Crop model based decision support for maize ( Zea mays L.) silage production in Portugal

Abstract

Maize ( Zea mays L.) silage is of major importance for milk production in the Northwest of Portugal. Farmers typically have a variety of maize hybrids to choose from according to cycle length and sowing date. The general recommendation regarding cultivar selection is to use long cycle cultivars for early sowing dates and vice versa. Cycle length, sowing date and temperature regime will determine the harvest date. Because weather regime is unknown at sowing date, there is a need to develop decision support based on historical weather series to help farmers optimize silage production. Production optimization occurs through a better matching of cycle length to sowing date to produce more and better silage at optimal harvest dates. The CERES-Maize crop model was used to establish decision support to help farmers identify the best cultivar and sowing date combinations. Cultivar parameters were estimated from 3-year field experiments involving five planting dates and six cycle lengths (FAO 200 to 700). The model was run with 39 years of historical weather data, simulating 18 sowing dates and 6 cycle lengths. Decision support was developed based on the analysis of simulation outputs and three integrated risk management strategies. Tactical use of guidelines is illustrated with examples. Current limitations of the model for maize silage simulation are also discussed.