干旱区生态环境知识资源中心

We present a simple simulation model as an aid to conventional expensive field experimentation and as a decision support system tool for animal production in irrigated pastures within semi-arid grazinglands. We also demonstrate and discuss the integration of results from conventional field experiments within the context of simulation models. We first develop a spreadsheet to estimate animal production under different stocking rates and pasture characteristics, and then incorporate functions from the spreadsheet and from the literature into a dynamic simulation model. The spreadsheet, which is programmed in EXCEL, is a static, deterministic series of calculations that estimates weight gain of grazing animals under different stocking rates and pasture conditions based on decision rules and functions reported by NRC and MAFF. The simulation model, which is programmed in STELLA (R) Research 6.0 for use on a personal computer, is dynamic and stochastic.

Simulation results were close to observed values for northern Mexico with regard to annual yield of dry standing crop (21,434 vs. 23,637 kg ha(-1)), dry matter digestibility (67-78/(, vs. 7074%), and optimum stocking rate as live weight (892-2567 vs. 2250 kg ha(-1)). Simulation results indicated the highest optimum stocking rate was 2567 62 kg ha(-1) in July and the lowest was 892 106 kg; ha-1 in January. The highest daily gain per head was 1.144 +/- 0.006 kg in February and the lowest was 0.480 +/- 0.052 kg in July. Daily forage allowance for the optimum stocking rate varied during the year from 4.05 +/- 0.053 in July to 4.84 +/- 0.008% of live weight in January. Producers should maintain a herd size that allows maintenance of the optimum stocking rate during months of low growth and, during months of high growth, should graze only the area required to cover the recommended forage allowance, and cut the rest of the forage. (c) 2005 Elsevier Ltd. All rights reserved.