Heterosis and combining ability studies by Line × tester analysis for fruit biochemical, morpho-physiological, and yield traits governing shelf life in tomato (Solanum lycopersicum L.)

Improving tomatoes keeping quality is crucial for reducing post-harvest losses. Knowledge on heterosis, and combining ability is pre requisite for breeding high yielding and good shelf life heterotic hybrids. An investigation was undertaken with each of 3 lines, testers, and 9 hybrids to identify desirable parents and crosses for 20 fruit biochemical, morpho-physiological, and yield traits and to elucidate nature of gene action for shelf life and its contributing traits through Line × Tester analysis. The lines contributed to most of hybrids variability than testers and fruit quality traits had higher degree of SCA variance as compared to GCA variance. pH, ascorbic acid, fruit rmness, and plant height governed by additive gene action. Lycopene, titratable acidity, TSS, calcium, magnesium, pericarp thickness, pulp content, locule number, fruit length, diameter, weight, shelf life, number of branches, number of clusters, number of fruit/cluster, and yield/plant were under the control of non-additive gene action. All the lines and Arka Saurabh were best general combiners and IIHR 2349 × Arka Vikas, IIHR 2349 × Arka Saurabh, IIHR 2358 × Arka Ahuti and IIHR 2357 × Arka Ahuti were the best specic combiner in producing heterotic hybrids. IIHR 2349 × Arka Vikas and IIHR 2349 × Arka Saurabh were promising hybrids for high yield and shelf life. The crosses involved both parents with high, one parent with high and other with low and both parents with low good overall general combining ability status respectively indicated the additive, non-additive and epistatic gene action in fruit quality and yield traits inheritance. on their per se performance for yield and shelf life. The hybrid IIHR 2349 × Arka Vikas recorded a signicantly higher yield potential of 2850.50 g/plant with signicantly higher standard heterosis over standard check Arka Rakshak (52.27%), and it had 24.70 days of shelf life. Higher heterosis for yield/plant in this hybrid may be attributed to higher mean performance for number of clusters/plant (11.05) which is further reiterated by signicantly higher standard heterosis for the same trait (8.87%). Another promising hybrid IIHR 2349 × Arka Saurabh recorded a signicantly higher shelf life of 29.40 days with signicantly higher standard heterosis over standard check Arka Rakshak (26.45%), and it had a yield potential of 2397.50 g/plant. Higher heterosis for shelf life in this hybrid may be contributed by signicantly higher mean performance for fruit magnesium content (17.79 mg/100g) and lycopene content (4.59 mg/100g) which is further evident by signicantly higher standard heterosis for magnesium content (154.14%) and lycopene content (128.36%). These promising hybrids need to be tested at preliminary yield trials and multilocation trials to assess their performance and stability before commercialization. The crosses, IIHR 2349 Arka Vikas and IIHR 2349 Arka Saurabh were promising hybrids for high yield and shelf life. Signicant differences among the genotypes indicated considerable genetic variability for fruit quality traits. The estimates of SCA variance was higher as compared to GCA variance. The non-additive gene action was exhibited for lycopene, titratable acidity, TSS, fruit length, diameter, weight, locule number, number of branches, number of clusters, number of fruit/cluster and yield/plant which can be exploited through hybrid development and additive gene action shown for pH, ascorbic acid, fruit rmness, and plant height can be exploited through varietal development. The study underlined the signicance of non-additive gene action for shelf life, pericarp thickness, calcium, magnesium, pulp content. The involvement of low combiners in hybridization signied over dominance and epistasis hence intensive selection for SCA in segregating generations will be effective. Involvement of high and low combiners in the crossing revealed the signicance of non-additive gene action which is a prerequisite for heterosis breeding. The involvement of high combiners in crossing signied the importance of additive gene action and simple selection in segregating generation will be effective with more reliance on mass selection and progeny selection.


Development of experimental material
The crosses were affected between high shelf life low yielding lines and high yielding low shelf life testers in Line × Tester (3 × 3) mating design. The nine rst lial generations were developed by pollinating the pollens from testers onto the stigmas of emasculated owers of lines in the early morning hours (6-8 AM) in protected poly house conditions during the 2017 summer. The seeds of lines, testers, and F 1 's were harvested individually which constituted the experimental material.

Evaluation of experimental material
The nine hybrids along with their parents and standard check were planted in the eld during the 2017 rainy season in two separate contiguous blocks in Randomized Complete Block Design with two replications. The standard crop production and protection practices were followed to raise healthy plants and were evaluated for fruit biochemical, morpho-physiological, and yield traits governing shelf life.

Sampling of plants and collection of data
The data were recorded from ve randomly selected plants avoiding border plants in each line, testers, and F 1 generations. The 20 fruit quality traits governing shelf life such as 7 fruit biochemical traits viz., TSS (%) (using Erma hand refractometer), pH (using Siemens pH meter), lycopene (mg/100g) (Lichtenthaler 1987), ascorbic acid (mg/100g) (2,6-Dichlorophenol indophenol method, Association of O cial Analytical Chemists 2006), titratable acidity (%) (Association of O cial Analytical Chemists 2000), calcium and magnesium content (mg/100g) (di-acid digestion and ICP-OES method), 8 morphophysiological traits viz., fruit length (cm), fruit diameter (cm) and pericarp thickness (mm) were measured with digital vernier caliper, fruit weight (g) (using digital weighing balance), fruit rmness (kg/cm 2 ) (using fruit penetrometer), pulp content (%) (Percent ratio of weight of pulp after removing fruit juice and seed to the fruit weight), locule number (manually counted), shelf life (Days) (counted as number of days taken by fruits harvested at breaker stage kept on shelf to show rst visible shrinkage on its fruit surface) and 5 yield attributing traits viz., plant height (cm) (using measuring scale), number of branches, number of clusters, number of fruit/cluster were manually counted, yield/plant (g) (using digital weighing balance) were recorded in the eld during harvest.
The fruit quality traits were recorded in the lab at the red ripe stage of ve randomly selected tomato fruits from each plant and the mean was computed.

Statistical analysis
The collected mean data on fruit quality traits of parents and hybrids were analyzed for percent heterosis over their mid-parent (mid parent heterosis or average heterosis), better parent (heterobeltiosis) and standard check (standard heterosis) (Turner 1953;Hayes et al. 1955), Analysis of Variance (ANOVA) (Panse and Sukhatme 1967), General Combining Ability (GCA) effects of lines, testers and Speci c Combining Ability (SCA) effects of hybrids using Line × Tester analysis (Kempthorne 1957;Arunachalam 1974), percent contribution of lines, testers and their interaction towards total variability in each character (Singh and Choudhory 1977), the overall status of a parent or a cross concerning GCA or SCA effects (Arunachalam and Bandopadhyay 1979) with slight modi cation as suggested by Mohan  using 'WINDOSTAT' statistical package.

Results
3.1 Mean performance of lines, testers, and hybrids for fruit biochemical, morphophysiological, and yield attributing traits The lines and testers recorded substantial variability for all traits. The performance of lines was higher compared to testers for pH, calcium, magnesium, fruit rmness, pericarp thickness, pulp content, locule number, shelf life, and number of clusters (Table I). Line IIHR 2349 was superior for pH, titratable acidity, calcium, magnesium, fruit length, diameter, weight, pulp content, number of branches, and number of clusters. Similarly, IIHR 2357 recorded higher TSS, lycopene, magnesium, number of fruits and yield/plant and IIHR 2358 exhibited higher ascorbic acid, fruit rmness, pericarp thickness, shelf life, and plant height. Among the testers, Arka Vikas recorded better mean for TSS, ascorbic acid, titratable acidity, magnesium, fruit diameter, locule number, number of clusters, number of fruits, and yield/plant. Arka Ahuti recorded higher pH, fruit length, rmness, pericarp thickness, pulp content, shelf life, and number of branches. Arka Saurabh exhibited higher lycopene, calcium, fruit weight, and plant height.
The hybrids were superior in their performance than parents for lycopene, fruit diameter, weight, rmness, pericarp thickness, locule number, plant height, number of clusters, number of fruits, and yield/plant.    (Table IV). The variance due to crosses was further portioned into variance due to lines, testers and line × testers. The mean sum of squares due to lines was highly signi cant for ascorbic acid, fruit rmness and pulp content whereas, mean sum of squares due to testers was signi cant for pulp content. The Line × Tester interaction variance was highly signi cant for lycopene, ascorbic acid, titratable acidity, calcium, magnesium, fruit length, diameter, weight, number of branches, number of clusters and yield/plant. The per cent contribution of lines towards total variation was higher for ascorbic acid followed by fruit rmness and least for titratable acidity. Similarly, the per cent contribution of tester towards total variation was higher for titratable acidity and least for fruit weight.

Variance due to combining ability effects
The variance due to general combining ability was highly signi cant for ascorbic acid, fruit rmness, and pulp content whereas, variance due to speci c combining ability was signi cant for pH, lycopene, ascorbic acid, titratable acidity, calcium, magnesium, fruit length, weight, plant height, number of clusters, number of fruits/cluster, and yield/plant (Table VI). It was evident that among twenty studied traits most manifested a higher degree of SCA variance as compared to GCA variance. The GCA/SCA ratio was less than one for all the characters except for pH, ascorbic acid, fruit rmness, and plant height.

Speci c combining ability effects and overall speci c combining ability effects of crosses
The hybrids from different combinations of the parents with high or low GCA effects are referred to as H × H (high × high), H × L (high × low), and L × L (low × low) combinations (Table X). Among nine hybrids, IIHR 2349 × Arka Vikas (L × H) was found good speci c combiner for number of clusters, and yield/plant (Table XI). The hybrid, IIHR 2349 × Arka Ahuti (L × L) was a good speci c combiner for calcium, magnesium, fruit length, and weight in a desirable direction. For lycopene, ascorbic acid, calcium, and magnesium, the hybrid IIHR 2349 × Arka Saurabh (L × L) was found superior in a desirable direction. Concerning titratable acidity, the hybrid IIHR 2357 × Arka Ahuti (H × L), was the best speci c combiner as it exhibited high positive signi cant SCA effects. As far as ascorbic acid, fruit length, and number of fruits/cluster concerned, the hybrid IIHR 2357 × Arka Saurabh (H × L) was the top speci c combiners with SCA effects in a desirable direction.
Out of nine crosses, four crosses had high overall speci c combining ability status. The crosses IIHR 2349 × Arka Vikas (H × L), IIHR 2349 × Arka Saurabh (L × L) were the best overall speci c combiner followed by IIHR 2358 × Arka Ahuti (L × L) and IIHR 2357 × Arka Ahuti (H × L) (Table XII). Based on SCA effects, crosses were classi ed into H × H (both the parents with high overall GCA status), H × L (one parent with high and other with low overall GCA status), and L × L (both the parents with low overall GCA status). Out of nine crosses, one cross was H × H (IIHR 2357 × Arka Vikas) involved both the parents with good overall general combining ability. Four crosses were H × L or L × H and the remaining four crosses were L × L.   Where, H = High overall GCA status, L = Low overall GCA status  The signi cance of mean sum of squares due to lines indicated the signi cance of additive variance for ascorbic acid, fruit rmness, and pulp content. The signi cance of Line × Tester interaction variance indicated the signi cance of dominance variance and the presence of heterosis for lycopene, ascorbic acid, titratable acidity, calcium, magnesium, fruit length, diameter, weight, number of branches, number of clusters, and yield/plant. The combining ability of lines differed based on the tester involved in the cross combination which was amply re ected by signi cant differences for crosses.
The lines contributed to the bulk of the variation observed in hybrids which were higher for TSS, ascorbic acid, calcium, magnesium, fruit length, diameter, weight, rmness, pulp content, shelf life, number of branches, number of clusters, number of fruits and yield/plant. Lines have an important role in the variation among the hybrids, while selecting the germplasm lines, emphasis should be given to the lines which are used as female parents. However, the same thing cannot be done concerning line × tester interaction as it is not under the control of the breeder, but it depends on the speci c manner in which the lines and testers interact. Therefore, most care should be taken while selecting the lines for hybridization.
The predominance of non-additive gene action exhibited by lycopene, titratable acidity, pericarp thickness (Dhatt et al. 2003;Joshi et al. 2005), yield/plant (Bhutani and Kalloo 1983;Das et al. 2020;Pavan and Gangaprasad 2022), TSS (Kalloo et al. 1974;Bhatt et al. 2001), calcium, magnesium, fruit length, diameter, weight, pulp content, locule number (Dhatt et al. 2003), shelf life (Roopa et al. 2001;Dhatt et al. 2003;Garg et al. 2008;Pavan and Gangaprasad 2022), number of branches, number of clusters, number of fruit/cluster (Pavan and Gangaprasad, 2022) which can be exploited through hybrid development programs as the GCA/SCA ratio was less than one. Contrasting ndings of additive gene action were cited for locule number, shelf life (Rodriguez et al., 2010), yield/plant (Katoch and Vidyasagar 2004), number of fruit/cluster (Mondal et al. 2009), and lycopene (Suo et al. 2010). The additive and non-additive gene effects for fruit quality and yield traits were reported by Gaikwad and Cheema (2009) and Akhtar and Hazra (2013). pH, ascorbic acid, fruit rmness, and plant height were governed by additive gene action which can be exploited through varietal development programs as the GCA/SCA ratio was more than one. The contrasting results of dominant gene action were reported for pH and plant height (Das et al. 2020;Pavan and Gangaprasad 2022), ascorbic acid, and fruit rmness (Bhatt et al. 2001;Pavan and Gangaprasad 2022).
The per se performance of a parent is not always a true indicator of its potentiality in a hybrid combination. Therefore, general combining ability, which is the breeding value of the parent expressed as a deviation from the population mean has proved as a useful tool for choosing the potential parents for hybrid development (Singh and Asati 2011). Among the parents with signi cant GCA effects, the one with a higher magnitude of GCA effects is considered superior to those with a lower magnitude (Technow 2019;Yu et al. 2020). An overall appraisal of GCA effects revealed that all the lines were good general combiners for lycopene, ascorbic acid, calcium, magnesium, number of clusters, and number of fruits/cluster. The Arka Saurabh had high general combining ability effects for pH, fruit length, plant height, number of clusters, and number of fruits/cluster proved to be the best combiner in producing heterotic hybrids. The high overall GCA status of the line IIHR 2357 and the tester Arka Vikas revealed that 33.33% of lines and testers proved to be high overall good general combiners which intern suggested their ability to transmit additive genes in the desired direction for the majority of the traits under study. However, these lines and testers should be evaluated further to con rm their superiority. Similar results were reported by Bhatt et al. (2001), Dhatt et al. (2003), Kansouh and Zakher (2011), Yogendra and Gowda (2013).
The speci c combining ability is used to designate those cases in which speci c combinations do relatively better or worse than would be expected from the average performance of the lines involved. The SCA is controlled by dominance and nonallelic gene interactions. The high GCA effects of parents may produce hybrids with low SCA effects and vice versa. Therefore, when selecting elite parents for crosses, understanding and accounting for the relationship between the GCA of the parents, the SCA of the crosses and the dominant type of combining ability are key to improving breeding e ciency based on combining ability (Liu et al. 2019). Both the crosses, IIHR 2349 × Arka Ahuti (L × L) and IIHR 2349 × Arka Saurabh (L × L) involved low combiners which showed the importance of overdominance and epistasis in the inheritance o f lycopene, ascorbic acid, calcium, magnesium, fruit length, and weight.
Involvement of high and low combiners in the crossing revealed the signi cance of non-additive gene action governing the traits. Hence, hybrid development will be effective. The ndings of Singh and Asati (2011), Yadav et al. (2013), andAdhi Shankar et al. (2014) would substantially support the present results.
Out of nine crosses, four crosses had high overall speci c combining ability status indicated that 45.55% of hybrids were assigned high (H) overall speci c combiners. The cross IIHR 2357 × Arka Vikas, involved both the parents with good overall general combining ability (H × H). Therefore additive gene action may be imperative and simple selection in segregating generation will be effective and the reliance should be placed on mass selection and progeny selection. Four crosses were H × L or L × H indicated the presence of non-additive gene action which is a prerequisite for launching a heterosis breeding. The remaining four crosses were L × L showed the signi cance of overdominance and epistatic gene action in the inheritance. Hence, intensive selection for SCA in segregating generations will be effective in the genetic improvement of fruit quality traits governing shelf life.
The two promising hybrids were identi ed based on their per se performance for yield and shelf life. The hybrid IIHR 2349 × Arka Vikas recorded a signi cantly higher yield potential of 2850.50 g/plant with signi cantly higher standard heterosis over standard check Arka Rakshak (52.27%), and it had 24.70 days of shelf life. Higher heterosis for yield/plant in this hybrid may be attributed to higher mean performance for number of clusters/plant (11.05) which is further reiterated by signi cantly higher standard heterosis for the same trait (8.87%). Another promising hybrid IIHR 2349 × Arka Saurabh recorded a signi cantly higher shelf life of 29.40 days with signi cantly higher standard heterosis over standard check Arka Rakshak (26.45%), and it had a yield potential of 2397.50 g/plant. Higher heterosis for shelf life in this hybrid may be contributed by signi cantly higher mean performance for fruit magnesium content (17.79 mg/100g) and lycopene content (4.59 mg/100g) which is further evident by signi cantly higher standard heterosis for magnesium content (154.14%) and lycopene content (128.36%). These promising hybrids need to be tested at preliminary yield trials and multilocation trials to assess their performance and stability before commercialization.

Conclusions
Page 17/19 The crosses, IIHR 2349 × Arka Vikas and IIHR 2349 × Arka Saurabh were promising hybrids for high yield and shelf life. Signi cant differences among the genotypes indicated considerable genetic variability for fruit quality traits. The estimates of SCA variance was higher as compared to GCA variance. The non-additive gene action was exhibited for lycopene, titratable acidity, TSS, fruit length, diameter, weight, locule number, number of branches, number of clusters, number of fruit/cluster and yield/plant which can be exploited through hybrid development and additive gene action shown for pH, ascorbic acid, fruit rmness, and plant height can be exploited through varietal development. The study underlined the signi cance of non-additive gene action for shelf life, pericarp thickness, calcium, magnesium, pulp content. The involvement of low combiners in hybridization signi ed over dominance and epistasis hence intensive selection for SCA in segregating generations will be effective. Involvement of high and low combiners in the crossing revealed the signi cance of non-additive gene action which is a prerequisite for heterosis breeding. The involvement of high combiners in crossing signi ed the importance of additive gene action and simple selection in segregating generation will be effective with more reliance on mass selection and progeny selection.