ABSTRACT

Cotton with higher yarn strength is needed to keep pace with faster weaving methods being adopted by the textile industry that require stronger yarn. In the Pee Dee breeding program, selection for high fiber strength (stelometer) in early generations followed by direct selection for yarn strength in more advanced generations has been successful in increasing yarn strength. In current populations, however, fiber strength is not consistently correlated with yarn strength. For example, the correlation between fiber strength and yarn strength among 285 random F4 lines was only 0.25 (p<0.05).

Cotton breeders face limitations in their fiber and yarn testing programs. The standard fiber properties consisting of fiber strength, fiber length, fiber elongation, and micronaire can be measured on large numbers of genotypes at a relatively cost efficient $ 4.00 per sample. Yarn strength is expensive, currently $ 28.50 per sample, and thus is not obtained until later generations. Consequently, direct selection for yarn strength in early generations is not possible. Since yarn strength is not consistently correlated with fiber strength in current Pee Dee populations, additional fiber properties may need to be considered to further improve yarn strength. As a first step to evaluate additional selection criteria for the improvement of yarn strength, the objective of this study was to elucidate the relationship of fiber properties and traits such as fiber fineness and maturity obtained from the arealometer instrument with yarn strength. Among 285 random F4 lines, none of the standard fiber properties including 2.5 and 50% length, stelometer strength, fiber elongation, or micronaire was highly correlated with yarn strength. A multiple regression model including micronaire, 2.5% fiber length, elongation, and stelometer strength accounted for only 9 % of the variation in yarn strength. Examination of the residuals suggested that a more complicated model including interaction or polynomial terms would not improve the R2. Apparently variation in yarn strength in this population was due to fiber trait(s) not measured. Among 25 random F5 Pee Dee lines grown in two tests in 1992 at Florence, SC, again none of the standard fiber traits was highly or consistently correlated with yarn strength within or across tests. Fiber properties obtained with the arealometer were, however, more consistently correlated with yarn strength. For example, the correlation between fiber cell wall thickness and yarn strength was -0.62 (p<0.01) in both tests as compared to the correlation between 2.5 % fiber length and yarn strength which ranged from 0.32 to 0.69. The utility of each of the standard fiber properties as well as traits such as wall thickness as selection critera to increase yarn strength was also assesed. Stelometer strength was no more effective than 2.5 % fiber length in identifying the 5 F5 lines with highest yarn strength. A selection index was also employed to simultaneously rank the 25 F5 lines for a combination of traits such as fiber strength, fiber length, and cell wall thickness. The index was no more efficient at identifying lines with superior yarn strength than individual traits. This preliminary analysis of fiber and yarn data among a set of unselected Pee Dee genotypes suggests that variation in yarn strength is not well-explained by individual standard fiber or arealometer traits. The magnitude and consistency of the correlations between the arealometer traits and yarn strength however, suggests that they be considered as selection criteria in the improvement of yarn strength. Research is underway to further elucidate relationships among fiber/arealometer traits with yarn strength.