摘要
森林的碳汇功能一直是国内外的研究热点,而对单木不同器官含碳量的分布规律研究较少。本文使用云南省普洱市思茅区64棵思茅松天然林解析木数据,对不同器官的含碳率进行了测定。应用Moran’s I及局域统计量(局域均值及局域标准差)在4个尺度(20m、100m、300m、700m)分别检验思茅松树枝、树叶、树干、树皮、根系及全树6个维度含碳量的空间相关性和空间异质性。同时以不同器官含碳量(C)为因变量,以胸径(DBH)、树高(H)、冠长(CL)、海拔(Elev)、坡度(Slo)5个变量以及胸径平方乘以树高(D~2H)1个复合变量为自变量分别构建普通最小二乘模型(OLS)和空间滞后模型(SLM),并以Akaike’s信息指数(AIC)值及拟合优度(R~2)为指标评判2个模型在拟合过程中的优劣。结果表明:(1)计算的Moran’s I均为正值,且Z值都大于1.96,P值除在20m带宽时为0.01,其余全部小于0.000 1,表示拒绝Moran’s I零假设,即思茅松各器官含碳量在一定的距离均具有空间效应;(2)由局域统计量可以看出思茅松单木的含碳量分布并不均匀,局域标准差在带宽700m时最大为191.23,即在该尺度下空间差异最大;(3)林分因子、地形因子都在不同程度上影响着思茅松不同器官含碳量的空间分布,相比较传统的普通最小二乘模型来说,空间滞后模型可以很好地解决模型构建中的空间效应。
Relevant study on forest carbon sink function has become a hot topic.However,few researches studied spatial distribution of carbon content in different organs of individual tree.There were a total of 64 sampling trees at Simao pine(Pinus kesiya var.langbianensis)natural forest investigated in Simao district of Puer,Yunnan Province,and the carbon content of different organs was determined.Moran's I and local statistics(local mean and local standard deviation)were used to explore the spatial autocorrelations and spatial heterogeneity of carbon content in 6 dimensions,namely branches,leaves,bark,trunks,roots and whole trees,at 4 bandwidths(20 m,100 m,300 m and 700 m).Ordinary least squares model(OLS)and spatial lag model(SLM)were established by selecting carbon content of different organs(C)as dependent variable,while diameter of living trees(DBH),height of tree(H),crown length(CL),elevation(Elev),slope(Slo),and the product of DBH squared and H(D~2H)as independent variables.And Akaike's information criterions(AIC)and goodness of fit(R~2)were used as indicators to evaluate the two models in the fitting process.The results showed that:(1)The calculated Moran's I were all positive,and all Z values were higher than 1.96,all P values were less than 0.000 1 except for 0.01 at 20 m bandwidth,indicating that the null hypothesis of Moran's I was rejectedIt means that the carbon content of each organ of Simao pine have a spatial effect at a certain distance;(2)The local statistics showed that the carbon content distribution of Simao pine was not uniform.The maximum local standard deviation was 191.23 at 700 m bandwidth,the spatial difference was the largest in this scale;(3)The distribution of the carbon content of different organs of Simao pine were influenced by stand and environmental factors,and the SLM can solve the spatial effect of the model construction compared with the conventional OLS.
引文
[1]Zhang L,Lin W,Wang Z,et al.Spatial distribution pattern of carbon storage in forest vegetation of Guangdong province[J].Ecology & Environmental Sciences,2010,19(6):1295-1299.
[2]Chang L,Lianjun Z,Fengri L,et al.Spatial modeling of the carbon stock of forest trees in Heilongjiang Province,China[J].Journal of Forestry Research,2014,25(2):269-280.
[3]Punchi-Manage R,Wiegand T,Wiegand K,et al.Effect of spatial processes and topography on structuring species assemblages in a Sri Lankan dipterocarp forest[J].Ecology,2016,95(2):376-386.
[4]Zhang C,Zhang S,He J.Spatial distribution characteristics of heavy metals in the sediments of Changjiang river system-Spatial autocorrelation and fractal methods[J].Acta Geographica Sinica,1998,1:87-96.
[5]李哈滨,王政权.空间异质性定量研究理论与方法[J].应用生态学报,1998,9(6):651-657.
[6]Brunsdon C,Fotheringham S,Charlton M.Geographically weighted summary statistics- a framework for localized exploratory data analysis[J].Computers,Environmental and Urban Systems,2002a,26:501-524.
[7]Ma Z,Benjamin Z,William F,et al.Use of localized descriptive statistics for exploring the spatial pattern changes of birds species richness at multiple scales[J].Applied Geography,2012,32:182-194.
[8]刘畅,李凤日,贾炜玮,等.基于局域统计量的森林碳储量空间分布变化的多尺度研究[J].应用生态学报,2014,25(9):2493-2500.
[9]Julie L G,Catherine B,Sandy D,et al.On the property of diffusion in the spatial error model[J].Applied Economics Letters,2005,12(9):533-536.
[10]Kehrer L.The central performance drop in texture segmentation:a simulation based on a spatial filter model[J].Biological Cybernetics,1997,77(4):297-305.
[11]Jesús M,Ana A.The spatial durbin model and the common factor tests [J].Spatial Economic Analysis,2006,1(2):207-226.
[12]Lambert D,Brown J,Florax R.A two-step estimator for a spatial lag model of counts:Theory,small sample performance and an application[J].Regional Science & Urban Economics,2010,40(4):241-252.
[13]Ghidey W,Lesaffre E,Eilers P.Smooth random effects distribution in a linear mixed model [J].Biometrics,2015,60(4):945-953.
[14]Zhang L,Shi H.Local modeling of tree growth by geographically weighted regression[J].Forest Science,2004,50(2):225-244.
[15]李果,王勇,冉敏,等.2002-2012年丰都县森林碳储量的动态变化[J].西部林业科学,2017,46(2):62-67.
[16]刘畅,李凤日,甄贞.空间误差模型在黑龙江省区域森林碳储量空间分布的应用[J].应用生态学报,2014,25(10):2779-2786.
[17]李伟,张翠萍,李士美.基于第8次森林资源清查数据的广西森林碳储量特征研究[J].西南林业大学学报,2017,37(3):127-133.
[18]云南森林编写委员会.云南森林[M].昆明:云南科技出版社&中国林业出版社,1988.
[19]Koenig W.Spatial autocorrelation of ecological phenomena[J].Trends in Ecology and Evoultion,1999,14:22-26.
[20]Brunsdon C,Fotheringham S,Charlton M.Geographically weighted local statistics applied to binary data[J].Geographic Information Science Lecture Notes in Computer Science,2002b,2478:38-50.
[21]Shi H,Zhang L.Local analysis of tree competition and growth[J].Forest Science,2003,49:938-955.
[22]Sawada M.Rookcase:an excel 97/2000 visual basic (VB) add-in for exploring global and local spatial autocorrelation[J].Bulletin of Ecological Society of America,1999,80(23):234.
[23]Brunsdon C,Fotheringham A S,Charlton M.Geographically weighted regression:A method for exploring spatial nonstationarity[J].Geographical Analysis,1996,28:281-298.
[24]欧光龙,王俊峰,肖义发,等.思茅松天然林单木生物量地理加权回归模型构建[J].林业科学研究,2014,27(2):213-218.
[25]Brown S.Measuring carbon in forests:current status and future challenges[J].Environ Pollut,2002,116:363-372.
