DairyMod在四川盆地多花黑麦草生产中的验证与应用

DairyMod Validation and Application for Lolium multiflorum Production in Sichuan Plain

介绍了DairyMod模型的结构和功能,应用刈割模式对多花黑麦草(Lolium multiflorum)的生长速率、干草总产量进行模拟和验证。在输入牧草实际生育期和调整生理参数的基础上,采用当地气象观测站1991—2012年气象数据作为驱动模拟。通过收集四川省雅安7年和洪雅4年多花黑麦草的生产数据,运用多个统计指标对干草总产量的模拟进行检验。两地的生长速率模拟值与实测值基本一致,干物质总产量模拟值与实际值显著相关,决定系数R^2分别为0.97和0.91,D值分别为0.94和0.58,RMSE分别为340和1132 kg·hm^(-2);NRMSE分别为2.3%和7.5%。模拟平均值比实测平均值分别高2.0%和4.5%,产量的高估可能是由于模拟施氮量大于生产施氮量而导致。结果表明,多花黑麦草在洪雅的生长速率和总产量大于在雅安,校准后的DairyMod对四川多花黑麦草的季节生产和总产量有较好的模拟和预测。

DairyMod,biophysical pasture simulation model,is composed of numerous modules with a common underlying structure.The mowing patterns were introduced and used to simulate the growth rate of annual ryegrass（Lolium multiflorum） and evaluate total yield in the growth season.The default parameter sets of the model were collected including pasture species,soil water characteristics,soil organic matter and nutrient characteristics,and livestock.Local meteorological climate data from 1991 to 2012 were used to run the model.The trial data of annual ryegrass（Lolium multiflorum） were collected in Yaan（7years） and Hongya（4 years） experimental stations,used to validate the simulation results in Sichuan province including growth speed and total dry matter yield in growth season.The results showed that the simulating growth rate of the model was consistent with the measured values in both Yaan and Hongya station,modeled dry matter production matched well with the observed ones（R2=0.97,D=0.94,RMRE=340 kg·hm-2,NRMRE=2.3%at Yaan and R2 =0.91,D=0.58,RMRE= 1132 kg· hm-2,NRMRE=4.5%at Hongya,respectively）.Annual ryegrass had higher potential production in Hongya than in Yaan.Though there was a slight overestimation according to the summary of model climate data,it could be used as an effective tool of herbage yield prediction in forage and livestock system.Therefore,DairyMod can accurately simulate the annual ryegrass growth rate and yield.It will provide an important technique for forage quantitative evaluation.