Plot size for progeny selection in maize (Zea mays L.).

Six progeny trials that included 147 half-sib progenies of maize (Zea mays L.) population ESALQ PB-5 were conducted for the purpose of studying plot size and its consequences in recurrent selection programs. The progenies were evaluated in three 7x7 duplicate simple lattice experiments using one-row plots of 5 m(2). At harvest each plot was partitioned into five sub-plots (sampling units), and data was collected from each sampling unit. At the same time and place the same progenies were evaluated in three 7x7 duplicate simple lattice experiments using 1-m(2) (linear row with 5 plants) plots. Data were collected for plant and ear height, ear diameter, total ear weight, and total grain yield. The data were combined by using adjacent sampling units, and the analyses were performed by considered five plot sizes in addition to those of the independent trials with 1-m(2) plots. The experiments with 1-m(2) plots were less efficient in discriminating for yield traits among progenies than those with 5-m(2) plots. The combination of plot size and number of progenies evaluated indicated that an optimum plot size for yield was between 3 and 4 m(2), or 15-20 plants per plot. With such sizes the expected gain was maximized for the four replications used in this study. If the total area covered by each progeny is constant, the maximum gain from selection, however, is attained by decreasing plot size and increasing the number of replications. The minimum size of plots is, however, limited by practical or theoretical criteria. Plot size affected the estimates of additive genetic variance, coefficient of heritability, and genetic coefficient of variation for all of the traits. No practical limitation was observed for conducting experiments with 1-m(2) plot.

Procedures for selecting composites based on prediction methods.

Procedures for selecting among parental varieties to be used in the synthesis of composites are discussed. In addition to the criterion based on the mean and variance of composites of the same size (k) proposed by Cordoso (1976), we suggest the index Ij=w1vj+w2 hj or I'j=(2/k) Ij for a preliminary selection among parental varieties. We show that by increasing k (size of the composite) I'j tends to gj, the general combining ability effect. Such a criterion is particularly important when n, the number of parental varieties, is large, so that the number of possible composites (Nc=2(n)-n-1) becomes too large to be handled when using the common prediction procedures. Yield data from a 9 × 9 variety diallel cross were used for illustration.

Predicting the potential of open-pollinating populations for the production of superior F1 hybrids.

The distributive properties of a single population or of a population resulting from a cross between two populations are reproduced when inbreds randomly extracted from the population itself or from the two parental populations are randomly paired. Hence, population parameters that are usually obtained during a breeding programme can be used to predict the performance of the F1 hybrids that can be derived from them at that stage. Multiple allelism, epistasis and deviations from Hardy-Weinberg equilibria should not cause biasis to the predictions. While in theory genotype x environment interaction and linkage disequilibrium may disturb the predictions, in practice they are unlikely to create problems that cannot be accommodated. Genotypic and phenotypic predictions of the proportions of the F1 hybrid distribution scoring above or below a given standard are made and analysed for three characters, weight of the ears, plant height and height of the ear, in two populations of maize per se and their interpopulational cross. Because no random inbred lines from the experimental populations are presently available we cannot check our predictions. However, genotypic and phenotypic predictions and observations of F1 hybrids obtained from populations created by computer simulation are provided to illustrate our procedures.