摘要： 被子植物木质部导管的分布格局非常多样, 并且与木质部的输水功能有密切的联系, 然而在木材解剖学中对导管分布格局往往采用定性描述, 不利于分析该特征与物种的水力功能、生态地理分布的关系。该文采用点格局分析方法, 依据木材孔型、导管空间排列和导管群集度三类木材宏观结构特征组合, 选取不同导管分布类型的17种代表性阔叶树种, 利用Strauss-Hardcore模型对其木质部横切面解剖影像进行定量分析。Strauss-Hardcore模型能够很好地拟合木质部中导管二维空间位点的分布特征, 该模型的3个参数: 硬核距离、局部聚集距离、点对交互作用强度(局部聚集指数)都有着明确的生物学意义。传统解剖学对导管构型的定性分类同模型相比不能准确表现被子植物的木质部导管空间分布特征, Strauss-Hardcore模型的局部聚集度指数主要受导管群集度影响, 尤其是复导管和导管团的存在都会增大导管小尺度聚集程度。对散孔材、半环孔材的生长轮及环孔材的晚材部分解剖图像分析表明, 导管以单导管为主且没有明显分布方向的散孔材树种, 其木质部导管点对交互作用强度为负值, 局部聚集指数一般小于0.4, 导管空间分布依次在3个局部尺度表现出排斥-排斥-随机格局; 而导管具有径向、弦向、锯齿形等明显目视识别特征的物种, 无论孔型和是否以单复导管为主, 其导管点对交互作用强度为正值, 局部聚集指数均大于0.4, 导管依次在3个局部尺度上表现出排斥-聚集-随机的分布格局。采用点过程模型有利于准确描述导管二维空间分布规律, 增强对导管空间格局形成机理的理解, 可有力地支撑木质部三维导管系统的理论研究和木质部结构-功能的实验研究。
被子植物木质部导管的分布格局非常多样, 并且与木质部的输水功能有密切的联系, 然而在木材解剖学中对导管分布格局往往采用定性描述, 不利于分析该特征与物种的水力功能、生态地理分布的关系。该文采用点格局分析方法, 依据木材孔型、导管空间排列和导管群集度三类木材宏观结构特征组合, 选取不同导管分布类型的17种代表性阔叶树种, 利用Strauss-Hardcore模型对其木质部横切面解剖影像进行定量分析。Strauss-Hardcore模型能够很好地拟合木质部中导管二维空间位点的分布特征, 该模型的3个参数: 硬核距离、局部聚集距离、点对交互作用强度(局部聚集指数)都有着明确的生物学意义。传统解剖学对导管构型的定性分类同模型相比不能准确表现被子植物的木质部导管空间分布特征, Strauss-Hardcore模型的局部聚集度指数主要受导管群集度影响, 尤其是复导管和导管团的存在都会增大导管小尺度聚集程度。对散孔材、半环孔材的生长轮及环孔材的晚材部分解剖图像分析表明, 导管以单导管为主且没有明显分布方向的散孔材树种, 其木质部导管点对交互作用强度为负值, 局部聚集指数一般小于0.4, 导管空间分布依次在3个局部尺度表现出排斥-排斥-随机格局; 而导管具有径向、弦向、锯齿形等明显目视识别特征的物种, 无论孔型和是否以单复导管为主, 其导管点对交互作用强度为正值, 局部聚集指数均大于0.4, 导管依次在3个局部尺度上表现出排斥-聚集-随机的分布格局。采用点过程模型有利于准确描述导管二维空间分布规律, 增强对导管空间格局形成机理的理解, 可有力地支撑木质部三维导管系统的理论研究和木质部结构-功能的实验研究。
Abstract: Aims Spatial patterns of vessel in xylem are diverse and closely related with water transportation functions in angiosperms. However, the pattern was generally described qualitatively in anatomy, which were unable to reveal their links to xylem functions and to species distribution. We used point pattern analysis to study vessel spatial pattern in xylem cross-sectional images to quantify their features. Methods Images of 17 types of vessel configurations were selected in terms of wood porosity, vessel arrangement, and vessel grouping. Optimum Strauss-Hardcore models for coordinates in the images were fitted. Correlations among vessel variables and model coefficients were tested. Important findings We found that (1) Strauss-Hardcore model fitted all the data well and its three parameters, i.e., hardcore distance, local aggregation distance, and point-pair interaction or point aggregation index, and had apparent biological significance. (2) Classifications of wood xylem by traditional anatomical indices could not precisely present the spatial pattern of vessels compared to spatial point analysis, and local aggregation index from Strauss-Hardcore model was mainly influenced by vessel grouping, especially frequency of radial multiples and vessel clusters. (3) Among the 17 vessel patterns analyzed, diffusive or semi-ring species with xylem consisting of solidary vessels showed negative point-pair interaction and aggregation index was less than 0.4, whereas species with obvious vessel arrangement and multiple or clusters of vessel grouping in xylem owned positive point-pair interaction and bigger aggregation index. (4) The former group of species demonstrated inhibition- inhibition-random pattern at three local scales while the latter species showed inhibition-aggregation-random pattern according to the fitted Strauss-Hardcore models. The findings showed that point process modeling could precisely describe vessel distribution features in 2-D xylem sections and provide insights on vessel development. Therefore, this method may support 3-D vessel system simulation and experimental studies on structure-function of angiosperm xylem.
Aims Spatial patterns of vessel in xylem are diverse and closely related with water transportation functions in angiosperms. However, the pattern was generally described qualitatively in anatomy, which were unable to reveal their links to xylem functions and to species distribution. We used point pattern analysis to study vessel spatial pattern in xylem cross-sectional images to quantify their features.
Methods Images of 17 types of vessel configurations were selected in terms of wood porosity, vessel arrangement, and vessel grouping. Optimum Strauss-Hardcore models for coordinates in the images were fitted. Correlations among vessel variables and model coefficients were tested.
Important findings We found that (1) Strauss-Hardcore model fitted all the data well and its three parameters, i.e., hardcore distance, local aggregation distance, and point-pair interaction or point aggregation index, and had apparent biological significance. (2) Classifications of wood xylem by traditional anatomical indices could not precisely present the spatial pattern of vessels compared to spatial point analysis, and local aggregation index from Strauss-Hardcore model was mainly influenced by vessel grouping, especially frequency of radial multiples and vessel clusters. (3) Among the 17 vessel patterns analyzed, diffusive or semi-ring species with xylem consisting of solidary vessels showed negative point-pair interaction and aggregation index was less than 0.4, whereas species with obvious vessel arrangement and multiple or clusters of vessel grouping in xylem owned positive point-pair interaction and bigger aggregation index. (4) The former group of species demonstrated inhibition- inhibition-random pattern at three local scales while the latter species showed inhibition-aggregation-random pattern according to the fitted Strauss-Hardcore models. The findings showed that point process modeling could precisely describe vessel distribution features in 2-D xylem sections and provide insights on vessel development. Therefore, this method may support 3-D vessel system simulation and experimental studies on structure-function of angiosperm xylem.
图1 Stewartia pseudocamelliashi的导管空间分布及其Strauss-Hardcore (SH)模型的包络检验。A, 木质部横切面的导管分布情况。每个圆圈代表一个导管, 圆圈直径为导管直径(µm)。B, 显著性水平为0.05的L函数包络检验。r, 点对距离; 纵坐标为公式(3)定义的L函数。黑色线代表实际数据拟合得到SH模型的L函数值, 红色线代表采用相同参数的SH理论分布模型的19次拟合平均值, 绿色和蓝色线分别代表19次理论模型拟合得到的2.5%和97.5%分位数值。
Fig. 1 Spatial distribution of vessels in xylem of Stewartia pseudocamelliashi and envelope test of fitted Strauss-Hardcore (SH) model. A, Distribution of vessels in xylem cross-section. Each cycle stands for a vessel with cycle diameter as vessel diameter (µm). B, Envelope test for L function at significance of 0.05. r, the distance of paired points; Y-axis is the L function defined in equation (3). Black line represents L value from data fitted SH model, red line for average L value from 19 simulation of theoretical SH model, green and blue lines represent 2.5% and 97.5% quantile of L value from 19 simulation of theoretical SH model respectively.
表1 不同导管构型代表物种的Straus-Hardcore模型拟合结果
Table 1 Strauss-Hardcore model fiting for representative species with various xylem vessel pattern
表2 用于导管构型定量分析的不同空间点-过程模型的特点
Table 2 Characteristics of spatial point-process models for vessel configuration analysis