A Wheat Grazing Model for Simulating Grain and Beef Production: Part II-Model Validation

Computer models must be thoroughly evaluated before being used for decision-making. The objective of this paper is to evaluate the ability of a newly developed wheat grazing model to predict fall-winter forage and winter wheat (Triticum aestivum L.) grain yield as well as daily weight gains of steer (Bos taurus) grazing on wheat pasture in Oklahoma. Experimental data of three independent field studies were used. The first was a variety trial in which fall-winter forage and grain yields were harvested. The second was a planting date experiment in which forage in the fall-winter period and grain yields were harvested. The third was a steer grazing experiment in which standing wheat biomass and steer weight gain were monitored. For the variety trials, the model efficiency (ME), which reflects how well model predictions match measured data (1 means a perfect match), was 0.102 for fall-winter forage prediction and 0.367 for grain yield. For the planting date experiment, the ME was 0.615 for predicting fall-winter forage yields and 0.409 for grain yields when a root downward extension rate of 20 mm d -1 was used. In the steer grazing experiment, the relationship between average daily weight gain and forage allowance was adequately represented by the model. For the total steer weight gains in a wide range of stocking rates and grazing durations, the ME was 0.616. Overall results show that the model, if well calibrated, has the potential to predict fall-winter forage and grain yields as well as mean daily weight gain per steer.