| Popis: |
A simulation study was conducted to compare diverse grazing strategies for steers grazing open woodlands in northeast Queensland. Simulations included a wide range of possible stocking rates and pasture utilisation levels using 108 years (1889–1996) of daily climate data for Charters Towers. Five strategies were compared in terms of steer liveweight gain per ha, risk of weight loss, pasture availability, frequency of burning and soil loss. The strategies included constant stocking, stocking in response to available feed, and stocking in response to predicted future feed availability based on a climate forecast. For strategies achieving an average annual liveweight gain per head of about 100 kg, the simulation studies indicated that a responsive stocking rate strategy in June using a forecast of the next year’s pasture growth would increase liveweight gain per ha by about 10%, reduce the risk of liveweight loss by 57%, reduce risk of low pasture yield, but would slightly increase the risk of soil loss (4%). Maximum LWG/ha was achieved at high utilisation rates (> 35%). However, at such high levels of utilisation burning was achieved in less than 10% of years and soil loss was 30–40% more than at levels of utilisation regarded as safe (≈20%). The simulations highlighted the potential value of achieving in June, the skill from seasonal forecasting that is now available in November using average SOI in the Aug-Oct period as the indicator of season type. Assumptions in the model development are outlined and future work required is discussed. Despite the complexity of the simulation analysis, it is concluded that there is a trade-off between production and environmental damage, and that improved forecasting may improve production and/or reduce damage. |
