Genetics of grains per spike in durum wheat under normal and late planting conditions

Parental, F₁, F₂, BC₁, BC₂, BC₁₁, BC₁₂, BC₂₁, BC₂₂, BC₁ self and BC₂ selfed generations of three crosses involving six cultivars of durum wheat (Triticum durum Desf.) were studied for grains per spike under normal and late sown environments to analyze the nature of gene effects. A 10-parameter model did not fully account for the differences among the generation means. In two cases more complex interactions or linkage were involved in the inheritance of grains per spike in durum. Both digenic and trigenic epistatic interactions had a role in controlling the inheritance of grains per spike, however, trigenic interactions contributed more than digenic interactions. Non-fixable gene effects were many times higher than fixable ones in all three crosses and in both sowing environments indicating a major role of non-additive gene effects in the inheritance of this trait. Duplicate epistasis between sets of three genes under both environments was recorded for the cross Raj 911 × DWL 5002. Epistatic interactions, particularly the trigenic ones, contributed the maximum significant heterosis. Epistatic interactions involving dominance in the F₂ generations caused significant inbreeding depression. Selective diallel mating and/or biparental mating could be used for amelioration of grains per spike in durum wheat.     

Genetic studies for seed size and grain yield traits in kabuli chickpea

Seed size, determined by 100-seed weight, is an important yield component and trade value trait in kabuli chickpea. In the present investigation, the small seeded kabuli genotype ICC 16644 was crossed with four genotypes (JGK 2, KAK 2, KRIPA and ICC 17109) and F₁, F₂ and F₃ populations were developed to study the gene action involved in seed size and other yield attributing traits. Scaling test and joint scaling test revealed the presence of epistasis for days to first flower, days to maturity, plant height, number of pods per plant, number of seeds per plant, number of seeds per pod, biological yield per plant, grain yield per plant and 100-seed weight. Additive, additive?×?additive and dominance?×?dominance effects were found to govern days to first flower. Days to maturity and plant height were under the control of both the main as well as interaction effects. Number of seeds per pod was predominantly under the control of additive and additive?×?additive effects. For grain yield per plant, additive and dominance?×?dominance effects were significant in the cross ICC 16644?×?KAK 2, whereas, additive?×?additive effects were important in the cross ICC 16644?×?JGK 2. Additive, dominance and epistatic effects influenced seed size. The study emphasized the existence of duplicate epistasis for most of the traits. To explore both additive and non-additive gene actions for phenological traits and yield traits, selection in later generations would be more effective.     

Genetics of grain yield and other agronomic characters in t'ef (Eragrostis tef Zucc. Trotter). I. Generation means and variances analysis

Quantitative genetics of grain yield and other agronomic characters of t'ef (Eragrostis tef) were studied using the F₁, F₂, BC₁, and BC₂ of the cross Fesho × Kay Murri. The study was carried out to estimate gene effects controlling the inheritance of grain yield and related agronomic characters. Significant additive [d] and dominance × dominance [l] interaction effects were detected for grain yield. The variations of yield per panicle and panicle weight were explained in terms of [d], dominance [h], and additive × additive [i] interactions. Non-allelic gene interactions were also detected for kernel weight, harvest index, tiller number, plant height, days to heading and days to maturity. The simple additive-dominance model explained the variation for panicle length, culm diameter and plant weight, allowing unbiased estimates of additive (D) and dominance (H) variance components. Large dominance variances (H) were estimated for grain yield, yield per panicle, and panicle weight. The additive variances for plant height, panicle length, days to heading and days to maturity were higher than the respective dominance variances. High narrow-sense heritability (h²) values (> 0.50) were estimated for plant height, panicle length, days to heading and days to maturity. The lowest h² (0.09) was obtained for kernel weight for which there was little variability. Since grain yield and several important agronomic characters of t'ef are influenced by non-allelic gene interaction, it is advisable to delay selection for yield to later generations with increased homozygosity. This revised version was published online in July 2006 with corrections to the Cover Date.     

Pattern of genetic inheritance of morphological and agronomic traits of sorghum associated with resistance to sorghum shoot fly, <Emphasis Type="Italic">Atherigona soccata</Emphasis>

Sorghum shoot fly, Atherigona soccata is an important pest of sorghum during the seedling stage, which influences both fodder and grain yield. To understand the nature of inheritance of shoot fly resistance in sorghum, we performed generation mean analysis using two crosses IS 18551 × Swarna and M 35-1 × ICSV 700 during the 2013–2014 cropping seasons. The F₁, F₂, BC₁ and BC₂ progenies, along with the parental lines were evaluated for agronomic and morphological traits associated with resistance/susceptibility to sorghum shoot fly, A. soccata. The cross IS 18551 × Swarna exhibited significant differences between the parents for shoot fly deadhearts (%) in the postrainy season. The progenies of this cross exhibited lower shoot fly damage, suggesting that at least one of the parents should have genes for resistance to develop shoot fly-resistant hybrids. Leaf glossiness, leafsheath pigmentation and plant vigor score during the seedling stage exhibited non-allelic gene interactions with dominant gene action, whereas 100 seed weight showed both additive and dominant gene interactions. Presence of awns showed recessive nature of the awned gene. Generation mean analysis suggested that both additive and dominance gene effects were important for most of the traits evaluated in this study, but dominance had a more pronounced effect.     

Inheritance of β-carotene concentration in durum wheat (Triticum turgidum L. ssp. durum)

Despite being one of the important characteristics in determining pasta quality in durum wheat (Triticum turgidum ssp. durum), there is no direct report on inheritance of β-carotene concentration. The objectives of this study were to determine the inheritance of β-carotene concentration and the number of genes involved in six crosses of durum. For the cross PDW-233 (P₁) × Bhalegaon-4 (P₂), F₁, F₂, BCP₁ and BCP₂ populations were developed. For all other crosses, only the F₁ and F₂ populations were developed. β-carotene concentration was determined for all populations and parents of each cross grown at Hol, Maharastra, India. The cross PDW-233 × Bhalegaon-4 was also evaluated at Dharwad, Karnataka, India. Low β-carotene concentration was partially dominant in most of the crosses. Broad sense heritability was 67 and 91% at Dharwad and Hol, respectively, for the cross PDW-233 × Bhalegaon-4 and varied from 74 to 93% for the other five crosses indicating the presence of additive gene effects. The frequency distributions of the trait in the F₂ populations were not normal and were skewed towards the lower parent. Segregation of β-carotene concentration in the six F₂ populations indicated that at least two major genes and two or three minor genes with modifying effects govern the trait. Analysis of variance indicated that environment had comparatively little influence on the trait and this should allow for easy selection. The joint scaling test revealed additive × additive, additive × dominance and dominance × dominance epistatic interactions in the cross PDW-233 × Bhalegaon-4. These authors contributed equally.     

Estimation of additive, dominance and digenic epistatic interaction effects for certain yield characters in Vigna sesquipedalis Fruw

Inheritance of yield and yield contributing characters were investigated using generation mean analysis, utilising the means of six basic populations viz., P₁, P₂, F₁, F₂, BC₁P₁ and BC₁P₂ in four crosses of Vigna sesquipedalis. The analysis reiterated that the importance of dominance (h) gene effects for pod yield/plant and pods/plant as compared to additive (d) gene effects. However, significant and positive additive effects were noticed for pod yield/plant, pods/plant, pod weight and seed weight in different crosses. The three types of gene interactions (additive, dominance and epistasis) were significantly involved for pods/plant in cross KU 7 ×KU 8. Among the digenic epistatic interactions, both additive ×additive (i) and dominance × dominance (l) contributed more for pod yield/plant and pods/plant, however, it varied among the crosses. Populations having earliness can be developed as indicated by reducing dominance effects. Pedigree selection and heterosis breeding is suggested to exploit the fixable and non fixable components of variation respectively in Vigna sesquipedalis. This revised version was published online in July 2006 with corrections to the Cover Date.     

Generation mean analysis of phosphorus‐use efficiency in freely nodulating soybean crosses grown in low‐phosphorus soil

Freely nodulating soybean genotypes vary in their phosphorus (P) uptake and P‐use efficiency (PUE) in low‐P soils. Understanding the genetic basis of these genotypes’ performance is essential for effective breeding. To study the inheritance of PUE, we conducted crosses using two high‐PUE genotypes, two moderate‐PUE genotypes and two inefficient‐PUE genotypes, and obtained F₁, F_2, BC₁ and BC₂ populations. The inheritance of PUE was evaluated using a randomized complete block design. A generation mean analysis of phenotypic data showed that PUE was heritable, with complex inheritance patterns and the presence of additive, dominance and epistatic gene effects. Seed P, shoot P, root P, P‐incorporation efficiency and PUE were largely quantitatively inherited traits. There were dominance, additive × additive and dominance × dominance gene effects on PUE, grain yield, shoot dry weight, 100‐seed weight, root dry weight and shoot dry matter per unit P for populations grown under low‐P conditions. Dominance effects were generally greater than additive effects on PUE‐related indices. These PUE indices can be used to select P‐efficient soybean genotypes from segregating populations.     

The inheritance of quantitative traits in Brassica napus ssp. rapifera (swedes): Augmented triple test cross analyses of production characters

Two F₂ triple test crosses, augmented with F₃s, produced from crosses between different inbred lines of swedes (Brassica napus ssp.rapifera L.) were assessed in field trials at Dundee in 1988 and 1989,respectively. This paper reports the analyses of resistance to powdery mildew, neck length, growth cracks, sugar content and hardness; analyses of yield have been published previously. Additive genetical variation was found for all traits while non-additive variation was less important, the highest degree of dominance being 0.44 for hardness. There was evidence of additive × dominance and dominance × dominance epistasis for mildew and additive × additive epistasis for neck length and hardness. Significant,consistent reciprocal differences were found and these were particularly large for neck length and growth cracks. Sugar determination was carried out on the basic generations of the second cross, the parental lines of which showed large differences in concentration of glucose, fructose and sucrose. Directional dominance was found for high glucose but not for fructose or sucrose. The implications of these results for swede breeding are discussed and it is proposed that inbred cultivars would be a more practical option than F₁ hybrids. This revised version was published online in August 2006 with corrections to the Cover Date.     

Genetic analysis of quantitative traits across environments in linseed (Linum usitatissimum L.)

Summary Fourteen lines of linseed (Linum usitatissimum L.) were evaluated for general and specific combining ability through line × tester cross analysis using five diverse testers in two different environments. Seventy F₁'s and nineteen parental genotypes were raised in completely randomized block design with two replications with respect to yield and yield related traits. Considerable genetic variation was observed for all the traits studied. The additive as well as non-additive gene effects played significant role in the inheritance of yield and related traits with preponderance of non-additive gene effects for all the traits studied. Higher proportion of general combining ability × environment interaction variance as compared to specific combining ability × environment estimates was recorded. Additive genetic variances were more sensitive than non-additive genetic variances to the changing environment. Low estimates of heritability (narrow sense) for primary branches per plant, capsules per plant, seeds per capsule, seed yield per plant, harvest index and oil content and medium heritability for days to 50% flowering, days to maturity, plant height and 1000-seed weight was observed. Among the female parents LCK-9816, Parvati, Himalini, KL-168 were good general combiners for yield and related traits along with oil content, whereas among male parents, Surbhi and KL-224 were good general combiners for yield related traits and oil content but KL-221 was good general combiner for seed yield and related traits.     

Genetic analysis of prolificacy in “Sikkim Primitive”—A prolific maize (Zea mays) landrace of North‐Eastern Himalaya

Prolificacy assumes significance for development of high‐yielding baby corn hybrids. “Sikkim Primitive” is a native landrace of North‐Eastern Himalaya, and is the highest prolific maize germplasm. So far, the genetics of prolificacy in “Sikkim Primitive” has not been deciphered. Here, a prolific inbred (MGU‐SP‐101) developed from “Sikkim Primitive” was crossed with four non‐prolific inbreds viz., LM13, BML7, HKI161 and HKI1128. Six generations (P₁, P₂, F₁, F₂, BC₁P₁ and BC₁P₂) of the crosses were evaluated at two locations during rainy season 2018. MGU‐SP‐101 possessed 2.50–3.78 ears per plant compared to 1.06–1.86 among non‐prolific inbreds. The variation for ears was the highest in F₂s (1–8), followed by BC₁P₁ (1–7) and BC₁P₂ (1–5). The quantitative inheritance pattern of prolificacy with prevalence of non‐allelic interactions of duplicate epistasis type has been observed. Dominance × dominance effect was predominant over additive × additive and additive × dominance effects. Total number of major gene blocks ranged from 0.41 to 2.86, thereby suggesting the involvement of at least one major gene/QTL governing the prolificacy. This is the first report of genetic dissection of prolificacy in “Sikkim Primitive”.     

Inheritance of Striga hermonthica adaptive traits in an early‐maturing white maize inbred line containing resistance genes from Zea diploperennis

Striga hermonthica can cause as high as 100% yield loss in maize depending on soil fertility level, type of genotype, severity of infestation and climatic conditions. Understanding the mode of inheritance of Striga resistance in maize is crucial for introgression of resistance genes into tropical germplasm and deployment of resistant varieties. This study examined the mode of inheritance of resistance to Striga in early‐maturing inbred line, TZdEI 352 containing resistance genes from Zea diploperennis. Six generations, P₁, P₂, F₁, F₂, BC₁P₁ and BC₁P₂ derived from a cross between resistant line, TZdEI 352 and susceptible line, TZdEI 425 were screened under artificial Striga infestation at Mokwa and Abuja, Nigeria, 2015. Additive‐dominance model was adequate in describing observed variations in the number of emerged Striga plants among the population; hence, digenic epistatic model was adopted for Striga damage. Dominance effects were higher than the additive effects for the number of emerged Striga plants at both locations signifying that non‐additive gene action conditioned inheritance of Striga resistance. Inbred TZdEI 352 could serve as invaluable parent for hybrid development in Striga endemic agro‐ecologies of sub‐Saharan Africa.     

Generation means analysis of plant architectural traits and fruit yield in melon

Unique architectural phenotypes have the potential for increasing yield in commercial melon (Cucumis melo L.). Therefore, a generation means analysis was conducted to investigate the inheritance of architectural traits (days to anthesis, primary branch number, fruit number and weight, and average weight per fruit). Progeny (F₁, F₂, BC₁P₁ and BC₁P₂) from a cross between US Department of Agriculture (USDA) line, USDA 846‐1 (P₁) and ‘TopMark’ (P₂) were evaluated at Arlington (AR) and Hancock (HCK), Wisconsin in 2001. Significant (P ≤ 0.05) environment effects and genotype × environment interactions (G × E) analyses necessitated analysis by location. Significant differences (P ≤ 0.05) among parents and generations were observed for all traits, and the two parental lines differed significantly for primary branch number, fruit number and average weight per fruit. Additive gene effects were most important in governing primary branch number and fruit number per plant, while dominance and epistatic genetic effects mainly controlled days to anthesis, fruit weight per plant and average weight per fruit. Narrow‐sense heritabilities were 0.62 (AR) for days to anthesis, 0.71 (AR) and 0.76 (HCK) for primary branch number, 0.68 (AR) and 0.70 (HCK) for fruit weight per plant, 0.33 (AR) and 0.45 (HCK) for fruit weight per plant, and 0.06 (AR) and 0.79 (HCK) for average weight per fruit. Estimations of the least number of effective factors for primary branch number were relatively consistent at both AR (approx. 4) and HCK (approx. 2). Results suggest that introgression of yield‐related genes from highly branched melon types (e.g. USDA 846‐1) into US Western Shipping germplasm may aid in the development of high‐yielding cultivars with concentrated fruit set suitable for machine and/or hand‐harvesting operations.     

Genetic analysis of grain mold resistance in white seed sorghum genotypes

Grain molds in rainy season sorghums can cause poor grain quality resulting in economic losses. Grain molds are a major constraint to the sorghum production and for adoption of the improved cultivars. A complex of fungi causes grain mold. Information on genetics of grain mold resistance and mechanisms is required to facilitate the breeding of durable resistant cultivars. A genetic study was conducted using one white susceptible, three white resistant/tolerant sources, and one colored resistant source in the crossing programme to obtain four crosses. P₁, P₂, F₁, BC₁, and BC₂, and F₂ families of each cross were evaluated for the field grade and threshed grade scores, under sprinkler irrigation. Generation mean analyses and frequency distribution studies were carried out. The frequency distribution studies showed that grain mold resistance in the white-grained resistance sources was polygenic. The additive gene action and additive × additive gene interaction were significant in all the crosses. Simple recurrent selection or backcrossing should accumulate the genes for resistance. Epistasis gene interactions were observed in colored resistance × white resistance cross. Gene interaction was influenced by pronounced G × E. Pooled analysis showed that environment × additive gene interaction and environment × dominant gene interaction were significant. The complex genetics of mold resistance is due to the presence of different mechanisms of inheritance from various sources. Evaluation of segregating population for resistance and selection for stable derivatives in advanced generations in different environments will be effective.     

Genetic analysis of grain mould resistance in coloured sorghum genotypes

Grain moulds are a major constraint to sorghum production and to adoption of improved cultivars in many tropical areas. Information on the inheritance of grain mould reaction is required to facilitate breeding of resistant cultivars. The genetic control of grain mould reaction was studied in 7 crosses of 2 resistant sorghum genotypes. P₁, P₂, F₁, F₂, BC₁ and BC₂ families of each cross were evaluated under sprinkler irrigation for field grade and threshed grade scores and subjected to generation mean analysis. Frequency distributions for grain mould reaction were derived and F₂ and BC₁ segregation ratios were calculated. Grain mould reaction in crosses of coloured grain sorghum was generally controlled by two or three major genes. Resistance to grain moulds was dominant. Significant additive gene effects were also found in all cross/season combinations. Significant dominance effects of similar magnitude to additive effects were also observed in five out of ten cross/season combinations. Gene interactions varied according to the parents with both resistant and susceptible parents contributing major genes. Choice of parents with complementary resistance genes and mechanisms of resistance will be critical to the success of resistance breeding. This revised version was published online in July 2006 with corrections to the Cover Date.     

玉米茎秆糖含量的遗传模式分析

较高的茎秆糖含量有助于提高青贮玉米的饲料品质和适口性,本研究以YXD053和98A-04两个高茎秆糖含量玉米自交系为母本,Y6-1低茎秆糖含量玉米自交系为父本,通过自交、杂交及回交产生2个组合的6个世代(P₁、P₂、F₁、F₂、BC₁和BC₂);运用主基因+多基因混合遗传模型6个世代联合分析方法,探明控制玉米茎秆糖含量的遗传模型,并进行遗传参数估计。结果表明,玉米茎秆糖含量遗传受2对加性-显性-上位性主基因+加性-显性-上位性多基因共同控制。YXD053×Y6-1及98A-04×Y6-1两个组合的主基因遗传率分别为53.50%和52.63%,多基因遗传率分别为7.96%和17.31%,总遗传率分别为61.46%和69.94%,显性度(h/d)均小于1。茎秆糖含量以主基因遗传为主,且主基因又以加性效应为主,但环境因素对茎秆糖含量的遗传有一定的影响。这一研究结果为玉米茎秆糖含量性状的基因定位和育种选择提供了理论依据。     

Genetic analysis of resistance to late leaf spot in interspecific groundnuts

Late leaf spot (LLS), caused by Phaeoisariopsis personata, is an important foliar fungal disease of groundnut (Arachis hypogaea L.), which causes significant economic losses globally to the crop. Inheritance of resistance to LLS disease was studied in three crosses and their reciprocals involving two resistant interspecific derivatives and a susceptible cultivar to refine strategy for LLS resistance breeding. The traits associated with LLS resistance, measured both in the field and under controlled conditions were studied following generation mean analysis. Results suggested that resistance to LLS is controlled by a combination of both, nuclear and maternal gene effects. Among nuclear gene effects, additive effect controlled majority of the variation. In JL 24 × ICG 11337 cross and its reciprocal only additive effects were important, while in JL 24 × ICG 13919 cross and its reciprocal, both additive and dominance effects contributed to the variation. Among digenic epistatic effects, additive × dominance interactions were significant. Additive–maternal effects were significant in both the crosses, while dominance–maternal effects also contributed to the variation in the crosses between the parents, JL 24 and ICG 13919. Due to significant contribution of additive effects of both nuclear and maternal inheritance to resistance to LLS, the parent, ICG 11337 would be a good donor in breeding programs. It would be worthwhile to use the resistance donor as female parent to tap maternal effects of resistance to LLS. Disease score is the best selection criterion in the field for use in breeding programs because of its high heritability and ease in measurement.     

Epistasis and heritability of resistance to <Emphasis Type="Italic">Phytophthora nicotianae</Emphasis> in pepper (<Emphasis Type="Italic">Capsicum annuum</Emphasis> L.)

Gene effects of resistance to two isolates of Phytophthora nicotianae in two crosses of pepper were investigated using separate generation means analysis. Additive-dominance models were inadequate in all cases. Digenic parameter models were adequate in three cases and the probability of goodness of fit of models was negatively correlated with the aggressiveness of the pathogen. None of these models explained variation among generation means in the combined cross Beldi × CM334 with P. nicotianae isolate Pn_2. Additive × additive, dominance × dominance and dominance × additive effects were significant in most cases. Additive and dominance effects (of negative sign) contribute more to resistance than to susceptibility. Additive variance was greater than environmental and dominance variance and ranged from 0.038 to 0.224. Narrow-sense heritabilities were dependent upon the cross and inoculate and ranged from 86 to 92%. The results of this study indicate that selection with more aggressive isolates of the pathogen will be useful for enhancing resistance in pepper.     

Inheritance of drought resistance related traits in two crosses of groundnut (<Emphasis Type="Italic">Arachis hypogaea</Emphasis> L.)

Groundnut (Arachis hypogaea L.) is an important oilseed crop grown in more than 100 countries across wide range of environments. Frequent occurrence of drought is one of the limiting factors adversely affecting groundnut productivity, especially in rainfed areas, and hence genotypes having high water use efficiency (WUE) under limited available water need to be developed. In groundnut, WUE is correlated with SPAD chlorophyll meter reading (SCMR) and specific leaf area (SLA). These two traits, SCMR and SLA, can be used as surrogate traits for selecting for WUE. In order to improve SCMR and SLA, and in turn WUE in groundnut, a good knowledge of the genetic system controlling the expressions of these traits is essential for the selection of the most appropriate and efficient breeding procedure. The present investigation was conducted to determine the gene action controlling the inheritance of SCMR and SLA in two crosses, ICG 7243 × ICG 9418 and ICG 6766 × Chico, and their reciprocals. Six generations of each cross (P₁, P₂, F₁, F₂, BC₁P₁, and BC₁P₂) were evaluated for SCMR and SLA at two stages of the crop growth viz., 60 and 80 days after sowing (DAS). For SCMR at 80 DAS, additive effects were important in both the crosses whereas predominance of dominance effects with duplicate epistasis was observed for SCMR at 60 DAS and SLA at both stages in both the crosses. Predominance of additive effect for SCMR at 80 DAS suggested effective selection could be practiced even in early generations whereas for SCMR at 60 DAS and SLA at both stages in both crosses, it would be better to defer selection to later generations. Further, recording of SCMR and SLA should be done between 60 and 80 DAS for screening the germplasm lines for drought tolerance.     

Inheritance of Maysin Content in Silks of Maize Inbreds Resistant to the Corn Earworm

Genetic information is needed on maysin, a compound in maize, Zea mays L., silks that is antibiotic to corn earworm larvae, Helicoverpa (formerly Heliothis) zea (Boddie), to assist in the process of breeding for resistance to this pest. This study was initiated to elucidate the inheritance of maysin content in the silks of a cross between two maize inbreds (GT114 and GT119) having resistance to the corn earworm. Reversed phase high performance liquid chromatography (HPLC) procedures were used to quantify the maysin concentration of silks from individual plants in the parent, F₁, F₂, and first backcross generations of the cross GT114 × GT119. A model assuming dominance at a single locus for low maysin content which is expressed only when a dominant modifier is present was proposed to explain the segregation ratios obtained in the F₂ and first back-crosses of GT114 × GT119. This model also provided a fit for maysin data previously generated from a cross between inbreds F6 and F44. Observations of selfed families in the backcross of GT114 × GT119 to the low maysin parent confirmed the two-locus model (one factor with dominance for low maysin expressed only when a modifier gene is also present) as a plausible explanation for inheritance of silk maysin in the cross. The process of selection for high maysin will be complicated somewhat by the presence of a gene which modifies the expression of dominance unless a method of easily identifying the gene, independent of maysin content, can be developed.     

Inheritance of Bolting in Three Sugar Beet Crosses after Different Periods of Vernalization

The induction of bolting after different periods of vernalization was studied in the parental, F₁, F₂ and backcross generations (F₂× P₁ and F₂× P₂) of three crosses between five sugar beet lines. The parental lines represented different levels of bolting resistance. The populations were evaluated in replicated field or greenhouse trials. Generation mean analysis was used to quantify the genetic effects of bolting. The additive (d) parameter was predominant for those levels of cold treatment that stimulated the bolting to occur. The dominance (h) parameter was also shown to be important in most cases. In one cross, the additive × additive (i) type of epistasis was significant. The chances of detecting genetic effects are increased by exposing the plants to vernalization conditions which maximize the difference between two parental lines of each cross. Hence, it would be sensible to exploit both additive and non-additive genetic effects in any selection programme.