Study of Heterosis, Combining Ability and Parental Diversity for Seed Cotton Yield and Contributing Traits using Diallel Data in Cotton (G. hirsutum L.)

Cotton is the foremost natural fibre crop of global importance grown commercially in around 75 countries with 2.9 million household engaged in its production, accounting for 25.9 million bales (170 kg) of lint production (DNFI, 2020). Cotton plays a key role in the national economy by way of its contribution in trade, industry, employment and foreign exchange earnings. India has largest area under cotton cultivation i.e. 11.6 mha, which is one third of the total world acreage (33.98 mha). India occupies first position in production followed by USA and China (Statista, 2019). The average productivity of cotton in India is very low compare to the other cotton-growing nations of the world so the knowledge of parental divergence and its impact on impact on yield and contributing traits can enhance the genetic improvement efforts for yield enhancement. The genetic improvement efforts limited by the low germplasm diversity in the upland cotton.
       
Upland cotton (Gossypium hirsutum L.) has a narrow genetic base due to genetic bottlenecks associated with polyploidy, domestication and modern plant breeding practices (Bowman et al., 1996; Iqbal et al., 2001). The low genetic base cotton has been validated by using molecular marker studies. Tyagi et al., (2014) have shown over 90% similarity based on a wide selection of several hundred cultivars and germplasm lines using simple sequence repeat (SSR) marker-based genotyping. Several other workers Lewis (2001) and May et al., (1995) suggested that narrow genetic diversity in cotton to impact the improvements in lint yield and fiber quality.
       
Looking into the challenge of narrow genetic base it is very much desirable to select germplasm from different geographical areas which exhibit maximum diversity. To improve productivity, one of the most important steps in a breeding programme is the detection of suitable parents which can produce heterosis. For the selected parents it is crucial be good combiner, which is required to gets its potential represented in hybrid. In combining ability, the entire genetic variability of each trait can be partitioned into GCA and SCA as defined by Sprague and Tatum (1942) and reciprocal effects as sketched by Griffing’s (1956). They stated that GCA effects administer the additive type of gene action whereas the SCA effects are shown due to genes which are non-additive (dominant or epistatic) in nature. Sayal et al., (1997) and Hassan et al., (1999) reported the importance of non-additive type of gene action for different cotton traits.

The parental material multiplication and crossing program for the research was undertaken at ICAR-CICR, Regional Research Station, Sirsa. The experimental material was comprised of eight G. hirsutum lines viz., RS-2013, RST-9, RS-810, F-1378, F-2164, F-2228, LH-2076 and LH-2108. These eight parents were crossed in diallel manner with reciprocals to produce 56 F₁ hybrids at ICAR-CICR Regional Research Station, Sirsa during kharif-2016. All the 56 F₁crosses along with eight parents were planted at three farmer field locations at Sirsa, Bathinda and Abhor. The entries were tested in randomized block design with a spacing of 105 × 60 cm row to row and plant to plant spacing respectively during the 2017-18 season. Data for different traits under study was randomly collected from five plants from each plot.
       
The Trial data was subjected to analysis of variance (ANOVA) using OP STAT computer software (Sheoran et al., 1998) to test the null hypothesis of no difference between various F₁ crosses and their parental lines. Estimates of GCA were computed according to (Griffing, 1956). The diallel analysis was used to evaluate traits that had a significant variation among parents. Simple additive-dominance model approach (Hayman, 1954, 1958) modified by Mather and Jinks (1982) was followed for genetic analysis and for estimation of the components of genetic variation.
       
The parental diversity was taken out using 20 Simple Sequence Repeats (SSR) markers. The DNA of the cotton samples were extracted from the tender leaves (collected from the testing plots during kharif 2017) using CTAB method (Zhang and Stewart, 2000).The purity and concentration of DNA were determined by agarose gel electrophoresis and spectrophotometric analysis. All DNA samples were diluted to a working concentration (50 ng/μL). Stock DNA samples were stored at -40°C and working DNA samples at 4°C until PCR amplification. This DNA sample was used to hybridize with the SSR markets. The similarity study was done using Jaccard coefficient. The Jaccard coefficient values were compared to calculate the similarity of dissimilarity between parents. A dendrogram based on these similarity coefficients was constructed by using unweighted pair group method of arithmetic means (UPGMA).
   
Heterosis
 
Heterotic percentages relative to the mid parents (Average heterosis) and the standard checks (Standard heterosis) for different characters were calculated using the procedure illustrated by Mather and Jinks (1971) as follows-

                                                                                                                                                                                 
 
Where
F₁ is the mean performance of the cross, MP is mean of the two inbred parents and SC is mean of the Standard check varieties.

Analysis of variance indicated highly significant differences due to treatments fulfilling the basic requirement to take the study forward (Table 1). The result suggests that there is enough variability in genetic material. The ANOVA for combining ability effects (Table 2) indicated that the mean square values of GCA were highly significant (p≤0.01) for all the traits except for GOT. The mean square values for SCA were also significant for all the traits except for the number of sympods. Though the maternal effects are non-significant for all the traits under study, but maternal interaction were seen significant for all the traits except for the number of sympods. Mean square values due to reciprocals were also found significant for all the traits except for boll weight.
 

 

 
Mean performance
 
Mean performance acts as the main criterion in selecting better hybrids as it reveals their real value. Shimna and Ravikesavan (2008) suggested that the per se performance of hybrids appeared to be a useful index in judging them. Gilbert (1958) reported that parents with good per se performance would result in good hybrids. Table 3 gives the trait means for the straight cross combinations as well as for reciprocal cross combinations. For number of monopods per plant straight and reciprocal crosses shared common mean whereas the trait range for reciprocal crosses was broader than straight crosses range. Number of sympods trait mean for straight and reciprocals was 12.6 and 12.2 respectively with straight crosses sharing a broader range than reciprocals. Boll weight trait mean and range was similar for straight and reciprocal crosses. For plant height the trait mean range was found highest among all the traits. Trait mean and range for straight crosses was 123.2 cm and range were107.0-137.7 cm whereas for reciprocal crosses mean and range was 125.9 cm and 110.1-137.8 cm respectively. Seed cotton yield exhibited the range between 1.135 - 1.894 kg/plot with mean of 1.500 kg/plot for straight crosses and 1.168 - 1.997 kg/plot and 1.613 kg/plot for reciprocal crosses. For fiber traits the trait means and trait ranges were very similar for straight as well as reciprocal cross combinations. The maximum range among fiber traits were observed for fiber strength followed by fiber length.
 

 
Average and standard heterosis
 
Heterotic effects were observed for all the ten given traits over mid and standard check but the extent of heterosis varied from trait to trait. Heterotic effects of straight as well as for reciprocal cross combination was computed over mid parent was termed as Average Heterosis and over the standard check was termed as Standard Heterosis. For the straight and reciprocal cross combinations, the mean heterotic effects are presented in Table 4. Average heterosis was found positive for all the traits in both straight and reciprocal cross combinations except for number of sympods for which mean for reciprocal effects was found to be negative. Several workers have reported similar type of results for economic traits in cotton (Choudhary et al., 2014, Solanki et al., 2014, Srinivas and Bhadru 2015).
 

       
Average heterosis mean for Number of Monopods for straight and reciprocal combinations crosses was found to be equal (2.2% in each case). For number of sympods mean heterotic effects were positive for straight crosses (0.6%) but it was found negative for reciprocal crosses (-2.0%). Maximum average heterosis was observed for seed cotton yield (70.9%, 83.2%) for straight and reciprocal crosses. Plant height followed seed cotton yield for leveraging wide range of heterosis (75.2%, 79.0%) for straight and reciprocal crosses. Positive average heterotic effects were obtained for both straight and reciprocal crosses.
       
For standard heterosis study, out of ten traits under study positive heterotic effects were found for fiber strength for both straight and reciprocals whereas for seed cotton yield positive effects were found for reciprocal combinations and GOT positive effects were exhibited by straight crosses only. Trait number of monopods per plant exhibited same extent for heterotic effects for straight and reciprocal crosses. Maximum standard heterosis was observed for seed cotton yield (-0.3%, 6.9%) followed by fiber strength (4.2%, 4.8%). Similar types of trends for standard heterosis were also found by earlier researchers (Ashok et al., 2013, Patel et al., 2014 and Usharani et al., 2015, Sawarkar et al., 2015, Hussain et al., 2020).
 
Genetic diversity of parents
 
Parents for this study were collected from 1) PAU Ludhiana, 2) RAU Regional Cotton Station Sri Ganganagar and 3) Regional Cotton Research Station Faridkot. The selection of parents was done based on their diversity for different agronomic and phenotypic traits. The origin and phenotypic traits of the parental lines are proved in Table 5. Based on SSR markers the distance matrix of eight parents (Table 8) revealed that the two parental lines P2 (RST-9) and P7 (LH-2076) had a genetic distance (GD) of 0.500, the highest among the eight genotypes used in this study. RST-9 also had higher GD (ranging from 0.0.462 to 0.500) with other lines. The second largest genetic distance was observed for P1 (RS-2013) with a range of 0.333 to 0.429 with all the other parental line used in the study. Parent P3 (RS 810), P4 (F 1378), P5 (F 2164) and P6 (F-2228) were most closely related or were found to be sister lines as the genetic distance between all the four were zero. Three of the four lines were originated from Faridkot (Punjab) while parent P3 has origin from Sri Ganganagar (Rajasthan). Parent P7 (LH 2076) was closely related with P3, P4, P5 P6 and P8 with GD of 0.071 whereas P8 (LH 2108) also depicted close relation with P3, P4, P5 and P6 with GD of 0.143. The close relatedness of Parents was understandable as they were all derived from Punjab except for P2. Parent P1 (RS 2013) and P2 (RST-9) highest GD from parental originated from Ludhiana (Punjab) followed by parental lines originated from Faridkot (Punjab). Tyagi et al., (2014), Lewis (2001) and May et al., (1995) also suggested narrow genetic diversity in cotton.
 

 

 
The results of hybrid performance per se shows a positive correlation between GD of parents and the performance of their hybrids for Seed cotton yield, Plant height, GOT and number of Monopods per plant. For these traits top 40 - 60% of the top hybrids involved distant parents (DP) i.e. RST 9 and RS-2013 as one of the parents. For number of monopods the distant parents contributed 6 hybrids in top 10 hybrids among straight crosses and 5 hybrids in reciprocal crosses. For plant height DP contributed 4 and 6 crosses among top 10 crosses. For seed cotton yield 4 and 5 crosses among top 10 hybrids constituted DP for straight and reciprocal crosses respectively whereas for GOT 4 and 6 cross combinations of DP were part of top ten best heterotic crosses for straight and reciprocal crosses. Further it was found that among the list of Top three best heterotic crosses for all the ten traits for straight crosses, out of total 30, 26 crosses have top four most diverse lines as one or both parents (Table 6). For Reciprocal crosses among the top three crosses for all the traits, top four diverse parents contributed towards 22 crosses out of total 30 (Table 7). Apart from these results it was also found that there were crosses in the top performer where the closely related parents were involved in hybrid production. This was in line with the earlier studies of Altaher and Singh (2003) and Kulkarni and Nanda (2006) where they found that moderate divergence can also produce good heterosis. Zhang et al., (2017) also found that the additive effects of parents contribute hybrid performance.
 

 

       
 Parental data for GCA indicates that Parent 6 (F 2228) is the best general combiner followed by Parent 8 (LH 2108) and Parent 5 (F 2164) But as per the distance matrix (Table 8) the best combiner parent (F 2228) is least diverse (0 to 0.462) from other parents involved whereas second best GCA parent (LH 2108) has medium divergence ranging from 0.071 to 0.462. Further the combining ability status of two most divergent parents i.e. P1 (RS 2013) and P2 (RST 9) was the lowest combining abilities, having positive combining ability effects for only one and three traits respectively. These results show that there is no correlation between the genetic diversity and combining ability of lines. Kamra et al., (2020) also found no correlation between combining ability and parental diversity in maize. The combining ability of the parents was reflected in better positive heterosis than the poor general combiner parents.

It was found that the diversity among the parents involved in the study was less and it was very visible that parents from different institutes/geography were distant from each other. Parents with good general combiner produce better positive heterosis compare to poor general combiner. The best combiners for different traits also contributed around 40-50% of the top ten most heterotic hybrids. RS 2013 and RST 9 were most diverse among all the eight parental lines. The genetic diversity is correlated positively with heterosis for Seed cotton yield, Plant height, GOT and number of Monopods per plant. Genetically diverse lines produced significant crosses among the top three best heterotic crosses for all the ten traits under study. The genetically diverse lines were not found to be good general combiner which indicated towards no correlation between divergence and combining abilities. It was also visible from the study that the heterotic effects of reciprocal cross combinations were similar in extent indicating absence of maternal effects.