摘要
【目的】研究模型和林分因子对区域尺度生物量碳储量转化和扩展系数(BCCEF)的影响,为区域尺度乔木林碳储量估算提供科学、合理的参数和估算方法。【方法】选择我国南方主要树种杉木,以福建、江西、湖南和广东4省固定样地为区组,选用区域内模型和区域外模型,按独立和兼容2种方式从单木到区域扩展,分一元和二元模型进行方差分析,确定稳定的模型;在此基础上,分起源、龄组,以林分平均胸径、平均高、胸高断面积和株数密度为定量因子,采用向后逐步回归和有交互作用的方差分析,确定林分因子与区域尺度地上部分和地下部分生物量碳储量转化和扩展系数之间的关系。【结果】扩展方式和选用模型对2种碳计量参数有显著影响,独立模型估计的参数大于兼容模型,选用模型会对参数造成8%~17%的误差,一元和二元模型对地上部分碳计量参数(BCCEFA)影响不显著,对地下部分碳计量参数(BCCEFR)影响显著;林分因子中起源对2种碳计量参数均有显著影响,天然林的碳计量参数大于人工林,龄组对BCCEFA有显著影响,从幼龄林到过熟林基本上呈下降趋势,而对BCCEFR影响不大,平均高、株数密度和胸高断面积对2种碳计量参数均有显著影响,对BCCEFR的影响大于对BCCEFA的影响,其中平均高、株数密度与2种碳计量参数呈负相关,胸高断面积与2种碳计量参数呈正相关,平均胸径对2种碳计量参数均影响不显著;使用区域内模型的湖南和广东间,除总BCCEFR外,2种总碳计量参数和分起源、龄组参数间差异均不显著,而同样使用区域内模型的福建和江西间,2种总碳计量参数和分起源、龄组的碳计量参数均差异显著。【结论】区域内二元兼容模型估计的2种总碳计量参数最稳定,林分起源、龄组、平均高、株数密度和胸高断面积对碳计量参数影响显著,即使使用同一区域内的模型,省际间林分因子的差异也可能造成碳计量参数的显著差异。
[ Objective] In order to provide scientific and reasonable parameters and estimation method for estimating the carbon storage of arbor forests at the regional scale, the effects of model and forest factors on biomass carbon storage transformation and expansion coefficient (BCCEF) were studied. [ Method] Taking Cunninghamia lanceolata as an example, thinking the permanent sample plots in Fujian, Jiangxi, Hunan and Guangdong provinces as four blocks, multi- ways analysis of variance were carried out to determine the most stable model. The factors included the selection of regional models or foreign models, the expansion ways from tree-level to regional scale with the independent model or the compatibility model with volume and the models with one or two variables. Then, six stand factors, which are two qualitative factor, stand origin and age group and four quantitative factors, mean DBH( diameter at breast height) , mean height, breast height basal area and stand density, were selected to conduct backward stepwise regression and analysis of variance with interaction. At last, the stand factors, which had significant influences on the parameters to carbon accounting at regional scale were screened out. [ Result] Expansion way and the selection of models had significant effects on two parameters to carbon accounting, the parameter estimated by independent model was greater than that estimated by the compatibility model with volume, and model selection could cause 8%- 17% of parameters error. The number of model variables had a significant influence on biomass carbon conversion and expansion factors for root part( BCCEFR ) , but had no significant effect on biomass carbon conversion and expansion factors for aboveground part(BCCEFA ). Stand origin had a significant effect on two parameters to carbon accounting, and the parameter for natural forest was greater than that for plantation. Age group had a significant influence on BCCEFA and minor effect on BCCEFR, the parameters for BCCEFA basically showed a declining trend from young forest to over-mature forest. The mean height, stand density and breast height basal area had significant effects on two parameters to carbon accounting with more influences on BCCEFR than BCCEFA. Among these factors, the mean height, stand density and the two parameters showed negative correlations, and breast height basal area and the two parameters were positively correlated. The mean DBH had minor effect on the parameters. Except for total BCCEFR, the differences of the two parameters grouped by stand origin and age group and total BCCEFA between Hunan and Guangdong provinces, in which a same regional biomass model was used, were no significant. Similarly, the differences of two total parameters and the two parameters grouped by stand origin and age group between Fujian and Jiangxi provinces were significant. [ Conclusion] The two total parameters to carbon accounting, estimated by the model compatible with volume and the model with two variables and regional model, are most stable. Stand origin, age group, the mean height, stand density and breast height basal area have significant effects on the parameters to carbon accounting. Even if using a same regional model, the differences of stand factors between provinces may lead to significant differences in the parameters to carbon accounting.
出处
《林业科学》
EI
CAS
CSCD
北大核心
2017年第9期55-62,共8页
Scientia Silvae Sinicae
基金
国家自然科学基金项目(31370634)
中国清洁发展机制基金赠款项目(2013014)
关键词
碳计量参数
扩展形式
模型选择
林分因子
杉木
parameter to carbon accounting
expansion way
selection of model
stand factors
Cunninghamialanceolata