福建含笑苗木生物量分配特征及其最优预估模型研究

Characteristics of Biomass Allocation and Its Optimal Prediction Model of Michelia fujianensis Seedlings

为了探究福建含笑Michelia fujianensis苗木生物量分配特征以及掌握更快捷、准确地估测福建含笑苗木生物量的方法,研究以1 a生福建含笑苗木为试验材料,采取整株收获法,测量苗木根、茎、叶以及单株的生物量,并以地径(D)、苗高(H)、地径与苗高的乘积(DH)、地径平方与苗高的乘积(D^(2)H)4个生长指标作为自变量,对生长指标与苗木生物量分配进行pearson相关性分析并通过回归方程拟合福建含笑苗木各器官及单株生物量的预估模型,选取最优模型并进行精度检验。结果表明:福建含笑苗木生物量分配特征为茎>根>叶,非光合器官生物量与生长之间相关性更强,体现了福建含笑苗期生存策略为集中投资茎生物量和根生物量,这除了能提高苗木自身传输效率,还有利于向垂直方向以及地下四周拓展营养空间,争夺土壤资源。D2H对于根、茎以及单株生物量拟合效果最好,而苗高对于叶生物量模型拟合效果最好,并且各器官最优拟合模型多为幂函数,其具有较大的决定系数和较小的估计值标准误差,因此在实际应用过程中,面对不同变量应该合理选择与之最适宜的预测指标进行模型的拟合从而提高预估精度。

To characterize the biomass distribution of Michelia fujianensis seedlings,as well as to acquire a quicker and more accurate method of estimating the biomass of M.fujianensis seedlings.In this study,we used 1-year-old M.fujianensis seedlings as experimental materials,and adopted the full stem harvesting method to measure the biomass of roots,stems,leaves,and single plants of the seedlings.Then,four growth indicators,namely,ground diameter(D),height(H),the product of the diameter of the ground and height of the seedling(DH),and the product of the diameter of the ground squared and height of the seedling(D^(2)H)were analyzed by the pearson correlation analysis with the allocation of biomass to the seedlings.The regression equations were also used to fit the predictive models for the biomass of each organ and individual plant of M.fujianensis seedlings,and the optimal models were selected and tested for accuracy.The results showed that the biomass distribution of M.fujianensis seedlings was characterized by stem>root>leaf,and the correlation between the biomass of non-photosynthetic organs and the growth was stronger,reflecting the survival strategy of M.fujianensis seedlings as a concentrated investment in stem biomass and root biomass,which not only improves the efficiency of seedling′s transmission,but also facilitates the expansion of the nutrient space vertically as well as around the underground,and competes for soil resources.D2H was the best fit for roots,stems,and leaves biomass,while seedling height H was the best fit for the leaf biomass model,and the optimal fitting model for each organ was mostly a power function with a large coefficient of determination and a small standard error of estimate.Therefore,in the process of practical application,in the face of different variables,we should reasonably choose the most suitable predictor to fit the model to improve the accuracy of prediction.