Changes in within-boll yield components explain cotton yield and quality variation across planting dates under a double cropping system of cotton-wheat

Context: Cotton (Gossypium hirsutum L.) yield is determined by whole-plant and within-boll yield components. Little information exists regarding within-boll yield component variations related to the planting date. Objective: This study aimed to investigate whether the differences in within-boll yield components will largely explain yield variation across planting dates under a double cropping system of wheat (Triticum aestivum L.) and short-season cotton.Methods: Two contemporary cotton cultivars were grown in the field with four different planting dates (PDMAY10, PDMAY20, PDMAY30 and PDJUN10) for the 2019-2021 growing seasons. Climatic variables and cotton growth pe-riods, plant mapping, extensive yield components and fiber quality assessments were recorded each year.Results: Compared to PDMAY10, the lint yields of PDMAY20, PDMAY30 and PDJUN10 were reduced by 3.8%, 14.7% and 44.1%, respectively. The significant yield loss in PDJUN10 could be explained by the variations in yield components. At the canopy level, boll density was the greatest contributor (75.6-90.8%) to yield loss. Moreover, within bolls, lint mass per seed (11.2-18.2%) and lint mass per unit seed surface area (11.9-16.2%) accounted for the most yield loss. In addition, seed number per boll, lint mass per seed and lint mass per unit seed surface area decreased by 1.0-6.2%, 7.3-15.0% and 5.5-13.4% for PDJUN10 relative to PDMAY20, PDMAY30 and PDJUN10, respectively. The number of fibers per seed and the seed index remained relatively stable among the planting dates across all three years. No significant differences in fiber length and fiber strength were observed among planting dates in 2019 and 2020, while shorter and weaker fibers were observed for the earliest planting date in 2021, possibly because of extreme rainfall that occurred at the end of July, indicating that a deteriorated environment could reduce the effect of the planting date. Among all climatic variables, the daily maximum temperature during the flowering and boll development stage had the strongest correlation with the lint yield.Conclusions: The first three planting dates generally produced comparable lint yields and fiber qualities. Therefore, it might be important to plant short-season cotton before May 30th in the Yellow River Valley to maintain relatively high yield and fiber quality. Implications: This study provided new results to assist decision making regarding cotton planting date in double cropping of wheat and cotton in the Yellow River Valley.