Bridging photosynthesis and crop yield formation with a mechanistic model of whole-plant carbon–nitrogen interaction

Abstract

Abstract Crop yield is determined by potential harvest organ size, source organ photosynthesis and carbohydrate partitioning. Filling the harvest organ efficiently remains a challenge. Here, we developed a kinetic model of rice grain filling, which scales from the primary biochemical and biophysical processes of photosynthesis to whole-plant carbon and nitrogen dynamics. The model reproduces the rice yield formation process under different environmental and genetic perturbations. In silico screening identified a range of post-anthesis targets—both established and novel—that can be manipulated to enhance rice yield. Remarkably, we pinpointed the stability of grain-filling rate from flowering to harvest as a critical factor for maximizing grain yield. This finding was further validated in two independent super-high-yielding rice cultivars, each yielding approximately 21 t ha−1 of rough rice at 14% moisture content. Furthermore, we revealed that stabilizing the grain-filling rate could lead to a potential yield increase of 30–40% in an elite rice cultivar. Notably, the instantaneous grain-filling rates around 15- and 38-day post-flowering significantly influence grain yield; and we introduced an innovative in situ approach using ear respiratory rates for precise quantification of these rates. We finally derived an equation to predict the maximum dried brown rice yield (Y, t ha−1) of a cultivar based on its potential gross photosynthetic accumulation from flowering to harvest (Apc, t CO2 ha−1): Y = 0.74 × Apc + 1.9. Overall, this work establishes a framework for quantitatively dissecting crop physiology and designing high-yielding ideotypes.

Plant phenotyping relevance

全植物の炭素・窒素動態と収量形成を推定する速度論モデルを開発し、耳の呼吸速度による粒充填速度の定量化手法も導入しているため、表現型取得・推定が中心的です。

Code and data availability

The paper's authors publicly released the WACNI model source code (the computational framework used for all simulations and analysis) on GitHub, with explicit availability language in the MODEL AND DATA AVAILABILITY section. Supplementary Data 2 contains literature-extracted experimental data but no separate public URL