Optimizing the seeding rate is critical for wheat productivity. However, excessive densification often compromises individual photosynthetic efficiency to maximize canopy light interception, creating a trade-off that constrains yield. The physiological mechanisms linking canopy structure, photosynthesis, and source-sink coordination under varying seeding rates remain insufficiently quantified. Based on a two-year field trial assessing four seeding rates, i.e., 22.5 × 10⁵ (D15), 30.0 × 10⁵ (D20), 37.5 × 10⁵ (D25), and 45.0 × 10⁵ (D30) plants ha⁻¹, this study evaluated how seeding rate regulates leaf photosynthetic performance, canopy architecture, dry matter accumulation and yield formation in winter wheat. Compared to D15, the D20 treatment maintained photosynthetic integrity. D25 and D30 triggered severe non-stomatal limitations, characterized by a 14.1%–28.1% decrease in net photosynthetic rate and a 10.4%–25.0% increase in intercellular CO2 concentration. This physiological decline negated the structural benefits of maximized light interception. Mechanistically, D20 optimized the source-sink relationship by promoting post-flowering dry matter accumulation, significantly increasing the number of spikes by 35.2%, thereby enhancing the yield by 30%. Crucially, it avoided the resource-dilution penalties typical of high seeding rates. Partial Least Squares Path Modeling confirmed that seeding rate regulates yield by reconciling the trade-off between population expansion and metabolic competence. Machine learning simulation further refined the optimal seeding rate to 30.7 × 10⁵ plants ha⁻¹, defining a high-yield stability window (26.4–35.1 × 10⁵ plants ha⁻¹) characterized by specific physiological indicators (LAI 4.6–5.3, upper IPAR 68%–73%, Pn 22.2–25.5 µmol m⁻² s⁻¹). Optimizing the seeding rate successfully reconciles the trade-off between canopy light capture and photosynthetic efficiency. By maintaining metabolic integrity while expanding population structure, the optimal seeding rate coordinates source-sink relationships to maximize yield. This study provides a mechanistic framework and quantitative physiological thresholds for precision management. Future applications should account for genotype-by-environment (G×E) interactions to refine these parameters across diverse wheat production systems.
Zhang et al. (2026) studied this question.