Crop size and sugarcane nitrogen fertiliser requirements: Is there a link?

The Australian sugarcane industry is under pressure to reduce nitrogen (N) fertiliser applications and hence N losses to the environment. One pathway suggested to reduce N applications is to match yield targets in N fertiliser recommendations to the yields achieved by farmers. This seems a sensible strategy: smaller crops generally grown by farmers (relative to current yield targets) 'should' need less N. Is it really that simple? We collated over 150 N response curves for ratoon crops from past experiments to investigate the amount of N (N-opt) needed to achieve 95% of maximum sugarcane yield (Y-95). There was little correlation between Y-95 and Nopt. For example, low yields (e.g. <50 t/ha) occurred at both low (<50 kg/ha) and high (>200 kg/ha) N-opt values. The low correlation was also seen in individual experiments and thus the results were not an artefact of amalgamating data from different locations. At one experiment, for example, across five years the N requirement varied three-fold. Given that soil and management were consistent across the years, the variation showed the climatic influence on N requirement. The results also showed that there was variation among sites and crops in both yield potential and the amount of N required to grow a tonne of cane. Rather than trying to improve N recommendations by changing concepts around target yields, we suggest it would be more beneficial to develop ways to predict N-opt directly. We simulated N responses with the APSIM model for one of the better characterised experiments in the database and derived N-opt from the response curves. Simulated N-opt was generally within the range of N-opt at the experiment. We conclude that direct prediction of N-opt through application the APSIM model, in combination with seasonal climate forecasts could be the basis of a future decision support system to define optimum N rates.