A SUNFLOWER SIMULATION-MODEL .2. SIMULATING PRODUCTION RISKS IN A VARIABLE SUBTROPICAL ENVIRONMENT

In highly variable environments, farmers require quantitative informat ion on production risk to make well informed farming decisions. In man y cases this can only be achieved through simulation studies. Thus, th e dynamic sunflower (Helianthus annuus L) model QSUN was used, in conj unction with long-term climate records, to quantify the impact of clim atic 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 thre e maturity types for planting times throughout the year and for two lo cations in Queensland, Australia (Dalby and Emerald). For both locatio ns, more than 100 yr of daily temperature and rainfall records were av ailable. Two hypothetical soil profiles were chosen to assess the effe ct on simulated yield likelihood of (i) maximum plant available soil w ater holding capacity of the soil (PAWC) and (ii) soil moisture conten t at sowing. The analysis showed that yields not restricted by water a vailability 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, respe ctively, with very little seasonal variation. At Emerald, median yield s (70-90 and 40-60 g m-2 for the full and half-full profile, respectiv ely) were generally lower and seasonal variations were apparent: highe st 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 eve n 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 co mparison among maturity types showed no clear advantage of maturity ty pe. These results provide farmers with an objective assessment of prod uction risk in environments, where even one lifetime of experience can be insufficient to sample the climatic variability adequately.