700 kg N ha⁻¹ to 500 kg N ha⁻¹, N2O emissions were reduced by 38%, and total N leaching losses were reduced by 39%. This demonstrates that farm management practices that reduce urine-N rate could be an effective way of reducing N losses from grazing systems. Lysimeter experiment 3 discovered that Italian ryegrass (Lolium multiforum)/white clover/plantain (Plantago lanceolate) (IRG/WC/P) and perennial ryegrass/white clover/plantain (PRG/WC/P) significantly reduced N leaching losses by 24% and 14%, respectively, compared with traditional PRG/WC. The reasons for these reductions were: (i) the higher plant N uptake, which decreased the soil mineral N content and subsequently reduced the amount of N available to be leached and; (ii) the greater herbage yield which increased transpiration, thus reducing drainage volume. This research programme has highlighted the importance of manipulating C inputs through winter plant growth and the manipulation of cattle diet in reducing N losses. Increasing plant N uptake over cooler months is potentially an economically viable and effective way of reducing farm N losses. This can be achieved by sowing more winter active crops/pastures, reducing time cattle spend on bare fallow soil and/or through the use of effective catch crops. Manipulating cattle diet by feeding stock low N feed and subsequently reducing the urine N loading rate is potentially an economically viable and effective way of reducing farm N leaching losses; especially if that low N feed such as fodder beet, can provide a BNI effect in the cattle urine.