Meng Yuan, Hongke Ding, Yunkai Yue, Chuanliang Zhang, Yuze Wang, Dandan Zhang, Shipeng Li, Ruiqi Nie, Jia Liu, Yanchuan Huang, Jinyu Han, Sufu Deng, Shengjie Liu, Chunlian Li, Weijun Zheng, Daojie Sun, Qiang Yao, Shouyang Liu, Shengwei Ma, Zifeng Guo, Jie Liu, Qingdong Zeng, Zhensheng Kang, Rui Yu, Dejun Han, Jianhui Wu
Sustaining wheat yield gains requires optimizing the spatiotemporal coordination of source (leaves), transport (stems), and sink (spikes) organs. However, the physiological mechanisms and underlying genetic networks orchestrating the dynamic development of these critical structures remain largely uncharacterized. Here, we leveraged high-resolution time-series phenotyping across 590 wheat accessions evaluated across three year-site environments (comprising two locations and two growing seasons) to dissect the genetic architecture of these biomass partitioning trajectories. To fully capture this spatiotemporal regulation, our analysis explicitly integrated both the temporal tracking across five floret developmental stages (Z39-Z65) and the spatial partitioning among these organ-specific dynamic systems. We identified 36 multi-stage stable dynamic quantitative trait loci (QTL) regulating five source-sink-related traits. By constructing genetic association and epistatic interaction networks, we prioritized two pivotal dynamic QTL, namely Qa.nw-7B.848 and Qa.nw-1D.96. Multi-omics integration pinpointed TraesCS7B03G1340600 as a key candidate gene for Qa.nw-7B.848. Furthermore, haplotype analysis uncovered distinct selection footprints, demonstrating how specific allelic combinations have been differentially selected to optimize yield components across diverse geographical environments. Collectively, this study moves beyond static trait analysis, offering a dynamic genetic framework and specific epistatic targets to precision-design wheat architecture for enhanced productivity.