摘要
【目的】根构型直接影响作物的水肥吸收,而定量根构型的相关指标多局限于二维分析,缺乏有效的3D分析指标。论文探讨计算分析根构型3D特征的指标与操作方法,用于定量稻麦轮作制不同耕作方式对小麦根构型的影响。【方法】使用自制的根构型数字化仪,测取田间小麦根系的真实空间拓扑数据,获得根系构型的空间坐标。然后运用Matlab编程实现小麦根构型拓扑数据的虚拟重构,令虚拟根系再现实体根系的空间拓扑。结合分形理论与软件的计算分析功能对虚拟根构型进行分形维计算,分别获取3D分形维数、3D分形丰度、2D分形维数、2D平面分形丰度和单株总根长5个特征指标,以此表达小麦根构型在不同年度、不同耕作方式处理下的时空动态。同时建立不同年度及耕作处理下单株总根长动态与根构型3D分形维数、3D分形丰度、2D分形维数、2D平面分形丰度间的相关关系。【结果】研究发现随着作物生长期的变化,不同年份及不同耕作处理下的小麦根构型指标都表现出稳定增长的趋势。不同之处在于2010—2011年度的小麦根构型指标平稳增长,而2011—2012年度的根系生长速率变化较为剧烈。对比两个年度间的小麦根构型指标发现,免耕和旋耕两种耕作方式对小麦根构型的影响效果相反,在2010—2011年度,旋耕处理方式下的根系指标优于免耕处理方式,而在2011—2012年度,免耕处理方式下的根构型指标表现更优。对于作物生长前期(0—98 d)而言,年度变化引起的根构型指标差异显著大于耕作处理引起的差异,在作物生长后期(98—112 d),年度变化和耕作处理方式对小麦根构型指标的影响较为相近。对比小麦根构型的3D分形维指标和平面分形维指标发现,3D分形维明显区别于平面分形维,这表明根系的三维分形是根构型的必要分析指标。在不同的年度与耕作措施下,单株小麦的总根长与3D分形维数、3D分形丰度、2D分形维数、2D分形丰度都满足指数模型,且显著相关,说明年度因素和耕作措施仅是影响模型的常量参数项。【结论】由计算机软硬件结合分形理论构建的田间小麦根构型的可视化和定量化分析手段是实现小麦根构型精确分析的保证,该分析过程真实再现了田间小麦根构型的时空动态。3D分形指标可以准确定量作物根构型真实的时空动态,在进行根系生长策略的选择及根土关系优化时需要考虑到田间作物根系的实际生长条件和耕作制度。
[Objective] Root system architecture （RSA） has a significant effect on water uptake and nutrient absorption. However, relevant indices for the quantification of crop RSAs are limited to 2D fractal analysis. Analytical tools for 3D fractal analysis on crop RSAs are lacking. Thus there is a need to investigate the related parameters and operational procedures suitable for the analysis of the 3D characteristics of crop RSAs. [Method] A self-fabricated digitizer for crop RSAs was used to measure the topological parameters of the field-grown wheat root, and the spatial dimensions of wheat RSAs were obtained. Virtual wheat RSAs were then modeled and reconstructed with Matlab programming, which guaranteed a realization of the real-world wheat RSAs with virtual reality. The fractal theory was then introduced into the computing software to calculate the fractal parameters of the modeled virtual wheat RSAs, including 3D fractal dimension, 3D fractal abundance, 2D fractal dimension, 2D fractal abundance and total root length. These parameters were used to quantify the dynamics of wheat RSAs, in both the 2 experimental years and the 2 tillage treatments. Correlations among 3D fractal dimension, 3D fractal abundance, 2D fractal dimension, 2D fractal abundance and total root length were also analyzed. [Result] It was found that all the RSA-related parameters were steadily increased along wheat developmental stages, in either different years or under different tillage treatments. Differences between the 2 years appeared as the 2010-2011 crop season revealed a steady increase of RSA-related parameters, while the 2011-2012 crop season observed a more radical increase of root elongation rate. A comparison between the 2 years revealed that tillage treatment had a contrasting effect from year to year, with a better crop performance under rotary till than no-till in the first year, whereas the no-till treatment in 2011-2012 outperformed the first year. At the early stage （0-98 d）, the crop season had pronounced influences on wheat RSAs, as compared with tillage treatments. At the ensuing stage （98-112 d）, however, annual difference of wheat RSA parameters was as similar as the tillage treatments. A comparison between 3D fractal parameters with the 2D parameters revealed that 3D parameters were markedly contrasted with the 2D parameters, indicating that introducing the 3D parameters for crop RSA analysis is necessary. Disregard annual difference and tillage treatment, all the dynamics of 3D fractal dimension, 3D fractal abundance, 2D fractal dimension, 2D fractal abundance and total root length satisfied power law functions and were all co-related significantly. This means that the effects of crop season and tillage treatment were only related to the coefficients of the power law models. [ Conclusion] It was concluded that the visualization and analytical tools developed with hardware and software integration and combined with fractal theory was a guarantee for precise quantification of crop root system architectures. Such an analytical tool allows recasting the spatio-and-temporal dynamics of field crop RSAs with modeled virtual roots. 3D fractal parameters could be used as a precision analytical tool for crop RSAs. In selecting root elongation tactics and optimizing the root-soil interactions an important consideration should be taken to match the crop root with its soil environment and the tillage system.
出处
《中国农业科学》
CAS
CSCD
北大核心
2017年第3期451-460,共10页
Scientia Agricultura Sinica
基金
国家重点研发计划"粮食丰产增效科技创新"重点专项(2016YFD0300900)
江苏省农机基金(201-051028)
关键词
稻茬麦根系
根构型可视化
3D分形维
平面分形维
根长动态
paddy wheat root system
visualization of root system architecture （RSA）
3D fractal analysis
2D fractal analysis
root length dynamics
