Chilli (Capsicum annuum L) is an important crop both as a vegetable and spice valued for its aroma, taste, flavour and pungency (Vikram et al., 2014). Chilli is one of the rich sources of Vitamin A (292 IU) and Vitamin C (111mg) per 100 g fresh weight. It is cultivated over 1.4 million ha with a production of 18.8 million tonnes (NHB, 2014). Yield being a complex character, which is not only influenced by its associated traits but also governed by number of genes and influenced by environment. So, to make selection effective, it is necessary to separate genetic variability from total variability, which enables breeder to adopt suitable breeding programme. Mere variability studies will not be of much helpful for improvement of yield, as it is associated with number of yield component characters. Association analysis of quantitative attributes would help in choosing component characters that are positively correlated (Kadwey, 2014). Therefore, it is essential to know the degree of mutual association (correlation) prevailing between yield and its component characters, to form the basis for selecting desirable genotypes. However, the correlation between yield and its component characters are often not real, because of inter relationship existing between the component characters themselves. Thus, analysis of inter component correlation is very essential to expose the direct and indirect contribution of each of the component, which in turn is determined by path-coefficient analysis (Wright, 1921). To plan appropriate breeding programme and to evolve high yielding cultivars the plant breeders must possess adequate knowledge on variability, character association patterns, the extent of contribution of each character to fruit yield and genetic divergence. To develop such programmes there is a need to evaluate the available germplasm with the following objectives: (i) To understand the degree and direction association between yield and yield components and their inter correlation among themselves and (ii) To assess the direct and indirect effects of component traits on yield.

1 Materials and Methods

Field experiment was conducted during 2015-16 to study the genetic correlations in chilli. A total of thirty diverse genotypes in chilli group were collected from different sources and used as study materials (Table 1). The trial was conducted in Department of Horticulture, Pandit Jawaharlal Nehru College of Agriculture and Research Institute, Karaikal situated at 10049’ and 110 01’ N latitude and between 780 43’ and 79052’ East longitude and at an altitude of 4 m above mean sea level. The mean annual rainfall is 1437 mm. The mean maximum and minimum temperature recorded are 31.95^oC and 25.52^oC respectively. The experiment was laid out in a Randomized Block Design with three replications and thirty genotypes as treatment. Plants were spaced at 60 cm between row and 45 cm between plants in a row. Recommended horticultural practices and plant protection measures were followed. Observations were recorded from ten randomly selected plants per replication in all genotype for plant height, number of primary branches, leaf area, days to first flowering, days to first fruit set, days to final harvest, number of fruit plant^-1, pedicel length, fruit length, fruit girth, fruit weight, number of seeds fruit^-1, 1000 seed weight, vitamin c, capsaicin content, capsanthin content, green fruit yield plant^-1 and dry fruit yield plant^-1. Data were subjected to ANOVA for all characters (Goulden, 1959). Correlation coefficients for yield and other traits in all 30 genotypes were worked out as suggested by Johanson et al. (1955). Path coefficient analysis as suggested by Dewey and Lu (1959) was used to partition the genotypic.

2 Results and Discussion

In the present investigation both genotypic and phenotypic correlation were worked out to find out the association between dry fruit yield of plant and its component traits (Table 2; Table 3). The correlation coefficient revealed that the dry fruit yield was significant and positively correlated at both genotypic and phenotypic levels with plant height; number of primary branches plant^-1, leaf area, number of fruits plant^-1, pedicel length, fruit girth, fruit weight, number of seeds fruit^-1, 1000 seed weight, capsanthincontent and green fruit yield indicating high heritable nature of the characters. These results were in consonance with early workers (Ibrahim et al., 2001); Sreelathakumary and Rajamony (2002); Bhradwaj et al. (2007) and Manna and Paul (2012). However days to first flowering, days to fruit initiation and days to final harvest showed non-significant negative correlation with dry fruit yield. Though this association showed negative correlation but these characters should be taken into consideration for improvement of chilli crop for earliness.

The results also showed that there was no much difference between genotypic and phenotypic correlations among characters studied. This indicated that the influence of environment had least effects. So, one can relay their selection on phenotypic values. Dry fruit yield plant^-1 is the combination of many characters and were polygenically controlled. For such traits, direct selection is difficult. Therefore, selection for any of these highly associated traits with dry fruit yield plant-1 will indirectly help in selecting the plants or genotypes with high yield. However in the present study, apart from the higher estimates of genotypic correlation observed for the above said eleven characters which were positively correlated with dry fruit yield directly, the other characters plant height, fruit length, pedicel length, fruit girth and capsaicin had inter correlated with dry fruit yield. The results were in agreement with the findings of Leaya Jose and Abdul Khader (2002), Mallikarjun et al. (2003), Nandadevi and Hosamani (2003) and Prabhudeva (2003) who also noticed positive association of number of branches and number of fruits plant^-1 with dry fruit yield plant^-1. The vitamin C content was negatively inter correlated with plant height, number of primary branches, leaf area, number of fruit, pedicel length, fruit length, fruit girth and  number of seeds but positively with fruit weight and 1000 seed weight.

Path coefficient analysis revealed that the number of fruit plant^-1 had maximum positive direct effect on dry fruit yield in chilli followed by days to first flowering, green fruit yield, number of seed fruit^-1 and days to final harvest (Table 4). As such these economic characters as well as earliness would have contributed direct influence on dry fruit yield therefore greater emphasis has to be given for these characters as selection indices for improvement of chilli. The importance of number of fruits plant^-1 has been emphasized by Korla and Rastogi (1977), Nair et al. (1984), Leaya Jose and Abdul Khader (2002) and Prabhudeva (2003). The path analysis of the present study also confirms the earlier studies that number of fruits plant^-1; fruit girth and number of seeds fruit^-1 were important traits that contributed to fruit yield.

Apart from the above said high direct positive effect of the characters on dry fruit yield, the fruit girth, pedicel length, capsaicin, vitamin C, plant height and leaf area exhibited low to negligible positive direct effects on yield. Due importance on these characters may also be given in selection programme for the improvement of chilli. The growth characters such as plant height, number of branches and leaf area exhibited indirect positive effect on number of fruits, number of seeds and green fruit yield (Leaya Jose and Abdul Khader, 2002; Verma et al., 2004).

Except the fruit length, all the economic fruit characters taken under this study had high positive indirect effect among each other’s along with days to first fruit set towards green fruit and dry fruit yield. Regarding quality parameters, the vitamin C exhibited low positive indirect effect with days to first flowering whereas capsanthin content, capsaicin content showed low to moderate positive indirect effect with green fruit yield and invariably negative effect with fruit length.

Residual effect determines how best the causal factor accounts for the variability of the dependent factor that is dry fruit yield plant^-1 in this study. It is noted from the present study that the residual effect was 0.1345 from the path coefficient analysis which clearly indicated that all the 18 characters taken under this investigation were sufficient for genetical analysis in chilli. The result was in accordance with Sharma et al. (2010).