Correlation and Path Coefficient Analysis in Indian Oil Palm genotypes

1ICAR-Indian Institute of Oil Palm Research, Pedavegi-534 450, Andhra Pradesh, India. 2Department of Biotechnology, Jawaharlal Nehru Technological University Hyderabad, Hyderabad, India. 3Department of Biotechnology, Sreenidhi Institute of Science and Technology, Yamnampet, Ghatkesar, Hyderabad-501 301, India. 4Department of Biotechnology, K L University, Greenfields, Vaddeswaram, Guntur District-522 502, Andhra Pradesh, India. 5Department of Biotechnology, Gokaraju Rangaraju Institute of Engineering and Technology, Hyderabad-500 090, India.

The Oil palm Elaeis guineensis (Jacq) is a native of Guinea Bissau of West Africa and belongs to the family Arecaceae and order Arecales. In India, Oil palm has been introduced as an irrigated crop. Use of palm oil is known for the last 500 years in India. Oil palm cultivation has now been extended to about 50,000 hectares in almost eleven states. Oil palm is the highest vegetable oil yielding crop which is the second most important vegetable oil crop next to soya. It is grown on a commercial scale throughout Malaysia, Indonesia and southern parts of India. The mesocarp is the source of palm oil and the seed yields palm kernel oil (Wood et al., 1984;Donough and Law, 1988;Chan et al., 1989).
It produces 4-6 tonnes of crude palm oil/ ha. As India is one of the highest consumers of palm oil, the yield did not meet the consumption, so India is still importing palm oil from other countries. An effective breeding programme for high yielding varieties can be developed by genetic variation in available resources, analysis of character association with FFBY and path analysis. As yield is complex, it is associated with number of component characters. These characters were interrelated. These interrelated characters may affect the yield which make correlation coefficient ineffective (Chitra et al., 2005). The path coefficient analysis provides direct and indirect causes of association of yield and their component characters. This will be helpful for In the present study, different characters of oil palm genotypes associated with Fresh fruit bunch yield (FFBY) was correlated with bunch dry weight, bunch index, average bunch weight, number of bunches, total dry matter, number of leaflets and oil to bunch ratio. Path coefficient analysis was carried out analyzing bunch dry weight, bunch index, average bunch weight, number of bunches and total dry matter.

Sample collection and parameters analyzed
In the present study fifty eight oil palm genotypes from different Indian villages of Andhra Pradesh and Telangana states were selected (Table.1). At every location the experiment was laid out in randomized block design (RBD) with 3 replications i.e., 3 palms (from each genotype). 24  and Kushairi et al. (1999). With the measuring tape height-HT of the palm was measured from the ground above 50 cm to 17th leaf base and girth-GT at base was measured at that 50 cm level from ground. Height increment was measured by comparing height for 5 years Total number of leaves-NL and number of spindles-NS produced per year was counted thrice per year by marking 17th leaf and 1st leaf with paint. Sex ratio-SR was obtained from the below formulae. Inflorescences in the leaf base was counted thrice in a year by marking 17th leaf and 1st leaf with paint. Number of leaflets-NLL, Rachis length-RL, Petiole width-PW, Petiole depth-PD, Leaflet length-LLL and Leaflet width-LLW were measured on 17th leaf. Bunch number-BN, fresh fruit bunch yield-FFBY and average bunch weight-ABW was recorded while harvesting the bunches. Leaf area-LA, Leaf dry weight-LDW, Specific leaf weight-SLW, Total leaf dry weight-TLDW, Trunk dry matter-TrDM, Vegetative dry matter-VDM, Bunch dry matter-BDM, Total dry matter-TDM and Bunch index-BI was measured using the indirect non destructive method and they are measured by below formulas (Corley et al., 1971 andKushairi et al., 1999).
Sex ratio-SR = number of female inflorescence/ number of male+female+hermophrodite inflorescences. Leaf area-LA = 0.455(NLL x LLL x LLW)-0.245 Leaf dry weight-LDW = 0.058(PW x PD)+0.00541 x RL-0.902 Total leaf dry weight-TLDW = LDW(NL + NS-1) SLW = LDW/LA Trunk dry matter-TDM = Diameter x volume x density of palm Vegetative dry matter-VDM = TLDW + TrDM Bunch dry weight-BDW = 0.53 x fresh fruit bunch weight-FFB Total dry matter-TDM = VDM + BDW Bunch index-BI = BDW/TDM Oil to bunch ratio was obtained using bunch analysis (Soxhlet method). Three bunches from each genotype was collected. Weights of bunch, fruits, spikelets and stalk was recorded. Fresh weight and dry weight of mesocarp, kernel, nuts was recorded for obtaining moisture percentage and the dry samples of mesocarp were kept in soxhlet apparatus where the petroleum ether was evaporated on heating and after cooling it was converted into vapours where it will flush the oil from the dry mesocarp samples and the weights of samples was recorded before and after the extraction in soxhlet. The oil to bunch ratio was calculated by oil/fresh mesocarp to mesocarp/fruit to fruit/bunch.

Data Analysis
The Morphological data from 3 replications of 58 genotypes was recorded on 24 characters was submitted to statistical package WINDOSTAT 9.2 (INDOSTAT services Ltd. Hyderabad, India) to carry out the Genotypic and Phenotypic correlation coefficients and path analysis. Genotypic and Phenotypic correlation was carried out by using the formulae suggested by Falconer (1964) and path analysis as described by Dewey and Lu (1959) was carried out to calculate direct and indirect contribution of various traits to fresh fruit bunch yield-FFBY.

RESULTS AND DISCUSSION
The analyzed data showed positive genotypic correlation with reference to the number of leaflets-NLL (0.1278), petiole width-PW (0.1091), petiole depth-PD (0.0867), rachis length-RL (0.2012), leaflet length-LLL (0.0253), leaflet width-LLW (0.0596), leaf area-LA (0.0845), leaf dry weight-LDW (0.1467), total leaf dry weight-TLDW (0.2766), specific leaf weight-SLW (0.0548) and vegetative dry matter-VDM (0.1107) ( Table.2; Fig.1). Number of leaves-NL had significant and positive correlation with bunch number-BN (Oboh and Fakorede, 1990). Number of leaves-NL was negatively associated in the present study with height-HT and average bunch weight-ABW but there was a positive association in the previous study . It was also reported that rachis length-RL was negatively correlated with number of leaves-NL and has no correlation with FFBY and oil to bunch ratio , but in the present study rachis length-RL was positively correlated with number of leaves-NL and negatively correlation with FFBY and oil to bunch ratio.
Vegetative dry matter-VDM was significantly associated with height-HT (G=0.5506;    . Petiole width-PW, petiole depth-PD, rachis length-RL, leaflet length-LLL, leaflet width-LLW, leaf area-LA, leaf dry weight-LDW, total leaf dry weight-TLDW, specific leaf weight-SLW and vegetative dry matter-VDM were interrelated among themselves and would bring in simultaneous improvement in addition to FFBY. The path analysis indicated (Table.4 & 5) that bunch number-BN, average bunch weight-ABW, bunch dry weight-BDW, total dry matter-TDM and bunch index-BI exerted direct positive effect on fresh fruit bunch yield-FFBY and for these traits direct selection can be made to improve the yield. Selecting FFBY alone could not get the superior hybrids, which can be improved by selecting high bunch index-BI and other vegetative character traits in breeding may get good superior high yielding and uniform palms (Corley et al., 1971). Average bunch weight-ABW had positive and indirect effect on yield via number of leaflets-NLL, leaflet width-LLW, bunch number-BN, bunch dry weight-BDW, total dry matter-TDM, bunch index-BI and oil to bunch ratio. For average bunch weight-ABW maximum indirect effect on yield was obtained through bunch index-BI, bunch dry weight-BDW, total dry matter-TDM and bunch number-BN. Hence selection through these characters may bring about improvement in fresh fruit bunch yield-FFBY.
Bunch number-BN exhibited positive indirect effect on yield through girth-GT, sex ratio-SR, number of leaflets-NLL, rachis length-RL, leaflet width-LLW, leaf area-LA, average bunch weight-ABW, bunch dry weight-BDW, total dry matter-TDM, bunch index-BI and oil to bunch ratio. Bunch dry weight-BDW had positive indirect effect on yield via, sex ratio-SR, number of leaflets-NLL, leaflet width-LLW, leaf area-LA, average bunch weight-ABW, bunch number-BN, total dry matter-TDM, bunch index-BI and oil to bunch ratio. Number of leaves-NL, sex-ratio-SR and other bunch characters, would be effective as indirect selection criteria for bunch number-BN and fresh fruit bunch yield-FFBY (Oboh and Fakorede, 1990). Total dry matter-TDM had indirect effect on yield via height-HT, girth-GT, height increment-HT incre, petiole width-PW, petiole depth-PD, number of leaflets-NLL, rachis length-RL, leaflet length-LLL, leaf area-LA, leaf dry weight-LDW, total leaf dry weight-TLDW, specific leaf weight-SLW, trunk dry matter-TrDM, vegetative dry matter-VDM, bunch number-BN, average bunch weight-ABW, bunch dry weight-BDW, bunch index-BI and oil to bunch ratio. Selection of high total dry matter-TDM character may increase the yields of the hybrids which had indirect effect on bunch number-BN (Rees, 1963). Total dry matter-TDM production is going to bunch number-BN in fraction or more. Path coefficient analysis results show that fresh fruit bunch yield-FFBY was directly influenced by average bunch weight-ABW and bunch number (Tanya et al., 2013). The characters fresh fruit bunch yield-FFBY and bunch number-BN shows highly positive correlation with large direct and indirect positive effects on oil yield and percent oil/bunch (Eksomtramage et al., 2001).

CONCLUSION
The trait fresh fruit bunch yield-FFBY can be improved through bunch number-BN, average bunch weight-ABW, bunch dry weight-BDW, total dry matter-TDM and bunch index-BI and these characters showed positive association with quality parameter like oil to bunch ratio and the major emphasis should be laid on selection process with high number of bunch number-BN, average bunch weight-ABW, bunch dry weight-BDW, total dry matter-TDM and bunch index-BI for better genotypes with superior oil quality and for realizing higher fresh fruit bunch yield-FFBY.