A Sunflower Simulation Model: II. Simulating Production Risks in a Variable Sub-Tropical Environment

In highly variable environments, farmers require quantitative information on production risk to make well informed farming decisions. In many cases this can only be achieved through simulation studies. Thus, the dynamic sunflower (Helianthus annuus L) model QSUN was used, in conjunction with long-term climate records, to quantify the impact of climatic variability on production. Simulation results will assist farmers in making important management decisions such as if or when to plant, or which maturity type to choose. Simulations were conducted for three maturity types for planting times throughout the year and for two locations in Queensland, Australia (Dalby and Emerald). For both locations, more than 100 yr of daily temperature and rainfall records were available. Two hypothetical soil profiles were chosen to assess the effect on simulated yield likelihood of (i) maximum plant available soil water holding capacity of the soil (PAWC) and (ii) soil moisture content at sowing. The analysis showed that yields not restricted by water availability were higher at Dalby (370 g m⁻²) than at Emerald (349 g m⁻²). At Dalby, simulated median yields for soil profiles either full or half-full (180 mm PAWC) at sowing were around 100 and 60 g m⁻², respectively, with very little seasonal variation. At Emerald, median yields (70–90 and 40–60 g m⁻² for the full and half-full profile, respectively) were generally lower and seasonal variations were apparent: highest simulated median yields were achieved for mid-summer sowings. Yield variability was assessed by comparing median yields to yields at the 75 and 25% probability level. For both locations, this showed that even at the 25% probability level, yields of more than 100 g m⁻² can only be achieved if the soil profile was fully charged at planting. The comparison among maturity types showed no clear advantage of maturity type. These results provide farmers with an objective assessment of production risk in environments, where even one lifetime of experience can be insufficient to sample the climatic variability adequately. Contribution from the Agric. Production Systems Res. Unit, Queensland Dep. of Primary Industries and Commonwealth Scientific and Industrial Research Organisation, PO Box 102, Toowoomba, Qld, 4350, Australia.