Estimates Of Dominance Variance Research Articles

Variability in a Silage Maize “Zea mays, L.” Base Population via Two Types of Families 3. Forage botanical composition and dry matter partitioning

The base population for the recent study which is a yellow seed synthetic variety of maize was provided by Dr. M. Abd El-Sattar Ahmed, professor of crop science, Crop Science Department, faculty of Agriculture, Egypt. Two types of families had formulated, as a step before selection for high yielding ability. Those were half-sib (H.S) families and the first generation of self-pollination (S1) families. Variability in the base population gene pool had estimated depending on vari-ance components of the families. Reduction in mean values associated with S1 families relative to half-sib families reached 0.182, 1.18, 0.98, 3.88, 2.75 and 5.72% for leaves percentage, stem percentage, ear percentage, grain percentage, cob per-centage and husk percentage, respectively. Genotypic variances among S1 families was larger in magnitude than the respec-tive value for half-sib families in each measured botanical composition and dry matter partitioning characters. Meanwhile, within family’s variance, (σ2f), environmental variances (σ2e) had reduced in S1 families relative to half-sib families in each measured character. Genotypic variation among half-sib and S1 families were less than 20% of the studied botanical compo-sition and dry matter partitioning characters mean values, except for cob percentage and husk percentage that had a genotyp-ic variation amount to 24% and 26%, respectively. The estimates of additive variance (δ2A) were positive for all studied botanical composition and dry matter partitioning characters. While, the estimates of dominance variance (δ2D) were nega-tive for all characters, the ratio of δ2A/δ2G indicated that additive genetic variance was more effective in controlling all stud-ied botanical composition and dry matter partitioning characters. Estimates of heritability were higher in magnitude when calculated from first selfing generation (S1) compared to half-sib families, for all studied characters. Stem percentage ex-pressed the least estimate heritability among all studied characters in each family type (0.798 and 0.414 for half-sib and S1 families, respectively). This might indicate the influence of environment in the expression of that character. Meanwhile husk percentage had highest genetic control expressed on heritability estimates from the two studied types of families.

SNP-based mate allocation strategies to maximize total genetic value in pigs

BackgroundMate allocation strategies that account for non-additive genetic effects can be used to maximize the overall genetic merit of future offspring. Accounting for dominance effects in genetic evaluations is easier in a genomic context, than in a classical pedigree-based context because the combinations of alleles at loci are known. The objective of our study was two-fold. First, dominance variance components were estimated for age at 100 kg (AGE), backfat depth (BD) at 140 days, and for average piglet weight at birth within litter (APWL). Second, the efficiency of mate allocation strategies that account for dominance and inbreeding depression to maximize the overall genetic merit of future offspring was explored.ResultsGenetic variance components were estimated using genomic models that included inbreeding depression with and without non-additive genetic effects (dominance). Models that included dominance effects did not fit the data better than the genomic additive model. Estimates of dominance variances, expressed as a percentage of additive genetic variance, were 20, 11, and 12% for AGE, BD, and APWL, respectively. Estimates of additive and dominance single nucleotide polymorphism effects were retrieved from the genetic variance component estimates and used to predict the outcome of matings in terms of total genetic and breeding values. Maximizing total genetic values instead of breeding values in matings gave the progeny an average advantage of − 0.79 days, − 0.04 mm, and 11.3 g for AGE, BD and APWL, respectively, but slightly reduced the expected additive genetic gain, e.g. by 1.8% for AGE.ConclusionsGenomic mate allocation accounting for non-additive genetic effects is a feasible and potential strategy to improve the performance of the offspring without dramatically compromising additive genetic gain.

Estimation of dominance variance for live body weight in a crossbred population of pigs.

The objective of this study was to estimate the dominance variance for repeated live BW records in a crossbred population of pigs. Data were provided by the Walloon Pig Breeding Association and included 22,197 BW records of 2,999 crossbred Piétrain × Landrace K+ pigs from 50 to 210 d of age. The BW records were standardized and adjusted to 210 d of age for analysis. Three single-trait random regression animal models were used: Model 1 without parental subclass effect, Model 2 with parental subclasses considered unrelated, and Model 3 with the complete parental dominance relationship matrix. Each model included sex, contemporary group, and heterosis as fixed effects as well as additive genetic, permanent environment, and residual as random effects. Variance components and their SE were estimated using a Gibbs sampling algorithm. Heritability tended to increase with age: from 0.50 to 0.64 for Model 1, from 0.19 to 0.42 for Model 2, and from 0.31 to 0.53 for Model 3. Permanent environmental variance tended to decrease with age and accounted for 29 to 44% of total variance for Model 1, 29 to 37% of total variance for Model 2, and 34 to 51% of total variance for Model 3. Residual variance explained <10% of total variance for the 3 models. Dominance variance was computed as 4 times the estimated parental subclass variance. Dominance variance accounted for 22 to 40% of total variance for Model 2 and 5 to 11% of total variance for Model 3, with a decrease with age for both models. Results showed that dominance effects exist for growth traits in pigs and may be reasonably large. The use of the complete dominance relationship matrix may improve the estimation of additive genetic variances and breeding values. Moreover, a dominance effect could be especially useful in selection programs for individual matings through the use of specific combining ability to maximize growth potential of crossbred progeny.

Genetic Variation and Gains in Resistance of Strawberry to Colletotrichum gloeosporioides

Anthracnose crown rot is an important disease of strawberry primarily caused by Colletotrichum gloeosporioides in Florida and North Carolina. Information on the magnitude of additive and nonadditive genetic variation is required to define breeding strategies and to estimate potential genetic gains. However, little is known about the genetic control of resistance and its utility in breeding. Our objectives were to obtain estimates of heritabilities and of components of genetic variances, genotype-environment interactions, and gains for resistance, and to examine the effects of locations and transplant types on the estimates. An incomplete diallel mating design generated 42 full-sib families, which were propagated in plugs from seed (seedling tests) and as bare-root runner plants (clonal tests) of different genotypes of the same families. Both seedlings and clones were inoculated with C. gloeosporioides under field conditions in North Carolina and Florida during the 2010-11 season. Narrow-sense heritability (h(2)) and broad-sense heritability (H(2)) for both clones and seedlings were higher at the North Carolina location (h(2) = 0.34 to 0.62 and H(2) = 0.46 to 0.85) than at the Florida location (h(2) = 0.16 to 0.22 and H(2) = 0.37 to 0.46). Likewise, the seedling tests showed higher genetic control than the clonal tests at both locations. Estimates of dominance variance were approximately one-third of the additive variance at North Carolina and were even larger at Florida. Epistasis was negative at both locations and assumed zero for heritability (H(2)) calculations. Genotype-environment interactions were different by transplant type, suggesting rank changes across locations. 'Pelican' was the most resistant parent at both locations, followed by 'NCH09-68' at the NC location and 'Winter Dawn' at the Florida location. Selection and deployment of the most resistant clone within each of the five best families is estimated to produce average genetic gains of 53.0 and 73.7% at the North Carolina and Florida locations, respectively.

Combining Ability Studies for Yield and Yield Attributing Traits in Sweet Pepper (Capsicum annuumL.)

Ten diverse lines of sweet pepper were crossed in diallel manner to generate forty five crosses. The crosses and parents were evaluated in randomized complete block design during Kharif, 2010 at three different locations. The observations were recorded on various yield and yield attributing traits. Analysis of variance for combining ability revealed that variances due to GCA and SCA were significant for all the traits suggesting the importance of both types of variances in the inheritance of the traits. GCA and SCA also interacted significantly with the environments. The estimates of dominance variance were higher in magnitude than corresponding estimates of additive variance indicating the preponderance of non-additive gene action. None of the parents exhibited desirable gca effects for all the traits simultaneously. However, overall ranking of genotypes revealed that the parents SP-82, SP-640 and California Wonder exhibited significant desirable GCA effects for most of the traits. None of the crosses exhibited significant desirable SCA effects for all the traits simultaneously. However, the crosses SP-29 x SP-640, California Wonder x SP-82 and SP-82 x SP-614 were found to be desirable for most of the traits.

Genetic analysis for fodder yield and other important traits in oats (Avena sativaL.)

The parents and 45 F1 crosses (excluding reciprocals) of forage oats (Avena sativa L.) were analyzed genetically to generate information on combining ability (general and specific), nature and magnitude of gene effects for forage yield and their attributing traits. The estimates of dominance variance were higher in magnitude than their corresponding additive genetic variance indicating preponderance of non-additive gene action except for days to 50% flowering. None of the parents was found to be a good general combiner simultaneously for all traits. However, SKO-208 was identified as a good general combiner for most of the traits.The most promising crosses that revealed highly positive and significant SCA effect for forage yield were SKO212, SKO213, SKO204 x SKO208, SKO204 x SKO211, SKO210 x SKO211 and SKO207 x SKO210.

Genetic effects of combining ability on forage quality traits and forage yield in diallel crosses of oats (Avena sativaL.)

Studies in oats (Avena sativa L.) were carried out to generate information on combining ability (general and specific), nature and magnitude of gene effects and heterosis for forage quality and yield. The experimental material comprised of 10 diverse parents and 45 F1 crosses (excluding reciprocals) generated through a 10 x 10 diallel mating design. Significant interaction between combining ability effects were observed for all the quality traits and yield. The estimates of additive variance were higher in magnitude than their corresponding dominance variance estimates indicating the role of additive gene action, except green forage yield and dry matter yield. The parent SKO-213 had highly significant and desirable GCA effects for crude protein content, SKO-210 for neutral detergent fibre, SKO-204 for acid detergent fibre, SKO-208 for crude fibre parent, SKO-207 for ash content and SKO-212 for both green forage and dry matter yields. Cross combinations viz., SKO-212 x SKO-213, SKO-204 x SKO-211, SKO-205 x Sabzaar, Sabzaar x SKO-212 and SKO-205 x SKO-211 exhibited high and desirable sca and heterobeltiosis effects simultaneously for all the quality traits, green forage yield and dry matter yield ha−1.

Gene action and combining ability for fodder yield and its attributing traits in oats (Avena sativa L.)

Studies in forage oats (Avena sativa L.) were carried out to generate information on combining ability (general and specific), nature and magnitude of gene effects for forage yield and their attributing traits. The experimental materials comprised of ten diverse parents and 45 F1 crosses (excluding reciprocals) generated through 10 x 10 diallel mating design. The experimental material was evaluated in a randomized complete block design. Observations were recorded on eight quantitative characters. Significant interaction between combining ability effects and environments were observed for all the traits. The estimates of dominance variance were higher in magnitude than their corresponding additive genetic variance indicating preponderance of non-additive gene action except for days to 50% flowering. Estimates of average degree of dominance also revealed preponderance of over dominance for all the traits except for days to 50% flowering that shows partial dominance. None of the parents was found to be a good general combiner simultaneously for all traits. However, SKO-207 and SKO-208 was identified as a good general combiner for most of the traits viz., days to 50% flowering, plant height (cm), number of tillers m-1, culm diameter (cm), leaf stem ratio, dry matter yield plant-1 (g). SKO-212 and SKO-204 was the best general combiner for green fodder yield plant-1. The most promising crosses that revealed highly positive and significant SCA effect for forage yield SKO-212 X SKO- 213, SKO-204 X SKO- 208, SKO-204 X SKO- 211, SKO-210 X SKO- 211 and SKO-207 X SKO- 210. Key words: Fodder oats, combining ability, GCA, SCA.

Estimation of the additive and dominance variances in SA Duroc pigs

The objective of this study was to estimate dominance variance for number born alive (NBA), 21-day litter weight (LWT21) and interval between parities (FI) in South African Duroc pigs. A total of 10,703 NBA, 6883 LWT 21 and 6881 FI records were analysed. Bayesian analysis via Gibbs sampling was used to estimate variance components and genetic parameters. Estimates of additive genetic variance were 0.554, 16.84 and 4.535 for NBA, FI and LWT21, respectively. Corresponding estimates of dominance variance were 0.246, 9.572 and 0.661 respectively. Dominance effects were statistically not significant for all traits studied. Further research utilizing a larger data set is necessary to make concrete conclusions on the importance of dominance genetic effects for the traits studied.

Genetic Variability for Grain Yield and Its Components in an Early Tropical Yellow Maize Population Under Striga hermonthica Infestation

A major problem that usually confronts the breeder is ensuring that there is sufficient genetic variability in a population for the trait being improved because this determines the breeding schemes to use for improvement and whether or not sufficient progress could be made. Three hundred full-sib families from cycle 3 of the early maturing yellow endosperm maize (Zea mays L.) population, TZE-Y Pop DT STR, were evaluated under artificial Striga hermonthica infestation at Mokwa and Abuja, Nigeria in 2003. The objectives of this study were to investigate the type of gene action involved in the inheritance of S. hermonthica resistance in TZE-Y Pop DT STR C3, determine the extent of genetic variability available in the population after three cycles of S1 recurrent selection, and the phenotypic and genetic correlation coefficients among the traits used for selecting for resistance to S. hermonthica. Estimates of dominance variances were larger than additive genetic variances for grain yield, plant height, ear height, number of ears at harvest, and Striga damage rating at 8 weeks after planting (WAP). Even though h2 estimates were generally low for most traits (< 0.40), moderate-to-large additive genetic variances, and wide ranges were obtained in TZE-Y Pop DT STR C3 for most traits, suggesting that there was adequate genetic variation for improving Striga resistance and grain yield in the population. Highly significant phenotypic correlation coefficients were obtained between grain yield and ears per plant, plant height, ear height, days to anthesis and silking, anthesissilking interval, and Striga damage score at 10 WAP. Recurrent selection methods that capitalize on both additive and dominance variances would be effective for improving the population for Striga resistance and grain yield.

Within-family genetic variation and epistasis in cassava (Manihot esculenta Crantz) adapted to the acid-soils environment

Little is known about the inheritance of agronomic traits in cassava, or on the relative importance of epistasis for most crops. A group of 10 clones, adapted to the acid-soils environment was used as parents in a diallel study. Thirty genotypes were obtained from each F1 cross and cloned. Each clone was represented by six plants, which were distributed in three replications at two contrasting locations. Genetic variability concentrated in the within-family component, which was statistically significant for all the variables analyzed (fresh root yield (FRY), fresh foliage yield (FFY), harvest index, root dry matter content, and plant type score (PTS)) except for the reaction to super elongation disease (SED). Estimates of dominance variance were considerably larger than those of additive variance for fresh root and foliage yields. The reverse was observed for harvest index, dry matter content, PTS and SED score. Epistasis played an important role only for fresh root and foliage productions. These results agree with those from similar studies targeting different environments. The common assumption of absence of significant epistatic effects frequent in many quantitative genetic designs is, therefore, challenged from the results for these two variables. Alternative breeding approaches are suggested according to the results obtained from this study.

Stichprobenumfang zur Schätzung der Dominanzvarianz bei hierarchischer Halb- und Vollgeschwisterstruktur

Abstract. Title of the paper: Sample size for estimation of dominance variance with hierarchical full- and half-sib structure The sampling size for the estimation of the ratio d2 of the dominance variance to the phenotypic variance is investigated for the case of hierarchical population structures corresponding to the relations in poultry and pig breeding. The standard error of the d2-estimation is calculated by using the Maximum-Likelihood method and the concept of parameter transformation. Estimation of the dominance variance requires much larger data sets than estimation of the additive variance. Assuming d2-values of 25, 17, 10 and 5% than the estimation of the ratio d2 required about 6, 10, 20 and 40 times more data than the estimation of the additive variance. For the case h2:d2=0.4:0.1 the estimation of d2 with 8320, 14720, 33040 and 58640 records yielded standard errors of 8, 6, 4 and 3%. In the situation of h2:d2=0.225:0.075 the same accuracies can be reached with records of 4400, 7760, 17520 and 31040 animals.

Estimation of dominance variance in purebred Yorkshire swine.

We used 179,485 Yorkshire reproductive and 239,354 Yorkshire growth records to estimate additive and dominance variances by Method Fraktur R. Estimates were obtained for number born alive (NBA), 21-d litter weight (LWT), days to 104.5 kg (DAYS), and backfat at 104.5 kg (BF). The single-trait models for NBA and LWT included the fixed effects of contemporary group and regression on inbreeding percentage and the random effects mate within contemporary group, animal permanent environment, animal additive, and parental dominance. The single-trait models for DAYS and BF included the fixed effects of contemporary group, sex, and regression on inbreeding percentage and the random effects litter of birth, dam permanent environment, animal additive, and parental dominance. Final estimates were obtained from six samples for each trait. Regression coefficients for 10% inbreeding were found to be -.23 for NBA, -.52 kg for LWT, 2.1 d for DAYS, and 0 mm for BF. Estimates of additive and dominance variances expressed as a percentage of phenotypic variances were, respectively, 8.8 +/- .5 and 2.2 +/- .7 for NBA, 8.1 +/- 1.1 and 6.3 +/- .9 for LWT, 33.2 +/- .4 and 10.3 +/- 1.5 for DAYS, and 43.6 +/- .9 and 4.8 +/- .7 for BF. The ratio of dominance to additive variances ranged from .78 to .11.

Influence of dominance relationships on the estimation of dominance variance with sire-dam subclass effects.

Two data sets from the USDA Livestock and Range Research Laboratory were analyzed to study dominance variance and the influence of dominance relationships. The first consisted of 4,155 birth weight (3,884 weaning weight) records of inbred USDA Line 1 Herefords. The second consisted of 8,065 birth weight (7,380 weaning weight) records from a line-cross experiment with five lines. Two models were used. Both included fixed effects of year-sex of calf and age of dam, and covariates for calving date, inbreeding of animal, and inbreeding of dam. For the second set, additional covariates were line composition and heterozygosity coefficients. Random effects were direct and maternal additive genetic, maternal permanent environment, sire-dam subclass, and residual. Model 1 considered sire-dam subclasses unrelated. Model 2 related sire-dam subclasses with a parental dominance relationship matrix. Variance components were estimated using REML. Differences between estimates with Model 1 and 2 were unimportant except for dominance variance. For the first data set, estimates with Model 2 of relative genetic direct and maternal variances, direct-maternal correlation, permanent environment, and dominance variances for birth weight were .35, .13, -.02, .03, and .25, respectively, and they were .39, .11, .04, .06 and .14 for the second data set. For weaning weight, the first data set estimates were .20, .15, -.37, .19, and .11, respectively, and they were .16, .20, -.07, .18, and .18 for the second data set. Changes, decreases and increases, in estimates of dominance variances may be due to increased information from relationships and family types other than full-sibs. The assumption of unrelated sire-dam subclasses might not be appropriate for estimation of dominance variance in populations with many dominance relationships among sire-dam classes.

A MARKER-BASED METHOD FOR INFERENCES ABOUT QUANTITATIVE INHERITANCE IN NATURAL POPULATIONS.

A marker-based method for studying quantitative genetic characters in natural populations is presented and evaluated. The method involves regressing quantitative trait similarity on marker-estimated relatedness between individuals. A procedure is first given for estimating the narrow sense heritability and additive genetic correlations among traits, incorporating shared environments. Estimation of the actual variance of relatedness is required for heritability, but not for genetic correlations. The approach is then extended to include isolation by distance of environments, dominance, and shared levels of inbreeding. Investigations of statistical properties show that good estimates do not require great marker polymorphism, but rather require significant variation of actual relatedness; optimal allocation generally favors sampling many individuals at the expense of assaying fewer marker loci; when relatedness declines with physical distance, it is optimal to restrict comparisons to within a certain distance; the power to estimate shared environments and inbreeding effects is reasonable, but estimates of dominance variance may be difficult under certain patterns of relationship; and any linkage of markers to quantitative trait loci does not cause significant problems. This marker-based method makes possible studies with long-lived organisms or with organisms difficult to culture, and opens the possibility that quantitative trait expression in natural environments can be analyzed in an unmanipulative way.

Estimates of Genetic Variances in Two Spanish Populations of Maize

Reports of estimates of components of genetic variance in open pollinated populations of maize (Zea mays L.) have been mostly on U.S. Corn Belt germplasm, whereas few studies have been conducted using European germplasm. This study was undertaken to obtain estimates of genetic parameters for important agronomic traits in two populations (Ribadumia and Tuy) from northwestern Spain. Full‐sib within halfosib groups were developed by the Design I mating system and evaluated at two locations in Spain in 1987 and 1988. The populations showed a similar genetic structure with very small differences. Moreover, the pattern of genetic variability was similar to other germplasm because the greatest proportion of total genetic variance was attributed to additive effects for most traits. In contrast, estimates of dominance variance were large for yield, ear, and plant height and ear length. In general, additive × environment interaction variances were significant while dominance × environment interaction variances were not significant for either population. The large range of variation and the high values of heritability for lodging and ear and kernels traits, as well as values of expected gain from selection for these traits, suggested good potential for improvement through an intrapopulation selection program. However, the large dominance effects and the small additive effects for grain yield indicated that an interpopulation selection program would be effective if the large intrapopulation dominance variance is expressed in population crosses. Yield did not show significant additive variance so estimates of additive correlation with any trait were not calculated. Correlated response between yield and any trait is not expected and improved grain yield based on indirect selection of yield components is not recommended.

A comparison of genetic information from open-pollinated and control-pollinated progeny tests in two eucalypt species

Genetic-parameter estimates and parental breeding-value predictions were compared from open-pollinated and control-pollinated progeny populations of Eucalyptus globulus and two populations of E. nitens. For E. globulus there were two types of open-pollinated populations (native stand open-pollinated and seed orchard open-pollinated) and two types of control-pollinated populations (intra-provenance and interprovenance full-sib families). For E. nitens there were two populations, a seed orchard open-pollinated population and intra-provenance full-sib families. Progeny tests were established across multiple sites and 2-year height and diameter were measured and volume calculated. Genetic parameters from native stand open-pollinated E. globulus were unlike the parameters from the other three E. globulus populations; heritability estimates were severely inflated, presumably due to high levels, and possibly differential levels, of inbreeding depression relative to the other populations. Estimates of dominance variance in the E. globulus full-sib populations were high, but were zero in the E. nitens population. Correlations among parental breeding values, predicted using data from the different populations, were generally low and non-significant, with two exceptions: predictions from the two E. globulus full-sib populations were significantly correlated (r=0.54, P = 0.001), as were predictions from the E. nitens seed orchard OP and full-sib population (r = 0.61, P = 0.08). There was some indication that superior parents of E. globulus native stand open-pollinated families also tended to have above-average breeding values based on the performance of intra-provenance full-sib offspring. The consequences of these results for exploitation of base-population collections from native stands are discussed.

Production Traits of Holstein Cattle: Estimation of Nonadditive Genetic Variance Components and Inbreeding Depression

Additive, dominance, and additive by additive components of genetic variance and inbreeding depression were estimated for production traits from a group of daughters of young sires from the Canadian Holstein population. First lactations of 92,838 cows were analyzed. Three sire and dam models (additive, additive plus dominance, additive plus dominance plus additive by additive genetic effects), all including regression of the trait on inbreeding coefficient of the cow, were used to estimate the effect of inbreeding on production traits. For all production traits, heritability in the narrow sense was overestimated with the simplest model, in which only the additive effect was fitted. Estimates of dominance variance were low for all traits, .9 to 3%. Additive by additive components were low for milk, 2.8%, and fat yield, 2.8%. but higher for protein yield, 6.8%, and for fat, 9%. and protein percentages, 8.9%. Estimates of inbreeding depression for the five traits were similar across all models (–25, –.9, and –.8kg; .05% and .05% per 1% increase in inbreeding for milk, fat, and protein production and fat and protein percentages, respectively). More accurate estimates of additive effects might be obtained with the inclusion of nonadditive effects for genetic evaluation. If the estimation of inbreeding depression is the only objective, simple models and small random samples of the population may be adequate.

Androgenic Somaclonal Variation in Tobacco and Estimation of its Value as a Source of Novel Genetic Variability1

In vitro anther culture of near homozygous tobacco (Nicotiana tabacum L.) cultivars yields haploids which, when chromosome doubled, result in populations of doubled‐haploid (DH) lines exhibiting great among‐line genetic variability with agronomic characteristics inferior to those of the cultivars. Studies were conducted to characterize the nature of this “induced” genetic variability and to determine its potential value in recurrent selection programs. Sixty‐four DH lines from a single plant of cultivar ‘NC 95’ were randomly assigned to eight sets of a North Carolina Design II mating design. The DH full‐sib families and selfed progenies of the parental plant were evaluated for seven variables in two field environments. Genetic variances in the DH population were estimated, heritabilities calculated, and genetic gains from practicing half‐sib and full‐sib family selection for yield improvement in the DH population were predicted. A DH population composite, a random‐mated DH population, and populations resulting from practicing half‐sib and full‐sib family selection, in the DH population were compared to two sublines of the original parental plant. Significant additive genetic variance was detected for six of seven variables measured in the DH population. Dominance variance estimates were zero or negligible. Random mating of DH lines resulted in a recovery of 26% of the yield loss of the DH population composite compared to the cultivar. One cycle of half‐sib and full‐sib family selection resulted in populations yielding 5% and 6.0% respectively, more cured leaf than the random‐mated DH population; both selected populations were lower yielding than the sublines of the original parental plant. Two or more cycles of recurrent full‐sib selection are estimated to be required to achieve a population equal in yielding ability to the parental cultivar. This requirement reduces the appeal of androgenic somaclonal variability as a germplasm source for tobacco improvement.

Diallel designs, analyses, and reference populations.

Some issues which evidently remain as areas of contention in diallel analysis are reviewed. In the estimation of genetic variance components for an ancestral reference population in equilibrium, analysis should be applied to F1 data only, and proceed as for a random effects model. No meaningful reference population is constituted by the parental sample itself, but an equilibrium reference population which could be derived from these parents can be defined. To give unbiassed estimates of parameters for this population, analysis must include S1 data, and the parents must be homozygous and in linkage equilibrium. Estimation of dominance variance must allow for the fact that the diallel population itself is not in linkage equilibrium by correcting the specific combining ability sum of squares for the mean S1 vs. F1 difference. Some different analyses of variance of the diallel table are discussed in the context of the above reference populations.