Drought-rehydration irrigation has an enhancing impact on rice yield, but the current research on its yield-increasing effect is mainly experimental and empirical, lacking mechanism theoretical support. Image-based machine vision is rapidly developing and can estimate crop physical and chemical properties. A novel image processing method has been purposefully carried out to detect the real-time response shape of rice drought-rehydration. By application of this method, two new types of morphological descriptors were proposed to characterize and quantify the vertical phenotypic heterogeneity of rice, in which the relative height of the plant centroid (RHC) locates the growth focus, while the leaf angle distribution model describes the vertical characteristics of the leaf phenotypic traits. We verified the response of the vertical traits to different water treatments through designed experiments. The results showed that the RHC and leaf angle distribution parameters followed divergent trends under water stress, reflecting the drought characteristics of rice at different growth stages. The newly developed indicators were sensitive to drought response at specific growth stages and also efficient for evaluating rice growth, including determination of radiation interception capacity and assessment of nutrient accumulation. Furthermore, through the measurement and analysis of vertical structural traits, we found that a short-term water deficit and reasonable rehydration during the rice heading period could help to extend the spike-growing time and improve photosynthetic efficiency, thus benefiting yield formation.
Keywords: drought response; drought-rehydration; image processing; rice vertical heterogeneity; vertical structural traits.
Copyright © 2022 Zhang, Zha, Jin, Wang and Qiao.