The objective of this work was to evaluate the profitability and risk of grain production systems combined with annual winter pastures, under no-tillage system. The experiment was carried out in Passo Fundo, RS, Brazil, in a six-year period (1990-1995). The systems were: system I (wheat/soybean, grazed black oats/soybean, and grazed black oats/soybean); system II (wheat/soybean and grazed black oats + grazed common vetch/corn); system III (wheat/soybean, grazed black oats + grazed common vetch/soybean, and grazed black oats + grazed common vetch/corn); and system IV (wheat/soybean, white oats/soybean, and white oats/soybean). An experimental design in blocks at random, with three replications and plots with 500 m², was used. Three types of analysis were applied to the net return of systems: mean-variance analysis, distribution of accumulated probability, and stochastic dominance. When mean-variance analysis was used no significant differences were observed in the systems studied. By the distribution of accumulated probability, the selection of an alternative depends solely and exclusively on the level of risk chosen by the decision-maker. By the stochastic dominance method, system II showed to be the best production alternative to be offered to farmers from both profit and lower risk standpoints.
production factors; costs; profitability; mathematical models; statistical analysis
