Nitrogen mineralisation in sugarcane soils in Queensland, Australia: II. From laboratory to field-based prediction

Using Australian sugarcane regions as a case study, we present an approach for prediction of in-field nitrogen (N) mineralisation over a crop season. The approach builds on the statistical modelling applied in Allen et al. 2019, which demonstrated good predictive ability on data from a laboratory incubation study (an external R of 0.84 in a cross-validation exercise), and adjusts those mineralisation rates according to soil moisture and temperature factors. The required field soil temperature and moisture conditions were simulated using a mechanistic model for the response of soil conditions to input climate data. We investigate drivers of variability in the predicted in-season mineralised N, and compare predictions with currently implemented N fertiliser discounts, which are based on a relationship with soil organic carbon content. The main purpose of this paper is to illustrate the potential use of the results in Allen et al. (2019) for calculating predictions of in-season mineralised N that could be applicable under field conditions in the Australian sugarcane regions. A thorough test to properly validate predictions has not yet been conducted, but collecting data to do so should be the focus of further work