Introduction

Genetic diversity is the base of all crop improvement programmes. Genetic diversity is related to manifestation of heterosis (Huyen, 2016), an important genetic phenomenon, which results into yield improvement, quality enhancement, elevated tolerance towards biotic and abiotic-stresses, and increased resilience to changing climatic conditions (Bhandari et al., 2017). The diverse lines are of special importance in plant breeding particularly in context of unpredictable climate change (FAO, 2015) as a source of favourable gene complexes conferring buffering capacity towards new threats. The genetic diversity or ‘within-species’ diversity confers the survival and adaptability (Rao and Hodgkin, 2002) in variable environments including different locations or different years in a particular location. The major strides in improvement of different crops have been possible as a result of utilization of genetic diversity existing in different crops. However, the commercial varieties of different crops released so far, especially in Indian context, have considerably narrow genetic base resulting into plateaued yield. This necessitates collection and evaluation of new germplasms and their judicious utilization in crop improvement programmes.

Conventionally, genetic diversity is assessed using phenotypic traits and quantitatively assessed using multivariate techniques like D² analysis and Principal Component Analysis. Genetic diversity assessment using phenotypic evaluation is easy and cheap as it does not require sophisticated tools and preparatory techniques (Sinha and Kumaravadivel, 2015). However, phenotypic measure of genetic diversity suffers from the limitation of few morphological markers and being influenced by environmental factors. In recent years, DNA markers are being increasingly used for assessment of diversity existing among genotypes at molecular level (Fu, 2015). These are considered more efficient on account of being neutral to environment influences (Ferreira, 2006). However DNA marker-based diversity analysis suffers from the limitation of being dependent on only one feature i.e. genomic sequences unlike phenotypic evaluation, which is based on several morphological and developmental features of the plant.

Both the methods of diversity analysis have their own benefits and limitations. Many reports are available detailing coherence of these two types of diversity in different crops. However, their comparative efficiency in realizing the combining abilities and heterosis for important economic traits are lacking. Sufficient reports are available indicating association between heterosis and phenotypic (Pandey et al., 2015) or molecular diversity individually. Simultaneous comparison of association between heterosis vs phenotypic diversity/molecular diversity is nowhere reported. Hence, the present investigation is an attempt to compare phenotypic- and molecular diversity in relation to gene effects and heterosis. This type of study can be helpful for plant breeder in deciding the diversity analysis tool to be used for achieving better performances in resultant hybrids.

Materials and Methods

The test material consisted of 66 genotypes of tomato including Indian and exotic lines. The tomato germplasm lines were evaluated in rabi, 2014 at Vegetable Farm, Institute of Agricultural Sciences, Banaras Hindu University, Varanasi (India). The experimental site is located in the middle Gangetic plains in the Eastern part of the state of Uttar Pradesh at 25°19” 59” N latitude, 83°00’ 00” E longitude and at elevation of 77 m above mean sea level. Phenotypic diversity was assessed using D² analysis (Mahalanobis, 1936) based on nine important economic traits viz., days to 50% flowering, plant height, number of primary branches plant^-1, number of fruits cluster^-1, number of fruits plant^-1, average fruit weight, number of locules fruit^-1, number of seeds fruit^-1 and fruit yield plant^-1. Molecular diversity among germplasm lines was assessed using 29 SSR primers covering all the 12 chromosomes. Clustering was done using the symmetric matrix of similarity coefficient and cluster obtained based on unweighted pair group with arithmetic mean (UPGMA) using SHAN module of NTSYS software version 2.0 (Rohlf, 1998).

Based on these studies, two sets of dendrograms were developed. The tomato germplasm (66) were clustered into five and seven groups based on D² analysis and molecular diversity analysis, respectively. One parent from each cluster from two dendrograms was selected separately and two sets of diallel crosses excluding reciprocals were attempted. The resultant progenies were termed as 1^st set for phenotypic diversity-based progenies and as 2^nd set for molecular diversity-based progenies. Both the sets of hybrids along with respective parents and a commercial hybrid check To 3150, (A Syngenta Private Ltd. hybrid.) were evaluated at three locations (Pratapgarh, Varanasi and Burdwan) for a period of two years (2015-16 and 2016-17). The locations Pratapgarh and Varanasi represented middle Gangetic plain zone, and the Burdwan location represented lower Gangetic plain zone of India. The data over two years per location were pooled and analysed.

The combining ability analysis was carried out by the procedure given by Griffing (1956). Commercial heterosis was estimated using the formulae as per Kempthorne (1957):

A relative term ‘Desirability Quotient’is being introduced in this study which can be used for comparing the two sets of parents/crosses for GCA/SCA effect and commercial heterosis. The Desirability quotient was calculated as below:

DQ=µ×p×100

Where µ=Mean of significant GCA effect, SCA effect or heterosis (in desirable direction), and p=Proportion of parents/crosses revealing significance (in desirable direction).

Results and Discussion

The diversity analysis using D² statistic and SSR markers grouped 66 genotypes into five and seven clusters, respectively. Accordingly, five and seven parents were selected from two dendrograms representing different clusters and half-diallel crosses were made. Desirability quotient of GCA effect, SCA effect and commercial heterosis for two sets of parents was calculated for six important yield traits viz., plant height, number of primary branches plant^-1, number of fruits plant^-1, average fruit weight, number of seeds fruit^-1 and fruit yield plant^-1. The results are presented in following paragraphs:

3.1.  General combining ability effect

The 1^st set of parents (phenotypically diverse parents) recorded higher desirability quotient for all six traits viz., plant height, number of primary branches plant^-1, number of fruits plant^-1_, number of seeds fruit^-1 and fruit yield plant^-1in all the locations (Table 1). Present investigation revealed that phenotypically diverse parents possessed higher desirability quotient of GCA effects than that of molecular diversity-based parents for all the traits in all locations. This indicated that the parents selected based on phenotypic diversity possessed higher proportion of additive gene action while, molecular diversity-based parents possessed higher proportion of non-additive gene action.

Table 1: Desirability quotient of gca effects for two sets of parents for different yield traits

3.2.  Specific combining ability effect

The 2^nd set of hybrids exhibited higher desirability quotient of SCA effect than that of 1^st set of hybrids for plant height in all the locations. For the trait, number of primary branches plant^-1, higher desirability quotient of SCA effect was noted by 2^nd set of crosses at Pratapgarh only (Table 2). These behavioural differences may be ascribed to differential expression of genotypes in different location. For the trait number of fruits plant^-1, 1^st set of hybrids recorded higher desirability quotient in all the locations. The 2^nd set of hybrids recorded higher desirability quotient of SCA effect for average fruit weight (at Varanasi and Burdwan) and number of seeds fruit^-1 (at Pratapgarh). For the trait fruit yield plant^-1, 2^nd set of hybrids recorded higher desirability quotient for SCA effect in all the locations. The molecular diversity-based hybrids recorded higher desirability quotient of SCA effect for plant height and fruit yield plant^-1 universally. In contrast, phenotypic diversity based hybrids recorded higher desirability quotient of SCA effect for number of fruits plant^-1 in all the locations.

Table 2: Desirability quotient of sca effects for two sets of crosses for different yield traits

3.3.  Heterosis

The 2^nd set of hybrids revealed higher desirability quotient than that of 1^st set of hybrids for plant height at Varanasi location only (Table 3). The 1^st set of hybrids revealed higher desirability quotient for number of primary branches plant^-1 and number of fruits plant^-1 in all the locations. In contrast, 2^nd set of hybrids revealed universally higher desirability quotient of heterosis for economically important traits like number of seeds fruit^-1, average fruit weight and fruit yield plant^-1. For the traits average fruit weight and fruit yield plant^-1, none of the crosses among 1^st set revealed desirable heterosis in any location when compared to the commercial check. As a result, 1^st set of crosses recorded null value of desirability quotient for these traits. Consequently, 2^nd set of hybrids consistently recorded higher desirability quotient than that of 1^st set of hybrids for these two traits.

Table 3: Desirability quotient of commercial heterosis for two sets of crosses for different yield traits

Higher desirability quotient of heterosis for major economic traits like number of seeds fruit^-1, average fruit weight and fruit yield plant^-1 was noted by crosses derived from crossing molecular diverse parents. This may be attributed to the specific heterozygosity of positive markers (Pheirim, 2017). The 1^st set of hybrids exhibited higher desirability quotient of heterosis for plant height, number of primary branches plant^-1 and number of fruits plant^-1. The 2^nd set of hybrids consistently recorded higher desirability quotient of heterosis for important yield traits namely average fruit weight, number of seeds fruit^-1 and fruit yield plant^-1. The higher desirability quotient of heterosis for 1^st set of hybrids for number of fruits plant^-1 in all locations may be attributed to the higher desirability quotient of SCA effect for the trait. The 1^st set of hybrids failed to record higher desirability quotient of heterosis for fruit yield plant^-1 despite higher desirability quotient for number of fruits plant^-1 indicating non-coherence between these traits. The better score of desirability quotient of commercial heterosis for number of seeds fruit^-1 and fruit yield plant^-1 may be linked to better score of average fruit weight. The study indicates that average fruit weight might have more contribution towards fruit yield.

The present investigation revealed higher efficacy of molecular markers in identifying diverse groups possessing higher non-additive gene action and ultimately the higher heterosis for important yield traits in tomato.  This is in accordance to the reports of Pandey et al. (2015) who established the necessity of molecular markers in determination of genetic diversity in context of non-consistent correlation between phenotypic diversity and heterosis. Similarly, Sud et al. (2010) indicated strong association of yield heterosis with SSR-based genetic diversity.  In contrast, Cebolla-Cornejo et al. (2013) emphasized the need of morpho-agronomical and quality characterization over molecular characterization in the ex-situ management of genetic resources in Spanish tomato landraces. Ezekiel et al. (2011) reported higher efficiency of molecular markers over phenotypic markers in determination of varietal identity. The present study establishes that fruit weight has greater association with fruit yield in tomato and is supported by reports of de Souza et al. (2012). In addition, the study indicated higher efficacy of molecular markers in identifying genotypes to be used in hybrid breeding programme.

Conclusion

The present study indicated higher efficiency of molecular diversity than that of phenotypic markers in attaining heterosis for yield traits in tomato.

Reference

Bhandari, H.R., Bhanu, A.N., Srivastava, K., Singh, M.N., Shreya, Hemnataranjan, A., 2017. Assessment of genetic diversity in crop plants-An overview. Advances in Plants & Agriculture Research7 (3), 279–286.

Cebolla-Cornejo, J., Rosello, S., Nuez, F., 2013. Phenotypic and genetic diversity of Spanish tomato landraces. Scientia Horticulturae162 (23), 150–164.

deSouza, L.M., Melo, P.C.T., Luders, R.R., Melo, A.M.T., (2012). Correlations between yield and fruit quality characteristics of fresh market tomatoes. Horticultura Brasileira30, 627–631.

Ezekiel, C.N., Nwangburuka, C. C., Ajibade, O. A. Odebode, A.C., 2011. Genetic diversity in 14 tomato (Lycopersicon esculentum Mill.) varieties in Nigerian markets by RAPD-PCR technique. African Journal of Biotechnology 10(25), 4961–4967.

FAO, 2015. Coping with climate change -The roles of genetic resources for food and agriculture. Rome, 109.

Ferreira, E.M., 2006. Molecular analysis of gene banks for sustainable conservation and increased use of crop genetic resources. In: Ruane, J., Sonnino, A. (Eds.), The role of biotechnology in exploring and protecting agricultural genetic resources. Food and Agriculture Organization of the United Nations, Rome, 121–128.

Fu, Yong-Bi., 2015. Understanding crop genetic diversity under modern plant breeding. Theoretical and Applied Genetics 128(11), 2131–2142.

Griffing, B., 1956. Concept of general and specific combining ability in relation to diallel system. Australian Journal of Biological Sciences 9, 463–493.

Huyen, N.T., 2016. Another explanation for the cause of heterosis. Journal of Genetics 95, 1065–1072.

Kempthorne, O., 1957. An introduction to genetic statistics. John Wiley and Sons. Incorporation, New York, 545.

Mahalanobis, B.C., 1936. On the generalized distance in statistics. Proceedings of National Academy of Sciences, India. 2, 49–55.

Pandey, P., Pandey, V.R., Kumar, A., Yadav, S., Tiwari, D., Kumar, R., Relationship between heterosis and genetic diversity in Indian pigeonpea [Cajanus cajan (L.) Millspaugh] accessions using multivariate cluster analysis and heterotic grouping. Australian Journal of Crop Science 9(6), 494–503.

Pandey, P., Pandey, V.R., Kumar,A., Yadav, S., Tiwari, D., Kumar, R., 2015. Relationship between heterosis and genetic diversity in Indian pigeon pea [Cajanus cajan (L.)Millspaugh] accessions using multivariate cluster analysis and heterotic grouping. Australian Journal of Crop Science 9(6), 494–503.

Pheirim, R., Niyaria, R., Singh, P.K., 2017. Heterosis prediction through molecular markers. Rising 1(1), 45–50.

Rao, V.R., Hodgkin, T., 2002. Genetic diversity and conservation and utilization of plant genetic resources. Plant Cell, Tissue and Organ Culture 68, 1–19.

Rohlf, F.J., 1998. NTSYSpc Numerical Taxonomy and Multivariate Analysis System Version 2.0 User Guide. Applied Biostatistics Inc., Setauket, New York.

Sinha, S., Kumaravadivel, N., 2015. Genetic Diversity Study in Sorghum Using Quantitative Morphological Traits. International Journal of Bio-resource and Stress Management 6(5), 552–557.

Sud, S., Bains, N.S., Nanda, G.S., Singh, K., Arya, L., 2010. Genetic diversity and heterosis in wheat. SABRAO Journal of Breeding and Genetics 42(2), 95–104.