Combining high-throughput phenotyping with overall growth measurements of indica rice (Oryza sativa L spp. indica) cultivars over the whole life cycle

Abstract

Importance of the work: High-throughput phenotyping systems containing non-destructive and non-invasive characterizations of phenotypic traits throughout the whole life cycle of plant development have prevailed over the conventional method.
Objectives: To evaluate the phenotypic characteristics of indica rice genotypes using red-green-blue (RGB) high-throughput phenotyping over the whole life cycle in relation to biomass and yield components.
Materials & Methods: Plant canopy width, canopy height and leaf area values of the rice cultivars RD41, Pathumthani1 (PT1), Homchonlasit, IR64, Riceberry and RD43 were measured using RGB imagery estimation together with actual measurements at 45 d after planting (DAP), 60 DAP, 75 DAP, 90 DAP, 105 DAP and 120 DAP.
Results: Canopy width and canopy height values obtained from actual measurements were linearly related to RGB-estimated values in all rice cultivars with values for the correlation coefficient (r) of 0.87–0.93 and 0.90–0.99, respectively. Notably, there was a positive relationship between plant projected area from the RGB imagery and the leaf area measurement, especially at the vegetative stage (r = 0.93–0.99). At harvest, there was also a positive relationship between aboveground biomass and total yield (coefficient of determination (R2) = 0.44). The agronomical traits and plant characterizations of RD41, PT1, Homchonlasit, IR64, Riceberry and RD43 were validated over the whole life cycle of rice crops.
Main finding: High-throughput phenotyping data collection should overcome conventional measurements due to its non-destructive, rapid and automated production for large amounts of data and high accuracy in indica rice crops.