Seed-score (S-score), a Method for Characterizing Seed and Lint Indices of Cotton Lines

Abstract

Increased lint yields of cotton (Gossypium hirsutum L.) cultivars during recent years have been accompanied by increased lint percentages and decreased seed size. Small seed size can be associated with low seed and seedling vigor and can contribute to ginning problems. In contrast, large seed size can be associated with thin seed coats and lower lint yields. Our objective was to develop an index that would characterize seed size and lint weight per seed. Seed-score (S-score) attempts to normalize seed index (SI) and lint index (LI) into a single index with penalties for both high and low SI values and no penalty for high LI values. Location × cultivar means (6,453 lines) for SI and LI extracted from the 1999 through 2020 Arkansas Cotton Variety Testing program produced mean SI of 10.17 ± 1.07 g and mean LI of 7.01 ± 0.90 g. These data were used to develop the normalization and weighting of factors for S-score. S-score was then calculated for transgenic cultivars evaluated in the 2015 through 2020 Arkansas Cotton Variety Tests. Within each year, cultivar was the major source of variation for SI, LI, and S-score. The 2015-2017 data set and the 2018-2020 data set produced 12 and 15 common cultivars, respectively. S-score among cultivars varied by more than 25 points in each data set and was relatively consistent over years. S-score will most likely be used as a secondary selection criterion in cotton cultivar development programs.