Idea Transcript
GENETIC PARAMETERS, PHYSIOLOGICAL AND MOLECULAR ANALYSIS OF ROOT AND SHOOT TRAITS RELATED TO DROUGHT TOLERANCE IN RICE (Oryza sativa L.)
VEERESH GOWDA R.P. PAK 50 45
DEPARTMENT OF GENETICS AND PLANT BREEDING UNIVERSITY OF AGRICULTURAL SCIENCES BENGALURU
2010
GENETIC PARAMETERS, PHYSIOLOGICAL AND MOLECULAR ANALYSIS OF ROOT AND SHOOT TRAITS RELATED TO DROUGHT TOLERANCE IN RICE (Oryza sativa L.)
VEERESH GOWDA R.P. PAK 50 45
Thesis submitted to the
University of Agricultural Sciences, Bengaluru In partial fulfillment of the requirements For the award of the degree of
Doctor of Philosophy (Agriculture) in
Genetics and Plant Breeding BENGALURU
JULY, 2010
Affectionately Dedicated to My Beloved Parents Smt. Laxmi Devi Sri. Po mpana Gowda R. and Friend, Kalmesh
DEPARTMENT OF GENETICS AND PLANT BREEDING UNIVERSITY OF AGRICULTURAL SCIENCES G.K.V.K. BENGALURU - 560 065 CERTIFICATE This is to certify that the thesis entitled “Genetic parameters, physiological and molecular analysis of root and shoot traits related to drought tolerance in rice (Oryza sativa L.)” submitted by Mr. VEERESH GOWDA R.P., ID No. PAK 5045 in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY (AGRICULTURE) in Genetics and Plant Breeding to the University of Agricultural Sciences, GKVK, Bengaluru is a record of research work done by him during the period of his study in this university under my guidance and supervision, and the thesis has not previously been formed the basis for the award of any degree, diploma, associateship, fellowship or other similar titles.
Bangalore
(Dr. H.E. SHASHIDHAR) Professor, Plant Biotechnology, UAS, GKVK, Bengaluru
July, 2008
APPROVED BY Chairman
Members
:
:
___________________________ (Dr. H.E. SHASHIDHAR))
: 1. ___________________________ (Dr. K. P. VISHWANATHA)
2.
___________ ________________ (Dr. D. L. SAVITHRAMMA)
3.
__________________________ (Dr. N. B. PRAKASH) __________________________ (Dr. VIJAYALAKSHMI)
4.
ACKNOWLEDGEMENT It is my heart’s turn to express my deepest sense of gratitude to all of those who directly and indirectly helped me in this endeavour. At the very outset, I fell inadequacy of words to express my profound indebtedness
and
deep
sense
of
gratitude
to
my
esteemed
chairman
Dr. Shashidhar H.E., Professor, Department of Plant Biotechnology, GKVK, University of Agricultural Sciences, Bengaluru for his esteemed stewardship, enabling guidance, cherishable counselling and personal affection for which I am greatly indebted to him. It was really a great pleasure and privilege for me to be associated with him during my M. Sc. Degree Programme. It gives me immense pleasure to express my heartfelt thanks to the supervisor’s of my research work Dr. S. Rachid Serraj, Senior Scientist, CESD, IRRI, Philippines. Dr. Ajay Kohli, Senior Scientist, CESD, IRRI, Philippines. Dr. Vincent Vadez, Senior Scientist, ICRISAT, India. Dr. Ram T, Senior Scientist DRR, India. for their valuable counsel, note-worthy guidance and cordial co-operation during the course of investigation. I would like to thank Dr. Shailaja hittalmani, Professor and Head, Department of Genetics and Plant Breeding, GKVK, University of Agricultural Sciences, Bangalore who provided all kind of support to me in completion of Doctoral study. I feel no words to express my heartfelt gratitude and respect to all his kindness. It also gives me immense pleasure to express my heartfelt thanks to the members of my advisory committee Dr. K. P. Vishwanath, Professor, GKVK, University of Agricultural Sciences, Bangalore. Dr. D. L. Savithramma, Professor, department of Soil
Science, and Dr. N. B. Prakash, Professor for their valuable counsel, note-worthy guidance and cordial co-operation during the course of investigation. I owe a lot to my parents, brother and sisters, without whose affection, support and sacrifice, this study would scarcely have been accomplished. My diction is poor to translate into words the sense of gratitude, heartfelt respect and affection to my father Mallappa Patne mother Kalavati M. Patne and sisters Anita, Sunita and Sangeeta and my brother Revansiddappa Patne and my grand mother Shantveeramma for their moral support, boundless love and building unshakable confidence in me which motivated as a force to fulfill this long cherished ambition. I bow my head with overwhelming respect and thanks to all people in ZARS, Hiriyur, Chitradurga especially to Field supervisors Basavaraj, Kumar and Basavaraj who helped me lot during my stay at Hiriyur. I use this opportunity to sincerely thank my dearest classmate’s Abdul, Rahamani, Chetan, Dhanajaya, Gopal, Madan, Matruthi, Nandhini, Prasanna, Praveen and Suresh, for their lovely friendship, help and care and for making the two year study very much enjoyable and memorable. Words could not help me when I need to thank my dear friends, Patil, Jaba, Kallihal, Mahesh Salimath, Sashi, Vishnu Vardhan, M. P, Yankareddy, Sharanappa, Gundappa, Sangappa and Naveen Dhama, for the great support they gave me. I fondly thank my senior friends Dr. Kalmeshwer Gouda Patil, Ramesh Patil, Kamala Kant and Muniraj who provided me their valuable guidance and to my room mates, Akhter and Jai Kumar for all their help. I am overwhelmed with gratitude to all my respondents, without whose whole hearted co-operation, this study would not have been fruitful.
Bangalore July, 2008
(Nagesh)
CONTENTS CHAPTER NO.
TITLE
I
INTRODUCTION
II
REVIEW OF LITERATURE
III
MATERIAL AND METHODS
IV
EXPERIMENTAL RESULTS
V
DISCUSSION
VI
SUMMARY AND CONCLUSION
VII
REFERENCES
PAGE NO.
LIST OF TABLES
TABLE NO.
TITLE
1.
List of OryzaSNP panel rice genotypes used for real time water uptake rates, root and grain yield parameters under both field and lysimetric experiments.
2.
Climatic conditions at experimental sites during 1-112 days after transplanting.
3.
Soil properties of experimental site used for screening OryzaSNP panel rice accessions during DS2008.
4.
Mean climatic conditions at experimental site during 1-112 days after transplanting
5.
List of parents of mapping population, donors and advanced breeding lines, of IRRI-India drought breeding network used for analyzing real time water uptake rates, root distribution and grain yield under both field and lysimetric experiments
6.
List of NILs and parents used for both gene expression and lysimetric experiments.
7.
List of LEA primers used for gene expression study and their sequence
8.
Means of root number (RN), root to shoot ratio (RSR) and root length density (RLD, cm cm-3) across different soil depths of OryzaSNP panel rice accessions under both drought stress and well- watered treatment in field experiment during DS 2008
9.
Means of root surface area (RSA, cm2) across different soil depths of OryzaSNP panel rice accessions under both drought stress and well-watered treatment in field experiment during DS 2008
10.
Means of root volume (RV, cm3) across different soil depths of OryzaSNP panel rice accessions under both drought stress and well watered treatment in field experiment during DS 2008 Means of root dry weight (RDW, g) across different soil depths of OryzaSNP panel rice accessions under both drought stress and well-watered treatment in field experiment during DS 2008
11.
PAGE NO.
LIST OF TABLES
TABLE NO.
TITLE
12.
Means of physiological character of OryzaSNP panel rice accessions measured under drought stress treatment in field experiment during DS 2008.
13.
Means of shoot parameters of OryzaSNP panel rice accessions under both drought stress and well-watered treatments in field experiment during DS 2008
14.
Means of grain yield (GY g/m2), straw biomass (SB, g/m2) and harvest index (HI) of OryzaSNP panel rice accessions under both well watered and drought stress treatments in field experiment during DS 2008
15.
Means of shoot and grain yield parameters of OryzaSNP panel rice accessions under both drought stress and well-watered in field experiment during DS 2009.
16.
Analysis of variance for root and shoot characters of OryzaSNP rice panel accessions under drought stress and well-watered condition in field experiment during DS 2008
17.
Analysis of variance for physiological characters of OryzaSNP rice panel accessions under drought stress condition in field experiment during DS 2008
18.
Analysis of variance for shoot and grain yield characters of OryzaSNP rice panel accessions under drought stress and well-watered condition in field experiment during DS 2009
19.
Mean, range and genetic parameters for shoot and physiological characters of OryzaSNP panel rice accessions under drought stress and well watered condition in field experiment during DS 2008.
20.
Mean, range and genetic parameters for root characters of OryzaSNP panel rice accessions under both drought stress and well-watered in field experiment during DS 2008.
PAGE NO.
LIST OF TABLES
TABLE NO.
TITLE
21.
Mean, range and genetic parameters for shoot and grain yield characters of OryzaSNP panel rice accessions measured under both drought stress and well-watered condition in field experiment during DS 2009.
22.
Phenotypic correlation co-efficients among root, shoot and grain yield traits of OryzaSNP panel rice accessions under drought stress in field experiment during DS 2008.
23.
Phenotypic correlation co-efficients among root distribution shoot and grain yield traits of OryzaSNP panel rice accessions under well watered condition in field experiment during DS 2008.
24.
Water uptake rates (g/plant) measured at different intervals during drought stress period using OryzaSNP panel accessions in lysimetric experiment during WS 2008
25.
Means of maximum root length (MRL, cm)), root number (RN), and root length density(RLD, cm/cm3) across different soil depths of OryzaSNP panel rice accessions under both drought stress and well-watered treatment in lysimetric experiment during WS 2008.
26.
Means of root surface area (RSA, cm2) across different soil depths of OryzaSNP panel rice accessions measured under both drought stress and well-watered treatment in lysimetric experiment during WS2008
27.
Means of root volume (RV, cm3) across different soil depths of OryzaSNP panel rice accessions measured under both drought stress and well-watered treatment in lysimetric experiment during WS 2008
PAGE NO.
LIST OF TABLES
TABLE NO.
TITLE
28.
Means of root dry weight (RDW, g) across different soil depths of OryzaSNP panel rice accessions measured under both drought stress and well-watered treatment in lysimetric experiment during WS 2008
29.
Means of root and shoot traits of OryzaSNP panel accessions under both drought stress and well-watered conditions in lysimetric experiment during WS 2008.
30.
Water uptake rate (g) measured at different intervals using OryzaSNP panel accessions under drought stress condition in lysimetric experiment during DS 2009.
31.
Means of root parameters of OryzaSNP panel rice accessions under both drought stress and well-watered conditions in lysimetric experiment during DS 2009.
32.
Means of shoot parameters of OryzaSNP panel accessions under drought stress and well-watered environment in lysimetric experiment during DS 2009.
33.
Water uptake rate (g/plant) measured at different intervals using OryzaSNP panel rice accessions under drought stress treatment in lysimetric experiment during WS 2009.
34.
Mean total water uptake (TWU, g/plant) during stress period under drought stress condition using OryzaSNP panel rice genotypes belonging to different rice types in three lysimetric experiments (WS2008 /IRRI; DS 2009 /ICRISAT; WS 2009/ IRRI).
35.
Analysis of variance for real time water uptake rates of OryzaSNP rice panel accessions under drought stress condition in three lysimetric experiments (WS 2008, DS 2009 and WS 2009)
PAGE NO.
LIST OF TABLES
TABLE NO.
TITLE
36.
Analysis of variance for root and shoot characters of OryzaSNP rice panel accessions under drought stress and well-watered condition in lysimetric experiment during WS 2008
37.
Analysis of variance root and shoot characters of OryzaSNP rice panel accessions under drought stress and well-watered condition in lysimetric experiment during DS 2009
38.
39.
Mean, range and genetic parameters for root characters of OryzaSNP panel rice accessions under drought stress and well-watered conditions in lysimetric trial during WS 2008. Mean, range and genetic parameters for shoot characters of OryzaSNP panel rice accessions under drought stress and well- watered conditions in lysimetric experiment during WS 2008.
40.
Mean, range and genetic parameters for root and shoot characters of OryzaSNP panel rice accessions under both drought stress and well-watered condition in lysimetric experiment during DS 2009.
41.
Phenotypic correlation co-efficients among water uptake, root distribution and shoot traits of OryzaSNP panel rice accessions under drought stress in lysimetric experiment during WS 2008
42.
Phenotypic correlation co-efficients among water uptake, root distribution and shoot traits of OryzaSNP panel rice accessions under well watered condition in lysimetric experiment during WS 2008.
PAGE NO.
LIST OF TABLES TABLE NO.
TITLE
43.
Phenotypic correlation co-efficients among root and shoot traits of OryzaSNP panel rice accessions under drought stress in lysimetric trial during DS 2009.
44.
Phenotypic correlation co-efficients among root and shoot traits of OryzaSNP panel rice accessions under well-watered condition in lysimetric experiment during DS 2009.
45.
Path analysis (phenotypic) indicating direct and indirect effects of component characters on shoot dry weight (SDW, g/plant) measured under drought stress condition using OryzaSNP panel rice accessions in lysimetric experiment during WS 2008
46.
Path analysis (phenotypic) indicating direct and indirect effects of component characters on shoot dry weight (g/plant) measured under well-watered condition using OryzaSNP panel rice accessions in lysimetric experiment during WS 2008.
47.
Grouping of OryzaSNP panel rice accessions based on D2 analysis using root and shoot traits measured in drought stress treatment during WS2008
48.
49.
50.
Grouping of OryzaSNP panel rice accessions based on D2 analysis using root and shoot traits measured in well watered treatment during WS 2008 Average intra and intercluster D2 values using root and shoot traits measured in drought stress treatment during WS 2008 Average intra and intercluster D2 values using root and shoot traits measured in well-watered treatment during WS 2008
PAGE NO.
LIST OF TABLES TABLE NO.
TITLE
51.
The nearest and farthest clusters from each cluster based on D2 values using root and shoot traits measured in drought stress treatment during WS 2008
52.
The nearest and farthest clusters from each cluster based on D2 values using root and shoot traits measured in well watered treatment during WS 2008
53.
Per cent contributions of twelve root and shoot characters towards diversity in drought stress and wellwatered condition.
54.
The mean values of clusters for twelve root and shoot characters measured under drought stress treatment in lysimetric experiment during WS 2008
55.
The mean values of clusters for root and shoot characters using root and shoot traits measured in wellwatered treatment during WS 2008
56.
Means of shoot and grain yield parameters of parents of mapping population, donors and advanced breeding lines of IRRI-India drought breeding network under both drought stress and well-watered condition in field experiment during DS 2009
57.
Analysis of variance for shoot characters under wellwatered and drought stress condition using parents of mapping population, donors and breeding lies of IRRIIndia drought breeding network in field experiment during DS 2009
58.
Mean, range and genetic parameters for shoot and grain yield characters of contrasting breeding lines and parents of mapping population measured under both drought stress and well-watered condition in field experiment during DS 2009.
PAGE NO.
LIST OF TABLES TABLE NO.
TITLE
59.
Water uptake rates (g/plant) of parents of mapping population, donors and advanced breeding lines of IRRIIndia drought breeding network in lysimetric experiment during DS 2009.
60.
Means of root parameters of contrasting parents of mapping population, donors and advanced breeding lines of IRRI-India drought breeding network under drought stress and well-watered conditions in lysimetric experiment during DS 2009
61.
Means of shoot parameters of parents of mapping population, donors and advanced breeding lines of IRRIIndia drought breeding network under both drought stress and well-watered condition in lysimetric experiment during DS 2009
62.
63.
64.
Analysis of variance for real time water uptake rates under drought stress condition using parents of mapping population, donors and breeding lies of IRRI-India drought breeding network in lysimetric experiment during DS2009 Analysis of variance for root and shoot characters under drought stress and well-watered condition using parents of mapping population, donors and breeding lies of IRRIIndia drought breeding network in lysimetric experiment during DS2009 Mean, range and genetic parameters for root and shoot characters of contrasting donors, parents of mapping population and breeding lines of IRRI –India measured under both drought stress and well-watered condition in lysimetric experiment during DS 2009.
PAGE NO.
LIST OF TABLES TABLE NO.
TITLE
65.
Phenotypic correlation co-efficients among root and shoot traits measured under drought stress in lysimetric trial during DS 2009 using contrasting parents of mapping population, donors and breeding lies of IRRIIndia drought breeding network.
66.
Phenotypic correlation co-efficients among root and shoot traits measured under well watered during DS 2009 using contrasting parents of mapping population, donors and breeding lies of IRRI-India drought breeding network.
67.
Water uptake rates at different time intervals using NILs and its parents in lysimetric experiment during WS 2008
68.
Means of root number (RN), maximum root length (MRL,cm) and root length density (RLD,cm cm-3) across different soil depth of Adeysel NILs along with parents under both drought stress and well-watered treatment in lysimetric experiment during WS 2008
69.
Means of root surface area (cm2) across different soil depth of Adeysel NILs along with parents under both drought stress and well watered treatment in lysimetric experiment during WS 2008
70.
Means of root volume (cm3) across different soil depth of Adeysel NILs along with parents under both drought stress and well watered treatment in lysimetric experiment during WS 2008
71.
Means of root dry weight (mg) across different soil depth of Adeysel NILs along with parents under both drought stress and well watered treatment in lysimetric experiment during WS 2008
PAGE NO.
LIST OF TABLES TABLE NO. 72.
73.
74.
75.
TITLE Means of shoot parameters of NILs and parents under both drought stress and well watered treatment in lysimetric experiment during WS 2008 Mean gene expression pattern under well water (1.0 FTSW) and severe drought stress (0.2 FTSW) conditions in different zones of shoot using IR64 and Dular. Mean gene expression pattern under well water (1.0 FTSW) and severe drought stress (0.2 FTSW) conditions using top zone of root in NILs of Adeysel X IR64 with IR64 and Dular. Mean gene expression pattern under well water (1.0 FTSW) and severe drought stress (0.2 FTSW) conditions using deep zone of root in NILs of Adeysel X IR64 with IR64 and Dular.
PAGE NO.
LIST OF FIGURES FIGURE NO.
TITLE
1.
Soil water potential measured by mercury-manometer tensiometers at soil depths of 15 and 30 cm. Values shown are mean ± s.e., n = 3. Measurements were not taken from days 72-94 as the field was irrigated for root sampling.
2.
Volumetric soil moisture at a soil depth of 70 cm, measured by diviner. Values shown mean ± s.e. , n = 6. Measurements were not taken from days 72-94 as the field was irrigated for root sampling.
3.
Rainfall pattern during stress period at experimental site during DS 2008, IRRI, Philippines.
4.
Volumetric soil moisture at a soil depth of 45 cm, measured by TDR. Values shown mean ± s.e. , n = 3. Measurements were not taken from days 22-24 as the field was irrigated.
5.
Rainfall pattern during stress period at experimental site during DS 2009, ICRISAT, India.
6.
Percent total root length distribution with depth for genotypes of Oryza SNP panel under drought stress treatment measured during DS 2008.
7.
Drought induced root growth at depth (30-45cm) in all OryzaSNP rice panel accessions
BETWEEN PAGES
LIST OF PLATES PLATE NO.
TITLE
1.
Overview of well- watered treatment during DS 2008 at IRRI, Philippines
2.
Overview of drought stress treatment during DS 2008 at IRRI, Philippines
3.
Method of root sampling under field condition using monolith sampler during DS 2008 at IRRI, Philippines
4.
Overview of well- watered treatment during DS 2009 at ICRISAT, India
5.
Overview of well- watered treatment during DS 2009 at ICRISAT, India
6.
Overview of lysimetric experiment during WS 2008 at IRRI, Philippines
7.
Lysimetric system used for real- time water uptake measurements during WS 2008 and WS 2009 at IRRI, Philippines
8.
Overview of lysimetric experiment during WS 2008 at IRRI, Philippines
9.
Method of root washing during DS 2009 at ICRISAT, India.
10.
Overview of pot experiment during WS 2009 at IRRI, Philippines
11.
Expression pattern revealed by semi quantitative RTPCR of LEA genes in top root tissues of drought resistant and susceptible checks
BETWEEN PAGES
LIST OF PLATES PLATE NO.
TITLE
12.
Expression pattern revealed by semi quantitative RTPCR of LEA genes in deep root tissues of drought resistant and susceptible checks
13.
Expression pattern revealed by semi quantitative RTPCR of LEA genes in top root tissues of drought resistant and susceptible NILs.
14.
Expression pattern revealed by semi quantitative RTPCR of LEA genes in deep root tissues of drought resistant and susceptible NILs.
15.
Expression pattern revealed by semi quantitative RTPCR of LEA genes in top root tissues of drought resistant and susceptible NILs
16.
Expression pattern revealed by semi quantitative RTPCR of LEA genes in deep root tissues of drought resistant and susceptible NILs.
BETWEEN PAGES
I. INTRODUCTION Rice is the world’s single most important food crop and a primary food for more than a one third of the world’s population. Rainfed lowland rice accounts 28% of the world’s rice growing area and it forms 18 per cent of world rice supply. More than 90% of the world’s total rainfed lowland area is in Asia. India and Bangladesh in south Asia and Indonesia, Thailand and Myanmar in south East Asia together accounts more than 80 per cent of the total area and production in Asia (IRRI, 1993). Despites its importance, rainfed lowland rice received very little attention from rice research community. Overall average yield of rainfed rice for Asia is 2.3 t/ha. In eastern India itself 13 million ha is grown to rice in the rainfed lowland condition. Even a small increase in the yield of these regions would add significantly to global rice production (Mackill et al 1996). Drought is a major abiotic stress, strongly limiting rice production system in several parts of the world. Drought resistance is genetically and physiologically complex. The plant uses different mechanisms to cope with drought stress namely, drought escape, drought tolerance, drought recovery and drought avoidance mechanisms (Blum, 1988). Achieving drought resistance in rice will be necessary for meeting the growing water shortage of the world and it will require deeper understanding of the possible mechanisms available for drought resistance. Drought stress needs more time to develop in lowlands than in uplands. But when stress occurs, it causes more loss of yield in lowland varieties than upland varieties because lowland varieties are not conditioned for such kind of situations. The drought resistant mechanisms, which are appropriate for upland system, may not be suitable for rainfed lowland and vice versa (Mackill et al 1996). Both systems even require different root phenotyping methods. Both theoretical and experimental studies illustrated that root system have role in water uptake and nutrient uptake (Andrews and Newman, 1970). Extensive reports are available, describing the role of roots in upland condition. However, information regarding
their development, distribution and role in drought resistance under lowland conditions is limited. Till today, there is no clear evidence of their contribution to grain yield under drought stress. Keeping this in mind, we have conducted a series of field and lysimetric experiments across years, using diverse genotypes of rice along with advanced breeding lines, donors, parents of mapping population and near-isogenic lines. LEA proteins play a special role in protecting cytoplasm from dehydration and storage of seeds and in whole-plant stress resistance to drought, salt, and cold. LEA proteins are expressed through all the developmental stages with different expression levels and no tissue specificity (Bo et al 2005). Gene expression analysis helps in identifying functionally important genes and pathways involved in root architecture under drought stress condition (Breyne et al. 2003). Keeping these points in view, the present investigation was taken up with the following objectives. 1. Characterization of OryzaSNP panel accessions for plant water uptake, root distribution and yield under rainfed lowland ecosystem, 2. Genetic parameters for root and shoot traits under irrigated control and stress, 3. Association between root and yield morphological traits under different moisture regimes, 4. Diversity studies using OryzaSNP panel accessions under well-watered and drought stressed conditions, 5. Screening of parents of mapping population, donors and advanced breeding lines with improved yield under drought for root and shoot characteristics under field and controlled conditions, 6. Dissection of drought tolerance mechanism using NILs of grain yield for drought stress and
7. Comparative expression of LEA genes in different zones of root and shoot under different soil water levels.
II REVIEW OF LITERATURE The literature pertaining to the present study in rice is reviewed and resented under the following headings. 1. Rainfed lowland ecosystem and drought 2. Genetic variation for root traits and their role in drought resistance 3. Role of physiological traits in drought tolerance 4. Genetic variability parameters, correlation, path coefficient and genetic divergence studies for root and shoot traits in rice 5. Genes conferring drought tolerance 6. Gene expression studies in rice 1. Rainfed Lowland Ecosystem and Drought Rice is produced in a wide range of locations under a variety of climatic conditions, occupying one-tenth of the global agricultural land. Rice is grown in four main ecosystems; irrigated lowland, rainfed lowland, deep-water and upland. As described by IRRI (IRRI 2000), in rainfed lowland, rice is transplanted or direct seeded in puddled soil, on level slightly sloping, bunded or dyked fields with variable depth and duration of flooding, depending of rainfall. Soils alternate from flooded to non flooded; yields vary depending on rainfall, cultivation practices and use of fertilizer. Most of rainfed lowlands are located in south and Southeast Asia. Modern high yielding rice varieties do not adapt well to these ecosystems, so farmers grow traditional varieties that yield about 2.3 t/ha, which is less than half of the irrigated rice. Garrity et al (1986) estimated that more than 50 per cent of rainfed lowland area could be classified as drought prone or highly drought prone. These areas may experience frequent and severe drought stress at any time during rice growth and may be subjected to uncertain rainfall distribution patterns. But in spite of those difficulties, around 20 per cent of global rice production occurs in rainfed land (Pandey et al., 2007). Rainfed lowland rice is grown on 46 M.ha out of the 132 M.ha of world rice area (MacLean et al., 2002).Yield improvements in this ecosystem not only make large
impact on world rice production but also on rural poverty in Asia as most of these areas are located in Asia (Garrity et al 1986; Khush, 1984). Drought is the most important abiotic stress limiting rice production in rainfed lowland condition (Widawsky and O’Toole 1990). Drought stress occurs when the combination of rainfall and soil water supply are inadequate to meet the demands of the crop with reference to transpiration from areal plant parts and evaporation from soil surface. Drought is highly heterogeneous in time (over seasons and years) and space (between and within locations) in most of the rainfed agricultural areas of world. The level of damage to crops due to drought stress depends on the genotypes ability to withstand drought, growth stage of the crop when it impacts, duration and intensity (which in turn depends several edaphic factors). 2. Genetic Variation for Root Traits and Their Role in Drought Resistance Roots are the principle plant organ for anchorage, nutrient and water uptake. Further, roots also influence shoot growth, water use efficiency, and overall productivity of the crop, as the roots constantly communicate with shoots and vice versa.
Studies on genetic variation for root traits in rice have been
ongoing for several decades. Only important and recent ones are reviewed hereunder. In upland conditions, Mambani and Lal (1983) reported a significant positive correlation between deep root growth and grain yield, and the authors clearly demonstrated that deep roots had a role in water uptake. O’Toole and Bland (1987) reviewed the genotypic variations in root systems and reported that plant root systems have capability of coping with the changes in environmental factors such as water status and temperature. A significant genotypic variation for root penetration ability was reported by Yu et al (1995) by using the wax layer method. Using the same method, Babu et al (2001) found that japonica accessions have a higher root penetration index (number roots penetrating the wax layer/ total number nodal and seminal roots)
than indica types, and these were used to develop double haploid mapping populations (CT9993/IR62266 and IR58821/IR52561) for mapping and tagging root traits, over the next decades. Sarkarung and Pantuwan (1999) reported a role of rooting depth and root thickness in determining drought tolerance of rice varieties under rainfed lowland conditions. In experiments simulating rainfed lowland conditions using pot plants, Wade (1999) noticed genotypic difference for water extraction and their relation with root distribution under drought stress. Genotypic differences for root mass, root length and distribution across lowland rice varieties were reported by Azhiri-Sigari et al (2000). Similarly Thanh et al (1999) and Kondo et al (2003) noticed large genotypic variation for root traits among upland varieties. The genotypic variation for root traits in different types of rice were studied by Lafitte et al (2001) and reported that Indica rice types had thin, highly branched superficial roots with narrow vessels and low root to shoot ratio, whereas japonica types had coarse roots with wider vessel, less branched long roots and a large root to shoot ratio, and aus types had intermediate root diameter, with a root distribution profile similar to that of japonica but with thin roots. Toorchi et al (2002) evaluated root morphology and related characters at three different stages and different moisture regimes using PVC cylinders. They notices significant genotypic differences for all root and shoot traits measured across different sampling. Venuprasad et al (2002), in a study involving simultaneous evaluation of root character and grain yield of under stress, and control conditions concluded that roots, that a rice plant produced prior to onset of stress, will enable a plant to tide through the stress situation and also produce better yield that a genotype that did not have the capacity to produce roots prior to the onset of stress. In a subsequent study Toorchi (2006) and Kanbar et al (2009), based on canonical correlation studies conducted under contrasting moisture regimes,
suggested that maximum root length, ration of root to shoot by weight and length, and
root dry weight (even when evaluated under well watered conditions)
conferred an advantage to grain yield under stress. 3. Role of Physiological Traits in Drought Tolerance Plant water status is estimated by several major variables such as water potential, turgor potential, and relative water content. Generally, maintenance of high relative water content has been considered to be a drought-resistance rather than drought-escape mechanism, and it is a consequence of adaptive characteristics such as osmotic adjustment and/or bulk modulus of elasticity. Leaf water potential is recognized as an index for whole plant water status (Turner, 1982) and maintenance of high leaf water potential is considered to be associated with dehydration avoidance mechanisms (Levitt, 1980). The maintenance of leaf water potential is determined by the interaction of numerous mechanisms. These include access to soil water and the pattern of soil water uptake by roots, loss to atmosphere controlled by stomatal conductance, canopy size, leaf rolling and death, and possible internal resistance to water transport. The maintenance of high leaf water potential minimized the effects of water deficit on spikelet sterility and consequently grain yield. Stomatal closure is the immediate response of plants to water stress in order to avoid the tissue dehydration during drought stress. Stomata regulate the flow of carbon dioxide and water between the dry atmosphere and the wet leaf interior. This method of water reservation helps maintain plant water potential and has been associated with reduced spikelet sterility and increased grain yield under flowering stage drought conditions in rice (Pantuwan et al., 2002). Various stomata characteristics such as low conductance, high sensitivity to leaf water status and ABA accumulation have been suggested as desirable traits in crop improvement for water-limited environment (Turner et al., 1986). However, with these traits the concurrent reduction in carbon dioxide movement,
photosynthesis and dry matter production is unavoidable, (Turner 1982) which in turn has a negative impact on yield. The sensitivity of stomata to leaf water status has been shown to have significant genetic variation in rice (Tuner et al., 1986; Dingkuhn et al., 1989; Dingkuhn et al., 1991 and Price et al., 1997). O'Toole and Chang (1979) observed that, leaf rolling under controlled conditions is related with the stomatal closure and decreases transpiration from rice leaves. In rice, cultivars with greater leaf rolling maintained higher leaf water potentials but this had no detectable effect on water transpired or dry matter produced over a ten-day period (Turner et al., 1986). Heritability of leaf rolling has been identified as moderate to high (Price et al., 2002). 4. Genetic Variability Parameters, Correlation, Path Coefficient and Genetic Divergence Studies for Root and Shoot Traits in Rice 4.1 Coefficients of Variation, Heritability and Genetic Advance There are several reports published for these genetic parameters, heritability and advance in rice. Only important and recent ones are reviewed hereunder. Armenta-Soto et al (1983) reported higher narrow sense heritability estimates for root thickness (62 per cent), root length (60 per cent), and root number (44 per cent). Similarly Mao (1984) reported that broad sense heritability for root length and thickness were high, root number and dry weight were moderately high and low respectively. Ekanayake et al., (1985a), using F1, F2 and F3 populations of cross between IR 20 (shallow, thin root system) and MGL-2 ( deep, thick root system), reported that root thickness, root dry weight and root length are polygenic traits with substantial proportions of additive variation and with narrow sense heritability’s greater than 50 per cent. They suggested that selection for these root traits based individual plant performance could be successful in early segregating generations. In another study, Ekanayake et al., (1985b) noticed low inheritance for root pulling force under lowland rice.
Chang et al., (1986) reported moderate to high heritability for maximum root length (61 per cent), root tip and root base diameters (62 per cent) by using aeroponics. They found that dominant genes controls root numbers, root depth and root mass where as root thickness is controlled by both dominant and recessive genes. Shashidhar et al (1990) reported high heritability estimates for five root traits in a study on twenty- four rice genotypes. High heritability was also reported for root length and root thickness (Das et al 1991). They also reported high environmental coefficient of variability for root length, root to shoot ratio, root dry weight, shoot dry weight and tiller number. In an F2 population, high phenotypic variability was observed for grain yield, panicle weight, number of productive tillers and total tillers. High heritability (broad sense) values were recorded for plant height and days to maturity in both direct sown rainfed (aerobic) and irrigated (anaerobic) conditions. The expected genetic advance was also high for panicle weight and productive tillers (Venkataravana 1991). Hemamalini (1997) reported moderate to high heritability to maximum root length, root number and root weight, highest expected genetic advance as per cent of mean for root volume and lowest for root diameter. She reported high and low heritability for root number and root thickness respectively. Latha (1996) reported high heritability for root dry weight (94.05 per cent), shoot dry weight (87.26 per cent), root number (86.52 per cent), root volume (80.50 per cent), number of tillers (77.63 per cent) and root length (72.35 per cent) and moderate heritability for other traits. She also reported high and low expected genetic advance as per cent of mean for rot dry weight and root thickness respectively. High heritability for root thickness and moderate heritability for root volume was reported by Price et al (1997). They also reported moderately high heritability for root length.
Vaithiyalingan and Nadarajan (2006) reported significant differences among the F2 populations for all the characters studied. Among the characters, grain yield showed high genotypic co-efficient of variation, heritability along with genetic advance as per cent of mean, followed by the characters viz., spikelet fertility per cent, productive tillers per plant and number of grains per panicle. These traits are highly amenable for selection while going for the crop improvement program of rice through inter sub-specific hybridization. High genotypic and phenotypic co-efficients of variability was reported for straw yield per plant, total biological yield per plant, number of fertile florets per panicle and number of branches per panicle (Panwar et al. 2007). The heritability estimates were highest for days to fifty per cent flowering, days to maturity and thousand grain weight. The genetic advance as per cent of mean were higher for number of branches per panicle, straw yield per plant, total biological yield per plant and grain yield per plant. Estimates of variability, heritability and genetic advance as per cent of mean were worked out in twenty four aromatic rice genotypes by Patil (2009) for yield and its attributing characters. They observed higher per cent of genetic and phenotypic co-efficient of variability for iron content, zinc content, test weight and length/ breadth ratio while, plant height, grain length, grain yield per plant, number of productive tillers per plant. Number of tillers per plant recorded moderate PCV and GCV values in the studied genotypes. The heritability values coupled with high genetic advance as per cent of mean were recorded for zinc content, iron content, test weight, length/breadth ratio, grain length, plant height and grain yield per plant. 4.2 Correlation and Path Co-efficient Analysis The literature regarding association and path co-efficient analysis in rice have been reviewed and presented in the following paragraphs. Ekanayake et al (1985a) reported that root thickness and root numbers were correlated with plant height, tiller number and shoot weight. They also found
that root length, thickness and root volume were significantly correlated with drought recovery. They observed positive association amongst root characters and reported significant correlation between plant height and root characters. Cruz et al (1986) reported strong linear relationship between total root dry mass and total root length and also between total plant dry mass and root dry mass. Shashidhar (1990) reported significant association of root weight with root length and root volume. Shahid et al (1994) reported positive correlations between root length, root dry weight, shoot dry weight, stomatal frequency and drought tolerance. Latha (1996) reported highest association between shoot dry weight and total dry weight. They also reported significant association amongst root traits. Hemamalini (1997) found positive correlation between all root characters under well watered conditions. They recorded highest correlation between root dry weight and root volume. Yadav et al (1997) also reported positive correlation among studied root traits. Thanh et al (1999) studied thirty three upland rice cultivars and observed significant correlation among all root traits except root number. Highest correlation among root characters was observed between maximum root length and total root dry weight. Plant height was also correlated with root thickness, maximum root length and total root dry weight. Venuprasad (1999) reported significant and positive association of grain yield with plant height, productive tiller number, panicle length, straw yield, total dry matter, harvest index and dry matter per day per plant at both phenotypic and genotypic levels. Gireesha (1999) found significant and positive correlation of plant height with total root number, root length, root dry weight, shoot dry weight and total dry weight. Kanbar (2002) reported significant positive correlation between plant height and root characters. Maximum root length, number of roots, root volume, root dry weight and number of tillers were observed to be interrelated. Prabuddha
(2002) reported root length showed significant positive association with root number, root volume, root dry weight, root thickness, plant height and shoot dry weight at both phenotypic and genotypic level. Significant positive association of grain yield with plant height, productive tillers per plant, dry matter per plant, leaf weight and harvest index was reported by Shashidhar et al. (2005). While, stem weight per plant, number of grains per panicle and flag leaf area showed significant positive association at genotypic level. Path co-efficient analysis revealed that, dry matter per plant had maximum positive direct effect followed by harvest index and plant height at phenotypic level. Kanbar et al (2009), based on canonical correlation studies conducted under contrasting moisture regimes, suggested that maximum root length, ration of Root to Shoot by weight and length, and root dry weight (even when evaluated under well watered conditions) conferred an advantage to grain yield under stress. 4.3 Genetic Diversity in Rice Genetic divergence was estimated by Biswas and Sasmal (1990) using Mahalanobis D2 statistic in seven rice varieties and their twenty one F1 hybrids. The twenty eight genotypes were grouped into six clusters but the grouping of parental genotypes did not follow geographical pattern. Jha et al. (1999) grouped twenty accessions of wild rice genotypes of Uttar Pradesh based on Mahalanobis D2 statistics into three clusters in which cluster I represented Oryza nivara with fourteen accessions, cluster II comprised of five accessions of Oryza sativa and cluster III with only one Oryza rifipogon accession. Das et al (2004) evaluated fifty land race collections of rice for genetic distance. The genotypes were grouped in ten clusters. Intra-cluster distance was highest in cluster IX followed by cluster I which included twelve genotypes of diverse origin. The maximum inter-cluster D2 value was recorded between clusters IV and IX. This was followed by cluster VIII and IX. The clustering pattern indicated that, the geographic diversity was not necessarily related with genetic
diversity. Days to fifty per cent flowering, grain yield per plant, grain length, kernel breadth and hundred-kernel weight were identified as potential characters that can be used as parameters while selecting diverse parents in the hybridization programme for yield and quality improvement. Assessment of genetic divergence using Mahalanobis D2 statistics was carried out on forty one high yielding and local genotypes of rice by Bhutia et al. (2005). The genotypes were grouped into six clusters. Cluster I had the highest number of genotypes (twenty seven) followed by cluster II with eight, and cluster III with three genotypes, respectively. An experiment was conducted with fifteen modern rice cultivars to estimate the contribution of different characters to the total divergence and the pair wise distance for each case to identify the right pair to be used in hybridization programme. Days to fifty per cent flowering had the greatest contribution to the total divergence, followed by thousand-grain weight and plant height. According to the distance matrix tables of Mahalanobis D2 analysis, BR-10 and BRRE - dhan 30 was the closest pair and BR-5 and BRRI-dhan 33 the most distant pair. The grain yield, tiller number per hill and filled grains per panicle were the least contributing characters towards the total divergence (Zaman et al., 2005). Patil (2009) evaluated genetic diversity in twenty four aromatic rice genotypes using D2 statistics. The varieties belonging to diverse ecological regions clustered together whereas, genotypes of the same region have entered widely into separate groups. Contribution of each character towards genetic divergence indicated that grain yield per plant contributed maximum towards the genetic divergence followed by grain length, zinc content and test weight. 5. Genes Conferring Drought Tolerance Many of the drought stress related genes have been isolated and characterized in the last two decades in a variety of crop species (Ramanjulu and Bartels 2002; Cattivelli et al., 2008). Most of the QTL studies have been undertaken in rice but in spite of such great effort on roots, no single QTL cloning has been achieved so
far in rice root. But recently in Arabidopsis one gene controlling root morphology/ architecture QTL has been identified by map- based cloning (Sergeeva et al 2006). In any study once a major QTL is identified and validated, positional cloning is the approach most commonly utilized to close the genotype-phenotype gap although alternative approaches based on candidate genes and linkage disequilibrium may represent an interesting shortcut to QTL cloning (Salvi and Tuberose 2005). Numerous transcription factors have been reported across crops and are responsible for the regulation of signal transduction and the expression of stress related genes which import stress resistance to plants. Transgenic rice with the transcription factor AtDREB1A or its orthologue OsDREB1A (DehydrationResponsive Element Binding gene) tested in pots which demonstrated improved resistance to simulated drought, high salt and low temperature stresses (Yamaguchi-Shinozaki and Shinozaki, 2004). In spite of their evident role in water uptake, so far roots have been less targeted in genetic engineering strategies to improve the performance of crops under water deficit conditions. Xiao et al (2007) have showed that over accumulation of LEA genes increases drought tolerance without yield penalty in field conditions. Recent work of Park et al (2005) demonstrated increase of root size by single gene transformation. AVP1 plays role in root development through the facilitation of auxin fluxes. Vinod et al (2006) identified candidate genes for root traits related to morphology, physiology. These were validated in the CT9993/IR62266 mapping population which was evaluated for roots traits under contrasting moisture regimes. Another work on peanut reported that the DREB1A transcription factor had a significant effect on root growth under drought stress conditions. In this case, transgenic plants showed twenty to thirty per cent higher water uptake than wild types and this water uptake was highly correlated with deep root dry weight (Vadez et al 2008).
Recently Norton et al (2008) proposed a novel approach for identification of positional candidate genes. This approach has three steps viz., initial meta analysis, then transcriptomic analysis and finally gene expression analysis. By using this approach in Bala x Azucena population they identified a reasonably good number of candidate genes with differential gene expression among genotypes. Although this approach is less accurate than fine mapping but it helps in identify positional candidate genes for small effect QTL with less cost and time. Prabuddha et al (2008) identified near isogenic lines for several root traits adopting an innovative strategy. The lines differing for the particular root trait also differed at the molecular marker loci cross validating the result. The candidate genes that were found to be associated with the particular root trait also validated with the isogenic lines. Improved performance under drought was observed in transgenic plants of the gene OsMT1a (metallothionein) that is predominantly expressed in roots and is induced by dehydration (Yang et al 2009). Another transgenic ONAC045 for transcription factor, NAC (NAM, ATAF1/2, CUC2), whose functions include a role in the development of lateral roots, was reported to have a greater survival rate in rice after drought and salt treatments (Zheng et al 2009). 6. Gene Expression Studies in Rice LEA proteins mainly play functions in dehydration tolerance and storage of seeds and in whole-plant stress resistance to drought, salt, and cold. LEA proteins are expressed through all the developmental stages with different expression levels and no tissue specificity. For instance, Em, RAB21 and dehydrins in seeds can be found in the root, stem, leaf, callus and suspension cultures of higher plants under ABA or/and NaCl induction (Federspiel, 2003). Gene expression analysis helps in identifying functionally important genes and pathways involved in root architecture under water deficit condition (Breyne et al. 2003). LEA protein gene expression in terms of time course starts from the late period of maturation and initiation period of drying reaches its peak in progressive
dehydration and sharply decreases after some hours of germination (Brands and David 2002). Many reports show that LEA protein gene expression has no tissuespecificity at the levels of tissues and organs as the gene can express in cotyledons, panicles of seeds and also in stems, leaves and roots (Federspiel, 2003). In fact, various factors and conditions and processes influence LEA protein gene expression, among which ABA is considered the most important, especially in reducing the harm caused by drought and is connected directly or indirectly with other regulatory circuits (Chaves et al. 2003; Shao et al.2005). Four steps at least are involved in LEA protein gene expression and regulation induced by drought: signal recognition, signal transduction, signal amplification and integration. Recently a few studies have been conducted in rice on tissue-specific gene expression patterns in different parts of the root system under drought stress condition. Yang et al. (2004) identified and cloned sixty six transcripts that were differentially responded in different types of roots tissue of Azucena under drought stress. Besides, they mapped four transcripts within interval containing QTL for root growth under water deficit in Azucena/IR1552 population. In another study, Wang et al 2007 noticed that the majority of genes expressed in upland rice and lowland rice are almost identical and Student’s t test showed that thirteen per cent of all the ESTs detected in leaves and seven per cent of that in roots expressed differentially in transcripts abundance between the two genotypes.
III. MATERIAL AND METHODS The particulars of the material used, methods and protocols followed and statistical tools used for analysis, in different experiments are presented under the respective experiments separately. 3.1 Experiment I Genetic Diversity and Assessment of OryzaSNP Panel Rice Accessions for Drought Tolerance Based on Water Uptake, Root Distribution and Shoot Characters under Different Moisture Regimes. 3.1.1 Field Experiment - Dry Season 2008 (DS 2008) 3.1.1.1 Plant Material The material for the study was obtained from gene bank, International Rice Research Institute (IRRI), Philippines. The genetic material used for this work was the OryzaSNP panel (McNally et al., 2009), which comprising 20 genotypes from the indica, japonica, and aus groups that have been completely mapped for SNP markers and selected as a mini-core collection representing genomic diversity of Oryza sativa (Table 1). 3.1.1.2 Experimental Site The experiment was conducted in lowland condition at the experimental farm of the IRRI, Los Baños, Philippines (14° 30’ N, 121° 15’E) during the 2008 dry season. The study included an irrigated control and a drained drought stress treatments. 3.1.1.3 Experimental Design and Crop Management Out of twenty genotypes only eighteen genotypes were evaluated due to poor germination. One seedling per hill was transplanted with 0.2m between hills and 0.25m between rows of 3m in length. The experimental design used was alpha lattice with three replications per treatment. Three rows per plot were used in both
the well-watered and drought treatments. The well-watered treatment was planted in a field neighboring the drought stress treatment (slightly far). Randomization and field layout for alpha lattice design was prepared using the CROPSTAT software. Soil was maintained saturated but without standing water for the first two weeks after transplanting to reduce risk of pest infestation. Two weeks after transplanting, both treatments were flooded to a standing water level of about 10 cm. This water level was maintained in the well-watered treatment by flood irrigation until one week before harvest. The drought treatment was drained at thirty days after transplanting, and no further irrigation was applied, except at seventy two days after sowing when drought plots were flooded to facilitate root sampling.
The fields were hand weeded two-three times.
Basal fertilizer
applications equivalent to 40 kg P and 40 kg K ha-1 were applied in the form of single super phosphate and potassium chloride, and 120 kg ha-1 of N in the form of ammonium sulphate was applied in three even splits around 21, 42 and 61 DAS. 3.1.1.4 Environmental Characterization Solar radiation, rainfall, and pan evaporation were recorded at agrometeorological stations operated by the IRRI Climate Unit located nearby (within 1.5km) of the experimental fields (Table 2 and Figure 3). The soil of the IRRI lowland soil was Aquandic epiaquall. General soil characteristics are presented in Table 3. Mercury manometer-type tensiometers (15 and 30cm soil depth) and poly vinyl chloride (PVC) tubes (4cm diameter, 1m long) for frequency domain reflectometry readings were installed at three locations within drought treatment field, as soon as the soil had solidified after draining Soil water potential based on height of the mercury column, and volumetric soil moisture at 10cm increments to a depth of 70cm (Diviner 2000, Sentek Sensor Technologies, Stepney SA, Australia) were monitored up-to three times weekly thereafter. Diviner readings were converted to volumetric soil moisture content (Ѳv) according to a calibration curve of Ѳv = 0.6007x + 35.66 (Figure 1 and 2).
3.1.1.5 Method of Root Sampling and Root Measurements Soil samples for root measurements were acquired with a 20cm x 20cm monolith sampler in all plots of the drought and well-watered treatment to a depth of 45cm at seventy two days after sowing. Before root sampling, the shoot parameters like, tiller number and plant height was measured. The monolith sampler was made of iron sheet with 20 X 20 X 50cm internal dimensions and these were fabricated at the IRRI. To ease the method of sampling, one day before root sampling, all the water was removed in well-watered treatment plots whereas, in stress treatment plots shallow irrigation was given. A rice plant observed to be healthy and representative of respective genotype was sampled in each replication. Root sampler was centered over the rice hill and then pushed into soil to a 45cm depth by hammering on a block of wood placed on top of the sampler. The surrounding soil was removed to dig out the sampler and then soil was sectioned with sharp cutting blade into 0-10, 10-20, 20-30 and 30-45cm depths. Roots were either washed from the soil immediately after sampling or stored at -4ºC until washing (within three weeks). Roots were washed by placing each soil sample on a 1mm screen and running water over the sample to remove the soil. All samples were stored in 50 per cent ethanol until scanning. Root samples were scanned at 400 dpi (Epson V700, California, USA). Scanned images were analyzed for architectural attributes using WinRhizo v. 2007d (Régent Instruments, Québec, Canada). A pixel threshold value of 150-175 was set for the analysis. 3.1.1.6 Observations Recorded 3.1.1.6.1 Physiological Traits 3.1.1.6.1.1 Leaf Water Potential (LWP, MPa) LWP (MPa) was measured at midday using pressure chamber. The uppermost fully expanded leaf (flag leaf) was cut approximately 2.0 cm below the leaf collar and placed in pressure chamber with the cut portion just protruding
through the seal on the top of the chamber into the atmosphere. Pressure was applied slowly to the leaf blade until a drop of water appears from the cut surface. The equivalent pressure was recorded from the gauge and this gave the approximate leaf water potential. The LWP measurements were taken twenty days after stress imposition. 3.1.1.6.1.2 Stomatal Conductance (SC, mol m-2 s-1) SC (mol m-2 s-1) was measured twenty days after stress imposition using a Li-1600 steady-state porometer (Li-Cor Biosciences, USA). Porometer readings were conducted mid-morning on the youngest fully expanded leaf blade of a main culm or primary tiller. 3.1.1.6.1.4 Photosynthesis Rate (PR, μmol mol-2 s-1) PR (μmol mol-2 s-1) was measured twenty days after stress imposition using a Li-1600 steady-state porometer (Li-Cor Biosciences, USA). Porometer readings were conducted mid-morning on the youngest fully expanded leaf blade of a main culm or primary tiller. 3.1.1.6.1.5 Transpiration Rate (TR, mmol mol-2 s-1) TR (mmol mol-2 s-1) was measured twenty days after stress imposition using a Li-1600 steady-state porometer (Li-Cor Biosciences, USA). Porometer readings were conducted mid-morning on the youngest fully expanded leaf blade of a main culm or primary tiller. 3.1.1.6.1.6 Relative Water Content (RWC, %) RWC was measured twice during the stress period. Leaves were sampled midday after the dew had dried. One uppermost fully expanded leaf per plot was sampled and placed in pre-weighed centrifuge tubes. Samples were stored on ice and weighed immediately upon return to the lab for the fresh weight. Tubes were then filled and stored overnight in the dark at 4°C. The next morning, leaves were blotted dry with paper towels using the standard procedure that required about 30s
per sample, and were weighed immediately. After recording fully turgid weight, leaves were dried at 70 °C to constant weight. 3.1.1.6.1.7 Leaf Rolling Score (LRS) One week after the start of drought, leaf rolling score were recorded based on standard evaluation system for rice. Leaf rolling scores ranged from 1 (no leaf rolling) to 9 (leaves completely rolled). Leaf rolling score measurements were taken between 11 am to 12 am at twenty days after stress imposition. 3.1.1.6.2 Root Traits 3.1.1.6.2.1 Root Number Number of all developed roots present at crown region was counted for root number. 3.1.1.6.2.2 Root Length (cm) Root length (cm) at different soil depths was measured using WinRhizo software. 3.1.1.6.2.3 Root Volume (RV, cm3) Root volume at different soil depths was measured using WinRhizo software. 3.1.1.6.2.4 Root Surface Area (RSA, cm2) Root surface area at different soil depths was measured using WinRhizo software. 3.1.1.6.2.5 Root Length Density (RLD, cm/cm3) Root length density (cm/cm3) was calculated by dividing the root length by the soil volume at each soil section.
3.1.1.6.2.6 Root Dry Weight (RDW, g) Root dry weight per plant (g) at different soil depths were recorded after oven drying for 72 hours 3.1.1.6.2.7 Root: Shoot Ratio (RSR) Root to shoot ratio was calculated by dividing the total root dry weight (g) with shoot dry weight (g). 3.1.1.6.2.8 Drought Induced Root Growth at Depth Drought induced root growth at 30-45cm was calculated by the difference of total root length (cm) in drought and well-watered conditions at depth. 3.1.1.6.3 Agronomic Traits 3.1.1.6.3.1 Shoot Dry Weight (SDW, g/plant) Shoot dry weight per plant was recorded after oven drying for 72 hours for the same plant which was used for root sampling 3.1.1.6.3.2 Plant Height (PH, cm) Plant height was measured in three plants randomly in each plot at maturity. The height of the plant was measured from base of main tiller to the tip of the panicle at the time of harvest and expressed in centimeter (cm). 3.1.1.6.3.3 Tiller Number (TN) Total number of tillers (both productive and non productive) was counted per plant at the time of harvest. 3.1.1.6.3.4 Straw Biomass (g/m2) Five plants from each replication were harvested from the middle 2 meters of all three rows, leaving 0.5m of the row at each ends. Harvested plants were oven dried for three days and used for straw biomass measurements (g/m2).
3.1.1.6.3.5 Grain Yield (g/m2) Ten plants from each replication were harvested from the middle two meters of all three rows, leaving 0.5m of the row at each ends. Harvested plants were threshed and air dried. Total grain weight for ten plants was measured in grams and converted to g/m2 at fourteen percent moisture content. 3.1.1.6.3.8 Harvest Index (HI) The ratio of the grain yield and the total dry matter of the plant computed. Grain yield (g/m2) Harvest index =
x 100 Total dry matter weight (g/m2)
3.1.2 Field Experiment – Dry Season 2009 (DS 2009) 3.1.2.1 Experimental Site The experiment was conducted at the Directorate of Rice Research field, International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Patancheru, India during the DS 2009. ICRISAT is located at 17°30′N latitude, 78°16′E longitude and at an elevation of about 549m above mean sea level. 3.1.2.2 Experimental Design and Crop Management Due to poor germination, only 13 genotypes were evaluated. One seedling per hill was transplanted with 0.2m between hills and 0.25m between rows of 4m in length. The experimental design used was alpha lattice with four replications in stress treatment and three replications in well-watered treatment. Three rows per plot were used in both the well-watered and drought treatments. The well-watered treatment was planted in a field neighboring the drought stress treatment. The drought treatment was drained at forty one days after transplanting, and further irrigation was applied after noticing sever rolling. Rests of crop management practices were same as in 3.1.1.3.
3.1.2.3 Environmental Characterization Soil moisture content was monitored by using Time Domain Reflectometry (TDR) (Tektronix Inc., Wilsonville, Oregon, USA) which gives the soil moisture in percent volume at different depths (10 to 40cm) (Figure 4). TDR readings were taken once in three days during drought stress in stress treatment.
Rainfall,
temperature, solar radiation and relative humidity of farm, during the stress period were obtained from ICRISAT website (Table 4 and Figure 5). The temperature recorded was also quite high (average maximum temperature was 39.57oc). Both drought and heat stress affected crop drastically. Because of this we were able to record only biomass in stress treatment. But in well-watered treatment, both grain yield and biomass was recorded. 3.1.2.4 Observations Recorded Ten plants were selected at random from each replication for recording plant height, number of tiller per plant, grain yield, biomass and harvest index observations. The averages of ten plants were considered for analysis. 3.1.3 Lysimetric Experiment- Wet Season 2008 (WS 2008) 3.1.3.1 Experimental Site The experiment was conducted in greenhouse (BG02) at IRRI, Los Baños, Philippines, during wet season 2008. 3.1.3.2 Experimental Design All genotypes were grown under two environments (well-watered and drought stress). The experimental design used was split-plot with five replications per treatment. Basal fertilizer applications equivalent to 40 kg P and 40 kg K ha_1 were applied in the form of single super phosphate and potassium chloride, and 80 kg ha_1 of N in the form of ammonium sulphate was applied in two even splits around 21 and 61 DAS. Irrigation was applied once in two days.
3.1.3.3 Method of Simulation of Lowland Condition. The experiment was conducted in PVC cylinder with a diameter of 19cm and height of 105cm. The cylinders (attached with plastic membrane inside) were filled with 29kg of sieved air dried upland sandy loam soil (bulk density of 1.28g of dry soil cm-3) to a level up to 80cm from the bottom of cylinder. The soil was shaken and compacted with circular metal plate (diameter 17cm) to bring soil to pre marked level in pipes i.e. 80cm. To achieve a uniform bulk density throughout the soil column, the compaction was carried out in increments of 5cm. The top 20cm of cylinder was filled with puddled lowland soil. The upland and lowland soil used in this experiment were collected from the top layer of upland and lowland fields of IRRI. Cylinders were painted with white to avoid the soil temperature and were placed in cement pits under greenhouse to mimic field growth condition. Two seedlings were transplanted per cylinder but after one week, one plant having uniform growth with those in the other cylinders was maintained. This method was intended to simulate lowland field conditions as closely as possible. Drought stress was imposed at thirty days after transplanting by opening the drainage at bottom. From then onwards no water was added. To prevent evaporation, each cylinder of drought stress was covered with two layer of plastic sheet. 3.1.3.4 Method of Root Sampling and Root Measurements For root sampling, plastic bags were pulled out of the cylinders and then soil columns were separated into four sections (0-30, 30-45, 45-60 and 60-100cm). Roots were separated from soil by gently spraying tap water on 1-mm screen until all soil washed through the screen. After cleaning, roots were stored in an alcohol solution (50 per cent isopropanol) in a refrigerator at 40C for later analyses. Measurements of roots were performed as mentioned in 3.1.1.5 except a pixel threshold value of 190 was set for the analysis.
3.1.3.5 Observations Recorded 3.1.3.5.1 Physiological Traits 3.1.3.5.1.1 SC (mol m-2 s-1) SC (mol m-2 s-1) was measured twice during stress using a Li-1600 steadystate porometer (Li-Cor Biosciences, USA). Porometer readings were conducted mid-morning on the youngest fully expanded leaf blade of a main culm or primary tiller. 3.1.3.5.1.2 Water Uptake (g/plant) Water uptake (g/plant) was calculated as the difference of the initial cylinder weight and the cylinder weight once in every week (7, 14, 21 and 28 days after stress) 3.1.3.5.2 Root Traits Root traits measured in previous experiment (3.1.1.5) were also measured in this experiment along with below mentioned trait. 3.1.3.5.2.1 Maximum Root Length (MRL, cm) The lowest visible root at the bottom, after removing the plastic bag 3.1.3.5.3 Agronomic Traits 3.1.3.5.3.1 PH (cm) One day before root sampling, plant height was measured from base of stem to top most leaves. 3.1.3.5.3.2 TN Total number of tillers (both productive and non productive) was counted per plant one day before sampling. 3.1.3.5.3.3 SDW (g/plant)
Plants were cut at crown level and shoot dry weight (g) were recorded after oven drying for 72 hours. 3.1.4 Lysimetric Experiment- DS 2009 3.1.4.1 Experimental Site The experiment was conducted under rainout shelter at ICRISAT, Patancheru, India during the dry season 2009. 3.1.4.2 Experimental Design The experimental design of this experiment was same as that of previous lysimetric experiment (3.1.3.2). 3.1.4.3 Experimental Method The light- gray PVC cylinder was 20cm diameter and 120cm in length with a PVC plate at the bottom. Initially bottom of cylinders were sealed with silicone adhesive.
The vertisol was collected from fields of ICRISAT. The
collected soil sample was air dried in sunlight, and big clods were broken and sieved with 2mm sieve to get uniform fine soil particles. Cylinders were filled with 44 kg of vertisol leaving the top 5cm empty. The soil was compacted with circular metal plate (diameter 19cm) to get bulk density of 1.22 g of dry soil cm-3. To achieve a uniform bulk density throughout the soil column, the compaction was carried out by evenly compacting the soil every 5kg while filling the cylinders. Cylinders were placed in cement pits to mimic field growth condition. Twenty-nine days old four seedlings were transplanted in each of cylinder. Plants were thinned to two individuals per cylinder at seven days after transplanting. All the plants were fully irrigated by watering every other day until the drought stress treatment. Drought stress was imposed after forty one days of transplanting by opening the silicone adhesive present all along the border of PVC plate at the bottom of cylinder. To prevent evaporation each cylinder of drought stress were covered 400g of plastic beads (low-density plastic beads from Reliance
Ltd). At the end of stress period i.e., thirty days after stress, plants were used for root phenotyping. 3.1.4.4 Method of Root Sampling and Root Measurements Root washing and root trait measurements were carried out in the same manner as in above Lysimetric experiment. Unlike in above cylinder trial, here entire root system was collected. Root measurements are same like previous experiment (3.1.1.5) except a pixel threshold value of 190 was set for the analysis. 3.1.4.5 Observations Recorded 3.1.4.5.1 Physiological Traits 3.1.4.5.1.1 Water Uptake (g/plant) Water uptake (kg/plant) was measured in same way like previous experiment but measured once in every three days in this experiment (5, 8, 12, 15, 18 and 23). 3.1.4.5.2 Root Traits Root traits measured in previous experiment (3.1.1.5) were also measured in this experiment. 3.1.4.5.3 Agronomic Traits Agronomic traits measured in previous experiment (3.1.3.5.3) were also measured in this experiment. 3.1.5 Lysimetric Experiment- Wet Season 2009 (WS 2009) 3.1.5.1 Experimental Site The experimental site of this experiment was same as that of previous lysimetric experiment (3.1.3.1). 3.1.5.2 Experimental Design
The experimental design of this experiment was same as that of previous lysimetric experiment (3.1.3.2). 3.1.5.3 Method of Simulation of Lowland Condition. The method was same as that of previous lysimetric experiment (3.1.3.3). 3.1.5.4 Observations Recorded 3.1.5.4.1 Physiological Traits 3.1.5.4.1.1 Water Uptake (g/plant) Water uptake (g/plant) was calculated as the difference of the initial cylinder weight and the cylinder weight once in every week (7, 14, 21, 25, 28, 33 and 35 days after stress imposition) 3.2 Experiment II Variation in Water Uptake, Root and Shoot Characters and Their Association with Drought Tolerance in Parents of Mapping Population, Donors and Advanced Breeding Lines of IRRI-India Drought Breeding Network. 3.2.1 Field Experiment - Dry Season 2009 3.2.1.1 Plant Material The IRRI-India drought breeding network has identified a large number of promising drought resistant breeding lines developed using diverse donors at IRRI and eight national institutes in India. Breeding lines with consistent superior performance over popular checks under severe and moderate stress and with no yield penalty under irrigated conditions, along with donors and popular checks, were selected for this experiment. This advanced breeding line’s along with donors was obtained from Dr. O.N. Singh, Director, Central Rice Research Institute, Cuttack, India. Whereas parents of mapping populations were collected from gene bank, IRRI, Philippines (Table 5).
3.2.1.2 Experimental Site The experimental site of this experiment was same as in previous experiment (3.1.2.1) 3.2.1.3 Experimental Design and Crop Management The experimental design and crop management was same as in previous experiment (3.1.2.2) 3.2.1.4 Environmental Characterization The environmental characterization of this experiment was same as in previous experiment (3.1.2.3) 3.2.1.5 Observations Recorded The observations recorded this experiment was same as in previous experiment (3.1.2.4) 3.2.2 Lysimetric Experiment- Dry Season 2009 3.2.2.1 Experimental Site The experimental site of this experiment was same as in previous experiment (3.1.4.1) 3.2.2.2 Experimental Design Forty nine advanced breeding lines, donors and parents of mapping population were grown under two environments (well-watered and drought stress). The experimental design of this experiment is same as in previous experiment (3.1.4.2). 3.2.2.3 Experimental Method The experimental method of this experiment was same as in previous experiment (3.1.4.3) 3.2.2.4 Method of Root Sampling and Root Measurements
The method of root sampling and root measurements of this experiment is same as in previous experiment (3.1.4.4) 3.2.2.5 Observations Recorded All physiological, root and agronomic traits measured in previous experiment (3.1.4.5) were also measured in this experiment. 3.3 Experiment III Physiological and Molecular Mechanism of Drought Avoidance Root Mechanisms in Adeysel NILs. 3.3.1 Studies on Water Uptake, Root Distribution under Different Water Levels 3.3.1.1 Plant Materials Two pairs of NILs [IR77298-14-1-2-10 (+), IR77298-14-1-2-13 (-), IR77298-5-6-18 (+) and IR77298-5-6-11(-)] along with their parents [IR64 and Adeysel] were used in our experiment. The material was obtained from Dr Arvind Kumar, IRRI, Philippines (Table 6). This experiment was conducted using lysimeters in wet season 2008 under greenhouse condition. Experimental site, experimental design, experimental method, method of root sampling and root measurements, and observations recorded are same as mentioned in experiment 3.1.3. 3.3.2 Comparative Expression of Four LEA Genes under Different Water Levels 3.3.2.1 Experimental Site The experiment was conducted in greenhouse (BG11) at IRRI, Los Baños, Philippines, during wet season 2009. 3.3.2.2 Experimental Design
Two pairs of NILs [IR77298-14-1-2-10 (+), IR77298-14-1-2-13 (-), IR77298-5-6-18 (+) and IR77298-5-6-11(-)] plus its recurrent parent (IR64) and drought resistance genotype (Dular) were grown under two environments (drought stress and well-watered). The experimental design used was randomized block design with eight replications per treatment. Four to five seeds of each entry were hand dibbled in the PVC pot filled with a mixture of sand and soil in 1:2 proportions. After germination, one healthy seedling was allowed to grow in each pot. Basal fertilizer applications equivalent to 40 kg ha_1 of N, 40 kg P and 40 kg K ha_1 were applied in the form of of ammonium sulphate, single super phosphate and potassium chloride respectively 3.3.2.3 Experimental Method and Drought Treatment Two kilograms of ground, air-dried soil mixture (2 soil: 1 sand) was placed in a PVC pot. Each pot was adequately fertilized and grown under greenhouse condition. All pots were watered to soil saturation. Ten to twelve seeds were sowed per pot and later thinned to 1 plant per pot after one week of sowing. All pots were irrigated twice daily to maintain the soil at saturation. Stress was imposed by initiating soil dry down protocol starting 25 days after sowing. The day before the start of progressive soil drying, all pots were fully watered (saturation). All pots were allowed to drain overnight by loosening the stoppers and were weighed early next morning to get the saturated weight. Then the pots were enclosed in white plastic bags, around the stem of the plant to prevent direct soil-evaporation. Thereafter, the pots were weighed every morning around 7.30 AM to know the soil moisture status. Daily transpiration (T) was computed as the difference in weight on successive days and fraction of transpirable soil water (FTSW) based on the proportion of soil water remaining in the pot. Well-watered control plants of each genotype are maintained at initial target weight by adding the daily water loss back to the pot. Feeder pipe will be inserted in all 1.0 FTSW treatment pots for
watering. The water stress will be continued, until each pot reaches the respective 0.2 FTSW. Lower limit for the same soil mixture was got from other experiment and it is 13 percent. Lower limit is the soil weight at 0 FTSW or when transpiration is minimal. Calculation of FTSW goes like this (below). TTSW (total transpirable soil water) = saturated soil weight- (total soil weight x 0.13 + dry soil weight) Pot weight at targeted 0.2FTSW = Pot weight at lower limit + TTSW x 0.2 3.3.2.4 Method of Root Sampling for RNA Isolation Five replications were used for root sampling. After removing the roots from pots, each root samples was washed quickly in a series of trays filling with clean tap water. The excess of moisture was removed by wiping with dry paper towel. After washing and cleaning all samples were stored in liquid nitrogen and later stored at -80oC for RNA extraction. -From each plant, 4 different tissue samples were collected. Out of 4 samples, 2 samples were collected from leaves (0-20cm, >20cm) and rest 2 samples were from roots (0-15cm, >15cm). All these process of sampling was done quickly to avoid degradation of RNA (less than one minute for each plant). 3.3.2.5 RNA Extraction and Gene Expression Analysis RNA is sensitive to ubiquitously present RNase in the surrounding environment so extreme care was taken to avoid RNase contamination. Total RNA was extracted from rice tissues using the TRIzol protocol, according to the instructions of the manufacturer (Invitrogen, UK). Tissues were ground using a chilled mortar and pestle. 100 mg of the homogenized tissue was transferred in 1 ml of TRIZOL Reagent (Invitrogen, Life technologies, Carlsbad, CA, USA) and the mixture incubated at room temperature for 10 min to allow complete dissociation of the nucleoprotein complexes. After the addition of 0.2 ml of chloroform, the tubes were shaken vigorously by hand for 15 sec and eventually
left for incubation for ~3 min at room temperature. Following the centrifugation at 12,000x g for 15 min at 4ºC, the aqueous-upper phase was transferred into a fresh tube and the RNA was precipitated using 0.5 ml of isopropyl alcohol. The mixture was centrifuged at 12,000x g for 10 min at 4ºC and the pellet formed was washed once with 75 per cent ethanol. Additional centrifugation at 7,500x g for 5 min at 4ºC lead to further formation of a pellet that was left to dry at room temperature for 10 min. Eventually the pellet was re-suspended in 20-50 µL of DEPC-water and the RNA concentrations were determined by nanodrop. RNA quality was assessed by fractionation on 1.5 per cent (w/v) agarose gel [1x Tris-Borate-EDTA (TBE)]. RNA was visualized by SyBR safe (5 µl/50ml) staining under UV illumination. The extracted RNA was treated with DNase I, Amplification Grade (Deoxyribonuclease I; Invitrogen, UK or Promega Corporation, Madison, WI, USA), for digestion of the DNA. The first strand of cDNA was synthesized as described in the SUPERSCRIPT II RNase H-Reverse Transcriptase kit (Invitrogen, UK), using 6 µg of total RNA and the oligo(dT) primer. RT-PCR was performed in a 50 µl reaction volume according to the instructions of the manufacturer (Invitrogen) using recombinant U-Taq DNA polymerase (SBS Genetech). The gene specific primers listed in Table 7 were used for amplification of LEA genes. As positive control the product of 18S ribosomal RNA was used (F: 5’– AAA CGG CTA CCA CAT CCA AG –3’ and R: 5’– TCA TTA CTC CGA TCC CGA AG –3’). The cycle conditions were as follows: preamplification denaturation at 94ºC for 3 min, 32 cycles of denaturation at 94ºC for 30sec, primer annealing at 57 ºC for 45 sec, and primer extension at 72ºC for 30 sec, and a final extension of RT-PCR products at 72ºC for 7 min. RT-PCR products were fractionated on 1.2
per cent agarose gels (TAE), and were
visualized by SYBR® Safe DNA Gel Stain (Invitrogen) staining under UV illumination.
3.4. Statistical Analysis 3.4.1 Analysis of Variance The analysis of variance for different characters was carried out using mean data in order to assess the genetic variability among genotypes as given by Cochran and Cox (1957). The level of significance was tested at 5 per cent and 1 per cent using F test. The model of ANOVA used is presented below.
Sl. No.
Source of variation
DF
M.S.S.
Expected M.S.S.
1
Replication
r-1
Mr
gσ2r + σ2e
2
Genotypes
g-1
Mt
σ2e + σ2g
3
Error
(r-1)(g-1)
Me
σ2e
Total
rg-1
Mr+Mt+Me
Where, r = number of replication g = number of treatments (genotypes) 3.4.2 Estimation of Genetic Variability Parameters 3.4.2.1 Phenotypic and Genotypic Variance Phenotypic and genotypic components of variance were estimated by using the formula given by Cochran and Cox (1957). MSS due to genotypes – MSS due to error (σ 2e) Genotypic variance (σ g2) = --------------------------------------------------------------r Phenotypic variance (σ p2) = Genotypic variance (σ 2g) + Environmental variance (σ 2e)
3.4.2.2 Co-efficient of Variability Both phenotypic and genotypic co-efficient of variability for all the characters were estimated using the formulae of Burton and De Vane (1953). Genotypic Co-efficient of Variability (GCV): Genotypic variance GCV per cent = ----------------------------------- x 100 Grand mean Phenotypic Co- efficient of Variability (PCV): Phenotypic variance PCV per cent = ----------------------------------- x 100 Grand mean PCV and GCV were classified as per Sivasubramanian and Menon (1973) and as shown below: 0-10 per cent - Low;
10 -20 per cent - Moderate; >20 per cent - High
3.4.2.3 Heritability in Broad Sense (h2) Heritability (broad sense) was estimated for all the characters as the ratio of genotypic variance to the total variance as suggested Lush, (1949) and Hanson et al. (1956). σ g2 h = --------σ p2 2
According to Robinson (1966) heritability estimates in cultivated plants can be placed in following categorizes. 5-10 per cent- Low;
10-30 per cent- Moderate;
30-60 per cent - High
3.4.2.4 Genetic Advance (GA) Genetic advance for each character was estimated by using the following formula of Johnson et al. (1955) GA = h2 x K x σ p Where, h2 = heritability estimate K = selection differential which is equal to 2.06 at 5 per cent intensity of selection. σ p = phenotypic standard deviation Further the genetic advance as per cent of mean was computed by using the following formula GA GA as per cent of mean = ---------------- x 100 Grand mean Genetic advance as per cent of mean was categorized according to Johnson et al. (1955), as given below. 0 -10 per cent = Low; 10 - 20 per cent = Moderate;
>20 per cent = High
3.4.3 Correlation Studies To determine the degree of association of characters with yield and also among the yield components, the correlation co-efficients were calculated. Phenotypic co-efficients of correlation between two characters were determined by using variance and covariance components as suggested by Al-Jibourie et al. (1958). Covp (xy) rp (xy)= ------------------------
σ2p (x) . σ2p (y) Where, rp (xy) is the phenotypic correlation co-efficient Covp is phenotypic co-variances of xy. σ2p phenotypic variance of x and y, The calculated value of ‘r’ was compared with table ‘r’ value with n-2 degree of freedom at 5 per cent and 1 per cent level of significance, where n refers to number of pairs of observations. 3.4.4 Path Co-efficient Analysis Path-co-efficient analysis was carried out using phenotypic correlation values of yield components on yield as suggested by Wright (1921) and illustrated by Dewey and Lu (1959). Standard path co-efficients which are the standardized partial regression co-efficients were obtained using statistical software package SPAR 1. These values were obtained by solving the following set of ‘P’ simultaneous equations by using the above package. P01 + P02 r12 + -----------+ P0P r1P = r 01 P01 + P12 r02 + -----------+ P0P r2P = r 02
P01 + r1P+ P02 r2 P -----------+ P0P = r 0P Where P01, P02, ---------------------------P0P are the direct effects of variables 1,2, -------------p on the dependent variable 0 and r12, r13 ,-------r1P------ rP(P-1 ) are the possible correlation co-efficients between various independent variables and r01, r02, r03 ------- r0P are the correlations between dependent and independent variables. The indirect effect of the ith variable via jth variable is attained as (Poj x rij). The contribution of remaining unknown factor is measured as the residual factor, which is calculated and given below.
P2ox = 1- [P2 01 + 2P01 P02 r12 + 2 P01 P03 r13 +----------+ P202 + 2P02 P03 r13 +……..+P20P] Residual factor = (P2ox)½. 3.4.5 Genetic Divergence Analysis The genetic divergence between genotypes was estimated using Mahalanobis’s D2 statistic (1936). The distance D from the sample was computed using the formula. D2p = d1 S-1d Where, D2p = Square of distance considering ‘p’ variables d =Vector observed differences of the mean values of all the characters (xi1- xi2) S-1 = inverse of variance and covariance matrix 3.4.5.1 Clustering of the D2 Values All the genotypes used were clustered into different groups following Tocher’s method (Rao, 1952). The intra and inter-distance were also computed the criterion used in clustering to the same cluster should atleast on the average, show a smaller D2 values than those belonging to different clusters. The device suggested by Tocher (Rao, 1952) was started with two closely associated populations and find a third population which had the smallest average of D2 from the first two. Similarly, the fourth was chosen to have a smallest average D2 value from the first three and so on. The permissible increase in D2 value shown by a population to the nearest population. If at any stage increase in average D2 value exceeded the average of already included, because of the addition of a new genotypes, than that genotypes was deleted. The genotypes that are included already in that group were considered as the first cluster. This procedure was repeated till D2 values of the other genotypes were exhausted
omitting those that were already included in the former cluster and grouping them into different cluster. 3.4.5.2 Intra-Cluster Distance The average intra-cluster distances were calculated by the formula given by Singh and Chaudhary (1977). Square of intra-cluster distance = ΣDi2 / n Where, ΣDi2 = sum of distance between all possible combinations. n = number of all possible combinations 3.4.5.3 Inter-Cluster Distance The average inter-cluster distance was calculated by the formulae described by Singh and Chaudhary (1977). Square of inter-cluster distance = ΣDi2 / ni nj Where, ΣDi2 = sum of distances between all possible combinations (ninj) of the entries included in the cluster study. ni = number of entries in cluster i nj = number of entries in cluster j 3.4.5.4 Contribution of Individual Characters towards Genetic Divergence The character contribution towards genetic divergence was computed using method given by Singh and Chaudhary (1977). In all the combination, each character is ranked on the basis of di = yij – yik values. Where, di = mean deviation
yij = mean value of the jth genotype for the ith character and yik = mean value of the kth genotype for the ith character. Rank ‘I’ is given to the highest mean difference and rank p is given to the lowest mean difference Where, P is the total number of characters. Finally, another table giving information on, number of times that each character appeared in the first rank is prepared and per cent contribution of characters towards divergence was calculated.
IV. EXPERIMENTAL RESULTS The results obtained from the present investigation on are presented under the following subheadings. 4.1 Genetic Diversity and Assessment of OryzaSNP Panel Rice Accessions for Drought Tolerance Based on Water Uptake, Root Distribution, Shoot and Yield Parameters under Different Moisture Regimes. 4.1.1 Field Experiments – Dry Season 2008 and 2009 4.1.1.1 Mean Performance The mean performance of all OryzaSNP panel rice accessions in respect of phenology, physiological, root, shoot and yield traits are briefly presented below. 4.1.1.1.1 Field Experiment – Dry Season 2008 4.1.1.1.1.1 Root Traits The mean performance of all OryzaSNP panel rice accessions in respect of root traits (RN, RSR, RLD, RSA, RV and RDW) across different soil depths under both drought stress and well-watered condition are presented below (Table 8-11). 4.1.1.1.1.1.1 RN Comparison of means (Table 8) showed that, Minghui 63 had highest RN (259.7) followed by SHZ 2 (243.5) and Dular (231.0) while, Azucena had the lowest RN (113.0) followed by N22 (115.67) and Cypress (120.5) under drought stress condition. Whereas, FR13A had highest RN (412.0) followed Sadu Cho (400.3) and Dom Sufid (357.5) while, LTH had the lowest RN (60.5) followed by N22 (132.5) and Moroberekan (153.0) under well-watered condition. 4.1.1.1.1.1.2 RSR
Under drought stress condition, Moroberekan and Swarna had highest RSR (0.15) followed by Minghui 63 and SHZ 2 (0.11). On the other hand, Pokkali recorded lowest RSR (0.05) followed by Sadu Cho and LTH (0.06). Whereas, under well-watered condition FR13A had highest RSR (0.34) followed by Azucena (0.25) and Tainung 67 (0.18). On the other hand, LTH and Pokkali recorded lowest RSR (0.05) followed by Zhenshan 97B (0.06) (Table 8). 4.1.1.1.1.1.3 RLD (cm/cm3) Means of RLD across different soil depths under both drought stress and well-watered are presented in Table 8. The highest RLD at 0-10cm soil layer under drought stress condition was recorded by IR64 (1.89 cm/cm3) followed by Dom Sufid (1.73 cm/cm3) and Aswina (1.51 cm/cm3). On the other hand, Azucena recorded lowest RLD (0.49 cm/cm3) followed by Tainung 67 (0.85 cm/cm3) and Cypress (0.87 cm/cm3). Under well-watered condition highest RLD
were
recorded by Dom Sufid (3.95 cm/cm3) followed by Sadu Cho (3.84 cm/cm3) and Swarna (3.64 cm/cm3). Whereas LTH was recorded lowest RLD (1.21 cm/cm3) followed by M 202 (1.23 cm/cm3) and Moroberekan (1.63 cm/cm3). Under drought stress condition, the highest RLD at 10-20cm soil layer was recorded by Swarna (1.45 cm/cm3) followed by SHZ 2 (1.02 cm/cm3) and Dom Sufid (0.99 cm/cm3). On the other hand, N22 recorded lowest RLD (0.16 cm/cm3) followed by Tainung 67 and LTH (0.29 cm/cm3). However in well-watered condition, the highest RLD at 10-20 cm soil layers was recorded by Minghui 63 (1.25 cm/cm3) followed Azucena (1.19 cm/cm3) and Swarna (0.71 cm/cm3). Whereas Tainung 67 was recorded lowest RLD (0.21 cm/cm3) followed by M 202 (0.22 cm/cm3) and IR64 (0.24 cm/cm3). The highest RLD at 20-30 cm soil layer was recorded by Swarna (1.21 cm/cm3) followed by M 202 (0.87 cm/cm3) and Azucena (0.67 cm/cm3) under drought stress condition. On the other hand, Cypress recorded lowest RLD (0.11 cm/cm3) followed by Zhenshan 97B (0.15 cm/cm3), LTH and Tainung 67 (0.17
cm/cm3). Under well-watered condition, the highest RLD at 20-30 cm soil layer was recorded by Minghui 63 (0.87 cm/cm3) followed SHZ 2 (0.35 cm/cm3) and Azucena (0.29 cm/cm3). Whereas M 202 was recorded lowest RLD (0.06 cm/cm3) followed by Tainung 67 (0.07 cm/cm3) and FR 13A (0.08 cm/cm3). Large variation was recorded for RLD at depth only under drought stress condition. At 30-45cm soil layer, highest RLD was recorded by Dular (0.24 cm/cm3) followed by SHZ 2 (0.20 cm/cm3) and Dom Sufid (0.17 cm/cm3). On the other hand, IR64 recorded lowest RLD (0.02 cm/cm3) followed by M 202 and Zhenshan 97B (0.03 cm/cm3). Under well-watered condition, all genotypes recorded lowest RLD at 30-45cm soil layer. Among them, Dular and Minghui 63 recorded highest RLD (0.06 cm/cm3) followed N22 (0.05 cm/cm3). Aswina, Swarna, M202, and Tainung 67 recorded lowest RLD (0.01 cm/cm3). 4.1.1.1.1.1.4 RSA (cm2) Means of RSA across different soil depths under both drought stress and well-watered condition are presented in Table 9. At 0-10cm soil layer in drought stress condition the highest RSA was recorded by IR64 (1036.77 cm2) followed by Dom Sufid (923.21 cm2) and FR13A (851.43 cm2). On the other hand, Azucena recorded lowest RSA (377.03 cm2) followed by Tainung 67 (453.10 cm2) and Cypress (555.67 cm2). Whereas under well-watered condition, Dom Sufid was recorded highest RSA (2053.87 cm2) followed by Sadu Cho (1958.67 cm2) and FR13A (1924.03 cm2). LTH was recorded lowest RSA (599.39 cm2) followed by M 202 (859.49 cm2) and Cypress (1012.98 cm2). Under drought stress condition, the highest RSA at 10-20cm soil layer was recorded by Swarna (587.52 cm2) followed by SHZ 2 (469.09 cm2) and Dom Sufid (430.62 cm2). On the other hand, N22 recorded lowest RSA (114.66 cm2) followed by M 202 (128.38 cm2) and Tainung 67(149.21 cm2). The highest RSA at 1020cm soil layer under well-watered condition was recorded by Azucena (658.87 cm2) followed Minghui 63 (545.84 cm2) and Aswina (497.05 cm2). Whereas
Cypress was recorded lowest RSA (152.02 cm2) followed by Tainung 67 (154.23 cm2) and M 202 (160.55 cm2). At 20-30cm soil layer in drought stress condition, the highest RSA was recorded by Swarna (372.33 cm2) followed by M 202 (315.04 cm2) and Dom Sufid (254.88 cm2). On the other hand, Cypress recorded lowest RSA (45.25) followed by LTH (59.33 cm2), and IR64 (59.77 cm2). Under well-watered condition, the highest RSA at 20-30 cm soil layer was recorded by Minghui 63 (361.92 cm2) followed SHZ 2 (290.50 cm2) and N22 (213.66 cm2). Whereas M 202 was recorded lowest RSA (41.97 cm2) followed by FR 13A (53.54 cm2) and Tainung 67 (54.50 cm2). Highest RSA at depth (30-45cm) under drought stress condition was recorded by Dular (123.73 cm2) followed by SHZ 2 (97.03 cm2) and FR13A (88.67 cm2). On the other hand, IR64 recorded lowest RSA (10.44 cm2) followed by M 202 (15.31 cm2) and Zhenshan 97B (16.57 cm2). All genotypes recorded drastically lower RSA at 30-45cm soil layer under well-watered condition. Among them Dular (57.55 cm2) and recorded highest RSA (0.06 cm2) followed Minghui 63 (54.96 cm2) and N22 (46.49 cm2). On the other hand, Tainung 67 recorded lowest RSA (7.68) followed by Swarna (9.43 cm2) and M 202 (10.27 cm2). 4.1.1.1.1.1.5 RV (cm3) Means of RV across different soil depths under both drought stress and well-watered are presented in Table 10. Under drought stress condition, the highest RV at 0-10cm soil layer was recorded by FR13A (14.28 cm3) followed by IR64 (11.46 cm3) and Dom Sufid (10.15 cm3). On the other hand, Tainung 67 recorded lowest RV (4.90 cm3) followed by Azucena (5.82 cm3) and M 202 (6.36 cm3). The highest RV at 0-10 cm soil layer under well-watered condition, was recorded by FR13A (25.00 cm3) followed by Aswina (22.19 cm3) and Dom Sufid (22.01 cm3). Whereas LTH was recorded lowest RV (5.90 cm3) followed by Azucena (11.41 cm3) and Cypress (11.82 cm3).
The highest RV at 10-20cm soil layer under drought stress condition was recorded by Swarna (4.88 cm3) followed by Aswina (4.52 cm3) and SHZ 2 (4.50 cm3). On the other hand, M 202 recorded lowest RV (1.19 cm3) followed by Tainung 67 (1.59 cm3) and LTH (1.62 cm3). Whereas the highest RV at 10-20cm soil layer under well-watered condition was recorded by Aswina (9.83 cm3) followed Azucena (7.94 cm3) and Moroberekan (5.33 cm3). Whereas Cypress was recorded lowest RV (2.04 cm3) followed by Tainung 67 (2.24 cm3) and M 202 (2.32 cm3). Under drought stress condition, the highest RV at 20-30 cm soil layer was recorded by Swarna (2.32 cm3) followed by Aswina (2.31 cm3) and M 202 (2.28 cm3). On the other hand, Cypress recorded lowest RV (0.39 cm3) followed by IR64 (0.41 cm3) and LTH (0.42 cm3). Under well-watered condition, the highest RV at 20-30 cm soil layer was recorded by SHZ 2 (4.98 cm3) followed N22 (3.91 cm3) and Dular (3.80 cm3). Whereas M 202 was recorded lowest RV (0.58 cm3) followed by FR13A (0.69 cm3) and Tainung 67 (0.82 cm3). Considerable variation was recorded under drought stress condition for RV at 30-45cm soil layer. Highest RV was recorded by Dular (0.86) followed by FR 13A (0.79 cm3) and Moroberekan (0.64 cm3). On the other hand, LTH, M 202 and Zhenshan 97B recorded lowest RV (0.11 cm3). Whereas under well-watered condition, all genotypes recorded lower RV at 30-45cm soil layer. Among them Minghui 63 recorded highest RV at 30-45 cm soil layer followed by Dular (0.71 cm3) and N22 (0.62 cm3). On the other hand, Tainung 67 recorded lowest RV (0.09 cm3) followed by Swarna and Sadu Cho (0.10 cm3). 4.1.1.1.1.1.6 RDW (g) Means of RDW across different soil depths under both drought stress and well-watered are presented in Table 11. Under drought stress condition, the highest RDW at 0-10cm soil layer was recorded by Azucena (1.97g) followed by FR13A (1.28g) and IR64 (1.25g). On the other hand, LTH recorded lowest RDW
(0.63g) followed by M 202 (0.66g) and N22 (0.72g). Under well-watered condition, at 0-10cm soil layer, FR13A was recorded highest RDW (2.88g) followed by Sadu Cho (2.19g) and Dom Sufid (1.80g). Whereas LTH was recorded lowest RDW (0.46g) followed by M 202 (0.82g) and Zhenshan 97B (1.08g). The highest RDW at 10-20cm soil layer under drought stress condition was recorded by Swarna and Moroberekan (0.56g) followed by Azucena (0.42g). On the other hand, N22 and M 202 recorded lowest RDW (0.13g) followed by IR64 and Tainung 67(0.17g). The highest RDW at 10-20cm soil layer under wellwatered condition was recorded by Azucena (0.66g) followed Aswina (0.50g) and Minghui 63 (0.48g). Whereas M 202 was recorded lowest RDW (0.12g) followed by Cypress (0.13g) and Tainung 67 (0.15g). Under drought stress condition, the highest RDW at 20-30 cm soil layer was recorded by M 202 (0.24g) followed by Aswina (0.22g) and Swarna (0.20g). On the other hand, LTH recorded lowest RDW (0.03g) followed by Cypress and IR64 (0.04g). Under well-watered condition, the highest RDW at 20-30 cm soil layer was recorded by Minghui 63 (0.29g) followed SHZ 2 (0.24g) and N22 (0.20g). Whereas M 202 was recorded lowest RDW (0.03g) followed by FR 13A and Tainung 67 (0.04g). At deep soil layer (30-45cm), Minghui 63 recorded highest RDW (0.05g) followed Dular, SHZ 2 and N22 (0.03g). Apart from these genotypes in wellwatered condition, most of other genotypes recorded lowest RDW. On the other hand considerable variation was recorded under drought stress condition for RDW at 30-45cm soil layer. Highest RDW was recorded by Aswina (0.11g) followed by Dular (0.07g). On the other hand, LTH, Zhenshan 97B and M202 recorded lowest RDW (0.01g).
4.1.1.1.1.2 Physiological Traits The mean performance of all OryzaSNP panel rice accessions in respect of PR (µmol mol-2 s-1), TR, SC, RWC, and LWP under both drought stress and wellwatered condition are presented below (Table 12). Azucena had highest PR (25.20 µmol mol-2 s-1) followed by Moroberekan (21.95 µmol mol-2 s-1) and M202 (21.05 µmol mol-2 s-1). On the other hand, FR13A recorded lowest PR (13.30 µmol mol-2 s-1) followed by SHZ2 (14.90 µmol mol-2 s-1) and IR64 (15.05 µmol mol-2 s-1). Highest TR was recorded in Azucena (3.20 m mol mol-2 s-1) followed by Moroberekan (2.97 m mol mol-2 s-1) and Dular (2.59 m mol mol-2 s-1). Whereas Pokkali recorded lowest TR (1.62 m mol mol-2 s-1) followed by SHZ 2 (1.69 m mol mol-2 s-1) and Tainung 67 (1.70 m mol mol-2 s-1). Among the genotypes evaluated, Aswina recorded lowest RWC of 60.83 % followed by SHZ 2 (61.85 %) and Sadu Cho (66.82 %). On the other hand, M 202 recorded highest RWC (83.21 %) followed by Dular (79.53 %) and IR64 (77.06 %) which were on par with one another. FR13A had highest SC (701.67 mol m-2 s-1) followed by Moroberekan (673.33 mol m-2 s-1) and N22 (651.67 mol m-2 s-1). On the other hand, Pokkali recorded lowest SC (376.33 mol m-2 s-1) followed by Aswina (383.33 mol m-2 s-1) and Zhenshan 97B (447.67 mol m-2 s-1). Dular recorded lowest LWP of -3.35 MPa followed by Moroberekan (-3.27 MPa) and Azucena (-3.08 MPa). On the other hand, Dom Sufid recorded highest LWP (-1.27 MPa) followed by Minghui 63 (-1.87 MPa) and IR64 (-2.17 MPa) Lowest LRS (1.0) was recorded for Aswina, Azucena, Moroberekan and Minghui 63 under drought stress condition. Whereas highest LRS was recorded in cypress (6.33) followed by LTH and Dular (5.67). 4.1.1.1.1.3 Phenology
The mean performance of all OryzaSNP panel rice accessions in respect of PH, TN and SDW under both drought stress and well-watered condition are presented below (Table 13). In drought stress condition, Swarna (39.17 cm) was the shortest followed by Minghui 63 (50.57 cm) and Zhenshan 97B (62.17 cm) while, Azucena was the tallest (129.78 cm) followed by Dom Sufid (113.89 cm) and Aswina (112.94 cm). Whereas under well-watered condition, M 202 (62.78 cm) was the shortest followed by Zhenshan 97B (64.61 cm) and Minghui 63 (75.00 cm) while, Azucena was the tallest (123.61 cm) followed by Aswina (118.94 cm) and Dom Sufid (117.78 cm). SHZ 2 had highest TN (26.33) followed by N22 (24.50) and Swarna (21.50) under drought stress condition while, Moroberekan had the lowest TN (7.00) followed by Azucena (9.00) and Dular (11.00). On the other hand Swarna had highest TN (38.00) followed by Minghui 63 (29.00) and SHZ 2(24.50) while, Azucena had the lowest TN (9.33) followed by Moroberekan (9.33) and Cypress (10.00) under well-watered condition. Pokkali had maximum SDW (29.51 g/plant) followed by FR 13A (24.79 g/plant) and Sadu Cho (22.06 g/plant) under drought stress condition. On the other hand, Tainung 67 recorded lowest SDW (10.27 g/plant) followed by Swarna (10.35 g/plant) and Azucena (11.43 g/plant). Whereas under well-watered condition, Sadu Cho had maximum SDW (38.47 g/plant) followed by Pokkali (34.54 g/plant) and Zhenshan 97B (23.97 g/plant). On the other hand, Azucena recorded minimum SDW (9.79 g/plant) followed by FR 13A (10.08 g/plant) and Tainung 67 (11.37 g/plant). 4.1.1.1.1.4 Agronomic Traits Means of all agronomical traits measured under both drought stress and well-watered are presented in Table 14.
4.1.1.1.1.4.1 SB (g/m2) Moroberekan had highest SB (4090.27g/m2) followed by FR13A (905.69 g/m2) and Azucena (852.44 g/m2) under drought stress condition. On the other hand, LTH recorded lowest SB (173.93 g/m2) followed by Zhenshan 97B (255.79 g/m2) and M 202 (289.40 g/m2). Cypress recorded lowest SB (224.68 g/m2) followed by LTH (272.57 g/m2) and Zhenshan 97B (304.60 g/m2) in well-watered condition. On the other hand, Swarna recorded highest SB (1470.30 g/m2) followed by IR 64 (881.68 g/m2) and FR13A (798.50 g/m2). 4.1.1.1.1.4.2 GY (g/m2) Under drought stress, Dular recorded highest GY of 254.05 g/m2 followed by N22 (252.13 g/m2) and Zhenshan 97B (166.58 g/m2). On the other hand, Minghui 63 recorded lowest GY (13.04 g/m2g) followed by Aswina (43.27 g/m2) and LTH (43.71 g/m2). Under well-watered condition, Cypress recorded lowest GY of 83.62 g/m2 followed by Aswina (85.30 g/m2) and Moroberekan (91.56 g/m2). On the other hand, IR64 recorded highest GY (409.93 g/ m2) followed by SHZ 2 (371.76 g/m2) and Minghui 63 (273.20 g/m2). 4.1.1.1.1.4.3 HI Zhenshan 97 B recorded highest HI (0.38) followed by N22 (0.34) and Dular (0.33) under drought stress condition. On the other hand, Moroberekan and Minghui 63 recorded lowest HI (0.03) followed by Aswina (0.06). SHZ 2 recorded highest HI (0.38) followed by Sadu Cho (0.34) in well-watered condition. Whereas, Aswina recorded lowest HI of 0.13 followed by Moroberekan (0.14) and Azucena (0.15). 4.1.1.1.2 Field Experiment – Dry Season 2009 4.1.1.1.2.1 Phenology and Grain Yield Parameters The mean performance of all OryzaSNP panel rice accessions under both drought stress and well-watered condition are presented below (Table 15).
4.1.1.1.2.1 PH (cm) In drought stress condition, Swarna (44.33 cm) was the shortest followed by IR64 (52.67 cm) and Zhenshan 97B (55.89 cm) while, Pokkali was the tallest (89.44 cm) followed by Dom Sufid (86.00 cm) and Sadu Cho (85.33 cm). Pokkali (103.78 cm) was the tallest followed by Dom Sufid (100.78cm) and Aswina (99.89 cm) under well-watered condition while, Zhenshan 97B was the shortest (70.44 cm) followed by Swarna (79.78 cm) and SHZ 2 (80.44 cm). 4.1.1.1.2.2 TN Rayada had highest TN (21.44) followed by Zhenshan 97B (20.44) and IR64 (20.22) in drought stress condition while, Moroberekan had the lowest TN (8.22) followed by Azucena (8.67) and Aswina (13.89). Highest TN was recorded by FR13A (21.33) followed by IR64 (20.89) and Rayada (20.11) in well-watered condition while, Dom Sufid had the lowest TN (10.89) followed by Zhenshan 97B (14.56), Sadu Cho and Aswina (15.11). 4.1.1.1.2.3 SB (g/m2) Under drought stress condition, Pokkali had maximum SB (555.00 g/m2) followed by FR13A (484.00 g/m2) and Azucena (450.67 g/m2). On the other hand, Dular recorded minimum SB (218.67 g/m2) followed by Sadu Cho (240.67 g/m2) and IR 64 (292.00 g/m2). Under well-watered condition, FR13A had maximum SB (1257.33 g/m2) followed by Dom Sufid (1050.67 g/m2) and Aswina (940.00 g/m2). On the other hand, Dular recorded minimum SB (480.00 g/m2) followed by Sadu Cho (574.00 g/m2) and Zhenshan 97B (578.67 g/m2). 4.1.1.1.2.4 GY (g/m2)) Among the genotypes evaluated, Swarna recorded highest GY of 473.33 g/m2 followed by FR13A (357.33 g/m2) and Zhenshan 97B (327.33 g/m2). On the other hand, SHZ 2 recorded lowest GY (126.67 g) followed by Dom Sufid (140.67 g/m2) and Rayada (148.67 g/m2) which were on par with one another. Most of the
entries did not produce any GYs in stress condition because of high temperature during grain filling and also due to severe drought stress. Because of this problem we measured GY only in well-watered condition. 4.1.1.1.2.5 HI As indicated above, HI was also calculated only in well-watered condition. Swarna and Zhenshan 97B had highest HI (0.36) followed by Sadu Cho (0.35). Dom Sufid recorded lowest HI of 0.12 followed by Rayada, SHZ 2 and Aswina with 0.16. 4.1.1.2 Analysis of Variance The mean sum of squares for root, shoot characters of both drought stress and well-watered condition (Table 16) and physiological traits of drought stress (Table 17) recorded in DS 2008 field experiment are presented in tabular form. Similarly mean sum of squares for all shoot and GY parameters recorded in DS 2009 field experiment are presented in Table 18. Highly significant differences among the genotypes were observed for all the characters in both experiments under both drought stress and well-watered condition (except RWC and SC under drought stress condition during DS 2008 experiment). 4.1.1.3 Variability Parameters Variability in respect of all the characters in all OryzaSNP panel accessions of rice are presented in tabular form and briefly described below. 4.1.1.3.1 Field Experiment – Dry Season 2008 4.1.1.3.1.1 Phenology, Physiological and Grain Yield Traits Variability parameters in respect of all shoot and physiological characters in all OryzaSNP panel accessions of rice under drought stress and well-watered are presented Table 19.
The overall mean PH of the genotypes was 83.65 cm with a range of 39.17 to 129.78cm under drought stress condition. High PCV and GCV values of 29.20 and 28.02 per cent respectively with high h2 estimate (92.10 %) and GA (55.39) as per cent of mean were recorded for this trait. PH ranged from 62.78 to 123.61cm with a mean value of 92.03 cm under well-watered condition. The genotypes showed high PCV (20.96 %) and GCV (20.30 %) values accompanied with high h2 (93.73%) and GA (40.48) as per cent of mean. TN varied from 10.27 to 28.84 with a mean value of 16.67 under drought stress condition. The genotypes showed high PCV (32.03 %) and GCV (30.71 %) values accompanied with high h2 (91.98 %) and GA (60.68) as per cent of mean. However genotypes varied largely (7.00-26.67) under well-watered condition with regard to TN and the overall mean for this trait was 15.74. High PCV (33.54 %) and GCV (32.38 %) were recorded for this trait with high h2 and GA as per cent mean of 92.23 % and 64.41 respectively. Under drought stress condition, the overall mean SDW of the genotypes was 10.00g/plant with a range of 5.00 to 24.00g/plant. High PCV and GCV values of 48.88 and 47.36 per cent respectively with high h2 estimate (93.90 %) and GA (94.95) as per cent of mean were recorded for this trait. SDW ranged between 9.79 to 38.47 g/plant with a mean value of 18.55g under well-watered condition. The PCV (43.20 %) and GCV (42.43%) co-efficients of variation were high accompanied with high h2 of 96.47 per cent and GA of 85.85 as per cent of mean. Genotypes exhibited a wide range of PR (13.30-25.20) with an average of 17.66 under drought stress condition. The values of PCV (19.26 %) and GCV (16.20 %) were moderate coupled with high h2 (70.74 %) and GA as per cent of mean (28.07) for this trait. TR ranged from 1.62 to 3.20 g with a mean value of 2.13 g under drought stress condition. The PCV (23.80 %) and GCV (20.90 %) values were high. This
trait also recorded high h2 (77.13 %) coupled with high GA (37.81) as per cent of mean. In drought stress condition, RWC varied from 60.83 to 83.21 with a mean value of 72.45. Moderate PCV (13.97%) with low GCV (2.30%) coupled with low h2 estimate (2.70 %) and GA (0.78) as per cent of mean were recorded for this trait. Genotypes exhibited a wide range of SC (376.33-701.67 mol m-2 s-1) with an average of 557.26 mol m-2 s-1under drought stress condition. High PCV (30.80%) with low GCV (7.09%) coupled with low h2 estimate (-5.30 %) and GA (-3.36) as per cent of mean were recorded for this trait. LWP ranged from -1.27 to -3.35 with a mean value of -2.64 under drought stress condition. High PCV (26.04%) with moderate GCV (16.45%) were recorded for this trait. Although this trait was recorded high h2 (39.90 %) coupled with high GA (21.40) as per cent of mean. Under drought stress condition, the overall mean LRS of the genotypes was 3.33 with a range of 1.00 to 6.33. High PCV and GCV values of 60.62 and 49.86 per cent respectively with high h2 estimate (67.66 %) and GA (84.49) as per cent of mean were recorded for this trait. The overall mean SB of the genotypes was 525.75g/m2 with a range of 173.93 to 905.69g/m2 under drought stress condition. High PCV and GCV values of 39.64 and 38.57 per cent respectively with high h2 estimate (94.66 %) and GA (77.30) as per cent of mean were recorded for this trait. SB ranged between 224.69 to 1470.30 g/m2 with a mean value of 576.77g/m2 under well-watered condition. The genotypes showed high PCV (49.01 %) and GCV (47.52 %) values accompanied with high h2 (94.03 %) and GA (94.93) as per cent of mean. Under drought stress condition, the overall mean GY of the genotypes was 122.30g/m2 with a range of 13.04-254.05g/m2. High PCV and GCV values of
53.54 and 52.56 per cent respectively with high h2 estimate (96.75 %) and GA (106.50) as per cent of mean were recorded for this trait. GY ranged between 83.62-618.16g/m2 with a mean value of 212.29g/m2 under well-watered condition. The PCV (65.42 %) and GCV (64.39%) co-efficients of variation were high accompanied with high h2 of 96.88 per cent and GA of 130.55 as per cent of mean. HI varied from 0.03 to 0.38 with a mean value of 0.20 under drought stress condition. The genotypes showed high PCV (57.96 %) and GCV (56.56 %) values accompanied with high h2 (95.23 %) and GA (113.71) as per cent of mean. The genotypes also varied largely (0.13-0.38) under well-watered condition with regard to HI and the overall mean for this trait was 0.27. High PCV (28.44 %) and GCV (25.13 %) were recorded for this trait with high h2 and GA as per cent mean of 78.05 % and 45.73 respectively. 4.1.1.3.1.2 Root Traits Variability parameters in respect of all root characters in all OryzaSNP panel accessions of rice under both drought stress and well-watered conditions are presented Table 20. 4.1.1.3.1.2.1 RLD (cm/cm3) In drought stress condition, genotypes exhibited wide range of RLD at 010cm soil layer (0.49-1.89 cm/cm3) with an average of 1.16 cm/cm3. High PCV (34.74 %) and GCV (24.25 %) coupled with high h2 (48.74 %) and GA as per cent of mean (34.88) were recorded for this trait. However genotypes varied largely (1.21-3.95 cm/cm3) under well-watered condition with regard to RLD at 0-10cm soil layer and the overall mean for this trait were 2.43 cm/cm3. High PCV (38.86 %) and GCV (33.45%) were recorded for this trait with high h2 and GA as per cent mean of 74.10 % and 59.33 respectively.
RLD at 10-20 cm soil layer ranged widely (0.16-1.45 cm/cm3) with a mean value of 0.63 cm/cm3 under drought stress condition. High PCV (53.57 %) and GCV (49.82%) were recorded for this trait with high h2 and GA as per cent mean of 85.92 % and 95.13 respectively. In well-watered condition, RLD at 10-20cm soil layer varied from 0.21 to 1.25 cm/cm3 with a mean value of 0.50 cm/cm3. The genotypes showed high PCV (66.14%) and GCV (59.64 %) values accompanied with high h2 (81.29 %) and GA (110.76) as per cent of mean. RLD at 20-30cm soil layer varied from 0.11 to 1.34 cm/cm3 with a mean value of 0.45 cm/cm3 under drought stress condition. The genotypes showed high PCV (73.53 %) and GCV (65.64 %) values accompanied with high h2 (79.69 %) and GA (120.71) as per cent of mean. Whereas, under well-watered condition genotypes varied from 0.06 to 0.87 cm/cm3 with regard to RLD at 20-30cm soil layer and the overall mean for this trait was 0.22 cm/cm3. High PCV (84.89 %) and GCV (81.16 %) were recorded for this trait with high h2 and GA as per cent mean of 91.40 % and 159.84 respectively. Genotypes varied largely from a range of 0.02 to 0.24 cm/cm3 with regard to RLD at 30-45cm soil layer and the overall mean for this trait was 0.09 cm/cm3. PCV (74.03 %) and GCV (72.20 %) values were high accompanied with high h2 (95.12 %) and GA as per cent of mean (145.06). However genotypes varied narrowly (0.01-0.06 cm/cm3) under well-watered condition with regard to RLD at 30-45cm soil layer and the overall mean for this trait was 0.03 cm/cm3. High PCV (65.62 %) and GCV (58.73 %) were recorded for this trait with high h2 and GA as per cent mean of 80.10 % and 108.10 respectively. RSA at 0-10cm soil layer ranged widely (377.03-1036.77 cm2) with a mean value of 686.36 cm2 under drought stress condition. High PCV (26.87 %) and GCV (23.05 %) were recorded for this trait coupled with high h2 and GA as per cent mean of 73.59 % and 40.74 respectively. RSA at 0-10cm soil layer also varied widely from 599.39-2053.86 cm2 with a mean value of 1389.05 cm2 in
well-watered condition. The genotypes showed high PCV (31.47 %) and GCV (27.51 %) values accompanied with high h2 (76.38 %) and GA (49.52) as per cent of mean. In drought stress condition, genotypes exhibited wide range of RSA at 1020cm soil layer (114.66-587.52 cm2) with an average of 296.15 cm2. High PCV (45.77 %) and GCV (40.58 %) coupled with high h2 (78.60 %) and GA as per cent of mean (74.12) were recorded for this trait. However genotypes also varied largely (152.02-658.87 cm2) under well-watered condition with regard to RSA at 10-20cm soil layer and the overall mean for this trait was 317.89 cm2. High PCV (45.46 %) and GCV (42.82%) were recorded for this trait with high h2 and GA as per cent mean of 88.74 % and 83.11 respectively. RSA at 20-30cm soil layer varied from 45.25 to 372.34 cm2 with a mean value of 164.46 cm2 under drought stress condition. The genotypes showed high PCV (59.47 %) and (56.88 %) values accompanied with high h2 (91.47%) and GA (112.06 cm2) as per cent of mean. Whereas, under well-watered condition genotypes varied from 41.97 to 361.92 cm2 with regard to RSA at 20-30cm soil layer and the overall mean for this trait was 143.58 cm2. High PCV (62.67 %) and GCV (56.33 %) were recorded for this trait with high h2 and GA as per cent mean of 80.78 % and 104.29 respectively. RSA at 30-45cm soil layer ranged widely (10.45 – 123.73 cm2) with a mean value of 49.75 cm2 under drought stress condition. High PCV (72.41 %) and GCV (68.28 %) were recorded for this trait coupled with high h2 and GA as per cent mean of 88.93 % and 132.65 respectively. RSA at 30-45cm soil layer also varied widely from 7.08 to 57.55 cm2 with a mean value of 23.77 cm2 in well-watered condition. The genotypes showed high PCV (66.72 %) and GCV (63.58 %) values accompanied with high h2 (90.82 %) and GA (124.82) as per cent of mean. Genotypes exhibited wide range of RV at 0-10cm soil layer (4.90-14.16 cm3) with an average of 8.50 cm3 under drought stress condition. High PCV
(29.46 %) and GCV (24.47 %) coupled with high h2 (69.03 %) and GA as per cent of mean (41.89) were recorded for this trait. Genotypes also varied largely (5.9025.00 cm3) under well-watered condition with regard to RV at 0-10cm soil layer and the overall mean for this trait was 16.84 cm3. High PCV (30.48 %) and GCV (27.21%) were recorded for this trait with high h2 and GA as per cent mean of 79.67 % and 50.03 respectively. RV at 10-20 cm soil layer ranged widely (1.20-4.89 cm3) with a mean value of 2.93 cm3 under drought stress condition. High PCV (42.67 %) and GCV (40.70 %) were recorded for this trait with high h2 and GA as per cent mean of 90.97 % and 79.97 respectively. In well-watered condition, RV at 10-20cm soil layer varied from 2.03 to 9.83 cm3 with a mean value of 4.44 cm3. The genotypes showed high PCV (45.83 %) and GCV (42.88 %) values accompanied with high h2 (87.51 %) and GA (82.63) as per cent of mean. RV at 20-30cm soil layer varied from 0.39 to 2.33 cm3 with a mean value of 1.29 cm3 under drought stress condition. The genotypes showed high PCV (55.86 %) and GCV (51.64 %) values accompanied with high h2 (85.48 %) and GA (98.36) as per cent of mean. Whereas, under well-watered condition genotypes varied from 0.58 to 4.65 cm3 with regard to RV at 20-30cm soil layer and the overall mean for this trait was 2.06 cm3. High PCV (60.98 %) and GCV (56.56 %) were recorded for this trait with high h2 and GA as per cent mean of 86.04 % and 108.08 respectively. Genotypes varied largely from a range of 0.11 to 0.86 cm3 with regard to RV at 30-45cm soil layer and the overall mean for this trait was 0.38.cm3 PCV (70.87 %) and GCV (66.08 %) values were high accompanied with high h2 (86.95 %) and GA as per cent of mean (126.94). However genotypes varied narrowly (0.09-0.83 cm3) under well-watered condition with regard to RV at 30-45cm soil layer and the overall mean for this trait was 0.29 cm3. High PCV (77.10 %) and
GCV (72.32 %) were recorded for this trait with high h2 and GA as per cent mean of 87.98 % and 139.75 respectively. RDW at 0-10cm soil layer ranged widely (0.63-1.97g) with a mean value of 0.94g under drought stress condition. High PCV (33.54 %) and GCV (31.80 %) were recorded for this trait coupled with high h2 and GA as per cent mean of 89.87 % and 62.10 respectively. RDW at 0-10cm soil layer also varied widely from 0.46 to 2.89g with a mean value of 1.44g in well-watered condition. The genotypes showed high PCV (37.19 %) and GCV (35.83 %) values accompanied with high h2 (92.84 %) and GA (71.12) as per cent of mean. Genotypes exhibited wide range of RDW at 10-20cm soil layer (0.13 – 0.56g) with an average of 0.30g under drought stress condition. High PCV (45.35 %) and GCV (42.65 %) coupled with high h2 (88.47 %) and GA as per cent of mean (82.65) were recorded for this trait. However genotypes also varied largely (0.12-0.66g) under well-watered condition with regard to RDW at 10-20cm soil layer and the overall mean for this trait was 0.31g. High PCV (50.94 %) and GCV (48.54%) were recorded for this trait with high h2 and GA as per cent mean of 90.82 % and 95.30 respectively. Under drought stress condition, RDW at 20-30cm soil layer varied from 0.03 to 0.35g with a mean value of 0.13g. The genotypes showed high PCV (65.53 %) and (62.46 %) values accompanied with high h2 (90.85%) and GA (122.65) as per cent of mean. Whereas, under well-watered condition genotypes varied from 0.03 to 0.29g with regard to RDW at 20-30cm soil layer and the overall mean for this trait was 0.12g. High PCV (61.60 %) and GCV (59.44 %) were recorded for this trait with high h2 and GA as per cent mean of 93.10 % and 118.15 respectively. At depth (30-45cm), RDW ranged widely under drought stress condition (0.01-0.11g) with a mean value of 0.04g. High PCV (76.89 %) and GCV (71.69 %) were recorded for this trait coupled with high h2 and GA as per cent mean of
86.93 % and 137.65 respectively. RDW at 30-45cm soil layer varied from 0.00 to 0.06g with a mean value of 0.02g in well-watered condition. The genotypes showed high PCV (80.10 %) and GCV (72.59 %) values accompanied with high h2 (82.12 %) and GA (135.50) as per cent of mean. Under drought stress condition, the overall mean RN of the genotypes was 187.34 with a range of 113.00 -259.67. High PCV and moderate GCV values of 32.88 and 19.59 per cent respectively with high h2 estimate (35.45 %) and GA (24.04) as per cent of mean were recorded for this trait. Under well-watered condition, RN ranged from 60.50 to 412.00 with a mean value of 231.18. The PCV (40.25 %) and GCV (38.71 %) were high accompanied with high h2 of 92.50 per cent and GA of 76.70 as per cent of mean. Genotypes exhibited wide range of RSR (9.33-38.00) with an average of 19.20 under drought stress condition. High PCV (38.00 %) and GCV (37.95 %) coupled with high h2 (37.14 %) and GA as per cent of mean (95.78) were recorded for this trait. However genotypes also varied largely (0.05-0.34) under wellwatered condition with regard to RSR and the overall mean for this trait was 0.12. High PCV (63.79 %) and GCV (60.48 %) were recorded for this trait with high h2 and GA as per cent mean of 89.91 % and 118.15 respectively.
4.1.1.3.2 Field Experiment – Dry Season 2009 4.1.1.3.2.1 Phenology and Grain Yield Parameters Variability parameters in respect of all shoot and physiological characters in all OryzaSNP panel accessions of rice under drought stress and well-watered are presented Table 21. The overall mean PH of the genotypes was 69.46 cm with a range of 44.33 to 89.44 cm under drought stress condition. High PCV and GCV values of 21.90 and 19.82 per cent respectively with high h2 estimate (81.89 %) and GA (36.94) as
per cent of mean were recorded for this trait. PH ranged from 70.44 to 103.78cm with a mean value of 91.85 under well-watered condition. The genotypes showed moderate PCV (19.83 %) and low GCV (4.57 %) values accompanied with low h2 (-5.33%) and GA (-2.17) as per cent of mean. TN varied from 8.22 to 21.44 with a mean value of 16.39 under drought stress condition. The genotypes showed high PCV (28.83 %) and GCV (24.32 %) values accompanied with high h2 (71.17 %) and GA (42.27) as per cent of mean. However genotypes varied largely (10.89-21.33) under well-watered condition with regard to TN and the overall mean for this trait was 16.56. High PCV (22.99 %) and moderate GCV (14.33 %) were recorded for this trait with high h2 and moderate GA as per cent mean of 38.85 % and 18.40 respectively. The overall mean SB of the genotypes was 360.67g/m2 with a range of 218.67-550.00 g/m2 under drought stress condition. High PCV and GCV values of 31.88 and 23.98 per cent respectively with high h2 estimate (56.59 %) and GA (37.16) as per cent of mean were recorded for this trait. Under well-watered condition, SB ranged between 480.00 to 1257.33 g/m2 with a mean value of 776.15g/m2. The genotypes showed high PCV (30.58 %) and GCV (25.17 %) values accompanied with high h2 (67.78 %) and GA (42.70) as per cent of mean. Under well-watered condition, GY ranged between 126.66-473.33g/m2 with a mean value of 244.41g/m2. The PCV (45.66 %) and GCV (40.70%) were high accompanied with high h2 of 79.46 per cent and GA of 74.75 as per cent of mean. HI varied from 0.12 to 0.36 with a mean value of 0.24 under well-watered condition. The genotypes showed high PCV (31.34 %) and GCV (25.84 %) values accompanied with high h2 (68.02 %) and GA (43.91) as per cent of mean. 4.1.1.4 Correlation Studies 4.1.1.4.1 Associations of Root and Shoot Characters with SDW and GY
Under drought stress condition, GY is positively correlated with deep root related traits such as RLD at 30-45cm, RSA at 30-45cm, RV at 30-45cm but was not significant. Similarly positive correlation of SDW per plant with deep root related traits i.e., RLD at 30-45cm, RSA at 30-45cm, RV at 30-45cm and RDW at 30-45cm also recorded along with its significant and positive correlation with RSA at 0-10cm and RDW at 0-10cm. On contrary GY is positively correlated with top root related traits and negatively related with deep root related traits under well-watered condition but was non significant in both the cases. Highly significant positive correlation is recorded between SB and GY (0.79) (Table 22 and Table 23). 4.1.1.4.2 Association among Root and Shoot Characters RLD at 0-10cm had highly significant and positive association with RSA at 0-10cm (0.74) under drought stress condition and with RSA at 0-10cm (0.77), RV at 0-10cm (0.59) and RDW at 0-10cm (0.59) and RN (0.73) under well-watered condition. RLD at 0-10cm also significantly correlated with RN (0.48) and RV at 0-10cm (0.55) under drought stress condition but at 5% significance level. Highly significant positive association of RLD at 10-20cm with RSA at 1020cm (0.87), RV at 10-20cm (0.65), RDW at 10-20cm (0.73) under drought stress and with RLD at 20-30cm (0.65), RSA at 10-20cm (0.84), RDW at 10-20cm (0.76) under well-watered condition were recorded. RLD at 10-20cm also significantly correlated with RLD at 20-30cm (0.53), RSA at 20-30cm (0.55), and RV at 20-30cm (0.51) under drought stress condition and with RSA at 20-30cm (0.50), RV at 10-20cm (0.54), and RDW at 20-30cm (0.51) under well-watered condition but at 5% significance level. Highly significant and positive correlation of RLD at 20-30cm with RSA at 20-30cm (0.88) and RV at 20-30cm (0.79) under drought stress condition and with RSA at 20-30cm (0.84), RV at 30-45cm (0.60), RDW at 20-30cm (0.81) and RDW at 30-45cm (0.65) under well-watered condition were recorded. At 5%
significance level, RLD at 20-30cm correlated with RSA at 10-20cm (0.46) under drought stress condition and with RLD at 30-45cm (0.51), RSA at10-20cm (0.55), RSA at 30-45cm (0.54) and RV at 20-30cm (0.50) under well-watered condition. At depth (30-45cm) RLD had highly significant and positive association with RSA at 30-45cm (0.92), RV at 30-45cm (0.83) and RDW at 30-45cm (0.69) under drought stress condition and with RSA at 30-45cm (0.92), RV at 30-45cm (0.89), RDW at 20-30cm (0.62) and RDW at 30-45cm (0.81) under well-watered condition. RLD at 30-45cm also significantly correlated with RV at 10-20cm (0.48) under drought stress condition and with RSA at 20-30cm (0.55) and RV at 20-30cm (0.55) under well-watered condition but at 5% significance level. Highly significant and positive correlation of RSA at 0-10cm with RV at 010cm (0.75) under drought stress condition and with RV at 0-10cm (0.81), RDW at 0-10cm (0.72) and RN (0.68) under well-watered condition was recorded. Highly significant positive association of RSA at 10-20cm with RV at 1020cm (0.83) and RDW at 10-20cm (0.82) under drought stress and with RV at 1020cm (0.83) and RDW at 10-20cm (0.90) under well-watered condition were recorded. RSA at 10-20cm also significantly correlated with RSA at 20-30cm (0.49) and RV at 20-30cm (0.49) under drought stress condition but at 5% significance level. RSA at 20-30cm had highly significant and positive association with RV 20-30cm (0.89) under drought stress condition and with RSA at 30-45cm (0.59), RV at 20-30cm (0.75), RV at 30-45cm (0.59), RDW at 20-30cm (0.93) and RDW at 30-45cm (0.66) under well-watered condition. RSA at 20-30cm also significantly correlated with RV at 10-20cm (0.46), RDW at 10-20cm (0.46) and RDW at 20-30cm (0.47) under drought stress condition but at 5% significance level.
At deeper soil layer (30-45cm) RSA had highly significant and positive association with RV at 30-45cm (0.90) and RDW at 30-45cm (0.79) under drought stress condition and with RV at 20-30cm (0.65), RV at 30-45cm (0.95) RDW at 20-30cm (0.67) and RDW at 30-45cm (0.80) under well-watered condition. RSA at 30-45cm also significantly correlated with RV at 10-20cm (0.48) and SB (0.50) under drought stress condition but at 5% significance level. At top soil layer (0-10cm) RV had significant and positive association with RV at 30-45cm (0.46, p