岷江干旱河谷西康扁桃地上生物量估测模型  

Aboveground biomass prediction models for Amygdalus tangutica(Batal.)Korsh.in the arid valley along the Minjiang River

作　　者：韦丹丹 包维楷[1] WEI Dandan;BAO Weikai(CAS Key Laboratory of Mountain Ecological Restoration and Bioresource Utilization&Ecological Restoration and Biodiversity Conservation Key Laboratory of Sichuan Province,Chengdu Institute of Biology,Chinese Academy of Sciences,Chengdu 610041,China;University of Chinese Academy of Sciences,Beijing 100049,China)

机构地区：[1]中国科学院成都生物研究所,中国科学院山地生态恢复与生物资源利用重点实验室,生态恢复与多样性保育四川省重点实验室,成都610041 [2]中国科学院大学,北京100049

出　　处：《应用与环境生物学报》2021年第2期315-324,共10页Chinese Journal of Applied and Environmental Biology

基　　金：国家重点研发计划项目(2017YFC0505106)资助。

摘　　要：西康扁桃[Amygdalus tangutica(Batal.)Korsh.]是西南干旱河谷区具有较高生态和经济价值的特色乡土灌木树种,为准确估测其地上生物量,采用全株收割法,获取了80丛西康扁桃灌丛及其588个萌生株的外部形态特征与地上生物量数据,分别建立灌木丛与萌生株地上生物量预测模型.选取8种[地径(D)、高度(H)、植冠面积(AC)等单变量因子以及植冠体积(VC)、DH、D2H、DH2、D2H2等复合变量因子]预测变量,5种数学方程(一元线性方程、二次多项式方程、三次多项式方程、幂函数方程、指数函数方程)构建模型,并采用决定系数(R2)、平均误差(ME)、平均绝对误差(MAE)、总相对误差(TRE)、平均系统误差(MSE)和平均绝对百分比误差(MPSE)等6个指标对模型进行拟合优度检验,以获取地上生物量最佳估测模型.结果显示,复合变量因子与地上生物量的相关性更强,其中以D、H构成的复合变量因子与地上生物量的相关性最高;西康扁桃灌丛与萌生株最优地上生物量方程的预测变量与方程形式不同,其中以DH、D2H2为变量参数的2种估测模型m1=2.022–13.627(DH)+80.626(DH)2–64.459(DH)3(R2=0.735,P<0.01)和m1=15.223(D2H2)0.583(R2=0.707,P<0.01)均能较好地反映西康扁桃灌木丛地上生物量的累计特征;以D2H2为变量参数的幂函数模型m2=0.083+283.693(D2H2)(R2=0.774,P<0.01)能较好地反映西康扁桃萌生株地上生物量的累计特征.本文构建的西康扁桃灌木丛与萌生株地上生物量模型精度较高、可靠易行,可为进一步研究旱生灌木碳储量及灌木在干旱河谷植物群落中的生态作用提供科学依据.(图3表8参53)Amygdalus tangutica(Batal.)Korsh.is a native shrub species with high ecological and economic value in the arid valley area of southwest China.To accurately estimate the aboveground biomass of A.tangutica,the external morphological parameters and aboveground biomass data of 80 shrubs and 588 individual ramets were obtained using the whole plant harvesting method to establish aboveground biomass estimation models.Single variable factors such as diameter(D),height(H),canopy area(AC),and compound variable factors such as canopy volume(VC),DH,D2H,DH2,and D2H2 were selected as predictors;univariate linear,quadratic polynomial,cubic polynomial,power function,and exponential function equations were selected to construct the model.The optimal model was that with the largest determination coefficient(R2)and the smallest mean error,mean absolute error,total relative error,mean sybranchatic error,and mean absolute percentage error.The results showed that(1)the correlation between compound variable factors and aboveground biomass was significantly higher than that of single variable factors,and the composite variables composed of D and H had the highest correlation with above-ground biomass;(2)the optimal above-ground biomass prediction model for A.tangutica shrubs and ramets has different variable factors and equation forms;and(3)m1=2.022–13.627(DH)+80.626(DH)2–64.459(DH)3(R2=0.735,P<0.01)and m1=15.223(D2H2)0.583(R2=0.707,P<0.01)could better reflect the characteristics of aboveground biomass accumulation in a single A.tangutica shrub,whereas m2=0.083+283.693(D2 H2)(R2=0.774,P<0.01)was the best estimation model for a single A.tangutica ramet.These biomass models have high precision and are easy to implement,providing a scientific basis for further research on the carbon storage of xerophytic shrubs and the ecological role of shrubs in the plant community in the arid valley along the Minjiang River.

关 键 词：西康扁桃 地上生物量 回归模型 岷江干旱河谷 

分 类 号：S662.9[农业科学—果树学]