Three cycles of recurrent selection for increased lint yield in Upland cotton (Gossypium hirsutum L) have been completed. Yield of the third cycle selections exceeded that of the base population by 29.7%. Response to selection was linear and is predicted to continue at approximately the same rate of gain for an additional cycle.Correlated responses to selection for yield were observed for the majority of the unselected traits measured. As selection increased lint yield, simultaneous increases were observed for lint percentage, number of seed per boll, earliness, fiber elongation, and fiber coarseness. Boll size, seed size, and fiber length and strength decreased. Weight of lint per seed showed little change.There was generally close agreement between predicted and observed responses to selection. This suggests that dominance effects were lacking or of relatively minor importance. Additive genetic effects would appear to predominate in this population, although additive ✕ additive effects cannot be excluded.

A population of switchgrass, Panicum virgatum L., consisting of 33 half‐sib families was investigated over 2 years to obtain estimates of genetic parameters for several agronomic and nutritive traits, including height, maturity, dry weight, in vitro dry matter disappearance (IVDMD), and percent N. Dry weight was less heritable than other traits, with an individual narrow‐sense heritability of only 0.25. Narrow‐sense heritability for dry weight calculated on a family mean basis was 0.59. Dry weight and IVDMD were not significantly correlated while the correlation between dry weight and percent N was highly negative. Maturity was also negatively correlated with dry weight. The IVDMD and percent N, both measures of forage quality, had family narrow‐sense heritabilities averaged over three measurements per year of 0.79 and 0.74, respectively. Mean values for both traits were highest early in the season and declined markedly with maturity. The IVDMD and percent N were not significantly correlated. The estimated genetic parameters were used to form several selection indices involving three traits: dry weight, IVDMD, and percent N.

I T HAS been generally accepted that broad gene base testers are more efficient than narrow gene base testers for the evaluation of general combining ability in inbred lines of maize (Matzinger (7), Grogan and Zuber (4), and others). The importance of performance level or, by inference, gene frequency of the tester as a criterion for choice of testers, however, has never been as clearly established. Hull (5) stated that theoretically the most efficient tester would be homozygous recessive at all loci and that homozygosity for the dominance alleles at any locus should be avoided. These conclusions were based on considerations of the constant parent regression method of analysis of single crosses. The regression of performance of offspring on the performance of the variable parents for a particular constant parent was shown to be largest when the gene frequency of the character for the constant parent was zero. The regression was zero when the gene frequency was at one for complete dominance or at equilibrium gene frequency for overdominance. A strong positive regression would be desirable as this would allow more discrimination among the variable parents. Green (2) tested Hull's hypothesis with respect to lodging resistance. A lodging resistant double cross and a lodging susceptible open-pollinated variety were used as testers for the F 2 progenies of three single crosses. When considerable lodging occurred, there was little difference between the testers. With little lodging, however, the lodging susceptible tester gave a much greater range for standability within all the F ;2 progenies than the lodging resistant tester. Keller (6) reasoned that if Hull's hypothesis were correct, the component of variance due to the interaction of lines with testers would be less for high combining testers than for low combining testers. Separate estimates of the line X tester component of variance were obtained for the high and low yielding groups in 22 single cross experiments. The averages of these estimates were essentially the same for the two groups. It was concluded that the data disagreed with Hull's hypothesis. The purposes of the study reported herein were to consider further the role of average gene frequency of the

No abstract

No abstract