VARIABILITY, PATH ANALYSIS AND GENETIC DIVERSITY IN GREEN GRAM {Vigna radiata (L.) Wilczek)

Transcription

1 VARIABILITY, PATH ANALYSIS AND GENETIC DIVERSITY IN GREEN GRAM {Vigna radiata (L.) Wilczek) A thesis submitted to the MAHATMA PHULE KRISHI VIDYAPEETH RAHURI , DIST: AHMEDNAGAR, MAHARASHTRA, INDIA By BABI RAMESH REDDY. N In partial fulfillment of the requirements for the Degree of MASTER OF SCIENCE (AGRICULTURE) In AGRICULTURAL BOTANY (CYTOGENETICS AND PLANT BREEDING) DEPARTMENT OF AGRICULTURAL BOTANY MAHATMA PHULE KRISHI VIDYAPEETH COLLEGE Of AGRICULTURE PUNE MAHARASHTRA, INDIA 2003

2 VARIABILITY, PATH ANALYSIS AND GENETIC DIVERSITY IN GREEN GRAM (Vigna radiata (L.) Wilczek) A thesis submitted to trie MAHATMA PHULE KRISHIVIDYAPEETH RAHURI DIST: AHMEDNAGAR, MAHARASHTRA, INDIA By BABI RAMESH REDDY. N In partial fulfillment of the requirements for the Degree of MASTER OF SCIENCE (AGRICULTURE) in AGRICULTURAL BOTANY (CYTOGENETICS AND PLANT BREEDING) Approved by ProfrO. B. Lad Chairman and Research Guide Assistant Professor of Agnl Botany,. College of Agriculture, Pune wh/v\^ ft Dr. H. B. Mudgse Dr. P. ATEFavale Committee Member Committee Member Professor of Agnl Botany i Associate Professor of Agnl Botany, College of Agriculture, Pune A/I. College of Agriculture, Pune Dr. M. JI Wattarnvtfar Committee Member Associate Professor of Agnl Statistics^ College of Agriculture, Pune MPK^" LIBRARY, fuh(j

3 Affectionately dedicated to my beloved parents, Sou. Satyavathi Nallamilli, Sri. Satti Reddy Nallamilli and family members. Ramesh Reddy

4 I CANDIDATE'S DECLARATION I hereby declare that, the thesis entitled "Variability, Path analysis and genetic diversity in green gram (Vigna radiata (L.) Wilczek)" or part thereof has not been submitted by me or any other person to any other University or Institute for a Degree or Diploma Pune: a.(u~sm*hl0- Oated y57 06/ 2003 ( Babi Ramesh Reddy. N)

5 Prof. D. B. Lad Assistant professor of Agril. Botany, Department of Agril. Botany, College of Agriculture, Pune Maharashtra State, India. II CERTIFICATE This is to certify that the thesis entitled "Variability, Path analysis and genetic diversity in green gram {Vigna radiata (L.) Wilczek) " submitted to the Faculty of Agriculture, Mahatma Phule Krishi Vidyapeeth, Rahuri, District Ahmednagar, Maharashtra State, for the award of degree of MASTER OF SCIENCE (Agriculture) in Agricultural Botany (Cytogenetics and Plant Breeding), embodies the results of a piece of bona fide research work carried out by Mr. Babi Ramesh Reddy. N, under my guidance and supervision, and that no part of thesis has been submitted for any other Degree or Diploma The assistance and help received during the course of investigation has been acknowledged Place : Pune Dated :3Sl *l 2003 (Prof. D. B. Lad) Research Guide and Chairman

6 Dr. R. V. Wuike Associate Dean, College of Agriculture, Pune Maharashtra State, India. in CERTIFICATE This is to certify that the thesis entitled "Variability, Path analysis and genetic diversity in green gram (Vigna radiata (L.) Wilczek)" submitted to the Faculty of Agriculture, Mahatma Phule Krishi Vidyapeeth, Rahuri, District Ahmednagar, Maharashtra State, in partial fulfillment of requirements for the degree of MASTER OF SCIENCE (Agriculture) in Agricultural Botany (Cytogenetics and Plant Breeding), embodies the results of a piece of bona fide research work earned out by Mr. Babi Ramesh Reddy. N, under the guidance and supervision of Prof D. B. Lad, Assistant Professor of Agnl. Botany, Department of Agricultural Botany, College of Agriculture, Pune-5 and that no part of thesis has been submitted for any other Degree or Diploma Place : Pune Dated :t A / 2003 '7 " 7 w *> - (Dr. R. V. Wuike)

7 ACKNOWLEDGEMENTS IV I wish to express my deep sense of gratitude and sincere thanks to my research guide and chairman of advisory committee, Prof D. B Lad, Assistant Professor of Agril Botany, College of Agriculture, Pune, whose guidance, support, timely help and erudite suggestions were instrumental in accomplishing this task I express my profound thanks to my Professor and member of advisory committee, Dr H B. Mungse, Professor of Agril Botany, College of Agriculture, Pune for his valuable advise and encouragement during the tenure of my study I am extremely grateful to Dr. P. A Navale, my member of advisory committee, whose uncountable practical suggestions during the conduct of research work I am also grateful to Dr M J Wattamwar, Associate Professor of Agril Statistics and Prof N D. Bangar, Assistant Professor of Agril Botany for his unfailing help throughout the period course work I am immensely grateful to Prof C A Nimbalkar, NARP, Ganeshkhmd, Pune for his help in analyzing statistical data, Prof. G D Mukhekar, Assistant Professor of Agril Botany, Prof. P D Khade, Assistant Professor of Agril Botany and Mrs Anita Kshirsagar, Agril Assistant for their profound and extensive help during my research work I express my profound thanks to Dr B M Jamadagni, Principal Scientist, Pulses Improvement Project, M P K V, Rahun, Senior Scientist, Agricultural Research Station, Badnapur and Chief Incharge, NBPGR, New Delhi, for providing me with green gram germplasm 1 specially thanks to Mr Kaule, Shn Chaudhan, Shn Han, Shn Narke, Shn Dewede Shn Kalamkar, Shn Sonawane, Shn Gole, Shn Bira, Shn Kakade, Smt Sawant, Smt Sable, Smt. Shinde and Smt Palaskar for their

8 love, intensive help during all field operations which made it easy for me to overcome difficulties in field work I thank my friends Naveen, Abhijit, Amol, Sandesh, Pratap, Leena, Sankalp, Sanobar, Venu, Santosh and all my seniors for their direct and indirect help and cooperation, which made my work so easy and fast I must express my indebtedness, appreciation and thanks to my parents, brothers and sister for providing me valuable opportunities and encouragement V Place Pune ( Babi Ramesh Reddy N ) Date -as-ma 003

9 vi TABLE OF CONTENTS Sr. No. <* Title Page No. CANDIDATE'S DECLARATION i CERTIFICATES 1. Research Guide ii 2. Associate Dean ACKNOWLEDGEMENT iv LIST OF TABLES -^ viii LIST OF FIGUERS ABSTRACT viii ix 1 INTRODUCTION 1 2 REVIEW OF LITERATURE Genetic Variability Heritability and Genetic Advance Correlation and Path analysis Genetic Diversity 22 3 MATERIAL AND METHODS Material Methods Statistical analysis 31

10 Sr. No. Title vii Page No. 4 EXPERIMENTAL RESULTS Analysis of Variance Mean Performance Parameters of Genetic Variability Correlation Path Analysis Divergence Cluster Means Per cent contribution of various characters for divergence 62 5 DISCUSSION Variability Heritability and Genetic Advance Correlation Path Analysis Divergence SUMMARY AND CONCLUSIONS LITERATURE CITED VITA 100

11 viii LIST OF TABLES Table Page Title No. No. 1 1 The proximate analysis of mung bean The genotypes and their source Analysis of variance for 13 characters in green gram Mean performance of 40 green gram genotypes for different characters. Parameters of genetic variability in 40 genotypes of green gram. Genotypic (above diagonal) and phenotypic (below diagonal) correlations of 13 characters in green gram Direct (diagonal) and indirect (above and below diagonal) path effects of different characters towards yield in green gram. 4.6 Distribution of 40 genotypes into different clusters Average intra and inter cluster D 2 and D values of 7 clusters from 40 genotypes of green gram. Mean performance of clusters for 13 characters in green gram. 4.9 Contribution of various characters to divergence Distribution of different cluster combinations into four divergent classes based on D values between them LIST OF FIGURES Figure No Title Path diagram showing nature of causal system variables with their coefficient for path analysis in green gram. A cluster diagram showing interrelationships between seven clusters. Between Pages

12 IX ABSTRACT Variability, Path Analysis and Genetic Diversity in Green gram (Vigna radiata (L.) wilczek) By BABI RAMESH REDDY.N A Candidate for the degree of Master of Science (Agriculture) in Agricultural Botany (Cytogenetics and Plant Breeding) Mahatma Phule Krishi Vidyapeeth, College of Agriculture, Pune Research Guide Discipline : Prof. D. B. Lad, Assistant Professor of Agril. Botany : Agricultural Botany (Cytogenetics and Plant Breeding) The present investigation entitled " Variability, path analysis and genetic diversity in green gram (Vigna radiata (L.) wilczek) "was undertaken to study the extent of genetic variability, heritability (b.s), genetic advance, correlation, path analysis and genetic divergence among 40 genotypes The material was evaluated in randomized block design with three replications during kharif, 2002 at Botany farm, College of Agriculture, Pune-5 Observations were recorded for 13 characters Significant treatment sum of squares for all characters studied revealed the presence of considerable amount of variability The magnitude of GCV and PCV were high for number of pods per plant, number of branches

13 X per plant and test weight indicating presence of variability for these characters The high magnitudmal differences between GCV and PCV were found for number of clusters per plant, number of pods per plant, seed yield per plant and number of branches per plant indicating the role of environment in phenotypic expression of these traits The lowest difference between GCV and PCV was found for days to maturity indicating the major role of genotype in the expression of this character High magnitude of hentability (b.s) was recorded for test weight followed by days to maturity, protein content, days to 50 per cent flowering, pod length and number of nodules per plant High hentability coupled with high genetic advance was observed for days to maturity and days to 50 per cent flowering Seed yield per plant was significantly and positively correlated with plant height and number of clusters per plant both at genotypic and phenotypic levels, whereas it was significantly and positively associated with number of seeds per pod, test weight, days to maturity and days to 50 per cent flowering at genotypic level and with number of pods per plant at phenotypic level Seed yield per plant was also significantly and negatively associated with harvest index. In path coefficient analysis, plant height, days to 50 per cent flowering and test weight recorded the highest direct effect in the desirable direction Their association with seed yield was significant and positive indicating the fact that there exists a true and perfect association between these characters

14 XI The D 2 values ranged between to suggesting the presence of considerable amount of genetic diversity All these 40 genotypes were grouped into seven clusters, in which cluster I was the largest (14) among all, followed by cluster II (9) cluster lll(8), cluster IV (4) and cluster V (3) and cluster VI and cluster VII were monogenotypic The maximum intra-cluster distance was observed for cluster IV followed by cluster III and cluster II, suggesting that genotypes present in these clusters possesing varied genetic architecture and might have originated from different genetic pool Whereas, maximum inter-cluster distance was observed between cluster V and VI followed by cluster II and VII and cluster II and VI, indicating wide divergence among these clusters The variance of cluster mean revealed that days to maturity, plant height, harvest index and days to 50 per cent flowering had maximum contribution towards divergence in the present material The following parents are suggested for tentative breeding programme based on divergence studies KOP PM 9342 PM 9376 PM 9380 BPMR132 VC3960 J-781 PM 9383 AKM 9801 Pages 1 to 100

15

16 INTRODUCTION

17 1. INTRODUCTION Mungbean (Vigna radiata (L.) Wilczek) commonly known as green gram, is one of the important grain legumes grown in India Among pulses, it has an important place as it contains more digestible proteins than the other pulses Pulse is defined as the split cotyledons of dry legume seeds boiled in excess of water, softened, macerated and used as a soup It is an excellent source of high quality protein in the diet of low income group in developing countries Mungbean is a short day, warm season crop, grown mainly in semiand to sub humid tropics and subtropics with 600 to 1000 mm annual rainfall, 22 to 35 C mean temperature during crop production and elevations not exceeding 1800 to 2000 m above mean sea level For high yield, a warm climate and deep well drained loam or sandy loam soils are desired Mungbean is rapidly growing, erect or sub-erect annual usually 40 to 120 cm in height It is frequently much branched with slight tendency of twining in its upper branches The leaves are trifoliate with large, ovate, entire or rarely lobed membranous leaflets with scattered hairs on both sides The pods are subcylindncal, long wide, straight or slightly curved, 10 to 20 cm, small, globular and ablong often green but may be yellow brown and speckled with black coloured seeds Narsimham (1929) reported that pollination in mungbean is completed in all cases by 1 30 p m with an interval of about 4 hours, between the dehiscence of anthers and opening of flower causing selffertilization

18 2 According to De Candolle (1884), Vavilov (1926) and Zukovskij (1962), mungbean (Vigna radiata (L) wilczek) has originated from Indian subcontinent Maximum diversity among the related species is limited to upper western ghats and deccan hills Roxburgh (1832) named mungbean as Phaseolus mungo, while wild and yellow variety (Sona mung) as Phaseolus aureus Roxb Bose (1939) referred to mungbean as Phaseolus radiatus Linn (Syn = Phaseolus aureus Roxb). Verdcourt (1970) proposed transfer of Asiatic Phaseolus spp to genus Vigna Wilczek (1954) named it Vigna radiata (L) Wilczek instead of Phaseolus aureus (L). The diploid mungbean (2n = 22, 24) belongs to family Leguminoseae, sub family papillionaceae, genus Vigna and species radiata In India green gram is grown in almost all states over an area at 25 3 lakh ha with the annual production of 8 6 lakh tonnes and average yield of 340 kg/ha during the year India is the primary mungbean producer contributing about 75 per cent of world production Among all the pulses grown in India, it ranks third after chickpea and red gram The major mungbean producing states are Andhra Pradesh, Maharashtra, Gujarat, Onssa and Tamilnadu About 70 per cent of mungbean is cultivated during Khanf season, while the remaining 30 per cent is grown in rabi and summer season In Maharashtra, it is grown over an area of 7 10 lakh ha with the annual production of 2 91 lakh tonnes and average yield of 409 kg/ha during the year Mungbean enriches soil fertility through biological nitrogen fixation and checks soil erosion as a cover crop Sometimes it is also used as green

19 3 manuring crop and fodder crop as well In view of low input requirement and short duration, its cultivation is quite economic Due to its short maturity, it can be accommodated well in multiple and relay cropping systems Compared to other crops, particularly the cereals, mungbean offers better chances of crop production under adverse conditions of moisture stress and low fertility Mungbean is rich in vitamin B and is regarded as a remedy for benben When it is allowed to sprout, ascorbic acid (vitamin C) is synthesized and a mount of riboflavin and thiamine are also increased It contains 24 per cent proteins with all essential amino acids It is consumed in a variety of ways Sprouted whole beans are used in South Indian diet for preparing curry and a Savoury dish and halwa Sprouted beans are widely used as fresh vegetables ir Chinese and Japanese food The proximate analysis of mungbean is given in Table 1 1 (Aionymous, 1970) Table 1.1 The proximate analysis of mungbean. Sr. No. Contents Whole dried seed. Dhal 1 Moisture (%) Protein (%) Fat (%) Minerals (%) Fibre (%) Carbohydrate (%) Ash(%) 36-8 Calcium (mg/100g) Phosphorous (mg/100g) lron(mg/100g) Vitamin A (I.U.) Calories

20 4 Though India is a leading producer of green gram in the world, its productivity is very low Therefore, there is a great scope for its improvement and to increase the productivity by developing high yielding, disease and pest resistance varieties with improved nutritional value Good amount of variability has been reported in green gram for various characters such as flowering, days to maturity, plant height, harvest index, pods per plant, protein content, number of clusters per plant, etc however, their utilization in breeding programmes resulted in identification and release of good number of varieties in green gram However, these released types can not be continued longer due to genetic erosion and susceptibility to disease and pest This demands replacement of old varieties by many developed ones To facilitate that it felt necessary to study the genetic variability, path analysis and genetic diversity in various genotypes of green gram, with the following objectives 1. To know the extent of variability for yield and yield contributing characters. 2. To study the association between different characters and to find out direct and indirect effect of important characters on yield. 3. To measure genetic divergence between different genotypes and group them into suitable clusters.

21

22 REVIEW OF LITERATURE

23 5 2. REVIEW OF LITERATURE The present investigation is aimed at finding out statistical parameters viz, genetic variability, correlation, path analysis and genetic diversity in forty cultivars of green gram (Vigna radiata (L.) Wilczek) The literature pertaining to variability, path analysis and genetic diversity in green gram is reviewed an i presented in this chapter under the following sub-heads 1 Genetic variability 2 Hentabihty and genetic advance 3 Correlation and path analysis 4 Genetic diversity 2.1 Genetic variability: Johanssen (1909) demonstrated the distinction between genotype and phenotype and Nilsson-Ehle (1908) and East (1916) gave the concept of multiple factor hypotheses and proved that quantitative characters are inher, ed according to Mendel's laws From these discoveries, it is clear that varial ility resulted from the joint action of genotype and environment (Fisher, 1930, Charles and Smith (1939) and Power et al (1950) partitioned genetic vana, ce from total variance by use of estimate of environmental variance from the non-segregating population This work made possible to use gene ypic coefficient of variation (GCV) of a relative magnitude of genetic van; tion in the material and helps to compare the genetic variability present for different characters The statistical methods to calculate the genetic

24 6 component of variance were given by Frankel (1947), Burton (1952) and Panse anclsukhatme (1985) Veeraswamy et al (1973a) reported high genotypic coefficients of variation for clusters per plant (35 3), number of pods per plant (34 5), number of branches per plant (331) and plant height (30 5) The genotypic coefficients of variation for days to maturity (7 2) and pod length (8 7) were very low Similar results were reported by Lakshmaiah et al (1989), Sudheer Kumar et al (1992) Maximum range of variation was recorded for pods per plant ( ) followed by hundred seed weight ( ) and pod length (6 4 - tj 72) by Rathnaswamy et al (1978) Their investigations further revealed that genotypic coefficient of variation was the highest for hundred seed weight (28.39) followed by pods per plant (19 14) The GCV estimate for seeds per pod (6 71) was the lowest From the variability studies in ninety green gram varieties paramasivan and Rajasekran (1980), reported that phenotypic variability was the highest for pod number (140 99) followed by plant height (94 68) and days to 50 per cent flowering (46 32) High G C.V estimates obtained for plant height (45 04), pod weight (43 40), cluster number (42 53), seed yield (41 61) and pod number (41 51) indicated that these traits were potentially variable Similar results were reported by Veeraswamy et al (1973b) for plant height in b^ack gram, Gupta and Singh (1969) for pod number and seed yield and Singh and Malhotra (1970) for pod number, seed yield and cluster number in green gram

25 7 While studying the variation for agronomic characters in hundred Vigna radiata varieties, Choi et al (1986) found that days to flowering ranged from 39 to 57 days, whereas the days to maturity ranged from 54 to BO days The plant height ranged from 47 to 105 cm, whereas the number of pods per plant varied from 4 to 22, Thus indicating a large amount of variation for the above characters Ali and Shaikh (1987) analysed 30 genotypes of mungbean grown at two sites in Bangladesh and revealed that seed yield per plant exhibited the highest genotypic (30.67) and phenotypic (49.66) coefficients of variation They also recorded the least genotypic variation for number of seeds per pod and seed size. Kalpana et al (1988) while studying the selection parameters for yield and yield components in 30 different varieties of mungbean, observed that pods per plant was the most variable character followed by seed yield per plant and flowers per plant They also reported high genotypic coefficient of variation for intemode length, branches per plant and plant height, The results of Natarajan et al. (1988) were in agreement with those of Ramana and Singh (1987), who studied the variability in 37 varieties of green gram along with two controls grown in Spring (S) and Khanf (R) seasons and reported that number of pods (S-24 7, R-22.8) and clusters per plant (S-19 7, R-2Q 7) had the highest genotypic coefficient of variation Mishra et al. (1995) also reported, the similar results Significant differences between 22 cultivars were noted for all the characters studied by lihamuddin and Tajammal (1989) They also recorded high genotypic and phenotypic variances for plant height and hundred seed

26 8 weight Genotypic and Phenotypic coefficients of variance were the highest for seed yield per plant The findings of Ginraj (1973) and Chowdhury et al (1971) were similar to that of above results Tiwari et al (1996) studied genetic variability in six parents along with 15 progenies of segregating generation of mungbean and observed that F 2 progeny showed high variability for days to maturity, clusters per plant, pod length and hundred seed weight Byregowda et al (1997) while studying the genetic variability in twenty five genotypes of green gram, reported that sufficient variability was present in the material for grain yield and pods per plant They also reported that characters such as pods per plant, seeds per pod and 100 seed weight should be given due importance while making selection for increased yield in green gram Das et al (1998) reported high genotypic coefficients of variation for plant height, branches per plant, pods per plant, pod length and yield per plant in 22 genotypes of green gram Manivannan (1999) evaluated Fi and F 2 progeny of 6 green gram crosses grown during rabi 1996 and Khanf "1997. Among the crosses, VGG 11 x PMB27 followed by Vamban 1 x PMB27, Vamban 1 x ML682 and VGG 11 x EC gave the highest values for number of pods and seed yield per plant With respect to the variability parameters, Vamban 1 x ME 682 and VGG11 x PMB27 recorded high genotypic coefficients of variation. Manivannan (2000) evaluated the extent of variability available for pods and seed yield in eight F3 populations of green gram during rabi season of Among the crosses, KM2 x ML131 ( and ) and

27 9 Vamban 1 x MGG 319 ( and ) recorded superior mean performance for pods and seed yield besides registering high genotypic coefficients of variation for these characters Loganathan et al (2001b) analysed fifty genotypes of green gram for genetic variability during rabi 1999 High phenotypic coefficient of variability indicated the favourable effect of environment for number of clusters per plant and seed yield per plant and high genotypic coefficient of variability suggested substantial amount of genetic variability for number of pods per plant and seed yield per plant Venkateswarlu (2001oJ analysed seventeen diverse genotypes of green gram for genetic variability. Genotypes differed significantly for all the characters studied except 100 seed weight Seed yield expressed high genotypic coefficient of variation coupled with high hentability and genetic advance Supnyo Chakraborty et al (2001) studied 24 genotypes of green gram for their variability parameters Relatively large differences between phenotypic and genotypic coefficients of variability were observed for root length, number of secondary roots per plant and root shoot ratio, indicating that these characters are influenced by environment 2.2 Heritability and genetic advance: Relative role of heredity in the expression of phenotypes is termed as heritability (Falconer, 1989 and Allard, 1961) It is also defined as the proportion of total variability that is due to genetic causes' Since, heritability indicates the possibility and extent to which improvement is possible through

28 selection, knowledge of the heritability of a character is Hnportant to the 10 breeder The genetic advance is the product of heritability, phenotypic standard deviation and standardized selection differential (Burton and Devene, 1953) This is a measure of expected genetic progress based on which selection procedure can be evaluated Singh and Malhotra (1970) reported that selection for improvement in yield of green gram should be based on hundred seed weight which had the highest genetic advance (35 48) as the percentage of mean. Genetic advance was also observed to be high for pod number, bunch number and seed yield but these characters had low heritability estimates The results of Sudheer Kumar et al. (1992) were similar to that of the above mentioned findings Veeraswamy et al (1973a) reported that days to first flower had the highest heritability (90 6%), followed by plant height (88 5%) and number of clusters per plant (83 8%) Seed yield per plant had the lowest heritability (47 2%) A high genetic advance as a per cent of mean was noted in case of number of clusters per plant (66 5), number of branches per plant (60 8), plant height (59.0) and number of pods per plant (57 0) In case of days to maturity, pod length, weight of pods and number of seeds per pod, though the heritability was moderate to high, the genetic advance was low Similar results were reported by Yadav and Rao (1986) and Singh et al (1980) According to Paramasivan and Rajasekaran (1980), the high heritability estimates were coupled with high genetic advance for pod length, hundred seed weight, number of pods per plant, number of clusters per plant and seed yield The results of lihamuddin and Tajammal (1989) were in confirmation with the above findings

29 11 Khan (1985) reported that number of fertile branches and number of pods per plant had high hentability values i e 70% and 67% respectively Whereas, All and Shaikh (1987) reported that seed size had the highest hentability estimate (76 8%) Malik et al (1985) revealed lack of heritable variation for days to flowering and seed number per pod Coheritability values of seed yield with other characters were greater than the hentability of yield itself, suggesting potential for improving seed yield by simultaneous selection for pairs of characters High hentability coupled with moderate to high genetic advance of seed yield per plant, internode length, pods per plant, branches per plant and plant height was recorded by Kalpana et al (1988) Similar findings were also reported by Lakshmaiah et al (1989), Naidu (1990), Joshi and Kabaria (1973), Gmraj (1973) and Chowdhary et al (1971) Ramana and Singh (1987) studied the hentability and genetic advance in thirty seven promising varieties of green gram, grown in two different seasons i e in the spring (S) and rainy (R) season and reported that plant height (S-97 4% and R- 89 5%) showed the highest broad sense hentability in both the seasons followed by 100 seed weight in spring (92 3%) and seed yield per plant in rainy season (89 4%) In both the seasons, pods per plant (S-74 3% and R %) and clusters per plant (S-71 6% and R-70 2%) showed moderate hentability values Seeds per pod (S-52 4% and R-66 8%) showed the lowest hentability in both the seasons In both the seasons, yield per plant (S-56 4 and R-63 1), pods per plant (S-44 0 and R-40 7), clusters per plant (S-34 3 and R-35 7) and plant height (S-30 8 and R-26 5) showed the highest genetic advance, while, days to first flower (S-11 5 and R-10 0),

30 12 seeds per pod (S-13 5 and R-12 6) and hundred seed weight (S-14 4 and R- 20.0) showed the lowest They further observed that the yield per plant, pods per plant, clusters per plant and plant height had high heritability coupled with high genetic advance, suggesting that additive effects were more important for these characters Seed weight and days to first flower had high heritability but low values of genetic advance indicating that the dominance or epistatic effects were of considerable values Natarajan et al. (1988) found that hundred seed weight showed the highest heritability (97 3%) followed by days to flowering (93 2%), plant height (73 7%) and pod length (69 7%) Number of clusters per plant had the lowest heritability (20 54%) Seed yield per plant (46.83%) showed the highest genetic advance as a percentage of mean followed by plant height (38 22%) These results were similar with those of Emping et al (1970) According to Imne and Butler (1982) the broad sense heritability estimate for seed yield was 0 07 Tiwan et al. (1996) studied the heritability estimates in parents and segregating populations and reported that the characters which showed high heritability in the parents did not show high values in segregating generations It was found that plant height, seed yield per plant and harvest index showed high heritability in the parental population, while in the F 2 generation, days to maturity, number of clusters per plant, pod length and hundred seed weight showed high heritability They further reported that high heritability estimates were generally associated with low genetic advance

31 13 Byregowda et al (1997) reported that high hentability associated with high genetic advance was observed for grain yield and pods per plant, which was mainly attributed to additive gene action. Das et al. (1998) reported that high hentability associated with high genetic advance over mean was observed for plant height, branches per plant, pods per plant and pod length It indicates that these traits were mostly controlled by additive gene action Seeds per pod and yield per plant recorded low hentability coupled with low and high genetic advance, respectively Loganathan et al. (2001b) reported low hentability, low genetic advance and non additive gene action for days to first flowering, plant height, number of branches per plant, pod length and 100 seed weight Supnyo Chakraborty et al (2001) reported moderately high hentability with genetic advance for seed yield per plant, nodules per plant and dry root weight suggesting the partial role of additive gene effects in their inheritance However, low hentability coupled with low genetic advance for root length, number of secondary roots per plant and root shoot ratio indicated that these traits were predominantly governed by non additive gene effects 2.3 Correlation and path analysis: The yield is a complex character dependent on many attributes of plants The correlation studies help in understanding the association between the traits The chief genetic cause of correlation is pleiotropy and linkage (Falconer, 1960), because of which it becomes difficult to get the actual idea about positive or negative effect of genes The path coefficient analysis is the

32 14 solution to such problem as it measures the direct as well as indirect effect of various traits on yield Malhotra et al (1974) in their studies on correlation and path analysis in 60 indigenous and exotic strains of green gram, reported that the association between seed yield and number of pods per plant was significantly positive and magnitudinally high (0 95), however, hundred seed weight showed a negative (-0.486) association with yield The path coefficient analysis indicated that number of pods per plant had the highest direct effect (0 844) on seed yield per plant The direct contribution of seeds per pod (0 165) and seed size (0 163) towards seed yield per plant was small but these characters influenced seed yield appreciably indirectly via pod number Upadhyaya et al (1980) in their studies on character association in 115 varieties of green gram of different maturity groups, reported that in the early maturity group, number of pods per plant, plant height and number of seeds per pod were the mam yield components, whereas in late maturity group number of pods per plant, 100 seed weight and number of branches per plant were the main yield components Thus, they concluded that the period from sowing to maturity greatly influenced correlations among characters Boomikumaran and Rathinam (1981) while studying the correlation and path coefficient analysis in forty nine genotypes of green gram under summer conditions found that clusters per plant, pods per cluster and pods per plant showed a significant and positive phenotypic correlation with grain yield but the corresponding genotypic correlation values were low indicating that the association might have been influenced by the environment The path

33 15 coefficient analysis showed that plant height exerted the maximum direct positive effect on grain yield. The indirect effects of other component characters on grain yield through plant height were also much appreciable, showing that plant height played an important role in determining the grain yield of green gram Sulaiman (1976) made similar observations in green gram Correlation and path coefficient analysis was carried out on 50 varieties of green gram by Malik et al (1982) and reported significant positive correlation of seed weight, pods per plant and days to maturity with seed yield Pods per plant (1 873) had low direct contribution but it contributed significantly through seeds per pod (4 557) Seed weight (3 0432) also had high direct contribution to yield. Similar conclusions were drawn by Meshram (1978). Deore (1983) while confirming the results of Kumani and George (1982) and Gupta et al (1982) observed that number of pods per plant, pod length, hundred seed weight and harvest index had significant positive association with seed yield, while negative correlation was observed between days to 50 per cent flowering and plant height Path coefficient analysis exhibited the highest positive direct effect of number of pods per plant on seed yield followed by hundred seed weight and seeds per pod for their major direct and indirect effects Liu et al (1984) revealed that seed yield per plant was related to number of days from emergence to flowering, values obtained being 4 52(g) in the 40 day group, 10 95(g) in the day group and intermediate in the greater than 61 day group Seed yield also increased sharply when the height

34 16 was greater than 47cm and reached maximum of per plant at 62 cm and decreased significantly when height was greater than 71 cm Thandapani and Rao (1984) studied the yield parameters and their significance in fifteen green gram genotypes in relation to yield and observed that clusters per plant (0 4355) had the greatest direct effect on yield, while pod length and seed weight were also directly associated with yield Number of seeds per pod, hundred seed weight and the fertility coefficient (ratio of pods flowers produced) had indirect effects on yield Similar results were reported by Ah and Shaikh (1987) and Panwar etal (1995) Khan (1985) reported that number of pods per plant and fertile number of branches should be given importance in construction of selection indices for green gram, as they had a high positive direct effect on yield Analysis of 30 elite strains in mungbean by Satyan et al (1986) revealed that number of primary branches (0 47) and number of fruiting branches (0 39) had greater direct effects on seed yield per plant, which was reflected in the higher phenotypic correlations (number of primary branches 0 60 and number of fruiting branches 0 58) The negative relationship between plant height and number of fruiting branches weakened the positive influence of plant height on pod length and seeds per pod However, these were also responsible for the reversal of roles of number of fruiting branches on these two traits Plant height, despite being highly correlated with number of primary branches and seed yield, could not have high direct effects either on pods per plant or seed yield because of strong influence of number of primary fruiting branches on seed yield and their negative relationship with plant height The above

35 17 findings were similar to the results reported by Sandhu et al (1980), Singh et al (1980) and Satyan etal (1980) Malik et al (1987) in correlation and path analysis studies found that seed yield per plant was positively and significantly correlated with plant height, primary branches per plant, pods per plant, clusters per plant and biological yield Days to pod initiation, plant height and biological yield had the highest direct positive effects on seed yield per plant. The direct effects of days to flower initiation, days to maturity, clusters per plant, pod length and seeds per pod on seed yield per plant were high and negative Ramana and Singh (1987) studied the character associations in 37 promising varieties and two checks of green gram grown in spring and rainy season and found that in both the seasons, seed yield per plant showed significant correlations with plant height, pods per plant, clusters per plant and seeds per pod In general, phenotypic correlations were smaller than the genotypic correlations in both the seasons Khan (1988) reported a strong positive association of seed yield with number of branches, plant height and seeds per pod. Path analysis revealed that the pod length had maximum positive direct effect followed by number of branches and plant height The direct effect of seeds per pod on seed yield was highly negative The number of pods and number of seeds per plant had negative direct effect on yield The indirect effects of number of pods via number of branches, number of clusters, number of seeds and pod length were high and positive In their studies on growth pattern in relation to yield of green gram a total of 37 Vigna radiata accessions were assessed for yield and yield

36 18 components in 3 successive seasons (early rainy, late rainy and dry) by Lampong ef al (1988) They concluded that early vegetative growth has no direct association with seed yield The degree of indeterminacy exerts a positive influence on yield They further suggested that selection for high yield should be based on the increase in plant height after anthesis and high dry matter production than harvest index Natarajan ef al (1988) in genetic association studies among forty five genotypes of green gram during khanf showed that seed yield was significantly and positively correlated with plant height, clusters per plant, pods per plant and seeds per pod Path analysis revealed that pods per plant (0 49) recorded the highest positive direct effect on seed yield followed by seeds per pod (0 41) The indirect effects of other characters like clusters per plant, plant height and pod length through pods per plant and seeds per pod were also much appreciable, indicating that these characters played an important role in determining yield Such parallelism was reported by Shamuzzaman et al (1981) Patil and Deshmukh (1988) reported significant and positive association of seed yield with hundred seed weight, seeds per pod, plant height, pods per plant, days to maturity and days to flowering in the eighty nine genotypes they studied, suggesting that all these characters had certain inherent relationship with seed yield Path coefficient analysis indicated that days to flower and 100 seed weight had maximum positive direct effect on seed yield Pods per plant had positive direct effect on seed yield and positive indirect effects via days to flower, seeds per pod, but it showed negative indirect effects via plant height, days to maturity, pod length and hundred

37 19 seed weight Singh and Malhotra (1970) and Chandel et al (1973) also reported positive direct effect of pods per plant on seed yield but Sharma and Gupta (1984) reported negative direct effect of pods per plant on seed yield Raut et al (1988) reported positive correlation of seed yield with number of seeds per pod, number of branches per plant and hundred seed weight at genotypic level However, number of pods per plant exhibited negative and non-significant association with seed yield. While studying the inter-relationships of yield and its components in F 3 progenies of a cross in mungbean, Singh et al (1988) reported significant association of seed yield with number of pods per plant, plant height, number of primary branches per plant, pod length, number of clusters per plant, seeds per pod and hundred seed weight Path analysis revealed that pods per plant, hundred seed weight, pod length, number of primary branches per plant and days to flower exerted positive direct effects on seed yield per plant According to Francisco and Maeda (1989), number of days to first flower was positively correlated with plant height Number of primary branches was negatively correlated with reproductive nodes, number of days to first mature pod, duration of reproductive stage and seed yield Lakshmaiah et al (1989) and Satyan et al (1989) showed that seed yield per plant had significant positive correlation with clusters per plant, plant height, number of pods per plant, days to flowering and maturity They concluded that for improving the yield of green gram, emphasis should be given on clusters per plant and pods per plant during selection Similar conclusions were drawn by Pokle and Nomulwar (1978)

38 20 The study of path analysis in eighty nine genotypes of green gram by Patil and Narkhede (1989) showed positive correlation of seed yield with plant height, days to maturity, number of pods per plant, seeds per pod and hundred seed weight indicating that seed yield is largely a function of these characters, which confirms the earlier findings of Joshi and Kabana (1973) and Yohe and Poehlman (1975) They further reported that hundred seed weight had maximum direct influence (0 79) on seed yield followed by pod length (0 58), pods per plant (0.48) and plant height (0 28) Seeds per pod and days to maturity showed negative direct effects on seed yield despite their positive correlation with yield This was attributed to a negative indirect effect via plant height, pods per plant and pod length respectively They further concluded that it would be rewarding to give more stress on hundred seed weight, pod length, pods per plant and plant height in the selection programme of mungbean However, Khan and Ahmed (1989) reported high positive direct effects of seeds per pod on seed yield, number of pods per plant had moderate and pod length had high negative direct effect on seed yield Khorgade etal (1990) in correlation and path analysis studies reported that grain yield was significantly and positively associated with number of pods per plant (0.924), number of clusters per plant (0 658) and plant height (0.759) at genotypic level They also noticed that number of pods per plant, plant height, number of seed per pod, hundred seed weight and shelling percentage exhibited positive direct effects Patel (1991) found that pod weight was positively and significantly correlated with seed yield Pod weight also had a high positive direct effect on seed yield

39 21 Holkar and Raut (1992) reported that days to flowering and maturity were negatively correlated with seed yield Path analysis of yield components in mungbean carried out by Mishra and Yadav (1992) revealed that harvest index, plant height, number of branches and biological yield per plant had a direct positive influence on seed yield Warn et al. (1992) while confirming the result of Choi et al (1986) and Khan (1991) concluded that number of pods per plant and number of clusters per plant should be considered as selection criteria for improvement of yield as these traits had significant positive correlation with yield The correlation and path coefficients were studied in 7 parents and F2 population of their 21 crosses in green gram by Ebenezer Babu Rajan et al (2000) They found that seed yield had significant positive genotypic correlation with number of secondary roots at maturity, dry weight of plants at maturity, plant height, clusters per plant, pods per plant, seeds per pod and hundred grain weight and harvest index Number of pods, clusters per plant and harvest index showed high positive correlation with gram yield and also with each other Path analysis revealed that pods per plant had the highest positive direct effect on grain yield, followed by hundred grain weight on gram yield Supnyo Cakraborty et al (2001) in correlation studies on root characters of 24 green gram genotypes revealed that seed yield was positively correlated with root length, nodules per plant and root dry weight Venkateswarlu (2001b) found that pods per plant, days to maturity plant 1 height, 100 seed weight, seeds per pod and pod length showed significant and positive association with seed yield Pods per plant and seeds

40 22 per pod had maximum positive direct effect on seed yield Days to maturity, clusters per plant, plant, height, 100 seed weight and seeds per pod exhibited high indirect effect on seed yield via pods per plant 2.4 Genetic diversity: Genetic differences as observed between the individuals or genetic stocks with respect to individual trait or an array of traits is termed as genetic diversity To select desirable parents for breeding programme, the knowledge of nature and degree of genetic divergence is useful D 2 statistics, as a tool for quantitative estimates of genetic divergence between the population was described by Mahalanobis (1936) While Rao (1948) suggested more flexible method, which would replace the measurement on large number of characters, all of which contribute in some degree towards discrimination by relatively few measurements Application of original concept of D 2 statistics of Mahalanobis (1928) for assessment of genetic diversity was suggested by Rao (1952) Moll et al (1962) reported no relationship between geographical distribution and genetic diversity and its reflection in expression of heterosis Murthy and Arunachalam (1966) revealed that geographical distribution and genetic diversity could not be directly related in any of the crops Ghaden et al (1979) in the multivariate analysis could not establish' any relationship between geographic and genetic diversity In their studies, number of pod per plant proved to be the most important yield component followed by number of seeds per pod Misra (1986) reported similar results

41 In the multivariate analysis, Thulasidas (1984) revealed that pods per plant, days to maturity, pod length and hundred seed weight contributed per cent of total variation in yield Days to maturity had the highest contribution (61%) towards divergence All the genotypes studied were grouped into 7 distinct clusters Ramana and Singh (1987) in genetic divergence studies in green gram during spring and khanf seasons grouped 39 genotypes in 8 clusters A total of 21 genotypes occurred in cluster 1 in the spring and 28 in cluster 1 in kharif A considerable number of genotypes were common to cluster 1 in both seasons Days to flowering and hundred seed weight contributed the most to genetic divergence in the khahf and spring season, respectively Singh and Pathak (1987) while confirming the results of Shanmugam and Sreerangaswamy (1982) grouped 20 genotypes in 6 clusters based on D 2 values The members of each cluster were geographically unrelated Cluster II with 8 genotypes had the maximum values for pods per plant and seed yield per plant These 8 genotypes together with four others were recommended for use in hybridization to obtain useful segregants in subsequent generations Natarajan et al (1988) grouped 45 genotypes in 5 clusters on the basis of D 2 analysis Seed weight and days to flowering showed the greatest contribution to divergence No correlation was observed between genetic divergence and geographical diversity Twenty genotypes from Punjab were scattered in 3 clusters indicating wide genetic variability in this material Singh (1988) studied the clustering of genotypes for selection for heterosis in yield and response to environmental variations in forty genotypes of mungbean, which were grown in four environments (2 seasons x 2

42 24 spacing) and analyzed for standardized distance (d 2 s (A) 11) for mean grain yield per plot All the genotypes were included in 9 clusters The criterion for grouping was that 2 populations belonging to the same cluster should at least on average show smaller d 2 s (A) H than those belonging to different clusters The greatest number of genotypes found in cluster 1 (21) followed by cluster III (8) and cluster II (5). The remaining clusters were monogenotypic Genotypes in the same cluster were similar in response to environmental changes but genetically diverse for yield Natrajan and Palaniswamy (1990) while doing multivariate and canonical analysis found that divergence between parents as measured by generalized distance, corresponded well with superiority of the Fi over the mid parental value for pod weight and seed yield Naidu (1990) in the genetic analysis of yield and yield components of mungbean suggested the need for multienvironment studies for quantitative assessments of genetic diversity Naidu and Satyanarayana (1991) while studying the 49 diverse genotypes under three different environments recorded data on yield and yield components and observed considerable genetic diversity They grouped the genotypes into 14, 11 and 8 clusters for three respective environments which showed the influence of environment on cluster pattern and also indicated the importance of studying material in more than one environment No relationship between geographic and genetic diversity was observed The parentage had no effect on clustering pattern in khanf and rabi uplands, but had effect in rice fallow Shoot dry weight at harvest and pods per plant made the highest contribution to cluster pattern in all three environments

43 ABSTRACT DIGITIZED 2 Raut era/. (1991) based on D 2 analysis grouped 55 genotypes into 9, 3 and 3 clusters for the spring and summer seasons and for pooled analysis respectively Clustering patterns were not associated with parental and geographical origin High levels of divergence in parents were not generally associated with high heterosis in the hybrids, except clusters per plant In the genetic diversity studies in mungbean, Naresh Chandra et al (1993) suggested that plant height, number of fruiting clusters per plant and number of pods per plant were important parameters contributing towards genetic diversity in mungbean Discriminant function analysis in mungbean by Deb et al (1994) revealed that selection for number of primary branches, leaf area and hundred gram weight was much more efficient (766 81%) than direct selection for yield alone. Mishra et al (1995) while confirming the results of Gupta and Singh (1970), Pokle and Nomulwar (1978) and Tawar and Mishra (1988) reported that the greatest divergence in mungbean was found for seed weight and pod length They further reported that the strains which did not coincide with their geographical distribution could be grouped into clusters Manivannan er al (1998) evaluated thirty green gram genotypes for genetic divergence studies Genotypes were grouped into 8 clusters based on their genetic diversity The highest inter cluster values were observed between clusters VI and VIII, VII and VIII and IV and VII Plant height contributed maximum towards divergence followed by pod length, seeds per pod and clusters per plant

44 26 D 2 analysis revealed a wide genetic variability among 40 genotypes of green gram studied by Venkataknshna Kishore et at (2000), where these genotypes were grouped into 15 clusters. There was no relationship between geographic and genetic diversity as genotypes chosen from same ecogeographical region were found in different clusters as well as in the same clusters. The maximum inter cluster distance was observed between clusters I and XIV (D 2 = ) and was followed by clusters VIII and IX (D 2 = ) clusters III and IX (D 2 = ) and clusters I and IX (D 2 = ) indicating wide divergence among these clusters, which also suggested that the genetic architecture of the genotypes in one cluster differed entirely from those included in other clusters The variance of cluster means revealed that number of pods per plant, days to maturity, days to flowering and plant height were the main characters contributing to the genetic divergence in the present material Genetic diversity was studied by using multivariate analysis among 42 F 3 and eight varietal genotypes of Vigna radiata by Loganathan et a/ (2001a) The grouping of material into seven clusters indicated the presence of wide range of genetic diversity among the genotypes The study indicated no definite relationship between geographic and genetic diversity and geographic diversity can not be used as an index of genetic diversity In general, genetic diversity among the parents was reflected in their progenies Seed yield per plant contributed maximum accounting for 41 4% of total divergence

45

46 MATERIAL AND METHODS

47 a? 3. MATERIAL AND METHODS The present investigation, "Variability, path analysis and genetic diversity in green gram (Vigna radiata (L.) Wilczek)" was conducted at Botany Farm, college of Agriculture, Pune The details of the material used and methods adopted to carry out the experiment and statistical procedures followed are described in this chapter 3.1 Material The experimental material consisted of 40 genetically diverse genotypes obtained from Pulses Breeder, Pulses improvement Project, Mahatma Phule Krishi Vidyapeeth, Rahun and Agricultural Research Station Badnapur Marathwada Krishi Vidyapeeth, Parbhani, Maharashtra The list of genotypes and their source is given in Table Methods Experimental design The experiment was conducted in a randomized block design with three replications Each plot consisted of a single row of 5.0m length with a spacing of 30 cm between rows and 10 cm between plants Sowing and cultural practices The land was prepared by ploughing followed by two cross harrowings and planking The basal dose of 25kg N/ha and 50kg PaOs/ha was applied at the time of sowing in the form of urea and single super phosphate, respectively

48 28 Table 3.1 The genotypes and their source Sr No Genotypes Source 1 AKM 8802 Agricultural Research Station, Badnapur, M S 2 AKM 9601 Agricultural Research Station, Badnapur, M S 3. AKM 9602 Agricultural Research Station, Badnapur, M S 4 AKM 9801 Agricultural Research Station, Badnapur, M.S 5 AKM 9910 Agricultural Research Station, Badnapur, M S 6. IC NBPGR, New Delhi 7 BPMR1 Senior Pulses Breeder, M.P K V, Rahun.M S 8 BPMR 128 Senior Pulses Breeder, M P.K.V, Rahun.M S 9 BPMR 132 Senior Pulses Breeder, M P K V, Rahuri.M S 10 BPMR 145 Senior Pulses Breeder, M P.K V; Rahun.M S 11. BPMR 207 Agricultural Research Station, Badnapur, M.S. 12 J 781 Agricultural Research Station, Badnapur, M S 13 KOP Agricultural Research Station, Badnapur, M S 14 PM Senior Pulses Breeder, M.P K V, Rahuri.M S 15. PM 9339 Senior Pulses Breeder, M P K V, Rahuri.M S 16 PM9340 Senior Pulses Breeder, M P K.V, Rahuri.M S 17 PM 9341 Senior Pulses Breeder, M P K V; Rahuri.M S 18 PM 9342 Senior Pulses Breeder, M P.K V, Rahun.M.S. 19 PM 9343 Senior Pulses Breeder, M P.K.V; Rahuri.M S 20 PM 9346 Senior Pulses Breeder, M.P.K V; Rahuri.M S 21 PM 9348 Senior Pulses Breeder, M P K V, Rahuri.M S 22 PM 9352 Senior Pulses Breeder, M P K V, Rahuri.M S 23 PM 9353 Senior Pulses Breeder, M P.K.V, Rahuri.M S 24 PM 9357 Senior Pulses Breeder, M P K V; Rahuri.M S 25 PM 9361 Senior Pulses Breeder, M P K V, Rahun.M S 26 PM 9362 Senior Pulses Breeder, M P.K V; Rahuri.M S 27 PM 9376 Senior Pulses Breeder, M P K V; Rahuri.M S 28 PM 9377 Senior Pulses Breeder, M P K V, Rahun.M S 29 PM 9378 Senior Pulses Breeder, M P K V; Rahuri.M S 30 PM 9380 Senior Pulses Breeder, M.P.K V, Rahuri.M S 31 PM 9381 Senior Pulses Breeder, M P K V, Rahuri.M S 32 PM 9383 Senior Pulses Breeder, M P K V, Rahuri.M S 33 PM 9384 Senior Pulses Breeder, M P.K V, Rahuri.M S 34 PM 9734 Senior Pulses Breeder, M P.K V, Rahuri.M S 35 PM 9877 Senior Pulses Breeder, M P K V, Rahuri.M S 36 VC 3960 Senior Pulses Breeder, M P K V, Rahuri.M S 37 VC6173 Senior Pulses Breeder, M P K V, Rahuri.M S 38 VC 6372 Senior Pulses Breeder, M P K V, Rahuri.M S 39 IC NBPGR, New Delhi 40 IC NBPGR, New Delhi

49 5 1?' The seeds were sown on 6 th July, 2002 The usual cultural practices like weeding, irrigation, plant protection measures etc were followed as and when required during the growth period of the crop Harvesting The pods were picked after attaining their physiological maturity in order to avoid shattering Observations recorded Following observations were recorded on five randomly selected plants from each treatment in each replication and averages were worked out Days to 50 per cent flowering (Nos.) Number of days required from date of sowing to the date on which 50 per cent of the plants flowered was recorded Days to maturity (Nos.) Number of days required from the date of sowing till the physiological maturity were considered as days to maturity Plant height (cm) Height was measured from ground level to uppermost tip of the plant in centimeters at maturity on randomly selected five plants Number of branches per plant The branches of the observational plant were counted and recorded at the time of harvest Pod length (cm) The length of five different pods from a single plant was measured with a scale and average was worked out and recorded as the pod length of that plant

50 so Similarly, pod length was recorded for five randomly selected plants and the average was estimated Number of pods per plant The number of pods during every picking were recorded and total number of pods per plant was calculated Number of seeds per pod The seeds of five different pods randomly selected from a single plant were counted and the average was worked out for recording number of seeds per pod Number of nodules per plant: The number of nodules during the flowering stage were recorded by uprooting the plant and total number of nodules per plant was calculated Number of clusters per plant Number of clusters per plant were counted at maturity Protein content (per cent) Per cent crude protein content of the green gram sample was estimated by determining nitrogen content of seeds adopting Macrokjeldahl Method (AOAC, 1975) Total nitrogen content was multiplied by factor 6 25 which gave protein content of that sample Harvest Index (per cent) The harvest index was recorded by taking the ratio of economic yield (seed yield) and biological yield (total dry matter produced by plant) as a percentage on randomly selected five plants

51 ai Test weight (g) This was recorded by taking the weight of hundred dry seeds from a random sample in each replication Seed yield per plant (g) The weight of seed of each of the five randomly selected plants was recorded and average was worked out to represent seed yield per plant 3.3 Statistical Analysis: The mean values of five randomly selected observational plants for thirteen different characters were used for statistical analysis The following statistical parameters were calculated for presentation of data on different quantitative attributes Analysis of Variance (ANOVA) - The analysis of variance was done as suggested by Panse and Sukhatme (1985) in the following form Sr. No. Source of variation D.F. Expected Mean Sum of squares 1 Replication r-1 a 2 e +1 a 2 r 2 Treatment t-1 a 2 e + r a 2 t 3 Error (r-1) (t-1) a 2 e Total (rt-1) Where, r = number of replications t = number of treatments

52 3R Genotypic Coefficient of Variation (GCV) It was estimated by the formula suggested by Burton (1952) GCV =V a 2 g / 8 x 100 Where, a 2 g = Vg = Genotypic variance 8 = General mean of the character Phenotypic Coefficient of Variation (PCV) It was estimated by the formula suggested by Burton (1952) PCV = V a 2 p / 8 x100 Where, a 2 p = Vp = Phenotypic variance 8 = General mean of the character Heritability percentage Heritability percentage in board sense was calculated as given by Burton (1952) h 2 (bs) = Vg/Vpx100 or h 2 (b s) = o^g/^pxioo Where, h 2 (b s) = Heritability percentage in broad sense a 2 g = Vg = Genotypic variance a 2 p = Vp = Phenotypic variance Genetic advance Genetic advance was calculated by the formula given by Johnson et al (1955)

53 33 Where, 2 / 2, G A = k x (erg / crp) x ap or G A = k x h 2 x ap k = selection differential which is 2 06 at 5 per cent selection intensity o 2 g = Vg = Genotypic variance a 2 p = Vp = Phenotypic variance ap = phenotypic standard deviation Correlation To understand the association among the characters, genotypic and phenotypic correlation coefficients were worked out by adopting the method described by Singh and Chaudhary (1977) Phenotypic Correlation Coefficient rp = Covanance X, Y (P) / V Variance X (P) Variance Y (P) Where, rp = Phenotypic correlation coefficient between character X and Y P = Phenotypic Genotypic Correlation Coefficient rg = Covanance X, Y (g_) / V Variance X (g) Variance Y (g) Where, rg = Genotypic correlation coefficient between character X and Y g = Genotypic

54 3tf Significance of correlation coefficients were tested by using "t" test (Panse and Sukhatme, 1985) Path analysis Path coefficient analysis was done according to the procedure suggested by Dewey and Lu (1959). If 'Y' is the effect and X^ is the cause, the path coefficient for the path from cause Xi to the effect Y is ox\l oy Direct and indirect effects were worked out by using genotypic correlations as below: Direct effect of Xi on Y = PX, Y Where, PXi = Path coefficient of X, on Y Similarly, direct effects of other attributes on yield were worked out Where, Indirect effect of X^ via X 2 on Y = PX 2 Y X r X 1 X 2 PX 2 Y = Path coefficient of the component character X 2 on Y r Xi X 2 = Genotypic correlation between Xi and X 2 Similarly, indirect effects in all possible combinations were calculated for all component characters The residual effect (R) was calculated as below R = [1- (PXiY. rxiy) - (PX 2 Y rx 2 Y)- - (PXnY rxny) f Where, PX^, PX 2 Y,., PXnY = Direct effects of respective characters on seed yield

55 3^ rx-iy, rx2y,, rxny = Correlation coefficient between respective characters and yield Mahalanobis generalized distance (D 2 ) The generalized distance between two populations was defined by Mahalanobis (1936) as 3f=s:z: A i j di dj Where, A i j = Reciprocal matrix to the common dispersion matrix. di = Difference between the mean values of two populations for i th character dj = Difference between the mean values of two populations for j th character Estimation of D 2 values from the above formula is very complicated in the present study Since, it requires the inversion of a thirteenth order determinant and then the evaluation of B (BH) / 2 terms whose sum is D 2 It was found convenient to work with a set of uncorrelated characters constructed from the original measurements. D 2 with such transformed variables reduces to the evaluation of a simple sum of squares. Transformation was done by using pivotal condensation method (Singh and Chaudhary, 1977). The coefficients for the transformation were obtained by dividing the first row of the reduced matrix by the square root of the corresponding pivotal condensation elements Determination of gene constellations Tocher's method as described by Rao (1952) was followed for cluster formation No formal rules can be laid down for finding the clusters because a

56 3& cluster is not a well defined term The only criterion appears to be that any two groups belonging to the same cluster should be at least on an average shows a smaller D 2 than those belonging to two different clusters A simple device suggested by K D Tocher is to start with two closely associated groups and find a third group which has the smallest average D 2 from the first two Similarly, the fourth is chosen to have the smallest D 2 from the first three and so on If at any stage the average D 2 of a group from those already listed appears to be high, then this group does not fit in with the former groups and is therefore taken outside the former cluster The groups of the first cluster are then omitted and the rest are treated similarly It is also useful to calculate the change in average D 2 within a cluster due to the inclusion of an additional group. If the changes are appreciable then the newly added group has to be considered as outside the cluster Average intra and inter cluster D 2 and D values Average intra cluster D 2 D 2 = Z D 2 i / n Where, Di is the sum of distances between all possible combinations (n) of the population included in a cluster Average inter cluster D 2 D 2 = Z Distances between the population of cluster I and J / ni nj where, ni = Number of populations in the cluster i nj = Number of population in the cluster j

57 3^ Average intra and inter cluster distance D = VD Cluster Means Cluster means were calculated for individual character on the basis of mean performance of the genotype included in that cluster Genetic diversity as an index for selecting desirable parents for hybridization The possible limits to parental divergence within which there were reasonably high chances for occurrence of heterosis were calculated following Aurunachalam and Bandopadhyay (1984) They advised to delineate the divergence among parents into four divergence classes To take into account the variable in parental divergence, the mean (m) and standard deviation (s) of the values of divergence were calculated The divergence classes were defined as follows DCT = D > or = (m+s) DC 2 = D < (m+s) and > or = m DC 3 = D > or = (m-s) and < m DC 4 = D < (m-s) They postulated that two parents whose genetic divergence falls between (m-s) and (m+s), i e in the classes DC2 and DC3 when crossed have higher chances of producing high frequency and magnitude of heterosis, when compared to a cross whose parental divergence falls outside the limits [(m-s), (m+s)].

58

59 EXPERIMENTAL RESULTS

60 4. EXPERIMENTAL RESULTS The present investigations on green gram (Vigna radiata (L.) Wilczek) were undertaken with a view to know the variability, path analysis and genetic diversity of forty green gram genotypes in Kharif 2002 The results obtained on various parameters are presented in this chapter 4.1 Analysis of variance The analysis of variance (Table 41) revealed highly significant differences among the genotypes for all the thirteen characters studied 4.2 Mean Performance The data on mean performance of 40 genotypes for thirteen characters is presented in Table Days to 50 per cent flowering The population mean for this character was days The genotypes PM 9352 (42 33) and VC 6372 (42 33) were the earliest to flower, followed by BPMR-132 (43 0), PM 9383 (43 0) and IC (44 0) Nineteen genotypes (47 5%) required less days to 50 per cent flowering than population mean The genotype PM 9342 (52 33) was late to flower Days to maturity The genotype VC 6372 (63 00) was the earliest to mature, followed by PM 9353 (65 00), PM 9383 (67 33), IC (69 33) and IC (69 67) Twenty one (52 5%) out of forty genotypes showed significantly early maturity

61 31 Table 4.1. Analysis of variance (M.S.S.) for 13 characters in green gram. Sr. No 1. Character Days to 50% flowenng(no) Mean sum of squares Replications Treatment Error (2) (39) (78) ** ** Days to maturity (no) ** ** Plant height (cm) ** ** No of branches per plant ** ** Pod length (cm) ** ** No of pods per plant ** ** No of seeds per pod * ** No of nodules per plant ** ** No of clusters per plant ** ** Protein content (%) ** ** Harvest index (%) ** ** Test weight (g) ** ** Seed yield per plant (g) ** ** denote significance at 5 and 1 per cent level of probability respectively. Figures in parentheses indicate the degree of freedom

62 *!& when compared with the population mean of days The late maturing genotypes included PM (85 00), BPMR-128 (84 00), PM 9342 (83 67), BPMR-145 (82 00) and PM 9734 (82 00) Plant height (cm) Sixteen (40 0%) out of forty genotypes were taller than the population mean of cm The genotype PM 9341 was the tallest which recorded maximum plant height of cm followed by PM 9339 (67 33 cm), PM 9381 (65 93 cm), PM 9383 (65 20 cm), PM 9342 (64 67 cm) and BPMR-145 (63 23 cm) The genotype IC was the dwarfest with only cm height Along with it, PM 9362 (43 33 cm), AKM 8802 (45 02 cm), AKM 9801 (48 00 cm), VC 6372 (48 49 cm) and PM 9348 (49 53 cm) recorded less than 50 cm height Number of branches per plant The population mean for this character was 4 0 branches and the variation ranged between 2 93 (VC 6372) and 6 10 (IC 00557) The genotypes IC (5 77), BPMR-1 (5 27), BPMR-128 (4 73) and BPMR-132 (4 70) produced high number of branches Seventeen (42 5%) out of forty genotypes, showed higher number of branches over population mean Pod length (cm) The population mean for this character was 8 11 cm and the variation ranged between 6 51 (VC 3960) and 9 87 cm (PM9340) Nineteen (47 5%) of the forty genotypes showed more pod length than the population mean of 8.11 cm The genotype PM9340 (9.87 cm) showed the maximum pod length followed by PM 9342 (9 59 cm), PM 9341 (9 41 cm), PM 9384 (9 29 cm) and

63 Table No. 4.2 Mean performance of 40 green gram genotypes for different characters. Genotypes Days to 50% flowering Days to maturity Plant height (cm) No. of branches/ plant Pod length (cm) No. of pods/ Plant No. of seeds/ pod No. of nodules /plant No. of clusters/ plant Protein content (%) Harvest Index Test weight (g) Seed yield/plant (g) AKM AKM AKM AKM AKM IC BPMR BPMR BPMR BPMR BPMR J KOP PM PM PM PM PM PM PM PM PM PM PM PM PM PM PM Cont *~

64 Genotypes Days to 50% flowering Days to maturity Plant height (cm) No. of branches/ plant Pod length (cm) No. of pods/ Plant No of seeds/ pod No of nodules /plant No. of clusters/ plant Protein content (%) Harvest Index Test weight (g) Seed yield/plant (g) PM PM j PM9381 J J PM PM PM PM VC VC VC IC IC Mean SE C D (5%) CVO/ ^

65 S PM 9376 (9 27 cm) The genotypes VC 3960 (6 51 cm), IC (6 57 cm), IC (6 69 cm) and BPMR-128 (6 71 cm) recorded less pod length Number of pods per plant Maximum number of pods were harvested from the genotype IC (32 50), followed by BPMR-128 (29 60), BPMR-1 (27 47), PM 9734 (24 87) and AKM 9801 (23 60) The genotypes PM 9362 (11 60) and AKM 8802 (12 4) recorded the lowest number of pods per plant Out of forty genotypes, twenty (50%) produced higher pods per plant than the population mean (20 53) Number of seeds per pod The population mean for this character was and the variation ranged between (IC 00114) and (PM 9361) Twenty one genotypes (52 5%) recorded higher seeds per pod than the population mean The genotypes PM 9361 (13 53) recorded maximum number of seeds per pod followed by BPMR-207 (13 37), PM 9348 (12 93), PM 9357 (12 80), PM 9340 (12 73) and PM 9346 (12 73) Number of nodules per plant The variation for this character ranged between (BPMR-132) and (PM 9339) Looking to the mean of nodules per plant, twenty two genotypes (55%) recorded higher nodules per plant than the population mean The genotypes PM 9339 (54 93), AKM 9910 (54 0), PM 9380 (53 13), PM 9353 (52 67), BMPR-1 (51.47) and PM 9376 (51 33) recorded high number of nodules per plant

66 ** Number of clusters per plant The genotype IC (9 7) recorded the maximum number of clusters per plant followed by PM 9381 (7.2), BPMR-1 (7 13), PM (7 13), PM 9734 (7 13) and PM 9339 (6 87) Only twenty two (55%) out of forty genotypes produced higher number of clusters than the population mean of 5 92 cluster per plant The genotype AKM 8802 (3 53) produced the least number of clusters per plant Protein content (%) The variation for protein content ranged between (PM 9381) and per cent (PM 9339) The genotype PM 9339 (25 04%) recorded highest protein content followed by AKM 9910 (24 09%), PM 9380 (24 88%), IC (24 81%), PM 9734 (24 45%) and PM 9377 (24 37%) Twenty two (55%) out of forty genotypes showed higher protein content than the population mean of per cent. The genotypes BPMR-132 (18 14%), AKM 9602 (18 83%) and PM 9346 (19 31%) recorded less protein content (%) Harvest Index (%) The variation for this character ranged between (PM 9340) and per cent (PM 9339) Looking to the mean of 37 54, seventeen genotypes (42.5%) recorded higher harvest index than the population mean The genotypes VC 3960 (43.43%), PM 9361 (43 28%), PM 9348 (42 64%) and AKM 9801 (42 1%) recorded high per cent of harvest index Test weight (g) Seventeen (42 5%) genotypes recorded higher values for test weight than the population mean of 3 97 g The genotype PM 9340 (5 47g) recorded the maximum test weight followed by PM 9362 (4 97g), PM 9383 (4 83g), PM

67 J* (4 80g) and PM 9376 (4 7g) The genotype IC (3 07g) exhibited the minimum test weight Seed yield per plant (g) The genotype PM 9376 recorded the highest seed yield per plant (9.33g) followed by KOP (9.31g), PM 9380 (8.08g), PM 9342 (8 03g) and BPMR- 132 (7.8g) The genotypes VC 3672 (5 11g) AKM 9910 (5 44g) and PM 9377 (5 71g) produced comparatively low yield. The variation ranged between 5 11 (VC 3672) and 9 33 (PM 9373) Nineteen (47 5%) genotypes recorded higher seed yield than population mean of 7 03g 4.3 Parameters of genetic variability The parameters of genetic variability viz., variability, range, GCV, PCV, hentability (bs), genetic advance, genetic advance as per cent of mean are summarized in Table 4 3 The important findings are discussed below Coefficients of variation (genotypic and phenotypic) It was observed that the estimates for genotypic coefficient of variation (GCV) were lower than the phenotypic coefficients of variation (PCV) for all the characters The highest GCV was recorded for test weight (14.8) followed by number of pods per plant (12 7), number of branches per plant (10 2), pod length (9 5), plant height (9 1) and protein content (8.3). The character harvest index (3 1) recorded magnitudinally low GCV, whereas 50 per cent flowering (6.3) recorded the lowest PCV Number of pods per plant (31 2), showed the maximum PCV followed by number of clusters per plant (29 8), number of branches per plant (26 2), seed yield per plant (24 1), harvest index (16 6), test weight (15 1), plant height (14 8) and pod length (12 3)

68 Table 4.3 Parameters of genetic variability in 40 genotypes of green gram. Sr. No. Character Range General mean GCV PCV 1 3ays to 50% llowenng Heritability (bs) Genetic advance G.A as a per cent of mean Days to maturity Plant height (cm) Number of branches per plant Pod length (cm) Number of pods per plant Number of seeds per pod Number of nodules per plant Number of clusters per plant , Protein content (%) Harvest index (%) Test weight (g) Seed yield per plant GCV = Genotypic coefficient of variation PCV = Phenotypic coefficient of variation bs = Broad sense G A = Genetic advance G\

69 ^ The highest magnitudinal difference between GCV and PCV was recorded for number of clusters per plant (24 4) followed by number of pods per plant (18 5), seed yield per plant (18 1), number of branches per plant (15 9) and harvest index (13 4) The lowest difference between GCV and PCV was found for days to maturity (0 154) Heritability percentage (broad sense) The heritability (bs) estimates varied between 3 2 (number of clusters per plant) to 96 0% (test weight) The highest heritability (bs) was obtained for test weight (96.0%) followed by days to maturity (95 7%), protein content (90 0%), days to 50 per cent flowering (87 5%), pod length (59 4 %), number of nodules per plant (53 3%) and plant height (37 3%) The lowest heritability was recorded for number of clusters per plant (3 2%) followed by harvest index (3 6%), seed yield per plant (6 0%), number of branches per plant (15 3%), number of pods per plant (16 5%) and number of seeds per pod (17 7%) Genetic advance The character days to maturity (10 6) showed the highest genetic advance followed by plant height (6 4), days to 50 per cent flowering (5 5), number of nodules per plant (4 8) and protein content (3 7) The lowest genetic advance was exhibited by number of clusters per plant (0 11) The other characters viz, seed yield per plant (0 21), number of branches per plant (0 33), number of seeds per plant (0 34), harvest index (0 46), test weight (1 19), pod length (1 22) and number of pods per plant (2.19) also showed low genetic advance However, the characters days to maturity and

70 *t* days to 50 per cent flowering recorded high genetic advance accompanied by high hentability estimates High values of genetic advance as per cent of mean was observed for test weight (30 0), followed by protein content (16 3), pod length (15 1), days to maturity (13 9), days to 50 per cent flowering (11 4) and plant height (11 4), whereas, harvest index (1 2) showed the lowest value 4.4 Correlation The phenotypic and genotypic correlation coefficients between yield and yield contributing characters are presented in Table Association between grain yield and its components The grain yield per plant showed highly significant positive association with plant height (0 82), number of clusters per plant (0 56), number of seeds per pod (0 42) and test weight (0 40) at genotypic level The characters days to maturity (0 37) and days to 50 per cent flowering (0 31) have also recorded significant positive association with seed yield per plant However, harvest index (-3 24) recorded highly and significantly negative association with seed yield per plant at genotypic level The characters, pod length (0 22), number of pods per plant (0 15) and number of nodules per plant (0 25) showed positive but non significant association with seed yield per plant The characters, number of branches per plant (-0.19) and protein content (-0 26) showed negative but non-significant association with seed yield per plant at genotypic level The phenotypic correlations of plant height (0 46), number of clusters per plant (0 43) and number of pods per plant (0 41) showed highly significant

71 and positive association with grain yield, while the characters days to 50 per cent flowering (0 15), days to maturity (0 18), number of branches per plant (0 12), pod length (0 09), number of seeds per pod (0 28), number of nodules per plant (0 02), harvest index (0 18) and test weight (0 19) showed positive but non-significant association with grain yield per plant at phenotypic level However, protein content (-018) showed negative but non-significant association with grain yield per plant Association between yield components The association between component characters at genotypic and phenotypic levels is presented below Days to 50 per cent flowering was significantly and positively correlated with days to maturity, number of clusters per plant, number of seeds per pod, plant height and number of pods per plant at genotypic level, whereas it showed significant and positive correlation with days to maturity at phenotypic level Days to 50 per cent flowering was significantly and negatively correlated with harvest index at genotypic level and showed negative but non-significant correlation at phenotypic level. Days to maturity was significantly and positively correlated with number of clusters per plant, number of pods per plant and plant height, whereas it showed significant negative correlation with harvest index at genotypic level Plant height was significantly and positively correlated with number of clusters per plant, test weight and pod length whereas it showed significant negative correlation with number of branches at genotypic level. Plant height was significantly and positively correlated with number of clusters per plant and test weight at phenotypic level Number of branches per plant was *?

72 Characters Table 4.4 Genotypic (above diagonal) and phenotypic (below diagonal) correlation of 13 characters in green gram. Days to 50% flowering Days to maturity Plant height (cm) No of branches/ plant Pod length (cm) No of pods/ plant No of seeds/ pod No of nodules/ plant No of clusters/ plant Protein content (%) Harvest index (%) Test weight (g) seed yield/ plant (g) ** * * ** ** ** * " * ** ** ** * ** * ** ** ** ** ** ** ** ** ** ** ** * ** ** ** ** * * ** ** ** ** * * ** ** * ** ** ** ** ** ** * ** * ** ** ** ** ** denotes significance at 5 and 1 per cent level of significance, respectively o

73 51 significantly and positively correlated with number of clusters per plant at both genotypic and phenotypic levels, where as it showed significant negative correlation with pod length and number of seeds per pod at genotypic level Pod length was significantly and positively associated with number of seeds per pod and test weight at both genotypic and phenotypic levels, however it was significantly and negatively correlated with number of pods per plant at both genotypic and phenotypic levels Number of pods per plant was significantly and positively correlated with days to maturity, days to 50 per cent flowering and negatively correlated with number of seeds per pod and harvest index at genotypic level, however it was significantly and positively correlated with number of clusters per plant at phenotypic level Number of seeds per pod was significantly and positively correlated with test weight, harvest index, pod length and days to 50 per cent flowering at genotypic level, whereas it was significantly and negatively correlated with protein content, number of pods per plant and number of branches per plant at genotypic level. Number of nodules per plant had significant and positive association with protein content at both genotypic and phenotypic levels, where as it showed significant and positive correlation with number of clusters per plant and harvest index at genotypic level Number of clusters per plant was significantly and negatively correlated with harvest index, pod length and test weight at genotypic level, however, it was significantly and positively correlated with protein content, number of nodules per plant, number of branches per plant, plant height, days to maturity and days to 50 per cent flowering at genotypic level Harvest index was significantly and negatively correlated with test weight, number of cluster per plant, number of pods per

74 plant, days to maturity and days to 50 per cent flowering at genotypic level Test weight was significantly and positively correlated with plant height, pod length and number of seeds per pod at genotypic level $a 4.5 Path analysis The direct and indirect contributions of each character towards seed yield per plant revealed by path analysis are presented in Table 4 5 Since the magnitude of genotypic correlation coefficient being more important, are only considered for path analysis Direct effects Among the 13 characters studied protein content (9 34) recorded magnitudinally the highest positive direct effect on grain yield per plant followed by plant height (6 58), test weight (4 7) and days to 50 per cent flowering (4 7). The other characters viz., pod length (-5 53), number of seeds per pod (-3 68) and number of nodules per plant (-7 67) showed negative direct effect of high magnitude Number of branches per plant (1 68) and harvest index (1 38) exhibited positive direct effect of low magnitude The characters plant height, days to 50 per cent flowering and test weight recorded the maximum and positive magnitude of direct effect on seed yield per plant and their association with seed yield was also highly significant and positive However, the character protein content had high magnitude of positive direct effect but negative and non-significant association with seed yield per plant. Number of seeds per pod and number of clusters per plant and days to maturity had negative direct effect but positive and significant association with seed yield per plant

75 Table 4.5 Direct (diagonal) and indirect (above and below diagonal) path effects of different characters towards yield in green gram. Characters Days to Days to Plant No of Pod No of No of No of No of Protein Harvest Test Correlation 50% maturity height branches/ length pods/ seeds/ nodules/ clusters/ content index weight (g) With Seed flowering (cm) plant (cm) plant pod plant plant (%) (%) yield/plant (g) * * ** ** ** * ** denotes significance at 5 and 1 per cent level of significance, respectively Ol OJ

76 5\ Indirect effects Looking to the indirect effects of various characters it was observed that the trait, days to 50 per cent flowering contributed indirectly via plant height (2 22) and test weight (1 21) Like wise days to maturity contributed via days to 50 per cent flowering (4.43) and plant height (2 36) Plant height played its role via test weight (2 16) and days to 50 per cent flowering (1.59) whereas number of branches per plant contributed via pod length (2 74) and number of nodules per plant (1 92) The pod length contributed via test weight (4 40) and plant height (2 22) Number of pods per plant contributed mainly through pod length (5 03) and number of seeds per pod (2 09) Number of seeds per pod contributed via test weight (3 82), number of nodules per plant (2 07) and days to 50 per cent flowering (2 04) Number of nodules per plant played its role via protein content (7.69) Number of clusters per plant contributed via pod length (4 81),plant height (3 73), protein content (3 12) and days to 50 per cent flowering (2 17) and test weight contributed via plant height (3 02) 4.6 Divergence: D 2 values estimated by Mahalanobis D 2 statistics, between all possible pairs of 40 genotypes ranged between (PM 9348 and PM 9380) to (PM 9340 and IC 00114) Cluster formation: The cluster formation was done by following Tochers method, as described by Rao (1952) All the 40 genotypes studied under investigation were grouped into seven clusters

77 #r Table 4.6 Distribution of 40 genotypes into different clusters. No of Cluster No. genotypes included 1 14 II 9 III 8 Genotypes AKM 8802, AKM 9602, J 781, KOP, PM 9339, PM 9343, PM 9348, PM 9352, PM 9361,PM 9376, PM 9380, PM 9877, VC 6173, IC AKM 9801, AKM 9910, BPMR 1, BPMR 128, BPMR 145, PM , PM 9377, PM 9734, VC 3960 BPMR 207, PM 9341, PM 9342, PM 9346, PM 9357, PM 9362, PM 9378, PM 9384 IV 4 IC 73430, PM 9353, VC 6372, IC V 3 AKM 9602, BPMR 132, PM 9381 VI 1 PM 9340 VII 1 PM 9383

78 Table 4.7 Average intra and inter cluster D and D values of 7 clusters from 40 genotypes of green gram. Clusters Cluster I Cluster II Cluster III Cluster IV Cluster V Cluster VI Cluster VII Cluster I (7.008) Cluster II Cluster III Cluster IV Cluster V Cluster VI Cluster VII (10.527) (7 025) (10.549) (12.741) (7.201) (11.407) (15 248) (18.396) (7.562) (11.811) (14 850) (14.836) (13.064) (6.104) (14.449) (18.797) (9.302) (14.951) J19.414) (0.000) (11.494) (19 043) (14.956) (12.550) (13 867) (14 475) (0.000) ON

79 6l- Cluster I with 14 genotypes emerged as the largest cluster followed by cluster II with 9 genotypes Likewise, cluster III contained 8 genotypes, cluster IV contained 4 and cluster V contained 3 genotypes whereas the cluster VI and VII were monogenotypic The distribution of 40 genotypes in different clusters are presented in Table Intra and inter cluster distances Intra and inter cluster D 2 and D values from divergence analysis are presented in Table 4 7 The highest intra-cluster distance was observed for cluster IV (D 2 = 57 19) followed by cluster III (D 2 = 51 86), cluster II (D 2 = 49 36) and cluster I (D 2 = 49 12) The lowest intra-cluster distance was observed for cluster V (37 92) The clusters VI and VII showed no intra-cluster distance being monogenotypic The maximum inter-cluster distance was observed between cluster V and VI (D 2 = ) followed by cluster II and VII (D 2 = ), cluster II and cluster VI (D 2 = ) and cluster III and IV (D 2 = ) The mtercluster distance between cluster III and VI (D 2 = 86 53), cluster I and II (D 2 = ), cluster I and III (D 2 = ) and cluster I and IV (D 2 = ) was comparatively low The cluster I was the most distant from cluster VI (D 2 = ), followed by cluster V (D 2 = ) and cluster VII (132 11) The D 2 values between cluster I and II (110 83) and cluster III (11128) were of comparatively low in magnitude. The distance between clusters II and VII (D 2 = ) was the highest followed by cluster VI (D 2 = ), cluster IV (D 2 = ) and cluster V (D 2 = ) while, it showed the lowest D 2 value

80 5% from cluster I (110 83) Cluster III was the most distant from cluster IV (D 2 = ) followed by cluster VII (D 2 = ) and cluster V (D 2 = ) Cluster III was closer to cluster VI (D 2 = 86 53) Cluster IV was closest to cluster I (D 2 = ) and cluster VII (D 2 = ), whereas it showed maximum distance from cluster VI (223 56) Cluster V showed the highest D 2 value from cluster VI (D 2 = ) It showed the shortest distance from cluster I (D 2 = ) Cluster VI was most distantly placed from cluster V (D 2 = ), followed by cluster II (353.36) and cluster IV (223 56) The minimum D 2 value of cluster VI was observed with cluster III (D 2 = 86 53) Cluster VII recorded maximum distance from cluster II (D 2 = ), followed by cluster III (D 2 = ), whereas cluster I (D 2 = ) and cluster IV (D 2 = ) recorded lower D 2 values from cluster VII 4.7 Cluster means Cluster means for 13 characters studied are presented in Table 4 8 It revealed wide range of variability for most of the characters Days to 50 per cent flowering The genotypes in the cluster VII were early for days to 50 per cent flowering (43 00) followed by cluster IV (43 417), cluster V (45 889) and cluster I (47 50) Whereas, the genotypes in cluster III (50 625) and cluster VI (50 333) were late for days to 50 per cent flowering

81 4.7.2 Days to maturity *f The cluster mean for days to maturity varied between (cluster IV) and (cluster III) The cluster IV was the earliest to mature (66 75) The cluster III was late taking days to mature, followed by cluster II (80 48) andclustervi(80 00) Plant height (cm) Cluster means for this character ranged between (cluster IV) and (cluster VII) The highest cluster mean of cluster VII was followed by cluster VI (61 86), cluster III (59 49) and cluster V (58 92) Number of branches per plant Cluster VI (4 40) showed the highest cluster mean while cluster VII (3.26) showed least mean value The highest cluster mean of cluster VI was followed by cluster IV (4 37), cluster I (4 03) and cluster II (3 95) Pod length (cm) The highest cluster mean for pod length was recorded by cluster VI (9 86) followed by cluster III (8 96), cluster VII (8 91) and cluster I (8 11) The cluster II (7.33), cluster V (7 60) and cluster IV (7 83) were recorded comparatively low values Number of pods per plant A wide variability was noted for this character. High cluster mean values were recorded by cluster II (23 68), cluster I (20 89) and cluster V (20.86), whereas lower values were observed for cluster VII, VI, IV and111 having values 16 66, 16 86,18 05 and 18 37, respectively

82 Table 4.8 Mean performance of clusters for thirteen characters in green gram. Cluster No. Days to 50% flowering Days to maturity Plant height (cm) No of branches per plant Pod length (cm) No of pods per plant No of seeds per pod No of nodules per plant No of clusters per plant Protein content (%) Harvest index (%) Test weight (g) Seed yield per plant (g) I II III ' IV V VI VII ON O

83 a Number of seeds per pod The variation for this character ranged between (cluster IV) and (cluster VI) The highest cluster mean value of cluster VI was followed by cluster III (12 71), cluster VII (12 53) and cluster I (12 37) Number of nodules per plant The highest cluster mean was recorded by cluster II (49 77), followed by cluster IV (49 19), cluster I (48.17) and cluster VII (47 33) The lowest cluster mean was recorded by cluster V (40 84), followed by cluster VI (45 0) and cluster III (46 66) Number of clusters per plant Cluster means for this character ranged between 5 26 (cluster VII) and 6 28 (cluster II). The highest cluster mean was recorded for cluster II followed by cluster V (6 06) and cluster I (6 05) Protein content (%) Cluster mean for this character was maximum for cluster IV (24 06) While the lowest cluster mean value was shown by cluster V (17 73) Remaining clusters exhibited intermediate cluster mean values with very low range of variability Harvest index (%) The highest cluster mean was recorded for cluster VII (39 38) followed by cluster I (39 28), cluster V (37 48) and cluster II (37 18) The lowest cluster mean was recorded for cluster VI (28 94) followed by cluster IV (36 32) and cluster III (36 35)

84 1% Test weight (g) The highest cluster mean was recorded by cluster VI (5 46), followed by cluster VII (4 83), cluster III (4 55) and cluster I (4 04) The minimum value was recorded by cluster II (3 37) followed by cluster IV (3.52) and cluster V (3 71) Seed yield per plant (g) This character showed variation, which ranged between 6 38 (cluster IV) and 7 35 (cluster I) The highest cluster mean shown by cluster I was followed by cluster III (7 17), cluster V (7 08) and cluster VII (7 02) Cluster IV (6 38) and cluster VI (6 59) recorded comparatively low values. 4.8 Per cent contribution of various characters for divergence All the 40 genotypes of green gram were studied for 13 characters and the data collected was used to determine divergence Out of 13 characters studied, the character days to maturity (34 61%) contributed the highest for divergence and was followed by test weight (25 25%), protein content (17 69%) and number of branches per plant (5 51%) However, the contribution of number of seeds per pod (0 89%), seed yield per plant (0 76%), number of pods per plant (1 66) and number of nodules per plant (1 66) was of low magnitude The character number of clusters per plant contributed zero per cent for divergence

85 63 Table 4.9: - Contribution of various characters to divergence Sr No Character Per cent contribution 1 Days to 50% flowering Days to maturity Plant height (cm) No of branches/plant Pod length(cm) No of pods /plant No of seeds/pod No of nodules/plant No of clusters/plant Protein content (%) Harvest index (%) Test weight (g) Seed yield/plant (g) Total 100

86

87 DISCUSSION

88 5. DISCUSSION Success of any plant breeding programme depends on selection of elite genotypes, which ultimately depends on knowledge of variability and genetic diversity of the germplasm Genotypic and phenotypic coefficients of variation measure the extent of variability (genotypic and phenotypic) present in a population for a particular character Hentabihty is an index of transmissibility of a character from parents to their offspring The concept of hentabihty is important in determining whether the phenotypic differences among the individuals are genetical or the result of environmental factors Genetic advance on the other hand measures the expected genetic gain from the selection applied in a population Hentabihty along with genetic advance gives the best picture of the efficiency of selection The correlation study provides the inter-relationships among the quantitative characters which facilitates the choice of suitable parents and the breeding method for the improvement in the crop, whereas, path analysis helps to know the direct and indirect effects of characters towards yield The genetic divergence enables the evaluation of parents without actual crossing and grouping the parental material into clusters in a significant pattern. The D 2 statistics suggested by Mahalanobis (1936) and clustering by Rao (1952) helps to select the diverse genotypes for hybridization programme.

89 The results obtained from the present investigation, "Variability, path analysis and genetic diversity in green gram" in 40 genotypes are discussed under the following sub-headings 5.1 Variability 5.2 Heritability and genetic advance 5.3 Correlations 5.4 Path analysis 5.5 Genetic divergence l>5 5.1 Variability In the present investigation, considerable amount of variability was observed for all the thirteen characters studied The variability observed for seed yield per plant ranged between 5 11g and 9 33g per plant with a mean of 7 03g Likewise, other yield related characters showed wide range of variability viz., days to 50 per cent flowering ( days), days to maturity ( days), plant height ( cm), number of branches per plant ( ), pod length ( cm), number of pods per plant ( ), number of seeds per pod ( ), number of nodules per plant ( ), number of clusters per plant ( ), protein content ( %), harvest index ( %), test weight ( g) Among the 40 genotypes studied PM 9342 (days to 50 per cent flowering), PM (days to maturity), PM 9341 (plant height), IC (number of branches per plant), PM 9340 (pod length),ic (number of pods per plant), PM 9361 (number of seeds per pod), PM 9339 (number of nodules per plant), IC (number of clusters

90 & per plant), PM 9339 (protein content), PM 9339 (harvest index), PM 9340 (test weight) and PM 9376 (seed yield per plant) recorded the highest per se performance for the respective characters The estimates of phenotypic coefficient of variation (PCV) were magnitudinally higher than the estimates of genotypic coefficient of variation (GCV) for all the characters studied indicating the influence of environmental factors on these traits The estimates of GCV and PCV were of high magnitude for number of pods per plant, number of branches per plant, indicating good scope for their improvement through selection These results confirmed the earlier findings of Ramana and Singh (1987), Natarajan et al (1988) and Mishra etal. (1995) The PCV estimates were high for number of pods per plant, number of clusters per plant, number of branches per plant and seed yield per plant Similar results were reported earlier by Natarajan et al (1988) The GCV estimates were high for test weight, number of pods per plant and number of branches per plant These results confirmed the earlier findings of Rathnaswamy et al (1978) for test weight and Veeraswamy et al. (1973a) for number of pods per plant and number of branches per plant The estimates of GCV for pod length, plant height and protein content were moderate These results confirmed the earlier findings of Natarajan et al (1988) for pod length and Veeraswamy et al. (1973a) for plant height The earlier findings of Veeraswamy et al (1973a), Paramasivan and Rajasekaran (1980), Lakshmaiah et al. (1989), Panwar et al. (1995) were similar to those observed in the present investigations for number of seeds

91 per pod, days to 50 per cent flowering and days to maturity, where genotypic and phenotypic coefficients of variation were of low magnitude The magnitude of GCV was lower than that of PCV for seed yield per plant, however, the findings of Singh and Malhotra (1970), Paramasivan and Rajasekaran (1980) and Ramana and Singh (1987) were contradictory to present findings Rathnaswamy et al (1978) and Natarajan et al (1988) reported low GCV and PCV estimates for seed yield The magnitudinal differences between genotypic and phenotypic coefficients of variation were maximum for number of clusters per plant, number of pods per plant and seed yield per plant, indicating that environment played a significant role in the expression of these characters Similar results were reported earlier by Natarajan et al. (1988) for number of clusters per plant and number of pods per plant and Kalpana et al. (1988) and lihammudin and Tajammal (1989) for seed yield per plant However, for the remaining characters, the magnitudial differences between genotypic and phenotypic coefficients of variation were minimum, indicating less role of environment on the genotypes in the phenotypic expression of these characters and one can rely on the phenotype alone while carrying out the selections w 5.2 Heritability and genetic advance High heritability estimate indicates the effectiveness of selection based on phenotypic performance, but does not necessarily mean a high genetic gain for that particular trait The characters with high genetic gain may be attributed to the additive gene effects (Panse, 1957), which can easily be improved by simple selection On the other hand, high heritability and low

92 i8 genetic advance may be attributed to non-additive gene action and such characters may be improved by hybridization Burton (1952) suggested that through genotypic coefficient of variation, the heritable variation cannot be estimated and as, genotypic coefficient of variation together with hentabihty would furnish most reliable information on the amount of genetic advance to be expected for selection. In the present investigation, the hentabihty estimates for test weight, days to maturity, days to 50 per cent flowering, protein content and pod length were of high magnitude, indicating major role of genotype and ultimately less environmental influence Similar results were reported by Emping era/ (1970) and Natarajan et al. (1988) for test weight, days to 50 per cent flowering and pod length Hentabihty was comparatively low for number of clusters per plant, harvest index, seed yield per plant and number of pods per plant, indicating the role of environment on the expression of these characters Similar results were reported by Natarajan et al (1988) for number of clusters per plant, Singh and Malhotra (1970), Veeraswamy et al. (1973a), Rathnaswamy et al (1978) and Imrie and Butler (1982) for seed yield per plant and number of pods per plant Johnson et al (1955) observed that the genetic gain will be low when there is non-additive gene action, whereas, the genetic advance would be high, when there is additive gene action In the present study high hentabihty for days to maturity and days to 50 per cent flowering was accompanied by high genetic advance indicating that the hentabihty is due to additive gene action and simple selection for such traits could be practiced for improving them Similar results were observed by Natarajan et al (1988) for days to 50

93 Cfl per cent flowering and Veeraswamy et al (1973a) for days to maturity In case of test weight, high hentabihty was accompained by low genetic advance, suggesting the presence of non-additive gene action for this character and it can be improved by hybridization Similar results were reported by Paramasivan and Rajasekaran (1980) and Natarajan et al (1988) The earlier findings to Singh and Malhotra (1970), Veeraswamy et al (1973a) and Rathnaswamy et al. (1978) were similar to the present results in case of seed yield per plant, where, low heritabihty was accompanied by low genetic advance, suggesting that it is not likely to respond favourably to selection However, the results of Ramana and Singh (1987) and Natarajan et al (1988) were contrary to the present findings In case of number of cluster per plant also, low hentabihty was accompanied by low genetic advance, suggesting that it is not likely to respond favourable to selection Similar results were reported by Natarajan et al. (1988) Thus, considering hentability and genetic advance together, it was evident that days to 50 per cent flowering and days to maturity were the characters where there exists additive gene effects and simple selection in the segregating generations would be effective for improving these characters, whereas, test weight and pod length, showed the presence of non-additive gene action implying that hybridization will be the best way to improve these characters 5.3 Correlations Knowledge about association between the yield and yield components facilitates the choice of suitable breeding method to be applied and selecting

94 ^0+ the parents for improving the crops The phenotypic and genotypic correlation have their own importance in breeding programme The phenotypic correlation coefficients helps in determining selection index, whereas genotypic correlations provide a close measure of association between characters and gives an indication of usefulness of characters in overall improvement of the crop They may also help to identify characters that have little or no importance in the selection programme In the present investigation, seed yield per plant was positively and significantly correlated with plant height, number of clusters per plant, number of seeds per pod, test weight, days to maturity and days to 50 per cent flowering at genotypic level Similar findings were reported by Joshi and Kabaria (1973), Malik era/ (1987), Natarajan et al. (1988), Patil and Deshmuk (1988), Singh et al (1988), Patil and Narkhede (1989), Khorgade et al (1990), Ebenezer Babu Rajan et al (2000) and Venkateswarlu (2001b) for plant height, Choi et al (1986), Malik et al. (1987), Natarajan et al (1988), Singh et al. (1988), Khorgade et al. (1990), Khan (1991), Warn et al (1992), and Ebenezer Babu Rajan et al (2000) for number of clusters per plant, Patil and Deshmukh (1988),Raut et al (1988), Singh et al (1988), Patil and Narkhede (1989),Ebenezer Babu Rajan et al (2000) and venkateswarlu (2001a) for number of seeds per pod, Raut et al. (1988), Singh et al (1988), Patil and Narkhede (1989), Ebenezer Babu Rajan et al (2000) and Venkateswarlu (2001b) for test weight, Patil and Deshmukh (1988), Patil and Narkhede (1989) Satyan et al. (1989) and Venkateswarlu (2001) for days to maturity, Pokle and Nomulwar (1978), Patil and Deshmukh (1988), Lakshmaiah et al. (1989) and Satyan et al (1989) for days to 50 per cent

95 ^ flowering The association between seed yield per plant and harvest index was significantly negative, however, the results of Ebenezer Babu Rajan et al. (2000) were contrary to the present findings Days to 50 per cent flowering was significantly and positively correlated with days to maturity, number of clusters per plant, number of seeds per pod, plant height and number of pods per plant These results were in accordance with the findings of Malhotra et al. (1974) and Upadhyaya et al. (1980), where as, Natarajan et al (1988) reported contradictory results Days to maturity showed positive and significant correlation with number of pods per plant, number of clusters per plant and plant height, its association with harvest index was significantly negative The character plant height showed positive and significant correlation with number of clusters per plant, test weight and pod length, whereas significant and negative correlation was found with number of branches per plant The earlier findings of Malhotra et al (1974), Upadhyaya et al (1980) and Ramana and Singh (1987) were similar to those observed in present findings Number of branches per plant showed negative and significant correlation with pod length, number of seeds per pod and plant height, where as significant and positive correlation was found with number of clusters per plant but Malhotra et al (1974) and Upadhyaya et al (1980) reported contradictory result for pod length and number of seeds per pod Number of pods per plant showed negative and significant correlation with number of seeds per pod, pod length and harvest index, whereas it showed positive and significant correlation with days to maturity and days to 50 per cent flowering

96 75- Malhotra et al. (1974), Upadhyaya et al (1980) and Natarajan ef al (1988) reported contradictory results Number of seeds per pod showed positive and significant correlation with test weight, pod length and days to 50 per cent flowering, whereas significant and negative correlation was found with protein content, number of pods per plant and number of branches per plant. The earlier findings of Malhotra et al, (1974) and Ramana and Singh (1987) were similar to those observed in the present findings Number of clusters per plant was significantly and positively correlated with protein content, number of branches per plant, number of nodules per plant, plant height, days to maturity and days to 50 per cent flowering, whereas its association with harvest index, pod length and test weight were significantly negative These results were in accordance with the findings of Malhotra et al. (1974) and Natarajan et al (1988) for test weight Test weight was significantly and positively correlated with number of seeds per pod, pod length and plant height, whereas it was significantly and negatively correlated with number of clusters per plant and harvest index Malhotra ef al (1974) and Upadhyaya ef al. (1980) reported similar results for number of clusters per plant, pod length and number seeds per pod While comparing the present results in light of earlier findings, it was observed that the yield contributing characters showing significant association with yield in desirable direction, viz, number of clusters per plant, plant height, test weight, number of seeds per pod, days to maturity and days to 50 per cent flowering could directly be useful to construct the selection index, which

97 / <3 will help in identifying high yielding and early maturity genotypes in green gram 5.4 Path analysis Path coefficient analysis is simply a standardized partial regression coefficient, which splits the correlation coefficient into direct and indirect effects In the present investigation path analysis was worked out by following Dewey and Lu (1959) to estimate the magnitude and direction of direct and indirect effects of various yield and yield contributing characters Correlation coefficients along with path effects provide more reliable information, which can be effectively used in various crop improvement programme If the correlation between a causal factor and direct effect is more or less of equal magnitude, it explains the true and perfect relationship between the traits and direct selection through these traits will be rewarding However, if the correlation coefficient is positive and the direct effect is negative or negligible, the indirect causal factors are to be considered in simultaneous selection (Singh and Kaker, 1977). Thus path analysis provide the information about characters and their relative importance The direct and indirect effects of yield components studied in 40 genotypes of green gram are presented in Table 4 5 and Figure 5 1 The high magnitudinal direct effect of plant height, days to 50 per cent flowering and test weight along with highly significant correlation in the desirable direction towards seed yield per plant indicated the true and perfect relationship between seed yield and these characters suggesting direct selection based on these characters would help in selecting the high yielding

98 Fig 5 1 Path diagram showing nature of causal system variables with their coefficient for path analysis in green gram {Vigna radiata (L)Wilczek) X r Days to 50% flowering X 5 - Pod length(cm) X 9 - No of clusters/plant X 2 - Days to maturity X 6 - No of pods /plant X10- Protein content (%) X 3 - Plant height (cm) X r - No of seeds /pod Xn- Harvest index X4- No of branches/plant X 8 - No of nodules/plant X 12 -Test weight R- Residual effect Y- Seed yield/plant

99 '1 ff genotypes in green gram These results were in agreement with the earlier findings of Khorgade et al (1990) and Mishra and Yadav (1992) for plant height, Patil and Deshmukh (1988) for days to 50 per cent flowering, Patil and Narkhede (1989), Khorgade et al. (1990) and Ebenezer Babu Rajan et al (2000) for test weight However, the results of Sandhu et al. (1980) and Thandapani and Sakharam Rao (1984) were contradictory to the present results Protein content showed high magnitudinal direct effects, and a nonsignificant negative association with seed yield per plant suggesting that under these circumstances a restricted simultaneous model is to followed However, the results of Malhotra et al. (1974) and SelYldhu et al (1980) were contradictory to the present findings The direct effect of number of seeds per pod, number of clusters per plant, and days to maturity was negative, though they were significantly and positively associated with seed yield per plant indicating that they played their role via indirect effects Number of seeds per pod played its role via test weight and protein content, number of clusters per plant via nodules per plant and pod length, days to maturity via days to 50 per cent flowering and plant height These results were in agreement with the earlier findings of Thandapani and Sakharam Rao (1984) for number of seeds per pod, Sandhu et al (1980) and Ebenezer Babu Rajan et al. (2000) for number of clusters per plant and Malhotra et al. (1974) for days to maturity However, the results of '&cr>ahu- et al. (198 ) and Natarajan et al. (1988) were contradictory to the present findings Thus it becomes clear that direct selection based on

100 33"" plant height, days to 50 per cent flowering and test weight can be made for the improvement of this crop 5.5 Divergence Selection of elite genotypes with high perse performance for yield and yield contributing characters with genetic divergence is important for starting any hydndization programme It would be possible to identify desirable genotypes from the estimates of genetic variability but it is difficult to expect any extraordinary results from their progeny unless we have sound knowledge about divergence between them Mahalanobis (1936) developed the concept of D 2 statistics which act as important tool for plant breeder Application of this technique for the assessment of genetic diversity in plant breeding was first suggested by Rao (1952) The degree of divergence between biological population at genotypic level and the relative contribution of different components to the total divergence at both intra and inter cluster levels can be evaluated by this method Clusters Intra and inter cluster distances and mean performances In the present investigation, D 2 values between all possible pairs of 40 genotypes ranged between (PM 9348 and PM 9380) to (PM 9340 and IC 00114). This high range of D 2 values showed the presence of good amount of diversity in the material used for the current study The 40 genotypes studied were grouped into seven clusters by using Tocher's method as described by Rao (1952) Cluster I with 14 genotypes emerged as the largest cluster followed by cluster II with 9 genotypes

101 ^r diagram showing interrelationships between seven clusters

102 Likewise, cluster III contained 8 genotypes, cluster IV contained 4 and cluster V contained 3 genotypes The clusters VI and VII were monogenotypic consisting of PM 9340 and PM 9383, respectively. These two genotypes included in the above clusters had wide variation from the rest as well as from each other These genotypes may have different genetic architecture from the others In the present study grouping of genotypes into seven clusters, itself indicated wide diversity among the genotypes Ramana and Singh (1987) grouped 39 genotypes in to 8 clusters. Similarly Manivannan et al. (1998) grouped 30 genotypes in to 8 clusters However, Venkatkrishna Kishore et al (2000) grouped 40 genotypes in to 15 clusters, with cluster VII the largest consisting of eight genotypes Loganathan et al. (2001a) grouped 42 F 3 and eight varietal genotypes into seven clusters. The maximum intra-cluster distance was observed in the cluster IV (D =7 56), followed by cluster III (D =7 20), cluster II (D =7 02) and cluster I (D =7 0) suggesting that genotypes included in these clusters might have different genetical architecture (Table 4 7) However, the lowest intra-cluster distance was observed in cluster V indicating that the strains of these cluster resemble one another genetically and appeared to have evolved from the common gene pool Maximum inter-cluster distance was observed between the cluster V and VI (D =19 41) and was followed by cluster II and VII (D =19.04), cluster II and VI (D=18 79) indicating wide divergence among the clusters This also suggests that the genetic architecture of the genotypes in one cluster differs entirely from those included in the other cluster /s

103 The minimum mter-cluster distance was observed between cluster III and VI (D =9 30) and was followed by cluster I and II (D =10 52) and cluster I and III (D =10 54) The lower D values between these clusters suggested that the genetic constitution of these genotypes in one cluster were in close -f proximity with the genotypes in other cluster of the pair The two monogenotypic clusters showed zero intra-cluster distance Similar results were reported earlier by Shanmugam and Sreerangaswamy (1982) who reported 3 monogenotypic clusters, whereas, Natarajan et al. (1988) reported only one monogenotypic cluster and Naidu and Satyanarayana (1991) reported 7 monogenotypic clusters, suggesting parallelism with the present study Based on mean performance of the clusters of the thirteen characters (Table 4 8), it was observed that the cluster I with maximum number of genotypes (14), recorded the highest seed yield per plant (7.35 g) and was characterized by the highest number of seeds per pod, pods per plant, more number of nodules per plant and high clusters per plant All these characters appeared to have played important role in determining the yield of this cluster Cluster III recorded high seed yield per plant showed late maturity, high number of pods per plant, high pod length and more seeds per pod Cluster VII, V, and II showed medium seed yield per plant as most of the yield components were intermediary in nature Cluster IV, which consisted of four genotypes, showed the lowest seed yield per plant among the clusters which was mainly due to early flowering and maturity, shorter pods, less seeds per pod and low test weight

104 t Relative contribution of different characters towards divergence The variance of cluster means provide information on relative importance of different characters towards yield In the present study, the variance of cluster means were high for days to maturity (32 02), plant height (21 625), harvest index (10 65) and days to 50 per cent flowering (8 67) indicating that these characters may be responsible for genetic divergence in the material under study. Similar results were observed by Ghaden ef al (1979), Thulasidas (1984) and Misra (1986) for days to maturity, Naresh Chandra et al. (1993) for plant height and Ramana and Singh (1987) and Natarajan et al. (1988) for days to 50 per cent flowering Pod length, test weight, number of clusters per plant, number of seeds per pod, number of branches per plant and seed yield per plant contributed the least to genetic divergence In the earlier findings, contradictory results were reported by Thulasidass (1984) for pod length, Natarajan et al. (1988), Deb et al (1994) and Mishra et al. (1995) for hundred seed weight and Singh (1988) for seed yield per plant Considering the present results and work done by earlier workers, it can be concluded that yield indicators responsible for divergence varied substantially This may be attributed to use of different sets of material and also due to role of environmental variability as suggested by Bains and Sood (1984) who also pointed out that a positive contribution of genetic divergence of yield components, like number of pods per plant and plant height, may be help in selecting genotype for yield and other economic traits

105 5.5.3 Genetic divergence as a measure of choosing potent parents for crossing Diversity is the basic need of crop improvement programmes The success of any crossing programme depends on selection of parents having high expression for the economically important characters. Among the different approaches of selecting parents, selection based on diversity has its own merit. Therefore, in the present investigation diversity among different genotypes was studied, which yielded valuable information that could be useful in suggesting potent parents for crossing Hays and Jahanson (1939) and East and Hays (1942) obtained greater heterosis from crosses between diverse parents than those between close related ones Timothy (1963) found that genetic divergence is one of the criteria for selecting the parents for hybridization, which may produce transgressive segregants in the later generations Bhatt (1970) advocated the use of multivariate analysis for the selection of parents. He also stated that statistical distance of all possible cluster combination may be considered arbitanly as a guideline and suggested that it would be logical to effect crosses between genotypes belonging to the clusters separated by high estimated statistical distance Arunachalam and Bandopadhyay (1984) suggested a method, which assert parent divergence into four divergence classes (DC) viz, DC1, DC2, DC3 and DC4 To take into account the magnitude of variation in parental divergence, the mean (M) and standard deviation (S) of the intra and inter cluster divergence (D) were calculated They conducted two experiments in =^f groundnut and one in rapeseed and concluded that the chance for the

106 Sv occurance of a high frequency of heterotic crosses with high heterosis values were more when the parents were chosen, whose divergence lies between interval M-S and M+S (DC2 and DC3) as compared to the crosses between parents whose divergence fall outside of this interval In the present investigation attempts were made to classify the cluster combinations into four divergent classes as per the method suggested by Arunachalam and Bandopadhyay (1984) The mean (M) calculated for inter cluster and intra cluster distances was with a standard deviation (S) of The minimum (X) and maximum (Y) divergence class values among these distances were and , respectively The divergence classes are presented in Table 5 1 Arunachalam and Bandopadhyay (1984) reported that crosses between divergent classes DC2 and DC3 will be more heterotic and promising than other combinations However, in the present investigation all seven clusters were included in DC2 and DC3 To reduce the risk from the heterosis point of view, high yielding genotypes from the selected clusters should be used However, while choosing genotypes from clusters, other practical considerations like yield, quality, resistance to disease should be taken into account The list of genotypes, which deserve to be considered as potent for crossing is as indicated below, Sr. No. Name of genotypes Clusters 1 KOP Cluster I 2 PM 9376 Cluster I 3 BPMR-132 Cluster V 4 J-781 Cluster I 5 AKM 9801 Cluster II 6 PM 9342 Cluster III 7 PM 9380 Cluster I 8 VC 3960 Cluster II 9 PM 9383 Cluster VII

107 81 Table 5.1 Distribution of different cluster combinations into four divergence classes based on D values between them. DC 4 DC 3 DC 2 DCi X M-S M M+S D I V E R G E N C E C L A S S E S (I,II),(I,IH),(I,IV),(I,V), (I,VII), (IV,II) (II,III),(III,VI),(IV,VII) (II,IV),(II,V),(III,V),(IV,V),(I,VI),(IV,VI), (III,VII),(V,VII),(VI,VII) (III,IV),(H,VI),(II,VII)

108

109 SUMMARY AND CONCLUSION

110 6. SUMMARY AND CONCLUSIONS The present investigation, "Variability, path analysis and genetic diversity studies in green gram (Vigna radiata (L.) Wilczek)" was undertaken with following objectives 1 To know the extent of variability for yield and yield contributing characters 2 To study the association between different characters and to find out direct and indirect effect of important characters on yield 3 To measure genetic divergence between different genotypes and group them into suitable clusters Forty genotypes of green gram collected from Pulses Improvement Project, M P K V Rahun and Agricultural Research Station Badnapur, Maharashtra The experiment was laid out in a randomized block design with three replications during khanf, 2002 The 40 genotypes were evaluated for 13 yield and yield contributing characters viz, days to 50 per cent flowering, days to maturity, plant height (cm), number of branches per plant, pod length (cm), number of pods per plant, number of seeds per pod, number of nodules per plant, number of cluster per plant, protein content (%), harvest index (%), test weight (g) and seed yield per plant (g) Significant treatment mean squares for all characters studied revealed the presence of considerable amount of variability in genotypes evaluated The magnitude of GCV and PCV were high for number of pods per plant, number of branches per plant and test weight, indicating the presence of good amount of variability for these characters However, moderate amount of GCV and PCV

111 S2, were observed for plant height, Pod length and number of nodules per plant The high magnitudinal differences between GCV and PCV were found for number of clusters per plant, number of pods per plant, seed yield per plant and number of branches per plant indicating the role of environment in phenotypic expression of these traits High magnitude of hentability (bs) was recorded for test weight (96%), followed by days to maturity (95 7%), protein content (90%), days to 50 per cent flowering (87 5%), pod length (59 4%) and number of nodules per plant (53 3%) High hentability coupled with high genetic advance was observed for days to maturity and days to 50 per cent flowering suggesting the role of additive gene effect and possibilities of achieving high genetic progress through selection The genotypic correlation coefficients were higher in magnitude than their corresponding phenotypic correlation coefficients for all the characters except number of pods per plant Seed yield per plant was significantly and positively correlated with plant height and number of clusters per plant both at genotypic and phenotypic levels, where as it was significantly and positively associated with number of seeds per pod, test weight, days to maturity and days to 50 per cent flowering at genotypic level and with number of pods per plant at phenotypic level indicating dependence of seed yield on these characters Seed yield per plant was also significantly and negatively associated with harvest index In path coefficient analysis, plant height, days to 50 per cent flowering and test weight recorded the highest direct effect in the desirable direction Their association with seed yield was significant and positive indicating the fact

112 S that there exists a true and perfect association between these characters This also suggested that direct selection for these characters will help in isolating early and high yielding genotypes The direct effect of number of seeds per pod, number of clusters per plant and day to maturity was negative but their association with seed yield per plant was positive, indicating that they played their role via indirect effects Correlation and path analysis revealed that plant height, test weight and days to 50 per cent flowering were good indicators of grain yield per plant in green gram and can be used for making direct selections for yield In the present investigation, the D 2 values between all possible pairs of 40 genotypes ranged between and The genotypes were grouped into seven clusters, following Tocher's method as described by Rao (1952). Cluster I had maximum number of genotypes (14), followed by clulster II with 9 genotypes Likewise, cluster III contained 8 genotypes, cluster IV contained 4 and cluster V contained 3 genotypes The maximum intra-cluster distance was observed for the clulster IV (D= 7 56) followed by cluster III (D = 7 20) and cluster II (D= 7 02), suggesting that genotypes present in these clusters possess varied genetic architecture and might have originated from different genetic pool The cluster VI and VII were monogenotypic The maximum inter cluster distance was observed between cluster V and VI (D= 19 41) followed by cluster II and VII (D= 19.04) and cluster II and VI (D= 18 79) indicating wide divergence among these clusters, which also suggested that the genetic architecture of the genotypes in one cluster differs entirely from those included in other clusters The minimum inter cluster distance was found between

113 ?r cluster III and VI (D= 9 3) The variance of cluster means revealed that days to maturity, plant height, harvest index and days to 50 per cent flowering were the main characters contributing to the genetic divergence in the present study The positive contribution of these yield components in genetic divergence may be of considerable help in selecting genotypes for yield and other economic traits The cluster combinations were classified into four divergence classes, following the method suggested by Arunachalam and Bandopadhyay (1984) Crosses were suggested between clusters in a pair from inter-cluster D values, which fall in divergence classes DC2 and DC 3,1 e having intermediate inter cluster distance between them, which would give higher chances of producing high frequency and magnitude of heterosis in future when they will be crossed. The following potent parents are suggested for tentative breeding programme, based on divergence studies 1) KOP 7)PM9380 2) PM9376 8)VC3960 3) BPMR132 9) PM ) J-781 5) AKM9801 6) PM9342

114

115 LITERATURE CITED

116 7. LITERATURE CITED Ah, M S and M A Q. Shaikh 1987 Variability and correlation studies in summer mungbean (Vigna radiata) Bangaladesh J agric 12(2) Allard, R W 1961 Relationship between genetic diversity and consistency of performance in different environments Crop Sci Anonymous 1970 Pulse crops of India. ICAR Pub, New Delhi, pp 148 Anonymous Directorate of Agriculture, M S, Pune Arunachalam, V and A Bandopadhyay 1984 Limits to genetic divergence for occurance of heterosis-expenmental evidence from crop plants Indian J. Genet 44(3) Bain, AS and K C Sood 1984 Resolution of genetic divergence for choice of parents in soybean breeding. Crop Improv. 11 (2) *Bhatt, G M 1970 Multivariate analysis approach to selection of parents for hybridization aiming at yield improvement in self pollinated crops Aust J agric Res Boomikumaran, P and M Rathinam Correlation and path analysis in green gram Vigna radiata (L) Wilczek. Madras agric J 68 (10) *Bose, R D 1939 Studies in Indian pulses IX Contributions to the genetics of mung (Phaseolus radiatus Linn., Syn Ph Aureus Roxb.) Indian J agric Sci *Burton, G W 1952 Quantitative inhentence in grasses Proceedings of the sixth International Grass land congress Vol

117 ?7 Burton, G W and E H Devene 1953 Estimating hentability in Jali Fesche (Festuca arundinaces) from replicated clonal material Agron J Byregowda, M, J Chandra Prakash, C S J Babu and P Rudraswamy 1997 Genetic variability and interrelationships among yield and yield components in green gram (Vigna radiata) Crop Res Hissar 13 (2) Chandel, K P S, B S Joshi and K C Pant 1973 Yield in mungbean and its components. Indian J. Genet 33(2) Charles, D R. and H H. Smith 1939 Distinguishing between two types at gene action in quantitative inheritance. Genetics 24 (3)* Choi, K J, H G Choi, H K Lim and D K Lee 1986 Studies on mungbean varieties 1 Variation in agronomic characteristics of mungbean varieties Crops 28 (1): Chowdhury, J B, R K Chowdhury and S N Kukar 1971 Studies on genetic variability in mung (Phaseolus aureus) J Res Punjab agncuniv 8(2) Das, S Y, Supnyo Chakraborty and S Chakraborty 1998 Genetic variation for seed yield and its components in green gram (vigna radiata (L.) Wilczek) Adv PI Sci 11 (1) Deb, S K and P K Bhaunik 1994 Application of discriminant function technique for selection in mung {Phaseolus aureus Roxb.) Annl Biol 10 (2) De Candolle, A Origin of cultivated plants Hofftner, New York pp 346

118 g<? Deore, A J Genetic variability, character association and path analysis studies of yield and its components in mungbean M Sc (Agric) thesis, Panjabrao Krishi Vidyapeeth, Akola Dewey, D R and K H Lu 1959 A correlation and path coefficient analysis of components of crested wheat grass seed production Agron J East, E M 1916 Studies on size inheritance in Nicotiana Genetics East, E M and H K Hays 1942 Heterozygosis in evolution and plant breeding, U.S D A Bureau of Plant Industry Bulletin Ebenezer Babu Rajan, R, D Wilson and Vijayaraghava Kumar 2000 Correlation and path analysis in the F2 generation of green gram (Vigra radiata (L.) Wilczek) Madras agric J, 87 (10-12) Emping, L T, R M Lantican and P B Escuro 1970 Heritability estimates of quantitative characters in mungbean (Phaseolus aureus Roxb) Crop Sci Falconer, D. S Introduction to quantitative genetics 3 rd Edn Longman, New York, USA Fisher, R. A 1930 The genetic theory of natural selection "Genetics" by Stnckberger, M W Published by Mc Milllan Co New York pp 772 Francisco, P B and K Maeda Agro-physiological studies on the yield performance of mungbean 1 Cultivar differences in earliness, flowering and their relationships with growth and seed yield Japanese J Crop Sci 58(4) Frankel, O H 1947 Plant collections J Aust Inst Agril Sci

119 m Ghaderi, A, M Shishegar, A Rezai and B Ehdaie 1979 Multivariate analysis of genetic diversity for yield and its components in mungbean J. Amer. Soc Hort Sci 104(6) Ginraj, K 1973 Natural variability in green gram (Phaseolus aureus Roxb.) Mysore J agnc Sci 7(2) Gupta, M P and R B Singh Variability and correlation studies in green gram Indian J agnc Sci Gupta, M P and R B Singh Genetic divergence for yield and its components in green gram Indian J Genet 30(1) Gupta, S N, S. Lai and L Rai Correlation and path analylsis in mungbean (Vigna radiata (L.) Wilczek) Haryana agnc Univ J Res 12 (2) Hays, H K. and I J Johanson 1939 The breeding of selfed improved lines of corn J. Amer. Soc. Agron Holkar, S and N O Raut 1992 Character association and path analysis in green gram J Maharashtra agnc Univ 17 (2) *hhamuddin and M A Tajammal Genetic and phenotypic variability in yield and other quantitative characters in mungbean (Vigna radiata (L.) Wilczek) Sarhad J agnc 5 (1) *lmne, B C and K C Butler 1982 An analysis of variability and genotype x environment interaction in mungbean Vigna radiata in South eastern Queensland Aust J. agnc. Res 33 (3) Johanssen, W L 1909 Elements der exlenten exblich Keitslenve Jena. Gustav. Fisher. Pertanika J Tropical agnc Sci 18(1) 63-69

120 cjo Johnson, H W, H F Robinson and R. E Comstock 1955 Genotypic and phenotypic correlation in soybean and their implication in selection Agron. J Joshi, S N and M. M Kabana 1973 Inter-relationship between yield and yield components in Phaseous aureus Roxb. Madras agnc. J. 60 (9/12) Kalpana, P, S S. Raghuvanshi and S P Singh 1988 Selection parameters for different traits in mungbean (Vigna radiata (L.) Wilczek) Narendra DevaJ agnc Res 3(1) *Khan, I A Correlation and path analysis of yield components in mungbean {Phaseolus aureus Roxb.) Bot Bull Acad Sinica 26 (1) Khan, I A 1988 Path coefficient analysis of yield attributes in mungbean Legume Res 11(1) Khan, I A 1991 Multiple correlation and regression analysis in mungbean (Vigna radiata (L.) Wilczek) Indian J agnc Res. 25 (4) Khan, I A and R. Ahmed Relationship of yield and its component characters in mungbean Legume Res 12 (4) Khorgade, P W, A H Nafade, M N Narkhede and S K Raut Some selection criteria in green gram. J. Maharashtra agnc Univ. 15 (2) Kumani, P K T and M K George 1982 Correlation and path analysis in green gram Agnc Res. J Kerala. 20 (2)

121 c i\ Lakshmaiah, K, P Ramesh Babu and D Lokandha Reddy 1989 Genetic parameters and correlations in green gram (Vigna radiata (L.) Wilczek) Andhra Pradesh agric Univ J 17 (4) Tampong, A N, S Pichitporn, S. Sinsingh and N Vanakijmongkol 1988 Mungbean growth pattern in relation to yield In mungbean proceedings of the second international symposium held at Bangkok, Thiland, *Liu, D J, H F Guo and K. P Zheng 1984 Studies on local cultivars of Vigna radiata in Fujian. I. Yield correlations and regression analysis J Fujian agric. College 13 (2): Loganathan, P ; K. Saravanan and J Ganesan. 2001a. Genetic divergence in green gram (vigna radiata (L.) Wilczek) Res Crops. 2 (3) Loganathan, P., K. Saravanan and J Ganeasan 2001b. Genetic variability in green gram (vigna radiata (L.) Wilczek) Res Crops. 2 (3) Mahalanobis, P. C 1928 A statistical study at Chinese head measurement J Asiatic Soc. Bengal 25' *Mahalanobis, P C On the generalized distance in statistics Proc Nat Ins Sci India Malhotra, V. V., S. Singh and K. B. Singh Yield components in green gram. {Phaseolus aureus Roxb.) Indian J agric Sci. 44 (3) Malik, B A; M Zubair, A H Choudhari, M Tahir and I A Khan 1987 Genetic variability, character correlation and path analysis of yield components in mungbean (Vigna radiata (L.) Wilczek). Pakistan J Bot. 19(1) 89-87

122 «?2 Malik, B A; M Tahir and M Zubair 1985 Coheritability among yield and yield components in green gram Pakistan J agric Res 6 (3) Malik, B P S ; V P Singh, B. D Chowdhary and P K Chowdhary 1982 Path coefficients and selection indices in green gram Indian J agric Sci 52(5) Manivannan, N, E Murugan, P. L Viswanathan and C V Dhanakodi 1998 Genetic divergence in green gram Legume Res 21(2) Manivannan, N Genetic variability for seed yield and its components of green gram (Vigna radiata (L.) Wilczek) Agric Sci Digest, Karnal 19 (2) Manivannan, N 2000 Variability studies in F3 populations of green gram (Vigna radiata (L.) Wilczek) Agric Sci Digest, Karnal 20 (2) Meshram, L D 1978 Genetic variability and correlation coefficients related to yield and other quantitative characters and the use of path coefficients in mung Third Int Sci Congr Adv. Breed Res Asia Oceania, Canberra, Australlia pp Mishra, A K,L N YadavandY M Indrapurkar 1995 Divergence studies in genotypes of mungbean and urdbean Indian J Pulses Res 8 (2) Mishra, R K and R K Yadav Path analysis of component factors influencing economical yield, harvest index and biological yield of mungbean Adv PI Sci 5 (2)

123 Mishra, R C 1986 Genetic divergence in green gram (Vigna radiata (L.) Wilczek) Gujarat agric Univ Res J 12(1) Moll, R H, W S Sathuana and H F Robinson 1962 Heterosis and genetic diversity in variety crosses of maize Crop Sci Murthy, B R and V Arunachalam 1966 The nature of divergence in relation to breeding system in some crop plants Indian J Genet 26 (1) ^s Naidu, N. V 1990 Genetic analysis of yield and yield components of mungbean (Vigna radiata (L) Wilczek) Ph D thesis submitted to Dept of Genetics and plant breeding college of Agriculture, A N G R A U, Rajendranagar, Hydrabad Naidu, N V and A Satyanarayana 1991 Studies on genetic divergence over environments in mungbean (vigna radiata (L.) Wilczek) Indian J Genet 51 (4) Naresh Chandra, R B Mehra and S N Chaturvedi Genetic diversity in mungbean Indian J Pulses Res 6(2) Narsimham, M A note on the pollination of black and green grams in the Godawan Dist Agric J India Natarajan, C; K Thiyagarajan and R Rathnaswamy 1988 Association and genetic diversity studies in green gram (Vigna radiata (L.) Wilczek) Madras agric J 75 (7-8) Natrajan, M. and S. Palaniswamy 1990 Genetic divergence and realized heterosis based on dry matter components in mungbean (Vigna radiata (L.) Wilczek) Indian J Genet. 50 (3) Milsson - Ehle, M 1908 Kreusing untersunchungaman Hater and Weizen Jands University Asseter N F Afal 2Bd N Sr

124 Panse, V G 1957 Genetics of quantitative characters in relation to plant breeding Indian J Genet Panse, V G. and P V Sukhatme 1985 Statistical methods for agricultural worker, ICAR, New Delhi, 4 th Edition Panwar, K S ; R K Kapila and R K Mittal 1995 Variability and character association in mungbean Indian J Pulses Res 8(1) Parmsivan, J and S Rajasekaran 1980 Genetic variability in green gram (Vigna radiata (L.) Wilczek) Madras agnc J. 67 (7) Patel, H. S 1991 Genetic variability, correlation and path coefficient analysis in green gram Thesis abstract 17(3) fy Patil, H. S and R B Deshmuk 1988 Correlation and path coefficient analylsis in mungbean J Maharashtra agnc Univ 13 (2): Patil, H S and B N Narkede Association and path analysis of yield attributes in mungbean J Maharashtra agnc Univ. 14 (2) Pokle, Y S and M T Nomulwar Correlation and discriminant function in mung College agnc Nagpur, Magazine Power, S L, L F Locke and J C Gareft 1950 Partitioning methods of genetic analysis applied to quantitative characters of tomato crosses U S A P Tech Bull Ramana, MV and D P. Singh 1987 Genetic parameters and character associations in green gram Indian J agnc Sci 57(9) Rao, C R The utilisation of multiple measurement in problems of biological classification J Roy. Stat. Soc 10 (2) *Rao, C. R 1952 Advanced statistical methods in Biometrical Research John Wiley and Sons, Inc, New York

125 Rathnaswamy, R, S Krisnaswamy, S lyermperumal and P V Marappan 1978 Estimates of variability, correlation coefficients and path analysis in early maturing green gram Madras agric J 65 (3) Raut, N D, Mridula Bargale and S D Billmore 1991 Divergence analysis and heterosis in mungbean Crop Improv 18(2) Raut, S K, M S Choudhari and P W Khorgade 1988 Character association and path analysis in green gram Ann PI Physiol 2 (1) *Roxburgh, W 1832 Flora Indica. Sepampore, pp 325 ir Sandhu, T S, B. S Bhullar and H S Cheema 1980 Path coefficient analysis for grain yield and its attribute in green gram Indian J agric Sci 10(3) Satyan, B A, K C N Amaranth and I Siddaraju 1989 Phenotypic and genotypic correlation coefficients of some quantitative characters in green gram Curr Res 18(2) Satyan, B A, A R G Ranganatha and N Aswathappa 1980 Interrelation between grain yield and its components in green gram Curr Res Satyan, B A, K S Prakash and A R G Ranganatha 1986 Yield structure analysis in mungbean Indian J Genet 46 (3) Shamsuzzaman, K M, M. R. H. Khan and M A Q. Shaikh Genetic variability and character association in mungbean (Vigna radiata (L.) Wilczek) Bangladesh J agric Res 8 (1) 1-5

126 ?4 Shanmugam, A S and S R Sreerangaswamy 1982 Genetic diversity for quantitative characters in green gram (vigna radiata (L.) Wilczek) Madras agnc J 69(10) Sharma, V G S and S C Gupta Path analysis in green gram Indian J agnc. Sci 54(2) Singh, I S,B D Singh and K K Singh 1988 Inter-relationships of yield and its components in F 3 progenies of cross in mungbean Crop Improv 15(2) Singh, K B and R. S. Malhotra Estimates of genetic and environmental variability in mung (Phaseolus aureus Roxb.) Madras agnc J Singh, R. K. and B. D. Chaudhary Variance and covanance analysis in "Biometrical methods in quantitative genetic analysis" Kalyani Public, New Delhi pp *Singh, R K and S N Kaker 1977 Control of individual trait means during index selection. Proc Third Cong SABRAO (Canberra), Singh, R P. and M M Pathak Multivariate analysis of divergence in mungbean (Vigna radiata (L.) Wilczek) Farm Sci J 2 (2): Singh, S P Clustering of genotypes for selection for heterosis in yield and response to environmental variations in mungbean (Vigna radiata (L.) Wilczek). a proposed method Genome 30 (6) Singh, V P, B P S Malik, R K Yadav and J B Chowdhary 1980 Component analysis of yield in mungbean. Crop Improv 7 (2)

127 97- Sudheerkumar, S, A Sudharshanam, S Vinod Kumar and V Narsimha Reddy 1992 Variability studies in green gram (Vigna radiata (L.) Wilczek) J Res Andhara Pradesh agric Univ 20(3) *Sulaiman, M S 1976 Studies on variability, correlation and path analysis in green gram M Sc (Agri) Thesis. Tamilnadu agnc Univ Coimbatore Supnyo Chakraborty, H K Borah and S Chakraborty 2001 Genetic variability and correlation among root characters in green gram (Vigna radiata (L.) Wilczek) Res Crops 2(2) Tawar, M L, A K Mishra, S K Rao and S M Sharma 1988 Genetic divergence in mungbean Lengume Res 11 (3) Thandapani, V and J S Rao 1984 Yield parameters and their significance in green gram (vigna radiata (L.) Wilczek) genotypes in relation to yield. Madras agnc. J. 71 (4) Thulasidas, G 1984 Multiple regression and classificatory analysis in Vigna radiata (L.) Fid Crop Res 9(3/4) Timothy, D H Genetic diversity, heterosis and the use of exotic stocks in maize in Columbia Symp Statist Genet PI Breed Raleigh, N Carolina, pp Tiwan, V K, Y Mishra, S R Ramgirj and G S Rawat1996 Genetic variability in parents and segregating generation of mungbean (Vigna radiata (L.) Wilczek) Adv PI Sci 9 (2) Upadhyaya, L. P, R B Singh and R K Agrawal Character association in green gram populations of different maturing groups Indian J agnc. Sci. 50 (6)

128 *Vavilov, N.I 1926 Studies on the origin of cultivated plants Trudy Byul Pnkl Bot Veeraswamy, R, R Rathnaswamy and G A Palaniswamy 1973a Genetic variability in some quantitative characters of Phaseolus aureus Roxb Madras agnc J. 60 (9/12) Veeraswamy, R, G A Palaniswamy and R Rathnaswamy 1973b Genetic variability in some quantitative characters of Vigna sinensis (L.) Madras agric J Venkatakrishna Kishore, P A. Navale and S. D Gandhi 2000 Genetic divergence for quantitative characters in green gram Crop Res Hissar. 19(3) Venkateswarlu, O. 2001a Genetic variability in green gram (Vigna radiata (L.) Wilczek) Legume Res 24 (1) Venkateswarlu, O. 2001b. Correlation and path analysis in green gram Legume Res 24(2) Verdcourt, B 1970 Studies in the leguminosae papilionoidae for flora of tropical East Africa Kew Bull Wani, S A, G H Zargar and H U Ahanger Path coefficient analysis in mungbean (Vigna radiata (L.) Wilczek) Adv PI Sci 5 (2) *Wilczek, R 1954 Vigna In Flore de congo beige, Yadav, H S and S. K Rao Genetics of yield and its components in mungbean (Vigna radiata (L.) Wilczek) Egyptian J Genet, and Cyto 15(1) ?g

129 Yohe, J M and J M Poehlman 1975 Regression, correlation and combining ability in mungbean Trop agnc *Zukovsklji, P. M 1962 Cultivated plants and their wild relatives Common wealth agnc. Bureau, London. 91 * Original articles not seen

130

131 VITA