A SUNFLOWER SIMULATION-MODEL .1. MODEL DEVELOPMENT

In dryland farming systems, opportunities to improve sunflower (Helian thus annuus L.) yields are mostly associated with management decisions made at planting. Dynamic crop simulation models can assist in making such decisions. This study reports the structure of QSUN, a simple an d mechanistic crop model for sunflower, and how it accounts for the dy namic interaction of the crop with the soil and aerial environment. Th e model incorporates several recent approaches to simulation of crop g rowth in dryland conditions. QSUN estimates growth, development, and y ield of a sunflower crop. Daily inputs of temperature and photoperiod drive a phenology submodel to predict stages of emergence, bud visibil ity, 50% anthesis, and maturity. Using these stages, the growth submod el, driven by daily inputs of radiation, rainfall, and temperature, es timates leaf area production and senescence and soil water extraction. Biomass production is calculated from the amount of radiation interce pted by leaves or from the amount of water accessible in the root zone , depending on whether radiation or water is limiting crop growth. See d yield is calculated from the allocation of biomass to the grain foll owing anthesis. Sensitivity testing of the model under several irrigat ion regimes indicated that QSUN was most sensitive to the rate at whic h partitioning of biomass to grain increased, the ratio of biomass pro duced to water transpired, and the rate of soil water extraction in a water limited situation. The model was tested against independent data , with actual phenological data and was able to satisfactorily predict leaf area index (r2 = 0.65), total biomass (r2 = 0.96), and grain yie ld (r2 = 0.93), thus providing a tool for use in simulations studies a nd to assist in management decision making.