Development and Utility of Q-score for Characterizing Cotton Fiber Quality

Cotton (Gossypium hirsutum L.) fiber quality characteristics may be improved with an index that incorporates weighted values of multiple fiber measurements. Our objectives were to describe the logic and calculation of a numerical index (Q-score), to evaluate the relationship of Q-score to loan value, and to illustrate the use of Q-score in cotton breeding and cultivar testing. The Q-score is calculated by first normalizing fiber properties from 0 to 1 and then algebraically combining the normalized values by qualityweighting factors based upon input from textile processing experts. Fiber properties (weights) for Q-score calculations included fiber length (50%), micronaire (25%), length uniformity index (15%), and strength (10%). Q-scores and loan values were calculated for the 2001-2007 Arkansas Cotton Variety Tests (1478 observations). Economic analysis included summary statistics and correlations for the parameters. Q-score values were normally (or near normally) distributed, while loan values followed a Poisson or chi-square distribution. Q-score and loan value were positively correlated, and similarly correlated to the fiber parameters. Obtaining optimum loan values was more likely as Q-score increased. Q-score was more conservative and discriminating than loan values. Data for 16 cultivars at four sites over three years indicated that Q-score values were relatively consistent over years and sites. Q-score was normally distributed in data extracted from the 2005-2008 Arkansas cotton breeding trials. This distribution facilitates subsequent development of superior cotton lines. These results indicate that Q-score may assist with characterizing fiber quality. However, application of Q-score is limited because relative weights of four fiber traits are subjectively assigned, and measurement of trash and color are not included.