Indirect selection can be a useful means of improving quantitatively inherited traits. The purpose of this experiment was to determine how the quantitatively inherited traits of cotton fiber quality are related, and if this information could be applied to the improvement of cotton fiber. Correlations between traits indicate relationships, but are frequently difficult to interpret. Analysis of commonality, which is analogous to path coefficients and based on analysis of variance, is one method used to make sense of relationships found through correlations. During the summers of 1993 and 1994, plants were grown in the vicinity of Stoneville, MS. The experimental design was a randomized complete block with two replicates. One hundred and seven randomly selected F2-derived lines of upland cotton (Gossypium hirsutum L.) from the cross, Acala 1517-75BR1 X DPL SR-383, were sampled for fiber in the F3 and F4 generations. Eight fiber traits were measured — elongation, maturity, micronaire, perimeter, 2.5% span length, strength, wall thickness, and weight fineness. The analysis of commonality identified perimeter as the only fiber trait that influenced all traits. Perimeter had the greatest effect on micronaire (76.5% of total sums of squares), but also affected slightly more than 35% (uniquely and in conjunction with other traits) of the variation in models explaining length and strength. Elongation had the least in common with other fiber traits. These results indicate the potential to decrease micronaire and increase fiber length and strength by selection for smaller perimeter. However, indirect selection for fiber quality may not be practical as measuring perimeter adds additional costs and perimeter's effects on length and strength may be too small for meaningful improvement in length and strength.
