Morphological characterization, combining ability and heterosis for important horticultural traits in snowball cauliflower (B. oleracea var. botrytis L.)

Abstract

Cauliflower is an important vegetable crop grown worldwide. Development and characterization of suitable cytoplasmic male sterile (CMS) lines and male fertile inbreds is instrumental in developing heterotic hybrids. A study was undertaken to estimate heterosis and combining ability in snowball cauliflower lines through line × tester analysis involving five Ogura CMS lines and seven male fertile testers. The assessment of F₁s along with their parental lines for different horticultural traits revealed a wide range of heterosis. Ogu-13-01 was the best general combiner for leaf length, leaf width and plant height, Ogu-13-85 for number of leaves/plant, Ogu-101 for days to 50% curd initiation, days to 50% curd maturity and net curd weight and Ogu-119 for harvest index. The lines with better general combining ability (GCA) were involved in majority of the heterotic hybrids. The tester Kt-18 was the best general combiner for leaf width, DB-1305 for number of leaves/plant and DB-187 for leaf length, plant height, gross plant weight and days to 50% curd initiation. However, GCA alone was not sufficient to determine and identify the potential parental lines. Hybrids Ogu-101 × DB-1305 and Ogu-119 × Suprimax Late were the best specific combinations for marketable curd weight, marketable curd yield, net curd weight and net curd yield. The hybrid Ogu-119 × Suprimax Late was the best heterotic combination for marketable curd weight, marketable curd yield, net curd weight and net curd yield, followed by Ogu-101 × DB-1305, Ogu-119 × Sel-26 and Ogu-101 × DB-187. Our findings will be instrumental to identify suitable parental lines and developing heterotic F₁ hybrids in cauliflower.

Introduction

Cauliflower (Brassica oleracea var. botrytis L.) is a very popular vegetable in the tropical and subtropical regions of the world and is grown throughout India both commercially as well as in kitchen garden for its tender white curd. It is used as a vegetable in curries, soup, pickles and various other dishes like pakoras. Snowball or European summer cauliflower is the main vegetable crop in Indian sub-continent cultivated during winter season. In cauliflower, F₁ hybrids are advantageous since they are uniform in maturity with high yield, early and have better curd quality with respect to compactness and colour, resistance to insect pests, diseases and unfavourable weather conditions. The magnitude of heterosis is high in cross-pollinated crops like cauliflower than in self-pollinated crops primarily due to greater genetic diversity. Mid-parent (MP) heterosis is being used widely to study the combining ability and determine the extent of heterosis in cauliflower (Dey et al., 2011). This approach was found to be useful in identification of parental lines with the potential to develop heterotic F₁ hybrids in cauliflower (Dey et al., 2013). Pusa Hybrid-2 is the first F₁ hybrid in cauliflower developed in India by using self-incompatible lines in November maturity group. In snowball cauliflower, self-incompatibility system is very weak or not present at all. Self-incompatible (SI) system poses a risk of occurrence of sibs in hybrid seeds besides difficulty in their maintenance and multiplication. In such situation, CMS system offers a good alternative for production of F₁ hybrid seeds. Hence, a speedy crop improvement can be brought about by assessing the genetic variability and exploitation of heterosis using CMS system in snowball cauliflower.

Identification of genetically superior parent is the founding stone for the development of any elite strain in cauliflower. Combining ability studies reveal the nature of gene action and lead to the identification of parents with high general combining ability effects and the cross-combinations with high specific combining ability effects. This in turn helps in choosing the parents to be included in hybridization or population breeding programme. Efforts to the improvement of this crop have been limited. For formulation of a sound breeding programme in snowball cauliflower, the present study was undertaken to find out the best general combiner and the best specific combinations for various morphological traits, yield and yield-attributing traits following a line × tester mating scheme.

Materials and methods

Plant materials

The plant materials include a set of five genetically diverse Ogura cytoplasm-based cauliflower, viz. Ogu 13-01, Ogu-101, Ogu-103, Ogu-119 and Ogu 13-85, which were developed through backcrossing of Snowball-16 (source of Ogura cytoplasm) with elite snowball cauliflower breeding lines at ICAR-Indian Agricultural Research Institute Station, Katrain, Kullu valley, Himachal Pradesh. For line × tester analysis, seven genotypes, viz. Kt-18, Kt-22, DB-1305, DB-187, Lalchowk Maghi, Sel-26 and Suprimax Late with unique traits were used as testers. The genotype DB-187 was resistant to Sclerotinia stalk rot and Sel-26 was partially resistant to black rot Xanthomonas campestris pv. campestris race 4. The lines Kt-18 and Kt-22 possess very compact snow-white curd.

Development of F₁ hybrids and their evaluation

Five CMS lines and seven male fertile lines were crossed in line × tester mating design (Kempthorne, 1957) using standard procedure of hand pollination to obtain 35 F₁ hybrid combinations during spring summer season at IARI Regional Station, Katrain, Himachal Pradesh, India (located at an elevation of 1600 m above mean sea level). The seedlings of 12 parents along with 35 F₁ hybrids were raised and transplanted in a randomized block design with three replications with an inter- and intra-row spacing of 45 cm during winter season at the Research Farm of Division of Vegetable Science, Indian Agricultural Research Institute, New Delhi. Each treatment comprised of five rows and eight plants per row, thus a total of 40 plants. Ten plants were taken at random and tag-labelled in each treatment in all the three replications for recording the observations. All the recommended agronomic practices along with plant protection measures were followed to raise a successful crop. Observations for 14 morphological qualitative traits, viz. seedling: anthocyanin colouration of hypocotyls, leaf attitude, leaf shape, leaf colour, leaf waxiness, curd covering by inner leaves, curd shape in longitudinal section, curd doming, curd knobbling, curd surface texture, curd compactness, curd anthocyanin colouration at maturity, riceyness and curd colour were recorded. Five vegetative traits, viz. (i) leaf length (cm), (ii) leaf width (cm), (iii) number of leaves/plant, (iv) plant height (cm) and (v) gross plant weight (kg); nine yield-attributing traits including yield, viz. (i) days to 50% curd initiation, (ii) days to 50% curd maturity, (iii) marketable curd weight (kg), (iv) net curd weight (kg), (v) marketable curd yield (t/ha), (vi) net curd yield (t/ha), (vii) curd polar diameter (cm), (viii) curd equatorial diameter (cm) and (ix) harvest index (%) were recorded.

Statistical analyses

Statistical analysis of combining ability in line × tester mating design was done following the method given by Kempthorne (1957) and later modified by Arunachalam (1974), using Statistical Package for Agricultural Research (SPAR 2.0) developed by ICAR-Indian Agricultural Statistics Research Institute, New Delhi. Significance of the combining ability effects was determined at 5 and 1% probability. MP heterosis was calculated based on the formulae:

$${\rm Average\ heterosis} = [ ( {{\rm F}_ 1 \hbox{-}{\rm mid\ parent}} ) {\rm /mid\ parent] \\times 100\ \% }.$$

Significance was tested through F test at 5 and 1% probability.

Results

Morphological characterization of lines, testers and F₁ hybrids

Lines and testers were assessed for 14 qualitative traits. Anthocyanin colouration of hypocotyls was present in all lines and testers except Kt-18 and Sel-2. Leaf attitude was erect, curd shape in longitudinal section was circular and curd surface texture was fine in all lines and testers. Leaf shape was elliptic for all lines but in testers it was broad elliptic for Kt-18, Kt-22 and DB-1305; narrow elliptic for Lalchowk Maghi and elliptic for DB-187, Sel-26 and Suprimax Late. Curd of all lines and testers was covered or partially covered by inner leaves except in Suprimax Late. Curd knobbling was medium in all lines and testers except Kt-18 and Lalchowk Maghi in which it was fine. Curd doming was strong in Ogu 101, Kt-18, Kt-22 and Lalchowk Maghi; medium in Ogu 13-01, Ogu 103, Ogu 119, DB-1305 and DB-187 and weak in Ogu 13-85 and Sel-26. Curd of all lines and testers was compact or medium compact and curd colour was white or snow white. Cauliflower is a thermo-sensitive crop and susceptible to several non-parasitic physiological disorders. Ogura CMS lines should be free from these disorders. In our study, the anthocyanin pigmentation of curd at maturity and riceyness were absent in all lines and testers (Supplementary material 1).

Observations of hybrids for all qualitative traits are given in Supplementary material 2. Anthocyanin colouration of hypocotyls is present in all hybrids except Ogu 13-01 × Kt-18, Ogu 101 × Kt-18, Ogu 119 × Suprimax Late, Ogu 13-85 × Kt-18 and Ogu 13-85 × DB-187. Leaf attitude is erect, curd shape in longitudinal section is circular and curd surface texture is fine in all hybrids. Leaf shape is elliptic for all hybrids but it is broad elliptic for Ogu 13-01 × Kt-18, Ogu 13-01 × DB-1305, Ogu 13-01 × Sel-26, Ogu 101 × Kt-22, Ogu 101 × Sel-26, Ogu 103 × Kt-18, Ogu 103 × Sel-26, Ogu 119 × DB-1305 and narrow elliptic for Ogu 13-85 × Lalchowk Maghi. Curd of all hybrids is covered or partially covered by inner leaves except in Ogu 101 × Kt-18, Ogu 101 × Lalchowk Maghi and Ogu 13-85 × Lalchowk Maghi and curd knobbling is medium in all hybrids except Ogu 13-01 × Kt-22, Ogu 13-01 × DB-1305, Ogu 13-01 × Lalchowk Maghi, Ogu 101 × DB-187, Ogu 101 × Sel-26, Ogu 101 × Suprimax Late, Ogu 119 × Kt-18, Ogu 119 × Lalchowk Maghi, Ogu 119 × Suprimax Late, Ogu 13-85 × DB-1305 and Ogu 13-85 × Suprimax Late in which it is fine. Curd doming is medium or strong in all hybrids except weak in Ogu 103 × Sel-26 and Ogu 13-85 × Sel-26. Leaf colour is dark green or bluish green except in two hybrids Ogu 119 × Kt-22 and Ogu 13-85 × Kt-18 where it is light green. Leaf waxiness is light or medium except in hybrid Ogu 101 × DB-187. Curd of all hybrids is compact or medium compact and curd colour is white or snow-white but creamy white in Ogu 103 × Kt-22, Ogu 103 × Sel-26, Ogu 119 × Kt-22 and Ogu 13-85 × Sel-26.

The hybrids based on Ogura CMS should be free of non-parasitic physiological disorders and should have good external quality. Hybrids were assessed for curd colour, compactness, curd surface texture, presence of riceyness and undesirable pigmentation in the curd during harvest maturity (edible maturity) stage. Thirteen hybrids were found promising with respect to external curd quality as per visual assessment. Out of these, three hybrids, viz. Ogu 13-01 × Lalchowk Maghi, Ogu 119 × DB 1305 and Ogu 119 × Lalchowk Maghi have compact, snow-white and anthocyanin- and riceyness-free curds with fine curd surface texture, circular curd shape and strong curd doming. Ten hybrids, viz. Ogu 101 × Kt-18, Ogu 101 × Kt-22, Ogu 101 × DB1305, Ogu 103 × Suprimax Late, Ogu 119 × Kt-18, Ogu 119 × DB-187, Ogu 119 × Suprimax Late, Ogu 13-85 × DB-187, Ogu 13-85 × Lalchowk Maghi and Ogu 13-85 × Suprimax Late have compact, white and anthocyanin- and riceyness-free curds with fine curd surface texture, circular curd shape and strong curd doming.

Anthocyanin pigmentation of curd at maturity is absent in all hybrids. Similarly, riceyness was absent in all hybrids except Ogu 103 × Kt-22, Ogu 119 × Kt-22 and Ogu 13-85 × Kt-22 (Supplementary material 2).

Analysis of variance for lines, testers and F₁ hybrids

The partitioning of mean squares into replications, lines, testers and line × tester interactions revealed that mean squares due to lines (female parents) were significant for all the parameters studied except gross plant weight, polar diameter, equatorial diameter and harvest index. Mean squares of testers (male parents) and lines × testers interactions were significant for leaf length, leaf width, number of leaves/plant, plant height, days to 50% curd initiation and days to 50% curd maturity (Supplementary material 3 and 4). Similarly, Pandey and Naik (1989) observed significant differences in the variance due to line × testers for leaf number, plant height, curd weight and curd size; Singh et al. (2005) for leaf number, leaf weight and leaf area; and Verma and Kalia (2011) for days to 50% curd initiation, days to 50% curd maturity and gross plant weight in cauliflower.

The estimates of variance components for different vegetative traits, yield and yield-attributing traits are presented in Table 1. The values of σ ² gca were lower than the σ ² sca for all the traits except gross plant weight, days to 50% curd maturity, polar diameter, equatorial diameter and harvest index. For number of leaves per plant and equatorial diameter, the σ ² gca was in negative direction. The proportions of σ ² gca/σ ² sca were less than unity in all traits except for days to 50% curd maturity and harvest index. Similarly, σ ²D was greater than σ ²A for all the traits except leaf width, gross plant weight, days to 50% curd maturity, polar diameter, equatorial diameter, net curd yield and harvest index.

Table 1. Genetic components of variance for different vegetative traits, yield and yield-attributing traits in snowball cauliflower

σ ²A, variance of additive genetic component; σ ²D, variance of dominance genetic component.

Combining ability

General combining ability of the parental lines and testers

The perusal of the data on general combining ability effects of the lines and testers (Tables 2, 3) indicated that, for leaf length, highly significant GCA effect was recorded in lines Ogu 13-01 (4.46) and Ogu 119 (2.06) and testers DB-187 (2.99) and Kt-18 (2.11). For leaf width, highly significant GCA effect was noted in CMS lines Ogu 13-01 (3.39) and Ogu 101 (0.83) and testers Kt-18 (1.74) and Sel-26 (1.59). For number of leaves/plant, maximum significant GCA effect in negative direction was recorded in line Ogu 13-85 (−0.77) and tester DB-1305 (−0.82). Among the lines, Ogu 13-01 (4.29) and Ogu 119 (1.82) and among the testers, DB-187 (2.72) and Kt-18 (1.69) were the best general combiners for plant height. For gross plant weight, DB-187 was the only tester, which proved to be a best general combiner but not a single line proved to be a good combiner for gross plant weight. Similarly, for harvest index, Ogu 119 (4.46) was the only line, which was a best general combiner with highly significant GCA effect but not a single tester was a good combiner for harvest index. Ogu 101 (−4.83), DB 187 (−3.12) and Lalchowk Maghi (−3.12) were the best general combiners for days to 50% curd initiation, whereas Ogu 101 (−6.00), Lalchowk Maghi (−4.31) and Kt 18 (−1.57) were the best general combiners for days to 50% curd maturity. No line and tester was found to have significant positive GCA effect for polar diameter, equatorial diameter, marketable curd weight, marketable curd yield and net curd yield. Among the lines and testers, significant positive GCA effect for curd weight and harvest index was noted in Ogu 101 (0.04) and Ogu 119 (4.46), respectively.

Table 2. GCA effects of genotypes for different vegetative traits in snowball cauliflower

CD, critical difference.

*^, ** Significant at 5 and 1% probability, respectively.

Table 3. GCA effects of genotypes for yield and yield-attributing traits in snowball cauliflower

CD, critical difference.

*^, **Significant at 5 and 1% probability, respectively.

Thus, the line Ogu 13-01 was the best general combiner for leaf length, leaf width and plant height, Ogu 13-85 for number of leaves/plant, Ogu 101 for days to 50% curd initiation, days to 50% curd maturity and net curd weight. The tester Kt-18 was the best general combiner for leaf width, DB-1305 for number of leaves/plant, DB-187 for leaf length, plant height, gross plant weight and days to 50% curd initiation and Lalchowk Maghi for days to 50% curd initiation and days to 50% curd maturity (Tables 2, 3).

Specific combining ability of the hybrids

The specific combining ability (SCA) effects for different vegetative traits and yield and yield-attributing traits were presented in Supplementary material 5 and 6, respectively. Ten, eight, 10 and one hybrids exhibited significant positive SCA effects for leaf length, leaf width, plant height and gross plant weight, respectively. Two hybrids each recorded significant positive SCA effects for equatorial diameter, marketable curd weight, marketable curd yield, net curd weight and net curd yield. No hybrid noted significant positive SCA effect for polar diameter and harvest index. Significant negative SCA effects were recorded in four, three and three hybrids for number of leaves/plant, days to 50% curd initiation and days to 50% curd maturity, respectively.

Maximum highly significant positive SCA effect for leaf length and plant height was recorded in hybrid Ogu 119 × DB-1305 (10.79, 11.21), followed by Ogu 13-01 × Kt-18 (9.15, 9.21), respectively. Maximum highly significant positive SCA effect for leaf width was noted in Ogu 103 × Sel-26 (3.72), followed by Ogu 13-85 × DB-1305 (2.93). The hybrid Ogu 13-01 × Sel-26 exhibited maximum highly significant negative SCA effect (−3.25) for number of leaves/plant, followed by Ogu 119 × DB-1305 (−2.16). For gross plant weight, only one hybrid, i.e. Ogu 103 × Lalchowk Maghi, recorded significant positive SCA effect (0.27).

Highest significant negative SCA effect for days to 50% curd initiation was noted in Ogu 13-85 × DB-187 (−12.02), followed by Ogu 101 × DB-1305 (−5.64) and Ogu 103 × Suprimax Late (−4.20). Maximum significant negative SCA effect for days to 50% curd maturity was recorded in Ogu 101 × Lalchowk Maghi (−3.93), followed by Ogu 13-01 × DB-187 (−3.89) and Ogu 101 × DB-1305 (−3.33). For equatorial diameter, highest significant positive SCA effect was noted in Ogu 103 × Lalchowk Maghi (1.83), followed by Ogu 119 × DB-187 (1.68). For marketable curd weight, marketable curd yield, net curd weight and net curd yield, maximum highly significant positive SCA effect was recorded in hybrid Ogu 101 × DB 1305 (0.24, 10.75, 0.12, 5.19), followed by Ogu 119 × Suprimax Late (0.19, 8.53, 0.11, 4.80), respectively. For polar diameter and harvest index, no hybrid recorded significant positive SCA effects.

For specific combining ability effects among 35 F₁ hybrids, 10 hybrids, Ogu 13-01 × Kt-18 (good × good general combiner), Ogu 13-01 × Suprimax Late (good × poor general combiner), Ogu 101 × DB-187 (average × good general combiner), Ogu 101 × Suprimax Late (average × poor general combiner), Ogu 103 × Kt-22 (poor × average general combiner), Ogu 103 × DB-187 (poor × good general combiner), Ogu 103 × Sel-26 (poor × average general combiner), Ogu 119 × DB-1305 (good × poor general combiner), Ogu 13-85 × DB-187 (poor × good general combiner) and Ogu 13-85 × DB-1305 (poor × poor general combiner) in term of superiority were the best specific combiners for leaf length. Eight hybrids, Ogu 13-01 × Kt-18 (good × good general combiner), Ogu 13-01 × Sel-26 (good × good general combiner), Ogu 101 × DB-187 (good × poor general combiner), Ogu 101 × Lalchowk Maghi (good × poor general combiner), Ogu 101 × Suprimax Late (good × poor general combiner), Ogu 103 × Kt-22 (poor × good general combiner), Ogu 103 × Sel-26 (poor × good general combiner) and Ogu 13-85 × DB-1305 (poor × poor general combiner) in term of superiority were the best specific combiners for leaf width. Only four hybrids, Ogu 13-01 × Sel-26 (average × poor general combiner), Ogu 119 × DB-1305 (average × good general combiner), Ogu 101 × DB-187 (poor × average general combiner) and Ogu 101 × Kt-18 (poor × average general combiner) were good specific combiners for number of leaves/plant. Ten hybrids, Ogu 13-01 × Kt-18 (good × good general combiner), Ogu 13-01 × Sel-26 (good × good general combiner), Ogu 13-01 × Suprimax Late (good × poor general combiner), Ogu 101 × DB-187 (poor × good general combiner), Ogu 101 × Suprimax Late (poor × poor general combiner), Ogu 103 × DB-187 (poor × good general combiner), Ogu 103 × Sel-26 (poor × good general combiner), Ogu 119 × DB-1305 (good × poor general combiner), Ogu 13-85 × DB-1305 (poor × poor general combiner), Ogu 13-85 × Lalchowk Maghi (poor × poor general combiner) were observed to be good specific combiners for plant height (Table 2, Supplementary material 5).

Three hybrids, Ogu 13-85 × DB-187 (average × good general combiner), Ogu 101 × DB-1305 (good × poor general combiner) and Ogu 103 × Suprimax Late (poor × poor general combiner) were found to be good specific combiners for days to 50% curd initiation. For days to 50% curd maturity, three hybrids Ogu 101 × Lalchowk Maghi (good × good general combiner), Ogu 13-01 × DB-187 (average × poor general combiner) and Ogu 101 × DB-1305 (good × poor general combiner) were noted to be good specific combiners. Two hybrids, Ogu 103 × Lalchowk Maghi (average × average general combiner) and Ogu 119 × DB-187 (poor × poor general combiner) were observed to be good specific combiners for equatorial diameter. The hybrid Ogu 101 × DB-1305 was noted to be the best specific combiner for marketable curd weight, marketable curd yield and net curd yield (average × poor general combiner) and net curd weight (good × poor general combiner). The hybrid Ogu 119 × Suprimax Late (average × average general combiner) was noted to be the second best specific combiner for marketable curd weight, marketable curd yield, net curd weight and net curd yield (Table 3, Supplementary material 6).

Heterosis

Heterosis for all the hybrids were calculated over MP and expressed as MP heterosis. Very high and significant heterosis in both directions was recorded for all the 14 quantitative traits under the present study (Tables 4, 5).

Table 4. Heterosis over mid-parent (%) for different vegetative traits in snowball cauliflower

CD, critical difference; *^, **significant at 5 and 1% probability, respectively.

Table 5. Heterosis over mid-parent (%) for yield and yield-attributing traits in snowball cauliflower

CD, critical difference; *^, **significant at 5 and 1% probability, respectively.

Heterosis for different vegetative traits is presented in Table 4. The range of MP heterosis for leaf length was −22.64–37.41%. Eleven hybrids recorded significant heterosis in positive direction over MP. The hybrid Ogu 13-01 × Kt-18 showed maximum heterosis in positive direction for leaf length (37.41%), followed by Ogu 13-01 × Suprimax Late (27.75%) and Ogu 13-01 × Sel-26 (27.36%). For leaf width, MP heterosis ranged from −24.03 to 59.90%. Ten hybrids recorded significant positive heterosis over MP. The hybrid Ogu 13-01 × Sel-26 showed maximum heterosis in positive direction (59.90%), followed by Ogu 13-01 × Kt-18 (48.68%) and Ogu 103 × Sel-26 (32.85%). The range of heterosis for number of leaves per plant was −19.06 to 58.24%. Nine hybrids recorded significant negative heterosis over MP. The hybrid Ogu 119 × Kt-18 had highest significant negative heterosis for number of leaves per plant in negative direction (−19.06%), followed by Ogu 119 × DB-1305 (−16.42%) and Ogu 13-85 × Kt-18 (−14.70%). For plant height, heterosis ranged from −21.08 to 29.91%. Ten hybrids recorded significant heterosis in positive direction over MP. The hybrid Ogu 13-01 × Kt-18 exhibited significant heterosis in positive direction (29.91%), followed by Ogu 119 × DB-1305 (25.59%) and Ogu13-01 × Sel-26 (24.96%) for this trait. The range of heterosis for gross plant weight was −23.41 to 54.61%. Twenty-three hybrids recorded significant heterosis in positive direction over MP. Maximum positive heterosis was observed in Ogu 119 × Suprimax Late (54.61%), followed by Ogu 13-85 × DB-187 (44.24%) and Ogu 13-01 × Suprimax Late (38.89%).

Heterosis for yield and yield-attributing traits is presented in Table 5. Heterosis for days to 50% curd initiation ranged from −27.05 to 6.67%. Twelve hybrids recorded significant negative heterosis over MP. The hybrid Ogu 13-85 × DB-187 showed highest heterosis in negative direction (−27.05%), followed by Ogu 101 × Lalchowk Maghi (−12.76%) and Ogu 119 × Kt-22 (−11.87%). Heterosis for days to 50% curd maturity ranged from −10.81 to 5.07%. Out of 35 hybrids, 11 showed significant negative heterosis over MP. The hybrid Ogu 119 × Suprimax Late recorded highest negative heterotic values of −10.81%, followed by Ogu 101 × Lalchowk Maghi (−10.45%) and Ogu 13-85 × Kt-18 (−9.73%).

For polar diameter, the extent of heterosis varied from −6.40 to 23.30%. Twenty-nine hybrids recorded significant positive heterosis over MP. The hybrid Ogu 103 × Lalchowk Maghi exhibited highest heterotic value (23.30%), followed by Ogu 103 × Kt-22 (21.42%) and Ogu 13-85 × DB-1305 (20.00%). For equatorial diameter, the extent of heterosis varied from −7.19 to 19.74%. Twenty-seven hybrids recorded significant positive heterosis over MP. The hybrid Ogu 103 × Lalchowk Maghi also exhibited highest heterotic value in positive direction (19.74%) followed by Ogu 119 × DB-187 (19.06%) and Ogu 13-01 × DB-187 (15.20%).

Heterosis for marketable curd weight and marketable curd yield ranged from −20.19 to 76.61%. Out of 35 F₁ hybrids, 28 and 25 exhibited significant heterosis over MP for marketable curd weight and marketable curd yield, respectively. For both marketable curd weight and marketable curd yield, the hybrid Ogu 119 × Suprimax Late also exhibited highest significant heterotic value in positive direction (76.61%), followed by Ogu 13-85 × DB-187 (56.36%) and Ogu 101 × DB-187 (46.81%). With respect to net curd weight and net curd yield, heterosis ranged from −25.58 to 94.12%. Out of 35 F₁ hybrids, 27 and 26 exhibited significant heterosis over MP for net curd weight and net curd yield, respectively. For both net curd weight and net curd yield, the hybrid Ogu 119 × Suprimax Late also exhibited highest significant heterotic value in positive direction (94.12%), followed by Ogu 13-85 × Suprimax Late (46.07%) and Ogu 119 × Sel-26 (43.59%). The promising heterotic hybrids are presented in Fig. 1.

Figure 1. Promising heterotic hybrids of snowball cauliflower.

The range of heterosis for harvest index was observed to be from −21.12 to 31.09%. Out of 35 F₁ hybrids, 14 exhibited significant heterosis in positive direction over MP. The hybrid Ogu 119 × Suprimax Late showed highest significant heterosis in positive direction (31.09%), followed by Ogu 119 × Sel-26 (29.73%) and Ogu 119 × Kt-22 (29.12%).

Discussion

The knowledge of combining ability together with per se performance of parents, hybrids and heterotic response helps the breeders in selecting suitable parents and crosses for their use in systematic breeding programme. The combining ability estimates also provide information to determine relative magnitude of additive and non-additive gene effects, and to develop breeding strategies for the development of heterotic hybrids.

Line × tester analysis (Kempthorne, 1957) helped in determining the suitability of the CMS lines in heterosis breeding based on the GCA effects of the parental lines and SCA effects of each cross. The estimate of GCA of a parent is an important indicator of its potential for generating superior breeding populations. A high GCA estimate indicates that the parental mean is superior or inferior to the general mean. This represents a strong evidence of favourable gene flow from parents to offspring at high frequency and gives information about the concentration of predominantly additive genes (Franco et al., 2001). The selected parental lines with better performance can be crossed in suitable combination to exploit heterosis. Such crosses with high SCA could be best utilized in heterosis breeding (Singh and Chaudhary, 1979).

In most of the cases, per se performance of parents and F₁ hybrids bears direct reflection of their respective GCA and SCA effects, i.e. parents and hybrids showing highest GCA and SCA effects for a trait were also observed to be a good performer with respect to that particular trait.

Based on the above results, it was evident that among five CMS lines used in the study, Ogu 101 took minimum number of days to 50% curd initiation and days to 50% curd maturity and had highest leaf length, leaf width, plant height and minimum number of leaves per plant; whereas Ogu 13-85 recorded maximum number of days to 50% curd initiation and days to 50% curd maturity. Ogu 13-85 also showed minimum value for polar diameter, equatorial diameter, marketable curd weight, marketable curd yield, net curd weight, net curd yield and harvest index. CMS line Ogu 101 showed maximum value for polar diameter, equatorial diameter, marketable curd weight and marketable curd yield; Ogu 103 had maximum net curd weight and net curd yield and Ogu 13-01 had maximum harvest index. Among seven testers, Lalchowk Maghi took minimum number of days to 50% curd initiation and days to 50% curd maturity, whereas Kt-18 and Suprimax Late took maximum number of days to 50% curd initiation and Suprimax Late maximum number of days to 50% curd maturity. Lalchowk Maghi and Suprimax Late recorded minimum value for polar diameter and equatorial diameter, respectively. DB-187 had minimum marketable curd weight and marketable curd yield and Suprimax Late had minimum net curd weight and net curd yield. Kt-18 noted minimum value for harvest index. Sel-26 and Kt-18 recorded maximum polar diameter and equatorial diameter, respectively. Kt-22 had maximum marketable curd weight, marketable curd yield, net curd weight and net curd yield. DB-1305 recorded maximum value for harvest index.

A perusal of the result revealed that for yield and yield-attributing traits, per se performance of the parents bears direct reflection of their respective GCA effects, i.e. parents showing highest GCA effects for a trait were also observed to be good a performer with respect to that particular trait. The results are in conformity with the findings of Singh et al. (1976), Sharma et al. (1988), Pandey and Naik (1989), Gangopadhyay et al. (1997), Singh et al. (1997), Thakur and Singh (1999), Thakur et al. (2001), Singh et al. (2005), Varalakshmi (2009) and Verma and Kalia (2011) in cauliflower.

In most of the cases, the hybrids which showed the best per se performance for yield and yield-attributing traits also possessed desirable significant SCA effects. This indicated the per se performance of hybrids had a direct relation with respective SCA effects. A comparison of the SCA effects of the hybrids and GCA effects of the parents involved indicated that in most of the cases GCA effects were reflected in the SCA effects of the cross-combination. It is apparent that in almost all the hybrids which showed the best SCA effects, the parental lines involved were at least one of the five outstanding parental lines, viz. Ogu 101, Ogu 13-01, Ogu 119, Kt-18, DB-187 which also recorded high GCA effects for one or more traits contributing towards yield. This indicated that there was strong tendency of transmission of higher gain from the parents to the offspring. The present findings corroborated the earlier work of Singh et al. (1976), Dhiman et al. (1983), Pandey and Naik (1989), Gangopadhyay et al. (1997), Thakur and Singh (1999), Singh et al. (2005) and Verma and Kalia (2011) in cauliflower.

Ogu 101 × DB-1305, Ogu 13-01 × DB-187, Ogu 101 × Lalchowk Maghi, Ogu 103 × Suprimax Late and Ogu 13-85 × DB-187 were the good specific combiners for earliness. It is apparent that hybrids which showed high SCA effects involved either good × poor, poor × good, good × average, average × good, good × average or good × good general combiners. This indicates that prediction of hybrids performance cannot be done in advance on the basis of GCA performance of parents. Similar observations were reported by Singh et al. (1976), Dhiman et al. (1983), Gangopadhyay et al. (1997), Thakur and Singh (1999), Singh et al. (2005) and Verma and Kalia (2011) who reported high SCA for earliness in cauliflower hybrids. Ogu 101 × DB-1305 and Ogu 119 × Suprimax Late were the best combiners for marketable curd weight, marketable curd yield, net curd weight and net curd yield. Ogu 101 × DB-1305 was also a best specific combiner for earliness.

The magnitude of SCA variance was higher than GCA variance for yield and yield-attributing traits. Thus, it may be concluded that there was a major role of dominance in exhibiting hybrid vigour of considerable magnitude in all 14 quantitative traits, thereby indicating that the improvement in snowball cauliflower can be made by heterosis breeding.

The study of heterosis revealed that the range of mean values in hybrids was higher than that of parents for all the traits studied except plant height and gross plant weight. Appreciable amount of heterosis was observed in positive and negative directions for all the traits. Heterosis in negative direction is desirable for characters like number of leaves per plant, days to 50% curd initiation and days to 50% curd maturity. Days to 50% curd maturity is considered an important trait as it determined final harvesting of curds. Earliness (indicated by negative estimates of heterosis) is a well-recognized and prime objective of any breeding programme as it helps the grower to reap a high market price earlier. The hybrids Ogu 101 × Lalchowk Maghi, Ogu 13-85 × Kt-18, Ogu 13-85 × DB-187, Ogu 119 × Kt-22 and Ogu 119 × Suprimax Late were found to be the most promising for earliness. Similar results were also reported by Pal and Swarup (1966), Singh et al. (1976) and Sharma et al. (1988). Dey et al. (2011) reported −16.30 and −9.41% heterosis for days to 50% curd initiation and days to 50% curd maturity, respectively. Plants having lower number of leaves are desirable in cauliflower as it will have low frame and thus, will accommodate more number of plants per unit area, enabling high-density planting.

High heterosis (50–80%) was obtained for traits leaf width, number of leaves per plant and gross plant weight. However, moderate heterosis (10–40%) was recorded for seven traits, viz. leaf length, plant height, days to 50% curd initiation, days to 50% curd maturity, polar diameter, equatorial diameter and harvest index which may be mainly attributed to the narrow genetic base of Indian snowball cauliflower. Similar results were reported by Dey et al. (2011) in snowball cauliflower.

Yield is the foremost trait to be considered in cauliflower breeding programme. It is a complex trait resulting from the interaction of its component traits of a crop. In cauliflower, marketable curd weight and net curd weight are the direct components of yield. The hybrid Ogu 119 × Suprimax Late was the best heterotic combination for marketable curd weight, marketable curd yield, net curd weight and net curd yield, followed by Ogu 101 × DB-1305, Ogu 119 × Sel-26, Ogu 101 × DB-187, Ogu 13-01 × Kt-18 and Ogu 119 × Lalchowk Maghi. The above results also indicate that maximum yield per hectare in the above-mentioned hybrids was attributed by maximum marketable curd weight and net curd weight. The above promising hybrids also have excellent external quality traits, viz. compact, white or snow-white and anthocyanin- and riceyness-free curds with fine curd surface texture and circular curd shape. In the present study, high significant heterosis in positive direction (50–94%) was noted for marketable curd yield and net curd yield and major yield-attributing traits, viz. marketable curd weight and net curd weight. Garg and Lal (2005) reported the crosses in cauliflower frequently exhibited high parent heterosis due to increased equatorial diameter and net curd weight. Therefore, yield can more accurately be estimated by the curd diameter and net curd weight and it would be possible to achieve yield improvement in this crop by manipulating these particular traits. Hence, breeder should concentrate mainly on curd diameter and net curd weight, in their efforts to increase yield. The results of present investigation are similar to the report of Garg and Lal (2005). They noticed 53.69 and 35.76% hybrid vigour for polar and equatorial diameter in cauliflower, respectively. They also reported 136.90% heterosis for net curd weight. Singh et al. (2009) found 10 hybrids in cauliflower were exhibiting maximum curd diameter and net curd weight than their parents. They reported 35.63 and 45.80% heterosis for curd diameter and net curd weight, respectively.

Conclusion

The present study revealed a wide range of heterosis for yield-related traits through selection of suitable CMS and male fertile inbreds. Among the CMS lines, Ogu 101 was the best general combiner for days to 50% curd initiation, days to 50% curd maturity and net curd weight and Ogu 119 was best for harvest index. The lines with better GCA were involved in majority of the heterotic hybrids. However, GCA alone was not sufficient to select suitable parental lines. Hybrids Ogu 101 × DB-1305 and Ogu 119 × Suprimax Late were the best specific combinations for marketable curd weight, marketable curd yield, net curd weight and net curd yield. The study revealed better suitability of the specific combing ability for most of the traits in selection of parents and development of heterotic hybrids. The results of this study will be instrumental to identify suitable parental lines and developing heterotic F₁ hybrids in cauliflower.