中文摘要：

灌木生物量模型是目前预测灌木生物量的重要方法之一。以浑善达克沙地3种灌木作为研究对象,每种灌木各取70个样本,实测其生物量。依据株高（H）和基径（D）的复合因子D-2H作为自变量,以实测生物量（W）为因变量构建回归分析模型。通过分析判别系数R2和估计值标准误差SEE筛选出最佳的生物量估测模型。结果表明：土庄绣线菊灌木生物量最优模型为W=1.772（D-2H）0.757;虎榛子灌木生物量最优模型为W=12.004＋1.006（D-2H）-0.014（D-2H）2＋0.001（D-2H）3;黄柳灌木生物量最优模型为W=16.359-0.009（D-2H）＋0.001（D-2H）2-7.415E-07（D-2H）3。经验证模型的预测值与实测值拟合率在72.08-88.72%,其预测效果较好。

英文摘要：

Shrub biomass model is an important method for predicting shrub biomass. We took three shrub species in Hunshandake sandy land as the study objects, measured the biomass of 70 samples of each species. Taking D2H which is composed of plant height and basal diameter （D） as independent variables and measured biomass as the dependent variables （w） to build the regression model. At last, we selected the optimal biomass prediction model by analyzing discriminant coefficient R2 and standard errors of estimated value. The results showed that the optimal shrub biomass model of Spiraea pubescens is W = 1. 772 （D2H）0.757 the optimal shrub biomass model of O. davidiana is W = 12. 004 ＋ 1. 006 （ D2 H） - 0.014 （ D2 H ） 2 ＋ 0.001 （ D2 H ） 3, the optimal shrub bio- mass model of Salix gordejevii is W = 16.359 - 0.009 （ D2 H ） ＋ 0.001 （ D2 H ） 2 - 7.415 E - 07 （ D2 H ） 3. The forecast results are good after testing the models whose fitting rate of predicable value and measured values are between 72.08 - 88.72%.