To determine the optimum parameters for extracting three carotenoids including zeaxanthin, lutein epoxide, and violaxanthin from pepper leaves by response surface methodology (RSM), a solvent of acetone and ethyl acetate (1:2) was used to extract carotenoids with four independent factors: ultrasound time (20–60 min); ratio of sample to solvent (1:12–1:4); saponification time (10–50 min); and concentration of saponification solution (KOH–methanol) (10–30%). A second-order polynomial model produced a satisfactory fitting of the experimental data with regard to zeaxanthin (R² = 75.95%, p < 0.0197), lutein epoxide (R² = 90.24%, p < 0.0001), and violaxanthin (R² = 73.84%, p < 0.0809) content. The optimum joint extraction conditions of zeaxanthin, lutein epoxide, and violaxanthin were 40 min, 1:8, 32 min, and 20%, respectively. The optimal predicted contents for zeaxanthin (0.823022 µg/g DW), lutein epoxide (4.03684 µg/g dry; DW—dry weight), and violaxanthin (16.1972 µg/g DW) in extraction had little difference with the actual experimental values obtained under the optimum extraction conditions for each response: zeaxanthin (0.8118 µg/g DW), lutein epoxide (3.9497 µg/g DW), and violaxanthin (16.1590 µg/g DW), which provides a theoretical basis and method for cultivating new varieties at low temperatures and weak light resistance.

Uneven seed placement in nursery trays reduces seedling uniformity and can reduce the reliability of mechanical transplanting in hybrid rice, but the optimum printed seeding density remains unclear. This study evaluated the effects of printed seeding density on seed distribution, seedling quality, transplanting performance, canopy productivity and yield formation in machine-transplanted hybrid indica rice. A two-year split-plot field experiment was conducted in Xuzhou, China, using Runliangyou 313 and Yangxianyou 903. Five printed seeding densities (1400–2600 printed points tray⁻¹) were compared with two local weight-based broadcasting controls, representing practical establishment systems rather than seed-number-matched contrasts. Moderate printed densities improved seed distribution uniformity, strengthened the seedling mat, reduced transplanting defects and supported productive tiller formation. T3 and T4 produced the highest harvested yields, increasing yield by 13.3–15.5% over the standard broadcasting control. These gains were associated with higher panicle number, greater post-anthesis dry matter accumulation and higher harvest index. The results indicate that moderate-density printed seeding can improve establishment quality and grain yield under wheat-rice rotation conditions.