Modelling decomposition of sugar cane surface residues with APSIM±Residue Peter J. Thorburn a,* , Mervyn E. Probert a , Fiona A. Robertson b a CSIRO Sustainable Ecosystems and CRC for Sustainable Sugar Production, 120 Meiers Road, Indooroopilly, Qld 4068, Australia b CRC for Sustainable Sugar Production and BSES, PMB 57, Mackay MC, Qld 4740, Australia Accepted 14 February 2001 Abstract Since residues on the soil surface affect the soil hydrological and nitrogen cycles, cropping system models need to accurately predict the decomposition of surface residues through time. Recent ®eld measurements of sugar cane residue decomposition have found that rates of residue decomposition are slower than those predicted by the agricultural productions systems simulator APSIM) cropping system model. To improve predictions of sugar cane residue decomposition in APSIM, a new function relating decomposition rate to mass of residues was developed, and parameter values in other functions modi®ed in the APSIM±Residue model. Predictions were compared to measurements of residue decomposition made over 1 year in ®ve ®eld experiments located at sites with contrasting climates in the Australian sugar industry. The new function was found to provide more accurate predictions of decomposition of large >10 t DM ha 1 ) masses of residue that are common with sugar cane crops. To better predict residue masses below 5±7 t ha 1 , the values of parameters describing the potential decomposition rate i.e., decomposition without environmental or residue quality constraints) and constraints to decomposition from sub- optimal temperatures were determined by applying a numerical optimisation technique to measurements made at two of the ®eld experiments. These two experiments were chosen because they had a large difference in initial residue mass 7.7 and 19.6 t ha 1 ) and included data on soil moisture. These latter measurements, which were not made in the other experiments, provided rigorous data for the function describing the moisture limitation to residue decomposition. Using the optimised parameter values, predictions of residue decomposition in the other three experiments were close to measured values. This study illustrates the potential importance of considering initial residue mass in predictions of decomposition in sugar cane systems. The function relating decomposition rate to mass of residues may be useful for other models applied in cropping systems with high residue masses. The study also provided further evidence that some parameters in APSIM±Residue, notably the potential decomposition rate, may be crop-speci®c. Crown Copyright # 2001 Published by Elsevier Science B.V. All rights reserved. Keywords: Cropping system; Crop residue; Immobilisation; Nitrogen; Simulation; Soil water; Temperature 1. Introduction The amount of crop residues present on the soil surface, and their rate of decomposition signi®cantly affect the soil hydrological and nitrogen N) cycles. Models that aim to predict the dynamics of crop growth under variable residue management practices also need to accurately predict residue dynamics Thorburn et al., 2001). The agricultural productions systems simulator APSIM; McCown et al., 1996) is one such model. It describes the dynamics of crop growth for various crops), soil water, soil N and Field Crops Research 70 2001) 223±232 * Corresponding author. Tel.: 61-7-3214-2316; fax: 61-7-3214-2325. E-mail address: peter.thorburn@cse.csiro.au P.J. Thorburn). 0378-4290/01/$ ± see front matter Crown Copyright # 2001 Published by Elsevier Science B.V. All rights reserved. PII:S0378-429001)00141-1