基于精确数据的大田玉米植株根系结构分析、建模及其应用

发布时间：2018-06-07 05:56

本文选题：拓扑结构 + 根直径　；参考：《中国农业大学》2016年博士论文

【摘要】：根系是植物吸收水分、养分资源和响应土壤环境变化的重要器官。三维根系结构模型能够精确描述根系在土壤中的空间分布。基于拓扑结构定量化研究侧根形态对探索根系功能-结构具有重要意义,并可为参数化根系结构模型提供重要依据。本论文基于整根取样扫描测定的侧根拓扑结构和几何形态数据,量化了各级侧根形态,并分析了各个根系性状间的关系；结合田间原位测定的轴根三维数据,构建了不同根构型的玉米根系三维结构模型；基于模型评估了不同根钻取样策略的可靠性和适用性。主要结果如下：(1)侧根的基部直径具有较高的变异性。侧根的基部直径部分决定于支撑根直径的大小。基部直径决定了根的潜在生长长度,然而,大部分根的长度远低于其潜在生长长度。这可能是由于伴随根生长根直径变化模式的差异性造成的。对于大部分根,从根基到根尖,直径大多呈现下降趋势,且直径下降梯度的斜率依赖于根基部直径。较长的根多为在生长过程中能维持(或增加)其直径的根。侧根的分枝密度也受到支撑根直径大小的影响。然而,支撑根段在父级根所处的位置也影响到根段上侧根的分支密度。较高的侧根分支密度往往出现在父级根的中间位置。(2)1级侧根的长度和密度并不总是相关于其在轴根上的生长位置,而是相关于轴根直径的大小。有侧根分枝的1级侧根占1级侧根总数不到50%,但其对根系总长的贡献高达85-91%。1级侧根上分枝总数和总长的变异性不仅依赖于1级侧根的长度,还依赖于侧根单元(LRU)的级别。LRU级别越高,各级侧根长度、直径和分枝个数越大。1级侧根基部直径与2级侧根最大直径的正相关关系以及1级侧根基部直径和长度的正相关关系的斜率随着LRU级别的增加而升高。(3)基于实测数据构建了个体和群体玉米根系三维结构模型,对常用的根钻取样策略进行了虚拟取样评估。结果表明,采用单点取样策略所估测的根长密度(RLD)都存在显著偏差；采用两点取样策略估算RLD同样存在显著偏差,只有在赋予两个取样点特定权重时,才能得到偏差较小的RLD估算值。本研究提出基于面积权重估算RLD的行间多点取样策略。结果表明,采用行间3点取样策略时,使用较大直径根钻(10和7.5 cm)能获得偏差较小的RLD估算结果。忽略行间最远的取样点,仅使用2点取样时,也能得到偏差较小的RLD估算结果。本研究基于根段尺度量化了侧根直径和分支,系统检测了根系形态特征和几何结构,揭示的各个根系性状间的相关关系为库-源模型中根系的生长和分枝的模拟提供了重要依据。基于拓扑结构量化的各级侧根形态特征及其相关关系充分体现了侧根分枝在形态学上的连续性。基于根系模型的虚拟评估在优化根钻取样策略方面具有重要价值,为大田中后期玉米根系RLD分布的估测提供了最优取样策略。
[Abstract]:Root system is an important organ for plants to absorb water, nutrient resources and respond to changes in soil environment. Three-dimensional root structure model can accurately describe the spatial distribution of root system in soil. Quantitative study of lateral root morphology based on topological structure is of great significance in exploring root function-structure and can provide important basis for parameterized root structure model. In this paper, based on the topological structure and geometry data of lateral roots measured by whole root sampling scanning, we quantify the lateral root morphology of all levels and analyze the relationships among root characters, and combine with the axial root 3D data measured in situ in the field. The 3D structure model of maize root system with different root configurations was constructed, and the reliability and applicability of different sampling strategies of root drill were evaluated based on the model. The main results are as follows: 1) the base diameter of lateral roots has high variability. The basal diameter of lateral roots depends partly on the diameter of the supporting roots. The diameter of base determines the potential growth length of root, however, the length of most roots is much lower than its potential growth length. This may be due to the variation of root diameter with root growth. For most roots, the diameters tend to decrease from root to tip, and the slope of diameters gradient depends on the diameter of root. The longer roots are the ones that can maintain (or increase) their diameters during growth. The branch density of lateral roots was also affected by the diameter of supporting roots. However, the position of the supporting root segment in the parent root also affects the branch density of the lateral root on the root segment. The higher lateral root branching density often appears in the middle position of the parent root. The length and density of the lateral root are not always related to the growth position of the lateral root, but to the diameter of the axial root. The number of lateral roots with lateral roots branching is less than 50, but its contribution to the total root length is as high as 85-91.1 lateral roots. The variability of the total number of branches and the total length of lateral roots depends not only on the length of the first lateral root, but also on the total length of lateral roots. The higher the level. LRU is also dependent on the lateral root unit (LRU), the higher the lateral root length at each level, The larger the diameter and the number of branches, the higher the slope of the positive correlation between the basal diameter of the lateral root of order .1 and the maximum diameter of the lateral root of the second order, and the positive correlation between the diameter of the base of the lateral root of the first order and the length of the lateral root of the first order, and the increase of the slope with the increase of the LRU level. Three dimensional structure model of individual and population maize root system was constructed. The common sampling strategy of root drill is evaluated by virtual sampling. The results show that there are significant deviations in the estimation of root length densities by single point sampling strategy, and significant deviations exist in estimating RLD by two-point sampling strategy, only when two sampling points are given specific weights, In order to get the small deviation of the RLD estimate. In this study, a multi-point sampling strategy based on area weight estimation (RLD) is proposed. The results show that the RLD estimation results with small deviation can be obtained by using the large diameter root drill (10 and 7.5 cm) using the three-point sampling strategy between rows. RLD estimation results with small deviation can be obtained when only two sampling points are used to ignore the farthest sampling points between rows. In this study, the diameter and branch of lateral root were quantified based on root segment scale, and the morphological characteristics and geometric structure of root system were systematically examined. The correlation between root traits revealed provides an important basis for the simulation of root system growth and branching in the sink-source model. Morphological characteristics of lateral roots based on topological quantization and their correlation fully reflect the morphological continuity of lateral root branches. Virtual evaluation based on root model plays an important role in optimizing the sampling strategy of root drilling and provides the optimal sampling strategy for estimating the RLD distribution of maize roots in the middle and late stage of the field.
【学位授予单位】：中国农业大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：S513

【参考文献】