中国农业科学 ›› 2023, Vol. 56 ›› Issue (8): 1456-1470.doi: 10.3864/j.issn.0578-1752.2023.08.003

• 耕作栽培·生理生化·农业信息技术 • 上一篇    下一篇

大田不同播种间距单株小麦根长密度动态研究

孙启滨1(), 王建楠2, 李毅念1, 何瑞银1, 丁启朔1()   

  1. 1 南京农业大学工学院/江苏省智能化农业装备重点实验室,南京 210031
    2 农业农村部南京农业机械化研究所,南京 210031
  • 收稿日期:2022-08-08 接受日期:2023-01-08 出版日期:2023-04-16 发布日期:2023-04-23
  • 联系方式: 孙启滨,E-mail:qibsun@stu.njau.edu.cn。
  • 基金资助:
    国家重点研发计划“粮食丰产增效科技创新”重点专项(2016YFD0300900); 江苏省苏北科技专项(SZ-LYG2017008)

Study on the Dynamics of Root Length Density in Soil Layers of Single Plant Wheat Under Controlled Seed-to-Seed Distance

SUN QiBin1(), WANG JianNan2, LI YiNian1, HE RuiYin1, DING QiShuo1()   

  1. 1 College of Engineering, Nanjing Agricultural University/Key Laboratory of Intelligent Agricultural Equipment of Jiangsu Province, Nanjing 210031
    2 Nanjing Institute of Agricultural Mechanization, Ministry of Agriculture and Rural Affairs, Nanjing 210031
  • Received:2022-08-08 Accepted:2023-01-08 Published:2023-04-16 Online:2023-04-23

摘要:

【目的】为探明单粒精播种植方式种间距对小麦根系的土层分布影响,构建了基于根系数字化仪实测根系3D拓扑结构数据下MATLAB分割分析复合型小麦根长密度(RLD)定量技术,获取大田条件不同种间距单株稻茬麦RLD在不同土层的分布特征和相对根长密度(NRLD)分布模型。【方法】选用宁麦13为试验材料,采用免耕等距单粒线播法,分别于2020和2021年进行稻茬小麦的免耕种植试验,设置单粒精播种间距 1.5、3.0、4.5、6.7、9.0 cm共5个处理(JT1.5、JT3、JT4.5、JT6.7、JT9),行距20 cm。RLD分析采用根系构型数字化仪实测根系3D拓扑结构配合Pro-E软件数字重构,辅以MATLAB实现基于“空间voxel元技术”的根系生长空间3 cm3精细分割和定量分析,跟踪各土层RLD分布动态和NRLD模型。【结果】不同处理的单株稻茬麦根长密度随土层深度的增加而逐步减小,0—9 cm土层深度内分布的稻茬麦根系达总根量95%以上,超过9 cm土层深度小麦根系急剧减小;单株小麦根系扩展面积随土层深度的增加先增加后减小,根构型以种子位作为中心点向四周拓展,且表现出明显的拓展方向性和约束性。群体稻茬麦RLD随种间距增加先增加后减小,且在JT4.5最大;RLD扩展面积随种间距的增大而不断增加,最大可达22 972 mm2。过高与过低的群体都造成不良的根构型影响,适宜的播种密度才能创建最佳的根系3D分布,实现土壤空间资源的高效利用。经归一化处理后0—20 cm土层NRLD分布同时符合三次多项式和指数模型,两模型的拟合效果均极好(R2>0.99,RMSE<0.1),但指数模型更符合随土层深度的RLD特征实际。【结论】融合根系数字化仪实测根系3D拓扑结构与MATLAB分割分析的复合型小麦根长密度定量技术实现了单粒精播稻茬麦不同种间距的单株及群体双尺度的根长密度分布动态的定量描述,所得结果可为今后开展小麦精确耕种、水肥精准运筹、根构型调控等研究提供方法学借鉴。

关键词: 稻茬麦, 种间距, 根系分割, 根长密度, 模型

Abstract:

【Objective】In order to quantify the influence of seed-to-seed distance on wheat root development in soil layers under single seed precision sowing, an integrated technique combining root architecture digitizer and MATLAB simulation was developed to quantify wheat root length density (RLD) and relative root length density (NRLD), as well as related models in each soil layer in the field.【Method】Ningmai 13 was used as experiment marital and the seed was sown with single seed precision sowing method in no-till paddy soil. The experiment was carried out in 2020 and 2021, respectively. Five treatments (JT1.5, JT3, JT4.5, JT6.7, and JT9) with row spacing of 1.5, 3.0, 4.5, 6.7 and 9.0 cm were introduced for field stand control. RLD was analyzed with combined technologies, i.e. root architecture digitizer and 3D root system architecture reconstruction with Pro-E, supplemented with MATLAB simulation, which facilitated fine segmentation and analysis of the rhizosphere dynamics under soil space voxel resolution of 3 mm3, and this further results quantified RLD distribution dynamics and the development of NRLD models along soil layers.【Result】The post-paddy wheat RLD decreased gradually along the soil layers under different treatments. As much as 95% of the root system was confined within the top soil layer in 0-9 cm, below which, root length decreased rapidly. The wheat root expansion area of a single plant first increased along the soil layers and then decreased. Root expansion started from the seed site as its central point, and revealed an obvious directional and constraining effects induced by the soil environment. With the increase of seed-to-seed distance, wheat RLD experienced first an increasing and then a decreasing trend, and the maximum value of which was found at JT4.5. The expansion area of wheat RLD increased with the increased seed-to-seed distance, and the maximum value of which was 22 972 mm2. Either the too high or the too low density stand was found adversely impacts the efficiency of root configuration. Only the most suitable sowing density led to the best 3D distribution of wheat root system, which has been considered as the primary mechanism for efficient utilization of soil spatial resources. The NRLD distribution within 0-20 cm soil layers satisfied both cubic polynomial and exponential models well (R2>0.99, RMSE<0.1), but when considered the field state root system architecture, it was found that the exponential model was more realistic and fit the field wheat RLD the best along the soil layers.【Conclusion】An integrated technique combining root architecture digitizer and MATLAB simulation was developed to quantify wheat RLD and NRLD in the field, which satisfactorily illustrated the influence of seed-to-seed distance on RLD and NRLD along the soil layers. The results showed that the proposed method could be applicable for studies of wheat precision cultivation, precise water and fertilizer management, root configuration regulation and so on in the future.

Key words: post-paddy rice, seed-to-seed distance, root segmentation, root length density, model development