Dynamic cropping strategies for risk management in dry-land farming systems

Low rainfall and soils with low water holding capacity are major features of the Mallee region of south-eastern Australia. To reduce risk related to likely water-deficits, many growers adopt a conservative cropping strategy on low-input cereal crops. This approach has, however, a substantial opportunity cost by missing the potential profit from more intense cereal cropping and high yielding, oilseed crops in more favourable seasons. We tested the hypothesis that whole-farm profitability could be enhanced by the adoption of a dynamic cropping strategy shifting from a cereal-only, conservative strategy in dry years, to a more risky strategy involving both cereals and canola in wet years. To test this hypothesis, we used 40-years rainfall series to (1) investigate rainfall features in 11 locations in the Mallee, (2) test the skill of simple rules to predict seasonal rainfall, as developed by local farmers, and (3) calculate whole-farm profit for conservative, risky and dynamic cropping strategies. Rainfall and profit were linked with a whole-farm model that estimates crop yield as a function of seasonal rainfall (i.e. rainfall from April to October) and water-use efficiency. Among locations, annual rainfall ranged from 259 to 358 mm. For each location, two types of seasons were defined: likely wet, when April rain was above the median, and likely dry otherwise. The strength of the association between April and seasonal rain varied widely among sites; it was stronger in locations with more marked rainfall seasonality. Contrasting whole-farm profit responses to cropping strategies were found in locations with annual rainfall below or above a threshold around 300 mm. For wetter locations (annual rain above the threshold), the more risky cropping strategy including canola was generally more profitable than the more conservative strategy. For farms in drier areas, the cereal-based conservative strategy outperformed the more risky strategy in seasons predicted to be dry, but was less profitable in wet seasons. The dynamic cropping strategy had a substantial effect on extreme years, alleviating economic losses associated with the risky strategy in dry seasons, while being able to capture the benefits of more favourable seasons. Analysis of rainfall patterns, development of a rainfall forecasting procedure, and quantification of whole-farm profit in response to cropping strategies, all highlighted the need for decision support tools that account for small-scale variation in rainfall characteristics.