Improving the prediction of potato productivity: APSIM-Potato model parameterization and evaluation in Tasmania, Australia

Crop growth models are required to be extensively evaluated against actual data from field grown plants in order to have confidence in their prediction of crop productivity under various management options or a future changed climate. We evaluated the ability of the APSIM-potato model to predict production, phenology, and Nuptake of potato (Solanum tuberosum L.) under Tasmanian conditions. On-farm monitoring plots were established in north-west Tasmania within four different well-managed potato fields grown during the 2012/13 cropping season. Detailed soil and crop data sets measured in the on-farm plots planted with two potato cultivars, 'Russet Burbank' and 'Moonlight' were used to parameterise and evaluate the model. The model ealistically reproduced the observed tuber yield with high precision (a mean N-RMSE of 15.4% and modelling efficiency of 1.0 for both cultivars). Measured mean tuber yield was 17 t ha^-1 for 'Russet Burbank' with a simulated yield of 20 t ha^-1. For 'Moonlight' simulated tuber yield was 16.0 t ha^-1 compared to measured yield of 15.1 t ha^-1. The simulation results provide insight on the model performance under Tasmanian conditions. The results suggest that the model has potential to be used for purposes such as simulating productivity under various management options and climate change impact studies. Additional experiments are however required to improve cultivar specific input parameters such as phenology, leaf area and leaf duration and other functions that needs further refinement to improve model ability to simulate plant organs beside the tuber