Appraisal of Genetics and Heterosis of Important Traits in Chilli Pepper Cultivated under the Influence of Chilli Leaf Curl Virus Disease

Chilli leaf curl virus (ChiLCV) is a severe threat to the successful cultivation of the chilli peppers. Besides, the genetics and heterosis for important traits under the influence of ChiLCV infestation is not well studied. Therefore here, were crossed seven genotypes (lines) with high yield and quality attributes with four highly resistant genotypes (testers) in a line (L) × tester (T) mating design to produce 28 F1 hybrids. Parents and their hybrids were evaluated for yield attributes and ChiLCV resistance. All the four testers were symptomless and among seven lines, two were moderately resistant and five were moderately susceptible to the ChiLCV. Among 28 F1 hybrids, 12 showed a moderate resistant reaction, 11 were moderately susceptible and five susceptible to the ChiLCV disease. Based on per se performance, most promising hybrid combinations were L3 × T2, L6 × T1, and L3 × T1 for yield attributes. Whereas the superior crosses for ChiLCV resistance based on heterobeltosis, standard heterosis and specific combining ability test (SCA) effects were L6 × T1 and L7 × T3. Lines and testers showed significant differences for all the characters except for the plant height. The general combining ability (GCA) effects for testers were significant for all the traits except for days to first harvest. The ratio of σ2GCA/σ2SCA indicated the predominance of non-additive gene effects in the inheritance of traits studied in this experiment. The contribution of lines was more compared to testers for all the characters except for primary branches plant1. Overall, this study provides valuable information regarding the genetics and heterosis in chilli with respect to ChiLCV infestation.

improvement of diverse quantitative traits [24]. Gene action is commonly measured in terms of components of genetic variance or combining ability effects and variances. The varieties or strains can be evaluated in several ways based on the combining ability of their parents and one of them is line × tester analysis [25,26]. By using this analysis promising lines could be selected from the germplasm. As compared to diallel technique, this approach could evaluate a greater number of breeding lines at once. This, in turn, suggests the breeder whether to go for F1 hybrid development or selection in subsequent generations to realize homozygous promising lines [27]. The information on combining ability effects (general and specific combining ability) are useful for the precise interpretation of the genetic basis of promising traits [28][29][30]. Keeping in view of these facts and need, the present investigation was planned with the following three objectives: 1. To determine the parental genotypes reaction to ChiLCV via natural and artificial screening.
2. Identification of parents with ChiLCV resistance based on mean performance and general combining ability (GCA) effects.
3. To identify superior performing ChiLCV resistant hybrids based on expressed heterosis and specific combining ability (SCA) effects.

Screening of the Parental Genotypes
Among the total 11 parents 7 were lines and 4 were tester genotypes. Out of seven lines, two were moderately resistant, and the remaining five showed a moderate susceptible reaction to ChiLCD ( Table  1). The genotypes (L1 and L4) which showed Moderately susceptible reaction under field conditions were susceptible under artificial graft inoculation (Table 1). In contrast, the coefficient of infection (CI) was highest for the parental genotype L7. Whereas, the lowest values of CI were recorded for the testers (< 5%) ( Table 1). The disease incidence was maximum for the parents ( Table 1). The percent of disease incidence (PDI) was above 69% for the lines and below 10% for the testers (Table 1).
In order to confirm the presence in the graft inoculated plants, the DNA extracted from the top young leaves was subjected to a polymerase chain reaction (PCR) using geminivirus universal primers (AV494/AC1048) for confirmation of ChiLCV (Table 1). All seven lines showed amplification of 560 bp DNA fragment specific to viral genome indicating the presence of the virus ( Figure 1). Whereas, out of 4 testers determined symptomless genotypes under field conditions, none were completely free from ChiLCV infection ( Table 1). All four genotypes showed highly resistant reaction under graft inoculation ( Figure 1). In the four testers, the disease symptoms appeared 32.00 (T1), 34.33 (T2), 33.33 (T3) and 34.33 (T4) days after graft inoculation (Table 1).

Figure 1.
Virus detection in graft inoculated parents (lines and testers) as confirmed by PCR amplification of the viral genome.

Field Screening of Parents and their Hybrids
The results of field screening of parent and their hybrids are presented in Table 2. All the four testers were symptomless under field screening experiment, and among seven lines, two were moderately resistant and the remaining five were moderately susceptible (Table 2). Where, among 28 F1 hybrids, none was completely free from ChiLCV incidence ( Table 2). Twelve hybrids showed moderately resistant reaction and the CI of disease ranged from 13.90% in the cross L3 × T2 to 18.13% in the cross L1 × T3 (Table 2).
Eleven hybrids were moderately susceptible, among them CI ranged from 28.14% in the cross L4 × T3 to 33.88% in the cross L3 × T4 ( Table 2). All of the four testers showed nil values for PDI and disease incidence (DI) ( Table 2). In the case of parents, the lowest value of PDI and DI was recorded for L4 (Table 2). Among  hybrid combinations L1 × T1, L3 × T2 and L6 × T3 were determined as promising with a lower value for  both PDI and DI (Table 2).

Analysis of Variance (ANOVA) for Experimental Design
The mean squares (MS) due to genotypes were highly significant (P≤ 0.01) for all the traits studied ( Table 3). The results further indicated that the MS due to replications were significant for fruit length, fruits plant -1 and fruit weight and non-significant for primary branches plant -1 , plant height, days to the first harvest, days to first flower, fruit girth, yield plant -1 , yield plot -1 , vitamin C, carotenoids and coefficient of infection (Table 3). Table 3. Analysis of variance for thirteen characters in 39 treatments (11 parents and 28 F1 hybrids).

Analysis of Variance (ANOVA) for Combining Ability and its Contribution
The results of ANOVA for combining ability for different traits are presented in Table 4. The MS due to  replication were non-significant for all the studied traits except for plant height, fruit length, fruits plant -1 ,  fruit weight and yield plant -1 (Table 4). Significant differences due to lines were found for all the traits ( Table  4). All of the four testers were significantly different for the studied traits except for the coefficient of infection (Table 4). The hybrids differed substantially for all of traits studied. Likewise, Lines vs Testers showed significant differences for all the traits except for plant height. Similarly, the MS due to parent vs crosses showed significant differences for all the traits ( Table 4). The GCA lines and SCA crosses were highly significant (P≤ 0.01) for all of the studied traits ( Table 4). The GCA testers were observed to be significant for all the traits except for days to first harvest. The ratio of σ 2 GCA/σ 2 SCA was less than unity for all the characters ( Table 5). The contribution of lines in the expression of traits was greater than the testers for all the characters except for the primary branches plant -1 ( Table 5).

General Combining Ability Effects
Based on general combining ability (GCA) effects, the line L1 was identified as a good general combiner for fruit weight, yield plant -1 , yield plot -1 and coefficient of infection (Table 6). L2 was identified as a good general combiner for plant height and carotenoids. Whereas, L3 for fruits plant -1 , yield plant -1 , yield plot -1 and vitamin C (Table 6). In the case of L4, it showed the high GCA effects for days to first flower and harvest, fruit length, fruit girth, carotenoids, vitamin C and coefficient of infection (Table 6). For L5 the high GCA effects were determined for days to first flower and harvest, fruit length, fruit girth and fruit weight; L6 for plant height, fruit girth and fruits plant -1 ( Table 6). The genotype L7 for plant height, fruits plant -1 , yield plant -1 , yield plot -1 , vitamin C and coefficient of infection. Among four testers, T1 was identified as a good general combiner for plant height, primary branches plant -1 , days to first flower, fruits plant -1 , yield plant -1 , yield plot -1 , carotenoids, vitamin C and coefficient of infection ( Table 6). The genotype T2 for fruit length, fruit weight and vitamin C; T3 for fruit girth, vitamin C and coefficient of infection (Table 6).

Specific Combining Ability Effects.
The SCA values for the cross combinations are presented in Table 7. The promising cross combinations identified for plant height included L1 × T2 (7.50), L1 × T1 (6.52), L3 × T1 (5.32) and L2× T3 (5.26) ( Table 7). The cross combinations L1 × T4 (-5.36), L3 × T2 (-3.22) and L7 × T1 (-2.11) were identified promising for days to first harvest (  (Table 8). All of these hybrids showed a significant and a positive SCA effects along with high per se performance suggested the importance of non-additive gene action (Table 7). In this direction, the top hybrid combination for yield plant -1 (L3 × T2) also showed substantial and desirable heterobeltiosis for primary branches, days to flower, days to harvest, fruit length, fruits plant -1 , yield plot -1 , vitamin C, carotenoids and coefficient of infection (Table 8). Whereas, the second beset hybrid, for yield plant -1 (L7 × T1) showed significant and desirable heterobeltiosis for plant height, days to first flower, days to first harvest, fruit length, fruits plant -1 , fruit weight, yield plot -1 , vitamin C, carotenoids and coefficient of infection ( Table 8).
The cross combination L6 × T1 also showed significant and desirable heterobeltiosis for plant height, primary branches plant -1 , days to first flower, days to first harvest, fruit length, fruit girth, fruits plant -1 , yield plot -1 , carotenoids and coefficient of infection (  (Table 8).

Discussion
Begomoviruses are known for infecting a large quantity of economically important dicot plants globally. The genus Begomovirus belongs to the family Geminiviridae vectored by the whitefly, Bemisia tabaci. In chilli pepper, the chilli leaf curl disease on chilli plant has been reported from India. A strain of Chilli leaf curl virus-Pakistan (ChiLCV-PK) was associated with chilli leaf curl disease in south-east Asia [31,32]. Meanwhile, tomato leaf curl Joydebpur virus, reported from tomato in Bangladesh, was also noticed to be associated with chilli leaf curl disease in India [33]. Till date genome sequence of four begomoviruses infecting chilli have been characterised from India viz., chilli leaf curl virus (ChiLCV), tomato leaf curl New Delhi virus (ToLCNDV), tomato leaf curl Joydebpur virus (ToLCJV) and recently chilli leaf curl Palampur virus (ChiLCPV). Their distinctive symptoms in infected plants are easily recognised. The typical symptoms consist of leaf curling, puckering, rolling, shortening of internodes and petioles, blistering of interveinal areas, thickening and swelling of the veins, older leaves turned out to be leathery and brittle, crowding of leaves and stunting of whole plants [14]. Moreover, the typical leaf curl symptoms and an increase in disease severity in infected plants are due to the presence of cognate betasatellites associated with the virus [34].
Identification of resistance sources is of utmost importance in any resistant breeding program. Keeping this in mind, natural field screening seemed best to eliminate the genotypes, which showed visible susceptible reaction under natural epiphytotic conditions [35]. The success of disease resistance breeding depends on the genetic variability and the reliable evaluation tests employed for identification of the resistant sources [36]. It is essential to use the most credible tests of resistance when dealing with destructive diseases like ChiLCV. In this direction, various methods have been employed to screen Capsicum germplasm for resistance to ChiLCV viz., screening under natural epiphytotic conditions and artificial inoculation (grafting inoculation and whitefly mediated inoculation) [37].
Breeding for ChiLCV resistance was started in the late sixties in India and natural field screening was mostly used to identify resistance sources based on disease incidence and severity. We, therefore, employed both natural and artificial screening methods in our research. In the present investigation, following graft inoculation, all ten genotypes showed symptom development. Further, these genotypes were confirmed for the presence of virus by amplification of 560 bp DNA fragment specific to the viral genome. Though the virus is present in all the graft inoculated plants, testers showed some level of resistance. Whereas under filed conditions all the four testers were symptomless and among seven lines, two were moderately resistant and the remaining five were moderately susceptible. Previously, Bhutia et al. (2015) reported the minimum per cent disease index (PDI) of 9.22% [38]. Kumar et al. [39] screened 321 chilli genotypes under field conditions and identified only four genotypes with a highly tolerant reaction against ChiLCV. Whereas, the sixty germplasm lines of Capsicum annuum L., one each of C. chinense, C. chacoense and C. baccatum and two of C. frutescens were screened and, none of the genotypes was found to be free from the disease [35].
Natural whitefly-mediated inoculation is the most commonly used technique which does not alter the natural virus-vector-host relationships but it's very difficult to control inoculum pressure [40].
In the present experiment, seventy genotypes were screened under natural disease conditions. The phenotypic observations suggested that the chilli plants infected at an early stage remained severely stunted. Their terminal and axillary shoots tend to stay erect, and their leaflets were reduced in size and abnormally shaped. Moreover, a wide variation in the leaf curl virus symptoms variability was noticed under natural field conditions. Establishment of advanced symptoms and in the severe form points out that the disease is present in epidemic form and field experiment for ChiLCV screening was successful. [41]. The differential response of genotypes to ChiLCV incidence and symptom expression could be attributed to the fact that the disease incidence and its spread are influenced by the occurrence and population dynamics of the vector whitefly and the weather conditions in the agro-ecosystem [42].
We found a significant difference among the parents and their hybrids for all characters studied. Naresh et al. [43] earlier reported significant differences among chilli genotypes. Significant differences among genotypes were reported also reported by Rodrigues et al. [44] and Geleta and Labuschagne [45] for various fruit and yield traits in chilli pepper. Predominating role of the nonadditive gene action makes it difficult to gather desirable genes because these genes are not fixed in the population [46]. Through the line × tester analysis, Payakhapaab et al. [47] found significant difference due to crosses and line × testers for plant height, fruits plant -1 , fruit weight, yield, fruit length and fruit width. The inheritance of these characters were apparently due to both additive and non-additive gene action. Singh et al. [48] observed the predominance of additive gene effects for fruit weight, fruit width, fruit length and days to flowering. Naresh et al. [43] reported that the mean sum of squares due to genotypes, parents and hybrids, and parent vs hybrids were highly significant for fruit length, fruit width, dry yield plant -1 and total carotenoids. Similar to our results of high values of GCA and SCA. High GCA and SCA effects were observed for plant height, days to flower, fruit length, fruit girth, fruit weight, and fruits per plant by various authors [49][50][51].
The traits with a lower GCA/SCA ratio values hold a non-additive genetic control. Whereas, the characters with a higher GCA/SCA ration possess an additive genetic control [26,52]. Our results indicated the possibility of the exploitation of heterosis along with picking up of superior segregants (pedigree method) in these crosses, especially for fruit yield important under ChiLCV, indicating the importance of exploiting hybrid vigour in these cross combinations. Krishnamurthy et al. [53] crossed five lines with 30 testers in line × tester mating design to develop 150 F1 hybrids, and they estimated the extent of mid-parent heterosis. For fruit yield plant-1 (green), seven hybrid combinations showed positive and significant mid-parent heterosis. Moreover, our results are in line with the previous study in tomato performed for the determination of heterosis in tomato under leaf curl virus infestation [54,55].
Overall, the four testers were highly resistant under artificial inoculated conditions. Moreover, lines vs testers showed significant differences for all the characters except for plant height. This suggested considerable differences exist among genotypes, i.e. parents (lines and testers) and their 28 F1 hybrids. The hybrids viz., L3 × T2, L7 × T1, L1 × T1, L6 × T3, L1 × T4, L4 × T2 and L7 × T3 were the most promising with desirable SCA effects, heterosis and per se performance for yield and quality attributes and they were moderately resistant to ChiLCV. The ChiLCV resistance could be improved through recombinant breeding or recurrent selection.

Plot Location, Soil and Weather Parameters
The present investigation was carried out at the Department of Vegetable Science, College of Agriculture, Kerala Agricultural University, Vellayani, India, during 2015-2018. The experimental site was located at 8.50° North-latitude and 76.90° East-longitude, at an altitude of 29.00 m above mean sea level. The soil properties were determined based on the methods defined elsewhere [56] and are presented in Table 10. Whereas, the climate during the entire crop season is presented in Figure 2.
Table10. Physicochemical properties of soil of the experimental field.

Plant Material
Seven female parents (lines) were crossed with the four testers (male parent) to produce twenty-eight F1 hybrids. The list of genotypes and their source of origin is given in Table 11. The seeds were sown in portrays by using the potting mixture. During summer season thirtyday old healthy seedlings (8-10 cm height) were transplanted in the summer season (March-May). The package and practices were followed, as described elsewhere [57]. The experiment was laid out as in a randomized complete block design (RCBD) as three replication. There were twenty plants in each replication. The plant to plant and row to row spacing was 45 cm.
However, to facilitate the attack of leaf curl virus disease in the experiment, plant protection measures were not used.

Artificial and Field Screening Against ChiLCV
Before crossing in order to determine the resistance potential of parental genotypes, the parental genotypes (lines and testers) were subjected to the artificial inoculation methods, namely whitefly mediated inoculation and graft inoculation technique. For the artificial inoculation, chilli seeds were sown in the plug trays filled with vermicompost and cocopeat in 1:1 proportion. The trays were kept in the insect-proof cage. Twenty days after sowing the seedlings were gently removed and transplanted into the plastic pot of size 14 × 10 × 13.5 cm filled with soil mixture with vermicompost and kept in an insect-proof cage for inoculation.
For feed acquisition of virus by the whiteflies, ChiLCV infected plant branches were inserted inside the bottles which contain non-viruliferous whiteflies. These flies were allowed to feed on the ChiLCV infected branches for 24 hours (Acquisition period). The healthy chilli seedlings (resistant lines) were inoculated at two-true leaf stage. Cage or single plant inoculation technique was followed for artificial whitefly inoculation [40,58].
Whereas, the graft inoculation was performed as the healthy test plants were grown in pots under insect-proof cage [59]. Small branches ( Figure 2). The base of the scion was trimmed to a wedge shape and inserted into a cleft made on the stem of the infected chilli rootstock plant. The graft was then tied firmly using a para film strip (Figure 2). To increase grafting success the plastic zip-lock pouch bags (10" × 12" inch) were covered over grafted plants (Figure 2).
The presence of ChiLCV from the infected plants was confirmed by PCR employing the degenerate primers [60]. The genotypes after graft inoculation were assessed for presence/absence of viral nucleic acid by Polymerase Chain Reaction (PCR) using a universal degenerate primer (AV494/AC1048) for identification of Geminivirus isolates (subgroup III) [60]. Subsequent observations were recorded at fortnightly intervals, as described by Muniyappa et al. [61]. Five plants were randomly selected from each treatment and the observation was taken at the fortnightly interval after transplanting (DAT) based on the scale defined in detail elsewhere [62]. The specific disease reaction was assigned for all the genotypes based on the coefficient of infection (CI) as suggested by Kumar et al. [63]. The scale used for the scoring of ChiLCV reaction is presented in Table 12.

Plants Characterization and Data Analysis
Observations were recorded from five randomly selected plants to determine the plant characteristics as discussed below.

Plant Height (cm)
Plant height was determined at before final harvest from five random plants.

Primary Branches Plant -1
At the end of the final harvest, the primary branches emerging from the main stem were recorded.

Days to First Flowering
From five randomly selected plants, the duration (days) taken to first flower opening from the date of transplanting were calculated and the mean worked out.

Days to First Harvest
The days taken from the date of transplanting to the first fruit harvest from five randomly plants were noted and the mean worked out.

Fruit Length (cm)
Ten matured green fruits were randomly selected from each tagged plant in the third harvest.
The fruit length (cm) was measured from pedicel attachment of the fruit to its tip end and the mean was worked out.

Fruit Girth (cm)
The girth of fruit was noted from the central fruit by using twine and scale. The same fruits which were used to measure length of fruit was used to measure fruit girth, and the average girth was noted.

Fruits Plant -1
The mature fruits number from each harvest were counted and recorded. Counted fruits were added and the average number of fruits per plant was worked out.

Fruit Weight (g)
The average weight of fruits was measured from 10 randomly picked fruits from third picking.
The weight of fruits was measured on an electronic balance.

Yield Plant -1 (g)
Yield plant -1 was computed by adding the mature green fruit weight from every harvest and dividing by the number of randomly selected plants (five); the mean weight is expressed in grams.

Yield Plot -1 (kg/6.48m 2 )
From each plot, harvested weight of fruits was calculated and expressed in kilograms.

Carotenoids (mg 100 g -1 )
The carotenoids fraction was determined using UV-visible spectrophotometric measurements at two characteristic wavelengths based on the method explained elsewhere [66]. Briefly, the dried red ripe fruits were selected and were ground into a fine powder. This powder (100 mg) was extracted with acetone (25 ml). This extract was transferred to a volumetric flask and the volume was made up to 50 ml by adding acetone. By using acetone as blank, the absorbance of the sample was recorded at two wavelengths (472 nm and 508 nm).
The Line × Tester genetic analysis was performed for the determination of GCA, SCA along with the variance components with the help of software program PBTools (PBTools-1.4, 2014) [67]. Heterosis magnitude was identified in relation to mid-parent, and better parent values and BP is the mean of of the parents and 100 × ((F1 − BP) / BP), where F1 is the hybrid mean the better parent.
Author Contributions: S.I., R.M. and V.S. conceived and designed the project. S.I. and R.M. supervised the study. V.S. performed the experiments. V.S. and P.K. analysed the data. V.S. and P.K. wrote the paper and corrected the final draft. All authors read and approved the final manuscript.
Funding: This research received no external funding.