Search

Recurrent selection for intrapopulation improvement in maize (Zea maysL.) using combined HS‐S1progeny evaluation was studied. The source population (3‐L Composite) originated from three inbred lines (W64A, W182E, A635). The F2generations of the three possible single crosses were planted in an isolated plot. At flowering time, ear shoots were covered on prolific plants. The lower ear was self‐pollinated to provide Si seed and the upper ear exposed to open‐pollination to provide half‐sib (HS) seed. Half‐sib and S1progenies were evaluated in separate trials using incomplete blocks of 10 entries each with two replications. Selection was carried out on a per block basis. Plants with superior HS and S1progeny performance, i.e., above average yield and equal or below average moisture in the grain at harvest in both trials, were selected. Selection intensity was 22 and 20% in CO and Cl, respectively. Each new cycle was initiated from remnant HS seed. Two selection cycles have been completed. The estimated degree of inbreeding was 34% in CO, 43% in Cl, and 44% in C2. Relative to CO, selection resulted in 5% yield improvement in Cl (based on 3‐year testing) and 20% improvement for C2 (2‐year tests). Also, Cl was 1 day earlier to flower, taller with slightly higher (103%) ear placement, and had less moisture in the grain at harvest (97%). C2 exhibited no further change for moisture in grain at harvest but was taller, ear height being 111% relative to CO. Thus far, selection has reduced inbreeding depression in the population without decreasing heterosis for yield in HS families.

In two different experiments, three prolific and three nonproliflc single‐cross hybrids of maize (Zea maysL.) growth were treated with growing regulating chemicals to determine their effects on ear development. In experiment A, the compounds used were N‐l‐naphthylphthalamic acid (NPA), p‐chlorophenoxyisobutyric acid (PCIB), and 6(γ,γ‐dimethylalllyamino) purine (i6Ade), all at concentration of 10‐4M. Treatments were applied as 0.5‐ml injections to the internode between the first and second ear shoots of nonprolific hybrids and between the second and third ear shoots of prolific hybrids. The auxin translocation inhibitor (NPA) significantly increased total yield of nonprolific hybrids and increased lower ear grain weight and ear number of all hybrids. The antiauxin, PCIB, had no discernible effect, while i6Ade (cytokinin) treatments decreased total yield of nonprolific hybrids and the yield of lower ears of all hybrids. In experiment B, upper ear shoots were excised on the day of silking and treatments of indolebutyric acid (IBA), gibberellic acid (GA3), and cytoklnin (i6Ade) were injected into the remaining ear shank. Gibberellic acid and combinations of IBA and GA3treatments sgnificantly reduced ear number. The results of these experiments indicate that indoleacetic acid (IAA), GA3, and possibly other hormones interact in a time‐dependent mode to inhibit lower ear development. The close relationship between sink size and total yield demonstrates the importance of an adequate sink for plants grown at low density.

The inheritance of prolificacy and the relationship of multiple ear development to time intervals between developmental stages of reproductive structures were studied in maize (Zea maysL.) in three different experiments using both quantitative and qualitative methods of genetic analysis. A six‐line diallel involving unrelated parents, F1's, and F2's were grown at population densities of 17,200 and 34,400 plants/ha over 2 years. General (GCA) and specific (SCA) combining ability effects were significant for both ear number and multiple‐ear‐weight index (MEWI). This index is defined as the total grain weight minus first ear grain weight divided by total grain weight. The ratio of GCA to SCA fixed‐effects variance ranged from 3.7:1 to 7.4:1 for ear number and from 2.5:1 to 7.2:1 for MEWI. Higher ratios were usually associated with the low population density. Phenotypic correlations and the correlations of GCA effects were strongly positive for the time intervals, first‐to‐second ear shoot emergence, and first‐to‐second silk. Also, silking interval was negatively correlated with both ear number and MEWI. In a second experiment, six unrelated parents and 15 F1's were dissected daily and the dates of initiation of tassel and ear shoot meristems recorded. General combining ability effects were highly significant for the time periods tassel‐to‐first ear shoot initiation and first‐to‐second ear shoot initiation. Only the time period between initiation of the first and second ear shoot was correlated with ear number (−0.57**) and MEWI (−0.63**). In an experiment involving related inbred lines, their F1's, F2's, and backcrosses, additive gene effects were highly significant for ear number and MEWI, and dominance effects were significant (0.05 level) for MEWI. Negative correlations were observed between both ear number and MEWI and the time periods, ear shoot emergence‐to‐pollen‐shed, first‐to‐second ear shoot emergence, and first‐to‐second silk. For the F1and segregating generations of one pair of related lines, a greater frequency of second ears occurred on plants having fewer nodes above the first ear. In both experiments prolificacy factors segregated in a manner predictable from the penetrance and genetic segregation inferred from parents and F1's. The genetic analysis on two pairs of related lines indicated that a single locus or group of linked loci account for the majority of the variation in ear number an their segregating generations.

A combined half‐sib (HS) and Slfamily evaluation procedure (HS‐S1) in a maize (Zea maysL.) population was compared with HS and S1per se evaluation and with a selection index of the form I = b1Xl+ b2X2where X1and X2are grain yields of the HS and S1families, respectively. The combined HS‐S1evaluation resulted in lower selection differentials than either HS or S1evaluation. Combined HS‐S1selection in the CO (original population) identified 81, 71, and 77% of the selected families in common with HS, S1and a selection index, respectively, and 60, 58, and 85% in CI. HS‐S1resulted in selection of genotypes better buffered to environmental differences than either HS or $1 alone. The evidence indicates that HS evaluation results in selection of genotypes with above average population heterotic response and greater than average inbreeding depression while S1evaluation has the opposite effect. Selection on combined HS‐S1evaluation resulted in genotypes with less than average inbreeding depression and average or better heterotic response. The selection index has a similar effect. Expected gains showed combined HS‐S1with resynthesis using S1progenies to be equal to or greater than other procedures.

Five open‐pollinated varieties each representative of a different racial group and 35 other populations developed from them (the five varieties selfed, 10 F1crosses, and the random mated and selfed progenies of the F1crosses), were grown in three environments in Mexico. Data were taken on grain yield, days to flower, plant height and number of ears per 10 plants. The phenotypic means of the 40 populations for each trait were analysed separately by analysis of variance technique, and the variance among populations was subdivided according to the model and analysis I outlined by Gardner and Eberhart (1966). In the case of yield, the greatest contributing factor to genetic variation was intravariety dominance effects (52.4% of the total). Additive gene effects contributed most to variation among means in days to flower (87.9%), plant height (79.3%) and ear number (60.9%). Average heterosis in the crosses explained 5.1% of the total variation in yield means whereas varietal and specific heterosis effects accounted for only a minor portion of the variation and were nonsignificant statistically. In the other traits, “the variation accounted for by heterosis effects was, in all cases, nonsignificant. Estimates of the gene effects were used to compute predicted means of the entries in the experiment, the advanced generations of crosses of three and four races and more precise estimates of heterosis. The highest yielding predicted crosses in advanced generations are suggested as basic genetic material to be used in a mass selection program prior to initiating any breeding procedure involving inbreeding. The data obtained are considered to fit the genetic model used. No evidence was found to indicate that epistasis contributes to heterosis in the traits studied even though genetically diverse varieties of maize were crossed. The possible presence of undetectable types of epistasis and the stability of the estimated gene effects in different environments are discussed.

A nested Design I consisting of full‐sib and half‐sib families was used to study changes in genetic variances resulting from two selection systems in the ‘Krug’ yellow dent variety of maize (Zea maysL.). Two populations derived from Krug on the basis of earlier Design I experiments were used. K (S1) syn‐2 was developed by intercrossing S1progenies of the male parent of selected half‐sib families. The second population, K (FS) syn‐2 was developed by intercrossing remnant seed of selected full‐sib progenies. Both selected populations, K (FS) syn‐2 and K (S1) syn‐2, were significantly higher yielding than the original variety, Krug. Design I matings were made within each of the two selected populations to provide evidence on changes in genetic variances that had occurred from the selection practiced. Selection resulted in a moderate decrease in genetic variance, with S1progeny and full‐sib selection producing, apparently, different changes in magnitude of genetic variances

Intra‐population genetic variances were obtained using design I matings in high and low selected S1lines (F2populations) representing extremes from a total of 169 KrugIIIsynthetic lines on the basis of evaluation as a) S1per se, b) testcrosses with the parental population and c) testcrosses with an unrelated population. The experiments were grown over a 2‐year period. There appeared to be more additive than non‐additive genetic variance for yield within lines selected as high yielding in testcross performance whereas the reverse was true in those selected as high yielding on the basis of line per seperformance. Low selected S1lines exhibited generally, more nonadditive than additive genetic variance. Standard errors were of such magnitude as to preclude establishment of clear cut differences among variance components. Performance trials of the derived synthetics failed to show superiority for any of the three line evaluation procedures.

The yield and plant and ear height data indicated that the relative importance of non‐additive sources of variation was quite small when compared to the variation explainable in terms of the additive model of gene action. Although the data preponderantly favored partial dominance, some indications of overdominance were obtained.

Four cycles of modified ear‐to‐row selection have been applied to the open‐pollinated variety, ‘Hays Golden.’ The objective of this research was to evaluate the effectiveness of combined selection for yield of grain. The method entails selection among and within half‐sib families. The family selection was based on the average performance in three replications, one at each of three locations. The selection of individuals within selected families was based upon phenotypic performance in a natural crossing block isolation. The mean yield increased an average of 9.44% per cycle relative to the parental variety. The predicted gain from selection based upon estimates of genetic and geneticenviromnental variances in the original Hays Golden variety was in good agreement with the observed response for grain yield. A comparison with results published earlier using mass selection in the same variety indicates that the modified ear‐to‐row procedure is more effective than mass selection alone in improving yield in the Hays Golden varietal population. Modified ear‐to‐row selection may be used to obtain a more rapid increase in the frequency of favorable genes prior to the extraction of inbred lines. The method can be used to advantage after some reduction in the additive genetic variance has resulted as a consequence of effective mass selection and a more precise measure of the genotypic effect is needed.

Lines selected for specific combining ability were found to represent a random sample with respect to general combining ability when compared with an unselected group of lines from the same original source using a broad gene base tester. The data support the common practice of testing new lines first for general combining ability, followed by tests for specific combining ability. Population improvement was greater when selection was based on a broad gene base tester.

Yields of 4 second cycle synthetics averaged 96% of hybrid U.S. 13 as check compared with 82% for the first cycle populations. Moisture at harvest averaged 105% and 103% of check for the second and first cycle synthetics. This improved performance indicates that the method might benefit areas where production of corn hybrids is not yet feasible, or where a stop‐gap procedure is needed during development of suitable hybrids.