Volume 44 No. 1 June 2012 - sabrao

SABRAO JOURNAL of BREEDING and GENETICS ISSN 1029-7073

VOL. 44 NO. 1 JUNE 2012 CONTENTS SABRAO Journal, Regional Secretaries and Editorial Board.......i

tolerance under laboratory conditions using PEG......................................28

Special announcements……………vii

Peyman S. Diallel analysis to study genetic parameters of rice salinity tolerance traits at germination stage.............................................42

Research Papers Krualee S, Sdoodee S, Eksomtramage T, and Sereeprasert V. Stability of fresh fruit bunch of oil palm cross (Elaeis guineensis Jacq.) in southern Thailand.........................................1 Jatoi WA, Baloch MJ, Khan NU, Kumbhar MB, and Keerio MI. Genetic analysis of physiological and yield traits under drought stress conditions in wheat.............................................9 Geetha A, Sivasankar A, Prayaga L, Suresh J, and Saidaiah P. Screening of sunflower genotypes for drought

Soomro ZA, Khan NU, Kumbhar MB, Khuhro MA, Ghaloo SH, Baloch TA and Mastungi MI. Deterioration of F2 heterosis in F3 generation in diallel cross of upland cotton.....................................58 Kajonphol T, Sangsiri C, Somta P, Toojinda T, and Srinives P. SSR map construction and quantitative trait loci (QTL) identification of major agronomic traits in mungbean (Vigna radiata L.) Wilczek)........................71 Rungnoi O, Suwanprasert J, Somta P, and Srinives P. Molecular genetic

diversity of bambara groundnut (Vigna subterranea L. Verdc.) revealed by RAPD and ISSR marker analysis......................................... 87

and Patanothai A. Relationship between chlorophyll density and SPAD chlorophyll meter reading for Jerusalem artichoke (Helianthus tuberosus L.)..149

Rasyad A, Gulat M.E. Manurung, and Van Sanford DA. Genotype x environment interaction and stability of yield components among rice genotypes in Riau province, Indonesia.....................................102

Yasmin F, Islam MR, Rehana S, Mazumder RR, Anisuzzaman M, Khatun H, Rayhan R, and Gregorio GB. Molecular characterization of inbred and hybrid rice genotypes of Bangladesh...................................163

Méndez-Natera JR, Rondón A, Hernández J, and Merazo-Pinto JF. Genetic studies in upland cotton. III. Genetic parameters, correlation and path analysis.......................................112

Erratum.......................................176

Bhadru D, Rao VT, Mohan YC, and Bharathi D. Genetic variability and diversity studies in yield and its component traits in rice (Oryza sativa L.)..............................................129 Choudhary AK, Iquebal MA, Nadarajan N. Protogyny is an attractive option over emasculation for hybridization in pigeonpea....................................138

Uzun A, Seday U, and Turkay C. Molecular marker based analysis of genetic diversity and relationships in some turkish and foreign olive cultivars and accessions...............................177* [*electronic version only] SABRAO Statutes (reprinted from Vol. 13(1) 1981 SABRAO Board ……………….xviii Instructions for authors ………......xx

Ruttanaprasert R, Jogloy S, Vorasoot N, Kesmala T, Kanwar RS, Holbrook CC,

SABRAO THE SOCIETY FOR THE ADVANCEMENT OF BREEDING RESEARCH IN ASIA AND OCEANIA Visit our website at: http://www.sabrao.org/

SABRAO JOURNAL OF BREEDING AND GENETICS ISSN 1029-7073 The SABRAO Journal of Breeding and Genetics is the official publication of the Society. Its objective is to promote the international exchange of research information on plant breeding, by describing new findings, theories, and/or achievements of a basic or practical nature. It also provides a medium for the exchange of ideas, news of meetings, and notes on personal and organizational achievements and developments among the members of the Society. Research articles, short communications, methods, reviews, tutorials and opinion articles will be accepted or invited for publication. Scientific contributions will be refereed and edited to international standards. The journal mainly publishes articles for SABRAO members and it is strongly preferred that at least one author should be a current member of the society. From January 2012, there is a US$50 publication fee for SABRAO members FOR ALL ARTICLES, which must be paid before publication (after acceptance of the article). This requirement is to cover journal printing costs and to maintain the website. Non-members may also publish in the journal for a publication fee of US$ 200 per article. ADVERTISEMENTS Advertising will be accepted from Universities offering courses of potential interest to students from SABRAO countries and from book companies or computer software suppliers whose products promote the aims of the Society. Prices are available on application to the Editorial Board.

SABRAO WEBSITE http://www.sabrao.org This website will contain information about the society, information about current officers and regional secretaries for society members, upcoming congresses, and issues of the SABRAO Journal of Breeding and Genetics. In order to improve access for authors and researchers, reprints of journal articles will be posted as soon as the journal issue is published.

i

REGIONAL SECRETARIES Regional Secretaries are elected by the members in each Region. They play an indispensable role in the operations of the Society by: • notifying members of Society announcements, e.g. from the SecretaryGeneral; • recruiting new members; • collecting the annual subscriptions and transferring these to the Treasurer after deducting expenses; • distributing the Journal issues to paying members if these are delivered to a region in bulk; • organizing other activities, such as local chapter newsletters and meetings, e.g., for the induction of new members; • keeping books of account and sending an audited statement to the Treasurer annually; and • providing the Secretary-General with a list of financial members in their region each year. In 2012, the Regional Secretaries are as follows: AUSTRALIA Dr. Phillip Banks Leslie Research Centre, 13 Holberton Street PO Box 2282, Toowoomba, Queensland 4350, Australia Email: [email protected] BANGLADESH Dr. Abul Kashem Chowdhury Professor DepartmentinGenetics and Plant Breeding Patuakhali Science and Technology University Patuakhali-8602, Bangladesh Email: [email protected] CHINA (PEOPLES’ REPUBLIC OF) Prof. Cheng Xuzhen Institute of Crop Sciences Chinese Academy of Agricultural Sciences 30 Bai Shi Qiao Road, Beijing 100081. Email: [email protected] INDIA Dr. Ramakrishnan M. Nair AVRDC - The World Vegetable Center Regional Center for South Asia ICRISAT Campus, Patancheru 502 324 Hyderabad, Andhra Pradesh Email: [email protected] INDONESIA Dr. Ismiyati Sutarto

ii

Horticulture/Agriculture, CRDIRT-BATAN Jl Cinere, Pasar Jumat, Jakarta 12440. Email: [email protected] JAPAN Prof. Kazutoshi Okuno Laboratory of Plant Genetics and Breeding Science Graduate School of Life and Environmental Sciences University of Tsukuba, Tennodai 1-1-1, Tsukuba, 305-8572. Email: [email protected] KOREA Dr. Kyu-Seong Lee, Reclaimed Land Agriculture Research Division NICS, RDA 570-080 #457 Pyeongdong-ro, IKSAN, Jeollabuk-do. Email : [email protected] MALAYSIA Dr. Abdul Rahim Bin Harun Malaysian Nuclear Agency Bangi 43000, Kajang, Selangor. Email: [email protected] PAKISTAN Prof. Hidayatur Rahman Department of Plant Breeding and Genetics NWFP Agricultural University, Peshawar. Email: [email protected] PHILIPPINES Prof. Teresita Borromeo Department of Agronomy, University of the Philippines Los Baños, College, Laguna. Email: [email protected] SRI LANKA Dr. Tissa Rajapakshe, Central Rice Research Station, Batalagoda, Ibbagamuwa. Email: [email protected] TAIWAN, REPUBLIC OF CHINA Dr. Hsun Tu, Rural Development Foundation, 5F, 7, Section 1, 4 Roosevelt Road, Taipei 100. Email: [email protected]

iii

THAILAND Dr. Kamol Lertrat Department of Plant Science and Agricultural Resources Faculty of Agriculture,Khon Kaen University Khon Kaen 40002, Thailand Email: [email protected] USA/CANADA. Dr. Georgia Eizenga USDA-ARS Dale Bumpers National Rice Research Center 2890 Hwy. 130 East, Stuttgart, AR 72160 Email: [email protected] VIETNAM Dr. Bui Chi Buu Institute of Agricultural Sciences for Southern Vietnam 121 Nguyen Binh Khiem, District I, Ho Chi Minh City. Email: [email protected]

SABRAO EDITORIAL BOARD SABRAO is delighted to announce the formation of the inaugural SABRAO Editorial Board in 2012. By establishing an editorial team co-ordinated by the Editor-in-chief, it is hoped that the efficiency, content and quality of the journal will dramatically improve. The main duty of associate editors will be processing manuscripts for publication in the journal. This involves finding reviewers, communicating with corresponding authors, following up completed reviews of manuscripts, checking revisions are thoroughly done, and editing/formatting. Each Associate Editor will be acknowledged as the “communicating editor” for the relevant article when it is published from 2012. Other minor duties include being a contact point for SABRAO members in their respective countries, providing new ideas for the journal (e.g. topics for special issues, ideas for website etc.), and assisting in the preparation and compilation of special issues and conference proceedings. Associate Editors: Dr. Sang-Nag Ahn Professor Department of Agronomy, College of Agriculture & Life Sciences Chungnam National University, Daejeon 305-764 REPUBLIC OF KOREA E-mail: [email protected] Area of expertise: QTL mapping, molecular genetics and breeding of rice Dr. CN Neeraja Principal Scientist, Biotechnology Unit Crop Improvement Section

iv

Directorate of Rice Research, Rajendra Nagar, Hyderabad – 500030 INDIA Email: [email protected] Area of expertise: molecular genetics and breeding Dr. C. Ravindran Assistant Professor Krishi Vigyan Kendra Agricultural College and Research Institute, Tamil Nadu Agricultural Univeristy, Madurai, Tamil Nadu, 625107 INDIA E-mail: [email protected] Area of expertise: breeding and genetics of horticultural and fruit species Dr. Naqib Ullah Khan Professor Department of Plant Breeding and Genetics Khyber Pakhtunkhwa Agricultural University Peshawar 25130 PAKISTAN Email: [email protected] OR [email protected] Area of expertise: plant breeding and quantitative genetics Dr. Sathiyamoorthy Meiyalaghan (Mei) Scientist Plant & Food Research Private Bag 4704, Christchurch, 8140 NEW ZEALAND Email: [email protected] Area of expertise: genomics and molecular breeding Dr. Cheng Xuzhen Professor Institute of Crop Sciences, Chinese Academy of Agricultural Sciences (CAAS) Beijing 100081, China Email: [email protected] Area of expertise: plant breeding in pulses Dr. Ramakrishnan M. Nair Vegetable Breeder - Legumes AVRDC - The World Vegetable Center ICRISAT Campus, Patancheru 502 324 Hyderabad, Andhra Pradesh, India Email: [email protected] Area of expertise: plant breeding and genetics research in pulses and pasture legumes Dr. Sivananda V. Tirumalaraju Research Associate II Soybean Breeding, Genetics and Genomics Program Department of Plant Science South Dakota State University

v

Brookings, SD, USA- 57006 Email: [email protected] Area of expertise: plant breeding (peanut, soybean and canola), molecular breeding, molecular marker technology Deputy Editor-in-Chief Dr. Sanun Jogloy Department of Plant Science and Agricultural Resources Faculty of Agriculture, Khon Kaen University Khon Kaen 40002 THAILAND Email: [email protected] Area of expertise: plant breeding, quantitative genetics, physiological traits Editor-in-Chief Dr. Bertrand (Bert) Collard Scientist International Rice Research Institute (IRRI) Los Banos, Laguna 4031 PHILIPPINES Email: [email protected] Alternative: [email protected] Area of expertise: plant breeding and genetics, QTL analysis, molecular breeding

vi

SPECIAL ANNOUNCEMENTS MOVING TO AN ELECTRONIC JOURNAL SYSTEM From 2013 onwards, the journal will completely move to an electronic format with open access. All articles will be published as pdf files on the website. This was a unamnimous decision made during the SABRAO General Meeting held in Chiang Mai, Thailand (January 2012) based on providing greater access for the journal and due to the financial status of the society. The December 2012 issue will be the last printed issue of the journal.

NEW SCOPE Plant breeding has changed considerably in the last 20 years. The Editorial Board proposes that from 2013 onwards, the scope of SABRAO J. Breed. Genet. will focus on more specific topics of breeding and genetics research that are of direct practical relevance to plant breeders. Classical quantitative genetics research will be considered in the context of how useful the research is to breeders. Authors conducting research in the following topics will be encouraged to submit their articles to the journal: • Classical quantitative genetics investigating genetic control of simple or oligogenic, trait heritabilities, combining ability (e.g. diallel analysis) • Use of agronomic, morphological or physiological traits in selection • Genetic diversity analysis – primarily using DNA markers • Multi-environment trial analysis • Germplasm evaluation • QTL mapping and validation • New methods (e.g. phenotyping methods) of broad interest to breeders Other topics may be submitted after consultation with the Editorial Board. A survey will be prepared and posted on the website later in the year, and disseminated by regional secretaries. The Editor-in-Chief welcomes any feedback or suggestions.

vii

SPECIAL ISSUE OF SABRAO JOURNAL OF BREEDING AND GENETICS IN 2012. To commemorate and document presentations of the 12th SABRAO Congress Plant breeding towards 2025: Challenges in a rapidly changing world (An International Conference to Celebrate His Majesty King Bhumibol’s 84th (7 Cycle) Birthday Anniversary), we are delighted to announce that selected presentations will be published in a special issue of the SABRAO Journal of Breeding and Genetics in 2012. All articles will be published online, from the Journal’s website (http://www.sabrao.org/).

SABRAO STATUTES At the General Meeting held during the 12th SABRAO Congress, it was noted that the SABRAO Statutes were not readily available, and hence they have been reprinted in this issue.

CALL FOR SOCIETY MEMBERS TO BE REVIEWERS FOR OUR JOURNAL The SABRAO journal continues to receive a large number of articles. Reviewers play a critical role for the journal by evaluating and editing manuscripts. Interested society members - especially on topics involving quantitative genetics and genetic diversity are encouraged to register as a potential reviewer for manuscripts submitted to the journal by emailing the Editor-in-Chief.

viii

RESEARCH ARTICLE

SABRAO Journal of Breeding and Genetics 44 (1) 1-8, 2012

STABILITY OF FRESH FRUIT BUNCH OF OIL PALM CROSS (Elaeis Guineensis Jacq.) IN SOUTHERN THAILAND SUDANAI KRUALEE1*, SAYAN SDOODEE1, THEERA EKSOMTRAMAGE1 and VINICH SEREEPRASERT1 1

Department of Plant Science, Faculty of Natural Resources, Prince of Songkla University, Hat Yai, Thailand Corresponding author email: [email protected]

SUMMARY This research aimed to study the interaction between seven crosses and three locations of 4-year old oil palm crosses (cross number 501, 506, 512, 514, 521, 523 and 530) were provided by Pao-Rong Oil Palm Company. These crosses were grown in three locations in southern of Thailand, including Nuea-Khlong District, Krabi Province; Ron-Phiboon District, and Cha-uat District; Nakhon Sri Thamarart Province. The experiment was arranged in a completely randomized design for each location during July 2009 – June 2010. Each pair of cross was set as treatment, and fresh fruit bunch of each treatment was harvested from five oil palm trees. The additive main effects and multiplicative interaction model (AMMI) were applied for the analysis of stability of fresh fruit bunch characteristics, given cross was fix factor and location was random factor. The result indicated that the influence of interaction between cross and location to differentiate the fresh fruit bunch was statistically significant. Specifically, each pair of cross grown in different locations exhibited different yields. The result showed that cross number 530 was the most stable due to similar fresh fruit bunches in the three locations. However, cross number 512, 514 and 501 were less stable, whereby fresh fruit bunch differed among locations. Cross number 512 yielded the most fresh fruit bunch at Ron-Phibun District (99.1 kg/palm/year). Cross number 501 yielded the most fresh fruit bunch at Cha-uat District (38.1 kg/palm/year) and cross number 514 yielded the most fresh fruit bunch at Nuea-Khlong District (57.9 kg/palm/year). Keywords: Genotype x environment interaction, Stability, AMMI, Oil palm, Fresh fruit bunch Manuscript received: January 10, 2011; Decision on manuscript: July 27, 2011; Manuscript accepted in revised form: August 21, 2011. Communicating Editor: Bertrand Collard

1

Krualee et al. (2012)

INTRODUCTION The yield experiments conducted in several environments is important for agricultural research (Gauch, 2006) because it helps to understand the response of genotypes to environment. Yield performance of genotypes is not always the same in different environments (Dabholkar, 1992). Thus, variance has affected yield. Variance is mostly caused by three major factors; environment, genotype and genotype by environment (G x E) interaction, which influences different yield. Each factor influences yield as follows. Variance caused by different environments is hardly important for breeding. This variation is divided into two terms i.e. predictable and unpredictable environmental variation. Predictable variation includes the permanent characters (climate, soil type) and the fluctuation characters which is determined by system and man (day length, sowing density), as for unpredictable variation includes fluctuation in weather such as amount and distribution of rainfall, temperature. Both variance effect to genetic constitution of population two ways as; long period it may lead to evolutionary changes and short period it help to segregate genetic material (Dobholhar, 1992). Genotypic variance is the difference in mean yield between varieties. These varieties have been classified as two groups; homogeneous population (i.e. pure line, single crosses) and heterogeneous population (i.e. composite variety, synthetic

variety) (Allard and Bradshaw, 1964). When these populations are planted on several environments, the G x E interaction has occurred. This case shows the homogeneous population tends to interact with the environment more than heterogeneous population because their genetic structure has susceptibility to environmental variation. But if it has not G x E interaction then only genotypic variance is useful (Annicchiario, 2002). The G x E interaction is differential genotypic expression across environments which reduce relation between phenotypic and genotypic value (Falconer, 1981). When genotypes are grown at several environments, each genotype ranking is not the same in each environment. This is an important factor because it is important for breeders in the process of developing improved varieties: (1) to help in determining a plan to cope with the effects of the interaction (Annicchiario, 2002); (2) to assist in parent selection from base population; and (3) to help in the evaluation of adaptation (Aina et al., 2007). In the context of plant breeding, adaptation indicates the ability of the varieties to be high yielding with respect to a given environment which has two contexts. Wide adaptation is the good performance in nearly all environments while specific adaptation is the good performance in a particular environment (Annicchiario, 2002). These adaptation evaluation methods are presented in several ways: (1) partitioning of variance; (2) regression analysis i.e. Finlay and

2

SABRAO J. Breed. Genet. 44(1): 1-8

Wilkinson (1963), Eberhart and Russell (1966); (3) non-parametric statistics; (4) multivariate techniques i.e. pattern analysis, the additive main effect and multiplicative interaction analysis (AMMI) introduced by Gauch (1988). The AMMI model is applied to describe relationships among sites and among genotype, using yield data from genotype x site matrices generated by breeding programs in each crop cycle. This model is more efficient than other methods due to four reasons. Firstly, AMMI can estimate yield more accurately than traditional methods. Secondly, AMMI provides the adjusted yield estimate which leads to corrected ranking of the genotype within each environment, so genotypes can be selected correctly. Thirdly, AMMI is can be represented as a biplot graph. This graph can clearly show a numerous complex yield pattern. Moreover, a biplot often captures 90 % of treatment variation and least 5 % to 10 % of treatment variation is noise. Fourthly, yield experiments have GER plot (include G genotype, E environment and R replication) for increasing the accuracy and scope of the trial but these experiments are expensive. This model help to economize the research cost because the AMMI model can use a few replications (Gauch, 1992a). This analysis is combination between analysis of variance (ANOVA) with principal components analysis (PCA). The ANOVA is used for considering main effect i.e. environment and genotype. This model helps to sort the variance due to different

sources and provides basis for test of significance (Singh and Chadhary, 1979). The PCA considers multiplicative effect i.e. G x E interaction as the ANOVA’s residual. The main idea of PCA is reduction of the dimensionality of a data set that may contain many highly correlated variables while retain possible of the variation in the data set (Gauch, 1992b). The result presentation is shown by biplot graph which allows information on both samples and variables of a data matrix to be displayed graphically. Samples are displayed as points while variables are displayed as vectors. Generally AMMI models are presented in two types. Firstly, a graph presents the relationship between principle component axis 1 and average yield that shows the interaction between genotype and environment. The points on the ordinate near zero have little interaction (stable) but the points plotted far zero are affected by interaction. Another graph presents the relationship between PCA axis 1 and PCA aixs 2, which implies the classification of suitable variety for any environment. If point of variety close to any environment, it will be considered as suitable for that environment. Fresh fruit bunch is one of the agronomic characters of oil palm which positively correlates with the oil yield, if oil palms produce fresh fruit bunch so much then we can extract many palm oil. But the fresh fruit bunch character has highly variance (Corley and Tinker, 2003). Consistently, Kushairi et al. (1993) studied the variation of 99 pairs of Tenera’ fresh fruit bunch and the result

3


showed that each cross yielded significantly different. The variation is influenced by many factors Obisesan and Fatunla (1983) indicated cross, age and interaction between cross and age influence to fresh fruit bunch of tenera. Additionally, Rafii et al. (2001) suggested cross, location and the interaction between cross and location influence to fresh fruit bunch of tenera. According to above reports, the interaction between cross and location influenced significantly to fresh fruit bunch of oil palm so it was important to oil palm breeding program. This research aims to study the influence of interaction between genotype and environment on fresh fruit bunch of tenera oil palm and identify cross that can yield stable or specific.

MATERIALS AND METHODS Data used in this experiment consisted of seven 4-years old oil palm crosses ( cross number 501, 506, 512, 514, 521, 523 and 530 ) aged 4 years from Pao-Rong oil palm company. These crosses were grown in 3 locations of southern Thailand; Nuea-Khlong District, Krabi Province; Ron-Phiboon District, and Cha-uat District; Nakhon Sri Thamarart Province. A completely randomize design was applied in each location. Initially, oil palm crosses were set as a treatment, and fresh fruit bunch harvested from each oil palm tree. There were 5 replications per treatment (1 tree / 1 replicate). Each oil palm tree was weighed during July 2009 – June 2010. To

achieve the above objectives, the experimental data was analyzed using the additive main effects and multiplicative interaction model which determining crosses are fixed factors and locations are random factors. This model can be written as; Yij = µ + αi + βj + Σλn ξinηjn + θij Where: Yij= Yield of genotype i in environment j µ = Grand mean αi = Genotype i mean deviation βj = Environment mean j deviation λn = Singular value for PCA axis n ξin = Genotype i eigenvector values for PCA axis n ηjn = Environment j eigenvector value for PCA axis n N = Number of PCA axis θij = Error RESULTS The results of variance analysis showed that cross (genotype) did not affect fresh fruit bunch. There were no differences in average yields from each cross for the different environments. While the influence of location (environment) affected to fresh fruit bunch significantly, average of all crosses were different in each location. The interaction between cross and location affected fresh fruit bunch significantly; fresh fruit bunch of each cross from each location was different, as presented in Table 1.

4


55.23 kg/palm/year and –5.13, respectively. Additionally, cross 530 had the least average fresh fruit bunch and PCA score was close to zero, 27.87 kg/palm/year and 0.03, respectively. The result of a cross’s stability analysis was represented as biplot graphs (Figure 1 and Figure 2).

Table 2 shows that cross 514 had the highest average fresh fruit bunch at Nuea-Khlong District. Cross 512 had the highest average fresh fruit bunch at RonPhibun District, and cross 501 had the highest average fresh fruit bunch at Cha-uat District. According to an averaged fresh fruit bunch from 3 locations, it was found that cross 512 had the highest average fresh fruit bunch and PCA score was far from zero,

Table 1. Analysis of additive main effects and multiplicative interaction of tenera’s fresh fruit bunch in 3 locations in Southern Thailand Source of variance Cross ( C ) Location ( L ) CxL

Degree of freedom

Sum of Square

6 2 12

1,279.61 2,692.53 2,333.22

213.27 1,346.27 194.44

1.10ns 60.43** 8.73**

PCA 1 PCA 2 Error Total

7 5 84 104

1,833.81 499.41 1,871.25 8,176.62

261.97 99.88 22.28

11.76** 4.48**

ns = Non significant difference

Mean Square

F-test

** Highly significant difference P ≤ 0.01

Table 2. Average fresh fruit bunch for each cross in each location, average fresh fruit bunch, variety PCA 1 score and variety PCA 2 score Cross 501 506 512 514 521 523 530 Mean PCA 1 Score PCA 2 Score

NueaKhlong 30.02 31.74 32.92 57.90 25.74 32.02 19.28 32.80

RonPhiboon 47.68 54.70 99.12 39.94 45.42 52.02 42.04 54.42

Cha-uat 38.14 21.86 33.66 22.74 29.38 31.84 22.28 28.56

4.06

-5.04

0.97

-2.51

-1.29

3.80

Mean 38.61 36.10 55.23 40.19 33.51 38.63 27.87

PCA 1 Score 0.75 -0.29 -5.13 3.95 0.41 0.28 0.03

PCA 2 Score 2.33 -1.03 -1.73 -2.97 1.45 0.78 1.16

5


6 NK

4

PCA 1

2 0

-2

20

25

CU 530

521

30

35

514

506

501 523

40

45

50

55

-4

60

Ron-Phiboon (RP) Cha-uat (CU) Nuea-Khlong (NK)

RP 512

-6

Mean (kg/palm/year)

Figure 1 Biplot graph between average fresh fruit bunch and PCA 1 score

6 CU

2

501 521 530 523

PCA 2

4

0

506

RP 512

-2

Ron-Phiboon (RP) Cha-uat (CU) Nuea-Khlong (NK) NK 514

-4 -6 -6

-4

-2

0 PCA 1

2

4

6

Figure 2 Biplot graph between PCA 1 score and PCA 2 score

6


DISCUSSION The results showed that location and interaction between cross and location influence fresh fruit bunch yield differently, which is consistent with the study by Rafii et al. (2001). However, the influence of cross did not affect fresh fruit bunch, which was different from the report of Kushairi et al. (1993); Obisesan and Fatunla (1983) and Rafii et al. (2001). This might be due to the fact that the the parents of these crosses were selected from the same population, so genotype of these crosses were similar and hence no statistically significant differences were detected for fresh fruit bunch. Regarding the interaction between cross and location or a multiple analysis, the results showed that PCA 1 and PCA 2 could separate sum of squares of the interaction between cross and location at 78.60 and 21.40 percentage, respectively. Furthermore, it could detect the difference of fresh fruit bunch significantly, implying different location yielded different fresh fruit bunch. According to the analysis, PCA 1 was considered as the stability of cross. The result showed that cross 530 was close to zero, implying this cross yield similar was in all three locations (i.e. highly stable), but fresh fruit bunch for this cross was low (27.87 kg/palm/ year ). The conclusion for this cross was that it was not suitable to be introduced for production. It is stable across environments hence it can yield something in areas where others fail, hence it is a good cross but need to be improved further

through breeding. Cross 523, PCA 1 close to zero (0.28), but fresh fruit bunch yielded moderately and consistently in all three locations (38.63 kg/ palm / year). This cross was more beneficial to agriculturists than cross 530. PCA 1 of cross 501, 512 and 514 was far from zero, indicating these crosses had high interaction between cross and location, but fresh fruit bunch yield was not consistent in all three locations. The cross 512 yielded the most fresh fruit bunch at RonPhibun District. Cross 501 yielded the most at Cha-uat District and cross 514 yielded the most at Nuea-Khlong District. Therefore, it is suggested that each cross is specifically suitable for each specific environment.

ACKNOWLEDGEMENTS We are grateful to the Graduate School and Oil Palm Research and Development Center, Prince of Songkla University for financial supported of this research. Equipments and facilities were provided by Department of Plant Science, Faculty of Natural Resources, Prince of Songkla University. Furthermore, we thank PaoRong Oil Palm Company for providing the oil palm crosses.

REFERENCES Aina OO, Dixon AGO, Akinrinde EA (2007). Additive main effects and multiplication interaction (AMMI) analysis for yield of Cassava in Nigeria. J. Biol. Sci. 7: 796–800. Allard RW, Bradshaw AD (1964). Implications of genotypeenvironment interaction. Crop Sci. 4: 503-507.

7


Annicchiario P (2002). Introduction. In: P. Annicchiario ed., Genotype x environment interaction challenges and opportunities for plant breeding and cultivar recommendations. Food and Agriculture Organization of The United Nations, Rome, pp. 1–4 . Corley RHV, Tinker PB (2003). Selection and breeding. In: Corley, R.H.V. and P.B. Tinker eds., The Oil Palm. Blackwell Science Ltd., Oxford, pp 133–200. Dabholkar AR (1992). Genotype environment interaction and stability parameters. In: AR Dobholhar Ed., Elements of Biometrical Genetics. Concept Publishing Company, New Delhi, pp. 326–365. Eberhart SA, Russell WA (1966). Stability Parameters for Comparing Varieties. Crop Sci. 6: 36–40. Falconer DS (1981). Correlated characters. In: D.S. Falconer, Quantitative Genetics Longman Inc., New York, pp. 281–300. Finlay KW, Wilkinson GN (1963). The Analysis of Adaptation in a PlantBreeding Program. Aust. J. Agric. Res. 14: 742–754. Gauch HG (1988). Model Selection and Validation for Yield Trials with Interaction. Biometrics 44: 705–715. Gauch HG (1992a). Introduction. In: H.G. Gauch, Statistical Analysis of Regional Yield Trials AMMI Analysis of

Factorial Designs. Elsevier Science Publisher B.V., Amsterdam, pp. 1–14. Gauch HG (1992b). AMMI and related models. In: HG Gauch, Statistical Analysis of Regional Yield Trials AMMI Analysis of Factorial Designs. Elsevier Science Publisher B.V., Amsterdam, pp 53–110. Gauch HG (2006). Statistical analysis of yield trial by AMMI and GGE. Crop Sci. 46: 1488–1500. Kushairi A, Rajanaidu N, Jalani BS, Zakri AH (1993). Variation Malaysian dura x pisifera planting materials I. bunch yield. Elaeis 6: 14–25. Obisesan IO, Fatunla T (1983). Genotype x environment interaction for bunch yield and its components in the oil palm (Elaeis guineensis Jacq.). Theor. Appl. Genet. 64: 133–136. Rafii MY, Rajanaidu N, Jalani BS, Zakri AH (2001). Genotype x environment interaction and stability analysis in oil palm (Elaeis guineensis Jacq.) progenies over six locations. Journal of Oil Palm Research 13: 11–41. Singh RK, Chaudhary BD (1979). Variance and covariance analysis. In: Singh, RK and Chaudhary BD Eds., Biometrical methods in Quantitative genetic analysis. Kalyani Publishers, New Delhi, pp. 39-69.

8

SABRAO J. Breed. Genet. 44(1): 1-8 RESEARCH ARTICLE

SABRAO Journal of Breeding and Genetics 44 (1) 9 -27, 2012

GENETIC ANALYSIS OF PHYSIOLOGICAL AND YIELD TRAITS UNDER DROUGHT STRESS CONDITIONS IN WHEAT WAJID ALI JATOI1*, MUHAMMAD JURIAL BALOCH1, NAQIB ULLAH KHAN2, MOULA BUX KUMBHAR1 AND MUHAMMAD IBRAHIM KEERIO3 1

Department of Plant Breeding and Genetics, Sindh Agriculture University, Tandojam, Pakistan 2 Department of Plant Breeding and Genetics, Khyber Pakhtunkhwa Agricultural University, Peshawar, Pakistan 3 SZABAC, Dokri, Larkana, Sindh-Pakistan *Corresponding author email: [email protected]

SUMMARY Breeding for drought tolerance requires the knowledge of gene action governing different physiological and yield traits. Six wheat genotypes were crossed in 6 × 6 half diallel fashion during 2009. The parents and their 15 F1 hybrids were evaluated in a randomized complete block (RCB) design with four replications under two irrigation treatments during 2010 at Sindh Agriculture University, Tandojam-Pakistan. All the physiological and yield traits were significantly affected by the water stress treatments. Physiological and yields traits like productive tillers plant-1, grains spike-1, grain yield plant-1, relative water content, stomatal conductance and leaf area were good indicators to distinguish drought tolerance of wheat genotypes. The wheat cultivars such as TD-1, Sarsabz and SKD-1 performed well under stress conditions. The mean squares for general and specific combining ability were significant for all the parameters suggesting that both genetic variances were substantial. Generally, in stress conditions, the GCA was predominant as compared to SCA variances. The GCA effects of the parents in non-stress and in water stress conditions revealed that TD-1, Sarsabz and Kiran were good general combiners, hence are potential parents to be used in hybridization and selection programmes for the development of drought tolerant breeding material. It was interesting to note that hybrid performance per se was largely reflected in their SCA effects. Among the fifteen F1 hybrids evaluated, the crosses TD1 × TJ-83, Kiran × Sarsabz and Sarsabz × Moomal were good specific combiners for majority of the traits, thus these hybrids may be considered as prospective hybrids under drought conditions.

Keywords: Diallel crosses, combining ability, water stress, drought tolerance, Triticum aestivum L. Manuscript received: June 25, 2011; Decision on manuscript: October 7, 2011; Manuscript accepted in revised form: March 20, 2012 Communicating Editor: Bertrand Collard

9

Jatoi et al.2012 INTRODUCTION The development of drought resistant wheat cultivars is a long, hard and complex process when the motive involves the incorporation of drought tolerance combined with higher grain yields. Several yield contributing parameters such as plant height, tillers plant-1, days to maturity, spikelets spike-1, grains spike-1 and 1000-grain weight has been associated with moisture stress tolerance of wheat (Munir et al., 2006). Drought stress reduces main stem height in crops by decreasing the number of nodes and internode length. It causes low dry matter production because of decreased stem height and stem diameter associated with limited leaf expansion and reduced tiller number. Drought also decreases the number of spikelets which is essentially linked to the decreased number of flowers produced. Reduced grains spike-1, grain weight and yield indicate that low dry matter accumulation was the main cause of smaller spikes and seed number. Drought at grain filling stage reduces the cell size and their number which result in shriveled grains and induces early maturity (Day and Intlap, 1970; Cooper et al., 1994; Rathore, 2005; Nisar et al., 2007). Besides yield traits, a physiological understanding of plant responses to drought has often been sought on the basis that such an understanding will assist plant breeders to develop higher yielding cultivars for water-scarce environments. However, despite the availability of an extensive literature on plant response to drought, there are still very few

documented examples where a physiological investigation of drought has identified the traits that limit yield under drought or where these have been used in successful crop improvement programs and have enhanced crop yields. It has been noted that direct selection for grain yield under water-stressed conditions has been hampered for several reasons, such as low heritability, polygenic control of traits, epistasis, significant genotype-by-environment (G × E) interaction and quantitative trait loci (QTLs)-by-environment (QTL × E) interaction (Piepho, 2000). Such complexity of drought tolerance mechanisms explains the slow progress of increasing yield in drought prone environments. Genetic improvement of crops for drought resistance requires a search of possible physiological and yield attributes and the exploitation of their genetic variation. Conventional breeding of grain crops continues to deliver improved cultivars to farmers with little evidence of leveling-off in yield (Brancourt-Hulmel et al. 2003). In general, these genetic increases in yield go hand-in-hand with improved crop cultivars through understanding the genetic basis of such characters. Genetic analysis to determine general combining ability (GCA) and specific combining ability (SCA) of parents for drought tolerance are very important attributes to help wheat breeders to develop new drought tolerant cultivars. The GCA is defined as an average performance of a genotype/parent in a series of hybrid combination while SCA indicates where as those instances where certain

10


hybrids are either better or poorer than would be expected on average performance in hybrid combinations. Thus SCA is important for hybrid crop development, whereas GCA is useful for hybridization and selection programs under water stress conditions. The significance of both GCA and SCA under nonstress and in water stress conditions for seeds spike-1, effective tillers plant-1 and grain yield plant-1 was reported (Drikvand et al., 2005; AlNaggar et al., 2007; Kamaluddin et al., 2007; Dhadhal et al., 2008). The type of gene action appeared to be different under drought than under non-stress, with additive effects being more important under drought and dominance effects in non-stress (Betran et al., 2003). On the contrary, Iqbal and Khani (2006), Dere and Yildirim (2006) and Iqbal et al. (2007) reported that, though GCA and SCA variances were significant for grain yield, nevertheless SCA was predominant and also found several parents as good general and as well as good specific combiners for grain yield. Farshadfar et al. (2000) studied some drought tolerance criteria in 8 × 8 diallel crosses and their results revealed that additive gene action was predominant for relative water content in leaf (RWC %). Therefore, the objective of this research was to investigate the combining ability of various parental cultivars and their F1 populations for physiological and yield traits of wheat under nonstress and water stress conditions.

MATERIALS AND METHODS Six wheat genotypes such as TD-1, Kiran, Sarsabz, Moomal, SKD-1 and TJ-83 with diverse origin and characters were crossed in a 6 × 6 half diallel mating fashion during 2009, thus 15 F1 hybrids were developed. These 15 F1 hybrids along with their parents were grown in a randomized complete block design (RCBD) with factorial arrangement in two irrigation treatments and four replications during 2010 at Sindh Agriculture University, TandojamPakistan. Two irrigations regimes were non-stress and water stress at anthesis. The irrigations were applied in measured quantities of 1506 and 1255 mm ha-1 in nonstress and water stress treatments, respectively through siphons. The data were analyzed for significance between treatments and genotypes according to Gomez and Gomez (1984), whereas combining ability analysis was carried out by adapting Griffing (1954), MethodII, Model-I (Singh and Chaudhary, 1985). The yield and physiological traits studied were; productive tillers plant-1, grains spike-1, grain yield plant-1 (g), relative water content in leaf (RWC %), stomatal conductance (mmol m-2 s-1) and flag leaf area (cm2). The RWC % was determined according to formula developed by Schonfeld et al. (1988) with the formula: RWC% = (fresh weight-dry weight)/(turgid weight-dry weight) x 100. The stomatal conductance was measured by Porometer AP4, Delta Devices, Cambridge, U.K. and leaf area by Portable Delta-T leaf Area Meter LI-3000/Lambda Instrument Corp. Lincolin,

11

Jatoi et al.2012 Nebraska, USA. Data analysis was performed using Mstatc software.

RESULTS Pooled mean squares from analysis of variance showed significant effects for water stress treatment on wheat genotypes and characters including productive tillers plant-1, grains spike-1, grain yield palnt-1, relative water content, stomatal conductance and leaf area (Table 1). This allowed conducting combining ability analysis in both environments. The GCA and SCA mean squares for all the traits were also significant (Table 2), yet for physiological and yield traits in stress conditions, the GCA was predominant for most of the traits as compared to SCA variances. The interactions of treatment × genotype for all the traits were also significant except stomatal conductance (Table 1) that implied variable response of genotypes over the stress treatments for various traits. Mean performance of parents and per se F1 hybrids The mean performance of parents and F1 hybrids per se for physiological and yield traits is presented in Table 3 whereas the relative difference of parents for different traits under drought stress and non-stress conditions is shown in Table 4. For productive tillers plant-1, though all the parents showed minimum declines due to water stress, nevertheless, parents, TD-1, Kiran and Sarsabz maintained their tillers largely. In F1 hybrids, the crosses TD-1 × Moomal, TD-1 × SKD-1, Sarsabz

× SKD-1, Sarsabz × TJ-83, Moomal × TJ-83 and SKD-1 × TJ83 showed minimum reductions in normal and drought conditions, hence indicating their being relatively more tolerant as compared to other hybrids in the test. For grains spike-1 and grain yield plant-1, the parents TD-1, Sarsabz, Moomal and SKD-1 generally provided minimum reductions from non-stress to stress conditions. In plant breeding, it is commonly assumed that when good performing parents are crossed, they are expected to produce better hybrids also but this assumption did not always hold true. Among the fifteen hybrids developed from 6 × 6 half diallel crosses, the F1 hybrids TD-1 × Kiran, TD-1 × Sarsabz, TD-1 × SKD-1, Kiran × Sarsabz and Kiran × Moomal gave minimum drop due to water stress in yield traits (Table 3). The parent, TD-1 showed maximum relative water content (89.3%) and stomatal conductance (473.3 mmolm-2s-1) in non-stress while in water stress conditions, in addition to TD-1, Sarsabz also maintained high relative water content with 44.6% and 44.8%, thus gave relatively small decrease from non-stress to water stress conditions (Table 4). The minimal stomatal conductance (139.0 mmolm-2s-1) was noted in TD-1 followed by SKD-1 (140.3 mmolm2 -1 s ) in stress whereas TD-1 also scored first in having wider leaves (24.25 cm2) followed by SKD-1 (23.3 cm2) and correspondingly their relative decreases in leaf area were -6.0 and -6.3 cm2, respectively from non-stress to water stress. Regarding the performance of F1 hybrids per se in

12


stress conditions TD-1 × TJ-83 recorded maximum RWC % (47.0%), with minimum stomatal conductance (128.0 mmolm-2s-1) and produced wider leaves (24.0 cm2) (Table 3). General and specific combining ability effects The mean squares for GCA and SCA were significant for all the physiological and yield traits in both non-stress and in stress conditions, nonetheless in stress conditions, most of the characters showed higher GCA against SCA variances (Table 2). The GCA and SCA effects for physiological and yield traits are depicted in Tables 5 and 6, respectively. Maximum GCA effects of 0.33 was expressed by the parent Sarsabz followed by TD-1 (0.27) for productive tillers plant-1 in non-stress (Table 5) and the same parents also showed higher GCA effects of 0.29 and 0.22 in stress conditions, respectively. It means about the same amount of GCA effects was expressed in both the environments. The F1 hybrids such as TD-1 × TJ-83, Sarsabz × Moomal and Kiran × Sarsabz recorded maximum SCA effects of 3.37, 2.53 and 2.09 in non-stress while in stress their SCA effects were 1.89, 0.64 and 0.73, respectively (Table 6). The hybrids that gave higher SCA effects in stress involved at least one of the parents exhibiting higher GCA effects also. Regarding grains spike-1, the maximum GCA effect was exhibited by the parent Kiran (1.13) followed by TD-1 (0.88) and Sarsabz (0.29) in non-stress condition whereas in stress condition, Sarsabz ranked 1st by

expressing the highest GCA effect of 2.78 while TD-1 (1.65) and Kiran (0.81) ranked 2nd and 3rd, respectively (Table 5). About the F1 hybrids, TD-1 × TJ-83 (6.28 and 9.11), Sarsabz × Moomal (5.42 and 7.92) and Kiran × Sarsabz (3.7 and 4.17) recorded higher SCA effects in non-stress as well as in stress conditions for grains spike-1, respectively (Table 5). At large, parents which expressed higher GCA values in stress as compared to non-stress conditions for grains spike-1 and also demonstrated higher SCA effects with other parents which otherwise exhibited lower GCA effects. In non-stress conditions, Kiran expressed maximum GCA effects (1.20) followed by Sarsabz (0.74) and TD-1 (0.64) for grain yield plant-1, while remaining parents exhibited negative GCA effects (Table 5). However, in stress conditions, due to change in rank, Sarsabz manifested the highest GCA effects (1.66) followed by TD-1 (1.44) and Kiran (0.29). For SCA effects, the hybrid TD-1 × TJ-83 gave maximum effects of 4.96 and 5.04 in nonstress and in stress respectively followed by Sarsabz × Moomal (3.40 and 3.38) and Kiran × Sarsabz (1.83 and 2.45) (Table 5). Nonetheless, maximum SCA effects of 4.59 and 2.30 were expressed by TD-1 × TJ-83 and Sarsabz × Moomal, respectively. The promising two parents, TD-1 and Kiran manifested higher GCA effects of 1.29 and 0.17 for relative water content (RWC %) in non-stress (Table 4) whereas in stress conditions, TD-1 (1.68) and Sarsabz (1.43) gave greater GCA effects, yet remaining parents

13

Jatoi et al.2012 displayed negative effects. The F1 hybrid TD-1 × TJ-83 gave maximum but similar SCA effects of 3.61 in both non-stress and in stress environments, respectively (Table 6). However, TD-1 × Kiran scored next maximum with 0.76 SCA effects in non-stress, hitherto Kiran × Sarsabz expressed next highest SCA effects (2.05) in stress at anthesis. By and large, the hybrids demonstrated greater SCA effects in stress over non-stress conditions for RWC %. With respect to stomatal conductance, the parents TD-1 and Sarsabz exhibited higher GCA effects of 16.04 and 3.16 in nonstress, respectively; however, negative GCA effects are preferable for stomatal conductance in stress conditions. Regarding GCA effects in stress conditions, three out of six parents i.e. Sarsabz (-9.13), TD-1 (-5.66) and Kiran (-4.41) gave negative effects showing less conductance (more resistance) in water-deficit conditions (Table 6). Among the F1 hybrids, TD-1 × TJ-83 (-42.53), Sarsabz × Moomal (-35.00) and Kiran × Sarsabz (-19.78) recorded maximum negative SCA effects demonstrating that these hybrids transpire less water in water stress conditions, hence showing more tolerance (Table 6). The higher SCA effects were relatively more in stress against in non-stress environments. The GCA effects of leaf area presented in Table 4 indicated that the parent TD-1 revealed highest GCA effect of 0.54 followed by Moomal (0.47) and Kiran (0.35) in non-stress conditions. Whereas in stress environment, the parents TD-1 and

Sarsabz exhibited larger GCA effects of 1.53 and 0.96, respectively, hitherto the rest of the parents such as kiran, Moomal, SKD-1 and T.J-83 expressed negative GCA effects in stress conditions. For leaf area, the maximum SCA effects was recorded by the F1 hybrid TD-1 × T.J-83 (2.33) followed by Moomal × T.J-83 (1.90) and SKD-1 × TJ-83 (1.87) in non-stress, whilst Kiran × Sarsabz manifested highest SCA effects of 3.55 followed by TD-1 × TJ-83 (3.05) and Sarsabz × Moomal (2.46) in water stress conditions (Table 6).

DISCUSSION Information about the genetic control of the target traits helps breeders in making important decisions about the choice of the parental material to be used and efficient breeding procedures to be adopted. These decisions could lead to a creation of new plant populations with superior agronomic traits in terms of drought tolerance. Genetic analysis or determining GCA and SCA of parents and their hybrids for drought tolerance is important to develop new drought tolerant breeding material. In the present study, the mean squares for GCA and SCA were significant for all the physiological and yield traits in non-stress and in stress treatments indicating that genotypes performed differently for various traits across the drought conditions. Generally, for yield and physiological traits in stress conditions, the GCA variances

14


were predominant for most of the traits as compared to SCA variances. These results suggested that in stress conditions, more stress responsive genes with additive genetic effects were involved in stress tolerance. Ayoub and Chowdhry (2000) determined genetic mechanisms of physiological and yield traits like flag leaf area, tillers plant-1, grains spike-1 and grain yield plant-1 under irrigated and drought stress conditions. They noted that additive genetic effects were more profound for these traits than nonadditive effects, however contrary to our results; their SCA variances were greater than GCA for yield plant-1. For productive tillers plant1 , the GCA and SCA mean squares were greater in non-stress than in stress conditions suggesting that stress environment caused a reduction for these traits.. The GCA effects were approximately equal in non-stress and in stress conditions suggesting that additive genes were governing this trait and the parents Sarsabz, TD-1 and Kiran exhibited greater and positive effects indicating persistence in additive genes and showed good GCA for productive tillers plant-1 (Table 5). If productive tillers were to be improved for increasing grain yield, then selection would be preferable from early segregating populations developed from the crosses of these best general combiners. The F1 hybrids like TD1 × TJ-83, Sarsabz × Moomal, Kiran × Sarsabz revealed maximum positive SCA effects in non-stress and in stress conditions as well; however, SCA effects in non-stress were generally higher

than in stress conditions (Table 6). There was also close association between per se hybrid performance and SCA effects (Table 6 and Figure 1) which further indicated that selection on the basis of per se hybrid could be equally effective as made on the basis of SCA effects for productive tillers plant-1.

15

Jatoi et al. (2012)

Table 1. Mean squares for morph-yield and physiological traits of parents and F1 hybrids of wheat under drought stress and non-stress conditions. Mean squares Replication Treatment (T) Genotypes (G) T×G Error D.F. = 3 D.F. = 1 D.F. = 20 D.F. = 20 D.F. = 123 Prod. tillers plant-1 35.95 331.52** 9.30** 2.16* 0.40 Grains spike-1 33.69 13285.92** 152.30** 14.89** 0.43 Grain yield plant-1 135.43 2065.01** 91.27** 23.53** 18.85 RWC% 8.89 82903.71** 76.20** 26.91** 0.36 Stomatal conductance 1.83 3.80* 6.20** 1.87N.S. 3.24 Leaf area 43.11 628.72** 31.39** 11.93** 0.44 RWC% = Relative water content in leaf, **, * = Significant at 1 and 5% probability levels respectively, N.S. = non-significant. Physiological and yield traits

Table 2. Mean squares of GCA and SCA for morph-yield and physiological traits of parents and F1 hybrids of wheat under drought stress and non-stress conditions. Non-stress Stress at anthesis G.C.A S.C.A Error G.C.A S.C.A Error D.F. = 5 D.F. = 15 D.F. = 60 D.F. = 5 D.F. = 15 D.F. = 60 Productive tillers plant-1 3.89** 9.45** 0.27 3.09* 3.49* 0.45 Grains spike-1 38.50** 63.44** 0.33 146.75** 97.32** 0.94 Grain yield plant-1 26.12** 30.27** 0.35 68.92** 35.17** 0.47 RWC% 19.89** 6.48** 0.54 51.86** 18.67** 0.66 Stomatal conductance 2905.27** 1691.25** 6.54 2044.45** 3210.74** 19.29 Flag leaf area 9.47** 10.29** 0.26 35.27** 18.17** 0.46 RWC% = Relative water content in leaf, **, * = Significant at 1 and 5% probability levels respectively. Physiological and yield traits

16

SABRAO J. Breed. Genet. 44 (1) 9-27

Table 3. Mean performance of parents and F1 hybrids for yield and physiological traits under drought stress (DS) and non-stress (NS). Parents / F1 hybrids

Productive tillers plant-1 NS DS

Grains spike-1 NS

DS

TD-1 Kiran Sarsabz Moomal SKD-1 T.J-83

7.8 7.5 6.8 6.8 6.8 7.5

6.5 6.3 5.8 4.8 5.3 5.5

66.8 66.0 61.8 60.0 67.5 61.0

50.5 45.5 50.8 40.8 48.5 43.5

TD-1 × Kiran TD-1 × Sarsabz TD-1 × Moomal TD-1 × SKD-1 TD-1 × T.J-83 Kiran × Sarsabz Kiran × Moomal Kiran × SKD-1 Kiran × T.J-83 Sarsabz × Moomal Sarsabz × SKD-1 Sarsabz × T.J-83 Moomal × SKD-1 Moomal × T.J-83 SKD-1 × T.J-83 Mean LSD (5%) (T) LSD (5%) (G) LSD (5%) (T × G )

8.8 7.8 6.8 8.5 11.8 10.8 8.8 7.5 7.8 10.8 8.5 8.3 7.8 7.5 6.5 8.1

5.5 4.3 3.8 5.5 7.5 6.5 5.5 4.5 4.5 5.8 5.5 5.5 4.5 4.8 4.0 5.3

67.0 66.0 59.5 67.5 75.0 71.0 67.0 67.5 66.0 70.0 66.8 66.5 67.8 64.5 60.5 65.9

50.5 49.5 40.0 51.0 58.5 56.0 48.5 50.5 47.8 56.0 47.0 49.5 47.5 40.8 39.5 48.2

0.2 0.6 0.9

0.2 0.7 0.9

Grain yield plant-1 (g) NS DS Parental cultivars 26.3 23.5 25.5 16.5 23.0 20.5 24.0 15.8 23.5 20.8 24.5 16.5 F1 hybrids 26.0 20.0 26.0 20.5 20.8 12.8 25.8 20.5 30.3 24.5 28.3 23.5 25.5 19.5 27.0 20.5 26.0 19.5 28.3 22.0 26.0 19.5 27.0 20.5 24.5 16.8 20.8 14.0 19.0 13.5 25.2 19.1 1.3 4.3 6.1

RWC (%)

Stomatal conductance (mmol m-2 s-1 ) NS DS

NS

DS

89.3 87.8 88.8 87.3 88.5 84.5

44.8 38.0 44.8 37.8 41.8 38.5

473.3 448.5 419.3 400.8 414.5 433.5

90.5 89.8 88.5 88.9 92.2 88.5 88.2 88.7 88.0 87.9 88.4 88.0 88.5 86.0 87.0 88.3

44.0 44.5 43.0 45.0 47.0 45.8 43.0 44.0 43.0 44.8 42.5 44.0 43.0 40.0 40.0 42.8

466.8 471.5 460.0 473.0 476.3 474.0 430.0 445.0 450.0 471.5 465.0 470.5 465.0 435.0 453.8 452.2

0.2 0.6 0.9

0.8 2.5 3.5

Leaf area (cm2) NS

DS

139.0 142.8 158.5 159.8 140.3 158.8

30.3 31.0 28.3 30.5 29.5 26.3

24.3 16.3 22.3 20.3 23.3 18.3

169.5 180.3 220.3 169.0 128.5 136.3 200.0 166.0 188.3 137.0 168.5 178.8 186.5 210.0 214.8 169.2

31.0 29.5 30.0 29.0 32.5 31.0 29.0 28.0 30.0 31.0 26.5 29.5 29.0 32.0 30.8 29.7

21.0 20.5 20.3 19.3 24.0 23.5 17.8 18.8 18.0 23.3 17.8 19.8 18.8 17.8 19.3 20.2 0.2 0.6 0.8

17

Jatoi et al. (2012)

Table 4. Characterization of wheat parents under drought stress and non-stress conditions. Productive tillers plant-1 Grains spike-1 Grain yield plant-1 RD* RD* Parents NonStress at Stress at Stress at Non-stress Non-stress stress anthesis anthesis anthesis TD-1 7.8 6.5 1.3 66.8 50.5 16.3 26.3 23.5 Kiran 7.5 6.3 1.2 66.0 45.5 20.5 25.5 16.5 Sarsabz 6.8 5.8 1.0 61.8 50.8 11.0 23.0 20.5 Moomal 6.8 4.8 2.0 60.0 40.8 19.2 24.0 15.8 SKD-1 6.8 5.3 1.5 67.5 48.5 19.0 23.5 20.8 T.J-83 7.5 5.5 2.0 61.0 43.5 17.5 24.5 16.5 Relative water content Stomatal conductance Leaf area (cm2) (%)1 (mmol m-2 s-1) RD* Parents R D* NonStress at Stress at Stress at Non-stress Non-stress stress anthesis anthesis anthesis TD-1 89.3 44.8 44.5 473.3 139.0 334.3 30.3 24.3 Kiran 87.8 38.0 49.8 448.5 142.8 305.7 31.0 16.3 Sarsabz 88.8 44.8 44.0 419.3 158.5 260.8 28.3 22.3 Moomal 87.3 37.8 49.5 400.8 159.8 241.0 30.5 20.3 SKD-1 88.5 41.8 46.7 414.5 140.3 274.2 29.5 23.3 T.J-83 84.5 38.5 46.0 433.5 158.8 274.7 26.3 18.3 RD* = Relative difference in drought stress and non-stress conditions.

RD* 2.8 9.0 2.5 8.2 2.7 8.0 RD* 6.0 14.7 6.0 10.2 6.2 8.0

18


Table 5. GCA effects of parents for morph-yield and physiological traits of wheat under non-stress and stress conditions. Parents / F1 hybrids

Productive tillers plant-1 Non-stress

TD-1 Kiran Sarsabz Moomal SKD-1 T.J-83 S.E (gi.) S.E (gi-gj)

0.27 0.20 0.33 -0.22 -0.57 -0.01 0.17 0.26

Stress at anthesis 0.29 0.22 0.22 -0.42 -0.33 0.01 0.22 0.34

Grains spike-1 Nonstress 0.88 1.13 0.29 -1.74 0.29 -0.86 0.19 0.29

Stress at anthesis 1.65 0.81 2.78 -2.87 -0.59 -1.78 0.31 0.48

Grain yield plant-1 Nonstress 0.64 1.20 0.74 -1.04 -0.85 -0.51 0.19 0.30

Stress at anthesis 1.44 0.29 1.66 -2.14 -0.17 -1.08 0.22 0.34

RWC (%) Nonstress 1.29 0.17 0.06 -0.55 0.06 -1.04 0.24 0.37

Stress at anthesis 1.68 -0.49 1.43 -1.31 -0.21 -1.09 0.26 0.41

Stomatal conductance (mmol m-2 s-1 ) Non-stress 16.04 -0.36 3.16 -12.83 -4.36 -1.64 0.83 1.28

Stress at anthesis -5.66 -4.41 -9.13 11.55 0.55 7.11 1.42 2.20

Leaf area (cm2) Nonstress 0.54 0.35 -0.52 0.47 -0.74 -0.11 0.16 0.25

Stress at anthesis 1.53 -1.21 0.96 -0.40 -0.12 -0.75 0.22 0.34

S.E (gi.) = Determines the significance of GCA effect if it is greater than S.E (gi) value. S.E (gi-gj) = Determines the significance if the difference between two GCA effects is greater than S.E (gi-gj) value.

19

Jatoi et al. (2012)

Table 6. SCA effects of F1 hybrids for yield and physiological traits of wheat under non-stress (NS) and drought stress (DS) conditions.

TD-1 × Kiran

NS 0.15

DS -0.33

NS -0.89

DS -0.20

NS -0.82

DS -0.83

NS 0.76

DS 0.01

Stomatal conductance (mmol m-2 s-1 ) NS DS -1.17 10.00

TD-1 × Sarsabz

-0.97

-1.58

-1.05

-3.11

-0.54

-1.71

0.10

-1.40

0.05

TD-1 × Moomal

-1.41

-1.42

-5.67

-6.95

-4.01

-5.65

-0.51

-0.17

TD-1 × SKD-1

0.68

0.23

0.29

1.77

0.80

0.13

-0.74

0.73

TD-1 × T.J-83

3.37

1.89

9.11

10.46

4.96

5.04

3.61

Kiran × Sarsabz

2.09

0.73

3.70

4.17

1.83

2.45

-0.08

Kiran × Moomal

0.65

0.39

1.58

2.33

0.37

2.26

0.31

Productive tillers plant-1

F1 hybrids

Grains spike-1

Grain yield plant-1

RWC (%)

Leaf area (cm2) NS 0.37

DS 0.52

24.22

-0.26

-2.20

4.55

44.78

-0.76

-1.08

9.08

4.53

-0.54

-2.33

3.61

9.62

-42.53

2.33

3.05

2.05

18.96

-19.78

1.43

3.55

1.98

-9.04

24.53

-1.57

-0.83

Kiran × SKD-1

-0.25

-0.71

0.04

2.05

1.68

1.29

0.22

1.88

-2.51

1.53

-1.35

-0.08

Kiran × T.J-83

-0.57

-1.05

-0.14

0.49

0.33

1.20

0.53

1.76

-0.23

17.22

0.02

-0.20

Sarsabz × Moomal

2.53

0.64

5.42

7.92

3.40

3.38

0.03

1.82

28.93

-35.00

1.30

2.46

Sarsabz × SKD-1

0.62

0.29

0.14

-3.36

0.96

-1.08

-0.06

-1.52

13.96

7.50

-1.98

-3.29

Sarsabz × T.J-83

-0.19

-0.05

1.20

0.33

1.62

0.82

0.58

0.85

16.74

11.19

0.40

-0.67

1.70

29.96

6.06

-0.48

-0.92

Moomal × SKD-1

0.43

-0.05

3.01

2.80

1.24

-0.02

0.75

Moomal × T.J-83

-0.38

-0.14

1.08

-2.76

-2.85

-1.87

-0.74

-0.42

-2.76

23.00

1.90

-1.29

SKD-1 × T.J-83

-1.04

-0.99

-4.96

-6.29

-4.79

-4.33

-0.36

-1.52

7.52

38.75

1.87

-0.04

S.E. (si.)

0.46

0.59

0.51

0.86

0.52

0.61

0.65

0.72

2.27

3.89

0.45

0.60

0.52

0.67

0.57

0.97

0.59

0.69

0.73

0.81

2.56

4.39

0.51

0.68

S.E (sij-sjk) +

0.96** 0.92** 0.94** 0.89** 0.94** 0.86** 0.76** 0.62** 0.64** 0.95** 0.90** Correlation (r ) ** = Significant at 1% probability level.+ = Correlation coefficient (r) between mean performance of F1 hybrids per se and SCA effects. S.E (si) = Determines the significance of SCA effect if it is greater than S.E (gi) value. S.E (gi-gj) = Determines the significance if the difference between two SCA effects is greater than S.E. (sij-sjk) value.

0.81**

20

Number of grains spike -1 70

60

0

TD-1 x Kiran TD-1 x Sarsabz TD-1 x Moomal TD-1 x SKD-1 TD-1 x T.J-83 Kiran x Sarsabz Kiran x Moomal Kiran x SKD-1 Kiran x T.J-83 Sarsabz x Moomal Sarsabz x SKD-1 Sarsabz x T.J-83 Moomal x SKD-1 Moomal x T.J-83 SKD-1 x T.J-83

8 7 6 5 4 3 2 1 0

Moomal x SKD-1 Moomal x T.J-83 SKD-1 x T.J-83

Sarsabz x Moomal Sarsabz x SKD-1 Sarsabz x T.J-83

Kiran x Moomal Kiran x SKD-1 Kiran x T.J-83

TD-1 x SKD-1 TD-1 x T.J-83 Kiran x Sarsabz

TD-1 x Kiran TD-1 x Sarsabz TD-1 x Moomal

Productive tillers plant -1


Hybrid per se

SCA effects

Hybrid per se SCA effects

50

40

30

20

10

2.5 2 1.5 1 0.5 0 -0.5 -1 -1.5 -2

Figure 1. Relationship between hybrid performances per se and SCA effects for productive tillers plant-1.

12 10 8 6 4 2 0 -2 -4 -6 -8

Figure 2. Relationship between hybrid performances per se and SCA effects for grains spike-1.

21

44 2

42 1

40 0

38 -1

36 -2

SKD-1 x T.J-83

Moomal x T.J-83

Moomal x SKD-1

Sarsabz x T.J-83

Sarsabz x SKD-1

Sarsabz x Moomal

Kiran x T.J-83

Kiran x SKD-1

Kiran x Moomal

Kiran x Sarsabz

Hybrid per se

SCA effects

25

20

10

SKD-1 x T.J-83

Moomal x T.J-83

Moomal x SKD-1

Sarsabz x T.J-83

46

Sarsabz x SKD-1

48

Sarsabz x Moomal

5 -6

0 -8

TD-1 x T.J-83

TD-1 x SKD-1

TD-1 x Moomal

TD-1 x Sarsabz

TD-1 x Kiran

Grain yield plant -1 30

Kiran x T.J-83

Kiran x SKD-1

Kiran x Moomal

Kiran x Sarsabz

TD-1 x T.J-83

TD-1 x SKD-1

TD-1 x Moomal

TD-1 x Sarsabz

TD-1 x Kiran

Relative water content (%)

Jatoi et al. (2012)

6

4

2

15 0

-2

-4

Figure 3. Relationship between hybrid performances per se and SCA effects for grain yield plant-1.

Hybrid per se 4

SCA effects 3

Figure 4. Relationship between hybrid performances per se and SCA effects for relative water content.

22


In stress conditions, the GCA and SCA variances for grains spike-1 were greater than in nonstress conditions indicating that both types of genes either additive or non-additive were associated in stress response (Table 2). The GCA effects of parents such as, Kiran, TD-1, Sarsabz were higher in non-stress conditions whereas in stress conditions, these parents changed their rank order with Sarsabz being 1st, TD-1 2nd and Kiran at 3rd positions (Table 5). These results demonstrated that genotypes Kiran, TD-1 and Sarsabz were less affected by stress conditions, hence were more drought tolerant and good general combiners for producing more grains spike-1. It may be further inferred that, selection from segregating populations in early generations could effectively improve the grains spike-1 in stress conditions. Farooqi et al. (2006) observed that GCA was greater than SCA for seeds spike-1 suggesting preponderance of additive genes for this trait. The F1 hybrids TD-1 × TJ-83, Sarsabz × Moomal and Kiran × Sarsabz recorded higher SCA effects in non-stress and in stress conditions also indicating their SCA for grains spike-1 (Table 6). The SCA effects of these particular hybrids highly reflected in per se hybrid performance (Figure 2) which indicated more reliability that if hybrid crop becomes a choice in water stress environment, then above three hybrids will perform better to increase the grains spike-1, consequently grain yield plant-1. The variability of GCA and SCA was much higher in stress against in non-stress environment,

yet GCA was predominant in stress for grain yield plant-1 suggesting that genes either additive or nonadditive contributing towards grain yield were less suppressed by the water stress conditions (Table 2). Similar to present results, AlNaggar et al. (2007) noted the significance of both GCA and SCA under non-stress and in water stress conditions for grain yield. The GCA effects of parents were greater in stress where maximum effects were expressed by Sarsabz and TD-1 being good general combiners with additive genes advocating grain yield plant-1. The higher GCA effects of these two parents may prove superior to improve this trait via hybridization and selection from early filial generations. Since majority of the hybrids displayed positive SCA effects, yet F1 crosses TD-1 × TJ83, Kiran × Sarsabz and Sarsabz × Moomal exhibited higher SCA effects in stress environment (Table 6) suggesting that these hybrid combinations may be used for hybrid crop development for increasing grain yield in stress environment. The potentiality of these F1 hybrids is more reliable in the sense that, there is close association between per se hybrids’ performance and their SCA effects (Figure 3). Similar to our findings, Betran et al. (2003) reported that additive genes were more important under drought conditions, while dominant effects were more important in non-stress environment. They also noted that the importance of additive effects increased by increasing the intensity of drought stress. This suggested the need of drought tolerance in both parental lines

23

Jatoi et al. (2012)

involved in crosses so as to achieve acceptable level of hybrid performance under severe drought conditions. Drikvand et al. (2005), Kamaluddin et al. (2007) and Dhadhal et al. (2008) also found that GCA effects were significant for seeds spike-1 and effective tillers plant-1. They reported that crosses displaying higher SCA effects for yield were derived high × high, high × low, low × low and medium × low GCA parents. Contrary to our findings, Iqbal and Khani (2006), Dere and Yildirim (2006) and Iqbal et al. (2007) reported that although GCA and SCA variances were significant for grain yield, SCA was predominant and also found several parents were good general or specific combiners for grain yield. Very little work has been previously done on genetic analysis of physiological traits, yet it is important for plant breeders to know the gene action for such traits so that effective breeding strategies can be planned while selecting plants in drought conditions. The GCA mean squares in both nonstress and specifically in stress conditions, relative water content in leaf (RWC %) was greater than SCA suggesting that this trait was primarily controlled by additive genes which could make the selection more effective under stress conditions and consequently for drought tolerance (Table 2). From six parents, the top three were; TD-1, Kiran and Sarsabz which recorded higher and positive GCA effects for RWC % in nonstress and in stress conditions as well indicating their good GCA, hence being potential parents for hybridization and selection

programmes to improve RWC % under drought environment (Table 5). Similar to our results, Farshadfar et al. (2000) studied drought tolerance in 8 × 8 diallel crosses and found that additive gene action was predominant for RWC % and the parent Plainsman was the best general combiner, whereas hybrid Plainsman × Kobomugi was the best specific combination under drought conditions. The F1 hybrid TD-1 × TJ-83 gave maximum SCA effects in non-stress and in stress conditions suggesting that this particular hybrid is suitable for hybrid crop development in stress conditions (Table 6 and Figure 4). The GCA and SCA variability for stomatal conductance was significant under both stress and non-stress conditions suggesting that additive and non-additive genes were controlling the trait. Among the parents, TD-1, Kiran and Sarsabz manifested negative GCA effects in stress implying that these parents gave less stomatal conductance, being drought tolerant, hence were more suitable for hybridization and selection programmes for drought stress conditions (Table 5). In F1 hybrids, crosses as TD-1 × TJ-83, Sarsabz × Moomal and Kiran × Sarsabz recorded negative SCA effects for stomatal conductance in stress conditions being more suitable crosses for drought conditions (Table 6). The parents involved in the above crosses, actually gave less stomatal conductance as per se in water stress conditions and likewise in the hybrid combinations (Figure 6). Saeed et al. (2001) also reported that hybrid Kohistan-97 × MH-97

24


exhibited negative SCA effects for stomata size. In stress conditions, though both GCA and SCA variances were greater than in nonstress for flag leaf area, yet GCA was about twice higher than SCA suggesting that leaf area was primarily under the control of additive genes. The parents TD-1 and Sarsabz gave higher GCA effects in stress, revealing good GCA, thus being potential parents for breeding drought tolerant genotypes (Table 5). The importance of additive genes in present studies are in consonance with those reported by Ayoub and Choudhry (2000) and Farooqi et al. (2006) who also found that additive genes were more predominant than non-additive genes under drought conditions. The maximum SCA effects for flag leaf area were recorded in F1 hybrids Kiran × Sarsabz, TD-1 × TJ-83 and Sarsabz × Moomal in stress at anthesis, thus this set of three hybrids were good specific combiners (Table 6 and Figure 6). This may be attributable to wider flag leaf area, thus permitting greater photosynthetic activity under drought conditions. Overall, parents Kiran, Sarsabz, TD-1 and TJ-83 were found to be the best general combiners for most of physiological and yield traits in non-stress conditions and therefore, these parents could be useful for hybridization and selection programmes for developing drought tolerant wheat genotypes. The F1 hybrid TD-1 × TJ-83 was the best specific combiner for most of the traits in non-stress as well as in stress conditions whereas for specific traits, Kiran × Sarsabz was

best for leaf area; hence these hybrids may be potential to be grown in drought prone environments.

CONCLUSION The physiological and yield traits of wheat were significantly affected by the water stress treatments and the significance of treatment × genotype interactions indicated that cultivars performed variably across irrigation treatments. The wheat cultivars such as TD-1, Sarsabz and SKD-1 performed well under stress conditions for productive tillers plant-1, grains spike-1, grain yield plant-1, relative water content, stomatal conductance and leaf area. Generally, in stress conditions, the GCA variance was predominant for most of the traits as compared to SCA variances. The higher GCA effects of the parents, especially in water stress conditions revealed that TD-1, Sarsabz and SKD-1 possessed additive genes, hence were good general combiners, and are potential parents to be used in breeding programmes for the development of drought tolerant lines. The hybrid performance per se was reflected in their SCA effects indicated their consistency in the performance. The F1 hybrids TD-1 × TJ-83, Kiran × Sarsabz and Sarsabz × Moomal proved to be good specific combiners for most of the traits, thus offering potential parents for hybrid crop development in drought conditions.

25

Jatoi et al. (2012)

REFERENCES Al-Naggar AM, Moustafa MA, Atta MM, Shehab-Eldeen MT (2007). Gene action of earliness and grain filling in bread wheat under two irrigation regimes. Egypt. J. Plant Breed. 11: 279297. Ayoub GMA, Choudhary MA (2000). Genetic studies in bread wheat under irrigated and drought stress conditions. Pak. J. Biol. Sci. 3: 1793-1798. Betrán FJ, Beck D, Bänziger M, Edmeades GO (2003). Genetic analysis of inbred and hybrid grain yield under stress and non-stress environments in tropical maize. Crop Sci. 43: 807817. Brancourt BH, Doussinault G, Lecomte L (2003). Genetic improvement of agronomic traits in winter wheat cultivars released in France. Crop Sci. 43: 3745. Cooper M, Byth DE, Woodruff DR (1994). An investigation of the grain yield adaptation of advanced CIMMYT wheat lines to water stress environments in Queensland. II. Classification Analysis. Aust. J. Agric. Res. 45: 985-1002. Day AD, Intlap S (1970). Some effects of soil moisture stress on the growth of wheat (T. aestivum L.). Agron. J. 62: 27-29. Dere S, Yildirim MB (2006). Inheritance of grain yield

per plant, flag leaf width, and length in 8 × 8 diallel cross population of bread wheat (T. aestivum L.). Turk. J. Agric. For. 30: 339-345. Dhadhal BA, Dobariya KL, Ponkia HP, Jivani LL (2008). Gene action and combining ability over environments for grain yield and its attributes in bread wheat (T. aestivum L.). Intl. J. Agric. Sci. 4: 66-72. Drikvand R, Farshadfar E, Nazarian F (2005). Genetic study of some morphophysiological traits in bread wheat lines under dryland conditions using diallel crossing. Seed and Plant 20: 429-444. Farooqi J, Habib I, Saeed A, Nawab NN, Khaliq I, Abbas G (2006). Combining ability for yield and its components in bread wheat (T. aestivum L.). Intl. J. Agric. Biol. 2: 207-211. Farshadfar E, Farshadfar M, Sutka J (2000). Combining ability analysis of drought tolerance of wheat over different water regimes. Acta Agron. Hungarica 48: 353-361. Gomez KA, Gomez AA (1984). Statistical Procedure for Agriculture Research. John Wiley and Sons Inc.2nd Ed, New York, USA. Griffing B (1956). Concept of general and specific combining ability in relation to diallel crossing system. Aust. J. Biol. Sci. 9: 463-493.

26


Iqbal

M, Khani AA (2006). Analysis of combining ability for spike characteristics of wheat (T. aestivum L.). Intl. J. Agric. Biol. 8 (5): 684-687. Iqbal M, Alireza N, Salmon DF (2007). Genetic analysis of flowering and maturity time in high latitude spring wheat: Genetic analysis of earliness in spring wheat. Euphytica 154 (1&2): 207218. Kamaluddin R, Singh M, Prasad LC, Abdin MZ, Joshi AK (2007). Combining ability analysis for grain filling duration and yield traits in spring wheat (T. aestivum L.). Genet. Mol. Biol. 30 (2): 411-416. Munir A, Akram Z, Munir M, Rauf M (2006). Physiomorphic response of wheat genotypes under rainfed conditions. Pak. J. Bot. 38(5): 1697-1702. Nisar A, Chowdhry MA, Khaliq I, Maekawa M (2007). The inheritance of yield and yield components of five wheat hybrid populations under drought conditions. J. Agric. Sci. 8 (2): 53-59. Piepho HP (2000). A mixturemodel approach to mapping quantitative trait loci in barley on the basis of multiple environment data. Genetics 156: 20432050. Rathore PS (2005). Techniques and Management of Field Crop Production. Agribios, India, pp. 525. Saeed A, Chowdhry MA, Saeed N, Khaliq I, Johar MZ (2001).

Line × Tester analysis for some morphophysiological traits in bread wheat. Intl. J. Agric. Biol. 3 (4): 444-447. Schonfeld MA, Johnson RC, Carver BF, Momhinweg DW (1988). Water relations in winter wheat as drought resistance indicators. Crop Sci. 28: 526-531. Singh RK, Chaudhary BD (1985). Biometrical Methods in Quantitative Genetic Analysis. (3rd Ed.), Kalyani Publishers, New Delhi, India.

27

RESEARCH ARTICLE


SCREENING OF SUNFLOWER GENOTYPES FOR DROUGHT TOLERANCE UNDER LABORATORY CONDITIONS USING PEG A. GEETHA1*, A. SIVASANKAR1, LAKSHMI PRAYAGA2, J. SURESH3 and P. SAIDAIAH4 1

Department of Plant Physiology, College of Agriculture, ANGR Agricultural University, Hyderabad – 500030, AP, India. 2 Division of Plant Physiology, Directorate of Oilseeds Research (DOR), Hyderabad – 500030, AP, India. 3 Department of Genetics and Plant Breeding, College of Agriculture, ANGR Agricultural University, Hyderabad – 500030, AP, India. 4 Department of Genetics and Plant Breeding, College of Horticulture, Andhra Pradesh Horticultural University, Mojerla – 509382, AP, India. *Corresponding author email: [email protected]

SUMMARY Fifty genotypes of sunflower (Helianthus annuus L.) were screened for seedling traits such as germination percentage, root length, shoot length, root to shoot length ratio and seedling dry weight for drought tolerance under laboratory conditions using PEG-6000 during 2008. Two levels of osmotic stress (-0.3 MPa and -0.9 MPa) were created and performances were monitored against a control. The results indicated that genotypes have wide genetic variability for all traits studied. PEG -6000 was successfully used in screening the seedlings for moisture stress at seedling establishment stage. Germination percentage, root length, shoot length, root/ shoot length ratio and seedling dry weight decreased with increasing osmotic stress. We propose the use of PEG -6000 for screening of sunflower genotypes for drought tolerance under lab conditions. In view of difficulties like uncontrolled climatic conditions, heterogeneity of soil, large amount of plant material to and time and labor consumption making field trials difficult for drought screening of genotypes, PEG-based screening of genotypes for drought tolerance may be useful as a complementary method to field screening. Key words: Drought tolerance, laboratory conditions, PEG, screening, sunflower Manuscript received: August 23, 2011; Decision on manuscript: March 3, 2012; Manuscript accepted in revised form: March 30, 2012. Communicating Editor: Bertrand Collard

28

Geetha et al. (2012)

INTRODUCTION Adequate water and nutrient supply are important factors affecting optimal plant growth and successful crop production. Water stress is one of the severe limitations of crop growth especially in arid and semi-arid regions of the world as it has a vital role in plant growth and development at all growth stages. However, depending upon plant species, certain stages such as germination and seedling or flowering could be the most critical stages for water stress. Seed germination is the first critical and the most sensitive stage in the life cycle of plants (Ashraf and Mehmood, 1990) and unfavorable environmental conditions like water stress may compromise the seedlings establishment (Albuquerque and Carvalho, 2003). Sunflower is an oil seed crop and is particularly susceptible to water shortage at germination stage. Sajjan et al. (1999) reported decrease in percentage germination and biomass accumulation in sunflower with increasing osmotic stress in germinating media, whereas mean germination time increased with increasing water deficit (El-Midaoui et al., 2001). Moreover, different genotypes of sunflower showed differential responses to all these stress treatments. Lenzi et al. (1995) reported that mutant seeds of sunflower showed a higher resistance to osmotic stress (i.e., germinating at more negative osmotic potentials). However, difficulties like uncontrolled climatic conditions,

insufficient homogeneity of soil, large amount of plant material and time and labor consumption make field trials difficult for drought screening of genotypes. Earlier research reports in favour of drought screening under laboratory conditions using drought inducing chemicals like PEG are gaining recently. Polyethylene glycols or PEGs are a group of neutral osmotically active polymers with a certain molecular weight. PEG 6000 (the number signifying molecular mass) is most frequently used in plant water deficit studies to induce dehydration by decreasing the water potential of the nutrient solution. Germination of sunflower was inhibited in presence of polyethylene glycol 6000, at osmotic pressure lower than -5 bars (Smok et al., 1993). Singh and Singh (1983) reported that decline in seedling vigour in sunflower under osmotic stress was due to fall in mobilization of reserves to plumule thus preventing their growth under stress. Sadasivam et al. (2000) screened 15 rice genotypes based on seed germination, seedling growth (root and shoot) under laboratory conditions using PEG. Osmotic stress was induced at four levels (0, -0.5,-0.75 and -1.00 MPa). Results indicated genotypes showed wide variation for germination, root growth and seedling growth. Manjula et al. (2003) screened eight castor varieties based on seed germination and seedling vigour under laboratory conditions using PEG. Osmotic stress was created at three levels (-3,-6 and -9 bars). Results revealed that stress level created at -6 bars was found to be

29


appropriate for screening of the genotypes. Ahmed et al. (2009) conducted a laboratory experiment to screen six sunflower genotype to drought stress based on seedling traits. Five water stress levels of zero (control), -0.35, -0.6, -1.33, and - 1.62 MPa were developed using polyethyleneglycol-6000 (PEG-6000). Results indicated germination percentage, dry weight and shoot growth declined with increase in water stress treatment while in contrast root length increased with water stress treatment. Rauf and Sadaqat (2008) studied response of roots to the drought in sunflower inbred lines and hybrids .The results indicated that higher the root growth better is drought tolerance. They also reported increase in root length in sunflower as result of drought stress is due to higher osmotic adjustment ability of drought tolerant genotypes. The adverse effects of water shortage on germination and seedling growth had also been well reported in different crops such as wheat (Dhanda et al., 2004), sugar beet (Sadeghian and Yavari, 2004), sorghum (Gill et al., 2002), and sunflower (Mohammad et al., 2002). In view of the above facts, a study was conducted to screen the sunflower genotypes for drought tolerance under laboratory conditions using PEG.

MATERIAL AND METHODS The experiment was carried out at Plant Physiology Laboratory, College of Agriculture, ANGRAU, Rajendranagar, Hyderabad, India during 2008. Fifty sunflower

(Helianthus annuus L.) genotypes including open pollinated varieties and hybrids widely grown in the region and inbred lines maintained at Directorate of Oilseeds Research, Rajendranagar, Hyderabad, Andhra Pradesh were colleted and tested against drought stress at germination and seedling stages under laboratory conditions (25±3○C) during November, 2008. The genotype SH491, a short duration cultivar with good centre filling, non lodging, compact and flat flower, high yielding and one of the drought tolerant cultivars widely cultivated under rainfed conditions in Andhrapradesh and several other sunflower growing states of India was used as drought tolerant check in this study. Polyethylene Glycol with a molecular weight of 6000 (PEG-6000, The Panchi Chemicals, Hyderabad, India) was used as a drought simulator and three water stress levels of zero (control), -0.3 MPa and -0.9 MPa were developed by dissolving 0, 8.6 and 28.9 g of PEG per 100ml distilled water (Michael and Kaufmann,1973). Seeds were surface sterilized with 10% sodium hypochlorite solution for five minutes and then washed three times with distilled water. Fifteen seeds of each sunflower genotype were planted in each petriplate containing filter paper. The experiment was laid out in a completely randomized design (CRD) with three replications for each experimental unit. Ten milliliters of the treatment solution was applied daily in each petri plate after washing out the previous solution. The number of seeds germinated was counted and germination percentage was calculated. A seed was considered germinated when both plumule and radicle had emerged to 5 mm. Root

30


length, shoot length, root to shoot ratio and dry weights were recorded at 14th day of the start of the experiment. Plant dry weights were recorded after drying at 70○C. From these measurements, the germination percentage, maximum root length, shoot length, root to shoot length ratio and dry weights were recorded. The data so collected was analyzed using analysis of variance (ANOVA) technique and WINDOWSTAT 7.5 (Indostat services, Hyderabad, India) software package was used for this purpose.

RESULTS AND DISCUSSION Polyethylene glycol (PEG) causes osmotic stress and could be used as a drought simulator (Ashraf et al., 1996; Turhan, 1997). In the present experiment, PEG-6000 was used to create the osmotic stress, as most of the researchers (Smok et al., 1993; Hu and Jones, 2004) utilized it for the development of water deficit in growth chamber studies. The sunflower genotypes differed in germination (Table 1) at three different levels of stress using PEG-6000 solutions of different osmotic potentials (0, -0.3 and -0.9 MPa). In the control (0 MPa), out of 50 lines tested, twenty five genotypes showed maximum of 90% seed germination. The seed germination decreased under osmotic stress. The average germination percentage of genotypes was 62.7 and 46.5 in 0.3 and -0.9 MPa treatments where as, it is 83.7% in control treatment. Genotypes showed significant variation for germination under PEG induced water stress. At -0.3 MPa stress (slight increase in the stress) resulted in more than 80%

seed germination in twelve genotypes, 60-80% germination in nineteen genotypes, 40-60% germination in eleven genotypes and less than 40% germination in eight genotypes. ASF-104 and GP4-187 showed superior germination of 100 and 95% respectively. Five genotypes NDR4, M-1024, SVG, GP-1284 and RHA-341 exhibited lower (2515%) germination. When maximum stress was induced (-0.9 MPa) only one genotype M-1013 recorded 80% germination, while others showed 60-80% germination in seventeen genotypes, 40-60% germination in twelve genotypes and less than 40% germination in twenty one genotypes. Substantial reduction in germination in different genotypes was recorded at -0.9 MPa. Overall rate of germination of all genotypes was consistently and significantly reduced whenever osmotic potential was reduced except in case SH-177 and DK3849, which showed only marginal decrease in germination percentage, while the check SH491 had 22-30% reduction compared to control treatments. Twenty nine genotypes at -0.3 MPa and 24 genotypes at -0.9 MPa showed better germination than check SH-491 used in the study. In general, germination percentage decreased with the increase in PEG-6000 concentration in all the genotypes. Such decline in the seed germination may be due to the inability of seed to imbibe water at low osmotic potential (Karan Singh and Africa, 1985). It has been reported that speed of germination and emergence increase with

31


increase in matrix potential (water held in micro capillaries or bound on surface of cell wall and other cell components) as long as oxygen does not become limiting (Hobbs et al., 1994). Decline in germination percentage under osmotic stress was reported earlier (Shantha Nagarajan and Jagadish Rane, 2000) in spring wheat (Sadasivam et al., 2000) in rice and (Manjula et al., 2003) in castor. The germination test is useful for identifying vigorous seed lots and genotypes capable of quickly establishing adequate population under low soil moisture conditions. Present study strongly support that germination % can be utilized to screen sunflower germplasm for osmotic stress under laboratory conditions. Mean root lengths (Table 2) were 3.3 cm, 2.9 cm and 2.2 cm under 0, -0.3 and -0.9 MPa of moisture stresses respectively. Reduction in the osmotic potential from 0 to -0.9 MPa led to a significant decrease in the root length in all the genotypes examined. In control, the difference in root length ranged from 0.32 (M-1019) to 8.01 cm (ASF-104), in -0.3 MPa, it ranged from 0.04 (M-3072) to 7.9cm (DSF-104). The maximum root length was reported in DSF-104 (7.9 cm) and minimum in M-3072 (0.04 cm), whereas at-0.9 MPa, it ranged from 0.2 to 6.7cm in NDR4 and SH-177, respectively. While check SH-491 reported 2.74 and 2.36 cm root lengths respectively. Repressing effect of PEG induced stress was observed on root length. With increase in the osmotic stress, the root length decreased in most of the genotypes but some of the

genotypes recorded increase in root length. Larger root systems of seedling may be useful in maintaining water availability under limited moisture supply. Increase in root length may be due to the diversion of dry matter to the root in search of moisture (Nicholas et al., 1995). Increase in root length as a result of drought stress has also been observed due to higher osmotic adjustment ability of drought tolerant genotypes (Rauf and Sadaqat, 2007; Rauf and Sadaqat, 2008). Chun et al. (2005) reported an increase in root length occurred at the expense of lateral root number.

32


Table 1. Effect of PEG induced stress (-0.3 MPa and-0.9 MPa) on seed germination. Control Genotype GP4-2704 GP-2793 GP2-1746 GP4-2935 M-1013 EC-512690 NBR-318 PS-4091 M-1029 GP-978 M-3072 RHA-274 GP-2035 NDR-7 GP4-2885 GP9-515-7-3 GP-247-4 RHA-6D1 M-1019 NDR-4 M-1024 GP4-2605 SVG GP4-1399 GP4-187 GP4-193 GP4-1018 GP-812-5 GP-776-2-B GP-69 GP9-846-4-4 GP9-38-C-2-1 GP-1080-2 RHA-341 GP-1284 LRHA RHA-340 DK-3849 RSF-101 TSF-103 ASF-107 DSF-114 SH-177 DSF-104 RSF-106 DSF-111 RSF-107 ASF-104 TSF-106 SH-491 Average CD (P=0.05)

0 MPa 100.00 93.33 100.00 93.33 90.00 66.66 73.30 90.00 75.00 73.33 83.33 65.00 83.33 88.33 88.33 75.00 83.33 73.33 70.00 45.00 40.00 93.33 50.00 100.00 95.00 90.00 81.67 78.33 73.33 100.00 90.00 93.33 50.00 50.00 65.00 80.00 86.67 100.00 100.00 80.00 100.00 98.33 100.00 100.00 93.33 100.00 90.00 100.00 96.67 100.00 83.70 3.21

Seed germination (%) Stress % decrease over control Mean -0.3 MPa -0.9 MPa -0.3 MPa -0.9 MPa 70.00 36.66 68.89 30.00 63.34 81.66 50.00 75.00 12.50 46.43 78.33 70.00 82.78 21.67 30.00 85.00 63.33 80.55 8.93 32.14 80.00 68.33 79.44 11.11 24.08 46.67 35.00 49.44 29.99 47.49 66.67 50.00 63.32 9.05 31.79 73.33 63.66 75.66 18.52 29.27 75.00 80.00 76.67 0.00 -6.67 45.00 26.33 48.22 38.63 64.09 35.00 25.00 47.78 58.00 70.00 45.00 35.00 48.33 30.77 46.15 30.00 33.33 48.89 64.00 60.00 66.67 63.33 72.78 24.52 28.30 70.00 51.33 69.89 20.75 41.89 46.67 40.00 53.89 37.77 46.67 53.33 38.66 58.44 36.00 53.61 50.00 25.00 49.44 31.82 65.91 43.00 25.33 46.11 38.57 63.81 25.00 16.66 28.89 44.44 62.98 23.33 26.66 30.00 41.68 33.35 58.33 53.33 68.33 37.50 42.86 23.33 15.00 29.44 53.34 70.00 60.00 46.66 68.89 40.00 53.34 95.00 51.67 80.56 0.00 45.61 75.00 30.00 65.00 16.67 66.67 30.00 18.33 43.33 63.27 77.56 78.33 30.00 62.22 0.00 61.70 56.67 38.33 56.11 22.72 47.73 81.67 56.33 79.33 18.33 43.67 70.00 50.00 70.00 22.22 44.44 76.67 66.67 78.89 17.85 28.57 40.00 18.33 36.11 20.00 63.34 15.00 20.00 28.33 70.00 60.00 21.67 26.66 37.78 66.66 58.98 55.00 36.33 57.11 31.25 54.59 68.33 35.00 63.33 21.16 59.62 86.67 75.00 87.22 13.33 25.00 86.67 65.00 83.89 13.33 35.00 65.00 60.00 68.33 18.75 25.00 78.66 71.67 83.44 21.34 28.33 83.33 43.33 75.00 15.25 55.93 91.67 75.00 88.89 8.33 25.00 68.33 53.33 73.89 31.67 46.67 75.00 70.00 79.44 19.64 25.00 83.33 60.00 81.11 16.67 40.00 86.67 61.67 79.45 3.70 31.48 100.00 60.00 86.67 0.00 40.00 60.00 41.67 66.11 37.93 56.89 76.67 70.00 82.22 23.33 30.00 62.73 46.46 64.30 4.22 4.87

33


Table 2. Effect of PEG induced stress (-0.3MPa and-0.9MPa) on root length (cm).

Genotype GP4-2704 GP-2793 GP2-1746 GP4-2935 M-1013 EC-512690 NBR-318 PS-4091 M-1029 GP-978 M-3072 RHA-274 GP-2035 NDR-7 GP4-2885 GP9-515-7-3 GP-247-4 RHA-6D1 M-1019 NDR-4 M-1024 GP4-2605 SVG GP4-1399 GP4-187 GP4-193 GP4-1018 GP-812-5 GP-776-2-B GP-69 GP9-846-4-4 GP9-38-C-2-1 GP-1080-2 RHA-341 GP-1284 LRHA RHA-340 DK-3849 RSF-101 TSF-103 ASF-107 DSF-114 SH-177 DSF-104 RSF-106 DSF-111 RSF-107 ASF-104 TSF-106 SH-491 Average CD(P=0.05)

Control 0M Pa 1.87 1.83 1.91 2.49 2.00 0.76 0.37 5.50 4.05 2.38 0.60 1.95 1.50 3.00 2.38 4.54 4.19 2.10 0.32 0.80 0.53 2.14 2.08 4.98 5.83 1.71 2.88 2.45 1.44 2.46 3.70 2.86 2.20 2.37 1.28 2.99 2.85 4.50 4.50 8.00 5.00 6.00 7.03 8.01 6.20 5.01 6.56 5.55 6.00 3.50 3.30 0.52

Root length (cm) Stress -0.3 MPa -0.9 MPa Mean 1.41 1.02 1.43 1.09 1.46 1.46 0.88 0.43 1.07 1.22 1.97 1.89 0.77 1.64 1.47 0.37 0.25 0.46 0.58 0.27 0.41 6.25 4.40 5.38 2.57 2.78 3.13 2.30 0.73 1.80 0.04 0.90 0.51 2.96 1.71 2.21 1.55 1.90 1.65 1.74 2.14 2.29 2.64 1.64 2.22 4.80 1.98 3.77 4.45 2.44 3.69 1.45 0.57 1.37 0.67 1.77 0.92 0.45 0.23 0.49 0.98 0.62 0.71 3.11 2.35 2.53 1.91 0.77 1.59 3.35 4.53 4.29 4.28 2.50 4.20 1.62 1.64 1.66 3.13 2.00 2.67 2.56 1.07 2.03 1.42 1.73 1.53 3.56 1.70 2.57 4.37 2.26 3.44 4.08 3.16 3.37 1.19 1.25 1.55 0.26 2.00 1.54 1.80 1.02 1.37 1.93 1.17 2.03 2.96 2.44 2.75 3.00 0.60 2.70 3.00 0.50 2.67 2.74 1.00 3.91 6.74 4.90 5.55 3.74 2.90 4.21 6.65 6.67 6.78 7.87 4.89 6.92 7.00 5.88 6.36 6.00 5.11 5.37 5.36 5.33 5.75 5.85 4.39 5.26 5.97 5.11 5.69 2.74 2.36 2.87 2.95 2.24 2.83 0.63 0.38

% decrease over control -0.3 MPa -0.9 MPa 24.60 45.45 40.44 20.22 53.93 77.49 51.00 20.88 61.50 18.00 51.32 67.11 -56.76 27.03 -13.64 20.00 36.54 31.36 3.36 69.33 93.33 -50.00 -51.79 12.31 -3.33 -26.67 42.00 28.67 -10.92 31.09 -5.73 56.39 -6.21 41.77 30.95 72.86 -109.38 -453.13 43.75 71.25 -84.91 -16.98 -45.33 -9.81 8.17 62.98 32.73 9.04 26.59 57.12 5.26 4.09 -8.68 30.56 -4.49 56.33 1.39 -20.14 -44.72 30.89 -18.11 38.92 -42.66 -10.49 45.91 43.18 89.03 15.61 -40.63 20.31 35.45 60.87 -3.86 14.39 33.33 86.67 33.33 88.89 65.75 87.50 -34.80 2.00 37.67 51.67 5.41 5.18 1.74 38.98 -12.90 5.16 -19.76 -2.00 18.23 18.64 -5.41 20.90 0.50 14.83 21.74 32.57

34


Table 3. Effect of PEG induced stress (-0.3MPa and-0.9MPa) on shoot length (cm).

Genotype GP4-2704 GP-2793 GP2-1746 GP4-2935 M-1013 EC-512690 NBR-318 PS-4091 M-1029 GP-978 M-3072 RHA-274 GP-2035 NDR-7 GP4-2885 GP9-515-7-3 GP-247-4 RHA-6D1 M-1019 NDR-4 M-1024 GP4-2605 SVG GP4-1399 GP4-187 GP4-193 GP4-1018 GP-812-5 GP-776-2-B GP-69 GP9-846-4-4 GP9-38-C-2-1 GP-1080-2 RHA-341 GP-1284 LRHA RHA-340 DK-3849 RSF-101 TSF-103 ASF-107 DSF-114 SH-177 DSF-104 RSF-106 DSF-111 RSF-107 ASF-104 TSF-106 SH-491 Average CD(P=0.05)

Control 0M Pa 6.12 6.56 6.45 6.98 4.96 4.65 5.68 5.66 6.89 3.54 6.54 6.21 6.57 6.00 5.32 4.57 7.12 4.56 4.38 4.87 6.00 6.00 7.21 6.02 5.87 6.89 4.65 5.98 5.76 4.67 5.99 9.65 8.97 4.85 6.00 4.63 6.00 8.50 6.75 8.73 5.56 7.50 9.25 7.56 7.21 5.81 9.23 9.57 7.14 3.89 6.31 0.59

Shoot length (cm) Stress -0.3 MPa -0.9 MPa Mean 4.24 2.22 4.19 4.75 2.66 4.66 4.63 2.64 4.57 4.02 2.03 4.34 3.86 1.87 3.56 3.36 0.75 2.92 3.65 2.78 4.04 3.56 1.76 3.66 5.01 2.99 4.96 3.46 3.96 3.65 4.68 2.64 4.62 4.32 2.31 4.28 4.36 2.30 4.41 4.09 1.73 3.94 4.32 0.96 3.53 2.57 0.21 2.45 3.45 1.43 4.00 2.71 0.35 2.54 2.53 2.17 3.03 3.87 0.66 3.13 4.15 2.37 4.17 4.52 2.26 4.26 5.36 3.47 5.35 4.17 6.28 5.49 4.67 2.13 4.22 5.04 1.90 4.61 2.80 0.91 2.79 4.88 2.24 4.37 3.91 2.02 3.90 4.82 3.93 4.47 4.14 1.13 3.75 6.03 2.99 6.22 5.98 2.99 5.98 3.32 0.51 2.89 4.67 1.67 4.11 3.30 0.29 2.74 4.67 0.99 3.89 7.89 6.67 7.69 6.00 2.99 5.25 4.20 1.78 4.91 4.75 4.11 4.81 4.88 2.46 4.95 6.50 3.88 6.54 6.30 5.88 6.58 6.33 5.91 6.49 5.54 5.65 5.67 4.70 2.54 5.49 9.17 6.88 8.54 7.08 3.87 6.03 2.50 1.06 2.48 4.59 2.60 0.25 0.47

% decrease over control -0.3 MPa -0.9 MPa 30.72 63.73 27.59 59.45 28.22 59.07 42.41 70.92 22.18 62.30 -15.27 83.87 35.74 -1.76 37.10 68.90 27.29 56.60 55.93 5.08 28.44 59.63 30.43 62.80 -12.02 64.99 31.83 71.17 -18.80 81.95 43.76 95.40 51.54 79.92 40.57 92.32 42.24 4.79 20.53 86.45 30.83 60.50 24.67 62.33 25.66 51.87 30.73 62.13 20.44 63.71 26.85 72.42 39.78 80.43 18.39 62.54 32.12 64.93 -24.63 -5.57 30.88 81.17 37.51 69.04 33.33 66.69 99.59 89.53 22.17 72.17 28.73 93.78 22.17 83.50 -43.42 -21.17 11.12 55.74 51.89 79.60 -57.50 -27.98 34.96 67.23 29.69 58.05 -23.07 8.95 -29.46 4.11 -12.56 37.18 49.11 72.49 4.13 28.10 0.89 45.82 35.74 72.74

35


Table 4. Effect of PEG induced stress (-0.3MPa and-0.9MPa) on root /shoot length ratio Genotype GP4-2704 GP-2793 GP2-1746 GP4-2935 M-1013 EC-512690 NBR-318 PS-4091 M-1029 GP-978 M-3072 RHA-274 GP-2035 NDR-7 GP4-2885 GP9-515-7-3 GP-247-4 RHA-6D1 M-1019 NDR-4 M-1024 GP4-2605 SVG GP4-1399 GP4-187 GP4-193 GP4-1018 GP-812-5 GP-776-2-B GP-69 GP9-846-4-4 GP9-38-C-2-1 GP-1080-2 RHA-341 GP-1284 LRHA RHA-340 DK-3849 RSF-101 TSF-103 ASF-107 DSF-114 SH-177 DSF-104 RSF-106 DSF-111 RSF-107 ASF-104 TSF-106 SH-491 Average CD(P=0.05)

Control 0M Pa 0.31 0.28 0.30 0.36 0.40 0.16 0.07 0.97 0.59 0.67 0.09 0.31 0.23 0.50 0.45 0.99 0.59 0.46 0.07 0.16 0.09 0.36 0.29 0.83 0.99 0.25 0.62 0.41 0.25 0.53 0.62 0.30 0.25 0.49 0.21 0.65 0.48 0.53 0.67 0.92 0.90 0.80 0.76 1.06 0.86 0.86 0.71 0.58 0.84 0.90 0.52 0.65

Root /shoot length ratio Stress -0.3 MPa -0.9 MPa Mean 0.33 0.46 0.37 0.23 0.55 0.35 0.19 0.16 0.22 0.30 0.97 0.54 0.20 0.88 0.49 0.11 0.33 0.20 0.16 0.10 0.11 1.76 2.50 1.74 0.51 0.93 0.68 0.66 0.18 0.51 0.01 0.34 0.15 0.69 0.74 0.58 0.36 0.83 0.47 0.43 1.24 0.72 0.61 1.71 0.92 1.87 9.43 4.10 1.29 1.71 1.19 0.54 1.63 0.87 0.26 0.82 0.38 0.12 0.35 0.21 0.24 0.26 0.20 0.69 1.04 0.69 0.36 0.22 0.29 0.80 0.72 0.78 0.92 1.17 1.03 0.32 0.86 0.48 1.12 2.20 1.31 0.52 0.48 0.47 0.36 0.86 0.49 0.74 0.43 0.57 1.06 2.00 1.23 0.68 1.06 0.68 0.20 0.42 0.29 0.08 3.94 1.50 0.39 0.61 0.40 0.58 4.06 1.76 0.63 2.46 1.19 0.38 0.09 0.33 0.50 0.17 0.44 0.65 0.56 0.71 1.42 1.19 1.17 0.77 1.18 0.92 1.02 1.72 1.17 1.25 0.83 1.05 1.11 0.99 0.99 1.08 0.90 0.95 1.14 2.10 1.32 0.64 0.64 0.62 0.84 1.32 1.00 1.10 2.23 1.41 0.64 1.25 0.81 0.38 0.47

% decrease over control -0.3 MPa -0.9 MPa -8.83 -50.37 17.74 -96.75 35.82 45.00 14.93 -172.04 50.53 -117.50 32.62 -103.95 -143.94 -49.10 -80.67 -157.27 12.73 -58.17 1.13 72.58 90.68 -271.59 -118.21 -135.74 -55.71 -261.83 14.91 -147.40 -36.60 -281.86 -88.00 -849.09 -119.18 -189.95 -16.18 -253.63 -262.48 -1016.45 29.22 -112.14 -167.33 -196.15 -92.91 -191.54 -23.52 23.08 2.89 12.80 7.72 -18.18 -29.51 -247.79 -80.49 -254.85 -28.04 -16.59 -45.27 -242.57 -40.21 17.88 -70.89 -224.36 -128.30 -256.84 18.86 -70.57 83.97 -705.67 -80.67 -186.30 9.44 -529.08 -33.44 -418.87 28.17 83.00 24.99 74.90 28.81 38.73 -57.66 -32.47 4.16 -47.47 -34.53 -126.06 -17.86 21.59 -28.52 -15.62 -25.60 -4.88 -60.67 -195.72 -9.94 -10.01 -0.39 -57.19 -21.77 -147.38 -22.56 -139.04

36


Table 5. Effect of PEG induced stress (-0.3MPa and-0.9MPa) on total plant dry matter (g). Genotype GP4-2704 GP-2793 GP2-1746 GP4-2935 M-1013 EC-512690 NBR-318 PS-4091 M-1029 GP-978 M-3072 RHA-274 GP-2035 NDR-7 GP4-2885 GP9-515-7-3 GP-247-4 RHA-6D1 M-1019 NDR-4 M-1024 GP4-2605 SVG GP4-1399 GP4-187 GP4-193 GP4-1018 GP-812-5 GP-776-2-B GP-69 GP9-846-4-4 GP9-38-C-2-1 GP-1080-2 RHA-341 GP-1284 LRHA RHA-340 DK-3849 RSF-101 TSF-103 ASF-107 DSF-114 SH-177 DSF-104 RSF-106 DSF-111 RSF-107 ASF-104 TSF-106 SH-491 Average CD (P=0.05)

Control 0 MPa 2.00 1.37 1.65 1.74 1.63 1.55 1.38 2.45 2.15 2.18 0.95 2.34 1.29 2.01 1.54 2.26 2.06 1.35 1.85 0.74 1.20 1.30 1.19 1.51 1.98 1.78 1.79 1.83 1.40 1.57 1.51 1.86 1.64 1.78 1.04 1.80 1.41 2.50 2.22 2.40 2.50 2.60 3.00 2.80 2.89 3.00 3.01 3.19 3.33 3.53 1.96 0.04

Total plant dry matter (g) Stress % decrease over control -0.3 MPa -0.9 MPa Mean -0.3 MPa -0.9 MPa 2.15 0.92 1.69 -7.50 17.50 1.38 0.69 1.15 -29.93 -3.65 1.60 0.34 1.20 -21.21 35.15 0.69 0.08 0.84 37.36 31.03 0.67 0.26 0.85 34.36 25.03 1.01 0.18 0.91 9.03 41.29 0.88 0.27 0.84 7.25 27.54 2.11 1.12 1.89 -2.45 -4.49 1.24 0.93 1.44 23.72 22.79 1.22 0.67 1.36 25.69 35.78 0.55 0.23 0.58 0.00 47.37 1.61 1.13 1.69 14.10 20.51 1.37 0.89 1.18 -37.21 -48.84 1.07 1.27 1.45 26.87 0.50 1.35 0.82 1.24 -13.64 -0.65 1.75 0.85 1.62 -39.38 3.54 1.47 0.79 1.44 9.22 26.21 1.31 0.47 1.04 -26.67 11.11 0.71 0.12 0.89 -30.59 -64.71 0.16 0.11 0.34 0.00 0.00 0.05 0.02 0.42 0.00 0.00 1.07 0.68 1.02 -13.08 -8.46 0.80 0.07 0.69 -0.84 32.77 1.20 0.17 0.96 -5.96 -5.96 2.11 1.10 1.73 -26.77 7.58 1.67 0.73 1.39 -16.29 17.98 1.49 0.90 1.39 -5.59 8.94 1.77 0.56 1.39 -73.22 29.51 1.03 0.21 0.88 -2.14 -6.43 1.11 0.67 1.12 -41.40 -33.76 1.33 0.52 1.12 -14.57 17.22 1.42 0.80 1.36 -18.82 17.74 1.45 0.68 1.26 -12.80 14.02 0.52 0.09 0.80 78.09 38.76 0.87 0.12 0.68 -69.23 -30.77 1.41 0.71 1.31 -0.56 20.00 0.63 0.21 0.75 -187.30 -49.21 1.53 0.15 1.39 22.96 90.00 1.66 1.07 1.65 7.21 18.92 1.93 0.77 1.70 3.00 37.50 2.06 1.21 1.92 1.52 -39.20 2.20 1.09 1.96 0.15 -46.92 2.12 0.78 1.97 50.00 83.33 2.33 1.47 2.20 2.50 21.43 2.46 1.09 2.15 0.90 -31.49 2.60 1.23 2.28 0.07 0.00 2.73 1.36 2.37 -4.05 -2.66 2.87 1.32 2.46 -2.38 5.99 2.40 1.67 2.47 -2.10 12.91 2.60 1.30 2.48 43.41 70.85 1.47 0.70 1.38 0.06 0.08

37


At maximum stress (Critical osmotic potential of 0.9MPa), reduction in total root length occurred. The present investigation is in concurrence with the report of Ahmed et al., 2009 in sunflower. Shoot length (Table 3) decreased under osmotic stress. The mean shoot length was 4.6 and 2.6 cm in -0.3 and -0.9 MPa osmotic stress treatments when compared to 6.3 cm in control. Genotypes showed variation in shoot length in response to the PEG induced stress. At-0.3 MPa osmotic stress treatment, ASF-104 was found to be superior (9.2 cm), while minimum shoot length was recorded by check SH-491 (2.5 cm). At -0.9MPa osmotic stress, two genotypes ASF-104 and DK3849 were at par and maintained superior shoot length over other genotypes and minimum shoot length was reported in genotype GP9-515-7-3. Shoot length decreased under osmotic stress. Genotypes showed variation in shoot length in response to the PEG induced stress. Singh and Singh (1983) reported that decline in seedling vigour was due to fall in mobilization of reserves to plumule thus preventing their growth under stress. The sunflower genotypes differed in root/shoot length ratio (Table 4) at three different levels of osmotic potentials (0, -0.3 and -0.9 MPa). The average root/shoot length ratio was 0.64 and 1.25 in 0.3 and -0.9 MPa where as, it is 0.52 in control. At the minimum stress of -0.3 MPa, the difference in root/shoot length ratio ranged between 0.02 (check SH-491) and 1.02 (GP-2034). When stress was

further increased to -0.9 MPa, the difference was ranged from 0.59 (GP-1284) to 1.2 (GP-2793). Among the genotypes GP-2035 (1.02) followed by M-1013 (0.99), GP4-2704 (0.99) and NDR-7 (0.97) were superior at -0.3 MPa osmotic stress. While genotypes GP4-2793 (1.19) followed by RSF-101 (1.14), GP-69 (1.13) and RSF-107 (1.12) proved to be superior at maximum stress of -0.9 MPa. The root to shoot length ratio showed higher values under osmotic stress. Increase in root to shoot was observed in almost all genotypes regardless of their drought tolerance, while drought tolerant inbred lines only showed increase in root length. There were two types of responses in osmotic stress regimes. The traits, root length and root to shoot ratio showed higher contribution under osmotic stress when compared with non- stress regime, while trait, shoot length showed decreased trend under water stress. This showed that root length and root to shoot length ratio can be used for the selection of genotypes under drought environment. Similar findings were made (Rauf and Sadaqat, 2007 and Rauf and Sadaqat, 2008) in sunflower. Total seedling dry matter (Table 5) decreased under osmotic stress. The average seedling dry matter was 1.47 and 0.70 g in -0.3 and -0.9 MPa as compared to 1.96 g in control. Significant variation in total seedling dry matter was observed under PEG induced water stress. At -0.3 MPa, total seedling dry matter differed from 0.05 (M1024) to 2.87 g (ASF-104). The genotypes ASF-104 followed by RSF-107 showed significantly

38


higher total seedling dry matter. At -0.9 MPa, total seedling dry matter varied from 0.02 (M-1024) to 1.67 g (TSF-106). The genotypes TSF106 followed by TSF-104 showed significant higher total seedling dry matter, while M-1024 recorded minimum value for this character at both the stresses (-0.3 and -0.9 MPa). None of the cultivars recorded control values and mean values superior to check SH-491 (2.48 g) with respect to total plant dry matter. Two genotypes viz., RSF-107 and ASF-104 recorded higher plant dry matter compared to check SH-491 at both the stresses (-0.3 and -0.9 MPa). The seedling dry matter decreased with the increase in water stress created by PEG-6000. However, different genotypes showed different responses under stress environment. Water stress inhibiting seedling growth is also reported by some researchers in other crop like wheat (Ashraf et al., 1996) and castor (Manjula et al., 2003) in castor. Genotypes (M1029 (80%), Dk-3849 (75%), ASF107 (72%), GP2 -1742 (70%), RSF-107 (70%) and SH-491 (70%)) with good germination percentage failed to record good total seedling dry matter. Similar findings were made by Shantha Nagarajan and Jagadish Rane (2000). According to Salisbury and Floor (1994) germination responses are not necessarily correlated with seedling growth response. There are many reports which are in agreement with the present findings indicating that drought stress induced by PEG severely reduced the growth and biomass of the plant. But some

genotypes had genetic potential to maintain the higher growth under stress conditions. Therefore, we propose the use of PEG-6000 for screening of sunflower genotypes for drought tolerance under laboratory conditions. However, the present study is preliminary and inferences drawn need confirmation. To conclude the genotypes with best performance under PEG induced water stress as drought tolerant one, the results need to be further validated ultimately in the field trials.

REFERENCES Ahmed S, Ahmad R, Ashraf MY, Ashraf M, Waraich EA (2009). Sunflower (Helianthus annuus L.) response to drought stress at germination and seedling growth stages. Pak. J. Bot. 41(2): 647-654. Albuquerque FMC, Carvalho NM (2003). Effect of type of environmental stress on the emergence of sunflower (Helianthus annuus L.), soyabean (Glycine max (L.) Merril) and maize (Zea mays L.) seeds with different levels of vigor. Seed Sci. Tech. 31: 465-467. Ashraf M, Mehmood S (1990). Response of four Brassica species to drought stress. Environ. Exp.. Bot. 30: 93100. Ashraf MY, Naqvi MH, Khan AH (1996). Evaluation of four screening techniques for drought tolerance in wheat (Triticum aestivum L.). Acta Agronomica. 44: 213-220.

39


Chun L, Guoho MI, Jiansheng, Fusuo Z (2005). Genetic analysis of maize root characteristics in response to low nitrogen stress. Plant Soil. 276:369-382. Dhanda SS, Sethi GS, Behl RK (2004). Indices of drought tolerance in wheat genotypes at early stages of plant growth. J. Agron. Crop Sci. 190:1-6. El-Midaoui M, Talouizte A, Benbella M, Serieys H, Griveau Y, Berville A (2001). Effect of osmotic pressure on germination of sunflower seeds (Helianthus annuus L.). Helia. 24: 129134. Gill RK, Sharma AD, Singh P, Bhullar SS (2002). Osmotic stress-induced changes in germination, growth and soluble sugar content of Sorghum bicolor (L.) Moench seeds. Bulgarian J. Plant Physiol. 28:12-25. Hobbs P, Woodhead T, Meisner C (1994). Soil physical factors limiting the productivity of the rice -wheat rotation and ways to reduce their impact through management .In: wheat in heat stressed environments: Irrigated, dry areas and rice – wheat farming system, CIMMYT, Mexico, DF. pp. 276-289. Hu FD, Jones RJ (2004). Effects of plant extracts of Bothriochloa pertusa and Urochloa mosambicensis on seed germination and seedling growth of Stylosanthes hamata cv. Verano and Stylosanthes scabra cv. Seca. Aust. J. Agric. Res. 48:1257-1264.

Lenzi A, Fambrini M , Barotti S, Pugliesi C, Vernieri P (1995). Seed germination and seedling growth in a wilty mutant of sunflower (Helianthus annuus L.): Effect of absisic acid and osmotic potential. Environ. Exp. Bot. 35:427-434. Manjula K, Sharma PS, Ramesh Tatikunta, Nageshwara Rao T (2003). Evaluation of castor (Ricinus communis L) genotypes for moisture stress. Indian J. Plant Physiol. 8(3): 319-322. Michel BE, Kaufmann MR (1973). The osmotic potential of polyethylene glycol 6000. Plant Physiol. 51:914-916. Mohammad MEL, Benbella M, Talouizete A (2002). Effect of sodium chloride on sunflower (Helianthus annuus L.) seed germination. Helia. 37: 5158. Nagarajan S, Rane J (2000). Relationship of seedling traits with drought tolerance in spring wheat cultivars. Indian J. Plant Physiol. 5(3): 264-270. Nicholas M E, Lambers H, Simpson R J, Dalling MJ (1995). Effect of drought on metabolism and partitioning of carbon in two wheat varieties differing in drought tolerance. Ann. Bot. (London) 55:727-742. Rauf S, Sadaqat HA (2008). Effect of varied water regimes on root length, dry matter partitioning and endogenous plant growth regulators in sunflower (Helianthus anuus L.). J. Plant Interactions. 2(1):41-51.

40


Sadasivam S, Chandrababu R, Ravindran N, Raja Jaj (2000). Genetic variation in seed germination root traits and drought recovery in rice. Indian J. Plant Physiol. 5(1):73-78. Sadeghian SY, Yavari N (2004). Effect of water-deficit stress on germination and early seedling growth in sugar beet. J. Agron. Crop Sci. 190:138-144. Sajjan AS, Badanur VP, Sajjanar GM. (1999). Effect of external water potential on seed germination, seedling growth and vigor index in some genotypes of sunflower. In: Proc. Symp. Recent Advances in Management of Arid Ecosystem, (Eds.): S.A. Faroda, N.L. Joshi, S.Kathju and A. Kar. Pp. 215-218. Salisbury PA, Floor RG (1994). Breeding for sowing seed quality. Plant Breeding. 64:1033-1043. Singh K, Afria BS (1985). Evaluation of moisture stress tolerance in castor cultivars during germination seedling growth and emergence. Proc. Indian Nat. Sci. Acad. Biol Sci. 51:364-368. Singh KP, Singh K (1983). Water uptake and germination of wheat seeds under different external water potentials in osmotic solutions. Seed Res. 11:13-19. Smok MA, Chojnowski M, Corbineau F, Come D (1993). Effects of osmotic treatments on sunflower seed germination in relation with temperature and oxygen. In: Proc. 4th Intl.

Workshop on seed: Basic and Applied Aspects of Seed Biology, (Eds.): D. Come and F. Corbineau. Angers, France, pp. 10331038. Turhan H (1997). Salinity studies in potato (Solanum tuberosum L.). PhD Thesis, The University of Reading, UK. pp. 247.

41

RESEARCH ARTICLE


DIALLEL ANALYSIS TO STUDY GENETIC PARAMETERS OF RICE SALINITY TOLERANCE TRAITS AT GERMINATION STAGE. SHARIFI PEYMAN Department of Plant Breeding and Agronomy, Rasht Branch, Islamic Azad University, Rasht, Iran. Corresponding Author email: [email protected]

SUMMARY A 6 × 6 diallel analysis was carried out according to Hayman to obtain the mean squares components. The plant materials used in this study were six rice varieties. The six parents and their F1 hybrids were arranged in a randomized complete block design with three replications for studying their reaction to salinity stress. Two different levels of salinity, 0 and 11.2 dS/m, were used. The data were expressed as percentage of reduction in average of radicle length (RL, %), coleoptile length (CL, %), germination percentage (GP, %) and germination rate (GR, %), obtained by comparing traits in two treatments for each genotype. Analysis of variance showed highly significant differences among genotypes for all of the traits. The significance of additive effects (a) indicated the presence of additive gene action in controlling of GP and RL, while dominance effect (b) component were significant for all of the traits and showed the presence of dominance gene action. The positive values of F component for all of the traits indicated excess of dominant alleles were present in the genetic material. The value (H1/D)0.5 of GP, GR and RL indicated over-dominance, while CL controlled by partial dominance. The low estimate of heritability was revealed the role of dominance gene action in controlling of studied traits. Key words: diallel, genetic analysis, quantitative traits, inheritance, rice. Manuscript received: August 26, 2011; Decision on manuscript: January 10, 2012; Manuscript accepted in revised form: January 21, 2012. Communicating Editor: Bertrand Collard

INTRODUCTION Rice (Oryza sativa L.) is the most important food for half of the world’s population. The soil salinity of reclaimed paddy fields is one of the stresses to limit rice growth and yield (Zeng et al., 2003). Seed germination is usually the most critical stage in seedling

establishment, determining successful crop production (Almansouri et al., 2001). Germination is the first stage of the plant life and appearance of a radicle is the first easily observable event entirely driven by endogenous respiration. In germinating seeds the energy for growth is provided by respiration.

42

Peyman (2012)

Universal inhibition of growth arises from inhibition of respiration since electron transport (mitochondrial and photosynthetic) in the membrane is inhibited by the inhibition of diffusion of the relevant quinone in the voids in the membranes (Mathai et al., 1993). In rice, tolerance to abiotic stresses, especially salinity can be important for direct seeding system. The transplanting of rice seedlings which is a highly labourintensive and expensive operation can be replaced by direct seeding that can reduce labour needs by more than 20 percent in terms of working hours required (Santhi et al., 1998.). Direct seeded rice (DSR) has received much attention because of its low-input demand. It involves dry seeding (sowing dry seeds into dry soil), wet seeding (sowing pre-germinated seeds on wet puddle soils) and water seeding (seeds sown into standing water) (Farooq et al., 2011). The adoption of a direct-seeded method for lowland rice culture would significantly decrease costs of rice production. With respect to yield, both direct seeding (i.e. wet, dry or water seeding) and transplanting had similar results (Kukal and Aggarwal, 2002). Before setting the breeding and selection methods we need to correct the character of genetic information. One way to obtain genetic information is the diallel cross analysis. Several methods have been proposed for diallel analyses (Jinks and Hayman, 1953; Hayman, 1954a, b; Dickinson and Jinks, 1956; Griffing, 1956; Gardner and Eberhart, 1966). Among these methodologies,

Hayman’s (1954a, b) approach has been used to determine gene action on different traits. Apart from additive and dominance gene effects, this method is efficient in detecting epistasis. Diallel method is a systematic experimental approach that is useful in identifying potential cross for the best selection in early generations (Syukur et al., 2010). Diallel analysis has been used for studying salinity stress of rice varieties at various stages. Lee et al. (2003) indicated the tolerance level of indica rice genotypes was higher than japonica, for traits contain height reduction, shoot weight reduction, root weight reduction, shoot Na+, shoot K+. Rice exhibits different salt tolerance at different growth stages and the researchers are focusing more on seedling stage in comparison to middle or late stages (Khatun et al., 1995; Gu et al., 1999). Mishra et al. (1990) suggested the involvement of both major and minor genes for Na/K ratio in rice for both alkali and saline soils. Gregorio and Senadhira (1993) reported overdominance effects for K+/Na+ discrimination in rice at germination stage. Kalaiyarasi et al. (2002) reported the preponderance of non additive gene action in terms of yield components from a study based on two line inter-subspecific rice hybrids. Mahmood et al. (2002) recorded high additive effects for plant height, panicle length and non-additive effects for panicle fertility, shoot dry weight and paddy yield. Deepa Sankar et al. (2008) reported that the traits contain days to 50 percent

43


flowering, productive tillers per plant, panicle exsertion, panicle length, panicle weight, filled grains per panicle, spikelet fertility, 100 grain weight, Na+: K+ ratio, harvest index and grain yield per plant were controlled by dominance gene action and plant height was controlled by additive gene action. In the present study, analysis was conducted to determine the genetic effects for salinity tolerance in rice at germination stage. The objectives of this study were too estimating genetic components of variance and heritability for studied traits.

MATERIALS AND METHODS Genetic material and crosses Six rice varieties were used in this study containing Shahpasand (P1), Hassani (P2), Sepidrod (P3), Neda (P4), Saleh (P5) and IRFAON215 (P6). Some of traits including origin, grain type and plant stature of selected parents were shown in Table 1. This table also revealed the salt tolerant reaction of parental varieties, which obtained from an experiment conducted by author on six level of NaCl (Sharifi, 2011). Among these materials, Shahpasand and Neda were as tolerant varieties and Sepidrod as a sensitive variety. The mating design used in this experiment was a complete diallel crosses. The crosses of parents was carried out in the farm of Rice Research Institute of Iran (RRII) in Rasht, located between 49˚E longitude and 37˚N latitude at an altitude of 7 m below the mean sea level in north of Iran.

Salinity tolerance evaluation was carried out in the laboratory of department of agronomy and plant breeding in Islamic Azad University of Rasht. Salinity tolerance evaluation The six parents and their F1 hybrids were arranged in a randomized complete block design with three replications. Hundred seeds of each of genotypes were surface sterilized in 5% Sodium hypochlorite solution for 5 min and then carefully rinsed with distilled water to remove the sterilizing agent. Two different levels of NaCl i.e., 0 and 11.2 dS/m were used. The concentration of 11.2 dS/m NaCl was choosing as salinity level according to the results of Sharifi (2011). Thirty seeds of each genotype were allowed to germinate in each Petri dish (diameter 9 cm) with two sheets of filter paper. The treatment solution in each Petri dish was changed every day so as to ensure the desired salt level. All Petri dishes were placed in an incubator at 25°C for 10 day with a 12-h light/12-h dark photoperiod. Seeds were considered to be germinated when their root length reached the seed length and shoot length half of the seed length (Wang et al., 2010). In each of these treatments the experiment was conducted in a randomized complete-block design with three replications, where the blocks corresponded to different shelves in the germination chamber. The germinated seeds were observed each day until 10 d, when almost all the seeds were germinated. The percentage of germinated seeds at 10 d was referred to as the final germination

44

Peyman (2012)

percentage (GP). Germination rate (GR) values were calculated by formula that described by Pieper (1952). The radicle and coleoptile length of the seedlings was measured from ten randomly selected seedlings in each treatment from each replication on the tenth day and average of it expressed in mm. The final data were expressed as percentage of reduction in average of these traits, obtained by comparing traits in two treatments for each genotype. Statistical analysis A diallel analysis was carried out according to Hayman (1954a) to obtain the mean squares components. The following statistical model was used: Yrm = m + Jr + Jm + Jrm + Kr – Km + Krm + Erm where, Jrm=L+Lr+Lm+Lrm; m= general mean; Jr, Jm= average deviation from general mean due to the rth and mth parents, respectively; Jrm= residual difference of the mean of the reciprocal crossings; Kr, Km = difference between the rth or mth parent effect used as male of female in the crossing; Krm = residual difference in the reciprocal difference in the mrth order; L = average dominance deviation; Lr = dominance deviation due to the rth parent; Lrm = deviation of the residual difference from the mean of the reciprocal crossings rm; Erm = average experimental error. The slope of the regression line ‘b’ and the Y-intercept ‘a’ were obtained from the relationship where:

Wr = a + b Vr; a = Wr - b Vr; and b = Cov. Vr. Wr / Var. Vr. Where, Vr = Variance of all the offspring in each parental array, when an array consisted of the parental mean and mean values of all the crosses involving that parent; Wr = the covariance between parents and their offspring of the offspring in each parental array with non-recurring parent. Calculation for genetic components, ratios, and estimators (Hayman, 1954a, b) were included in the DIAL98 program (Ukai, 2006). The estimators were calculated only when the genetic components in the respective ratios were significantly different from zero. The significance of the different components of variation was verified by the t-test, by the division of the estimates according to the respective standard deviations. Estimates of genetic components of variation were divided by standard deviations, and significance was determined by the t-test, considering values greater than 1.96, at a significant probability of 5% (Singh and Chaudary, 1979). To fulfill the assumptions of absence of epistasis, no multiple allelism and independent gene distribution, data were subjected to the uniformity test (t2) as described by Singh and Chaudhary (1979).

RESULTS Analysis of variance showed highly significant differences among genotypes for all of the

45


traits. The replication effect was not significant for those traits except RL (Table 2). Reduction of GP, GR, RL and CL were inversely proportional to salinity tolerance, i.e. the higher reduction in the traits revealed lower salinity tolerance or more sensitivity to salinity stress. The reduction of GP due to 11.2 ds/m NaCl concentration in comparison with control was high in the all of the parents. The highest and lowest reduction of GP was revealed in Sepidrod and Neda, respectively. The lowest reduction in GP, RL and CL was revealed in Shahpasand and the highest reduction in these traits was showed in Hassani, IRFAON215 and Sepidrod, respectively (Table 3). The hybrid trait values ranged from 73.04% (Shahpasand × Sepidrod) to 97.5% (Shahpasand × Saleh) for GP, 46.61% (Shahpasand × Sepidrod) to 74.18% (Shahpasand × Neda) for GR, 1.49% (Shahpasand × IRFAON215) to 20.13% (Sepidrod × Neda) for RL and 9.66% (Sepidrod × IRFAON215) to 44.13% (Sepidrod × Neda) for CL (Table 3). In the crosses contain SH × H, SH × SP, H × SH, H × SP, N × SP and N × IR the hybrids presented a lower mean for GP than the parents. In the crosses contain SH × SP, H × SP, SA× SP and IR × N, the hybrids mean for GR was also lower than the lower parents (Shahpasand). Crosses contain SH × IR and SP × IR indicated lower reduction of RL in the hybrids than their parents. Any of hybrids didn’t show lower reduction of CL than parents. The ‘t2’ values were nonsignificant for all of the traits (Table 5), indicating the validity of

assumptions underlying the diallel analysis and additive-dominance model is an adequate description of data for both the characters. Table 4 contains the significance levels of the diallel analysis of variance components. The significance of additive effect (a) indicated the presence of additive gene action for GP and RL, while dominance effect (b) component were significant for all of traits and showed the presence of dominance gene action in controlling of these traits. The directional dominance effect (b1), was significant for RL and CL

46

Peyman (2012)

Table 1. Information of some of important traits on parental varieties. Varieties

Shahpasand Hassani Sepidrod

Neda

Saleh

IRFAON215

Origin

Grain type

Stature

Germination rate (ratio)

Tall

Germination percentage (%) 80.13

5.58

Radicle length (cm) 3.74

Coleoptile length (cm) 2.95

Iran (land race) Iran (land race) Iran (Improved cultivar) Iran (Improved cultivar) Iran (Improved cultivar) IRRI (line)

Long Coarse

Tall

77.25

2.62

2.63

2.06

Long

Dwarf

66.13

2.15

2.55

1.92

Medium

Dwarf

85.38

4.28

2.98

2.84

Long

Semidwarf

75.13

3.17

2.65

2.34

Coarse

Semidwarf

73.63

4.33

1.88

2.26

Table 2. Analysis of variance in a 6×6 complete diallel. Mean Square Source of variance1 Replications Genotype Error

d.f.

2

GP(%)

GR (%)

RL (%)

12.61ns

97.57ns

33.51*

35

43.65**

71.39

70

9.17

34.92

*

47.94

CL (%) 4.15ns

**

6.45

384.34** 53.06

ns

: not significant; *: Significant at 5% of probability level**: significant at 1% of probability level. RL, radicle length reduction; CL, coleoptile length reduction; GP, germination percentage reduction; GR, germination rate reduction.

47


Table 3. Averaged performance of four germination related traits in the six parents and their crosses. Male Traits Female SH H SP N SA IR SH 91.25 87.5 73.04 92.94 97.5 90 H 84.69 91.39 86.5 93.5 94 96.5 SP 95.39 91.5 96.5 95.44 93.39 96.89 N 96.5 91 87 88.5 90.5 86.35 GP (%) SA 93.5 93.5 95 95 92.5 89.28 IR 96.33 93.39 94 94.27 92 95.94 LSD5% 6.14 LSD1% 8.25 SH 52.1 56.26 46.61 74.18 73.3 54.43 H 55.53 61.32 46.71 57.19 57.53 56.33 SP 55.37 56.4 58.62 55.13 64.02 51.32 N 61.5 53.69 50.6 57.6 53.61 56.19 GR (%) SA 55.68 58.13 48.36 50.7 59.18 61.38 IR 53.71 55.56 59.02 47.84 52.64 58.81 LSD5% 11.99 LSD1% 16.09 SH 3.23 8.61 5.56 9 11.41 1.49 H 6.76 17.34 8.4 14.95 10.54 9.3 SP 10.62 16.27 8.94 22.13 5.48 2.41 N 16.59 9.24 3.66 11.39 4.17 3.59 RL (%) SA 8.85 8.24 9.26 8.27 3.39 5.02 IR 7.08 6.89 14.46 6.69 6.52 18.62 LSD5% 5.16 LSD1% 6.91 SH 7.56 22.58 19.08 41.29 86.5 13.74 H 36.43 28.41 27.12 37.68 31.69 30.39 SP 23.98 32.53 29.04 45.13 21.31 9.66 N 19.26 15.43 22.48 15.35 30.62 25.33 CL (%) SA 17.61 13.65 33.83 34.54 7.63 15.69 IR 25.13 26.3 37.06 21.46 29.78 22.43 LSD5% 14.79 LSD1% 19.84 SH, Shahpasand; H: Hassani; SP: Sepidrod; N: Neda; SA: Saleh; IR: IRFAON-215. RL, radicle length reduction; CL, coleoptile length reduction; GP, germination percentage reduction; GR, germination rate reduction.

48

Peyman (2012)

Table 4. Diallel analysis of variance for four germination related traits in rice. MS Source of Variation Replication Additive effect (a) Dominance effect (b) Directional dominance effect (b1) Gene distribution among the Parents (b2) Effects of specific genes (b3) maternal effects (c) reciprocal effects (d) Error

d. f. 2 5 15

13.15 27.25* 40.05**

GR (%) 97.68 42.15 76.77*

1

6.24

38.85

31.23*

5 9 5 10 70

23.04* 53.25** 85.32** 36.84** 9.15

68.14 85.78* 82.06 72.62 34.92

81.49** 20.82** 46.52** 52.72** 6.46

GP (%)

RL (%)

CL (%)

33.48* 58.36** 41.74**

4.15 66.28 335.24** 971.11** 333.66** 265.46** 673.11** 472.7** 53.01

**: Significant at the 1% probability level. *: Significant at 5% of probability level RL, radicle length reduction; CL, coleoptile length reduction; GP, germination percentage reduction; GR, germination rate reduction

Table 6. Estimation of dominance ratios of four germination related traits in rice Parents

GP (%)

GR (%)

RL (%)

CL (%)

SH

0.49

0.66 0.07 0.92 0.95 0.68

0.47 0.93 0.85 0.74 0.91 0.96

0.86

H SP N SA IR

0.73 0.85 0.91 0.91 0.47

0.39 0.95 0.91 0.86 0.44 0.93

SH: Shahpasand; H: Hassani; SP: Sepidrod; N: Neda; D: deilamani; SA: Saleh; IR: IRFAON-215. RL, radicle length reduction; CL, coleoptile length reduction; GP, germination percentage reduction; GR, germination rate reduction

49


Table 5. Estimates of genetic parameters of four germination related traits in rice. Genetic parameters Additive variance (D) Dominance variance (H1) Proportion of dominance due to positive and negative effect of genes (H2) Relative frequency of dominant and recessive alleles (F) Square of difference parents versus all (h2) Environmental variance, whole (E) average degree of dominance ((H1/D)0.5) Proportion of dominance genes (kd/(kd+kr)) Average direction of dominance (h) Broad-sense heritability (h2b) Narrow-sense heritability (h2n) Uniformity test (t2) Proportion of dominance and recessive genes (H2/4H1)

GP (%)

GR (%)

RL (%)

CL (%)

4.71±6.42

-7.78±16.60

40.62±11.05

68.37±48.91

35.53±13.91

52.93±39.57

60.29±14.71

375.71±108.44

31.15±10.83

42.83±28.65

35.46±8.73

283.65±76.13

6.32±10.08

2.09±30.02

56.99±16.08

159.71±86.92

-0.59±5.86

1.90±24.37

7.03±8.14

256.24±126.96

5.57±1.07

17.46±4.29

3.23±0.75

26.53±6.27

2.75±3.13

1.23±7.09

1.22±0.13

0.86±0.97

0.62±0.12

0.0001±0.4505

0.7879± 0.0237

0.7482± 0.0453

-1.32±1.64

-3.28±3.08

-2.95±1.34

16.42±3.82

0.67±0.08

0.38±0.13

0.80±0.05

0.73±0.06

0.10±0.07

0.11±0.08

0.26±0.07

0.01±0.04

0.97

1.53

3.84

0.56

0.22

0.21

0.15

0.20

RL, radicle length reduction; CL, coleoptile length reduction; GP, germination percentage reduction; GR, germination rate reduction

50

Peyman (2012)

Figure 1. Relationship covariance (Wr) and variance (Vr) of reduction in germination percentage (GP, %). 1: Shahpasand; 2: Hassani; 3: Sepidrod; 4: Neda; 5: Saleh; 6: IRFAON-215.

Figure 2. Relationship covariance (Wr) and variance (Vr) of reduction in germination rate (GR, %). 1: Shahpasand; 2: Hassani; 3: Sepidrod; 4: Neda; 5: Saleh; 6: IRFAON215.

51


Figure 3. Relationship covariance (Wr) and variance (Vr) of reduction in radicle length (RL, %). 1: Shahpasand; 2: Hassani; 3: Sepidrod; 4: Neda; 5: Saleh; 6: IRFAON-215.

Figure 4. Relationship covariance (Wr) and variance (Vr) of reduction in coleoptile length (CL, %). 1: Shahpasand; 2: Hassani; 3: Sepidrod; 4: Neda; 5: Saleh; 6: IRFAON-215.

52

Peyman (2012)

The gene distribution among the parents (b2) was significant for GP, RL and CL, which showed symmetrical distribution of genes. The effect of specific genes (b3) was significant for all of the traits. The maternal effect (c) was significant for GP, RL and CL. The reciprocal effect (d) was significant for GP, RL and CL. With regard to the genetic components estimated by the diallel analysis (Table 5), the additive component (D) was significant for RL that confirmed the additive effects of the genes. Dominance component (H1) was significant for GP, RL and CL, which showed dominance effects of genes. Proportion of dominance due to positive and negative effect of genes (H2) was also significant for GP, RL and CL. The positive values of the relative frequency of dominant and recessive alleles (F) for all of the traits indicated excess of dominant alleles were present in the genetic material. The nonsignificance average direction of dominance (h) confirmed that dominance was not unidirectional. The amount of dominance effects can be seen from the value (H1/D)0.5. The value (H1/D)0.5 of GP, GR and RL was more than one indicating over-dominance, while the value of (H1/D)0.5 on CL was less than one indicating a partial dominance. The ratio H2/4H1, estimated the frequency of negative versus positive alleles at loci exhibiting dominance, it was less than 0.25, which indicated that the additive components did not contain all the dominance effects.

Estimates of the ratio of dominance to recessive genes in the parents [KD/(KD+KR)] were less than 1.0 for all of the traits, which indicated the presence of an excess of recessive genes in the parents. The more than 1.0, indicated the presence of an excess of dominant genes in the parents. Estimated value of broad sense heritability (h2b) for GP, GR, RL and CL were 0.67, 0.38, 0.80, and 0.73, respectively. The estimated value of narrow sense heritability (h2n) for four traits were low, namely 0.10, 0.11, 0.26, and 0.01, respectively. Figure 1 shows the distribution of the parent varieties for reduction in germination percentage (GP, %). Because the regression line of Wr (covariance between parents and their offspring in each parental array with nonrecurring parent) cuts in the negative area, one can assume an over-dominance of the genes controlling this trait. This result comes from the dominance ratio, which is larger than one (Table 5). Genotypes contain Neda and Saleh had the most dominant genes, while Shahpasand and Sepidrod had the most recessive genes; these results were revealed in Table 6. The order of dominance of parents to germination rate reduction was IRFAON-215, Hassani, Saleh, Sepidrod, Neda and Shahpasand (Table 6 and Figure 2). Shahpasand was the most recessive genes containing control GR, because the most distant from zero. IRFAON215 contained the most dominant gene, because the closest to zero. The regression line on the graph Wr - Vr has a value of intercept a =

53


−2.85, thus cutting the regression line with the vertical axis (Wr) below the origin (0). Cutting points in these positions indicated that there was an over-dominance gene action (Figure 2). Regression analysis revealed over-dominant type of gene action for reduction in radicle length (Figure 3). Placement of array points as displayed by position of the genotypes along the regression line shows that Sepidrod had the maximum dominant genes for RL being closest to the origin, where as IRFAON-215 had the least dominant gene effect being farthest from the origin. Figures 4 showed that the line intersects the ordinate above the origin, suggesting the occurrence of partial dominance in coleoptile length. The relative position of array points on the regression line depicted that IRFAON-215 had the most dominant genes, while Shahpasand had the most recessive genes for CL (Figure 4).

DISCUSSION Uniformity of Wr and Vr test (t2) revealed the absence of non-allelic interaction giving complete adequacy of the data for the additive-dominance model for all the traits (a significant t2 would indicate the failure of the hypothesis). However, as suggested by Hayman (1954a), even if a trait exhibits a partial failure of assumption, analysis could be carried for such traits, though the results would not be as reliable as they would have been had all assumptions been fulfilled.

Estimation of genetic parameters using diallel analysis carried out for rice germination related traits reacted to salinity, because there are significant differences among genotypes based on the F test of the observed variables (Singh and Chaudhary, 1979). In order to increase rice tolerance to salinity stress, breeders must understand the mechanism of inheritance of related traits and the genetic worth of the parents or the hybrid vigour of the crosses. The efficiency of rice breeding programs based on salinity tolerance can be increased with the knowledge of the genetic inheritance mechanism. The significance of additive effect (a) for GP and RL and dominance effect (b) for all of traits showed the presence of additive and dominance gene action in controlling of mentioned traits. This result is similar to report of Moejopawiro and Ikehashi (1981) and Gregorio and Senadhira (1993) that indicated rice salinity at germination stage were controlling by additive and non-additive gene actions. Genetic effects were observed for all of four traits, indicating the greater importance of non-additive effects in comparison to additive effects in the studied crosses. These results were in agreement with reports of Lee et al. (1996) that indicated in japonica rice, more dominant alleles were present in salt tolerant parents. In contrast, Ray and Islam (2008) indicated the importance of additive effects in the inheritance of salinity tolerance at germination stage. The low estimate of heritability was revealed the role of dominance gene action in

54

Peyman (2012)

controlling of these traits. This offers a lot of scope for improvement of the character through individual plant selection. According to Hayman (1957), epistasis can decrease or increase degree of dominance, which also effect on heritability estimates. Dominance variance (H1) was larger than proportion of dominance due to positive and negative effect of genes (H2) for all of the traits. The proportion of positive genes will be seen from the value of H1 against H2. If H1 > H2 then the genes were more positive genes, on the other hand, if H1 < H2 then the genes were more negative genes (Syukur et al., 2010). Genes involved more heavily in determining GP, GR, RL and CL were positive genes reflected in the value of H1 > H2 (Table 3). An autogamous crop like rice, exploitation of non-additive genetic variance would be impractical. However, using biparental or recurrent selection in segregating material, followed by conventional selection, is likely to lead to substantial trait improvement. Further, advancing of segregating material through bulk, pedigree, single seed descent or single seed descent methods, as suggested by Gupta and Dahiya (1986), would be rewarding. According to Roy (2000), if the estimated value of the narrowsense heritability is high, the selection is done in the early generations because of the character of a genotype heritage to the offspring. Variance of additives and narrow sense heritability is low in the population studied and trait development program can be done

through a selection of the individual or mass selection method and hybrid programs can be used for improving of these traits. In conclusion, the majority of the investigated traits in this study are inherited non-additively. So, the selected genetic material would be oriented to create F1 hybrids. Besides, parents P1 (Shahpasand) had the lowest value of reduction of traits were as a tolerant variety and can be used as a parental variety for improving of salinity tolerance.

REFERENCES Almansouri M., Kinet JM, Lutts S (2001). Effect of salt and osmotic stresses on germination in durum wheat (Triticum durum). Plant Soil. 231: 245–256. Deepa S, Subbaraman PN, Narayanan SL, (2008). Evaluation of temperature sensitive genic male sterility based rice hybrids for adaptation in salt affected environments by AMMI analysis. J. Ecobiol. 23: 263274. Dickinson AG, Jinks JL (1956). A generalized analysis of diallel crosses. Genetics 41:65-78. Farooq M., Kadambot HMS, Rehman H, Aziz T, Lee DJ, Wahid A (2011). Rice direct seeding: Experiences, challenges and opportunities. Soil and Tillage Research 111: 87–98 Gardner CO, Eberhart SA (1966). Analysis and interpretation of the variety cross diallel and

55


related populations. Biometrics 22: 439- 452. Gregorio, G.B., and D. Senadhira. 1993. Genetic analysis of salinity tolerance in rice (Oryza sativa L.). Theor. Appl. Genet. 86:333–338. Griffing B (1956). Concept of general and specific combining ability in relation to diallel crossing systems. Aust. J. Biol. Sci. 9: 463-493. Gu XY, Yan XL, Zheng SL, Lu YG (1999). Influence of salinity on genetic variation of agronomic traits in rice. Scien. Agric. Sinica 32:1-7. (in Chinese). Gupta KR, Dahiya BS (1986). Inheritance of pod yield traits in pea. Crop Improv. 13:4548. Hayman BI (1954a). The theory and analysis of diallel crosses. Genetics 39:789-809. Hayman BI (1954b). The analysis of variance of diallel crosses. Biometrics 10: 235-245. Hayman BI (1957). Interaction, heterosis and diallel crosses. II. Genetics 42: 336-355. Jinks JI, Hayman BI (1953). The analysis of diallel crosses. Maize Genet. News Lett. 27:48-54. Kalaiyarasi R, Palanisamy GA, Vaidyanathan P (2002). The potentials and scope of utilizing TGMS lines in intersubspecies crosses of rice (Oryza sativa L.). J. Genet. Breed. 56:137-143. Khatun S, Rizzo CA, Flowers TJ (1995). Genotypic variation in the effect of salinity on fertility in rice. Plant and Soil 173: 239-250. Kukal SS, Aggarwal GC (2002). Percolation losses of water in relation to puddling intensity

and depth in a sandy loam rice (Oryza sativa) field. Agric. Water Manag. 57:49–59. Lee KS, Choi WY, Ko JC, Kim TS, Gregorio GB (2003). Salinity tolerance of japonica and indica rice (Oryza sativa L.) at the seedling stage. Planta 216: 1043–1046. Lee KS, Senadhira D, Gregorio GB, KyuSeong L (1996). Genetic analysis of salinity tolerance in japonica rice. SABRAO J. Breed. Genet. 28: 7-13. Mahmood T, Shabbir G, Sarfaraz M, Sadiq M, Bhatti MK, Mehdi SM, Jamil M, Hassan G (2002). Combining ability studies in rice (Oryza sativa L.) under salinized soil conditions. Asian J. Plant Sci.1: 88-90. Mathai JC, Sauna ZE, John O, Sitaramam V (1993). Rate limiting step in electron transport: osmotically sensitive diffusion of quinines through voids in the bilayer. J. Biol. Chem. 268: 1544215454. Mather K, Jinks JL (1971). Biometrical Genetics. Chapman and Hall Ltd., London, UK. Mishra B, Akbar M, Seshu DV (1990). Genetics studies on salinity tolerance in rice towards better productivity in salt-affected soils. Rice Research Seminar, International Rice Research Institute, Los Baños, Philippines, 12 July. Moejopawiro S, Ikehashi H (1981). Inheritance of salt tolerance in rice. Euphytica 30: 291-300. Pieper A (1952). Das Saatgut V.P. D arey Berlin, Hamburg, Germany.

56

Peyman (2012)

Ray PKS, Islam MA (2008). Genetic analysis of salinity tolerance in rice. Bangladesh J. Agric. Res. 33:519-529. Roy D (2000). Plant breeding, analysis and exploitation of variation. New Delhi: Narosa Publishing House. Santhi PK, Ponnuswamy, Chetty NK (1998). Effect of seeding methods and efficient nitrogen management practices on the growth of lowland rice. J. Ecobiol. 10: 123-132. Singh TK, Chaudary BD (1979). Biometrical methods in quantitative genetic analysis. Kayani Puplishers, Ludhiana, 304 p. Sharifi P (2012). Evaluation of Salinity stress on some of characters in Germination of rice. Plant Eco. J. (submitted). Syukur M, Sujiprihati S, Yunianti R, Undang G (2010). Diallel Analysis using Hayman Method to Study Genetic Parameters of Yield Components in Pepper (Capsicum annuum L.). HAYATI J. Biosci. 17:183188. Ukai Y (2006). DIAL98. A package of progams for the analyses of a full and half diallel table with the methods by Hayman (1954), Griffing (1954) and http://lbm.ab.a.uothers. tokyo.ac.jp/~ukai/dial98.html. Wang Z, Wang J, Bao Y, Wang F, Zhang H (2010). Quantitative trait loci analysis for rice seed vigor during the germination stage. J. Zhejiang Univ. Sci. 11: 958–964. Zeng L, Poss JA, Wilson C, Drez ASE, Gregorio GB, Grieve CM (2003). Evaluation of salt tolerance in rice genotypes in rice genotypes

by physiological characters. Euphytica 129: 281-292.

57

RESEARCH ARTICLE


DETERIORATION OF F2 HETEROSIS IN F3 GENERATION IN DIALLEL CROSS OF UPLAND COTTON Z.A. SOOMRO1*, N.U. KHAN2, M.B. KUMBHAR1, M.A. KHUHRO1, S.H. GHALOO1, T.A. BALOCH3 AND M.I. MASTUNGI3 1

Department of Plant Breeding and Genetics, Sindh Agriculture University, Tandojam, Pakistan 2 Department of Plant Breeding and Genetics, Khyber Pakhtunkhwa Agricultural University, Peshawar, Pakistan 3 Directorate of Agriculture Extension, Balochistan, Pakistan *Corresponding author e-mail: [email protected]

SUMMARY An experiment was laid out in randomized complete block design with four replications to assess heterosis, heterobeltiosis and inbreeding depression for yield and yield components in 5 × 5 F2 and F3 diallel cross of Gossypium hirsutum L. during 2002-05 at Sindh Agriculture University, Tandojam, Pakistan. Heterosis was manifested by F2 populations for all the characters but extent of heterosis varied among the characters. It may be assumed that heterosis in one or more than one yield components conferred heterosis for seed cotton yield. The F2 hybrids Reshmi × TH-3/83, Reshmi × Mc-Niar-3150, CIM-109 × Mc-Niar-3150, CIM-109 × TH-3/83 and Mc-Niar-3150 × CIM-109 displayed positive heterosis and heterobeltiosis for seed cotton yield along with one or more yield components character. Reduction in F2 and F3 heterosis was due to decrease in heterozygosity through allelic fixation. Generally, expected inbreeding depression was quite higher than observed for all quantitative traits except seed index, the discrepancy between observed and expected inbreeding depression may be attributable to several reasons that involve linkage disequilibrium, epistasis and abnormal segregation at meiosis due to high ploidy level. Keywords: heterosis, inbreeding depression, diallel cross, F2 and F3 generations, upland cotton Manuscript received: December 30, 2011; Decision on manuscript: March 22, 2012; Manuscript accepted in revised form: April 1, 2012. Communicating Editor: Bertrand Collard

58

Soomro et al. (2012)

INTRODUCTION The most important goal in cotton breeding programs is to improve the yielding capacity of the crop. Breeding for yield includes genetic manipulation of the yield components. One method of increasing yielding potential in cotton is to breed cultivars that make better use of heterosis. Heterosis in seed cotton yield has been observed for many years, although its level of expression is highly variable. Heterosis varies with the parental genotypes and environmental conditions under which the hybrids are grown (Knobel et al., 1997; Larik et al., 2004; Soomro et al., 2006). F2 hybrids have greater heterogeneity and genetic variation and might result in an enormous range of acclimatization relative to their parents and F1 hybrids (Meredith and Brown, 1998; Wu et al., 2004; Khan et al., 2010). F2 performance was highly correlated with F1s, and in some cases the F2 heterosis has been found equivalent to F1 heterosis due to lower inbreeding depression, and some where F2 hybrids also showed superior performance than welladapted cultivars (Meredith, 1990; Khan et al., 2010). The F2 heterosis in cotton has also been reported in previous studies (Wang and Pan, 1991; Tang et al., 1993; Li et al., 2000; Xing et al., 2000; Han and Liu, 2002; Khan et al., 2007, Khan, 2011). Average heterosis of F2 over mid parent, based on population means, suggested that little inbreeding depression exists for F2 and F3 generations and it is possible to screen and select high yielding F2 hybrids for further use

(Yuan et al., 2002; Khan, 2011). However, Khan et al. (2009) suggested that combined performance of F1 and F2 hybrids could be a good indicator to identify the most promising populations to be utilized either as F2 hybrids or as a resource population for further selection in advanced generations. A critical pre-requisite for successful production of hybrid cotton is that sufficient heterosis is available through specific parental combinations. The primary objectives of this study were to determine heterosis over mid and better parents, and inbreeding depression for seed cotton yield and its components in F3 generation of upland cotton. MATERIALS AND METHODS Five approved upland cotton cultivars (TH-3/83, NIAB-78, CIM109, Mc-Niar-3150 and Reshmi) chosen for their diverse genetic background were used in diallel crosses. The experimental material consisted of the five parents and 20 F2 and F3 diallel combinations. The F2s was sown in 2003-04 to obtain F3 seeds. The F3s seeds along with parents were sown in 2004-05 in a randomized complete block design (RCBD) with four replications. Seed rate of 35 kg ha-1 was applied. Before first irrigation seedlings were thinned to maintain plants and rows spacing’s of 30 and 75 cm, respectively. The adequate requirements of fertilizer and irrigations were applied. The interculturing and other post cultural operations to keep the crop free from

59


weeds, pests and disease were adopted. Ten plants from both F2 and F3 populations and parents in each replication was keenly studied and data were recorded for plant height, sympodia and bolls per plant, boll weight, seed index and seed cotton yield per plant. The standard method of analysis of variance according to Steel and Torrie (1980) was used to workout the mean differences among all the genotypes including parents, their F2 and F3 hybrids for various traits. Heterosis over mid and better parents was calculated according to Larik et al. (2004) using following formula.

(ID) estimates in F3 hybrids were computed by using the method as suggested by Griffing (1950).

Inb. Depression (%) = F 3 F 2 × 100 F2 The expected Inbreeding depression of F3 hybrids was also formulated as Falconer (1989).

Exp. Inb. Depression (%) =

1 ( P1 + P2 + F2 ) 4

P1, P2 and F2 are the mean performance of parents and their F2 hybrids.

RESULTS AND DISCUSSION

- MP × 100 Heterosis (% ) = F 1 MP

- BP Heterobeltiosis (% ) = F 1 × 100 BP Significant difference of F2 means from mid and better parental values were also tested by using “t” test as suggested by Sharma (1994).

S .E ( F 1 - MP ) = S 2 x 3 / 8 t=

F 1 - MP S .E ( F 1 - MP )

S .E ( F 1 - BP ) = S 2 x 3 / 8 t=

F 1 - BP S .E ( F 1 - BP )

where s2 is the sum of square from ANOVA of the particular cross combination. Inbreeding depression

Analysis of variance revealed that all the genotypes differed significantly (P ≤ 0.01) for all traits, indicating presence of considerable genetic variability among parents and their F2 and F3 progenies for further evaluation. Performance of parents, F2 and F3 populations In the F2 generation, the hybrids TH3/83 × Reshmi produced tallest plants (111.3 cm), while McNair3150 × NIAB-78 had the shortest plants (77.6 cm) (Table 1). In F3 generation, cross NIAB-78 × McNair-3150 produced tallest plants (88.3 cm), while parent cultivars CIM-109 and TH-3/83 manifested shortest plants of 53.5 and 53.4 cm, respectively as compared to the rest of the genotypes. For sympodia per plant, NIAB-78 × CIM-109 displayed highest number of fruiting branches (22.18) and Reshmi × McNair-3150 expressed lowest number of fruiting branches per plant (17.07) in F2 generation. The F3 hybrid

60


CIM-109 × NIAB-78 displayed 17.9 sympodia per plant, while parent TH-3/83 expressed lowest number of fruiting branches (12.3). In case of bolls per plant, parent NIAB-78 expressed maximum bolls per plant (33.9) as compared to other genotypes, while lowest bolls per plant were expressed by cross TH3/83 × Reshmi (17.4) in F2 generation. In the F3 generation, the cross CIM-109 × TH-3/83 produced maximum bolls per plant (20.1) while lowest were shown by parent TH-3/83 (8.56). The parent Reshmi gave maximum boll weight (3.55 g) while lowest boll weight was shown by Mc-Nair-3150 (2.36 g) in F2 generation. In F3 generation, NIAB78 × Mc-Nair-3150 produced maximum boll weight (3.22 g) and lowest boll weight was produced by parent NIAB-78 (2.19 g). Maximum seed index was manifested by F2 hybrid Mc-Nair3150 × Reshmi (6.77 g) and lowest by NIAB-78 × CIM-109 (5.17 g). In the F3 generation, hybrid NIAB-78 × Mc-Nair-3150 displayed maximum seed index (7.65 g) and TH-3/83 × Reshmi gave minimum seed index (5.62 g). The F2 hybrid Reshmi × TH-3/83 provided higher seed cotton yield per plant (91.18 g), while cross Mc-Nair-3150 × NIAB78 provided lowest seed cotton yield (51.01 g). In F3 generation, NIAB78 × Mc-Nair-3150 produced higher seed cotton yield per plant (50.14 g), whereas parents TH-3/83 and CIM109 expressed lowest seed cotton yield (21.45 g). Results revealed that two hybrids NIAB-78 × Mc-Nair3150 and CIM-109 × NIAB-78 displayed overall better performance for six quantitative traits. Results further revealed that F2 populations revealed higher estimates as

compared to F3 hybrids for all traits showing deterioration of heterosis in F3 generation. Heterosis over mid and better parents There was a general tendency for an increase in plant height in F2 hybrids over mid-parent (MP) and betterparent (BP), respectively in direct as well as reciprocal crosses (Table 2). The F2 hybrid TH-3/83 × CIM-109 displayed maximum significant MP (+19.96%) and BP (17.24%) heterosis. It is concluded that increase height in these F2 hybrids over MP and BP was may be due to interaction of complementary growth genes for tallness. Therefore, these F2 hybrids could be exploited in developing tall hybrids which suppress weeds to some extent and produce more sympodia per plant which consequently enhance seed cotton yield per plant. Out of 10 F2 crosses, only two hybrids i.e. NIAB78 × Reshmi and CIM-109 × Reshmi showed maximum negative MP heterosis (-0.575% and 6.023%) in direct crosses, while three F2 reciprocal crosses exhibited negative MP heterosis and four crosses exhibited negative heterobeltiosis. The F2 hybrids McNair-3150 × TH-3/83 and TH-3/83 × CIM-109 displayed maximum significant positive heterosis over MP (+19.79%) and BP (+17.24%). The expression of significant positive heterosis in these F2 hybrids indicated the preponderance of additive gene action for this trait (Larik et al. 2004, Solangi et al., 2002). Larik et al. (2004) showed fair degree of heterosis for plant height over MP and BP, indicating over dominance of additive gene action. They explained the

61


deterioration in F2 generation may be due to epistasis, linkage disequilibrium, abnormal segregation at meiosis and higher polyploidy of cotton crop. In case of sympodia per plant, majority of F2 hybrids displayed negative MP and BP heterosis (Table 2). The F2 hybrid NIAB-78 × CIM-109 expressed maximum positive heterosis (09.04 %) and heterobeltiosis (7.51%) in direct crosses. Similarly, maximum negative heterosis over MP (-9.93%) and BP (-11.52%) was shown by F2 hybrid NIAB-78 × Reshmi. In reciprocals, 50% and 60% of F2 hybrids displayed negative heterosis and heterobeltiosis, respectively (Table 2). Positive heterosis was considered inevitable for sympodia per plant by several workers (Baloch et al. 1999; Solangi et al., 2001; Larik et al. 2004; Khan, 2011) because fruiting branches were found mostly positively correlated with bolls per plant and seed cotton yield. For bolls per plant, F2 cross CIM-109 × Mc-Nair-3150 showed maximum heterosis (+16.11%) and heterobeltiosis (+13.09%) (Table 2). However, 50% and 80% of F2 hybrids expressed negative heterosis and heterobeltiosis, respectively for bolls per plant (Table 2). Maximum negative heterosis (-21.14%) and heterobeltiosis (-33.25%) was shown by F2 cross TH-3/83 × Reshmi and its reciprocal cross displayed maximum significant heterosis (+45.17%) and heterobeltiosis (+22.87%). In case of inheritance of bolls per plant only 33% hybrids over MP and 25% hybrids over BP manifested dominant gene action, while rest of the crosses showed additive type of

gene action. These results are in accordance with the findings of Baloch et al. (1993) and Khan et al. (1999). In case of boll weight, 60% of F2 hybrids expressed positive heterosis over MP and 90% displayed negative heterobeltiosis (Table 2). Maximum positive heterosis (+7.23%) and heterobeltiosis (+1.94%) was shown by F2 hybrid CIM-109 × Mc-Nair3150, while maximum negative heterosis (-12.09%) and heterobeltiosis (-21.28%) was shown by F2 hybrid TH-3/83 × Reshmi in direct crosses. In reciprocals, F2 hybrid Reshmi × NIAB-78 exhibited maximum significant heterosis (+12.35%), while F2 hybrid CIM-109 × TH-3/83 displayed maximum heterobeltiosis (4.55%). These results reinforce other researcher’s conclusion that inbreeding depression starts from F2 to F3 generation (Baker and Verhalen, 1975). The F2s performance was highly correlated with F1s, and in some cases F2 hybrids also showed superior performance than well-adapted cultivars (Meredith, 1990; Khan et al., 2010). Although it is assumed that F2 hybrids consisted of very heterogenous populations and heterogeneity may result in a greater range of adaptation as compared to their parents, however, there was little evidence in our study to support the hypothesis. However, F2 hybrids have greater genetic variation and might result in an enormous range of acclimatization relative to their parents and F1 hybrids (Meredith and Brown, 1998; Wu et al., 2004; Khan et al., 2007 and 2010; Khan, 2011).

62


Table 1. Mean performance and ANOVA for yield and yield components of parents and F2 and F3 populations of upland cotton. F2 generation Source of Variation

D.F.

Plant height (cm)

Sympodia plant-1

Replications 3 733.761 25.288 Genotypes 24 289.128** 5.264 Error 72 70.688 3.356 Parents and their F2 and F3 populations TH-3/83 94.45 18.72 NIAB-78 98.45 20.05 CIM-109 90.16 20.63 Mc-Niar-3150 83.12 19.39 Reshmi 110.6 19.34 TH-3/83 × NIAB-78 97.83 19.14 TH-3/83 × CIM-109 110.7 19.71 TH-3/83 × Mc-Niar-3150 104.4 18.96 TH-3/83 × Reshmi 111.3 19.5 NIAB-78 × CIM-109 110.3 22.18 NIAB-78 × Mc-Niar-3150 98.28 19.98 NIAB-78 × Reshmi 103.9 17.74 CIM-109 × Mc-Niar-3150 96.65 19.89 CIM-109 × Reshmi 94.32 18.84 Mc-Niar-3150 × Reshmi 102.1 17.21 NIAB-78 × TH-3/83 93.45 18.44 CIM-109 × TH-3/83 100.6 20.78 Mc-Niar-3150 × TH-3/83 107.0 19.63 Reshmi × TH-3/83 109.4 20.14 CIM-109 × NIAB-78 99.62 20.08 Mc-Niar-3150 × NIAB-78 77.60 20.3 Reshmi × NIAB-78 103.3 18.96 Mc-Niar-3150 × CIM-109 95.67 20.51 Reshmi × CIM-109 107.4 19.68 Reshmi × Mc-Niar-3150 99.02 17.07 LSD(0.05) 7.749 1.633 ** = Significant at p≤0.01, N.S. = Non-Significant

F3 generation

1

Boll weight (g)

Seed index (g)

93.947 73.098** 18.792

0.258 0.321** 0.093

0.441 0.792** 0.094

Seed cotton yield plant-1 (g) 236.764 557.58** 150.707

26.07 33.9 22.06 23.27 18.06 33.64 23.95 27.18 17.4 23.1 23.38 22.56 26.32 20.55 21.37 25.09 25.75 22.75 32.04 24.53 19.92 24.75 26.96 18.74 25.07 6.111

2.81 2.59 2.62 2.36 3.55 2.51 2.57 2.66 2.79 2.56 2.59 2.86 2.67 2.82 3.06 2.43 2.94 2.69 2.86 2.57 2.61 3.45 2.52 2.69 2.98 0.2733

6.26 5.39 5.51 5.66 6.57 5.47 5.79 6.08 6.16 5.17 5.66 6.55 6.51 6.02 6.77 5.41 6.35 5.94 6.19 5.53 5.51 6.09 5.48 6.23 5.57 0.3628

72.24 87.03 56.33 54.56 65.52 84.69 61.52 72.14 49.37 58.8 57.92 63.7 70.87 57.25 64.11 61.31 74.77 58.91 91.18 62.35 51.01 84.85 64.72 52.97 79.57 17.30

Bolls plant-

Plant height (cm)

Sympodia plant-1

Bolls plant-1

Boll weight (g)

Seed index (g)

1700.61 323.116** 69.241

59.846 8.774** 2.616

10.894 40.679** 10.332

0.170 0.239** 0.10

1.656 0.612** 0.24

Seed cotton yield plant1 (g) 1142.325 299.766** 63.795

53.43 60.75 53.48 71.59 73.08 67.65 65.75 69.86 66.31 70.25 88.34 77.53 62.13 75.53 64.65 77.56 76.01 61.52 61.92 82.41 78.33 82.10 63.37 73.69 79.91 15.57

12.25 14.68 12.5 16.69 15.56 16.21 13.71 15.88 15.37 15.80 17.79 15.65 15.21 15.43 14.90 17.5 16.59 15.8 15.24 17.93 15.18 17.75 13.84 16.88 14.89 3.026

8.56 12.43 9.40 18.65 13.75 11.88 10.85 11.61 11.09 10.88 16.09 13.11 11.42 15.56 8.79 17.35 20.14 11.16 11.25 18.15 13.03 15.79 10.00 14.31 10.74 6.014

2.70 2.19 2.64 2.42 2.79 2.94 2.37 2.84 2.72 2.85 3.21 2.89 2.40 2.85 2.91 2.48 2.42 3.03 2.58 2.80 2.43 2.48 2.51 2.72 2.82 0.5916

6.36 6.81 6.82 6.65 6.90 6.65 6.57 6.52 5.62 6.30 7.65 7.29 6.56 6.62 6.92 6.22 6.06 6.90 6.48 6.57 6.83 6.39 6.61 6.70 6.87 0.9165

21.45 27.80 21.45 42.16 37.80 31.86 25.53 31.89 28.66 30.63 50.14 37.99 27.30 43.95 25.89 43.95 47.12 33.50 28.54 50.11 31.41 39.98 23.89 39.03 28.85 1.494

63


Table 2. Heterosis and heterobeltiosis in F2 populations for seed cotton yield and its components in upland cotton. F2 populations

Plant height MP (%)

BP (%)

Sympodia plant-1 MP (%)

BP (%)

Bolls plant-1 MP (%)

BP (%)

Boll weight MP (%)

BP (%)

Seed index MP (%)

BP (%)

Seed cotton yield plant-1 MP (%)

BP (%)

TH-3/83 × NIAB-78

1.43

-0.63

-1.30

-4.57

12.18

-0.77

-6.93

-10.49

-6.16

-12.64**

6.35

-2.68

TH-3/83 × CIM-109

19.97**

17.24**

0.17

-4.46

-0.48

-8.14

-5.48

-8.61

-1.58

-7.47*

-4.31

-14.85

TH-3/83 × Mc-Niar-3150

17.63**

10.57

-0.54

-2.25

10.18

4.26

2.88

-5.26

2.04

-2.83

13.77

-0.15

TH-3/83 × Reshmi

8.59

0.67

2.47

0.83

-21.14

-33.26

-12.09*

-21.29**

-3.97

-6.25

-28.32*

-31.66*

NIAB-78 × CIM-109

16.98**

12.06*

9.04

7.51

-17.45

-31.86

-1.74

-2.28

-5.19

-6.18

-17.97

-32.44**

NIAB-78 × Mc-Niar-3150

8.26

-0.17

1.31

-0.35

-18.22

-31.04

4.43

-0.19

2.43

0.00

-18.19

-33.45**

NIAB-78 × Reshmi

-0.58

-6.02

-9.93

-11.52

-13.16

-33.44

-7.004

-19.54**

9.55**

-0.26

-16.48

-26.80**

CIM-109 × Mc-Niar-3150

11.55

7.19

-0.63

-3.61

16.11

13.09

7.23

1.94

16.43**

14.86**

27.82*

25.82

CIM-109 × Reshmi

-6.02

-14.69

-5.74

-8.68

2.45

-6.84

-8.54

-20.50

-0.33

-8.38*

-6.04

-12.63

Mc-Niar-3150 × Reshmi

5.44

-7.64

-11.14

-11.25

3.39

-8.18

3.53

-13.78*

10.68**

3.03

6.78

-2.15

NIAB-78 × TH-3/83

-3.11

-5.07

-4.88

-8.03

-16.32

-25.98

-10.08

-13.51

-7.24*

-13.65**

-23.01*

-29.54**

CIM-109 × TH-3/83

9.01

6.54

5.61

0.73

6.99

-1.24

8.13

4.55

7.88*

1.42

16.31

3.49

Mc-Niar-3150 × TH-3/83

19.79**

13.25*

2.99

1.22

-7.79

-12.75

4.04

-4.19

-0.28

-5.03

-7.09

-18.46

Reshmi × TH-3/83

6.74

-1.03

5.83

4.14

45.17**

22.87

-9.96

-19.38**

-3.46

-5.75

32.37**

26.21*

CIM-109 × NIAB-78

5.64

1.19

-1.30

-2.69

-12.33

-27.64

-1.39

-1.94

1.43

0.36

-13.01

-28.36**

Mc-Niar-3150 × NIAB-78

-14.52*

-21.18**

2.91

1.22

-30.32**

-41.24**

5.24

0.58

-0.35

-2.72

-27.95*

-41.38**

Reshmi × NIAB-78

-1.12

-6.53

-7.29

-8.93

-4.76

-27.01

12.34*

-2.81

1.80

-7.31*

11.25

-2.49

Mc-Niar-3150 × CIM-109

10.42

6.11

2.48

-0.59

18.93

15.84

1.14

-3.85

-1.82

-3.14

16.73

14.89

Reshmi × CIM-109

7.03

-2.84

-1.53

-4.59

-8.42

-16.72

-12.84*

-24.24**

3.19

-5.14

-13.05

-19.15

Reshmi × Mc-Niar-3150

2.25

-10.44

-11.86*

-11.97

21.29

7.73

0.66

-16.17**

-8.98**

-15.27**

32.52*

21.44

Average

4.65

-3.43

-0.57

-3.39

6.76

-5.04

0.66

-1.34

0.06

-2.76

3.73

3.67

64


Table 3. Inbreeding depression in F3 populations for seed cotton yield and its components in upland cotton. F3 populations

Plant height Exp. (%)

Obs. (%)

Sympodia plant-1 Exp. (%)

Obs. (%)

Bolls plant-1 Exp. (%)

Obs. (%)

Boll weight Exp. (%)

Obs. (%)

Seed index Exp. (%)

Obs. (%)

Seed cotton yield plant-1 Exp. (%)

Obs. (%)

TH-3/83 × NIAB-78

80.88

-30.84

12.46

-15.29

18.91

-64.67

2.61

17.16

6.38

21.56

50.23

-62.38

TH-3/83 × CIM-109

85.52

-40.63

13.57

-30.39

18.95

-54.66

2.91

-7.55

6.84

13.42

57.14

-58.49

TH-3/83 × Mc-Niar-3150

87.39

-33.10

14.48

-16.21

25.29

-57.29

2.78

6.91

6.71

7.23

69.8

-55.78

TH-3/83 × Reshmi

88.86

-40.42

11.99

-21.17

18.28

-36.25

3.16

-2.79

7.23

-8.79

57.63

-41.94

NIAB-78 × CIM-109

86.18

-36.32

12.91

-28.74

19.26

-52.87

2.6

11.07

6.22

21.81

53.32

-47.89

NIAB-78 × Mc-Niar-3150

80.15

-10.11

13.73

-10.96

21.91

-31.17

2.34

24.07

6.22

35.05

50.89

-13.42 -40.35

NIAB-78 × Reshmi

84.51

-25.38

11.71

-11.79

16.17

-41.89

2.89

1.26

6.74

11.19

46.93

CIM-109 × Mc-Niar-3150

82.95

-35.72

14.41

-23.49

20.77

-56.58

2.56

-10.31

6.51

0.79

50.59

-61.48

CIM-109 × Reshmi

90.43

-19.92

12.97

-18.06

16.49

-24.29

3.07

0.96

7.01

9.96

50.12

-23.21

Mc-Niar-3150 × Reshmi

88.64

-36.68

13.09

-13.40

22.2

-58.82

3.03

-5.09

7.03

2.14

62.13

-59.61

NIAB-78 × TH-3/83

79.50

-17.00

12.61

-5.10

17.82

-30.83

2.59

2.06

6.43

15.07

47.53

-28.31

CIM-109 × TH-3/83

80.64

-24.46

12.97

-20.15

15.53

-21.78

2.86

-17.42

6.92

-4.53

44.36

-36.98

Mc-Niar-3150 × TH-3/83

85.88

-42.48

14.19

-19.51

23.19

-50.94

2.71

12.58

6.8

16.07

60.99

-43.13 -68.69

Reshmi × TH-3/83

88.53

-43.40

11.77

-24.31

20.95

-64.88

3.25

-9.80

7.17

4.65

68.51

CIM-109 × NIAB-78

82.07

-17.27

13.11

-10.67

16.03

-26.00

2.69

9.05

6.33

18.81

44.28

-19.62

Mc-Niar-3150 × NIAB-78

81.40

0.94

12.6

-25.20

24.02

-34.58

2.42

-6.98

6.25

23.98

58.63

-38.41

Reshmi × NIAB-78

88.26

-20.55

12.38

-2.79

20.68

-36.19

3.07

-28.24

6.86

4.95

66.59

-52.88

Mc-Niar-3150 × CIM-109

84.59

-33.76

14.72

-32.49

21.78

-62.90

2.55

-0.55

6.49

20.48

53.44

-63.08

Reshmi × CIM-109

90.14

-31.39

12.81

-14.21

17.6

-22.09

3.11

1.04

7.06

7.50

54.26

-26.31

Reshmi × Mc-Niar-3150

90.95

-19.29

13.5

-12.75

23.02

-57.13

2.98

-5.37

6.98

23.31

63.61

-63.74

Exp. = Expected inbreeding depression, Obs. = Observed inbreeding depression

65

In seed index, majority of F2 hybrids displayed negative heterotic effects. In direct crosses, 50% of F2 hybrids displayed negative heterosis with maximum values of -6.15% and -12.64% shown by F2 hybrid TH-3/83 × NIAB-78 over MP and BP, respectively. Maximum positive significant heterosis over MP (+16.43%) and BP (+14.85%) was manifested by F2 hybrid CIM-109 × Mc-Nair-3150. Similarly, significant negative MP (-7.24%) and BP (13.65%) heterosis was expressed by F2 hybrid NIAB-78 × TH-3/83. Heterosis for seed cotton yield per plant demonstrated general tendency for negative heterotic effects (Table 2). In direct crosses, 40% of hybrids expressed positive heterotic effects over MP and 10% over BP. Maximum significant positive heterosis over MP (+27.82%) and BP (+25.81%) was shown by F2 hybrid CIM-109 × McNair-3150. Similarly, maximum significant negative MP (-28.32%) and BP (-33.45%) heterosis was exhibited by F2 hybrids TH-3/83 × Reshmi and NIAB-78 × Mc-Nair3150, respectively. In reciprocal crosses, 50% and 40% F2 hybrids revealed positive heterosis and heterobeltiosis, respectively. Hybrid Reshmi × TH-3/83 displayed maximum significant positive heterosis (+32.36%) and heterobeltiosis (+26.21%). Results supported by Solangi et al. (2001 & 2002) and Baloch et al. (2002) and reported that F2 hybrids showed 50% inbreeding depression but surprisingly were superior in few characters even from their F1’s. F2 hybrids may even express only 50% of economic heterosis shown by F1 hybrids, and even less when heterosis is defined in terms of

higher yielding parent, which can lead also to cultivar improvement (Meredith, 1990; Khan et. al., 2010). Results indicated additive gene effects, which inferred that continued progress could be possible in exploiting these hybrids for genetic improvement of seed cotton yield per plant. The yield heterosis of these F2 hybrids was also in agreement with those of Baloch et al. (2002), Larik et al. (2004), Soomro et al. (2006) and Khan (2011). Results suggested that F2 hybrids have considerable economic potential to be exploited for breeding hybrid cotton genotypes. However, Khan et al. (2009) suggested that combined performance of F1 and F2 hybrids could be a good indicator to identify the most promising populations to be utilized either as F2 hybrids or as a resource population for further selection in advanced generations. However, large number of studies has shown that heterosis is based on directional dominance and epistasis, but there is little evidence of real over dominance (Jinks, 1956). It is concluded that sympodia per plant, bolls per plant and boll weight significantly contributed towards enhancement of seed cotton yield in F2 hybrids. Inbreeding depression Inbreeding is a system of mating that leads to an increase in homozygosity, resulting decline in vigor and productivity. The degree of inbreeding in any generation is equal to degree of homozygosity in that generation (Baloch et al., 1991). Results demonstrated that F3 hybrids suffered considerable amount of

66


inbreeding depression (Table 3). Generally expected inbreeding depression was higher than observed values for all traits except seed index which confirms the involvement of multigenic epistasis. The expected inbreeding depression ranged from 44.28 to 69.8, while observed depression varied from -13.42 to 68.69 in seed cotton yield per plant. The discrepancy between observed and expected inbreeding depression could be explained by two factors, linkage disequilibrium and epistasis interaction. Gardner et al. (1953), Gardner (1963) and Baloch et al. (1991) reported that repulsion phase linkages can cause upward or positive biases in estimation of dominance variance in F3 populations where linkage effects are expected to be maximum. Comstock and Robinson (1948) and Soomro et al. (1996) suggested that if multigenic epistasis is present, the epistasis of dominance will biased downwards and ultimately the expected inbreeding depression will rise than the observed. Inbreeding depression in polyploids has been found to exceed what is expected by the coefficient of inbreeding. Aycock and Wilsic (1968) reported that in alfalfa, an autotetraploid, the yield decreased twice as much as predicted and that response may be attributed to decrease in favourable interactions among multiple alleles due to inbreeding and abnormal segregation at meiosis due to higher ploidy. However, maximum positive values for inbreeding depression were observed for boll weight and seed index, where F3 values exceeded F2 means. Results suggested that in such inter-varietal crosses the maximum accumulation

of genes resulted in increased fertility in F3 populations consequently resulted in formation of bold seed (Paul et al. 1987). Inbreeding depression may be defined as the reduction in vigor and fertility as a result of inbreeding. The character wise inbreeding depression in F3 generation is given below: For plant height, almost all F3 crosses showed inbreeding depression, highest values -40.63% occurs in TH-3/83 × CIM-109 while lowest magnitude of – 10.11% exhibited by F3 hybrid NIAB-78 × Mc-Niar-3150 (Table 3). In case of reciprocal crosses, only one cross Mc-Niar-3150 × NIAB-78 showed no inbreeding depression while in rest of crosses, the F3 hybrid Reshmi × TH-3/83 showed highest inbreeding depression (-43.40%). In sympodial branches per plant, F3 hybrid TH-3/83 × CIM-109 showed highest inbreeding depression (-30.39%) and lowest (10.96%) by NIAB-78 × Mc-Niar3150. While in reciprocals, McNiar-3150 × CIM-109 displayed highest value –32.49 and Reshmi × NIAB-78 displayed lowest value – 2.79%. The F3 hybrid TH-3/83 × NIAB-78 displayed highest values of inbreeding depression (-64.67%) for bolls per plant (Table 3), and CIM-109 × Reshmi presented lowest magnitude of inbreeding depression (-24.29%). In the case of reciprocals, F3 hybrid Reshmi × TH-3/83 (–64.88%) revealed highest inbreeding depression which excelled all other reciprocals, while CIM-109 × TH3/83 showed lowest inbreeding depression (-21.78%).

67

In case of boll weight, maximum inbreeding depression was expressed by F3 hybrid Reshmi × NIAB-78 in reciprocals, whereas hybrid CIM-109 × Mc-Niar-3150 displayed 10.31% depression in direct crosses. For seed index, out of 20 F3 crosses, only two crosses displayed inbreeding depression, in which maximum value (-8.79%) was shown by F3 hybrid TH-3/83 × Reshmi. In seed cotton yield per plant, all F3 hybrids manifested considerable amount of inbreeding depression ranged from -13.43% (NIAB-78 × Mc-Niar-3150) to 68.68% (Reshmi × Th-3/83). Results fully supported by findings of Larik et al. (1999) as reported considerable amount of inbreeding depression in F2 hybrids of bread wheat. Average heterosis of F2 over mid parent, based on population mean, suggested that little inbreeding depression exists for F2 and F3 generations and it is possible to screen and select high yielding F2 hybrids for further use (Yuan et al., 2002). Khan et al., (2007) mentioned that majority of F2 populations displayed inbreeding depression and it was high for seed cotton yield followed by bolls per plant, lint % and staple length. Inbreeding depression was higher in high performing hybrids than in low performing hybrids. However, Meredith (1990) determined that F2 hybrids with lower inbreeding depression in yield revealed superior performance than well-adapted cultivars.

REFERENCES Aycock MK, Wilsic CP (1968). Inbreeding in Medicago sativa L. by sib-mating. II. Agronomic traits. Crop Sci. 8: 481-485. Baker JL, Verhalen LM (1975). Heterosis and combining ability for several agronomic and fibre properties among selected lines of upland cotton. Cotton Grow Rev. 52(3): 209-223 Baloch MJ, Lakho AR, Soomro AH (1993). Heterosis in interspecific cotton hybrids. Pak. J. Bot. 25(1): 13-20. Baloch MJ, Lakho AR, Bhutto HU (1999). Line-tester analysis for estimating genetic components of some quantitative traits in G. hirsutum L. Sindh Bal. J. Plant Sci. 1(1): 28-34. Baloch MJ, Lakho AR, Bhutto HU, Rind R (2002). Seed cotton yield and fiber properties of F1 and F2 hybrids of upland cotton. Asian J. Plant Sci. 1(1): 48-50. Baloch MJ, Tunio GH, Lakho AR (1991). Expression of heterosis in F1 and its deterioration in intrahirsutum F2 hybrids. Pakphyton 3: 95-106. Comstock RE, Robinson HF (1948). The components of genetic variance in populations of biparental progenies and their use in estimating the average degree of dominance. Biometrics 4: 254-266. Falconer DS (1989). Introduction to Quantitative Genetics. 3rd ed. Longman Scientific and Technical Publ. U.K.

68


Gardner CO (1963). Estimates of genetic parameters in cross fertilizing plant and their implication in plant breeding. In: Statistical genetics and Plant Breeding (ed.) W.D. Hanson and H.F. Robinson pp. 225-252. Gardner CO, Harvey PH, Comstock RE, Robinson HF (1953). Dominance of genes controlling quantitative characters in maize. Agron. J. 51: 524-528. Griffing JB (1950). Analysis of quantitative gene action by constant parent regression and related techniques. Genet. 35: 303-312. Han XM, Liu XX (2002). Genetic analysis for yield and its components in upland cotton. Acta Agron. Sin. 28: 533-536. Jinks JL (1956). The F2 back cross generation from a set of diallel crosses. Heredity 10: 1-30. Khan MA, Khan MA, Lodhi TE (1999). Genetic study of yield and yield related components in cotton (G. hirsutum L). J. Animal and Plant Sci. 9(1): 73-75. Khan NU (2011). Economic heterosis for morpho-yield traits in F1 and F2 diallel crosses of upland cotton. SABRAO J. Breed. Genet. 43 (2): 144-164. Khan NU, Hassan G, Kumbhar MB, Marwat KB, Khan MA, Parveen A, Aiman U, Saeed M (2009). Combining ability analysis to identify suitable parents for heterosis in seed cotton yield, its components and lint % in upland cotton. Ind. Crops Prod. 29: 108-115.

Khan NU, Hassan G, Kumbhar MB, Kang S, Khan I, Parveen A, Aiman U, Saeed M (2007). Heterosis and inbreeding depression and mean performance in segregating generations in upland cotton. Eur. J. Sci. Res. 17: 531-546. Khan NU, Basal H, Hassan G (2010). Cottonseed oil and yield assessment via economic heterosis and heritability in intraspecific cotton populations. Afr. J. Biotechnol. 9(44): 74187428. Knobel HA, Labuschagne MT, VanDeventer CS (1997). The expression of heterosis in the F1 generation of a diallel cross of diverse hard red winter wheat genotypes. Cereal Res. Comm. 25(4): 911-916. Larik AS, Mahar AR, Kakar AA, Shaikh MA (1999). Heterosis, inbreeding depression and combining ability in T. aestivum L. Pak. J. Agric. Sci. 36(1-2): 39-44. Larik AS, Soomro ZA, Ghaloo SH (2004). Expression of hybrid vigor and dominance in inter-varietal crosses of cotton. The Indus Cottons 1(1): 4-8. Li WH, Hu XY, Shen WW, Song YP, Xu JA (2000). Selection crosses with heterosis for F2 generation of hybrids in upland cotton (G. hirsutum L.). Acta Agron. Sin. 26: 919-924. Meredith WR (1990). Yield and fiber-quality potential for second generation cotton hybrids. Crop Sci. 30: 10451048.

69

Meredith WR, Brown JS (1998). Heterosis and combining ability of cottons originating from different regions of the United States. J. Cotton Sci. 2: 77-84. Paul NK, Johnson TD, Eagles CF (1987). Heterosis and inbreeding depression in forage rape (B. napus L). Euph. 36: 345-349. Sharma JR (1994). Heterosis and gain in vigor. Published by Tata McGraw Hill Publ. Co. Ltd. New Delhi, pp. 122140. Solangi MY, Baloch MJ, Bhutto HU, Lakho AR, Solangi MH (2001). Hybrid vigor in interspecific F1 hybrids of G. hirsutum × G. barbadense for some economic characters. Pak. J. Biol. Sci. 4(8): 945-948. Solangi MY, Baloch MJ, Solangi MH, Chang MS, Lakho AR (2002). Performance of intra-hirsutum F1 hybrids. Asian J. Plant Sci. 1(2): 126-127. Soomro AR, Nachnani GH, Soomro AW, Kalhoro AD (1996). Heterosis for yield in tetraploid cotton hybrids and their reciprocals. The Pak Cottons 40(3&4): 18-23. Soomro ZA, Larik AS, Kumbhar MB, Khan NU (2006). Expression of heterosis in the F1 generation of a diallel cross of diverse cotton genotypes. Sarhad J. Agric. 22(3): 427-431. Steel RGD, Torrie JH (1980). Principles and procedures of statistics: A biological approach. Second Edition, McGraw Hill, Book Co., New York, Toronto, London.

Tang B, Jenkins JN, McCarty JC, Watson CE (1993). F2 hybrids of host plant germplasm and cotton cultivars: I. Heterosis and combining ability for lint yield and yield components. Crop Sci. 33: 700-705. Wang XD, Pan JJ (1991). Genetic analysis of heterosis and inbreeding depression in upland cotton. Acta Agron Sin. 17: 18-23. Wu YT, Yin JM, Guo WZ, Zhu XP, Zhang TZ (2004). Heterosis performance of yield and fiber quality in F1 and F2 hybrids in upland cotton. Plant Breed. 123: 285-289. Xing CZ, Jing SR, Guo LP, Yuan YL, Wang HL (2000). Study on the heterosis and combining ability of transgenic Bt (Bacillus thuringiensis) cotton. Cotton Sci. 12: 6-11. Yuan YL, Zhang TZ, Guo WZ, Pan JJ, Kohel RJ (2002). Heterosis and gene action of boll weight and lint percentage in high quality fiber property varieties in upland cotton. Acta Agron. Sin. 28: 196-202.

70

RESEARCH ARTICLE


SSR MAP CONSTRUCTION AND QUANTITATIVE TRAIT LOCI (QTL) IDENTIFICATION OF MAJOR AGRONOMIC TRAITS IN MUNGBEAN (Vigna radiata (L.) Wilczek) TANAPORN KAJONPHOL1, CHONTIRA SANGSIRI2, PRAKIT SOMTA3, THEERAYUT TOOJINDA4, and PEERASAK SRINIVES3* 1

Program in Plant Breeding, Faculty of Agriculture at Kamphaeng Saen, Kasetsart University, Kamphaeng Saen, Nakhon Pathom 73140, Thailand 2 Agricultural Science Program, Mahidol University Kanchanaburi Campus, Sai Yok, Kanchanaburi 71150, Thailand 3 Department of Agronomy, Faculty of Agriculture at Kamphaeng Saen, Kasetsart University, Kamphaeng Saen, Nakhon Pathom 73140, Thailand 4 Rice Gene Discovery Unit, National Center for Genetic Engineering and Biotechnology, Kasetsart University, Kamphaeng Saen, Nakhon Pathom 73140, Thailand * Corresponding author email: [email protected]

SUMMARY Mungbean (Vigna radiata) (2n = 2x = 22) is an important annual legume in Asia. It is widely grown in South and Southeast Asia, as well as China. The aim of this research was to use SSR markers to construct a linkage map and identify chromosome regions controlling some agronomic traits in mungbean. The mapping population comprised 186 F2 plants derived from a cross between an annual cultivated mungbean line ‘KUML29-1-3’ (Vigna radiata var. radiata) and an Australian wild perennial mungbean accession ‘W021’ (Vigna radiata var. sublobata). A total of 150 SSR primers were composed into 11 linkage groups, each containing at least 5 markers. The map spans 1,174.2 cM with the average distance between the adjacent markers of 7.8 cM. Comparing the mungbean map with azuki bean (Vigna angularis) and blackgram (Vigna mungo) linkage maps revealed extensive genome conservation between the three species. Twenty QTLs controlling major agronomic characters includingdays to first flower (FLD), days to first pod maturity (PDDM), days to harvest (PDDH), 100 seed weight (SD100WT), number of seeds per pod (SDNPPD) and pod length (PDL) were located on to the linkage map. Most of the QTLs were located on linkage groups 7 and 5.

Keywords: mungbean, Vigna radiata, comparative genome mapping, agronomic traits, simple sequence repeat, quantitative trait loci Manuscript received: December 14, 2011; Decision on manuscript: January 14, 2012; Manuscript accepted in revised form: January 28, 2012. Communicating Editor: Bertrand Collard

71

Kajonphol et al. (2012)

INTRODUCTION Mungbean (Vigna radiata (L.) Wilczek: 2n = 2x = 22) is one of the most important annual legumes. It is native to India and becomes an economic crop in many countries in Asia, Africa and South America. Mungbean seed is consumed as a protein source for human and animals. Mungbean plants can be made into hay and green manure. It is usually cultivated in cropping systems. The production of mungbean grain in the world is around 3.5 to 4 M tons per year (Weinberger, 2003). Products from mungbean seed are rich in vitamins, minerals and easily digested proteins. However, the average yield of mungbean is still low due to susceptibility to pests and diseases, its indeterminate growth habit and photoperiod sensitivity (Fernandez and Shanmugasundaram, 1988). All mungbean cultivars are annual crop with two broad growth stages, vegetative (V) and reproductive (R). Pookpakdi et al. (1992) proposed a system that further classifies both stages based on soybean growth stages published by Fehr et al. (1971). V stages are determined by counting the number of developed nodes on the main stem, beginning with the unifoliolate nodes as the first nodes (stage V1) and the final node is the node which has fully developed trifoliate leaf (stage Vn) when the leaf at the node above is unrolled sufficiently. R stages are determined from R1 (beginning bloom), R2 (beginning pod), R3 (beginning seed), R4 (full seed), R5 (beginning maturity), R6 (first harvest), and R7 (second harvest).

Difference in number of dates specifying to each growth stage can affect seed yield. Khattak et al. (1995) found that number of days to flowering is negatively correlated with number of pods per plant and total seed weight with especially strong direct effect on total seed weight. To perform a breeding process effectively, inheritance of dates specifying different growth stages should be investigated in order to manipulate developmental stages of mungbean through selection. In addition, molecular markers associated with the traits should be determined in order to save time used in selection cycles through marker-assisted selection. Mungbean has a very small genome with the size of 579 Mbp/1C (Arumganathan and Earle, 1991). Thus one would expect short chromosomes in each linkage group. The first linkage map of mungbean was constructed from RFLP markers by MenancioHautea et al. (1992). Then Young et al. (1992 and 1993) located a major bruchid insect resistance and powdery mildew disease resistance genes onto this map. However, the RFLP marker map has not been further used due to limitation that it requires a large amount of good quality DNA for analysis. The technique is time-consuming and expensive, making it less suitable for large-score screening programs in plant breeding. Moreover, the RFLP markers are not distributed throughout the genome. Although, Lambrides et al. (2000) added RAPD markers into the map, there was no report of its further use. The main reason is that the RAPD markers are dominant markers and

72


thus cannot distinguish between homozygous and heterozygous genotypes. Yet, RAPD technique is not always repeatable. This made a specific PCR-based marker, especially Simple Sequence Repeat or SSR marker the marker of choice. SSR is the variable in short tandem repeat of DNA bases, giving co-dominant markers which can distinguish between homozygote and heterozygote. Nowadays, there are research reports using SSR makers for mapping the mungbean genome and locating QTLs. Kasettranan et al. (2010) located QTLs conferring resistance to powdery mildew disease on a SSR partial linkage map of mungbean. Chankaew et al. (2011) reported a QTL mapping for Cercospora leaf spot (CLS) resistance in mungbean. Recently, Zhao et al. (2010) reported construction of a mungbean genetic linkage map by combining 76 RFLP markers from Humphry et al. (2002) and 103 new loci consisting of 97 SSR, 4 RAPD and 2 STS markers. The number of PCR-based markers in their map was too low to make use of the map. The objectives of this study were: (1) to estimate heritability of yield components and dates specifying growth stages in mungbean; (2) to construct a SSR linkage map of mungbean; and (3) to map QTLs controlling the traits.

MATERIALS AND METHODS Mapping population and DNA extraction

The population used in this study was an F2 population of 186 plants developed from an inter-subspecies cross between cultivated mungbean line ‘KUML29-1-3’ (V. radiata var. radiata) (hereafter called KUML) and a wild mungbean accession ‘W021’ (V. radiata var. sublobata). KUML29-1-3 was developed from the Project on Genetics and Breeding of Field Legumes for Thailand, Kasetsart University, Kamphaeng Saen Campus. The line has high and stable seed yield. W021 was obtained from the National Institute of Agrobiological Sciences (NIAS), Tsukuba, Japan. It is a small-seeded wild perennial mungbean with long vegetative and reproductive growth stages. Young leaves of 7 days old from parental lines and F2 plants were extracted for DNA using a CTAB method (Lodhi et al., 1994). The DNA concentration was estimated by comparing with λ DNA standard on agarose gel electrophoresis. The DNA concentration was adjusted to 1ng/µl for PCR amplification. Phenotyping and data analysis Twenty plants each of KUML, W021, F1 (KUML x W021) and F1r (W021 x KUML), and all 186 F2 plants were individually grown in 12-inch pots each filled with 5 kg of soil. All plants were placed in a net house during February to October 2010 at Kasetsart University, Kamphaeng Saen, Thailand (14° 01′ N, 99° 59′ E, 7.5m ASL). Data were recorded from individual plants on days to first flower (FLD), days to first pod maturity (PDDM), days to harvest (PDDH), pod width (PDW) in mm,

73


pod length (PDL) in cm, number of seeds per pod (SDNPPD), total number of pods per plant (PDTN), 100 seed weight (SD100WT) in g, and total seed weight (SDTWT) in g. The data were analyzed by an analysis of variance (ANOVA) of a completely randomized statistical design (CRD). Difference between trait means is declared by Duncan’s Multiple Range Test (DMRT) at P ≤ 0.05. Variance of each trait was calculated in KUML, W021, F1, F1r and F2 populations and used to estimate broad-sense heritability (h2) based on the equation h2 = σ2g/σ2p. Where σ2g is the genotypic variance component and σ2p is the phenotypic variance component. In this experiment, σ2g was estimated from VF2 – (VP1+VP2+VF1+V F1r)/4; where VF2, VP1, VP2, VF1 and VF1r are the variation between plants within the specified genotypes, and σ2p was estimated from VF2 (Fehr, 1987). Phenotypic correlation coefficients between traits were calculated from 186 F2 plants using software R-program v. 2.8.1 (http://www.r-project.org/).

(2008). The DNA product was electrophoresed on 4.5% polyacrylamine gel with 0.5x TE buffer for 1-2 h. The DNA bands were visualized by silver staining.

SSR analysis Nine hundred and forty-five SSR markers were screened to detect polymorphism between the parents. Of these, 628 markers were developed from mungbean (Kumar et al., 2002a and 2002b; Gwag et al., 2006; Somta et al., 2008 and 2009; Seehalak et al., 2009 and Tangphatsornruang et al., 2009), 191 were from azuki bean (Wang et al., 2004), 119 were from common bean (Blair et al., 2003 and Buso et al., 2006) and 7 were from cowpea (Li et al., 2001). PCR reaction and amplification were the same as described by Somta et al.

RESULTS

Linkage map construction and QTL analysis A linkage map was developed by JoinMap 3.0 program (van Ooijen, 2009). A minimum LOD score of 3.0 was used as a threshold value for grouping the markers. Genetic distance between markers was calculated using Kosambi map function (Kosambi, 1944). Linkage groups were named after azuki bean linkage groups (Han et al., 2005). QTL analysis for each character was performed using composite interval mapping (CIM) by WinQTL Cartographer 2.5 program (Wang et al., 2007). A permutation test (Churchill and Doerge, 1994) was run for 2,500 times at the significance level of P = 0.01 to determine a LOD score threshold for declaring a significant QTL.

Phenotypic data and broad-sense heritability Mean and standard deviation of the parents and F2 population were presented in Table 1. All traits were different among the parents but not different between F1 and F1r, revealing no maternal effect conditioning these traits. KUML showed determination, while W021 showed indetermination in growth habit. Days to first flower (FLD) of KUML was only 31 while that of W021 was 65. The same relationship was also found in days

74


to first pod maturity (PDDM) (47 vs 82), and days to harvest (PDDH) (76 vs 140). Pod length (PDL) of KUML was longer than W021 (8.2 vs 4 cm), and pod width (PDW) was also wider (4.7 vs 3.1 mm). Total number of pods per plant (PDTN) of KUML was lower than W021 (25 vs 109), while number of seeds per pod were not different. The F2 population can be classified into different classes according to days to first flower, days to first pod maturity, days to harvest, 100 seed weight, number of seeds per pod, total number of pods per plant, pod length, pod width and total seed weight (Figure 1). All traits, except 100 seed weight showed transgressive segregation. Days to first flower, pod maturity and harvest of the F2 population ranged respectively from 29-76, 44-96 and 79-178 days, demonstrated skewing toward KUML. The progenies showed positive segregation when compared with W021 (Figure 1a, 1b and 1c). Yield components such as number of seeds per pod, total seed weight and pod length showed transgressive segregation, while 100 seed weight fell between the parents. The F2 plants had 100 seed weight of 1.0 g to 3.0 g whereas W021 and KUML had 0.6 and 4.2 g per 100 seeds, respectively (Fig. 1d). When compared with KUML, total seed weight and pod width showed positive transgressive segregation but number of seeds per pod showed negative segregation. The progenies also showed positive transgressive segregation in total number of pods per plant when compared with W021.

Broad-sense heritabilities (h2) as calculated from the F2 data of each trait were presented in Table 1. The broad-sense heritability of flowering dates, viz. FLD, PDDM and PDDH were 88.6%, 91.2% and 86.8%, respectively, which were considered highly heritable. The heritabilities of yield components were high in PDL (92.4%), PDW (97.5%), SDNPPD (91.2%) and SD100WT (90%), and mediumhigh in PDTN (77.0%), SDTWT (65.1%). Correlation between agronomic characters The phenotypic correlation coefficients among 9 quantitative traits are given in Table 2. FLD showed positive correlation with PDDM (r = 0.966**) and PDDH (r = 0.693**). These traits tended to correlate negatively with yield components and total seed weight. 100 seed weight had positive correlation with SDTWT (0.535**), PDL (0.574**) and PDW (0.376**). Correlation between yield components were high in number of seeds per pod with pod length (0.781**), and 100 seed weight with pod length (0.574**). This result indicated that yield depended on seed size and pod size (pod length and pod width), number of seeds per pod, and 100 seed weight. In this experiment, pod length, pod width and 100 seed weight were positively correlated to total seed weight, while days to flower was negatively correlated with seed weight. Thus an optimum number of days to flower should be considered as a selection criterion together with yield components.

75


(a)

(b) Days to first flower

Days to first pod maturity 60

Number of F2 plants

50

50

40

Number of F2 plants

29-1-3 30

W021

20 10 0

29-33 34-38 39-43 44-48 49-53 54-58 59-63 64-68 69-73 74-78 Days

(c)

W021

20 10 0 44-49 50-55 56-61 62-66 67-71 72-76 77-81 82-86 87-91 92-96 Days

100 seed weight

Days to harvest W021

50

70 60

Number of F2 plants

29-1-3

50

40

40

30

29-1-3

W021

30

20

20

10

10

0 76-88

89-101 102-114 115-127 128-140 141-153 154-166 167-179 Days

(e)

0 1.0-1.2

1.3-1.5

2.5-2.7

2.8-3.0

80

29-1-3

70 60 50 40 30 20 10 0 3-5

6-8

9-11

29-1-3

12-14

50 40 30 20 10 0 40

Grams

(g)

(h) Pod width

Pod length 80

50 45 40

W021 29-1-3

10 5 0 3.4-3.7 3.8-4.3 4.4-4.8 4.9-5.3 5.4-5.8 5.9-6.3 6.4-6.8 6.9-7.3 7.4-7.8 7.9-8.3 Centimeters

Number of F2 plants

Number of F2 plants

2.2-2.4

Total seed weight

W021 60

Number of F2 plants

W021

90

35 30 25 20 15

1.6-1.8 1.9-2.1 Grams

(f) Number of seeds per pod

Number of F2 plants

29-1-3

30

(d)

60

Number of F2 plants

40

70 60 50 40 30

W021

29-1-3

20 10 0 2.7-3.0 3.1-3.4 3.5-3.8 3.9-4.2 4.3-4.6 4.7-5.0 5.1-5.4 5.5-5.8 Millimeters

76

>5.8


(i) Total number of pods per plant Number of F2 plants

45 40 35 30

W021

29-1-3

25 20 15 10 5 0 >10

10-30

31-50

51-70

71-90

91-110

111-130

131-150

151-170

>170

Pods

Figure 1. Frequency distribution of the F2 population derived from the cross ‘KUML29-1-3 x W021’: (a) days to first flower, (b) days to first pod maturity, (c) days to harvest, (d) 100 seed weight, (e) number of seeds per pod, (f) total seed weight, (g) pod length, (h) pod width, (i) total number of pods per plant.

Linkage map construction Nine hundred and forty-five SSR markers were screened between the two parents and 152 markers (16.08%) were found polymorphic. One hundred and fifty markers could be assigned into 11 linkage groups of mungbean chromosomes plus a small linkage group (CEDG144 and CEDG149) with the total coverage of 1,174.2 cM, giving the average chromosome length of 97.9 cM. The average distance between SSR loci on the map is 7.8 cM (Figure 2). Each chromosome was tagged with five or more markers. Of these 150 markers, 75 are mungbean loci, 61 are azuki bean loci, 13 are common bean loci and one is cowpea locus. QTL analysis Twenty putative QTLs of agronomic traits were detected by CIM with WinQTL Cartographer 2.5 (Figure 2; Table 3). Four, 3, 3, 6, 2 and 2 QTLs were detected for FLD, PDDM, PDDH, SD100WT, SDNPPD and PDL, respectively (Table 3). The amount of phenotypic variation in each trait explained by its respective QTLs ranges from 6.3 to 28.6%. The

number of QTLs per trait range between 2 and 6 loci. QTLs for different traits were co-located on the map (Figure 2). These include Fld2, Pddm2 and Pddh2 on LG2 and Fld4.1, Pddm4.1 and Pddh4.1 on LG4, to name a few. Comparative linkage map between mungbean, azuki bean and black gram The mungbean linkage map was compared with azuki bean linkage map (Han et al., 2005), and black gram linkage map (Chaitieng et al., 2006). Sixty-two and 21 SSR markers were common between azuki bean and mungbean, and between black gram and mungbean, respectively. Most of the common markers were mapped on the same linkage groups and orders, with a few exceptions (Fig. 3). Marker order in our study was 42 out of 62 loci (68%) colinear with azuki bean, and 19 out of 21 loci (90%) colinear with black gram. Reverse regions were identified between mungbean map and azuki bean map on LG1, 2 and 5. One inversion region between mungbean map and black gram map was tagged by BW212 and CEDG056 on LG9.

77


Table 1. Mean and standard deviation of major agronomic traits observed from parents and progenies from the cross between an annual cultivated mungbean line ‘KUML291-3’ and a wild perennial mungbean accession ‘W021’, with their variances and corresponding heritabilities. Generations/ parameters

FLD

PDDM

PDDH

PDL (cm)

PDW (mm)

SDNPPD

PDTN

P1(29-1-3)

31.2c ± 1.6 65.5a ±6.5 42.5b ±2.2 47.8b ±2.1 47.7b ±11.0 11.1 120.3 13.7 88.6

47.4c ± 1.1 81.8a ±6.5 57.0bc ±1.5 62.7b ±2.0 64.2b ±12.0 12.1 143.6 12.6 91.2

76.4b ± 0.5 139.6a ±19.2 124.0a ±6.0 123.5a ±5.4 123.4a ±28.6 28.1 818.2 108.3 86.8

8.2a ±0.1 4.0b ±0.2 5.3ab ±0.3 5.3ab ±0.4 5.7ab ±2.7 3.0 1.1 0.1 92.4

4.7a ±0.2 3.1c ±0.1 4.0ab ±0.1 4.0b ±0.1 3.9bc ±0.7 0.7 0.4 0.01 97.5

12.0a ±0.4 9.7ab ±0.9 8.8b ±3.0 8.4b ±1.0 10.1ab ±2.6 2.9 6.8 0.6 91.2

25.0c ±6.5 109.5ab± 14.8 149.6a ±21.2 151.4a ±29.1 61.2bc ±41.2 52.9 1695.9 390.1 77.0

P2(W021) F1 F1r F2 LSD.05 VF2 VE h2(%)

SD100 WT (g) 4.2a ±0.1 0.6c ±0.04 1.9b ±0.1 2.0b ±0.2 1.8b ±0.4 0.4 0.2 0.02 90.0

SDTWT (g) 9.6a ±2.7 1.4b ±0.2 12.0a ±4.1 12.1a ±6.1 7.7ab ±6.6 8.1 43.8 15.3 65.1

Means of each trait followed by the same letter are not different as compared by DMRT at P ≤ 0.05 FLD = days to first flower, PDDM = days to first pod maturity, PDDH = days to harvest, PDL = pod length (cm), PDW = pod width (mm), SDNPPD = number of seeds per pod, PDTN = total number of pods per plant, SD100WT = 100 seed weight (g) and SDTWT = total seed weight (g)

Table 2. Correlation between number of days in each growth stage and yield components of the F2 plants. PDDM PDDH PDW PDL SDNPPD PDTN SD100WT FLD 0.966** 0.693** -0.152* -0.286** -0.244** -0.278** -0.373** PDDM 0.700** -0.162* -0.349** -0.314** -0.319** -0.395** PDDH -0.146* -0.303** -0.272** -0.021ns -0.312** PDW 0.328** 0.190** 0.134* 0.376** PDL 0.781** 0.232** 0.574** SDNPPD 0.235** 0.226** PDTN 0.295** SD100W T ns,** non significant and significant at 0.01 level of probability (df=184), respectively. FLD = days to first flower, PDDM = days to first pod maturity, PDDH = days to harvest, PDL = pod length (cm), PDW = pod width (mm), SDNPPD = number of seeds per pod, PDTN = total number of pods per plant, SD100WT = 100 seed weight (g), SDTWT = total seed weight (g)

78

SDTWT -0.293** -0.347** -0.118ns 0.212** 0.522** 0.472** 0.822** 0.535**


Figure 2. SSR linkage map of mungbean constructed from the F2 population. Cumulative distances in centiMorgans (Kosambi’s) and marker names are shown on the left and right sides of the linkage group, respectively. QTL intervals detected at LOD ≥ 2.0 are presented as boxes on the left of the linkage groups.

.

79

Figure 3. A comparative linkage map between mungbean from this study Vs azuki bean (left) (Han et al. 2005) and black gram (right) (Chaitieng et al. 2006), based on azuki common markers.

80


Table 3. QTLs conditioning six traits detected in the F2 population of a cross between cultivated mungbean line ‘KUML29-1-3’ and accession ‘W021’. Traits

QTL names

Days to first flower (FLD)

Days to first pod maturity (PDDM)

Days to harvest (PDDH)

100 seed (SD100WT)

weight

Markers in F2 population

Position (cM)

LOD score

QTL effect Additive (a)

Dominance (d)

R2 (%) 15.88%

Fld2

2

VR0364

72.70

13.3

-6.20

3.08

Fld4.1

4

CEDG241-VR-SSR019

9.85

6.5

-3.98

2.34

7.39%

Fld4.2

4

DMB-SSR199-CEDG107

69.27

20.1

-7.91

-4.37

28.57%

Fld11

11

VR0216-CEDG168

14.01

5.3

-4.73

-0.59

6.28%

Pddm2

2

VR0364

72.70

10.2

-6.04

2.76

12.58%

Pddm4.1

4

CEDG241-VR-SSR019

9.85

7.1

-4.8

2.00

8.43%

Pddm4.2

4

DMB-SSR199-CEDG107

69.27

18.5

-8.46

-4.67

27.83%

Pddh2

2

VR0364

72.70

8.4

-12.95

8.93

15.60%

Pddh4.1

4

CEDG241-VR-SSR019

9.85

6.5

-10.19

10.58

11.73%

Pddh4.2

4

VR0313

17.3

5.9

-9.61

11.60

11.63%

Sd100wt2.1

2

VR078-CEDG065

4.01

6.6

0.19

-0.04

14.56%

(g)

Number of seeds per pod (SDNPPD) Pod length (cm) (PDL)

Linkage groups

Sd100wt2.2

2

VR17-VR0200

19.13

7.7

0.19

-0.01

12.99%

Sd100wt4

4

VR0366-VR035

35.08

4.9

0.18

0.14

11.96%

Sd100wt8

8

VR-SSR031-VR0225

52.22

5.0

0.15

0.05

8.16%

Sd100wt9

9

CEDG259-CEDG166

19.21

4.5

0.15

-0.03

7.22%

Sd100wt11

11

MB-SSR104-VR-SSR011

53.08

5.2

0.16

-0.03

8.74%

Sdnppd1.1

1

VR-SSR015-VR-SSR018

34.69

5.7

1.69

1.05

12.29%

Sdnppd1.2

1

VR0194-VR0198

46.18

5.6

1.7

0.80

12.08%

Pdl7

7

CEDG111-VR0126

66.86

5.2

0.41

-0.12

10.74%

Pdl8

8

VR-SSR005-VR-SSR031

50.43

4.9

0.36

0.23

9.26%

81


DISCUSSION In this study, broad-sense heritabilities of days to first flower (FLD), days to first pod maturity (PDDM) and days to harvest (PDDH) were high (88.6, 91.2 and 86.8%, respectively). Sriphadet et al. (2007) studied inheritance of agronomic traits and their interrelationship in RIL mungbean lines obtained from the cross between wild mungbean ‘ACC 41’ and the cultivated ‘Berken’. They found that flowering date skewed towards ACC 41, but the narrowsense heritability was high at 88.0 %. They also reported an abnormal distribution in FLD, PDDM and PDDH data. Similar results were also reported by Siddique et al. (2006) that there were high heritabilities in days to first flower and days to harvest. Rohman et al. (2003) reported that days to first flower, days to harvest, 100 seed weight and plant height had high heritability, while total number of pods per plant and number of seeds per pod were low in heritability. Correlation analysis in this study revealed that days to flower showed positive correlation with days to first pod maturity and days to harvest. Days to flowering and days to maturity showed negative correlation with yield components such as 100 seed weight. According to Khattak et al. (1995), days to flowering was positively correlated with days to maturity, but negatively correlated with total number of pods per plant and total seed weight, while days to maturity was negatively correlated with total seed weight. In contrast, Rohman et al. (2003) found that days to flowering showed negative

correlation with days to maturity but showed positive correlation with 100 seed weight and total seed weight. Rajan et al. (2000) worked in mungbean and found similar results to ours that total seed weight had positive genotypic correlation with pods per plant, seeds per pod and one hundred seed weight. Thus the genetic of total seed weight can be improved by indirectly selecting characters showing positive correlation (PDTN, SD100WT, PDL and SDNPPD), as well as negative correlation (FLD and PDDM). Our research can assign 150 SSR markers into 11 linkage groups, corresponding to the haploid number of mungbean chromosomes. In the previous research, Menancio-Hautea et al. (1992) used 171 RFLP markers to construct a map grouping into 14 linkage groups that span a total of 1,570 cM with an average distance of 9 cM. Humphry et al. (2002), clustered 255 RFLP probes into 13 linkage groups, with a total length of the map spanned 737.9 cM at an average distance between markers of 3.0 cM and a maximum distance between linked markers of 15.4 cM. While our SSR map has covered 1,174.2 cM, with the average distance between adjacent loci of 7.8 cM. Han et al. (2005) analyzed azuki bean genetic linkage map from a backcross population of (V. nepalensis × V. angularis) × V. angularis. They used 486 markers comprising 205 SSR, 187 AFLP and 94 RFLP to saturate the map. Their map covers altogether 11 linkage groups as our results, but spanned 832.1 cM with an average marker distance of 1.85 cM. Our result showed longer

82


genome coverage than both maps, with the longer average marker distance. Recently, Zhao et al. (2010) constructed a mungbean integrated map including 97 SSRs, 76 RFLPs, 4 RAPDs and 2 STSs. Their prime objective was to locate the bruchid-resistance Br1 gene. Among the SSR markers located on the map, 91 were from azuki bean, blackgram, common bean and cowpea. The linkage map spans 1,831.8 cM with the average marker distance of 10.2 cM. However, their marker names were presented as codes and thus were not available to the public. Considering the low polymorphism found in mungbean germplasm, our work is the most successful in developing an SSR linkage map resolving 11 mungbean linkage groups. Azuki bean SSR markers were used for constructing a black gram linkage map (Chaitieng et al. 2006), and mungbean linkage map in this study. Our map construction capitalized the co-linearity between genomes of the Asian Vigna, viz. mungbean, black gram and azuki bean. Although many azuki bean markers can be assigned into mungbean genome, the number of common markers were more conserved between mungbean and black gram than that between mungbean and azuki bean. This result supported the propose of Tomooka et al. (2002) that mungbean and black gram are in the same section Ceratotropis, while azuki bean is in section Angulares. Chaitieng et al. (2006) later reported highly co-linearity (88%) between black gram and azuki bean, revealing the use of

cross-species genetic markers among Vigna species. Our results showed internal insertion/deletion on LG1, 2 and 5 between mungbean and azuki bean maps, and on LG9 between mungbean and black gram maps. These results indicated that genomes in the subgenus Ceratotropis have accumulated a small number of insertions/deletions. The chromosome aberrations detected between mungbean, black gram and azuki bean linkage maps may play important roles in evolution among these species.

ACKNOWLEDGEMENTS This research was supported by the Commission on Higher Education under the Strategic Scholarships for Frontier Research Networks for Thai Doctoral Degree Program, Ministry of Education, Thailand and the National Science and Technology Development Agency, Thailand. We also thank the Center for Agricultural Biotechnology, Kasetsart University, Kamphaeng Saen Campus, Thailand for lab facilities.

REFERENCES Arumganathan K, Earle ED (1991). Nuclear DNA content of some important plant species. Plant Mol. Biol. Rep. 9: 208. Blair M, Pedraza F, Buendia H, Gaitán-Solís E, Beebe S, Gepts P, Thome J (2003). Development of a genomewide anchored microsatellite map of common bean (Phaseolus vulgaris L.). Theor. Appl. Genet. 107: 1362-1374.

83


Buso G, Amaral Z, Brondani R, Ferreira M (2006). Microsatellite markers for the common bean Phaseolus vulgaris. Mol. Ecol. Notes 6: 252-254. Chaitieng B, Kaga A, Tomooka N, Isemura T, Kuroda Y, Vaughan D (2006). Development of a black gram [Vigna mungo (L.) Hepper] linkage map and its comparison with an azuki bean [Vigna angularis (Willd.) Ohwi and Ohashi] linkage map. Theor. Appl. Genet. 113: 1261-1269. Chankaew S, Somta P, Sorajjapinum W, Srinives P (2011). Quantitative trait loci mapping of Cercospora leaf spot resistance in mungbean, Vigna radiata (L.) Wilczek. Mol. Breed. 28: 255-264. Churchill GA, Doerge RW (1994). Empirical threshold values for quantitative trait mapping. Genetics 138: 963-971. Fehr W (1987). Principles of Cultivar Development, vol. 1. Theory and Technique. McGraw-Hill Inc, New York. Fehr W, Caviness C, Burmood D, Pennington J (1971). Stage of development descriptions for soybeans, Glycine max (L.) Merrill. Crop Sci. 11: 929-931. Fernandez G, Shanmugasundaram S (1988). The AVRDC mungbean improvement program: The past, present and future. In S. Shanmugasundaram and B.T. McLean, eds., Mungbean: Proceeding of the Second International

Symposium. AVRDC, Shanhua, Taiwan, pp. 58-70. Gwag J, Chung J, Chung H, Lee J, Ma K, Dixit A, Park Y, Cho E, Kim T, Lee S (2006). Characterization of new microsatellite markers in mungbean, Vigna radiata (L.). Mol. Ecol. Notes 6: 1132-1134. Han OK, Kaga A, Isemura T, Wang X, Tomooka N, Vaughan DA (2005). A genetic linkage map for azuki bean (Vigna angularis (Willd.) Ohwi & Ohashi). Theor. Appl. Genet. 111: 12781287. Humphry ME, Konduri V, Lambrides CJ, Magner T, McIntyre CL, Aiken EB, Liu CJ (2002). Development of a mungbean (Vigna radiata) RFLP linkage map and its comparison with lablab (Lablab purpureus) reveals a high level of colinearity between the two genomes. Theor. Appl. Genet. 105: 160-166. Kasettranan W, Somta P, Srinives P (2010). Mapping of quantitative trait loci controlling powdery mildew resistance in mungbean, Vigna radiata (L.) Wilczek. J. Crop Sci. Biotech. 13(3): 155-161. Khattak G, Srinives P, Kim D (1995). Yield partitioning in high yielding mungbean (Vigna radiata (L.) Wilczek). Kasetsart J. (Nat Sci) 29: 494-497. Kosambi D (1944). The estimation of map distances from recombination values. Ann. Eugen. 12: 172-175. Kumar S, Tan S, Quah S, Yusoff K (2002a). Isolation of

84


microsatellite markers in mungbean, Vigna radiata. Mol. Ecol. Notes 2: 96-98. Kumar S, Tan S, Quah S, Yusoff K (2002b). Isolation and characterization of seven tetranucleotide microsatellite loci in mungbean, Vigna radiata. Mol. Ecol. Notes 2: 293295. Lambrides CJ, Lawn RJ , Godwin ID, Manners J, Imrie BC (2000). Two genetic linkage maps of mungbean using RFLP and RAPD markers. Aust. J. Agric. Res. 51: 415 – 425. Li C, Fatokun C, Ubi B, Singh B, Scoles G (2001). Determining genetic similarities and relationships among cowpea breeding lines and cultivars by microsatellite markers. Crop Sci. 41: 189-197. Lodhi MA, Ye GN, Weeden NF, Reisch BI (1994). A simple and efficient method for DNA extraction from grapevine cultivars and Vitis species. Plant Mol. Biol. Rep. 12: 6-13. Menancio-Hautea D, Kumar L, Danesh D, Young N (1992). A genome map for mungbean (Vigna radiata (L.) Wilczek) based on DNA genetic markers (2n=2x=22). In Genetic Maps Locus Maps of Complex Genomes. Edited by Stephen O’Brien. Cold Spring Habor Laboratory, pp. 6.259-6.261. Pookpakdi A, Promkham V, Chuangpetchinda C, Pongkao S, Lairungrueng C, Tawornsuk C (1992).

Growth stage identification in mungbean (Vigna radiata (L.) Wilczek). Kasetsart J. (Nat Sci) 26: 75-80. Rajan R, Wilson D, Vijayaraghava K Correlation and (2000). path analysis in the F2 generation of green gram (Vigna radiata (L.) Wilczek). Madras Agri. J. 87: 590-593. Rohman M, Hussain AI, Arifin S, Akhter Z, Hauzzaman M (2003). Genetic variability, correlation and path analysis in mungbean. Asian J. Plant. Sci . 2: 1209-1211. Seehalak W, Somta P, Sommanas W, Srinives P (2009). Microsatellite markers for mungbean developed from sequence database. Mol. Ecol. Resour. 9: 862-864. Siddique M, Faisal M, Malikand A, Awan S (2006). Genetic divergence, association and performance evaluation of different genotypes of mungbean (Vigna radiata). Int. J. Agri. Biol. 6: 793795. Somta P, Musch W, Kongsamai B, Chanprame S, Nakasathein S, Toojinda T, Sojjapinun W, Seehalak W, Tragoonrung S, Srinives P (2008). New microsatellite markers isolated from mungbean (Vigna radiata (L.) Wilczek). Mol. Ecol. Resour. 8: 1155-1157. Somta P, Sommanas W, Srinives P (2009). Molecular diversity assessment of AVRDC-The World Vegetable Center elite-parental mungbeans. Breed. Sci. 59: 149-157. Sriphadet S, Lambrides J, Srinives P (2007). Inheritance of agronomic traits and their

85


interrelationship in mungbean (Vigna radiata (L.) Wilczek). J. Crop. Sci. Biotech. 10: 249-256. Tangphatsornruang S, Somta P, Uthaipaisanwong P, Chanprasert J, Sangsrakru D, Seehalak W, Sommanas W, Tragoonrung S, Srinives P (2009). Characterization of microsatellites and gene contents from genome shotgun sequences of mungbean (Vigna radiata (L.) Wilczek). BMC Plant Biol. 9: 137. Tomooka N, Vaughan D, Moss H, Maxted N (2002). The Asian Vigna: Genus Vigna subgenus Ceratotropis genetic resources. Kluwer Academic Publishers, Dordrecht. Van Ooijen J (2009). JoinMap®3.0, software for the calculation of genetic linkage maps. Plant Research International, Wageningen, Netherlands. Wang X, Kaga A, Tomooka N, Vaughan DA (2004). The development of SSR markers by new method in plants and their application to gene flow studies in azuki bean (Vigna angularis (Willd.) Ohwi & Ohashi). Theor. Appl. Genet. 109: 352-360. Wang S, Basten CJ, Zeng Z-B (2007). Windows QTL Cartographer 2.5. Department of Statistics, North Carolina State University, Raleigh, NC. Weinberger K (2003). Impact analysis on mungbean research in South and Southeast Asia. Final Report GTZ Eigenmassnahme no.

99.9117.5. AVRDC, Shanhua, Taiwan. Young N, Kumar L, MenancioHautea D, Danesh D, Talekar N, Shanmugasundarum S, Kim D (1992). RFLP mapping of a major bruchid resistance gene in mungbean (Vigna radiata (L.) Wilczek). Theor. Appl. Genet. 84: 839844. Young N, Danesh D, MenancioHautea D, Kumar L (1993). Mapping oligogenic resistance to powdery mildew in mungbean with RFLPs. Theor. Appl. Genet. 87: 243-249. Zhao D, Cheng X, Wang L, Wang S, Ma YL (2010). Constructing of mungbean genetic linkage map. Acta Agron. Sin. 36(6): 932-939.

86

RESEARCH ARTICLE


MOLECULAR GENETIC DIVERSITY OF BAMBARA GROUNDNUT (Vigna subterranea L. Verdc.) REVEALED BY RAPD AND ISSR marker ANALYSIS O. RUNGNOI1, J. SUWANPRASERT2, P. SOMTA3* and P. SRINIVES3 1

Department of Plant Production Technology, King Mongkut’s Institute of Technology Ladkrabang, Bangkok 10520, Thailand 2 Songkhla Agricultural Research and Development Center, Hat-Yai, Songkhla 90110, Thailand 3 Department of Agronomy, Faculty of Agriculture at Kamphaeng Saen, Kasetsart University, Nakhon Pathom 73140, Thailand *Corresponding author email: [email protected]

SUMMARY A set of 363 Bambara groundnut (Vigna subterranea L. Verdc.) accessions from five geographical regions and unknown origin was assessed with 65 loci generated from inter simple sequence repeat (ISSR) and random amplified polymorphic DNA (RAPD) markers to provide more information on genetic diversity and origin. Higher percentage of polymorphic DNA bands, gene diversity and Shannon’s diversity index were found in accessions from Cameroon/Nigeria and West Africa regions but not much different from those found in accessions from West Africa. Genetic distance revealed close genetic relationships between accessions from Cameroon/Nigeria and West Africa regions. Hierarchical cluster analysis revealed that Thai Bambara groundnuts are clustered with both West and East Africa regions. Principal coordinate analysis demonstrated that accessions from Cameroon/Nigeria and West Africa were separated into two groups, where accessions from the two regions were mingled with each others. Although most accessions from East Africa were grouped with a cluster of Cameroon/Nigeria and West Africa accessions. Population structure analysis grouped the germplasm into two subpopulations where most accessions were in one subpopulation and 11 accessions were in the other subpopulation. Altogether, the results support the views that West Africa (including Cameroon and Nigeria) is the center of diversity of Bambara groundnut. The crop was introduced from West to East Africa, from where the Thai accessions were introduced. Key words: African Vigna; evolution; domestication; genetic diversity; DNA markers; legume Manuscript received: December 29, 2011; Decision on manuscript: April 4, 2012; Manuscript accepted in revised form: April 9, 2012. Communicating Editor: Ramakrishna Nair

87

Rungnoi et al. (2012)

INTRODUCTION The genus Vigna (Family Leguminosae) is an important legume taxon. It comprises about 90 described species of which seven species are cultivated as economic crops in various regions, while several species are cultivated as minor crops and some wildly grown species are harvested for food and feed. The important crops in the genus Vigna include azuki bean [Vigna angularis (Willd.) Ohwi and Ohashi], Bambara groundnut [Vigna subterranean (L.) Verdc.], blackgram [Vigna mungo (L.) Hepper], cowpea [Vigna unguiculata (L.) Walps], moth bean [Vigna aconitifolia (Jaqc.) Maréchal] mungbean [Vigna radiata (L.) Wilczek] and rice bean [Vigna umbellata (Thunb.) Ohwi and Ohashi]. Bambara groundnut and cowpea are originated in Africa, while the other five species are of Asian origin. Bambara groundnut is the third most important food legume of Africa. Its production is ranked after peanut and cowpea. The crop is usually intercropped with cereals, root and tuber crops in the marginal lands of Africa. It is one of the most drought-tolerant cultivated legumes, and is more tolerant to infertile soil than any other legumes (Chomchalow, 1993). The crop is also grown in some Asian countries such as India, Malaysia, the Philippines and Thailand. In Thailand, the crop is grown exclusively in the southern part of the country especially in the sandy soil areas where other crops fail to yield. Dried seeds of Bambara groundnut

contain 18-25% protein, 55-72% carbohydrate and 6-7% oil. The crop is a very important source of dietary protein for poor people in Africa who cannot afford expensive animal protein (Baryeh, 2001). Thus it has high potential for food security in unpredictable drought regions. Bambara groudnut was previously classified into the genus Voandzeia and later reclassified as the member of the genus Vigna (Verdcourd, 1970), although it’s morphological characters are much different from other Vigna species. A single pod of Bambara groundnut contains only one or two seeds and the pods are set underground. Average yield of Bambara groundnut is rather low compared with other cultivated Vigna crops. This is due mainly to the fact that all of Bambara groundnut cultivars grown are landrace. No improved cultivars were developed by a selective breeding program because an efficient hybridization technique just has been recently developed (Suwanprasert et al., 2006), and this enables cultivar development of Bambara groundnut become more feasible. Nonetheless, before setting up a breeding program for Bambara groundnut, understanding on its genetic diversity is necessary. Like many other orphan crops, there are only a few studies on genetic diversity in a large set of Bambara groundnut germplasm. Goli et al. (1997) and Olukolu et al. (2012) studied diversity based on seed patterns in 1,384 and 1,973 accessions, respectively. Olukolu et al. (2012) found that Bambara groundnut from Cameroon/Nigeria region has higher diversity than

88


those from the other geographical regions. Diversity study in 124 accessions using 28 quantitative traits and in 40 accessions using 554 Diversity Arrays Technique (DArT) markers by the same authors revealed the highest diversity in Cameroon/Nigeria region. The results support the view of Hepper (1963) that center of origin/domestication of Bambara groundnut is in the Cameroon/Nigeria region. In contrast, Somta et al. (2011) studied diversity in a collection of 240 Bambara groundnut accessions using 22 simple sequence repeat (SSR) markers found highest diversity in West African (excluding Cameroon and Nigeria). Therefore, the center of diversity and origin of Bambara groundnuts is still inconclusive and more evidence is needed to elucidate them. In this study we evaluated diversity in a collection of 363 Bambara groundnut accessions from several geographical origins using inter simple sequence repeat (ISSR) and random amplified polymorphic DNA (RAPD) markers with the objective to provide more evidences on diversity, and genetic relationships among different origins of the Bambara groundnut.

MATERIALS AND METHODS Bambara groundnut germplasm and DNA extraction Three hundred and sixty three landraces of Bambara groundnut accessions were used in this study, of which 327 accessions were from

14 African countries, 5 were from Southeast Asia (Thailand) and 31 were from unknown origin (Table 2, Supplementary data 1). The accessions were grouped based on geographical origins into six populations, i.e. Cameroon/ Nigeria, West Africa, East Africa, Central Africa, Southeast Asia, and unknown origin. This grouping was the same as those reported by Olukolu et al. (2012) and Somta et al. (2011). This enabled us to compare our results with these two studies. Total genomic DNA of each accession was extracted from young leaves using a CTAB method. DNA concentration was determined by comparing with a known concentration of λ DNA and adjusted to 20 ng/μl for PCR amplification. RAPD and ISSR marker analysis Initially, 95 RAPD and 20 ISSR primer pairs were tested for amplification ability and polymorphism using DNA from 4 accessions of Bambara groundnut TVSu 349, TVSu 351, TVSu 368 and TVSu 370. Primers that produced repeatable DNA bands and polymorphism were further used to analyze all 363 accessions. The PCR amplification of RAPD was performed as follow: PCR mixture was prepared in a total volume of 10 µl containing 20 ng of genomic DNA, 0.5 µM primer, 1× Taq buffer (750 mM Tris-HCl, 200 mM (NH4)2SO4, 0.1% Tween 20), 100 µM dNTPs, 3.125 mM MgCl2, and 1 U Taq DNA polymerase (Fermentas, Vilnius, Lithuania). The PCR cycling profile was 94oC for 2 min followed by 45 cycles of 94oC for

89


30 s, 35oC for 30 s (RAPD) or 58oC for 15 s (ISSR), 72oC for 1 min, ending with final extension at 72oC for 10 min. The amplification was performed using PTC-200 thermocycler (MJ Research, Waltham, USA). The PCR products were electrophoresed on 1.5% agarose gel in 0.5x TBE buffer and stained with ethidium bromide. DNA bands were visualized under UV light and photographed using a Gene Genius gel documentation system (Syngene, Frederick, MD, USA). Data analysis Polymorphic DNA bands were scored as 1 (present) or 0 (absent). Percentage of polymorphic bands (%P), gene diversity (HE; Nei, 1973), Shannon’s information index (I; Lewontin, 1972) of each marker locus was calculated using software PopGene32 version 1.32 (Yeh et al., 2000). Nei’s genetic distance (DS; Nei, 1972) among accessions from different regions and countries were computed using software GENALEX version 6.41 (Peakall and Smouse, 2006). To determine genetic relationship among accessions, an unweighted pair group method with arithmetic mean (UPGMA) tree was constructed based on Jaccard’s similarity coefficients (Jaccard, 1908). The coefficients were further used in a principal coordinate analysis (PCoA). The UPGMA and PCoA analyses were performed with software NTSYSpc 2.2 (Rohlf, 2005). To infer clusters of individuals, clustering was performed using a model-based Bayesian algorithm of software STRUCTURE 2.3.2 (Pritchard et

al., 2007) with recessive allele model for dominant markers (Falush et al., 2007). Twenty independent simulations of the STRUCTURE were run using number of subpopulations (K) between 1 and 10 with a burn-in period of 10,000 and 50,000 iterations of Markov Chain Monte Carlo (MCMC) algorithm. The optimum K was determined by ΔK method as described by Evanno et al., (2005). Then, the STRUCTURE was run based on the determined optimum K with burn-in period of 100,000 and 500,000 iterations of MCMC algorithm to assign each accession to a cluster.

RESULTS Marker polymorphism Among 95 RAPD primers screened in 4 Bambara groundnut accessions, 19 pairs produced polymorphic DNA bands of which 14 were consistently reproducible (data not shown). For ISSR, 3 out of 20 primers produced polymorphic DNA bands. Seventeen primers, 14 RAPD and 3 ISSR, were chosen based on the number of polymorphic bands, and used for further analysis (Table 1). RAPD analysis in the 363 Bambara groundnut accessions revealed that number of reproducible DNA bands per primer ranged from 3 (OPF-12) to 11 (OPAA-16) with an average of 5.42. The number of polymorphic bands per primer was between 2 (OPF-12) and 6 (OPAA-16 and OPAB-14) with an average of 3.64 T 1). Percentage of polymorphic DNA bands varied from 50% to

90


100% with an average of 70.1%. ISSR analysis in the 363 accessions showed that number of DNA bands per primer ranged from 5 (ISSR-11) to 7 (ISSR-18 and ISSR-20) with an average of 6.33 (Table 1). The number of polymorphic bands per primer was between 3 (ISSR-11) and 6 (ISSR18) with an average of 4.67. Percentage of polymorphic DNA bands varied from 60% to 85.7% with an average of 72.4%. Diversity within populations and genetic distance among populations The percentage of polymorphic DNA bands produced by all 17 primers analyzed in 363 Bambara groundnut accessions was 58.21% (Table 2). The percentage of polymorphic DNA bands was highest in the Cameroon/Nigeria population (90.77%) followed by West Africa (75.38%) and East Africa (72.31%) populations, and was lowest in the Southeast Asia population (21. 50%). The Cameroon/Nigeria population also had greatest gene diversity (HE = 0.266) and Shannon’s diversity index (I = 0.417), followed by West Africa population (HE = 0.229, I = 0.357) and East Africa population (HE = 0.198, I = 0.315) (Table 2). Southeast Asia population had lowest HE and I being 0.090 and 0.129, respectively. The values were more or less similar to those of Central Africa population, 0.104 and 0.159, respectively. Within Cameroon/Nigeria region, the Nigerian accessions had more diversity than the Cameroonian accessions (%P = 84.62, HE = 0.255, I = 0.398 vs.

%P = 63.08, HE = 0.208, I = 0.318). In the West African region, the Ghanaian accessions possessed highest diversity with %P = 70.77, HE = 0.269 and I = 0.398. They also showed greater gene diversity than accessions from the other countries. In the East African region, the Zimbabwean accessions possessed highest gene diversity and Shannon’s diversity index. However, the diversity was not much different from those Zambian accessions. Genetic distance among the six populations is shown in Table 3. The least genetic distance was between Cameroon/Nigeria and West Africa populations (0.010) indicating a close genetic relationship between the two populations. While greatest genetic distance was between Central Africa and Southeast Asian populations (0.063) indicating distant genetic relatedness between these two populations. Among the African populations, the highest genetic distance was between East Africa and Central Africa populations being 0.044, although the figure was almost the same as that of between West Africa and Central Africa populations being 0.042. Among the African populations, the East African population showed the least genetic distance with Southeast Asian population. Genetic distance among accessions from different countries is summarized in Table 4. Within the Cameroon/Nigeria region, genetic distance between accessions from Nigeria and Cameroon was relatively low (DS = 0.030). Accessions from Cameroon were most closely related to those

91


from Central Africa Republic, followed by those from Togo, but most distantly related to accessions from Madagascar. Accessions from Nigeria were most closely related to accessions from Togo (DS = 0.016), followed by Burkina Faso (DS = 0.025), and most distantly related to those from Madagascar. For the West African accessions, Burkina Faso and Togolese generally showed close genetic relationship with accessions from other countries, whereas Guinean accessions always displayed great genetic differentiation from accessions from the other countries. Of East African accessions, the Madagascan accessions were more genetically differentiated from those of other regions than Tanzanian, Zambian and Zimbabwean accessions. They were most similar to accessions from Zambia and least similar to accessions from Ghana. Zambian and Zimbabwean accessions were closely related to both Nigerian and Burkina Faso accessions. Central African accessions were most genetically differentiated from Madagascan accessions, but most closely related to accessions from Cameroon. Thai accessions had the highest genetic distance from Guinean and Madagascan accessions, and showed closest genetic relationship with Zambian and Zimbabwean accessions. UPGMA, principal coordinate and STRUCTURE analyses A dendogram generated by UPGMA cluster analysis failed to illustrate clear pattern of germplasm groups (Figure 1). In most cases, accessions from different regions or countries were

clustered with one another. However, it demonstrated that three of the five accessions (THA1, THA4 and THA5) from Southeast Asia clustered with accessions from West Africa, whereas the other two accessions (THA2 and THA3) were clustered with accessions from East Africa (data not shown). It also showed close relatedness between THA1 and THA4. Principal coordinate analysis (PCoA) was performed to reveal genetic relationship among the Bambara germplasm. The first, second and third PCs together accounted for 90.3% of the total variation with each PC explaining 87.0%, 2.0% and 1.3% variation in that order. PCoA plot based on PC1 and PC2 showed that the Bambara groundnut accessions are separated into two major groups, each at the right and left of the plot (Figure 2a). About each half of the accessions from Cameroon/Nigeria and from West Africa were at the right and left of the PCoA plot (Figure 2, b&c). Most of the accessions from the two regions were overlapped demonstrating close genetic relationship. However, the West African accessions showed wider distribution on the plot. Of the Cameroon/Nigeria region, Nigerian accessions scattered wider than Cameroonian accessions. For the West African region, Togolese accessions were the most diverse. Majority of the accessions from Central Africa located at the lower left of the plot and overlapped with Cameroon/Nigeria and West Africa populations (Figure 2d). One Central African accession was clearly distinct from the others.

92


Majority of the East African accessions were at the right of the plot, and overlapped with the accessions from Cameroon/Nigeria and West Africa (Figure 2e). The Southeast Asian (Thai) accessions located at the upper-middle of the PCoA plot (Figure 2f). One accession was relatively distinct from the other four accessions. Accessions with unknown origins scattered widely on the plot (Figure 2f). Determination of optimum K using ΔK method after simulation runs of the STRUCTURE clustered the

germplasm into two subpopulations (Figure 3). Majority of the accessions were clustered into subpopulation I, whereas only 13 accessions were clustered into subpopulation II (Figure 3). These 13 accessions were from 7 countries (data not shown). Since the ΔK method for STRUCTURE analysis does not support the evaluation of the populations at K = 1 (see Evanno et al. 2005 for details) it is possible that only a single population represents the 363 Bambara groudnut accessions.

Table 1. List of polymorphic RAPD and ISSR primers used for diversity assessment in 363 Bambara groundnut accessions. Primer name Primer sequence No. of No. of (5’-3’) reproducible polymorphic DNA bands DNA bands (%) produced OPA-04 AATCGGGCTG 5 3 (60) OPAA-16 GGAACCCACA 11 6 (54.6) OPAB-14 AAGTGCGACC 8 6 (75) OPAL-08 GTCGCCCTCA 3 3 (100) OPAU-03 ACGAAACGGG 4 3 (75) OPC-02 GTGAGGCGTC 6 3 (50) OPC-05 GATGACCGCC 4 4 (100) OPC-08 TGGACCGGTG 5 4 (80) OPD-20 ACCCGGTCAC 6 3 (50) OPF-12 ACGGTACCAG 3 2 (66.7) OPF-13 GGCTGCAGAA 5 4 (80) OPH-14 ACCAGGTTGG 5 3 (60) OPH-15 AATGGCGCAG 6 3 (50) OPK-16 GAGCGTCGAA 5 4 (80) ISSR-11 AGCAGCAGCAGCAGC 5 3 (60) ISSR-18 CACACACACACACACA 7 6 (85.7) ISSR-20 CGTCGTCGTCGTCGTCGTG 7 5 (71.4) Total 95 65 (68.4) Average 5.6 3.8 (70.5)

93


Figure 1. A UPGMA tree generated by Jaccard’s similarity coefficients showing relationships among 363 Bambara groundnuts.

94


Figure 2. A PCoA plot depicting relationship among 363 Bambara groundnut accessions from different regions. See supplementary data 1 for abbreviation of the country names.

95

Figure 3. Clusters of populations of 363 Bambara groundnut accessions based on STRUCTURE analysis with 65 RAPD and ISSR loci.

96


Table 2. Percentage of polymorphic DNA bands (P), Shannon’s diversity index (I) and gene diversity (HE) in 363 Bambara groundnut accessions detected by 51 RAPD and 14 ISSR marker loci. Region

Country

N1

P (%)

HE

I

108

90.77

0.266

0.417

Cameroon (CMR)

34

63.08

0.208

0.318

Nigeria (NGR)

74

84.62

0.255

0.398

79

75.38

0.229

0.357

Cameroon/Nigeria

West Africa

Central Africa

Benin (BEN)

1

-

-

-

Burkina Faso (BFA)

21

40.00

0.133

0.202

Ghana (GHA)

20

70.77

0.269

0.398

Guinea (GIN)

3

12.31

0.043

0.065

Mali (MLI)

4

16.92

0.066

0.096

Senegal (SEN)

5

18.46

0.070

0.103

Togo (TGO)

25

44.62

0.155

0.234

Central Africa Republic (CAF)

18

33.85

0.104

0.159

121

72.31

0.198

0.315

2

16.92

0.070

0.102

East Africa Madagascar (MDG)

4

16.92

0.063

0.093

Zambia (ZMB)

73

63.08

0.171

0.274

Zimbabwe (ZWE)

42

58.46

0.193

0.294

Southeast Asia

Thailand (THA)

5

21.50

0.090

0.129

Unknown

NA

32

56.92

0.193

0.294

363

100

0.179

0.227

Total 1

Tanzania (TZA)

number of accessions

97


Table 3. Nei’s genetic distance (DS; Nei 1972) among Bambara groundnut populations from 6 geographical regions. Cameroon/Nigeria West Africa Central Africa East Africa NA1 West Africa 0.010 Central Africa 0.032 0.042 East Africa 0.022 0.025 0.044 NA 0.019 0.018 0.042 0.018 Southeast Asia 0.046 0.047 0.063 0.036 0.045 1

unknown origin

Table 4. Nei’s genetic distance (DS; Nei 1972) among Bambara groundnut populations from 14 countries. CMR

NGR

BFA

GHA

GIN

MLI

SEN

TGO

CAF

MDG

TZA

NGR

0.030

BFA

0.038

0.025

GHA

0.034

0.030

0.040

GIN

0.058

0.041

0.035

0.072

MLI

0.038

0.036

0.026

0.050

0.030

SEN

0.042

0.033

0.028

0.064

0.027

0.037

TGO

0.029

0.016

0.017

0.033

0.029

0.025

0.026

CAF

0.025

0.044

0.048

0.060

0.061

0.044

0.042

0.045

MDG

0.065

0.064

0.077

0.100

0.092

0.075

0.086

0.065

0.094

TZA

0.052

0.056

0.049

0.094

0.056

0.040

0.054

0.048

0.058

0.063

ZMB

0.034

0.031

0.029

0.063

0.057

0.045

0.042

0.031

0.050

0.056

0.023

ZWE

0.036

0.026

0.032

0.051

0.065

0.052

0.045

0.034

0.043

0.071

0.048

NA THA 1

ZMB

ZWE

NA1

0.015

0.032

0.023

0.017

0.043

0.046

0.037

0.031

0.020

0.042

0.068

0.045

0.020

0.026

0.057

0.047

0.046

0.076

0.091

0.065

0.055

0.048

0.063

0.084

0.079

0.039

0.037

0.045

unknown origin

98


DISCUSSION Previous diversity studies using morphological characters (seed traits) and/or DNA markers in Bambara groundnuts showed highest diversity in the germplasm from Cameroon/Nigeria region (Goli et al., 1997, Olukolu et al., 2012) or West Africa (Somta et al., 2011). In this study, although RAPD and ISSR analyses revealed comparatively more diversity in Cameroon/Nigeria region, the level of diversity is not much different from that in West Africa region. It is worth noting that the number of accessions from Cameroon/Nigeria was relatively higher than that from West Africa. These results support the suggestion by Somta et al., (2011) that the center of diversity of Bambara groundnut is in West Africa (including Cameroon and Nigeria). Based on the 201 DArT marker analysis in a set of 429 cultivated Bambara groundnut accessions from various origins, Stadler (2010) showed that the Bambara groundnuts are unambiguously divided into two subgenepools, viz. upper equator and lower equator. RAPD and ISSR analysis in the present study differentiated 363 Bambara groundnuts into two major groups and showed association of Bambara groundnuts from upper equator (Cameroon/Nigeria, West Africa and Central Africa) and lower equator regions (East Africa). This is in line with our previous finding using 22 SSR markers analysis in 240 accessions (Somta et al., 2011). The contrast results may stem from the different DNA markers, number of loci, and

germplasm used. Nonetheless, all results support the view of Somta et al. (2011) that Bambara groundnut is a noncentric or oligocentric crop. This is further supported by the results from population structure analysis. It has been suggested that East African Bambara groundnuts originate from West Africa (Hepper, 1963; Olukolu et al., 2012; Somta et al., 2011). Our results based on PCoA scattered plot support the suggestion in that the East African accessions were separated into two (major and minor) groups, both of which were clustered with West African accessions (including Nigeria and Cameroon). This is also supported by small genetic distance between East African accessions and West African or Cameroonian/Nigerian accessions. Among the African Bambara groundnuts, accessions from Madagascar are the most isolated from the center of diversity/origin. In this study, Madagascan Bambara groundnuts showed lowest genetic distance with Zambian Bambara groundnut, suggesting that Madagascan accessions are probably introduced from Zambia. This is likely due to the proximity of the two countries. Similar result is reported by Stadler (2010). Bambara groundnut is also cultivated in some restricted parts of India, Indonesia, Malaysia, the Philippines, and Thailand. In this study, Bambara groundnuts from Thailand were included to trace for their probable origins. Two and three accessions from the country were found associated with East and West African accessions,

99


respectively. Similarly, Somta et al., (2011) illustrated that Thai Bambara groundnuts originate from both West Africa (Nigeria) and East Africa. These results imply that Bambara groundnuts was introduced to Thailand more than once. However, Stadler (2010) showed that landrace Bambara groundnuts from Indonesia have a closest genetic relationship with East African (Tanzanian) accessions. The results together suggest different routes of Bambara groundnut introductions from Africa to Southeast Asia. Additional study using more germplasm from different Asian countries can provide necessary evidence to support this view. Since the RAPD and ISSR markers revealed low genetic base within the landrace Bambara groundnuts, cultivar development by hybridization using such germplasm is difficult. Thus techniques such as mutation breeding and interspecific hybridization can be alternative means to broaden genetic variations in breeding for new cultivars of this crop.

ACKNOWLEDGEMENTS This study was financially supported by Center for Advanced Studies for Agriculture and Food, Kasetsart University under the National Research University Program of Office of the Higher Education Commission, Ministry of Education, Thailand. P. Somta is thankful to the Thailand Research Fund for providing him the Research Grant for New Scholar.

REFERENCES Baryeh

EA (2001). properties of

Physical bambara

groundnuts. J. Food Eng. 47: 321–326. Chomchalow N (1993). Bambara groundnut. In Proceedings of the FAO/UNDP project RAS/89/040 workshop on underexploited and potential food legumes in Asia. Ed. by Gowda CLL, Laosuwan P, Food and Agriculture Organization of the United Nations Regional Office for Asia and the Pacific, Bangkok, Thailand, pp. 30– 34. Evanno G, Regnaut S, Goudet J (2005). Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Mol. Ecol. 14: 2611–2620. Falush D, Stephens M, Pritchard JK (2007). Inference of population structure using multilocus genotype data: dominant markers and null alleles. Mol. Ecol. Resour. 7: 895–908. Goli AE, Begemann F, Ng NQ (1997). Characterization and evaluation of IITA’s bambara groundnut collection. In Proceedings of the workshop on conservation and improvement of bambara groundnut (Vigna subterranea (L.) Verdc.). Eds Heller J, Begemann F, Mushonga J, Institute of Plant Genetics and Crop Plant Research, Gatersleben/Department of Research and Specialist Services, Harare/International Plant Genetic Resources Institute, Rome, Italy, pp. 101–118.

100


Hepper FN (1963). Plants of the 1957–58 West African expedition. II. The Bambara groundnut (Voandzeia subterranea) and Kersting’s groundnut (Kerstingiella geocarpa) wild in West Africa. Kew Bull. 16: 395–407. Jaccard P (1908). Nouvelles recherches sur la distribution florale. Bul. Soc. Vaudoise Sci. Nat. 44: 223–270. Lewontin RC (1972). The apportionment of human diversity. Evol. Biol. 6: 381– 398. Nei M (1973). Analysis of gene diversity in subdivided populations. Proc. Natl. Acad. Sci. USA. 70: 3321– 3323. Nei M. (1972). Genetic distance between populations. Am. Nat. 106: 283–292. Olukolu BA, Mayes S, Stadler F, Ng NQ, Fawole I, Dominique D, Azam-Ali SN, Abbott AG, Kole C (2012). Genetic diversity in Bambara groundnut (Vigna subterranea (L.) Verdc.) as revealed by phenotypic descriptors and DArT marker analysis. Genet. Resour. Crop Evol. 59: 347– 358 Peakall R, Smouse PE (2006). GENALEX 6: genetic analysis in Excel. Population genetic software for teaching and research. Mol. Ecol. Notes 6: 288– 295. Pritchard JK, Wen X, Falush D (2007). Documentation for STRUCTURE software: Version 2.2. University of Chicago, Illinois.

Rohlf

FJ (2005). NTSYS-pc: numerical taxonomy and multivariate analysis system, version 2.2. Exeter Software, Setauket, NY. Somta P, Chankaew S, Rungnoi O, Srinives P (2011). Genetic diversity of the Bambara groundnut (Vigna subterranea (L.) Verdc.) as assessed by SSR markers. Genome 54: 898–910. Stadler F (2010). Analysis of differential gene expression under water-deficit stress and genetic diversity in bambara groundnut [Vigna subterranea (L.) Verdc.] using novel high-throughput technologies. Ph.D. Thesis, Technology University of Munich, Germany. Suwanprasert J, Toojinda T, Srinives P, Chanprame S (2006). Hybridization technique for Bambara groundnut. Breed. Sci. 56: 125–129. Verdcourt B (1980). The correct name for the Bambara groundnut. Kew Bull. 35: 474. Yeh FC, Yang R, Boyle TJ, Ye Z, Xiyan JM (2000). POPGENE 32, Microsoft Window-based Freeware for Population Genetic Analysis, version 1.32. Molecular Biology and Biotechnology Centre, University of Alberta, Edmonton, Canada.

101

RESEARCH ARTICLE SABRAO Journal of Breeding and Genetics 44 (1) 102-111, 2012

GENOTYPE X ENVIRONMENT INTERACTION AND STABILITY OF YIELD COMPONENTS AMONG RICE GENOTYPES IN RIAU PROVINCE, INDONESIA ASLIM RASYAD1*, GULAT M.E. MANURUNG1 and DAVID A. VAN SANFORD2 1

2

Department of Agrotechnology, University of Riau, Indonesia Department of Plant and Soil Sciences, The University of Kentucky, USA * Corresponding author email: [email protected]

SUMMARY Genotype by environment (GE) interactions and genotype stability of a trait in rice (Oryza sativa L) are very important information for plant breeders in making decisions regarding the development and evaluation of new cultivars as well as for farmers in selecting suitable cultivars to be planted for commercial purpose. Four locally adapted cultivars and one high yielding cultivar of rice were evaluated at three environments in Riau, Indonesia. Traits measured included panicles plant-1, number of grains panicle-1, filled grain percentage, seed weight and grain yield. A regression coefficient and deviation from regression as proposed by Eberhart and Russel (1966) were used to determine stability of a genotype. There were significant effects of environments on yield and some yield components, except panicle number plant-1. The cultivars differed significantly in all yield components and grain yield. Cisokan and IR64 had b value did not significantly (P < 0.05) different from 1.0 and had s(d) approaching to zero and were considered as stable for all environments. Kwatik Putih was considered as unstable genotype and specifically adapted to less favorable environment. Key words: genotype x environment interaction, yield components, stability, rice cultivars Manuscript received: December 24, 2011; Decision on manuscript: April 9, 2012; Manuscript accepted in revised form: April 28, 2012. Communicating Editor: Bertrand Collard

INTRODUCTION In Riau, the majority of rice is grown by smallholder farmers and varieties used are locally adapted cultivars which are developed by farmers with unknown pedigree; as

a result a specific cultivar may be a mixture of different genotypes. High yielding cultivars released by plant breeders are particularly unacceptable to farmers in Riau due to adaptability, growth habit and consumer’s need for rice with

102

Rasyad et al. (2012)

good taste. In Riau and other central parts of Sumatra, consumers favor to consume aromatic unsticky rice varieties. Generally, farmers in Riau plant rice in tidal swampy and peat soil, so cultivars must meet certain criteria to perform well under these conditions, such as tall stature, erect stem and rapid growth. As soil type and weather varies from one area to another, there is possibility of variation in environments; therefore, yield performance of the cultivars might be influenced by the environments and to some degree by genotype by environment (GE) interaction. In addition, because the ancestors of the cultivars were unknown, yield performance may vary among genotypes depending on how they used cultural practice and how they preserved seed for next planting season. Recently, Reichardt, et al. (2001) and Witt, et al. (2001) presented evidence confirming a greater role of environment and cultural practices in rice yield. This evidence seemed to support the perception of most of the community in Riau and the central part of Sumatra who generally believe that location has strongly determined yield and yield quality in rice cultivars (Anhar and Leilani, 2001). Several workers have reported the importance of GE interaction on yield of various cereal crops (Saeed and Francis., 1984; Oosterom et al., 1993; Harsanti et al., 2003). Specifically, Ceccarelli (1987) reported that relatively low temperature and low irradiance reduced yield of barley significantly indicating the important role of environment on

yield. Under these conditions, he reported that GE interactions are quite large. Saeed et al., (1984) after evaluating some genotypes of sorghum had also reported the important role of GE interaction which changed the relative performance of cultivars from one environment to another. When superiority of the cultivar differs within different environment, it is important to quantify GE interaction in order to develop an efficient testing and selection procedure. Ideally, such a procedure could be used to develop cultivars that can be grown in a wide environment. Yield stability and effect of GE interaction has been mentioned in several studies in Indonesia. Recently, Harsanti et al. (2003) noticed the occurrence of GE interaction in yield of rice genotypes and found two of his mutant lines that were stable under a wide range of environment. Haryanto et al. (2008) revealed that yield stability of some aromatic upland rice genotypes differed over different environments. He noticed that four lines had high stability and were potential as candidates of new aromatic upland rice. More recently, Suwarto and Nasrullah (2011) after evaluating iron content on rice genotypes in ten environments in central java, found that environments effect contribute more than genetic and GE interaction and at least one cultivar i.e. Barumun had highest iron content as well as its stability. This study was intended to determine the magnitude of environment and GE interaction variance components and its contribution on yield and yield

103


components on locally adapted cultivars of rice and to look at the stability of the cultivars on different environments. The stability measure was calculated according to Eberhart and Russell (1966) in which ideal genotypes would have a high mean yield performance over a wide of environments, coefficient of regression equal to one and deviation of mean square from regression aproaching zero. This method was utilized due to its simplicity and being able to minimize interaction between genotype and its environment.

MATERIALS AND METHODS Four local genotypes commonly grown in Riau due to their aroma and good taste and one high yielding variety of rice were planted in three environments. The genotypes including Cisokan, Karya, Kwatik Putih, Seratus Hari and an IRRI released genotype (IR64) were considered fixed factor. The local cultivars were farmer’s derived genotypes with unknown pedigree by continually preserving seed from their own crop, so a specific cultivar may be a mixture of different genotypes. A field experiment was established by arranging genotypes in a completely randomized block design with three replicates in every environment. The environments were in Bunga Raya, Kampar and Rimba Melintang which have diverse environmental condition (Table 1). Kampar is characterized by slightly low average daily temperature and precipitation, located at more than

30 m above sea level (ASL) and the soil was wetland paddy soil. Bunga Raya has a peat swampy soil, and Rimba Melintang has a tidal swampy soil. Both of these are located below 10 m ASL, with higher daily mean temperature and precipitation compared to Kampar. The experiments were conducted in dry planting season of 2007; however, the area used for the experiments had enough irrigation system which enables sufficient supply of water during the plant growth. Seedlings of each genotype were kept at wet seedbed nursery for 21 days and then transplanted to a plot of three meters wide and four meters length. Planting density was 16 plants per m-2. Nitrogen fertilizer was applied at the rate of 90 kg N ha-1 in the form of urea, split into three applications in equal amount, where one third was given at the time of transplanting, another portion at 21 days after transplanting (DAT) and the remaining was applied at 35 DAT. Phosphorous and potassium, at the rate of 60 kg P2O5 and 60 kg K2O ha-1 respectively, were applied at the time of transplanting. At harvest, we obtained number of panicles plant-1, number of grain panicle-1, filled grain percentage, and 1000-grain weight from each plot. Grain yield was also observed in a plot basis then was converted into yield per hectare. Analysis of variance was established for individual environment to test the homogeneity, then a combined analysis of variance was performed, considering both environments and genotypes as fixed by using SAS package (SAS,

104


2004), so that significance of all effects were tested against mean square of error. Stability was determined by calculating the linear regression of a genotype’s mean on the overall mean yield of each environment as described by Eberhart and Russell (1966) and by arranging yield rank in every environment.

RESULTS AND DISCUSSION Mean squares of panicles number plant-1, number of grains panicle-1, filled grain percentage, 1000-grain weight, and grain yield are presented in Table 2. Environment effect was significant for grain panicle-1, filled grain percentage, grain weight and grain yield, but not for panicle number plant-1. Genotypes differed significantly for all traits, and the environment x genotypes interaction effect was significant for all traits. These results implied that genotypes and GE interaction accounted for a greater contribution to these traits than environment effect. Greater role of GE interaction may suggest that the genotypes performed differently under diverse environments and their performance was unpredictable across environments as stated by Ceccarelli et al. (1994). Since GE interaction was significant in this study, farmers need to be cautious in selecting a variety to be grown in their area of production. They must decide whether to plant a widely adapted or a locally adapted cultivar. To choose a widelyadapted variety, the farmers need a variety which is stable across locations.

For yield components evaluated, the average number of panicles plant-1 was not significantly different between environments for all genotypes (Table 3). Grain number panicle-1 was significantly different between environments for each genotype. Filled grain percentage was significantly different between environments for Karya and IR64 but not for the other genotypes, while grain weight was only significant between environments for IR64. Both Karya and IR64 had significantly lower percentage of filled grain in Rimba Melintang compared to the two other environments. It is probable that the high mean daily temperature followed by frequent rainfall with low light intensity in Rimba Melintang compared to Bunga Raya and Kampar (Table 1) reduced grain filling and limited the percentage of filled grains of the two genotypes. This is consistent with previous reports by Seshu and Cady (1984), Singh (2005) and Kobata et al. (2006) that high night temperature reduced assimilates supply to the seed of some genotypes during grain filling especially when followed by low light intensity. Yield performance revealed wide variation in cultivars between locations (Table 3). Karya, IR64 and Seratus Hari produced significantly greater grain yield in Bunga Raya and Kampar than that in Rimba Melintang. Meanwhile, there was no yield difference among locations for Cisokan and Kwatik Putih. Reduced filled grain percentage for Karya and IR64 and number of grain per panicle for Seratus Hari

105


contributed to lower grain yield of the cultivars in Rimba Melintang. Cultivar mean yield ranged from 4.89 to 6.06 in Kampar, from 4.77 to 6.82 in Bunga Raya and from 3.86 to 4.52 tons per hectare in Rimba Melintang, respectively (Table 4). The yield ranks of IR64 and Cisokan were in the top two across three locations thus they may be considered stable, high yielding cultivars. According to Eberhart and Russell’s (1966) stability measure, ideal cultivars would have a high mean yield performance over a wide of environments, a coefficient of regression equal to one and deviation of mean square from regression equal to zero. In the analysis conducted on these cultivars, the regression of cultivar mean yield on the environmental index resulted in regression coefficients ranging from -0.276 to 1.699 and standard deviations ranging from 0.234 to 1.242. Of the five cultivars, only two cultivars had b value not significantly (P < 0.05) different from 1.0 and had s(d) approaching to zero (Table 5). Therefore; these cultivars, Cisokan and IR64 were confirmed as the ideal cultivars to be grown in three locations, and Kwatik Putih was the ideal cultivar for a less favorable environment such as Rimba Melintang. Karya and Seratus Hari were two cultivars which have low yielding ability, are less stable and would not be ideal to be grown even in less favorable environment such as in Rimba Melintang. IR64 is a high yielding variety developed by IRRI which is broadly adapted and is recommended for most Indonesian

rice production area. It was understandable that the performance of IR64 were better than that local varieties as confirmed studies by Harsanti, et al. (2003) and Suwarto and Nasrullah (2011). Even though IR64 was a stable cultivar according to our stability analysis, in reality the variety has not been grown by most farmers in this area of production. This cultivar has physical characters such as stickiness when cooked and an aroma unacceptable to consumers in Riau. Furthermore Cisokan is the only stable cultivar which could be recommended for the area. This observation supports the idea that involvement of farmers in a plant breeding program is very crucial in this region. This sort of approach to plant breeding is called participatory plant breeding which involves farmer-researcher collaboration. The interaction between farmers and researchers/breeders can increase the efficiency of the breeding program (Ceccarelli and Grando, 1997, 2005).

106


Table 1. Altitude, daily temperature and precipitation at the time of rice plant growth to grain maturity stage in three environments. Environmental condition

Kampar

Bunga Raya

32.0

9.0

Rimba Melintang 7.5

23 30 97 3507 10

25 32 114 2781 22

26 33 169 2677 17

Altitude Avg. daily temperature • Minimum (0C) • Maximum (0C) Precipitation (mm month-1) Mean annual rainfall (mm) Raindays (days month-1)

Table 2. Mean square from analysis of variance for grain yield and yield components of several rice cultivars evaluated at three environments. Source of Variation Environment (E) Rep/E Genotypes (G) GxE Error

df

Panicles plant-1

2

0.45ns

6 4 8 24

11.41 52.90** 15.02** 3.06

Mean squares Filled Grain grains weight (%) (g 1000-1) 957.18** 50.01* 7.53** Grains panicle-1

123.58 6021.97** 697.22** 125.84

15.27 85.03** 106.66** 15.44

1.61 94.56** 7.02** 1.43

Grain yield (Mg ha-1) 10.23** 0.29 3.34** 2.03** 0.48

** - significant (P < 0.05); ns - not significant

107


Table 3. Genotypes mean of yield components and grain yield of rice cultivars evaluated at three environments Genotypes

Environment

Cisokan

Kampar Bunga Raya Rimba Melintang Kampar Bunga Raya Rimba Melintang Kampar Bunga Raya Rimba Melintang Kampar Bunga Raya Rimba Melintang Kampar Bunga Raya Rimba Melintang

Kwatik Putih

Karya

IR64

Seratus hari

Panicles plant-1

Grains panicle-1

22.4 a 22.3 a 21.0 a

119.2 a 108.9 b 99.5 b

Filled grains (%) 78.5 a 85.3 a 88.4 a

20.4 a 22.2 a 23.3 a

154.9 b 169.0 a 185.2 a

22.8 a 22.3 a 20.0 a

Grain weight (g 1000)

Grain yield (t.ha-1)

19.0 a 23.7 a 21.7 a

5.45 a 5.59 a 4.83 a

85.2 a 80.5 a 72.6 b

17.7 a 15.6 a 16.6 a

5.51 a 4.89 a 5.52 a

127.6 b 139.9 a 114.1 b

87.0 a 80.7 b 76.4 b

20.3 a 21.5 a 21.3 a

4.89 a 5.28 a 3.86 b

18.3 a 15.7 a 14.4 a

125.3 b 164.9 a 165.4 a

86.9 a 85.5 a 68.7 b

25.7 a 26.7 a 22.7 b

6.06 a 6.82 a 4.12 b

17.83 a 19.44 a 23.86 a

126.46 a 114.41 b 99.41 b

85.85 a 88.92 a 89.04 a

22.56 a 24.16 a 23.86 a

5.42 a 4.77 a 3.90 b

For each genotype, mean followed by the same letter among environments is not significantly different (P < 0.05) using LSD.

108


Table 4. Average grain yield of some genotypes of rice and its rank within each environment

Genotype Cisokan Kwatik Putih Karya IR64 Seratus hari

Kampar Yield Rank 5.56 2 5.51 3 4.89 5 6.06 1 5.42 4

Environment Bunga Raya Yield Rank 5.50 2 4.89 4 5.28 3 6.82 1 4.77 5

Rimba Melintang Yield Rank 4.23 3 4.52 1 3.86 5 4.38 2 3.90 4

Table 5. Genotypes means and stability measurements (β; Sdi) of grain yield of five genotypes of rice Cultivar Cisokan Kwatik Putih Karya IR64 Seratus Hari

Mean yield (t ha-1) 5.256 5.497 4.577 5.677 4.427

β

Sdi

0.868 1.699* -0,276* 1.059 0.650

0.234 1.153* 0.298 0.762 1.242*

* β significantly different from 1 and Sdi significantly different from 0 (P < 5%)

We conclude that farmers in these regions are aware of their traditional/common knowledge in agricultural production systems, which has already been adapted to specific environment such as peat swampy and tidal swampy land. Secondly, the farmers are aware of their own and consumer’s preference that make unreasonable to grow cultivars that lack these traits. Finally, farmers in Riau have attempted to maintain biodiversity in their communities by continually propagating material from their own production or farmer-tofarmer exchange.

REFERENCES Anhar

A, Leilani I (2001). Sustainability of locally adapted cultivars of rice fter Agriculture intensification program: A Case Study in Solok Regency, West Sumatra. Saintek 3: 129-138 Ceccarelli S (1987). Yield potential and drought tolerance of segregating populations of barley in contrasting environments. Euphytica 36: 265-273. Ceccarelli S, Erskine W, Grando S, Hamblin J (1994). Genotype x Environment Interaction and International Breeding

109


Programmes. Exp. Agr. 30: 177–187. Ceccarelli S, Grando S (1997). Increasing the efficiency of breeding through farmer participation. In: Ethique and Equity in Conservation and Use of Genetic Resources for Sustainable Food Security. Proceeding of Workshop to Develop Guideliness for CGIAR. pp 116-121. Ceccarelli S, Grando S (2005). Decentralized participatory plant breeding: A case from Syria. In: Gonsalves J; Becker T; Braun A; Campilan D; De Chavez H; Fajber E; Kapiriri M; Rivaca-Caminade J; Vernooy R (ed.) Participatory Research and Development for Sustainable Agriculture and Natural Resource Management: A Sourcebook : Understanding Participatory Research and Development. CIP-UPWARD, Laguna, The Philippines and IDRC, Ottawa, Canada. Vol 1: 193199. Eberhart SA, Russell WA (1966). Stability parameter for comparing varieties. Crop Sci. 6:36-40. Harsanti L, Hambali, Mugiono (2003). Stability and adaptability of 10 mutant lines of lowland rice based on yield evaluation in twenty locations for two planting seasons. Zuriat. 14: 1-7. Haryanto TAD, Suwarto, Yoshida T (2008). Yield Stability of aromatic upland rice with high yielding ability in Indonesia. Plant Prod. Sci. 11: 96-103. Kubota T, Nagano T, Ida K (2006). Critical factors for grain

filling in low grain ripening rice cultivars. Agron. J. 98 : 536-544. Oosterom EJ, Kleijn D, Ceccarelly S, Nachit MM (1993). Genotype by environment interaction of barley in the Mediterranean Region. Crop Sci. 33: 669-674. Reichardt W, de Jesus R, Man LH, Kunnot L (2001). Soil biochemical and microbiological clues to the sustainability of intensive rice intercropping systems in Southeast Asia. In S.Peng and B. Hardy (ed.) Rice Research for Food Security and Poverty Alleviation. IRRI. 459 – 467. SAS Institute (2004). SAS/STAT User Guide: Statistics. Version 6.01. SAS Institute, Cary, NC. Saeed M, Francis CA (1984). Association of weather variables with genotype x environment interaction in grain sorghum. Crop Sci. 24: 13 - 16. Saeed M, Francis CA, Rajewski JF (1984). Maturity effect on genotypic environment interaction in grain sorghum. Agron. J. 76: 5558. Seshu, DV, Cady FB (1984). Response of rice to solar radiation and temperature estimated from international yield trials. Crop Sci. 24:649655. Singh S (2005). Effect of low light stress at various growth phases on yield and yield components of two rice cultivars. Central Agric. Res. Inst. Bull. 36 – 39. Suwarto, Nasrullah (2011). Analysis of effect of genotype x

110


environment interaction on rice grain’s iron content in Indonesia using graphical GGE-biplot method. Electron. J. Plant Breed. 2: 288-294. Witt C., Dobermann A, Simbahan GC, Gines HC (2001).

Balanced nutrient management and beyond. In S.Peng and B. Hardy (ed.) Rice Research for Food Security and Poverty Alleviation, IRRI. pp 469-478.

111

RESEARCH ARTICLE

SABRAO Journal of Breeding and Genetics 44 (1) 112 - 128, 2012

GENETIC STUDIES IN UPLAND COTTON. III. GENETIC PARAMETERS, CORRELATION AND PATH ANALYSIS Jesús Rafael MÉNDEZ-NATERA, Abelardo RONDÓN, José HERNÁNDEZ and José Fernando MERAZO-PINTO Departamento de Agronomía, Escuela de Ingeniería Agronómica, Núcleo de Monagas, Universidad de Oriente, Maturín, Estado Monagas, Venezuela. E-mail: [email protected]

SUMMARY The experiment was carried out to determine coefficient of phenotypic, genotypic and environmental correlations and path coefficients (at genotypic level) between seed cotton yield and its components in six upland cotton cultivars and their 15 hybrids, and to determine the coefficient of variation, heritability and genetic advance of these traits at Jusepín, Monagas State, Venezuela. Traits with larger coefficients of genotypic variation were set flowers (24.5%), sympodial branches (20.3%) and at phenotypic level were set flowers (33.3%) and bolls/plant (32.3%). Broad-sense heritability estimates were highest for blooming initiation (96.9%), fiber fineness (80.0%) and stem diameter (64.6%). Expected genetic advances as percentage of the mean were highest for set flowers (37.0%) and sympodial branches (32.1%). Seed cotton yield ha-1 was significantly positively correlated with bolls/plant at both phenotypic and environmental level, while at genotypic level the correlation was significant and positive with fiber length, bolls/plant, flowers/plant, boll weight, sympodial branches/plant, 100-seed weight, and negative with fiber strength. Path analysis indicated that components with maximum direct effects on seed cotton yield were sympodial branches (5.380) and effective boll set (4.993), but these characters being annulated each other. The character that showed a positive correlation (0.532) and a direct positive effect (2.397), being not altered by rest of the components was boll weight, indicating its potential use as a selection criterion to increase the seed cotton yield.

Keywords: cotton, Gossypium hirsutum, genetic advance, correlation, path analysis, heritability Manuscript received: August 26, 2011; Decision on manuscript: December 29, 2011; Manuscript accepted in revised form: February 1, 2012 Communicating Editor: Naqib Ullah Khan

112

INTRODUCTION Knowledge of the relation to yield and its components is invaluable to the plant breeder in selecting desirable strains. Since a change in one character is often accompanied by changes in several others, and practical application cannot be drawn from simple correlation because they do not provide the causal basis of such an association. Hence path coefficient analysis helps to measure the direct and indirect effects of characters influencing yield and also permits a critical appraisal of factors influencing a given correlation. Generally, the magnitude of heritability is influenced by variability between populations, the extent to which a particular character is affected by prevailing environmental conditions of that experiment. Because of this limitation, some workers questioned the wisdom of estimating heritability. However, it can be argued that if a number of estimates for a certain character were made under different environments, a general idea could be formulated about range and magnitude of heritability for that character. Such general knowledge would be useful in indicating the ease or difficulty in attaining effective selection on the basis of phenotypic performance. Heritability in itself provides no indication about the genetic progress that would result from selection. However, at a fixed selection pressure, the amount of advance varies with magnitude of heritability. Genetic advance in a population cannot be predicted

from heritability alone, the genetic gain for specific selection pressure has to be worked out. Several studies had been carried out on correlation and path coefficient analysis. Rauf et al. (2004) computed path coefficients to estimate the contribution of individual characters to yield in cotton, and their findings indicated that bolls plant and sympodial branches had significant positive correlation with seed cotton yield at genotypic level. However, bolls per plant had maximum positive direct effect on seed cotton yield per plant followed by boll weight; whereas, internodal length had maximum negative correlation and direct effect on seed cotton yield. Iqbal et al. (2003) conducted a study on correlation and path coefficient analysis of earliness in upland cotton, they found that node of first fruiting branch, monopodial and sympodial branches per plant, flowers and bolls per plant, boll weight, fiber fineness and fiber strength were positively and significantly correlated with yield. Similarly path coefficient analysis also revealed that sympodial branches, flowers, bolls per plant and boll weight had maximum direct positive effect on seed cotton yield, whereas, monopodial branches per plant, ginning outturn percentage and staple length had the direct negative effects on seed cotton yield. Haidar and Khan (1998) studied path coefficient analysis to understand the contribution directly as well as indirectly of each character to yield of seed cotton, and their findings indicated that

113


bolls per plant and boll weight had the maximum direct effect on seed cotton yield. Bolls per plant had negative indirect effect on boll weight. However, boll weight had positive indirect effect on ginning outturn percentage and negative indirect effect on bolls per plant and seeds per boll. The seeds per boll had negative direct effect on yield while lint % had very small positive effect on seed cotton yield. Also, many investigations had been made on heritability for seed cotton yield and other traits. Basbag and Gencer (2004) indicated that seed cotton weight per boll, 100 seed weight, fiber fineness and fiber strength had high heritability; bolls per plant had low heritability, while other characters had moderate heritability. The characters with high heritability suggested some possibilities in obtaining required genotypes by selection in early segregating generations (F2, F3); while selection for improvement was delayed due to low heritability for some characteristics. Basal and Turgut (2005) mentioned that moderate heritability estimates were observed for earliness ratio (0.53), fiber strength (0.50) seed cotton weight per boll (0.42) and lint % (0.40), however, bolls per plant and seed cotton weight per plant showed low heritability estimates, 0.33 and 0.22, respectively. Khan et al. (2009) stated that identification and use of cotton genotypes with better genetic potential is a continuous prerequisite for synthesis of physiologically efficient and genetically superior genotypes showing promise for increased production per unit area under a

given set of environments. Hence, a comprehensive study of genetic mechanism of the control of plant characters under different environmental conditions is also an obligation. Therefore, the present study was planned to determine the coefficient of genotypic, phenotypic and environmental correlation and path coefficients (at genotypic level) between seed cotton yield and its components of six varieties and their 15 hybrids (no reciprocals) and to quantify the coefficient of variation, heritability and genetic advance of these traits.

MATERIALS AND METHODS The experiment was carried out at Jusepín Town, Monagas state, Venezuela. In this trial six commercial varieties of cotton were used, viz., L1 = Deltapine-16; L2 = Tamcot-SP-21; L3 = Cabuyare; L4 = Stoneville; L5 = Ospino and L6 = Acala-90-1. The experiment consisted of two phases: Phase I: The first phase was denominated phase of crosses, it was carried out from August to November (post rainy season) in the Estación Experimental de Sabana (Experimental Station of Savanna) of the Universidad de Oriente, Venezuela, located in the Table of Piedemonte of the plateau of the Oriental Plains, in a savanna soil previously cultivated. The preparation of the soil consisted of three passes of harrow, lime application of 1000 kg/ha 30 days before sowing. The experiment was

114

Méndez-Natera et al. (2012)

fertilized with 15-15-15 in a dose of 600 kg/ha, applied in deep bands, urea was applied 30 days after sowing, in a dose of 200 kg/ha, applied in superficial bands. Two hills of plants were sown for each one of the used varieties. The distance between plants was 0.20 m and between hills was 1.0 m. The hills, with a length of 35.0 m were divided in 6 segments of 5.0 m each one, separated to each other by 1.0 m. One of the segments was dedicated for the self fertilization and intra-varietal crosses and the five remaining hills were used to carry out the crosses with the other varieties, this with the purpose of obtaining the hybrid seeds, by means of diallel crosses excluding reciprocals to be used in the second phase. The crosses began 36 days after sowing, following the methodology described by Poehlman (1981). The crosses were made one day before it was expected, the flowers opened up. For that, in first place, the corolla of the flower of the plant that would serve as female parent was cut, with a small scissors, then the anthers were emasculated, the stigma protected to avoid the possibility of cross pollination. The following day, in the morning, the pollination was carried out, collecting the male parent's pollen in a small piece of closed straw in one of its ends. This straw piece partially full with pollen was placed on the exposed stigma after removing the protector of the emasculated flower. To assure the pollination, the flower bracts were rose and they were placed surrounding the straw piece, getting together with a fine wire so

that they stayed in their place. Finally, the cross was marked with a color tape in order to facilitate its identification. Approximately 40 crosses were made for segment with an approximate percentage of success (75%) as reported by Poehlman (1981). For these crosses, there was a bigger emphasis to select flowers located in middle third of the plant, which assure a bigger germination percentage and vigor (Arturi, 1984). The mature bolls from the crosses and from the self fertilization were harvested in two harvests, 119 and 135 days after sowing. Phase II: The second phase or evaluation phase was carried out from May to August (rainy season) in the Estación Hortícola de Producción Vegetal (Horticultural Station of Crop Production of the Universidad de Oriente. Among the cultural practices carried out in this phase were: the preparation of the soil consisted of two passes of plow, three passes of harrow and lime application in dose of 1000 kg/ha, 30 days before sowing. The last pass of harrow was made one day before sowing, jointly with a pass to make the furrows. Three hills of the six upland cotton cultivars and their 15 hybrids were sown at a distance of 0.20 m between plants and 0.80 m between hills for an equivalent final population to 62500 plant/ha. Supplementary irrigation was made to meet the water requirements of crop due to absence of rains during development of the crop. The cotton plants were fertilized with 15-15-15 in a dose of 600 kg/ha,

115


placed in deep bands, urea was applied in superficial bands 30 days after sowing in a dose of 200 kg/ha. Weeds were controlled chemically using Round-up in dose of 4 l/ha in pre-sowing. H1-2000 was also applied in post-emergence in dose of 2 l/ha, 20 days after sowing. Two manual cleans were carried out, 35 and 82 days after sowing. The randomized complete blocks design was used with three repetitions and 21 treatments (six upland cotton cultivars and their 15 hybrids). Each treatment was constituted by three hills of plants, the two outer two hills were considered as borders and the central hill was used for evaluation of the different parameters. Coefficients of phenotypic, genotypic and environmental correlations and partitioning of genotypic correlations of seed cotton yield with their component characters were made by following Singh and Chaudhary (1977). On the basis of Table 1, phenotypic and genotypic variances, their coefficients, heritability and genetic advance were calculated as follows: Genotypic variance (σ2 ) = g

M2 - M3 r

=

Heritability (H %) =

σ2

p

Phenotypic coefficient of variation (P.C.V.) =

___

x 100 or =

x 100

X

σ2g

Genotypic coefficient of variation (G.C.V.) =

___

x 100

X

Genetic advance (G.A.) =

σ2g σ 2p

σ2g xK σp

xKσ = p

K σ2 Genetic advance as percentage of the mean =

σp

g

x

100 ___

X

___

Where X is the general mean of a character, K σp is the selection differential expressed in phenotypic standard deviations and K value was 2.06 at 5% selection intensity.

(σ2 + r σ2 ) - σ2 g g g r

M2 Phenotypic variance (σ2 ) = = p r

σ2g

σ 2p

(σ2 + r σ2 ) e g r

M2 - M3 x 100 M2

116


RESULTS For all the characters, the genotypic coefficients of variation were higher in magnitude as compared to phenotypic coefficients of variation (Table 2). High genotypic coefficients of variability were obtained for set flowers, sympodial branches, plant height, seed yield and stem diameter, while low for fiber properties viz., index of fiber uniformity, fiber length and strength and height of first sympodial branch. However, high phenotypic coefficients were obtained for set flowers, bolls/plant, seed cotton yield, seed yield and sympodial branches, and low for fiber properties, viz., fiber length, strength and fineness and index of fiber uniformity. Heritability (broad sense) varied from 0% (index of fiber uniformity) to 96.9% (blooming initiation). Except for blooming initiation, fiber fineness, stem diameter, sympodial branches, plant height and set flowers, all other characters gave heritability estimates with less than 50%, with low to moderate heritability. Four characters, viz., set flowers, sympodial branches, plant height and stem diameter exhibited high magnitude of genetic advance, expressed as a percentage of the mean, whereas, index of fiber uniformity, height of first sympodial branch and fiber length gave low values. These two groups of traits were also characterized by high and low genotypic coefficients of variation, respectively. In other words, the traits which had the least genotypic coefficients of variation also had

lowest expected genetic improvement. Seed cotton yield had a low heritability estimate and genetic advance of 18.8 and 12.3%, respectively. Coefficients of genotypic, phenotypic and environmental correlations have been given in Tables 3, 4 and 5. Significant genotypic, phenotypic and environmental correlations were 38, 14 and 7, respectively, and overall genotypic correlation were higher than phenotypic and environmental correlations, thus phenotypic expression seems to have been lessened by the influence of environment. At genotypic level, seed cotton yield was found to be positively and significantly correlated with fiber length, bolls/plant, set flowers, boll weight, sympodial branches and 100-seed weight and negatively correlated with fiber strength. At phenotypic and environmental levels, seed cotton yield was only found to be positively and significantly correlated with bolls/plant. Path coefficient analysis at genotypic level (Table 6) indicated that sympodial branches had the maximum direct effect (5.380) on seed cotton yield followed by boll set (4.993), boll weight (2.397), 100-seed weight (-2.309) and fiber length (1.610), and the remaining characters showed negligible direct effects and less than residual value (1.424).

DISCUSSION Apart from fiber properties, height of first sympodial branch, 100-seed weight and seeds/boll, all other

117


characters were amenable to improvement through selection due to their moderate to high values of genotypic and phenotypic coefficients of variation. Similarly, some of the previous workers reported a wide range of genotypic and phenotypic variability for most of the characters. Seth and Singh (1984) reported a wide variation for boll number, yield and number of primary and secondary monopodia. Zhou (1986) reported that genetic coefficient of variation for bolls/plant, boll weight and lint % were 17.2, 10.6 and 4.3%, respectively. Ali et al. (2009) found that genotypic, phenotypic and environmental coefficients of variability were highest for fiber fineness (8.0, 13.8 and 11.2%, respectively) follow by cotton seed yield, staple length and fiber strength. Broad sense heritability was calculated in order to estimate the proportion of total variance attributed to genotypic differences. The values obtained were generally moderate to high, with heritability estimates exceeding 20% in 12 out 18 characters studied and exceeding 50% in six of the 18 characters. Results indicated that six characters viz., blooming initiation, fiber fineness, stem diameter, sympodial branches, plant height and set flowers, possessed a wide range of genetic variability and improvement could be achieved with mass selection alone. Heritability estimates encountered in present study were also common in previous literature. Sinde and Deshmukh (1985) observed that heritability was high for boll number, ginning

percentage, days to flowering and yield in G. arboreum. Bhatade and Bhale (1984) grown 7 x 7 diallel crosses of G. arboreum at 4 sites and found moderate to high heritability for ginning outturn and halo length, but low for seed and lint indices. Seth and Singh (1984) in a study with parents, F1, F2 and F3 from each of five crosses involving six cotton varieties of G. hirsutum reported that boll number, yield and number of primary and secondary monopodia showed high broad sense heritability and genetic advance. Ali et al. (1998) indicated that broad sense heritability estimates were prominent for bolls per plant, boll weight and yield of seed cotton and suggested that improvement in these traits can be made through selection in early segregating generations. Naveed et al. (2004) indicated that broad sense heritabilities were low to moderate; and the values were 22% and 23% for boll weight and lint %, respectively. For seed cotton yield, plant height and bolls, the estimates were being 33%, 35% and 38%, respectively and suggested that rigorous plant selection is required to identify desirable plants in F2 generation. Basbag and Gencer (2004) findings revealed that seed cotton weight per boll, 100 seed weight, fiber fineness and fiber strength had high heritability, bolls per plant had low, while other traits had moderate heritability, and characters with high heritability would speed up the obtaining desirable genotypes by selection in early segregating generations; while selection was delayed where the heritability was low. Basal and Turgut (2005)

118


found highest heritability for earliness ratio, fiber strength, seed cotton weight per boll and lint %, and moderate for bolls and seed cotton weight per plant. Also they reported that genetic advance as mean percentage was higher for fiber fineness followed by staple length. In the present study, heritability (%) and genetic advance were 16.7 and 3.7% for seeds per boll and 31.4 and 5.2% for seed index. These results were not in agreement with those reported by Khan et al. (2010) who stated that heritability estimates with expected selection response for seeds per boll were moderate and desirable with presence of significant positive correlation, revealed that cultivars have the genetic potential to boost up the seeds per boll, and high heritability for seed index with valuable expected selection response and positive correlation with yield, exhibited that seed size was administered through genetic variance and there is space for improvement in seed size. Batool et al. (2010) found that heritability for boll weight (0.97) was high and genetic variances were found greater than environmental variances and along with high heritability it authenticated that cultivars have the potential to enhance the boll weight which is the second main contributor (after boll number) to seed cotton yield. Ali et al. (2009) found that among fiber quality parameters, fiber elongation possessed highest broad sense heritability (81.3%) followed by fiber uniformity ratio (77.4%), staple length (73.7%) and fiber strength (62.9%). High heritability

advocated great amount of fixable and additive gene action in phenotypic expression of these characters. However, it was low for fiber fineness probably due to involvement of non-genetic effects. According to Johnson et al. (1955) high heritability and genetic advance are usually more helpful in predicting gain under selection than heritability estimates alone. According to Osman and Khirdir (1974) usually low heritability coupled with low genetic advance is an indication of non-additive gene effects and consequently a low genetic gain is expected from selection. On the other hand, an association of high heritability with high genetic advance is indicative of additive gene effects and consequently a high genetic gain from selection would be anticipated. In this study, high heritability and genetic advance were found for blooming initiation, plant height, stem diameter, sympodial branches and set flowers. It is therefore suggested that high magnitude of genetic gain would result for these characters. Seed cotton yield/ha had a low heritability estimate and moderate advance genetic, so the selection for high yields could be ineffective. There were more significant correlations at genotypic level, followed for phenotypic ones and low environmental correlations. Results were in agreement with Desalegn et al. (2009) who stated that in most cases, the phenotypic correlation was lower than genotypic correlation, showing that traits were mainly governed by genetic effects. Similarity, Qayyum et al. (2010) conducted a diallel

119


experiment with eight cotton varieties and found that genotypic correlation coefficients were higher than phenotypic which indicated less involvement of environmental effects and genetic causes were more pronounced in expression of associations among traits. Yield was positively and significantly correlated with fiber length, bolls/plant, set flowers, boll weight, sympodial branches and 100-seed weight and negatively correlated with fiber strength. It revealed that yield potential was increased with more prolific plants (more bolls and flowers per plant) with more sympodial branches, heavier seeds and bolls with larger fibers. With few exceptions, the total correlation coefficients between yield and yield components were comparable with those reported by various workers. Méndez-Natera (1996) found that seed cotton yield was positively and significantly correlated with bolls/plant, boll weight and fiber index. Méndez-Natera et al. (1992) in a step wise regression study found that sympodial branches/plant contributed with 63% of seed cotton yield variation, adding 100-seed weight, an increment of 15% was obtained in order to predict yield. El-Helw et al. (1988) and Sandhu et al. (1986) found a positive and significant correlation for seed cotton yield with bolls/plant and boll weight. Iqbal et al. (2003) indicated that node of first sympodial branch, monopodial and sympodial branches plant, flowers and bolls per plant, boll weight, fiber fineness and fiber strength were positively and significantly correlated with yield. The positive

correlation of seed cotton yield with fiber length have been reported by Haidar and Khan (1998), however, indicated a negative correlation for seed cotton yield and seed index, being contrary to our results. Positive correlation of seed cotton yield with bolls/plant had also been found by Hussain et al. (2000) and Naveed et al. (2004). More recently, Desalegn et al. (2009) found that seed cotton yield was highly and significantly correlated with lint yield, seeds per boll, boll weight, lint index, bolls per plant; and moderately and significantly correlated with lint %, while lint yield was highly and positively correlated with lint % and seeds per boll, and moderately correlated with lint index and negatively correlated with seed index. Thiyagu et al. (2010) found that seed cotton yield was significantly positively correlated with seven traits i.e. bolls per plant, sympodial branches per plant, plant height, 2.5 per cent span length, bundle strength, seed index and elongation percentage. Following these characters, lint index and boll weight recorded positive correlation with yield. Hence, selection for these characters will help in selecting genotypes with high seed cotton yield per plant. Path coefficients are an excellent means to study the direct and indirect effects of interrelated components of a complex character as seed cotton yield/ha. Path analysis indicated that the components with the biggest direct effects on seed cotton yield were sympodial branches (5.380) and effective boll set (4.993), but these characters being annulated each

120


other. The character that showed a positive correlation (0.532) and a direct positive effect (2.397), being not altered by rest of components was boll weight, indicating its potential use as a selection criterion to increase the seed cotton yield/ha. The moderate residual value (1.424) obtained from path analysis indicated that influence of other characters may also be evaluated. Results agreed with those reported by various workers. Schwendiman (1977) indicated that bolls/plant and boll weight were the more important components of seed cotton yield. Waldia et al. (1979) in a study with 19 cotton varieties found that seeds/locule, bolls/plant and boll weight exhibited a positive direct effect on seed cotton yield. According to Thiyagu et al. (2010) path coefficients are the subdivision of genotypic correlation coefficients of individual characters with seed cotton yield. Path coefficient analysis was done in order to study the direct and indirect effects of individual component characters on dependent variable, seed cotton yield per plant. Path coefficients study enables breeders to concentrate on the variable which shows high direct effect on seed cotton yield and ultimately reduce the time wastage in looking for more component traits by restricting selection to one or few important traits. Rauf et al. (2004) found that bolls per plant had maximum positive direct effect on seed cotton yield per plant followed by boll weight; whereas, internodal length had maximum negative direct effect on seed cotton yield. Haidar and Khan (1998) revealed that

bolls per plant and boll weight had the maximum direct effect on yield of seed cotton per plant and bolls per plant had negative indirect effect on boll weight, while boll weight had positive indirect effect on lint % and negative indirect effect on bolls per plant and seeds per boll, they suggested that selection for higher productivity should be based on higher bolls per plant and boll weight. Iqbal et al. (2003) indicated that path coefficient analysis revealed that sympodial branches, flowers and bolls plant and boll weight had maximum direct positive effect on yield of seed cotton, whereas, the traits monopodial branches per plant, lint % and staple length had the direct negative effects on seed cotton yield, they indicated that for evolving a superior genotype possessing all the three basic characteristics (earliness, high yield and improved fiber quality of international standard) the breeder had to use the reciprocal recurrent selection method or modified back cross or three way cross within genetic material under study. The contradictory findings were reported by Méndez-Natera (1996) who found a negative direct effect (-1.63) of boll weight on seed cotton yield. However, Thiyagu et al. (2010) observed a very high positive direct effect for bolls per plant (1.030) and boll weight (0.411). The remaining characters namely 2.5 per cent span length (0.065), sympodial branches per plant (0.055), ginning percentage (0.040), seed index (0.029), fiber fineness (0.025), uniformity ratio (0.024), days to 50% flowering (0.005) and fiber elongation percentage (0.002) recorded

121


positive effect on seed cotton yield, and most of these results were in agreement with our findings.

CONCLUSIONS Traits with larger coefficients of genotypic variation were set flowers and sympodial branches and at phenotypic level were set flowers and bolls/plant. Broad sense heritability estimates were highest for blooming initiation, fiber fineness and stem diameter. Expected genetic advances were highest for set flowers and sympodial branches. Seed cotton yield/ha was significantly positively correlated with bolls/plant at phenotypic and environmental level, while at genotypic level the correlation was significant and positive with fiber length, bolls/plant, flowers/plant, boll weight, sympodial branches/plant and seed index, and negative with fiber strength. Path analysis indicated that yield related traits with maximum direct effects on seed cotton yield were sympodial branches and effective boll set, but these characters being annulated each other. However, boll weight showed positive correlation and direct positive effect on yield and not altered by rest of components and can be used as a selection criterion to increase the seed cotton yield.

122


Table 1. Analysis of variance and the expectations of the components of variance. S.O.V.

d.f.

M.S.

Replications

(r-1)

M1

Expected value of M.S.* --

Genotypes

(g-1)

M2

rσ2g + σ2e

Error

(r-1)(g-1)

M3

σ2e

Total

(rg-1)

M1 + M2 + M3

*σ2g = Genotypic variance, σ2e = Error variance, r = Number of replications

Table 2. Phenotypic and genotypic coefficients of variation, heritability and genetic advance for various traits of 6 commercial varieties and 15 hybrids of upland cotton. Heritability Characters GCV PCV G.A. G.A. (%) (%) Blooming initiation 7.9 8.0 96.9 9.6 days 15.9 Plant height 16.9 22.4 57.0 25.3 cm 26.3 Stem diameter 14.3 17.8 64.6 0.3 cm 23.6 First fruit branch height 3.5 9.0 14.7 0.7 cm 2.7 Fruit branches 20.3 26.5 58.8 4.9 branches 32.1 100-seed weight 4.5 8.0 31.4 0.6 g 5.2 Seeds per boll 4.4 10.6 16.7 1.1 seeds 3.7 Bolls per plant 12.8 32.3 15.7 1.3 bolls 10.4 Boll weight 7.3 15.0 22.0 0.6 g 6.8 Fruit set 11.2 19.6 32.7 4.2% 13.2 Set flowers 24.5 33.3 53.9 11.4 flowers 37.0 Seed cotton yield 13.8 31.8 18.8 186.3 kg/ha 12.3 Seed yield 14.5 31.0 21.9 135.9 kg/ha 14.0 Fiber content 4.4 9.9 19.4 1.4% 4.0 Fiber length 2.6 4.0 41.2 0.03 inches 2.9 Index of fiber uniformity 0.3 5.4 0.00 0.02% 0.1 Fiber strength 3.1 5.0 37.3 2890 lb/inch2 3.9 Fiber fineness 4.8 5.3 80.0 0.37 mic 8.8 GCV, PCV: Genotypic and Phenotypic coefficient of variation, G.A.: Genetic advance

123


Table 3. Coefficients of genotypic correlation of seed cotton yield with others traits of six commercial varieties and 15 hybrids of upland cotton. Blooming 100-seed Fruit Seeds per Boll Fiber Set Bolls per Boll set Fiber Characters Initiation weight branches boll Weight content (%) flowers plant (%) length Seed cotton yield -0.203 0.794 * 0.572 * 0.126 0.532 * -0.347 0.645 * 0.755 * 0.112 0.484 * Fiber strength -0.263 -0.400 -0.780 * -0.618 * -0.679 * -0.843 * -0.837 * -0.764 * 0.591 * 0.297 Fiber fineness -0.182 0.413 0.218 -0.419 0.357 0.297 0.123 0.190 -0.472 * -0.040 Fiber length -0.300 0.352 -0.298 0.099 -0.760 * -0.800 * -0.298 0.165 0.566 * Boll set -0.413 -0.437 * -0.877 * -0.014 -0.952 * 0.776 * -0.808 * -1.000 * Bolls per plant 0.077 0.573 * 1.000 * -0.143 1.000 * -1.000 * 1.000 * Set flowers 0.329 0.728 * 0.998 * 0.148 1.000 * -0.923 * Fiber content -0.249 -0.312 -0.972 * 0.072 -0.757 * Boll weight 0.400 0.359 1.000 * 0.231 Seeds per boll 0.497 * -0.180 -0.075 Fruit branches 0.289 0.706 * 100-seed weight 0.155

Fiber fineness -0.034 0.093

Table 4. Coefficients of phenotypic correlation of seed cotton yield with others traits of six commercial varieties and 15 hybrids of upland cotton. Blooming 100-seed Fruit Seeds Boll Fiber Set Bolls per Boll Fiber Fiber Characters Initiation weight branches per boll weight content (%) flowers plant set (%) length fineness Seed cotton yield -0.057 0.047 0.345 0.240 -0.158 0.253 0.364 0.653 * 0.298 0.130 0.003 Fiber strength -0.154 -0.389 -0.378 -0.219 -0.388 0.330 -0.375 -0.141 0.249 0.241 0.052 Fiber fineness -0.162 0.125 0.173 -0.067 0.206 0.044 0.132 0.066 -0.346 0.035 Fiber length -0.179 0.156 -0.159 0.577 * -0.385 0.357 -0.119 0.133 0.172 Boll set -0.240 -0.301 -0.547 * 0.182 -0.679 * 0.560 * -0.520 * 0.068 Bolls per plant 0.032 0.059 0.595 * 0.171 0.056 -0.037 0.629 * Set flowers 0.264 0.117 0.916 * 0.030 -0.548 * -0.495 * Fiber content -0.092 -0.067 -0.458 * 0.229 -0.795 * Boll weight 0.185 0.272 0.532 * -0.190 Seeds per boll 0.134 -0.043 -0.033 Fruit branches 0.236 0.211 100-seed weight 0.063 * Significant at P ≤ 0.05.

Fiber strength -0.805 *

Fiber strength -0.173

124


Table 5. Coefficients of environmental correlation of seed cotton yield with others traits of six cultivars and 15 hybrids of upland cotton. Blooming 100-seed Fruit Seeds per Boll Fiber Set Bolls per Boll set Fiber Fiber Characters initiation weight branches boll Weight content (%) flowers plant (%) length fineness Seed cotton yield 0.186 -0.194 0.268 0.264 -0.333 0.394 0.260 0.634 * 0.445 0.002 0.038 Fiber strength 0.027 -0.384 -0.025 -0.089 -0.277 0.146 0.000 0.055 0.062 0.206 0.007 Fiber fineness -0.069 -0.188 0.087 0.032 0.141 -0.161 0.161 0.003 -0.270 0.153 Fiber length 0.056 0.050 0.009 0.045 -0.234 0.192 0.036 0.180 -0.054 Boll set -0.031 -0.235 -0.301 0.250 -0.585 * 0.494 * 0.319 0.392 Bolls per plant 0.018 -0.240 0.467 * 0.230 -0.272 0.227 0.509 * Set flowers 0.226 -0.322 0.812 * -0.233 0.238 -0.322 Fiber content 0.099 0.013 -0.225 0.263 -0.717 * Boll weight 0.007 0.243 0.252 -0.291 Seeds per boll -0.411 -0.003 0.017 Fruit branches 0.160 0.172 100-seed weight -0.155 * Significant at P ≤ 0.05.

Fiber strength 0.055

125


Table 6. Direct (diagonal) and indirect effects of yield components on seed cotton yield through path coefficients of six cultivars and 15 hybrids of upland cotton. Characters

Fiber strength

Fiber fineness

Fiber strength (0.364) 0.123 Fiber fineness 0.034 (1.327) Fiber length 0.108 -0.053 Boll set 0.215 -0.626 Bolls per plant -0.278 0.252 Set flowers -0.305 0.163 Fiber content -0.307 0.354 Boll weight -0.247 0.474 Seeds per boll -0.225 -0.198 Fruit branches -0.284 0.289 100-seed weight -0.146 0.548 Blooming initiation -0.096 -0.242 * Significant at P ≤ 0.05. Residual Value : 1.424,

Fiber length

Boll set Bolls per Set Fiber (%) plant flowers content (%)

0.478 2.951 -0.102 0.639 -0.064 -2.357 0.025 -0.094 (1.610) 2.826 0.022 0.228 0.911 (4.993) -0.133 0.617 0.266 -4.993 (0.133) -0.763 -0.480 -4.034 0.133 (-0.763) 1.288 3.874 -0.133 0.703 -1.223 -4.753 0.133 -0.763 0.159 -0.070 -0.021 -0.113 -0.480 -4.379 0.133 -0.762 0.567 -2.182 0.076 -0.556 -0.483 -2.062 0.010 -0.251 Gen. corr : Genotypic correlation

-0.416 0.147 0.395 0.383 -0.493 -0.455 (0.493) -0.373 0.036 -0.479 -0.154 -0.123

Boll Seeds per weight boll 1.628 0.856 -1.822 -2.282 2.397 2.397 -1.814 (2.397) 0.554 2.397 0.860 0.959

0.370 0.089 -0.059 0.008 0.086 -0.089 -0.043 -0.138 (-0.598) 0.045 0.108 -0.297

Fruit branches

100-seeds weight

Blooming initiation

Gen. corr. with yield

-4.196 1.163 -1.603 -4.718 5.380 5.365 -5.229 5.380 -0.403 (5.380) 3.798 1.555

0.924 -0.954 -0.813 1.009 -1.323 -1.681 0.720 -0.829 0.416 -1.630 (-2.309) -0.358

-0.312 -0.216 -0.355 -0.489 0.091 0.390 -0.295 0.474 0.589 0.342 0.184 (1.185)

-0.805 * -0.034 ns 0.484 * -0.112 ns 0.755 * 0.645 * -0.347 ns 0.532 * 0.126 ns 0.572 * 0.794 * -0.203 ns

126


REFERENCES Ali B, Khan IA, Aziz K (1998). Study pertaining to the estimation of variability, heritability and genetic advance in Upland cotton. Pak. J. Biol. Sci. 1(4): 307-308. Ali MA, Khan IA, Nawab NN (2009). Estimation of genetic divergence and linkage for fibre quality traits in upland cotton. J. Agric. Res. 47(3): 229-236. Arturi, M (1984). El algodón. Mejoramiento genético y técnica de su cultivo. Editorial Hemisferio Sur, Buenos Aires, Argentina. 179 pp. Basal, H, Turgut I (2005). Genetic analysis of yield components and fiber strength in Upland cotton (G. hirsutum L.). Asian J. Plant Sci. 4(3): 293-298. Basbag, S, Gencer O (2004). Investigations on the heritability of seed cotton yield, yield components and technological characters in cotton (G. hirsutum L.). Pak. J. Biol. Sci. 7(8): 1390-1393. Batool S, Khan NU, Makhdoom K, Bibi Z, Hassan G, Marwat KB, Farhatullah, Mohammad F, Raziuddin, Khan IA (2010). Heritability and genetic potential of upland cotton genotypes for morphoyield traits. Pak. J. Bot. 42(2): 1057-1064. Bhatade, SS, Bhale NL (1984). Estimate of gene effects for seed and fiber characters in desi cotton (G. arboreum L.). Madras Agric. J. 71(2): 7177. Desalegn Z, Ratanadilok N, Kaveeta R (2009). Correlation and heritability for yield and fiber quality parameters of Ethiopian cotton (G. hirsutum L.) estimated from 15 (diallel)

crosses. Kasetsart J. (Nat. Sci.) 43: 1-11. El-Helw MR, Younis SEA, Sherif THL, Omara MK, Taghian AS (1988). Heterosis for yield and its components in crosses among Egyptian and Russian cotton varieties. Assiut J. Agric. Sci. 19(2): 27-39. Haidar S, Khan MA (1998). Path coefficient analysis of some yield traits in cotton (G. hirsutum L.). Pak. J. Biol. Sci. 1(2): 115-116. Hussain SS, Azhar FF, Mahmood I (2000). Path coefficient and correlation analysis of some important plant traits of G. hirsutum L. Pak. J. Biol. Sci. 3(9): 1399-1400. Iqbal, M, Chang MM, Iqbal MM, Hassan M, Nasir A, Islam N (2003). Correlation and path coefficient analysis of earliness and agronomic characters of upland cotton in Multan. J. Agron. 2(3): 160168. Johnson HW, Robinson HF, Comstock RE (1955). Estimates of genetic and environmental variability in soybean. Agron. J. 47: 34-38. Khan NU, Hassan G, Marwat KB, Farhatullah, Batool S, Makhdoom K, Khan I, Khan IA, Ahmad W (2009). Genetic variability and heritability in upland cotton. Pak. J. Bot. 41(4): 16951705. Khan NU, Marwat KB, Hassan G, Farhatullah, Batool S, Makhdoom K, Ahmad W, Khan HU (2010). Genetic variation and heritability for cotton seed, fiber and oil traits in G. hirsutum L. Pak. J. Bot. 42(1): 615-625. Méndez-Natera JR, Merazo JF, Jiménez E (1992). Correlación y regresión múltiple entre el rendimiento

127


de algodón en rama/ha y algunos caracteres de la planta y de la bellota en 10 cultivares de algodón (G. hirsutum L.) en Jusepín, Edo. Monagas. In II Congreso Científico de la Universidad de Oriente. Guatamare, Estado Nueva Esparta, Venezuela. p. 249-250. Méndez-Natera JR 1996. Análisis de los coeficientes de correlación lineal y de los coeficientes de trayectoria en algodón (G. hirsutum L.). SABER 8(1): 56-62. Naveed MF, Azhar M, Ali A (2004). Estimates of heritabilities and correlations among seed cotton yield and its components in G. hirsutum L. Int. J. Agric. Biol. 6(4): 712714. Osman HE, Khidir MO (1974). Estimates of genetic and environmental variability in sesame. Exp. Agric. 10: 105112. Poehlman J. 1981. Mejoramiento genético de las cosechas. Editorial LIMUSA, México, 453 pp. Qayyum A, Murtaza N, Azhar FM, Iqbal MZ, Malik W (2010). Genetic variability and association among oil, protein and other economic traits of G. hirsutum L. in F2 generation. J. Agric. Res. 48(2): 137-142. Rauf S, Khan TM, Sadaqat HA, Khan AI (2004). Correlation and path coefficient analysis of yield components in cotton (G. hirsutum L.). Int. J. Agric. Biol. 6(4): 686-688. Sandhu BS, Arora RL, Mangat NN, Singh G (1986). Association of yield components in arboreum cotton. Crop Improv. 13(2): 189-192. Schwendiman J (1977). Modifications induced by completely

replacing the A6 pair of chromosome of G. hirsutum by their homologue from G. barbadense. Cotton Fibres Trop. 30(3): 283-291. Seth S, Singh DP (1984). Studies on heritability and variability for yield components in Upland cotton (G. hirsutum L.). Haryana Agric. Univ. J. Res. 14(3): 313-317. Sinde VK, Deshmukh MD (1985). Genetic variability for yield and character association in desi cotton. J. Maharashtra Agric. Univ. 10(1): 21-22. Singh RK, Chaudhary BD (1977). Biometrical methods in quantitative genetic analysis. Hissar, India. 318 p. Thiyagu K, Nadarajan N, Rajarathinam S, Sudhakar D, Rajendran K (2010). Association and path analysis for seed cotton yield improvement in inter-specific crosses of cotton (Gossypium spp). Electron. J. Plant Breed. 1(4): 1001-1005. Waldia RS, Jatasra DS, Dahiya BN (1979). Correlations and path analysis of yield components in G. arboreum L. Indian J. Agric. Sci. 49(1): 32-34. Zhou YY (1986). Yield components in upland cotton. Acta Agric. Univ. Pekin. 12(3): 269-274.

128


GENETIC VARIABILITY AND DIVERSITY STUDIES IN YIELD AND ITS COMPONENT TRAITS IN RICE (Oryza sativa L.) D. BHADRU1, 2*, V. TIRUMALA RAO, Y. CHANDRA MOHAN and D. BHARATHI 1

Rice Research Scheme, Regional Agricultural Research Station, Poalasa, Jagtial, Karimnagar Dist. Andhra Pradesh 2 Present Address: Agricultural Research Station, Kampasagar, Babau sai pet (Post), Thripuram mandal, Nalgonda District. Andhra Pradesh, 508 207 *Corresponding author email: [email protected]

SUMMARY A total of 21 rice genotypes (resistant to gall midge biotype 3 and BPH were evaluated for their variability and genetic divergence. The highest genotypic and phenotypic coefficient of variation, heritability and genetic advance % of mean corresponded to grains per panicle, seed yield, 1000 grain weight and plant height and direct selection for these traits would be useful for yield improvement in rice. The D2 values were significant among the 21 genotypes, which were grouped into 6 clusters. Most of the genotypes with same pedigree either male or female parent involved cross combination came under the same cluster and few genotypes in different cluster and genotypes of quite different pedigree may all into the same cluster. For getting desirable transgressive segregants for the development of early duration (Genotypes from the clusters II and Cluster VI) and medium duration (from the clusters I and IV), coarse and medium slender high yielding gall midge and BPH resistant varieties genotypes could be utilized in the hybridization programme. Key words: Rice, genetic divergence, cluster analysis and variability Manuscript received: September 13, 2011; Decision on manuscript: March 29, 2012; Manuscript accepted in revised form: April 28, 2012. Communicating Editor: Bertrand Collard

INTRODUCTION Rice occupies a pivotal place in Indian agriculture, as it is the staple food for two thirds of the population, providing 43 % of calorific requirement and 20-25% of agricultural income. It is grown in more than 23% of gross cropped

area in India (about 43.5 M ha) which is the largest in the world among all the rice growing countries. Annual production of rice in the country is around 90 million tones, which is the second largest in the world after China. Among the several constraints in rice production, biotic stress as

129

Bhadru et al. (2012)

caused due gall midge (Orseolia oryzae) and brown plant hoppers (BPH) are very important. It is estimated that these pests can cause an economic loss varying from 20100% depending the level of incidence. The cheapest, easy and most successful plant protection measure against this pest is the host plant resistance. Many varieties are available for cultivation either with resistance to gall midge or BPH. However, varieties are not available with resistance to both pests with fine and medium slender grain type. Genetic variability and divergence are of great interest to plant breeders as they play a vital role in forming a successful breeding programme. In general, the genetically diverse parents are utilized to obtain the desirable recombinants in segregating generations. Multivariate analysis is an important tool for the assessment of genetic divergence. Thus, it is utilized to assess genetic divergence along with the relative importance of different traits in the total divergence. Earlier studies reported that the clustering pattern of the genotypes from different sources/origin clustered together Chaturvedi and Maurya (2005) and Sabesan (2008) indicated that there was no association between geographical distribution of genotypes and genetic divergence. The possible reason for grouping of genotypes of different region in one cluster could be the free exchange of germplasm among the breeders of different regions, or unidirectional selection practiced by breeder in tailoring the promising cultivars of different regions. Genotypes from the same

center of origin were distributed in different clusters (Kandamoorthy and Govindarasu (2005), Senapathi and Sarkar (2005), Sabesan et al., (2009)) and Banumathy et al. (2010) which may be due to differential adaptation to varied agro-ecosystems. The present study was carried out to: (1) assess the extent of variability, heritability, genetic advance; (2) genetic diversity and clustering pattern among the genotypes derived from versatile crosses with resistance to gall midge biotype 3 and BPH, so that efforts can be utilized to develop high yielding rice varieties with resistance to gall midge and BPH with fine grain type.

MATERIALS AND METHODS Nineteen Jagtial cultures resistant to gall midge biotype 3 and two well adopted brown plant hopper (BPH) resistant varieties viz., MTU1010 and MTU1001 comprise the basic material for study (Table 1). These twenty one entries were evaluated in randomized complete block design with 4 m length of 20 rows each with 20 x 15 cm spacing during kharif 2007, 2008 and 2009 with two replications at Regional Agricultural Research Station, Jagtial (Latitude 18° 48¹ N and Longitude 78° 24¹ E) for Northern Telangana agro- climatic zone of Andhra Pradesh, South India. Observations were recorded on five plants at random in each replication for the characters plant height (cm), panicle length (cm), filled grains per panicle and panicle bearing tillers per m2 and seed

130


yield per m2 and 1000 grain weight. However for days to 50% flowering data was recorded on whole plot basis. Statistical analysis: Data obtained from the three years were subjected to pooled analysis of variance as per Panse and Sukhatme (1985) using WINDOSTAT statistical software. Genetic parameters were calculated by pooling data three years’ data. Coefficients of variation were estimated as per Burton and Dewane (1952). Heritability (broad sense) as per Hanson et al. (1956) and genetic variance as per Johnson et al. (1955). D2 analysis (Mahalanobis, 1928) as elaborated by Murty and Arunachalam (1966). The genotypes were grouped into different clusters by Tochers method as described by Rao (1952).

RESULTS AND DISCUSSION In the present investigation, 21 genotypes were subjected to pooled analysis of variance for seven characters (Table 2). The analysis of variance for stability revealed that the genotypes were significant for all the traits studied except for plant height and panicle bearing tillers whereas, days to 50% flowering, panicle length panicle bearing tillers and number of grains per panicle for environments, indicating the diversity among the genotypes and environments studied. The genotypes and environment interactions were significant for grain yield/ m2. Significant genotype x environment

interactions implies differential behavior of genotypes under different environments as also revealed by Deshpande and Dalvi (2006), Saidaiah et al. (2011) and Sridhar et al. (2011). The analysis of variance showed highly significant differences between genotypes for the characters studied indicated the presence of considerable genetic variation in the experimental material. The means, phenotypic and genotypic coefficients of variation and heritability estimates are presented in Table 3. A wide range of variability was observed for all the traits. Phenotypic coefficient of variation (PCV) were higher than those of genotypic coefficient of variation (GCV) for all the traits studied, indicating that they all interacted with the environment with the same degree. Similar findings were reported by Zahid et al. (2006) and Abdus Salman Khan et al. (2009). Moderate to high GCV and PCV values recorded for all the characters studied, indicating presence of high amount of variability indicating a low influence environment on their expression. Sharma and Sharma (2007) also reported selection based on phenotypic performance of these traits would be effective to bring about considerable improvement in these characters.

131


Table 1: Cultures and their parentage used for study S. No 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Entries JGL 15204 JGL 15230 JGL 15185 JGL 13621

Parentage MTU-4870/JGL 1798 MTU 4870/JGL 420 MTU 4870/JGL 418 JGL 418/Vijetha

JGL 15663

JGL 410/Betangamblin

JGL 16257

JGL 1851/IET 8585

JGL 16259

JGL 1851/IET 8585

JGL 16261


JGL 16268


JGL 16284

JGL 384/CR 311-34

JGL 16277

JGL 1798/Twin Rice

JGL 11118

IET 8585/JGL 1798

JGL 11459

JGL 1798/G.Beton

JGL 17004

WGL 14377/JGL 3855

JGL 17025

WGL 14377/JGL 3844

JGL 17196

JGL 326/MTU 1010

JGL 17197

JGL 326/MTU 1010

JGL 17204

JGL 1797/MTU 1010

JGL 384

BPT 5204/Kavya

MTU 1010

MTU 2077/IR 64

Grain Type* MS MS MS MS MS

Resistance Gallmidge biotype 3 Gallmidge biotype 3 Gallmidge biotype 3 Gallmidge biotype 3 Gallmidge biotype 3

MS

Gallmidge biotype 3

MS

Gallmidge biotype 3

MS

Gallmidge biotype 3

MS

Gallmidge biotype 3

MS

Gallmidge biotype 3

SB

Gallmidge biotype 3

MS

Gallmidge biotype 3

MS

Gallmidge biotype 3

MS

Gallmidge biotype 3

MS

Gallmidge biotype 3

MS

Gallmidge biotype 3

MS

Gallmidge biotype 3

MS

Gallmidge biotype 3

MS

Gallmidge biotype 3

LS

Brown Plant Hopper

21

LB Brown Plant Hopper MTU 1001 Vajram/MTU 7014 * MS-Medium slender; SS-Short Bold; LS-Long slender; LB-Long bold

Table 2. ANOVA for yield and yield components for stability in rice

Source

Environment (E)

20

2

40

120

237.39**

177.57**

6.98

0.5

1301.65

333.26

1196.35

3.79

Effective bearing tillers/ m

2265.25

12010.59**

1437.46

213.43

Panicle length (cm)

4.64 **

22.55***

0.94

0.23

Test weight (g)

20.04**

1.23

0.83

0.27

4991.34**

16694.25**

763.37

301.31

22950.25**

1927.38

9642.27**

753.26

df Days to 50% flowering Plant height (cm)

GxE

Pooled Error

Genotypes (G)

2

Number of grains per panicle 2

Seed yield/m (g)

132


Table 3. Components of genetic variability in rice.

Days to 50% flowering Days to maturity Panicle bearing tillers / m2 Panicle length (cm) 1000 grain weight (g) Grains/ panicle Grain yield/ m2 (g)

Genetic advance (% of mean)

Range of variation

General mean

GCV (%)

PCV (%)

Heritability (broad sense)

Genetic advance

85-111

102

6. 9

9.29

0.55

13.72

13.50

81.5- 122.6

103.2

9.7

12.9

0.57

19.94

19.32

263.1-395.1

319.8

9.2

15.8

0.34

44.87

14.03

22.45-26.3

24.5

4.82

7.5

0.36

1.74

7.10

14.7-22.5

17.1

12.8

16.2

0.60

4.34

25.45

144.8-298.9

206.9

18.9

27.6

0.47

70.89

34.25

521.6-851.3

690.8

11.6

18.0

0.42

136.67

19.78

Characters

Table 4. Clustering pattern of 21 rice cultures. Cluster No.

No. of Genotypes

I

11

JGL15204, JGL15663, JGL16257, JGL16268, JGL16277, JGL16284, JGL16259, JGL384 ,JGL15230, JGL16261, JGL11459

II

6

JGL17004, JGL17025, JGL17196, JGL17197, JGL17204, JGL11118

III IV

1 1

JGL13621

V

1

MTU1001 JGL15185

VI

1

MTU1010

Genotypes

Table 5. Contribution of different characters towards genetic divergence Characters % Contribution Days to 50% flowering 11.9 Days to maturity 29.5 Panicle bearing tillers /m2 7.1 Panicle length (cm) 8.1 1000 grain weight(g) 24.8 Grains/ panicle 4.8 Grain yield/ m2 (g) 13.8

133


Table 6. Cluster mean values for 7 characters of 21 in rice Cluster No. I II III IV V VI

Days to 50% flowering 109.9 93.4 109.3 111.7 109.7 97.0

Plant height (cm) 86.0 103.2 112.4 78.6 107.3 94.2

panicle bearing tillers/ m2 342.4 298.4 360.0 405.0 226.0 426.0

Panicle length (cm) 22.8 24.9 22.1 21.6 23.1 23.5

Table 7. Average intra (bold) and inter cluster distance valves I II III IV V Cluster No. I 3.9 3.1 3.9 3.2 2.0 II 3.9 3.5 6.3 4.3 2.3 III 3.1 3.5 5.2 3.7 0.0 IV 3.9 6.3 5.2 5.6 0.0 V 3.2 4.3 3.7 5.6 0.0 VI 3.7 5.1 4.7 2.6 4.8

Stanfield (1971) reported that heritability is considered high when the value is greater than 0.50 and medium between 0.20 to 0.50. High heritability with moderate genetic advance (percentage) mean were observed for 1000 grain weight (0.60, 25.45), days to 50% flowering (0.57, 19.32), plant height (0.55, 13.50) indicating the additive genetic effects. Prasad et al. (2001), Borbora et al. (2005) and Jaiswal et al. (2007), Verma (2010) and Sangam Kumar Singh et al. (2011) reported high heritability coupled with high genetic advance indicating preponderance of additive genetic effects. The highest genotypic and phenotypic coefficient of variation, heritability and genetic advance per cent of mean corresponded to grains per panicle, seed yield, 1000 grain weight and plant height and

Test weight (g) 16.0 16.9 16.5 23.8 14.2 25.9

Number of grains per panicle 198.3 166.3 242.3 131.0 155.0 102.7

Grain yield/m2 (g) 692.2 692.4 828.0 868.7 418.7 622.0

VI 3.7 5.1 4.7 2.6 4.8 0.0

direct selection for these traits would be rewarding for yield improvement in rice. Johnson et al. (1955) also suggested that high GCV along with high heritability and genetic advance gave a better indication for the selection of the genotypes. Similar results were also reported by Sarkar et al. (2007), Anbanandan et al.(2009) and Sangam Kumar Singh et al. (2011). All 21 genotypes were grouped into six clusters, using the Tochers method, in such a way that all the genotypes within the cluster had smaller D2 values among themselves than those belonging to different clusters (Table 4). Pattern of distribution genotypes among various clusters reflected the considerable genetic variability present in the genotypes under study. Cluster I and II each

134


comprised of maximum number of genotypes 11 and 6 respectively. Four clusters III (JGL 13621), IV (MTU 1001), V (JGL 15185) and VI (MTU 1010) with monophyletic genotypes. Most of the Jagtial cultures fell within cluster I and II. Interestingly most of the genotypes with same pedigree either with male or female parent involved cross combinations were within the same cluster and few genotypes in different cluster. The possible reason could be due to unidirectional selection practiced by the breeders in developing new elite lines. It was also observed that genotypes of quite different pedigree may fall into the same cluster, due to unidirectional selection pressure that could yield the genotypes, which were genetically closer than their parents. Likewise, it is also true that selection produce genetically diverse genotypes of same pedigree. This indicates that the pedigree record may not necessarily be an indicator of genetic divergence. Plant height (29.5%) followed by 1000 seed weight (24.8%), grain yield (13.8%) and days to 50% flowering (11.9%) contributed maximum towards the total divergence (Table 5), indicating that these characters contributing most of the divergence and importance should be given for effective selection and the choice of parents for hybridization. The data on character means for six clusters indicated (Table 6) that, cluster IV showed the highest mean for two characters viz., days to 50% flowering and grain yield. Cluster III recorded highest mean for number of grains per panicle

and plant height, and panicle bearing tillers/m2 and test exhibited by cluster VI. This indicated that the parents selected for hybridization on the basis of these characters are represented to be genetically diverse. Nayak et al. (2004) and Baradhan and Thangavel (2011) also reported the selection and choice of parents mainly depends upon contribution of characters towards divergence. The maximum inter-cluster distance of 6.30 existed between cluster II and IV followed by between cluster IV and VI (5.6), between cluster III and IV (5.2) and between cluster II and VI (5.1) (Table 7). The lowest inter-cluster distance (2.6) was found between cluster IV and VI, indicating a close relationship between them. The genotypes grouped into same cluster displayed the lowest degree of divergence from one another, and in the case where crosses are made between genotypes belonging to the same cluster; no transgressive segregants are expected from such combinations. Therefore, hybridization programmes should always be formulated in such a way that the parents belonging to different clusters with maximum divergence could be utilized to get desirable transgressive segregants. Although, for final selection of the parents for breeding programme, the genotypes to be used may be selected almost without exception or its proven performance in the areas of intended use including quantitative characters and include in crossing with the existing varieties for their further improvement (Allard, 1960). Genotypes from the clusters II and

135


Cluster VI could be utilized in the hybridization programme for getting desirable transgressive segregants for the development of early duration, resistance to gall midge and BPH. Similarly, the genotypes from the clusters I and IV could be utilized in the hybridization programme for obtaining desirable transgressive segregants for medium duration with resistance to BPH and gall midge. All the above clusters may also useful for the getting medium and long slender grain types.

REFERENCES Abdas Salam Khan, Muhammad Imran, Muhammad Asfaq (2009). Estimation of genetic variability and correlation of grain yield components of rice (Oryza sativa L.). American Eurasion J. Agric and Env. Sci. 6(5): 585-590. Allard RW (1960). Principles of Plant Breeding. John Wiley and Sons Inc., New York. pp 115-128. Anbanandan V, Saravanan K, Sabesan T (2009). Variability, heritability and genetic advance in rice (Oryza sativa L.). Intl. J. Plant Sci. 3(2): 61-63 Banumathy S, Manimaran R, Sheeba, A, Manivannan N, Ramya B, Kumar D, Ramasubramanian GV (2010). Genetic diversity analysis of rice germplasm lines for yield attributing traits. Electron. J. Plant Breed. 1(4): 500-504. Baradhan G, Thangav P (2011). D2 analysis in rice (Oryza

sativa L.). Plant Archives 11(1): 373-375. Borbora TK, Hazarica G N, Medhi AK (2005). Correlation and path analysis for plant and panicle characters in rice (Oryza sativa L.). Crop Research 30: 215-220. Burton GW, Dewane EH (1952). Estimating heritability in Jali Fescue (festuca arundinances) from replicated clonal material. Agron. J. 45: 478-481. Chaturvedi HP, Maurya DM (2005). Genetic divergence analysis in rice (Oryza sativa L.). Adv. Plant Sci. 18(1): 349353. Deshpande VN, Dalvi VV (2006). Genotype/Environment interactions in hybrid rice. Oryza 43 (4): 318-319. Jaiswal HK, Srivastava AK, Dey A (2007). Variability and association studies in indigenous aromatic rice (Oryza sativa L.). Oryza, 44(4): 351-353. Johnson HW, Robinson HE, Comstock RE (1955). Estimate of genetic and environmental variability in soybean. Agron. J. 47: 314318. Kandamoorthy S, Govindarasu R (2005).Genetic divergence in extra early rice (Oryza sativa L.) under two culture systems. Indian J. Genet. 65(1): 43-44. Mahalanobis PC (1928). On test and measures of group divergence. J. Proc. Asia. Soc. Bengal., 26: 541-588. Murty BR, Arunachalam V (1966).The nature of genetic diversity in relation to breeding system in crop

136


plant. Indian J. Genet. 26: 188–198. Nayak AR, Chaudhury D, Reddy JN (2004). Genetic divergence in scented rice. Oryza, 41: 79-82. Panse VG, Sukhatme PV (1985). Statistical methods for agricultural workers, ICAR, New Delhi, 4th Edn . Prasad B, Parway AK, Biswas PS (2001). Genetic variability and selection criteria in fine rice (Oryza sativa L.). Pak. J. Biol. Sci. 4: 1188-1190. Rao CR (1952). Advanced statistical methods in biometrical research. 1st Edn. Wiley and Sons Inc., New York. Sabesan T, Suresh R, Saravanan K (2009). Genetic variability and correlation for yield and grain quality characters of rice grown in coastal saline low land of Tamilnadu. Electron. J. Plant Breed. 1: 56-59. Sabesan T, Saravanan K (2008). Genetic divergence analysis in rice (Oryza sativa L.). Paper resented in Golden Jubilee Commemorative National seminar on “Fifty years of Indian agriculture: Problems, Prospects and Future thrusts”, Annamalai University, Tamil Nadu, India. 20-21 March, 2008. pp 14. Saidaiah P, Sudheer Kumar S, Ramesh MS (2011). Stability analysis of rice (Oryza sativa) hybrids and their parents. J. Agric. Sci. Vol. 81(2): 111-115. Sangam Kumar Singh, Chandra Mohan Singh, Gil GM

(2011). Assessment of genetic variability for yield and its component character in rice. Research In Plant Biol. 1(4):73-76. Sarkar KK, Bhutia KS, Senapathi BK, Roy SK (2007). Genetic variability and character association of quality traits in rice (Oryza sativa L.). Oryza 44(1): 6467. Senapati BK, Sarkar G (2005). Genetic divergence in tall indica rice (Oryza sativa L.) under rainfed saline soil of Sundarban. Oryza, 42(1): 70-72. Sharma AK, Sharma R N (2007). Genetic variability and character association in early maturing rice. Oryza, 44(4): 300-303. Sridhar S, Dayakar Reddy T, Ramesha MS (2011). Genotype/Environment interaction and stability for yield and its components in hybrid rice cultivars (Oryza sativa L.). Int. J. Plant Breed. Genet. 5(3): 194-208. Stanfield WD (1971). Genetica. Teoria Y 400 problems resuelets. Serie Scham, Mcgraw hill, Mexico.405pp. Verma U (2010). Genetic diversity analysis in exotic rice genotypes. M.Sc. thesis, Dept. of GPB, Alhabad. Zahid MA, Akhtar M, Sabir M, Manzoor Z, Awan T (2006). Correlation and path analysis studies in yield and economic traits in basmati rice (Oryza sativa L.), Asian J. Pl. Sci., 5(4): 643-645.

137


PROTOGYNY IS AN ATTRACTIVE OPTION OVER EMASCULATION FOR HYBRIDIZATION IN PIGEONPEA A. K. CHOUDHARY1,*, M. A. IQUEBAL2, N. NADARAJAN1 1

Indian Institute of Pulses Research, Kanpur 208 024, India Indian Agricultural Statistics Research Institute, New Delhi 110 012, India * Corresponding author: [email protected]

2

SUMMARY Hybridization is the initial step for creating genetic variability in conventional breeding programmes. Pigeonpea is one of the crops in which some genetic mechanisms including protogyny tends to promote natural outcrossing. Considering these facts in view, an investigation was performed to assess the feasibility of crossing without emasculation. Sixteen crosses were made with and without emasculation during the year 2006-2007 and 2007-2008. The number of crossed pods was significantly greater under crossing without emasculation than with emasculation. “Selfs” were observed in both the schemes, which could be ascribed to chance events and could further be brought down to zero per cent by selecting appropriate buds depending on environmental conditions. The results showed that protogyny-mediated hybridization is an alternative method to crossing involving emasculation in pigeonpea. Key words: Cajanus cajan, crossing, emasculation, protogyny, selfs. Manuscript received: January 10, 2011; Decision on manuscript: April 4, 2012; Manuscript accepted in revised form: April 28, 2012. Communicating Editor: Bertrand Collard

INTRODUCTION Pigeonpea (Cajanus cajan L. Millspaugh) is a major food legume of the tropics and subtropics. Globally, it is grown in more than 80 countries of Asia, Eastern and Southern Africa (ESA), Latin America and the Caribbean on 4.86 million hectares (M ha) with an annual production and mean productivity of 4.10 million tons (Mt) and 844 kg/ha, respectively (FAOSTAT, 2011).

India has the largest acreage under pigeonpea (3.90 M ha) with a total production and productivity of 2.89 Mt and 741 kg/ha, respectively (DAC, 2011). Despite its main use as de-corticated, dried split peas (dal), the use of immature seeds is very common as fresh vegetable in some parts of India such as Gujarat, Maharashtra and Karnataka. Besides this, in the tribal areas of various states, the use of pigeonpea as green

138

Choudhary et al. (2012)

vegetable is also common (Saxena et al., 2010). Pigeonpea is considered as a drought tolerant crop with a large temporal variation for maturity period. As a result, it is widely adapted to a range of environments and cropping systems (Choudhary et al., 2011). Broadly, four maturity groups are recognized in pigeonpea: extra early (90 – 120 days), early (>120 – 150 days), medium (>150 – 200 days), and late (>200 – 300 days). These variations for maturity have direct relevance on the survival and fitness of the crop in different agroecological niches (Choudhary, 2011). The cleistogamous flowers of pigeonpea predominantly favour self-fertilization. However, considerable extent of natural outcrossing has been reported (Saxena et al., 1990). According to Onim (1981), pigeonpeas shed pollen while flowers are still in the bud stage, and they do not start germinating until the flowers start to wither 24-48 hr after their anthers dehiscence. It has also been observed that germination and development and growth of native pollen tube down the style is very slow due to the presence of weak self-incompatibility (Dutta and Deb, 1970; Choudhary, 2011) which may be more of sporophytic than gametophytic in essence, taking 54 h to reach the base of the ovary. As a result of these two mechanisms, considerable outcrossing takes place mostly due to pollen transfer by different species of honey bees (Choudhary, 2011). Anthesis in pigeonpea usually occurs during 8.0 – 17.0 h and flowers remain open for 36 to

48 h. The stigma becomes receptive 68 h before anther dehiscence (protogyny). However, the receptivity of stigma continues for further 20 h after anthesis (Prasad et al., 1977), thereby opening up avenues for formation considerable “crossed” seeds. Therefore, the presence of weak self- incompatibility (Choudhary, 2011) and protogyny (Reddy and Mishra, 1981) provides scope for crossing even without emasculation in pigeonpea. The present investigation presents the comparative results of crossing with and without emasculation and examines the efficiency of both methods in pigeonpea.

MATERIALS AND METHODS The experimental materials comprised 12 genotypes of pigeonpea, viz., ‘IPA 7-1’, ‘IPA 73’, ‘IPA 7-5’, ‘Kudrat 3’, ‘Bahar’, ‘NA 1’, ‘IPA 92’, ‘MAL 13’, ‘T 7’, ‘IPA 06-1’, ‘BDN 1’, ‘BDN2’, ‘Ranchi Local’ and ‘UPAS 120’. These genotypes were selected on the bases of different morphological markers (Table 1). These were “pure” lines for one or more marker traits. These genotypes are maintained at the Indian Institute of Pulses Research (IIPR), Kanpur through selfing. A total of 16 crosses were attempted during the winter season of the year 2006-07. In all the crosses, female and male parents were qualitatively different for one or more morphological traits such as stem colour, plant type, petal (standard) colour, pod colour, and the like. Each cross was made following two schemes: one with

139


emasculation and the second without emasculation. In the first scheme, fully developed buds (likely to open after 2 days) were emasculated during evening hour and hybridization was performed the next morning between 9-10 h. In the second scheme, pollinations on selected buds were done on the same day during the same period. In each scheme, pollinations were performed on 100 floral buds on the same female plant (10 buds/ day) for a given cross. The same pollen source was used for pollination in both the methods. Pods developed on female plants of individual cross under both the crossing schemes were counted and picked upon maturity, threshed and kept separately. In the following year (2007-2008), seeds of the crosses of each set were sown in the field along with the respective female parents to compare individual plants of putative crosses for plant types, petal colour, pod colour and other marker traits. On the bases of morphological markers, “selfs” (if present at all) and true F1s were identified and tagged. Just before flowering, these plants were covered with nylon net of size 110 × 90 cm2 with 2 mm mesh size to control its pollination system, thereby ensuring 100% selfpollination. Pods from “selfs” and true F1’s were harvested and threshed and seeds were kept separately to observe their breeding behaviour in the coming season. The same set of crosses (as attempted during the year 20062007) under both the schemes of crossing was attempted again during this year (2007-2008) to

confirm the results of the previous year. During the cropping season 2008-2009, seeds from “selfs” and true F1s of the preceding year were sown along with the parents to compare the breeding behaviour of descendants with respect to marker traits. The same exercise as done in the year 2007-2008 for assessing the breeding behaviour of putative crosses was also repeated during this year for crosses attempted during the year 2007-2008. Seeds obtained from “selfs” and true F1s were kept separately and sown during 2009-2010 to confirm the results of the preceding year in the same manner as practiced during the year 2008-2009. Statistical analysis The statistic “t-test” was applied to compare the average difference in the pod setting between two schemes of crossing (with and without emasculation). The same statistic was also applied for test of significance of average differences in the percentage of “selfs” obtained with two schemes of crossing (with and without emasculation). The analysis was carried out using SPSS version 16.0 software.

RESULTS AND DISCUSSION The number of crossed pods set on emasculated plants varied from 5 (BDN 1 × Kudrat 3 and BDN 2 × Kudrat 3) to 23 (T 7 × Kudrat 3) and from 0 (MAL 13 × Kudrat 3) to 12 (T 7 × Kudrat 3) during the year 2006-2007 and 2007-2008, respectively with the mean value (over two years) ranging from 3

140


(BDN 2 × Kudrat 3) to 17.5 (T 7 × Kudrat 3). However, average pod setting (over two years) in the second scheme ranged from 3.5 to 22.5 for the cross ‘BDN 2’ × ‘Kudrat 3’ and ‘IPA 7-3’ × ‘IPA 75’, respectively (Table 2). The difference in pod setting under the two schemes of crossing was statistically significant (t=2.33; P < 0.03). Pods set under crossing without emasculation was significantly greater than that under crossing with emasculation. These crosses were grown in the succeeding cropping season along with the female parents to confirm whether crosses so-derived were true hybrids or “selfs”. The percentage “selfs” ranged from 0.0 (IPA 92 × IPA 6-1, BDN 2 × Kudrat 3 and NA 1 × MAL 13) to 13.33 (IPA 7-3 × IPA 7-5) and from 0.0 (IPA 7-1 × IPA 7-3, IPA 6-1 × Bahar, T 7 × NA 1, Kudrat 3 × MAL 13, Ranchi Local × T 7, BDN 2 × Kudrat 3, MAL 13 × IPA 6-1 and NA 1 × MAL 13) to 20 (Bahar × Kudrat 3) in the years 2007-2008 and 2008-2009, respectively when crosses were made following emasculation. Under the second scheme of crossing (without emasculation), the range of “selfs” were observed from 0.0% (IPA 92 × IPA 6-1, T 7 × NA 1 and MAL 13 × Kudrat 3) to 22.22% (Bahar × T 7) and from 0.0% (IPA 6-1 × Bahar, IPA 92 × IPA 6-1, Ranchi Local × T 7, BDN 2 × Kudrat 3 and MAL 13 × Kudrat 3) to 25% (Kudrat 3 × MAL 13 and BDN 1 × Kudrat 3) during the year 2007-08 and 200809, respectively. The mean

percentage “selfs” (over two years) varied from 0.0 (IPA 6-1 × Bahar, BDN 2 × Kudrat 3 and NA 1 × MAL 13) to 14.81 (IPA 7-3 × IPA 7-5) and from 0.0 (IPA 92 × IPA 61and MAL 13 × Kudrat 3) to 20.00 (Bahar × T 7) under the two schemes of crossing, respectively. The difference in mean percentage “selfs” obtained under the two schemes was statistically nonsignificant (t=1.93; P > 0.06). However, “Selfs” under the second system of crossing (crossing without emasculation) were numerically high (Table 3). Selfed progenies from both “selfs” and true F1s derived under both schemes of mating were assessed for their breeding behaviour. All the selfed progenies from “selfs” bred true to the type, that is, all resembled their female parents. However, hybrid progenies segregated for different marker traits such as petal colour, plant type, pod colour, stem pubescence, days to maturity, and the like (Table 4). It is a known fact that conventional plant breeding depends on hybridization between diverse parents for creation of new genetic variability. In pigeonpea like other crops, this is accomplished by emasculating flowers on female plants followed by placing pollens from male plants onto the stigma of emasculated flowers. However, during the process much injury is caused to the ovary and style of the emasculated flowers, resulting in reduced pod setting.

141


Table 1. Description of pigeonpea genotypes. Genotypes IPA 7-1

Pedigree Selection from a landrace ‘Kudrat 3’

IPA 7-3

Selection from a landrace ‘Kudrat 3’

local

IPA 7-5

Selection from a landrace ‘Kudrat 3’

local

Kudrat 3

A local landrace, selected from Varanasi area of U.P. (India) Selection from a landrace of Motihari district in Bihar (India) Selection from a landrace of Faizabad district of U.P. (India) Selection from a local collection, ‘98-3’ (MA 2 × MA 160) × Bahar

Bahar

NA 1

IPA 92 MAL 13

local

T7

Selection from a landrace of Lucknow district of U.P. (India)

IPA 06-1

Selection from a landrace of Etwah district of U.P. (India)

BDN 1

Selection from local landrace ‘Bori’ (India)

BDN 2

Selection from ‘Bori II – 132-A-1’(India) A land race of Ranchi, Jharkhand (India) Selection from ‘P 4768’

Ranchi Local UPAS 120

Distinguishing (marker) characters Tall and compact plant type, dorsal surface of standard petal is dark red in colour Medium height and semi-spreading, standard petal is golden yellow, pods are green with sparse black streaks Medium height and semi-spreading, purple pods (unripe), standard petal is dark red Medium height and compact, dorsal surface of standard petal is pink Compact plant type with golden yellow colour of standard petal and purple pods (unripe) A long-duration variety with dense red streaks on dorsal surface of standard petal with green pods A late pigeonpea line of spreading growth habit and green stem colour A large seeded (14g/100 seeds), spreading, long-duration variety with constricted pods and greenish yellow standard petal A very late (280-300 days) and tall (2.5-3.0 m) variety of long-duration pigeonpea with compact plant type and green stem colour A large seeded (> 15 g/100 seeds) pigeonpea line having purple stem colour, only 3-4 primary branches, very prominent strophiole, dark red standard and high sensitivity to low temperature An old variety of medium maturity group (180 days) with yellow standard Indeterminate variety of medium maturity with white seeds A large seeded pigeonpea line of medium duration with pink standard A short-duration variety (120-150 days) with yellow standard

142


Table 2. Differences in pod setting on plants pollinated with and without emasculation. Crosses

IPA 7-1 × IPA 7-3 IPA 7-3 × IPA 7-5 Bahar × Kudrat 3 Bahar × IPA 6-1 Bahar × T 7 IPA 92 × IPA 61 T 7 × Kudrat 3 T 7 × NA 1 UPAS 120 × Kudrat 3 Kudrat 3 × MAL 13 Ranchi Local × T7 BDN 1 × Kudrat 3 BDN 2 × Kudrat 3 MAL 13 × IPA 6-1 MAL 13 × Kudrat 3 NA 1 × MAL 13 Mean SE (Mean) CD (P = 0.05) t value (for differences) P value for t-test

No. of pods set on emasculated plants 2006- 2007- Mean 07 07 19 4 11.5

No. of pods set on plants without emasculation 2006-07 2007Mean 08 24 12 18.0

13

7

10.0

31

14

22.5

11

3

7.0

13

6

9.5

12 12 09

3 8 6

7.5 10.0 7.5

12 17 10

3 12 12

7.5 14.5 11.0

23 07 18

12 8 11

17.5 7.5 14.5

25 09 29

13 9 14

19.0 9.0 21.5

10

2

6.0

12

4

8.0

11

5

8.0

11

6

8.5

05

5

5.0

13

8

10.5

05

1

3.0

05

2

3.5

13

3

8.0

17

9

13.0

10

0

5.0

14

8

11.0

10 11.8 1.21 2.47 --

9 5.4 0.87 1.78 4.22

9.5 8.6 0.90 1.85 --

10 15.8 1.89 3.92 --

12 9.0 0.98 2.03 3.16

11.0 12.4 1.34 2.73 2.33

--