Rice is the world’s most important food crop and a primary source of food for more than half the world’s population. India accounts for nearly one-fourth (22%) of the rice produced in the world with the first place occupied by China. World rice production now is around 597.8 million tones grown over 151 million hectares with a productivity of 3.96 tones/ha. India has an area of 50.00 million hectares under rice cultivation with an output of 104.10 million tones (Anon., 2012), which averages to around 4.00 tones/ha.Dietary intake surveys from China and India reveal an average adult intake of about 300 g of raw rice per day.India produces some of the best quality rices in the world. The superiority of basmati cultivars over other premium rice varieties is its superfine grains which have a distinct aroma, excellent elongation ability and the soft, flaky texture of the cooked rice. India’s revenue from the export of 8.44 lakh tones basmati rice during 2006-2007 was Rs 2213 crore.

Besides the much sought after basmati types which get high price in international markets, the country also abounds with hundreds of indigenous short grain aromatic cultivars and landraces grown in pockets of different states. Almost every state has its own collection of aromatic rices that perform well in native areas (Shobha Rani and Krishnaiah, 2001). These aromatic rices also possess exemplary quality traits like aroma, fluffiness and taste. However, the improvement of these rices is very much neglected as they lack export value per se.

The aroma of rice plays a role in its consumer acceptability. More than 100 compounds that contri- bute to the aroma of rice have been identified. Some of these volatile compounds contribute to consumer acceptance of certain types of rice, whereas other compounds contribute to consumer rejection. The popcorn-like smell of aromatic rice stemming prima- rily from its 2- acetyl-1-pyrroline (2-AP) content is preferred by many consumers. Several methods are used to detect aroma like biting kernels, smelling vegetative tissue after warming or soaking in KOH and eating cooked rice (Sood and Siddiq, 1978;). In recent years, actual quantification of 2-AP has been done using different distillation and extraction procedures. However, these methods are laborious, expensive, time consuming and require lot of sample for estimation.

The inheritance of quantitative traits is often influenced by variation in other characters, which may be due to pleiotropy or genetic linkage. Hence, knowledge of association between yield and its attributes obtained through estimation of genotypic and phenotypic correlation helps in determining the extent of improvement that could be brought in the characters and also in selecting suitable genotypes.

Correlation coefficients merely describe the existence of association between characters. It is rather difficult to explain a system of correlation as the indirect association of the character increase. The method of path coefficient developed by Wright (1921) is helpful in assessing whether association of character with yield is having direct or indirect effect on yield or is a consequence of indirect effect through some other trait.

All these parameters help to identify superior genotypes with high yielding ability to increase productivity for a successful breeding programme. Keeping eye on these objectives the present investigation was carried out to know the extent of genetic association among yield and yield attributing traits in genetically diverse group of genotypes.

Materials and Methods
The investigation was carried out during Kharif 2010 at Agricultural Research Station, Mugad, the only center working exclusively on rainfed drill sown rice in South India. The Research Station is located at an altitude of 697 meters above mean sea level (MSL), 15°15′ North latitude and 70°40′ East longitude, which falls in Agro-climatic Zone No.8 of Karnataka. Observations were recorded on thirteen quantitative traits viz., days to 50% flowering, plant height, number of tillers for plant, number of productive tillers for plant, panicle length, panicle weight, grain length, grain breadth, L/B ratio, test weight, iron and zinc content and grain yield per plant. The analysis of variance was done as suggested by Panse and Sukhatme. Correlations of various biometrical chara- cters were undertaken as per the procedures suggested by Al-Jibouri et al and with path coefficient analysis by Dewey and Lu.

Results and discussion
The success of any crop improvement programme depends on the magnitude of genetic variability and the extent to which the desirable trait is heritable. The estimate of variability of yield and yield contributing characters and their heritable components in the material is more important in any crop breeding programme. The presence of genetic variability in breeding material has been emphasized by Falconer (1981), so as to exercise critical selection pressure. The information on the nature and magnitude of variation in segregating population of a cross where selection is actually practiced will be more meaningful and it is of immediate practical utility. Moreover correlation studies provide information about the relative contribution of various component traits on grain yield per plant and help in effective identification and selection of superior plants.

Since yield is polygenically controlled and highly influenced by environment, selection based on yield alone is not effective. Therefore, improvement in yield can be brought about by effecting indirect selection through yield attributes whose heritability is high and show strong association with yield. 

Selection for specific character is known to result in correlated response in certain other characters. Generally, plant breeders make selection for one or two attributes at a time. Then it becomes important to know the effect on other characters. Improvement on grain yield per plant, the most important target character in many cereal crops, it can be achieved by direct selection through other easily observable characters. But, this needs a good understanding of association of different traits with grain yield per plant and their possible associations among themselves.

The analysis of variance was presented in Table 1 and all traits found to be significant. The phenotypic and genotypic correlations of grain yield per plant with other quantitative characters in the 42 aromatic genotypes studied are presented in Table 2, Table 3 and figure1.

Table 1 Analysis of variance for yield and yield related characters in forty two aromatic rice genotypes

Table 2 Genotypic correlation co-efficients among yield, yield attributing traits in forty two aromatic rice genotypes

Table 3 Phenotypic correlation co-efficients among yield, yield attributing traits in forty two aromatic rice genotypes

Figure 1 Graph representing correlation coefficients of yield component traits on grain yield per plant and their direct effects

Highly significant and positive correlation was observed for grain yield per plant with number of tillers per plant, number of productive tillers per plant, panicle weight, grain breadth and test weight. This suggests that these characters should be considered while selecting plants for grain yield improvement. Reports of Shashidhar et al (2005) and Girish et al (2006) are in conformity with the above results.

While, significant and negative correlation was observed for days to per cent flowering and plant height with grain yield per plant. The characters, which were negatively correlated with grain yield per plant, were negatively correlated among themselves. Thus grain yield may be improved considerably through selecting its components characters like plant height, number of productive tillers per plant, panicle weight and test weight. These observations were in accordance with the reports of Lanceras et al (2004).

Among yield attributing traits, positive significant association of plant height was observed with panicle length, zinc content, iron content and grain breadth; number of tillers per plant with number of productive tillers per plant; grain length with L/B ratio and test weight; grain breadth with test weight; iron content with zinc content. This suggests that inter relation among the traits should be considered while breeding genotypes for different purposes. These results were in confirmation with the findings of Shivapriya (2002).

Path coefficient analysis was worked out to partition the correlation coefficient into direct and indirect effects. In the present study direct and indirect effect of twelve traits with grain yield per plant results are presented in Table 4, Table 5 and Figure1.

Table 4 Genotypic path coefficient analysis of grain yield per plant versus yield components in 42 aromatic rice genotypes

Table 5 Phenotypic path coefficient analysis of grain yield per plant versus yield components in 42 aromatic rice genotypes

The relationship between yield and yield components may be negative or positive but it is the net result of direct effect of that particular trait and indirect effects via other traits. Hence, it is necessary to determine the path co-efficients which partition the observed corre- lation in to direct and indirect effects and also reveals the cause and effect relationship between yield and their related traits.

The genotypic and phenotypic path indicated high positive direct effect of number of tillers, number of productive tillers per plant, panicle weight and test weight on grain yield per plant. Yogameenakshi et al (2004) and Girish et al (2006) also reported high direct effect of number of tillers per plant on grain yield while Suman et al (2006) reported high positive direct effect of number of productive tillers per plant on grain yield. Suman et al (2006) also reported high positive direct effect of panicle weight and test weight on grain yield per plant. Hence, desirable improvement may be brought about by selecting genotypes with higher number of productive tillers per plant, panicle weight, and test weight.