Modelling Pasture Growth in the Mitchell Grasslands

Abstract: The Mitchell grasslands are some of the most productive grazing lands in Australia. However, the decision making environment for graziers is often challenging and risky within the naturally high rainfall and pasture growth variability experienced. Many graziers adjust their stock numbers based on the amount of standing feed and expected rainfall. Pasture models are useful if they explain the majority of variability in measured pasture growth, because long time series of measured growth are usually unavailable. The WinGrasp pasture model has been used to successfully simulate pasture growth in other plant communities (e.g. spear grass), however it is yet to be validated using independent data on Mitchell grasslands. If the model is useful in simulating Mitchell grasslands, forecasts of pasture growth can be used to help make better management decisions. Pasture growth forecasts based on El Niño Southern Oscillation (ENSO) have the potential to be more useful than rainfall forecasts for graziers in western Queensland. The accuracy of WinGrasp to simulate and forecast the growth of Mitchell grass needs testing. In this paper we compare the output from WinGrasp to measured total standing dry matter (TSDM) data from long-term grazing trials at Toorak Research Station, Julia Creek in the northern Mitchell grasslands and Rosebank Research Station, Longreach in the central Mitchell grasslands of Queensland. Using observed data from independent sites, the model explained 64 to 81 % of the variability of observe