Simulating fababean development, growth, and yield in Australia

The capability to simulate fababean (Vicia faba L.) production across the range of environments in which it is grown in Australia provides a tool for assessing agronomic and management options for the crop. This paper describes the building and testing of a model of fababean (cv. Fiord) development and growth, designed for use in the cropping systems simulator, APSIM. Parameters describing leaf area expansion, radiation interception, biomass accumulation and partitioning, root growth, water use, and nitrogen accumulation were sourced from the literature or developed from experiments conducted by the authors. In addition, parameters defining phenological development in response to temperature and photoperiod were derived from a comprehensive dataset of times to flowering and maturity. Routines for pre-flowering phenology predicted time to flowering ranging from 43 to 94 days with a root mean squared deviation (RMSD) of 4.3 days. Post-flowering development could not be satisfactorily predicted using thermal time alone; analysis of the collated data suggested that the lengths of some post-flowering phases were related to photoperiod. With incorporation of a photoperiod effect on post-flowering development, time from sowing to maturity was simulated with an R(2) of 92% and an RMSD of 6.7 days. The model was tested over a diverse range of latitudinal and climatic conditions within Australia, using data from experiments in which sowing date, crop density, and water supply varied. Observed grain yield (n=42) varied from 500 to 5600 kg/ha. In general, observed biomass and yield patterns within a growing season were reproduced well. Simulated grain yield explained 87% of the variance in observed yields (RMSD=466 kg/ha). Apart from demonstrating the capability of the model over a wide range of growing conditions, the tests highlighted a number of areas for future improvement, including accounting for lodging, variation in harvest index under wet conditions, and accurate simulation of the response of leaf area expansion to mild levels of water deficit.