关键词: 水稻, 根系形态, 定量分析, WinRhizo, 图像分析

Abstract:

The objective of this study was to quantitatively analyze the change patterns of rice root morphological characteristics based on image analysis method in pot experiments with different nitrogen rates, water regimes, and rice cultivars in different years. WinRhizo software (WR) was used as the root image analysis tool, and two modes of LM and T200 were used to measure the root diameter and adventitious root length under two ranges of root diameter as 0–0.30 mm and 0.25–1.80 mm, respectively. The measurement results from WR were compared with those from Image-Pro Plus 6.0 software (IP). Then, the diameter ranges of different types of root were determined. On the basis of the dataset from WR, the statistical analysis were performed on rice root morphological characteristic indices, including dry weight, total volume, total surface area, total adventitious root length and number per plant at different growth stages. The results were as follows: (1) The average NRMSEs between WR (LM and T200) and IP for root diameters were 5.44% and 10.95%. (2) The diameter range of fine lateral root was circumscribed within 0.03–0.10 mm for all the experiments, coarse lateral and adventitious root were circumscribed within 0.10-0.30 and 0.30-1.40 mm for V1 (Nipponbare) and V2 (Wuxiangjing 14), but 0.10–0.25 and 0.25–1.65 mm for V3 (Yangdao 6). Following this method, the NRMSEs for adventitious root length measured by WR and IP were 10.78%–12.01%. (3) All indices increased rapidly in the growth process and reached or approached the maximum at heading, then increased slowly or decreased after heading. There were significant differences between treatments with different nitrogen rates from tillering to maturing on all indices. Increasing nitrogen rate accelerated the growth of roots, and especially enhanced the proportion of adventitious roots. Properly control of irrigation could promote the growth of roots and increase the proportion of lateral roots. All indices of V3 were significantly higher than these of other cultivars, and there were no significant differences between V1 and V2. The results indicated that it was feasible to measure morphological indices of different types of rice root by WR with different modes, and the methods used in this paper have a good accuracy and reliability. These results could provide a support for the quantitative analysis on root morphology of rice or other crops.

Key words: Rice, Root morphology, Quantitative analysis, WinRhizo, Image analysis