A polyfactoral experiment was established in the Research Centre of the University of Debrecen with randomised design and four replications to examine the effects of different tillage methods (shallow spring tillage, autumn ploughing and spring ploughing) of the various factors affecting maize production. Based on the results obtained during three consecutive dry and warm crop years (2012, 2013, 2014) of the 30-year-long long-term experiment, it was examined in irrigation treatments which tillage methods provide higher yield in the whole yield range. It is especially important to examine when the different tillage methods can provide the given yield level at the lowest risk. In order to carry out this examination, stochastic dominance and expected value variance risk programming criterion methods were used.

Based on the research findings obtained in the multiple-decade-long long-term field experiment, it was concluded that it is possible to set up an order of tillage methods over the whole yield range in accordance with the secondary stochastic dominance to effectively help decision-makers.

Based on the use of the expected value variance criterion, a given yield can be produced with the lowest risk – in the range of expected yields resulting from various tillage methods – by parametrically altering yield if the proportion of shallow spring tillage is the highest (above 8.7 t ha-1).

Keywords: risk assessment, tillage methods, irrigation