Effects of plant growth regulators on seed yield, quality, and physiological traits of seed-producing sugar beet

During sugar beet seed production, excessive stem elongation and lodging can adversely affect seed yield and quality. However, the effects of plant growth regulators (PGRs) on lodging resistance, physiological traits, and seed production performance remain insufficiently understood. A 2 year field experiment (2023–2024) was conducted to evaluate the effects of 1,1-dimethyl piperidinium chloride [DPC; 300 (D1), 500 (D2), and 900 g ha −1 (D3)] and chlormequat chloride [CC; 600 (C1), 1000 (C2), and 1800 g ha −1 (C3)] on plant architecture, physiological traits, seed yield (SY), and seed quality of sugar beet. Correlation analysis and Random Forest modeling were used to identify variables associated with SY and germination percentage (GP), and entropy weight–TOPSIS was used for comprehensive evaluation. Both PGRs reduced plant height (PH) and lodging rate (LR) without reducing the number of effective branches. Compared with CK, PH under D3 and C3 was reduced by 13.6% and 17.5%, respectively. LR reductions under D3 and C3 were 45.8% and 55.9% in 2023 and 33.7% and 63.7% in 2024, respectively. Both PGRs also increased total chlorophyll content, antioxidant enzyme activities, and soluble regulatory compounds, while reducing relative electrical conductivity and H 2 O 2 content; however, these responses were not accompanied by significant changes in MDA. Across the two growing seasons, C3 increased SY, thousand-seed weight, GP, and kernel plumpness rate by 32.1%, 32.9%, 19.8%, and 25.9%, respectively, compared with CK. Correlation analysis and Random Forest modeling identified APX, soluble reducing sugar content, and LR as variables associated with variation in SY and GP rather than direct causal determinants. C3 had the highest TOPSIS score in both years, but C2 and C3 showed comparable responses for several important yield and quality traits. Thus, the TOPSIS result represents a relative comprehensive ranking rather than evidence of an agronomic optimum.