Can the SPUR rangeland simulation model enhance understanding of field experiments?

This article assesses whether the rangeland ecosystem simulation model SPUR2.4 can predict the response of key parameters measured in a field experiment and enhance our understanding of the experiment by conducting elements of that experiment that could not be conducted in the field due to logistic or other constraints. Questions posed are: (1) does Simulation of Production and Utilization of Rangeland (SPUR)2.4 adequately predict the treatment differences measured in a rangeland grazing experiment, and (2) if the soils, slopes and vegetation had been identical between treatment areas in the grazing experiment, would the results of the experiment be different? With both forage and animal production, the differences in the simulation experiment due to soil and slope composition in each pasture were very close in order of magnitude to the differences measured in the field and ascribed to treatment effects.. In the simulations, differences due to treatment were confined to stocking rate effects. SPUR2.4 simulates a number of parameters well enough to provide credible evidence that differences previously ascribed to grazing treatment were probably due to the differences in the soils and slopes present in each treatment area rather than to the treatments imposed. By not taking the significant soil differences between treatment pastures into account, it is probable that incorrect conclusions were drawn from this field experiment. This exercise illustrates that simulation models probably have a very useful role to perform in enhancing our understanding of field experiments provided they have been validated and demonstrate adequate accuracy.