Past cropping research in the semi-arid tropics of northern Australia has shown that in this climate and on the predominantly sesquioxidic soils, recovery of fertiliser nitrogen (N) by crops is often low. Conceptually, no-tillage, legume ley farming offers features for coping better with the constraints of climate, soil and high fertiliser transport costs to this remote region. This paper summarises the N cycle in a system in which pastures provide N for successive crops, and mulch at the time of crop establishment is provided by the killing of new pasture growth. The aim was further to provide a sound foundation for managing N supply in relation to demand in a climate that causes high variation and uncertainty for pasture N2 fixation and sequestering, the amount of early season re-growth (mulch), rate of mulch decomposition, nitrate leaching losses, and crop growth and N demand. The research approach combined field studies with simulation modelling. A series of field studies that included bare fallow and grass and legume pasture leys on clay loam and sandy loam soils, were conducted at Katherine over 4 wet seasons to study subsequent mineralisation of N. Experimental results were used to test the performance of a simulation model for predicting the observed variations consequent upon the various management options. Experimental results showed that the carbon (C) : N ratio of the residue and soil texture were important factors in determining N mineralisation, immobifisation, and nitrate leaching following chemical kill of pasture leys. However, the greatest variation was between seasons. A modified version of the CERES-Maize N model was able to simulate the accumulation of nitrate following a bare fallow and following pasture leys with high levels (above and below ground) of freshly killed residues with widely differing C:N ratio, the downward movement of nitrate-N in the soil and the interaction of these processes with seasonal rainfall. Despite a capability for simulation of the soil N dynamics in a cropping phase following pasture leys, ex~erimental results indicated how nitrate distribution following leys is influenced by pasture growth during the ley, and how this varied greatly with season and soil texture. The simulation capability reported here has been incorporated elsewhere into the development of a full system model, embracing both the ley phase and the crop phase.
Seasonal simulation of water, salinity and nitrate dynamics under drip irrigated mandarin (Citrus reticulata) and assessing management options for drainage and nitrate leaching