The area of canola is expanding in the wheat-based farming systems of the summer-dominant rainfall zone of the Australian wheatbelt. Despite this, there is little information on yield and oil content expectations in relation to rainfall, location, and soil type and the reliability of the crop in a region characterised by high climatic variability. In this paper we assess variation in profitability of canola production in locations with different rainfall in the north-eastern wheatbelt by using long-term simulations at 4 locations spanning the climatic range of the region (Gunnedah, Moree, Walgett, Roma) with a validated model (APSIM-Canola). A new semi-mechanistic method for simulating oil content, accounting for temperature and water deficit effects during grain filling, is described and tested. Key agronomic determinants of reliable grain yield and oil content are identified.

Long-term simulations showed strong effects of location, plant-available soil water at sowing (PAW), and in-crop rainfall on grain yield expectations. Yield was negatively related to sowing date, particularly in those situations of high water supply (PAW and in-crop rainfall). Grain yield was positively related to in-crop rainfall up to 300 mm, with water use efficiency in most seasons falling between 6 and 12 kg/ha.mm. Variation in oil content was most strongly affected by sowing date, followed by location, with PAW having a minor effect. Importantly, the price bonus cut-off for oil content of 42% was exceeded in 25, 40, 40, and 55% of seasons for Roma, Walgett, Moree, and Gunnedah, respectively. Negative and falling phases of the SOI in April–May were associated with lower grain yield and oil contents, whereas positive and rising phases with higher grain yield and oil content. This suggests that the choice to sow canola over other alternatives could be a tactical decision that depends upon the seasonal climate outlook.

The approach used in this paper can be applied to the analysis of canola production risk (yield and oil content) and profitability in other prospective environments.