The International Rice Research Institute (IRRI) was established in 1960 by the Ford and Rock- efeller Foundations with the help and approval of the Government of the Philippines. Today, IRRI is one of the 15 nonprofit international research centers supported in part by more than 40 donors: members of the Consultative Group on International Agricultural Research (CGIAR – www.cgiar. org), other government funding agencies, foundations, the private sector, and nongovernment organizations.
The responsibility for this publication rests with the International Rice Research Institute.
Copyright International Rice Research Institute 2009
This publication is copyrighted by the International Rice Research Institute (IRRI) and is licensed for use under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 License (Unported). Unless otherwise noted, users are free to copy, duplicate, or reproduce, and distribute, display, or transmit any of the articles or portions of the articles, and to make translations, adaptations, or other derivative works under the following conditions:
Attribution: The work must be attributed, but not in any way that suggests endorsement by IRRI or the author(s).
NonCommercial: This work may not be used for commercial purposes.
ShareAlike: If this work is altered, transformed, or built upon, the resulting work must be distributed only under the same or similar license to this one. l For any reuse or distribution, the license terms of this work must be made clear to others. l Any of the above conditions can be waived if permission is obtained from the copyright holder. l Nothing in this license impairs or restricts the author’s moral rights. l Fair dealing and other rights are in no way affected by the above. l To view the full text of this license, visit http://creativecommons.org/licenses/by-nc-sa/3.0/
IRRI Limited Proceedings Series
The series allows IRRI scientists and partners to quickly share information with specialized institu- tions and individuals. It consists of proceedings from conferences, meetings, and workshops. To permit rapid publications, the review and editing may not be as rigorous as with formal proceed- ings.
Mailing address: IRRI, DAPO Box 7777, Metro Manila, Philippines Phone: +63 (2) 580-5600 Fax: +63 (2) 580-5699 Email: [email protected] Web: www.irri.org. Rice Knowledge Bank: www.knowledgebank.irri.org Courier address: Suite 1009, Security Bank Center 6776 Ayala Avenue, Makati City, Philippines Tel. +63 (2) 891-1236, 891-1174, 891-1258, 891-1303
Suggested citation: Hossain M, Bennett J, Mackill D, Hardy B, editors. 2009. Progress in crop improvement research. Los Baños (Philippines): International Rice Research Institute. 112 p.
ISSN 1607-7776 L P 2009 No. 14
Progress in Crop Improvement Research
M. Hossain, J. Bennett, D. Mackill, and B. Hardy, Editors
Contents
Preface...... v
Drought-prone ecosystems...... 1 Breeding rice for drought-prone areas of eastern India: accomplishments in. . . 3 the recent past and current scenario M.N. Shrivastava and S.B. Verulkar
Deepwater/boro ecosystmes ...... 13 Breeding rice for flood-prone, deepwater, and boro areas ...... 15 of eastern Uttar Pradesh J.L. Dwivedi Breeding for rainfed lowland, deepwater, and boro land in Bihar, India:...... 21 achievements and challenges R. Thakur, N.K.Singh, and J.N. Rai Breeding rice for deepwater and flood-prone areas of Thailand...... 28 Wilailak Sommut, Kalaya Kupkanchanakul, Prayote Charoendham, Udompan Promnart, and Suthep Nuchsawasdi
Shallow-flooded (submergence-prone) aman areas of South Asia ...... 43 Breeding rice for submergence-prone and aman areas of India ...... 45 S. Mallik, J. Ahmed, S.K. Bardhan Roy, J.N. Reddy, and G. Atlin Breeding rice for submergence-prone and aman areas of Bangladesh ...... 57 M.A. Salam
Tidal wetlands/problem-soil ecosystems ...... 63 Rice breeding for the tidal wetlands of Indonesia ...... 65 S. Sulaiman, I. Khairullah, and T. Alihamsyah Rice breeding for acid-sulfate soils in Vietnam...... 75 Bui Chi Buu and Nguyen Thi Lang Breeding rice for salt-affected areas of India ...... 78 R.K. Singh, B. Mishra, A.M. Ismail, and G.B. Gregorio
Upland ecosystems ...... 91 Breeding rice for the sloping uplands of Yunnan...... 93 D. Tao, F. Hu, G.N. Atlin, S. Pandey, P. Xu, J. Zhou, J. Li, and X. Deng Progress of upland rice breeding in Indonesia since 1991...... 98 Suwarno, E. Lubis, and B. Kustiano Breeding rice for the Indian plateau uplands ...... 104 P.K. Sinha, M.Variar, and N.P. Mandal
Breeding rice for drought-prone areas of eastern India: accomplishments in the recent past and current scenario iii iv Shrivastava and Verulkar Preface
This limited proceedings on the progress of crop improvement research highlights rice breeding research in various ecosystems in South and Southeast Asia. Specifically, these ecosystems cover drought, deepwater/boro, shallow-flooded (submergence-prone), tidal wetlands/problem soils, and upland areas. The drought-prone area involves rice breeding in eastern India, whereas deepwa- ter/boro highlights rice breeding in eastern Uttar Pradesh and Bihar in India and some areas in Thailand. India and Bangladesh are the cases presented for submergence-prone areas, while Indonesia, Vietnam, and some salt-affected areas of India fall under the tidal wetlands/ problem-soils ecosystems. As for upland ecosystems, the focus is on Yunnan’s sloping uplands, rice breeding in the uplands of Indonesia since 1991, and the Indian plateau uplands. We hope that this limited proceedings will contribute significantly to the further- ance of rice breeding in these different ecosystems, providing research updates that will enhance ongoing research and encourage future undertakings in these ecosystems.
Breeding rice for drought-prone areas of eastern India: accomplishments in the recent past and current scenario vi Shrivastava and Verulkar Drought-prone ecosystems
Breeding rice for drought-prone areas of eastern India: accomplishments in the recent past and current scenario Shrivastava and Verulkar Breeding rice for drought-prone areas of eastern India: accomplishments in the recent past and current scenario
M.N. Shrivastava and S.B. Verulkar
Rice is the staple food of millions of people and it plays a predominant role in determining their livelihoods. Eastern India, which comprises the states of Assam, Bihar, Jharkhand, Chhattisgarh, Madhya Pradesh, Orissa, eastern Uttar Pradesh, and West Bengal, and smaller states of the northeast region, accounts for 63.3% of the total rice area in the country. About 80% of this area is rainfed and is often exposed to abiotic stresses such as drought, low soil fertility, flood, and stagnant water. Although eastern India accounts for about three-fifths of the total rice area, it produces only 48% of the total rice produced in India. The rainfed rice ecology in eastern India can be broadly classified into upland and lowland. Weed competition, frequent drought, low input application, high incidence of blast and brown spot diseases, and the lack of suitable varieties are the major problems faced in the upland ecosystem. In the lowland ecosystem, especially in the shallow lowland areas, high incidence of insect pests and diseases, weeds, terminal-stage drought, and unavailability of good-quality cultivars are the major constraints. Seedling-stage drought and submergence are major issues under semideep- and deepwater situations. The coastal saline soils are often affected by iron and zinc deficiency, which causes chlorotic and low-tillering plants. The varietal improvement programs in the different eastern Indian states, which started in the first or second decade of the 20th century after the establishment of the Imperial Council of Agricultural Research in 1892 (and reoriented periodically), have led to the release of many improved varieties. Strategies for developing drought-tolerant, high-yielding varieties using conventional and molecular approaches are discussed in this paper.
Rice is the staple food of millions of people, playing a pre- rice cropped area in the country but produces only 48% of the dominant role in determining their economy, survival, migra- total rice. About 80% of the rice area of eastern India is rain- tion, social structure, and religious beliefs and rituals. Besides fed and exposed to abiotic stresses such as drought, low soil being cultivated in 122 countries, the crop is the lifeline of fertility, flood, and stagnant water (Singh and Singh 2000). the three most populous countries of the world—China, In- Even though the region receives good rainfall, yield losses dia, and Indonesia. Of all the rice commercially produced, caused by drought every year reach 2.9 million t annually 90% is directly consumed as food. A majority of the poor (Widawsky and O’Toole 1990). The yield losses caused by in several Asian countries spend as much as half of their drought at various stages combined are highest among all the income on rice. They depend on rice for around two-thirds biotic and abiotic stresses and about double those caused by of their calories and some 60% of their protein. Hence, any weeds (Widawsky and O’Toole 1990). Some of the poorest technology that improves the productivity and profitability rice farmers live in this region. However, this area has great of rice and reduces risk under adverse growing conditions potential—productivity can go up significantly. can make a big difference in the lives of millions of poor The yield potential of rice cultivars under favorable people all over Asia. environments appears to have reached its respective plateaus Water is the basic requirement of life. In agriculture, for more than 20 years, regardless of tremendous breeding availability of water is a major concern for farmers around the efforts. Over the next 50 years, the world population is pro- world and drought stress has been identified as one of the most jected to increase by some 3 billion, mainly in developing pervasive threats posed to agriculture by the environment. countries. The projected rice demand and targets of rice yield Drought is a major abiotic stress that limits plant growth and in the eastern region are shown in Figure 1. productivity and is a major cause of yield instability. With The percentage of area under rice in India has remained limited irrigation, eastern India is highly exposed to the static: in 1950-51, it was 23.6 million ha; it was 23.3 million vagaries of the monsoon. Eastern India comprises the states ha in 2000-01. Overall, rice productivity is showing signs of of Assam, Bihar, Jharkhand, Chhattisgarh, Madhya Pradesh, decline, rice area expansion is limited, investments in irriga- Orissa, eastern Uttar Pradesh, and West Bengal, and smaller tion have virtually ceased, increased fertilizer use threatens states of the northeast region. Here, rice is grown in the basins the environment, and good rice land is being diverted to of the Ganga, Yamuna, and Mahanadi rivers (and their tribu- other purposes. The only option left is to increase yields taries); this area has the highest intensity of rice cultivation of existing unfavorable lands in the years to come. This in the country. Eastern India accounts for 63.3% of the total challenge, associated with the existence and well-being of
Breeding rice for drought-prone areas of eastern India: accomplishments in the recent past and current scenario 60 Demand (million tons ) The rainfed rice ecosystem is usually classified into 50 Y ield targets (q ha –1) upland and lowland. Various constraints contribute to lower 40 the yield. These include the following: (1) moisture stress due to erratic and often inadequate rainfall, high runoff, poor 30 soil, and lack of facilities for rainwater and soil moisture 20 conservation; (2) flash floods and submergence due to poor 10 drainage, low-lying physiography, and high rainfall in sub- mergence-prone lowlands (as in Assam, West Bengal, and 0 1996 2000 2005 2010 2015 Bihar); (3) accumulation of toxic decomposition and iron Year toxicity; (4) continuous use of traditional varieties due to nonavailability of improved genotypes and lack of farmers’ Fig. 1. Projected rice demand and targets of rice yield in eastern India. Source: Singh and Singh (2000). awareness; (5) low soil fertility due to soil erosion leading to nutrient stress and imbalance in the use of fertilizers; and (6) heavy infestation by weeds and insect pests and lack of knowledge of control measures. Poor crop establishment (due to several factors) and poor adoption of improved crop millions of poor people in Asia, requires an integrated and production technologies (due to inappropriate technology systematic approach; a better understanding of type of rice and economic condition of farmers) (Singh and Singh 2000) culture, hydrology, climate (rainfall and temperature), soil further aggravate the situation. constraints, biological constraints (weeds, diseases, and insect pests), and socioeconomic factors (Mackill et al 1996); and Upland an appreciation of the importance of the sustainable use and Rainfed upland rice is characterized by the absence of stand- management of natural resources. ing water in fields a few hours after the rain stops. The total area under rainfed upland in the country is about 6 million Rainfed rice production ecosystems ha, which accounts for 13.5% of the total area under rice cultivation. Most rainfed upland areas are found in eastern The rice crop is grown in a complex set of socio-physical India. The productivity of upland rice is very poor (0.9 t ha–1). and biological environments that determine the performance Weed competition, frequent drought, low input application, and adaptation of a variety. It necessitates enhanced efforts high incidence of diseases (such as blast and brown spot), toward research prioritization and technology development, and lack of suitable varieties are the major constraints in delivery, and impact assessment activities. Before defining this ecosystem. any strategy aimed at changing a genotype for a particular environment, breeders must be acquainted with and should Lowland understand fully the target environment (Fischer et al 2003) The lowland areas in the eastern region of the country are in terms of type of rice culture, soil hydrology, rainfall pattern located in Assam, West Bengal, Bihar, Orissa, Chhattisgarh, and temperature, soil type, predominant pest complexes, as and eastern Uttar Pradesh. The total area under this ecosystem well as social structure. A broad classification of rice envi- is estimated to be around 14.4 million ha, which accounts ronments in different eastern Indian states is given in Table for 32.4% of the country’s total area under rice. The aver- 1 (see also Fig. 2). age productivity is low and ranges from 1.0 to 1.2 t ha–1.
Table 1. Rice area (million ha) under different ecosystems in eastern India.
Rainfed lowland Deep water Very deep State Irrigated Upland (50−100 water Total 0−30 30−50 cm) (>100 cm) cm cm
Assam 0.2 0.2 0.9 0.5 0.4 0.1 2.3 Bihar 1.5 0.5 1.7 0.5 0.4 0.7 5.3 Orissa 1.1 0.7 1.7 0.5 0.4 0.1 4.5 Chhattisgarh 0.6 0.7 2.2 0.1 – – 3.6 Uttar Pradesh 1.0 0.7 1.9 0.3 0.2 0.5 4.6 West Bengal 1.3 0.9 1.7 0.5 0.4 0.7 5.5 Total 5.7 3.7 10.1 2.4 1.8 2.1 25.8
Source: Singh and Singh (2000).
Shrivastava and Verulkar Ar ea (million ha) 2.5 Drought-prone 2.0 Drought- and submergenc e-prone Fav orable Submergenc e-prone 1.5
1.0
0.5
0 A s s am Bihar Oris s a Chhattis garh Uttar W es t Prades h Bengal
States Fig. 2. Rice areas (million ha) under different ecosystems in eastern India.
The lowland can be further classified into shallow water, emphasis shifted from selecting within types to semideep water, and deep water. The major constraints to selecting between types, as well as to combining higher yield under shallow water include high incidence of characters through hybridization. Attention was also insect pests and diseases, weeds, terminal-stage drought, diverted toward developing varieties with better and unavailability of a wider choice of cultivars with good grain quality, including aroma. Improving locally quality. Seedling drought and submergence are major issues cultivated landraces through pureline selection, im- under semideepwater and deepwater situations. proving grain yield potential through hybridization, and evolving quality rice continued to be the major Coastal saline area themes during the early period of the third quarter In eastern India, the coastal saline area is situated in West of the past century (1951-66) as well. The breeding Bengal and Orissa. Yield is about 1 t ha–1. The coastal saline objectives then were (1) earliness, (2) deepwater soils are often deficient in iron and zinc, which results in and flood tolerance, (3) lodging resistance, (4) chlorosis and reduced tillering. drought tolerance, (5) nonshattering grains, and (6) higher response to heavy manuring. Pureline Hills selection was mainly used for genetic improvement, These areas lie in the hilly regions of Chhattisgarh. The total which resulted in the development of 394 varieties. area is around 5% of the total cropped area. The productivity Breeders also tried to create new variability through of this region is very low (<1.0 t ha–1). The major problems hybridization, followed by the pedigree method of encountered are cold injury, blast, frequent drought spells, selection. This resulted in 51 varieties. All these im- and a very short cropping season. proved traditional varieties were high-yielding under low inputs, were resistant to prevailing insects and A varietal improvement program diseases, and were bred before the exploitation of the dwarfing gene, which resulted in very high-yielding, l First phase. The varietal improvement program in fertilizer-responsive cultivars, thereby changing the different eastern Indian states started in the first or rice breeding scenario. second decade of the 20th century after the estab- l Second phase. A new phase in rice breeding started lishment of the Imperial Council of Agricultural with the introduction of the semidwarf plant type in Research in 1892. Its focus has been reoriented 1965. The establishment of the International Rice periodically. The approach during this period was Research Institute (IRRI) in the Philippines and the to identify better purelines among similar types with coming into existence of the All India Coordinated the same name within each major group under cul- Rice Improvement Project ushered in a new era in tivation. The pureline selection method was mostly varietal improvement. Between 1966 and 1975, followed, which resulted in the release of cultivars several dwarf varieties came from IRRI and central such as N22, HB10, T141, T1242, Safri 17, Jhona institutes such as the Central Rice Research Institute, 349, Basmati 370, and others in different states. Cuttack; Indian Agricultural Research Institute, New Also, salt-tolerant lines such as SR26B and flood- Delhi; Directorate of Rice Research, Hyderabad; tolerant varieties such as FR13A were developed. and various agricultural universities made recom- During the second quarter of the 20th century, the mendations at different stages. The more important
Breeding rice for drought-prone areas of eastern India: accomplishments in the recent past and current scenario Table 2. Varieties developed before 1991 that are Table 4. Number of rice varieties released for still popular with farmers in eastern Indian states. different ecosystems in India, 1965-2000.
State Varieties Varieties released (no.) Ecosystem Chhattisgarh Kranti, Aditya, Tulsi, K3, IR36 Up to 2000 Until 2005 Madhya Pradesh Poorva, Anupama, Jawahar 75 Irrigated 314 422 Uttar Pradesh Narendra-1,2,97,118 Orissa Rudra, Subahdra, Annanda, Kalinga 3 Upland 84 283 Jharkhand Vandana, Brisa Dhan 103, 201 Lowland 123 Bihar Rajshree, Jaishree Semideep water 30 Deep water 14 Table 3. Area covered by dwarf variet- Saline alkaline 15 ies in some states of India, 2000-01. High altitude 33 Zone State Percentage Scented fine grained 19 Eastern states Assam 56.0 Total 632 705 Bihar 62.5 Source: DRR (2000), AICRIP Annual Workshop (2005). CG 69.1 Orissa 72.6 Southern states AP 77.9 Karnataka 87.0 ecology as the local parents used in hybridization quite often TN 96.5 Northern states Haryana 64.7 possessed a good amount of tolerance for drought (Table Punjab 96.0 4). Breeding materials, however, were rarely subjected to planned drought situations. The mechanisms and traits as- sociated with drought tolerance and recovery were poorly understood. Varieties such as IR36 played an important role in bringing drought-prone area under HYVs. Nevertheless, varieties were Taichung Native 1, IR8, Jaya, Bala, semidwarf cultivars spread considerably in rainfed drought- Cauvery, Sona, Ratna, IR20, IR28, IR36, Phalguna, prone environments. An excellent example is Chhattisgarh, Surekha, and Sarjoo 49. A number of donors were where semidwarf varieties occupied as high as 90% of the identified and varieties were bred for resistance cultivated area in plain regions, although more than 80% of to diseases such as blast, bacterial leaf blight, and the area was rainfed (Fig. 3). Most of the changes came only tungro virus and insects such as gall midge, brown in the late 1980s or early 1990s, when separate trials were planthopper (BPH), and whitebacked planthop- constituted for rainfed conditions at the all-India level by per. Resistance breeding received rice breeders’ directorates of rice research under the All India Coordinated considerable attention. It also became a priority as Rice Improvement Project. various biotic factors were seriously limiting yield, particularly where no other control measures were Current efforts available. It is important to understand the impact of these varieties. As an example, in Chhattisgarh, Mackill (1986) expressed the view that rice breeders were much emphasis was given to incorporating the genes well aware of the importance of drought as a major constraint for gall midge resistance into improved varieties. to yield and its stability. However, only a few crosses or The extensive cultivation of these varieties has selections were made because of the lack of knowledge and substantially reduced losses due to this pest. The confidence in drought screening and selection protocols. other pest, BPH, was not important initially and not Breeding for drought tolerance was simply not a part of much effort was made to control this menace. The their mainstream rice improvement program. The irrigated result was that this insect gained prominence as the environment usually requires fewer genotypes, resulting in most damaging pest in rice fields. A list of varieties maximum phenotypic expression. In contrast, the rainfed that evolved prior to 1991 and that remain popu- lowlands have a range of environments and involve so much lar is given in Table 2. Adoption of high-yielding diversity (moisture regime, fertility, flooding incidence, dwarf varieties was 96.5% and 96%, respectively, biotic factors, daylength, interaction among each other, in Tamil Nadu and Punjab, followed by Karnataka interaction with genotypes, etc.), resulting in complex phe- and Andhra Pradesh (Table 3). notypic expression and requiring a range of genotypes for Most of the varieties, which became popular under maximum phenotypic expression (Fischer 1996). It is critical the drought-prone rainfed ecology in the eastern states, that the factors contributing to low yield under the rainfed particularly Madhya Pradesh, Chhattisgarh, Uttar Pradesh, ecosystem be understood, particularly in the eastern Indian and Orissa, were spillovers from work done for the irrigated context. These include low crop yield at original sites, co-
Shrivastava and Verulkar Area (% of total rice area) 80
70 60 50 40 30 20 10
0 1965 1970 1975 1980 1985 1990 1995 2000 2004 Year Fig. 3. Approximate area under modern (dwarf) varieties in Chhattisgarh.
evolution of pests leading to a larger number of pests, wide be made to get higher yields from the same or reduced area topographical/ecological variation within a small geographic planted to the crop and perhaps with a lesser quantity of avail- area, tropical/subtropical areas with lower yield potential, a able water. This requires evolving a well-planned strategy to narrow genetic base, inefficiency of selection (more so in meet the challenges ahead. unfavorable conditions), abiotic stress undefined in time and space, and a few but vital biotic stresses. The other problems Components of the strategy are inadequate funding for breeding programs and training of breeders for specific environments. The following represent key components of a breeding strategy: Occurrence of drought and a breeding strategy l Recommending and popularizing early-duration varieties to enable farmers to cope with terminal- Availability of appropriate moisture is necessary to obtain stage drought. optimum yield from any crop plant. Under the rainfed rice l Using more local germplasm: The exploitation of ecosystem of eastern India, inadequate moisture, which is a narrow genetic base in breeding programs has linked to erratic rainfall during crop growth, is a crucial fac- resulted in reduced gain in improvement. Plants tor that affects overall yield. Although mean rainfall is quite have evolved several mechanisms to combat drought good, seasonal distribution and year-to-year variation are damage. It is believed that local landraces and wild high, as the rains may cease for any number of days at any species still have a good number of untapped genes. growth stage from germination to maturity. However, the Thus, it is important to explore new donors that rainfall pattern in the past 12 years at Raipur indicates that exhibit stable performance under water stress and terminal-stage drought is very frequent and medium-duration have good yield potential. The different mechanisms varieties are prone to terminal drought almost every second for drought resistance can bring in potential new year. Developing very early duration varieties is not a good genes/QTLs into the breeding pool. option as these are prone to damage by rain at maturity and l Identifying morphological and physiological traits a yield penalty is incurred due to early maturity. Therefore, associated with drought tolerance: Considering the developing drought-tolerant, 110–115-d-maturing, high- complexity of the rainfed ecosystem, it is imperative yielding, and stable genotypes is one of the major objectives to understand the myriad physiological interactions of rice breeding. and reduce that enormous complexity to a mana- Terminal-stage drought can be managed to a good gable level. There is a need to give information on extent by recommending varieties of appropriate duration in how it might be manipulated genetically to achieve the context of the water-holding capacity of the soil. Based practical results (Fischer 1996). Work on this aspect on this, breeders have developed and recommended many is in progress at different places. The work at IGAU varieties in different states (Appendix 1). These varieties has resulted in understanding interactions—for ex- have played a big role in improving yield in the eastern states ample, delay in flowering under drought conditions and bringing self-sufficiency on the food front. The situa- was related to low water status and was an indicator tion, however, is likely to worsen in view of ever-increasing of drought susceptibility; delay in flowering was also population, current rice and food crises, and a slowdown associated with higher spikelet sterility. Some geno- in productivity increases. Concentrated efforts will have to types increased root growth at reproductive-stage
Breeding rice for drought-prone areas of eastern India: accomplishments in the recent past and current scenario Table 5. Promising hybrids identified for the drought and yield stability was attained by maintain- rainfed ecology. ing higher plant water status. This maintenance of Hybrid Yield Yield advantage higher water status under water-limiting conditions (t ha–1) (%) over check was a key to achieving drought tolerance. KRH-2 3.7 31 l Standardization of drought-screening protocols and development of rapid screening protocols: The prob- PSD-1 3.6 28 PHB-71 3.6 27 lem of within-field heterogeneity is a major one as it CORH-2 3.6 26 reduces selection efficiency and heritability of traits PA 6201 3.6 25 under stress. The well-managed screening protocol Best check 2.8 – in the field needs to be standardized to increase the heritability of drought-related traits and repeatability of results. Agronomic adjustments such as sowing and transplanting may be made to increase the prob- the pedigree method is most widely used in most ability of exposing test material to drought. situations, and the backcross method for incorporat- l Proper selection of field: When water is drained from ing QTLs is now increasingly used. Since drought a field, the water level becomes uneven; this imposes situations are quite varied and location-specific, test- heterogeneity in the field and genotype performance ing of material in farmers’ fields is now being done. becomes unpredictable. Physical (recording the Involvement of farmers at various stages of breeding depth of freely available water in soil), biological would ensure better adoption of lines when these (resistant and susceptible checks), and statistical are finally released. Multilocation and multiyear tools can be used to overcome the problem of within- testing (though a regular component of all breeding field heterogeneity. There is also an urgent need to programs) is much more pertinent in the case of develop rapid screening techniques for large-scale developing varieties for rainfed environments than screening of segregating populations. for irrigated areas. l Incorporating resistance to major diseases: The l The use of hybrids in less favorable rainfed environ- drought-prone environments are quite often as- ments: Because of their genetic plasticity and faster sociated with incidence of particular diseases and growth, hybrids can yield better under adverse con- insects. Blast and brown spot are two such diseases. ditions (Table 5). Their adaptation under such con- It is necessary to incorporate at least a good level ditions is to be encouraged. Considerable progress of resistance to blast as this is controlled by major has been noted recently in some areas, particularly genes. in Chhattisgarh. l Under rainfed conditions, weeds are a major yield Most of the abovementioned approaches are known constraint. The genotypes need to have early vigor and are being used by breeders, but, now that productivity to compete with weak growth. in favorable irrigated areas has already reached a plateau, no l Water-use efficiency and stay-green traits: These stone is left unturned to improve yields in rainfed environ- two traits have not yet received adequate attention. ments. It is heartening to note that most breeders are adopting Their role in raising the yield potential of genotypes more and more systematic approaches and modern tools to under drought situations needs to be understood and achieve their targets. used to the extent possible. l Exploiting molecular tools: Current advances in References molecular biology have raised the hopes of making a breakthrough in improving tolerance of plants DRR (Directorate of Rice Research). 2000. Annual progress report. for biotic and abiotic stresses. Identifying major as Fischer KS. 1996. Improving cereals for the variable rainfed system: well as minor genes (quantitative trait loci [QTLs]) from understanding to manipulation. In: Singh VP et al, editors. and/or candidate genes has become possible. These Physiology of stress tolerance in rice. Proceedings of the Inter- technologies may lead to some basic understanding national Conference on Stress Physiology of Rice, 28 Feb.–5 as well as an increase in selection efficiency, opening March 1994, Lucknow, Uttar Pradesh, India. p 1-10. up possibilities of developing transgenic lines with Fischer KS, Fukai S, Lafitte R, McLaren G. 2003. Know your target increased stress tolerance. environment. In: Fischer KS, Fukai S, Lafitte R, Atlin G, Hardy l Developing local mapping populations can be an B, editors. Breeding rice for drought-prone environments. Los important approach to identifying genes that may Baños (Philippines): International Rice Research Institute. p impart adaptability to local adverse conditions re- 5-11. lated to the environment or the soil. Mackill D. 1986. Varietal improvement for rainfed lowland rice in l Following appropriate breeding methodologies: South and Southeast Asia: results of a survey. In: Progress in Adoption of appropriate breeding methods is es- lowland rice. Los Baños (Philippines): International Rice Re- sential for combining genes of interest. Therefore, search Institute. p 115-144.
Shrivastava and Verulkar Mackill DJ, Coffman WR, Garrity DP. 1996. Rainfed lowland rice Notes improvement. Los Baños (Philippines): International Rice Research Institute. 242 p. Authors’ address: Indira Gandhi Agricultural University, Raipur Singh,VP, Singh RK, editors. 2000. Rainfed rice: a source book for best 492006, India. practices and strategies in eastern India. Los Baños (Philippines): International Rice Research Institute. 292 p. Widawsky DA, O’Toole JC. 1990. Prioritizing the rice biotechnology research agenda for eastern India. New York: The Rockefeller Foundation.
Appendix 1
Rice varieties developed/released in Chhattisgarh (1991 onward)
Featuresc Variety Parentage Suitabilitya Duration (d) Yield (t ha–1) Grain typeb (resistance to or Year released tolerance of)
IR64 IR5657-32-2-1/IR2061- RF, Irr 112 4.0−4.5 LS Bl, BLB 1992 465-1-5-5 Shyamala R60-2712/R 2389 RF, Sh. LL 130 4.5 LS – 1993 Mahamaya Asha/Kranti RF, LL 125 5.5−6.0 LB GM, BLB, drought 1995 Poornima Poorva/IR8608 RF, DS 110 3.5−4.0 LS GM, BLB, drought 1996 Danteshwari Samridhi/IR8608 RF UpL & Sh. LL 105 3.5−4.0 LS GM, drought 2001
Rice varieties developed/released in Madhya Pradesh (1991 onward)
Variety Parentage Suitability Duration (d) Yield (t ha–1) Grain type Features Year released
JR3-45 Lodhi/IR36 RF, DS 100 2.5–3.0 SB, red Drought 1997 ker JR353 IR36/JR75 RF, DS 110 3.0–3.5 MS LS, Dr 1998 JR201 – RF, DS 105 2.5–3.0 LS GM, BLB 2001 Richa – RF, DS 100–105 LS MT 2004
Rice varieties developed/released in Jharkhand (1991 onward)
Variety Parentage Suitability Duration (d) Yield (t ha–1) Grain type Features Year released
Birsadhan Fine Gora/IET RF Sh. LL 95–110 2.2–3.0 LB BM, SB, Bl, BLB, BS 1992 103 2832 Vandana C 22/Kalakeri Upland 90–95 3.0–3.5 LB SB, termite, Bl, BS 1992 Hazaridhan IR42/IR5853- RF Sh. LL 115–120 4.0–4.5 LS SB,WBPH, Gundhi, Bl, 2003 118-5 BLB, BS, drought Sadabahar BRRI SAIL/ RF Sh.LL 105–110 3.5–4.0 LB SB, Bl, BS, BLB, 2003 IR10181- drought 58-3-1 aRF = rainfed, Irr. = irrigated, LL = lowland, Sh. LL = shallow lowland, UpL = upland, DS = dry season. bLS = long slender, LB = long bold, SB = short bold, MS = medium slender, LM = long medium, MB = medium bold. cBl = blast, BLB = bacterial blight, GM = gall midge, SB = sheath blight, WBPH = whitebacked planthopper, RTV = rice tungro virus, LF = leaf folder, SR = sheath rot.
Breeding rice for drought-prone areas of eastern India: accomplishments in the recent past and current scenario Varieties released in Bihar (1991 onward)
Variety Parentage Suitability Duration (d) Yield (t ha–1) Grain type Features Year released
Kamini Katarni RF LL – – MS BPH. BLB, drought 1992 Shakuntala Pankaj/BR 8 RF UpL 145–150 3.0–4.5 LM BPH, BLB, Bl 1994 Turant Dhan Rasi/Sattari Timely or 70–75 3.0–4.0 SB SB, Bl, drought 1994 late sown Vandehi Beldar Drought- – 3.0–4.5 LB Bl, BLB, RTV, 1995 prone LL drought Prabhat IR2033-521- Early and 95–100 3.5–4.5 MS SB, other major 1997 1/IR261- late sow- insect pests, BLB, 264-2//IR3 ing BS Satyam RD19/Desaria RF LL 145–150 3.0–6.0 LS BPH, WBPH, LF, 1997 8 BLB, Bl, BS, BLB, Bl Kishori IR8/Barogar RF LL 140–145 6.0–7.0 LB LF, BPH BLB, BS 1998 Richharia Pusa 33/ Upland 100–105 3.5–6.0 Long fine SB,BLB, BS 2000 IET7464 Santosh Pankaj/BR34 Intermedi- 145–150 3.5–5.0 LS WM, Gundhi bug, 2001 ate land BLB, BS, SR Saroj Gautam/ Fav. upland, 115–120 4.5–5.0 LS SGP 2001 Type 3 medium land
Varieties released/developed in West Bengal (1991 onward)
Variety Parentage Suitability Duration (d) Yield (t ha–1) Grain type Features Year released
Bhupen C22/IR26 and Drought- 110–114 2.0–2.4 LS SB, drought 1993 C22/054 prone areas Jamini BG 280-12/ Drought- 95–110 3.2 LB Bl, BLB, ShR, ShB 1996 PTB 33 prone laterite Khanika Jaya/CR 237-1 Drought- 75–85 3–4 LS BPH, BS, ShB 1996 prone DS Shantabdi CR 10-114/CR Drought- 112–145 3.5–4 LS ShB 2000 10-115 prone DS
Varieties release/developed in Eastern UP (1991 onward)
Variety Parentage Suitability Duration (d) Yield (t ha–1) Grain type Features Year released
Narendra 7 N 22/Ratna Drought- 85–90 4.5–5.0 MS Bl 1992 prone Barani (-1064-5?IR Drought- 95–100 4.5–5.0 MB Drought 2001 Deep 9129-320- prone 3-3-3/IR54) – –
10 Shrivastava and Verulkar Varieties release/developed in Assam (1991 onward)
Variety Parentage Suitability Duration (d) Yield (t ha–1) Grain type Features Year released
Moniram Pankaj/Mah- Sh. LL 155 5–5.5 MS BLB, Bl 1992 suri Ketekijoha (Badsabhog Sh. LL 160 3–3.5 SB GM, SB, BLB, ShB 1994 /Savitri)/ Badsabhog Satyarajan IET9711 Medium 130 – MB GM (biotypes 1&2), 1996 /IET11162 lands BPH, WBPH, SB, LF, Bl.
Rice varieties developed/released in Orissa (1991 onward)
Variety Parentage Suitability Duration (d) Yield (t ha–1) Grain type Features Year released
Badami Suphala/An- Uplands/bialy 90–95 3.5 MB GLH, SB, BPH, BLB, 1992 napurna lands BS Ghan- IR2061-628/N Irr, RF UpL 90–95 3.3 MB SB, GM, BLB, SR, 1992 teswari 22 ShB, RTV Khandagiri Parijat/ Irr, RF UpL 90–95 3.2 MS BPH, GM, Bl, SR, 1992 IR13429- ShB, BLB, RTV 94-3-2-2 Nilagiri Suphala/DZ Medium 90–95 3.0 MB SB, GM, BPH BLB, 1992 192 lands, LS SR Gajapati OR 136-3/ Medium 125 4.4 MS GM, SB, BPH, 1999 IR13429- lands WBPH, Bl, BS, 196-1-20 ShR, BLB, RTV Kharavela Daya/ Medium 125 4.4 MS SB, GM, BPH, 1999 IR13240- lands WBPH, Bl, RTV, 108-2-2-3 ShR, ShB, BLB Konark Lalat/OR 135- Medium 125 4.5 LS GM, SB, BPH, Bl, 1999 3-4 lands BS, ShR, BLB, RTV Lalitagiri Badami/ Upland 95 3.2 MB SB, GM, WBPH, BS, 1999 IR19661- ShR, Bl, Dr 364 Udayagiri Savitri/IRAT Upland 95 3.5 MB SB, BPH, WBPH, Bl, 1999 138/IR RTV, ShB, ShR, 13543-66 BLB, Dr Bhanja IR36//Hema/ RF, Irr me- 135–140 4.0 MB GM,SB, BLB, BS, 1992 Vikram dium lands, Bl, ShR, lodging, Sh. LL shattering Birupa ADT 27/IR 8// RF, Irr me- 130–135 4.0 Coar. SB, GM, cutworm, 1992 Annapurna dium lands WBPH, ShB, SR, BLB, Bl, RTV Mahalaxmi Pankaj/Mah- Irr, RF Sh. LL 145–155 4.0 MB BPH, ShB, FSm, LF, 1992 suri ShB, BLB Manika CR 1010/OBS RF, Irr LL, LS 155–160 4.5 MB SB, GM, BPH, ShB, 1992 677 RTV,BLB Meher (OBS-677/ RF, Irr me- 135–140 4.0–6.5 MB GM, SB, BPH, 1992 IR2071)/ dium lands WBPH, BLB, SR (Vikram/W 1263) Samanta T90/IR8//Vi- RF, Irr me- 140 4.5 MB GM (biotypes 1&4), 1992 kram///Siam dium lands SB, WBPH, BLB, 29/Mahsuri BS, Bl, SR, ShB
Breeding rice for drought-prone areas of eastern India: accomplishments in the recent past and current scenario 11 Rice varieties developed/released in Orissa (1991 onward)
Variety Parentage Suitability Duration (d) Yield (t ha–1) Grain type Features Year released
Santep Pankaj/Sigadis RF, Irr Sh. 150–155 4.5 MB GM, BLB, SR, ShB 1992 Heap 3 LL Urbashi Rajeswari/Ja- RF and Irr 145 4.0 MB GM, SB, BLB, SR, 1992 jati Sh. LL, LS ShB, RTV Bhoi Gouri/RP 825- 120 3.9 MB GM,WBPH, BPH, SB, 1999 45-1-3 Bl, BS, BLB, ShB, RTV Indravati IR56/OR 142- Sh. LL 150 4.1 MS BPH, WBPH, SB, Bl, 1999 99 BS, ShR, RTV Mahanadi IR19661-131- Sh. LL 150 4.4 MB SB, WBPH, LF, Bl, 1999 1-3/Savitri BS, ShR, BLB, RTV Prachi IR9764-45- Sh. LL 155 4.3 MB BPH, WBPH, BS, 1999 2-2/CR BLB, RTV 149-3-2 Ramchandi IR17494-45- – 155 4.0–6.5 MB Shallow water and 1999 2-2/jagan- flash floods nath Sebati Daya/IR36 Medium – 4.1 MS GM, Bl, BS, BLB, 1999 lands RTV, ShB, ShR Surendra OR158-5/Rasi Medium 135 4.7 GB SB, GM, LF, BPH, 1999 lands WBPH, Bl, BS, SB, ShR, RTV, BLB
12 Shrivastava and Verulkar Deepwater/boro ecosystems
Breeding rice for flood-prone, deepwater, and boro areas of eastern Uttar Pradesh 13 14 Dwivedi Breeding rice for flood-prone, deepwater, and boro areas of eastern Uttar Pradesh
J.L. Dwivedi
Flood-prone rice areas in India are mainly located in various eastern states such as Assam, Bihar, Orissa, Uttar Pradesh, and West Bengal. Of the total of 2.3 million hectares of these lands, eastern Uttar Pradesh has 0.39 million ha. These areas are more vulner- able than the other rainfed ecosystems as their yields are low and quite unstable. In eastern Uttar Pradesh, four different flood-prone subecosystems exist: deep, semideep, flash-flood-prone, and off-season deep stagnant. Many types of rice varieties have been selected by farmers over the centuries that meet the varied and harsh conditions of this ecology. These traditional types give stable but low yields. The IRRI-ICAR-NDUAT collaborative project on flood-prone rice established in 1991 resulted in a characterization of the environment followed by strong work on genetic improvement of rice grown in various conditions of this fragile ecology. This paper deals with the accomplishments of this collaboration with IRRI, with special reference to breeding rice for flood-prone, deepwater, and boro areas of eastern Uttar Pradesh. The technologies have significantly helped to increase on-farm productivity, meet household needs, and generate income for resource-poor farmers of the flood-prone areas of eastern Uttar Pradesh.
India has approximately 2.4 million hectares of flood-prone yielding traits is the basis for developing improved varieties rice (FPR), of which 80% are now in eastern India. The extent for this ecosystem. and distribution of rice areas under different ecosystems in The IRRI-ICAR-NDUAT collaborative project on FPR India as well as eastern India are provided in Table 1. Eastern since 1991 has resulted in the characterization of rice environ- Uttar Pradesh shares about 390,000 ha of FPR area. This area ments, development of nondestructive screening techniques suffers from uncontrolled flooding and drought. Low rice for assessing elongation ability, identification of suitable yields in this area, as a rule, have yet to meet the projected donors, and development of improved varieties for different needs for the future, and an increase in rice productivity will subecological FP conditions, besides limited validation of likely have to come from this fragile ecosystem. Rice farming improved germplasm in the target environment. in eastern Uttar Pradesh is complex, diverse, and risk-prone. Increased production in this area depends on submergence- Characterization of the rice environment tolerant varieties and other appropriate technologies. Many types of rice have been selected by farmers over A good understanding of the socioeconomic and biophysical the centuries that meet the varied harsh conditions of this aspects of the environment is crucial to breeding research. ecosystem. Combining their adaptive mechanisms with high- The IRRI-ICAR collaborative project helped in the systemat- ic mapping of the extent and distribution of rice under excess water conditions, supported by field observations. In eastern Uttar Pradesh, four different flood-prone subecosystems ex- Table 1. Extent and distribution of rice area under different ecosystems in India. ist: deep, semideep, flash-flood-prone, and off-season deep Total area (million ha) stagnant flooding. In the deep subecology, maximum water Ecosystem depth is usually >100 cm and could increase up to 300 cm. All India Eastern India The semideepwater situation includes areas with maximum Upland 7.0 5.2 water depth up to 100 cm. The rise of water is slow and Rainfed lowland 12.1 9.4 the water is not turbid. Flash-flood-prone is the largest area Shallow water (0–30 cm) (8.1) (5.9) among the four subecosystems mentioned. Flooding can Intermediate (30–50 cm) (4.0) (3.5) occur up to 9 times per season, with depth ranging from 30 Deep water 5.4 4.6 to 120 cm depending on rainfall (Table 2). Damage caused Semideep (50–100 cm) (3.0) (2.5) by flash flood can be quite severe. Off-season deep stagnant Deep (> 100 cm) (2.4) (2.1) rice grown in flood-prone depressed areas during the dry Irrigated 17.8 7.1 season is called boro.
Source: CRRI, Cuttack, India.
Breeding rice for flood-prone, deepwater, and boro areas of eastern Uttar Pradesh 15 Table 2. Flood spells at Ghagharaghat during the wet seasons.
Year Water inception date Flood duration (d) Peak water depth (cm)
2000 (9)a 8 June 4 35 22 June 3 40 2 July 3 11 8 July 3 21 12 July 14 36 29 July 10 70 8 Aug 2 30 11 Aug 12 72 30 Aug 15 36 2001 (3) 15 July 6 55 27 July 10 52 21 Aug 3 16 2002 (3) 12 Aug 3 12 22 Aug 5 23 6 Sept 6 36 aNumbers in parentheses indicate number of flood spells.
Genetic improvement and Senadhira (1996) studied the genetics of these traits and reported a single major gene with additive Collaboration strengthened rice breeding research by devel- effects involved in this trait. IR31142-14-1-1-3-1- oping nondestructive screening techniques for elongation 1-2, IR31406-333-1, and IR40931-33-1-3-2 were ability and refining the submergence screening test, identi- identified as good donors for submergence tolerance. fying donors for better elongation ability and submergence Rice yields are severely lowered by flash flood in tolerance, and studying the inheritance of these flood-prone addition to stagnant flooding. traits during the early stage of collaboration. In addition, substantial progress has been made in this subecology in Development of varieties developing improved germplasm and varieties. Brief accom- plishments related to these aspects are described below. Flood-prone deepwater ecology l Screening techniques: Efforts were made to develop The Rice Research Institute of Thailand took the responsibil- nondestructive screening techniques for assess- ity for developing and distributing improved deepwater (DW) ing elongation ability in flood-prone varieties as rice in Southeast Asia in 1992. Because of the difference in the screening test used for elongation ability was latitude, breeding materials developed in Thailand were not destructive. Elongation is essential for survival in suitable for most DW areas of India as Thai genotypes were floating rice. Floating rice elongates rapidly under late in India. To solve this problem, selections were made in submergence. Accordingly, three nondestructive Thailand and at IRRI. With this improvement in selection, techniques, GA3 application at the 3–5-leaf stage, as- the adaptability and acceptability of germplasm provided sessing elongation without flooding, and elongation by IRRI and Thailand were enhanced. It is believed that this under shallow water, have been developed. These collaboration resulted in improving grain quality traits. To techniques could be used in assessing segregating further streamline FPR research in northeastern Thailand, populations without losing nonelongating types. special research and training workshops were held in Thai- They were successfully used to determine the ge- land in 1992, 1994, and 1995, in which FPR breeders from netics of elongation ability (Dwivedi and Senadhira this region also participated. 1996). Substantial progress was made in this fragile ecology. l Genetic studies: Dwivedi and Senadhira (1996) Since 1993, four improved varieties—Jalnidhi for floating found elongation to be controlled by two dominant ecology (1993); Jalpriya and Jal Lahari for semideep and complementary genes and heritability was also high. shallow deep water, respectively; and Barh Avarodhi (1995) Submergence tolerance is necessary for rice grown for flash-flood areas—were released. These varieties are in medium-deepwater areas. Complete submergence spreading in their respective areas. could occur at any time during crop growth. Dwivedi
16 Dwivedi Table 3. Performance of promising submergence-tolerant breeding lines/varieties for yield and other traits.
Genotype Date to 50% Plant height Panicle No. of tillers Yield flowering (cm) length (cm) hill–1 (kg ha–1) Madhukar 15 Oct. 180.1 24.4 8.2 3,400 Barh Avarodhi 27 Oct. 175.7 26.2 10.2 3,900 NDGR 24 30 Oct. 161.3 25.7 6.7 3,100 FRG 10 27 Oct. 179.7 26.0 8.4 3,636 FRG 13 27 Oct. 181.4 27.2 10.2 4,040 NDGR 60 21 Oct. 181.5 24.3 8.6 3,800 NDGR 63 20 Oct. 171.6 25.4 6.8 3,750 NDGR 65 19 Oct. 181.7 26.4 8.2 4,200 NDGR 80 27 Oct. 188.6 26.3 9.0 4,350 NDGR 81 4 Nov. 156.8 25.8 9.2 4,667 NDGR 82 27 Oct. 163.8 27.5 8.2 3,437 NDGR 83 28 Oct. 182.4 26.4 8.6 3,850 NDGR 84 29 Oct. 176.7 27.2 8.4 3,700 NDGR 85 28 Oct. 181.8 26.8 8.6 3,850 NDGR 86 30 Oct. 181.3 27.4 8.8 4,440 NDGR 87 30 Oct. 188.6 26.4 8.2 3,840 NDGR 88 23 Oct. 186.4 26.8 8.4 3,900 NDGR 37 25 Oct. 188.7 24.4 8.2 3,750 NDGR 104 16 Oct. 190.4 26.0 7.4 3,867 NDGR 105 28 Oct. 163.7 24.4 8.0 3,667 NDGR 106 15 Oct. 184.7 26.2 8.4 3,637 NDGR 107 25 Oct. 183.7 24.4 8.2 3,737 NDGR 108 26 Oct. 191.4 23.8 7.8 3,850 NDGR 109 14 Oct. 186.7 24.6 8.0 3,950 NDGR 110 22 Oct. 175.7 25.8 8.6 4,267 NDGR 111 28 Oct. 184.7 24.6 8.6 3,900 NDGR 103 21 Oct. 176.7 26.8 8.2 3,037 NDGR 70 22 Oct. 181.4 27.0 8.4 3,900 FR13(A) 20 Oct. 156.4 24.8 8.2 3,850 Sabita 30 Oct. 176.9 25.2 7.4 4,040
Two of the most promising floating lines, NDGR 426 Off-season deep stagnant/boro rice (IET 11873) and NDGR 427 (IET 11871), have been recom- Because of submergence coupled with waterlogging in the mended for multiplication of seeds and minikits for the areas rice season, the flood-prone environment is subject to a high where water depth surpasses 100 cm for at least 30 days, in risk of crop loss. Off-season deep stagnant boro rice grown addition to NDGR 433, 444, 445, 448, and 449. Promising in the dry season is one promising option for such deeply lines possess fast elongation ability besides kneeing ability flooded areas. About 50,000 ha of potential boro area exist and nodal tillering. NDGR 231, 272, 279, 260, 268, 283, in the state. However, at present, barely 4,000–5,000 ha of 290, and 269 for semideep water have been identified. These land are under boro rice. Most of the improved germplasm promising lines have moderate elongation. grown in the boro season is cold-sensitive. No systematic In eastern Uttar Pradesh, damage due to flash flood is breeding program was undertaken in the past to improve boro quite high. Yields are highly unstable and depend on flooding varieties although there is immense potential to increase the depth, duration of crop submergence, and stage of the crop rice productivity of such lands. when submerged. With increased submergence tolerance, Accordingly, efforts were made in 2001 and 2003 to yields of this subecology could be increased. To identify identify the potential boro rice-growing areas in eastern Uttar genotypes possessing submergence tolerance, 25 genotypes Pradesh with the aim of evaluating the available technology along with tolerant variety FR13A and susceptible variety and improved rice germplasm developed in other states IR42 were evaluated following the Standard Evaluation through a scientist-farmer participatory approach (Dwivedi System developed by IRRI. NDGR 83, 84, 85, 86, and 109 1996). showed submergence tolerance on a par with FR13A and High-yielding cold-tolerant varieties developed at Sabita. High survival of 75–80% was recorded in these RAU, Pusa, Bihar, and Vivekanand Parvatiya Krishi Anu- promising lines. These breeding lines were also assessed for sandhan Sansthan, Almora, Uttaranchal, were evaluated at days to 50% flowering, plant height, number of tillers per the Crop Research Station, Ghagharaghat, Bahraich, U.P., hill, and panicle length. Some of these breeding lines showed during 2000. Three boro rice varieties, Richharia, Prabhat, little elongation coupled with high survival, indicating they and Dhan Laxmi, from RAU, Pusa, and 20 cold-tolerant are true submergence-tolerant types (Table 3). lines from Almora were planted. Though all the entries had
Breeding rice for flood-prone, deepwater, and boro areas of eastern Uttar Pradesh 17 Table 4. Performance of some improved varieties during boro season, 2002.
50% flowering 50% flowering Days to Yield (kg Variety Seedling vigor from the date from the date of maturity ha–1) of transplanting nursery (days) (days)
NDR-97 Moderate 63 130 158 3,823 Barani Deep Moderate 64 130 160 3,676 Gautam Good 69 137 175 3,241 Prabhat Very good 69 137 160 5,210 Richharia Good 54 131 170 4,203 Saroj Good 72 145 171 4,071 Dhan Laxmi Good 75 147 170 5,021 Joymati Average 63 140 167 3,648 Boro Dhan Good 60 130 180 4,040
Date of nursery: 23 November 2001 CD: 0.682 Date of transplanting: 28 January 2002
Table 5. Overall scores for farmer opinion and average yield of good germination, seedlings could not tolerate the chilling breeding lines/varieties evaluated under flash-flood conditions cold prevailing at this location. As a result, high mortality during 2000-02 across locations in Barabanki and Bahraich of was recorded at the seedling stage. However, VL 93-288, VL eastern Uttar Pradesh. Dhan 81, and VL 88-1011 were promising. Farmers Average score Average yielda (q ha–1) During 2002, eight improved boro rice varieties were Variety (no.) for opinion evaluated in farmers’ fields in the traditional boro rice- 2000 2001 2002 growing areas of Gorakhpur and Sant Kabir Nagar districts. NDGR 70 9 2.4 1,930 2,310 – Seedling vigor, days to maturity, phenotypic acceptability NDGR 24 12 2.9 1,700 2,040 – score (range of 1 = excellent to 9 = poor), and grain yield FRG 13 9 3.0 1,730 2,030 1,040 of these varieties are given in Table 4. Prabhat (5.21 t ha–1) NDGR 63 9 2.9 1,550 1,930 – was the outstanding variety, followed by Dhan Laxmi (5.02 Madhukar 9 4.0 1,380 1,760 870 t ha–1) and Richharia (4.20 t ha–1). Other varieties showed Barh Avarodhi 15 2.3 1,630 2,240 1,090 Sugapankhi 9 3.1 1,280 1,650 1,510 moderate performance. Gautam was slightly late in maturity. NDGR 88 5 3.0 – – 1,070 These varieties were also assessed on larger plots in farmers’ NDGR 82 7 3.0 – – 820 fields involving 13 farmers from Ramnagar Kajraha Village in Gorakhpur District and 4 farmers from Jhumia Village a2000 WS = excess rain coupled with flooding, 2001 WS = normal year, 2002 WS = drought year. in Sant Kabir Nagar District. Similar patterns in yield and maturity duration were observed. This indicates that there is immense scope to increase flood-prone ecologies, on-farm trials were conducted using the rice productivity of boro areas, where the main rice lines/varieties suitable for flash-flood cultivated areas in crop fails due to excess water, by providing integrated crop eastern Uttar Pradesh using participatory selection during the management practices such as cold-tolerant, short-duration 2000-02 WS. On-farm trials included seven varieties/lines varieties and nursery management to reduce the effects of (Table 5). These activities were implemented in two districts cold injury. of eastern Uttar Pradesh, Bahraich and Barabanki, involving three villages and 5–7 farmers from each district, where most Validation of improved germplasm in target environments of the farmers were small to marginal. Farmers’ field evaluations were done in researcher-de- Rice yields are severely reduced by flash floods in addition signed and -managed trials. Farmers were provided with only to stagnant flooding. Tolerant rice varieties can make a sig- seed as an input. The rest of the inputs had to be supplied by nificant difference in rice yield. the farmers. Farmers were asked to grow the lines/varieties Improved varieties released in the past have been found using their own methods. Seeds were given to the farmers limiting in adaptation to varying on-farm conditions probably along with a questionnaire with varietal characteristics to be because of the very wide variability in this ecology, besides scored in comparison with the local popular variety. These farmers’ needs and preference. It is therefore necessary to characteristics were vigor, lodging, reaction to diseases and consider the farmers’ participation in varietal evaluation. pests, yield, grain quality, threshability, and overall farmers’ The farmers’ participatory varietal selection approach opinions and reasons. Scores given to characteristics were 1 = was followed for selecting appropriate varieties by taking very good, 2 = good, 3 = medium, equal to farmers’ variety, into account farmers’ preferences. To stabilize rice yield in 4 = bad, and 5 = very bad.
18 Dwivedi Table 6. Number of scores of farmer opinion given to each entry and corresponding remarks made by farmers.
Number of farmers giving Variety scorea of Mean Farmers’ remarks
2 3 4
NDGR 70 7 – 2 2.4 Moderate submergence tolerance NDGR 24 4 3 3 2.9 Bold shining grain FRG 13 4 1 4 3.0 Stiff straw NDGR 63 4 2 3 2.9 Moderate survival Madhukar – – 9 4.0 Susceptible to neck blast, red pericarp Barh Avarodhi 13 – 2 2.3 Bold shining grain, better survival Sugapankhi 4 – 5 3.1 Drought-tolerant and stable NDGR 88 – 5 – 3.0 Moderate performance NDGR 82 – 7 – 3.0 Good survival and medium slender grain
aScores for farmers’ opinions: 1 = very good, 2 = good, 3 = medium (equal to farmers’ variety), 4 = bad, 5 = very bad.
Table 7. Average yield of varieties/lines under stagnant flooding and over- all score of farmers in participatory varietal trials conducted in different districts of eastern Uttar Pradesh during WS 2002 and 2003.
Average yield (q ha–1) Average score Variety for opiniona Barabanki Bahraich Basti Faizabad
2002 2003 Ranjit 2,360 1,870 2,746 2,493 2,429 2.4 Bahadur 1,700 2,610 2,510 2,490 3,088 3.0 Rajshree 2,800 2,436 2,200 2,776 3,583 2.3 B. Avarodhi 2,573 2,785 1,956 2,558 2,498 2.4 Jalpriya 1,766 – 2,137 – 3,031 3.1 Jal Lahari – – – – 4,141 2.3
aScores for farmers’ opinions: 1 = very good, 2 = good, 3 = medium (equal to farmers’ variety), 4 = bad, 5 = very bad.
The yields of flood-prone rice are highly unstable and tween states, it can be disseminated to other states/countries depend on flooding depth, duration of crop submergence, having similar ecological conditions. and stage of the crop when submerged. There is not much choice in the released varieties of Yield of tested entries was variable during 2000-02 Uttar Pradesh for stagnant deep flooding. Hence, to expedite because of changes in rainfall distribution and flooding the interstate flow of promising rice varieties developed pattern (Table 2). The lowest yield was recorded during the elsewhere in the country for deepwater conditions, rice varie- exceptional drought year (2002) and entries yielded higher ties Rajshree from Bihar, Ranjit and Bahadur from Assam, in 2000 and 2001. However, yield trends were similar. Most along with Barh Avarodhi and Jalpriya were evaluated fol- of the farmers scored variety Barh Avarodhi as good and lowing the farmers’ participatory approach during the 2002 it was rated first in both districts for its bold and shining WS in four districts—Bahraich, Barabanki, Faizabad, and grain (Tables 5, 6). Even other varieties such as NDGR 70 Basti—involving three farmers from each district. Varietal and NDGR 24 yielded more, along with NDGR 63, which performance differed among the test sites. Rajshree yielded flowered 20 days earlier without sacrificing yield. the highest (3,583 q ha–1), followed by Barh Avarodhi (2,468 Rice varieties performed well in Assam and Bihar under q ha–1), Ranjit, and Bahadur (2,324 q ha–1) considering overall the multilocational farmers’ participatory varietal selection performance across locations (Table 7). program in five states of eastern India with flood-prone Most of the farmers scored the varieties good in com- conditions. Since the farmers’ perceptions on adoption and parison with their own variety because of their higher yield. acceptability of Barh Avarodhi were positive within and be- Among these, Rajshree was rated first because of desired
Breeding rice for flood-prone, deepwater, and boro areas of eastern Uttar Pradesh 19 Table 8. Number for scores for farmer opinion given to each variety and corresponding remarks made by farmers.
Number of scores Variety Mean Remarks 1 2 3 4 5
Ranjit – 7 – 2 – 2.4 Brown spot infestation Bahadur – 4 1 4 – 3.0 Lodging tendency, brown spot and stem borer Rajshree – 13 – 2 – 2.3 Timely maturity, good yield under shallow water B. Avrodhi – 7 – 2 – 2.4 Slightly late, good survival, bold shining grain Jalpriya – 4 1 4 – 3.0 Fine grain, neck blast Jal Lahari – 13 – 2 – 2.3 Good tillering and high yield, trace of bacterial blight
maturity and lowest infestation of diseases and pests (Ta- References bles 7, 8). The majority of the farmers preferred Jal Lahari because of its high tillelring ability and good yield, followed Dwivedi JL. 1996. Problem and prospects of boro rice in Eastern U.P. by Rajshree. Paper presented at International Flood-prone Rice Workshop The most promising lines rated highly by most farmers held at Rajendra University, Pusa, Bihar, India, 28-31 October and that produced relatively higher yield were Jal Lahari and 1996. Rajshree. Bahadur, although it produced a higher yield (3,088 Dwivedi JL, Senadhira D. 1996. Genetic control of elongation ability q ha–1), was not rated good because of its lodging tendency in flood-prone rice. Paper presented at the 2nd International and moderate infestation of stem borer. It was also moderately Crop Science Congress held 7-24 November 1996, New Delhi, susceptible to brown spot. India. (Abstract.) Though modest, the varieties have significantly helped to increase on-farm productivity. It is not envisaged that Notes the varieties of the future will alone enhance productivity but they should have stable performance. Therefore, efforts Author’s address: Senior rice breeder, Crop Research Station, N.D. should address yield stability and greater cropping diversity University of Agriculture and Technology, Masodha, Faizabad in varietal development besides yield alone for this complex 224 133, Uttar Pradesh, India. fragile ecosystem.
20 Dwivedi Breeding for rainfed lowland, deepwater, and boro land in Bihar, India: achievements and challenges
R. Thakur, N.K.Singh, and J.N. Rai
Rice in Bihar covers about 3.8−3.9 million ha of land under diverse ecological conditions. More than 70% of this area is rainfed. The rainfed lowland is the most important ecological system, followed by deep water. The state has two geographi- cal divisions, north and south, and in both divisions rice is the main crop. Because of the wide diversity in riceland in Bihar, productivity is poor. However, in recent times and in collaboration with IRRI, there has been significant progress in developing improved varieties for different situations, with the breeding program being reoriented on the basis of existing challenges and needs. Rainfed lowlands in Bihar fall within two broad categories: unfavorable and favorable. The favorable ecosystem virtually resembles the irrigated system, whereas the unfavorable ecosystem is highly variable. Varietal requirements for these two ecosystems are different; consequently, selection and evaluation criteria are modified. New entries are now evaluated jointly by farmers and scientists using farmers’ management practices. Through this process, a pureline selection, Vaidehi, was developed, while two other high-yielding varieties, Satyam and Kishori, were bred under the shuttle breeding program. Both have multiple resistance to biotic and abiotic stresses and are adapted to late planting. Santosh, a variety with good grain quality, was developed through the farmers’ participatory breeding project. In deepwater rice land, excess water causes serious damage and more tolerant and higher yielding varieties are needed. Boro (dry-season) rice, grown from Oct.-Nov. to Apr.-May, is highly productive, but germplasm with tolerance for cold stress at the seedling stage is required to expand the boro area. Recently, an EMS mutant genotype, Rasi, was developed with high cold tolerance and high yield potential (more than 8 t ha–1). It was released as Gautam. In subsequent years, Richharia, Dhanlaxmi, and Saroj (all with different maturity durations) were developed for the boro season. These varieties are becoming more popular and cultivated area is progressively expanding in traditional and nontraditional boro regions. There is now a subtle change in the attitude of farmers: yield is no longer the sole criterion for adopting new varieties; they also assess other traits such as cooking quality and an attractive look. This shift in preference could help popularize new varieties with good grain quality, even in the rainfed ecosystem.
Keywords: rainfed lowland, deepwater rice, boro rice, elongation, submergence, chaur, cold tolerance
Bihar is a typical eastern state of India where rice is cultivated Physiographic features on 3.8–3.9 million ha of land under diverse agroecologi- cal conditions. Rice is cultivated on more than 60% of the The Ganga River divides Bihar into two distinct broad cropped area. Agriculture is the most dominant sector of the regions: the north Bihar alluvial plains and the south Bihar state economy and 82% of the population is directly depend- alluvial plains. ent on it. Consequently, rice production determines food North Bihar alluvial plains. This region is practi- security and stability in the state. Rice as a food crop is well cally plain land sloping toward the southeast. Many natural known for its unique adaptation. It is grown as an upland crop riversGandak Ghaghra, Burhi Gandak, Kamal Balan, Bag- on precarious moisture conditions and in water as high as 2–3 mati, Koshi Balan, Mahananda, and their tributariesflow m, just like in deepwater areas of Bihar. There are several through this area. The region has two distinct agroclimatic intermediate conditions between these two extremes. Apart zones: northwest and northeast. The northwest (zone 1) cov- from these, rice is also grown under diverse climatological ers Chapra and Tirhut divisions. The zone slopes toward the and edaphic situations, which are often not conducive to southeast, as seen in the direction of the river flow. Vast wa- proper growth and productivity. More than 70% of the rice terlogged areas in the districts of Saran, Vaishali, Samastipur, area in Bihar is rainfed, with productivity determined mainly Muzaffarpur, and Begusarai are found. Due to the near-flat by the pattern of the monsoon. landscape and saucer-shaped depressions (chaurs) between the rivers, vast areas become flooded during the rainy sea- son in practically all districts in this zone. When floodwater recedes from most areas, the saucer-shaped depressions and the abandoned channels of the rivers and lakes remain flooded
Breeding for rainfed lowland, deepwater, and boro land in Bihar, India: achievements and challenges 21 for various durations. Due to the construction of irrigation requirements are different. The constraints encountered in infrastructure and canal systems under the Gandak project, these ecologies are dealt with separately in this paper. most of the natural drainage systems had been disturbed. Rainfed lowland. This ecosystem constitutes more This caused more flooding and waterlogging in the central than 50% of the rice area in Bihar and is the most important, and western parts of the zone. Embankments and roads also However, it faces the following constraints: worsened the waterlogging problem. The Ganga River serves l The hydrology is irregular. When monsoon rain is as the major drainage channel in which all riverflow from heavy, there is excess water in the field; when rain various points meets. Total rice area in the region is about is scanty, drought is experienced at any growth 1.5 million ha. period. The northeast plain (zone II), covering Kosi Division, l Flash flood could inundate the fields for at least a generally slopes toward the southeast; Kosi, Mahananda, and week at any time during the season. Ganga are the major rivers. This zone is full of abandoned l The monsoon pattern determines sowing and plant- beds and dead channels of the Kosi River and its tributaries. ing time. If rain is timely, sowing is normal. Planting In addition, small lakes, oxbow lakes, and patches or marshy is delayed when rain is delayed. This happens in grounds are frequently encountered. The total area of rice in canal-irrigated systems also. this zone is about 0.79 million ha. l Photoperiod-sensitive tall cultivars are predomi- South Bihar alluvial plains. This region has two divi- nantly cultivated; these have weak culms and are sions and slopes toward the northeast. Important rivers other prone to lodging. The yield potential of these culti- than the Ganga are the Sone, Punpun, Falgu, Badua, and Chir. vars is low, but they are adapted to prevailing abiotic Except for the Sone River, all others are seasonal. The region stresses. They seem to tolerate both excess water has comparatively higher irrigation resources and more than and drought and are adapted to delayed sowing and 60% of the land is normal, not like the situation in the north transplanting, which occur because of variation in Bihar plains. Total rice area is about 1.5 million ha. the onset of the monsoon season. Deepwater rice. Deepwater rice land, locally called Challenges and constraints chaur, is found all over north Bihar, covering about 0.83 million ha (Catling 1992). This is the main source of liveli- When high-yielding varieties (HYVs) with features such as hood of millions of resource-poor farmers. Water on the dwarf habit; short, erect leaves; and better response to N fer- peripheral parts of the chaur remains shallow, medium in the tilizer were developed in the late sixtieswhich triggered the middle, and very deep in the center. Varieties are used on the Green Revolution in India and elsewherebreeding methods basis of this water depth gradient. Tall, photoperiod-sensitive were modified everywhere to accommodate these HYVs. cultivars with no elongation ability are grown on the rela- The general belief was that the Green Revolution technology tively shallow periphery, and taller elongating cultivars are could provide solutions to all rice production problems. Rice grown in the middle. Cultivars with high elongation ability breeders were consequently made to develop short-stature, are grown in the deeper central parts. However, almost all N-responsive varieties for all rice ecosystems, including current cultivars are traditional. The transplanting method deepwater areas (Morshima-Okinow 1964). The success is used in the shallow areas and direct seeding is practiced of the efforts was not tangible. It was soon realized that the in the deeper areas. The major constraint in this ecosystem Green Revolution technology was best suited for the more is the unpredictable rainfall during different growth stages favorable ecosystems such as irrigated lands but had little and reasonable yield is thus not assured. If a crop survives relevance to the more heterogeneous and unfavorable rainfed and matures in the central portion, harvesting is usually done production systems. A few varieties were developed through using boats, as the water remains in the field for a very long this approach for the rainfed lowland and deepwater ecosys- period, even up to March or April. tems. Their average yields were very high in managed trials Boro rice. Boro rice is grown during winter in low-ly- and they were therefore recommended for release. However, ing areas, especially in river deltas in both eastern India and none of them were widely adopted. Mansarowar was released Bangladesh. It is commonly grown in deeply flooded areas in the late eighties for the rainfed lowlands based on yield of northeastern Bihar, West Bengal, Assam, eastern Uttar data from the all India coordinated trials. When this variety Pradesh in eastern India, and in many areas in Bangladesh. was included in frontline demonstrations in farmers’ fields in Groundwater in these areas is shallow and waterlogging/ typical lowland areas and was planted a little late, it did not water stagnation prevails in some places during winter. The flower. Such varieties can better express their yield potential winter rice crop grown in Myanmar also resembles this boro in a favorable environment. Because the rainfed system is rice. It suffers from damage due to low temperature of vary- highly heterogeneous, varieties for this ecosystem must be ing degrees at the seedling stage. Where temperature during able to withstand local environmental fluctuations. Rainfed winter is not too low, boro rice covers large areas as in As- lowland and deepwater ecologies are parts of a continuum sam (Pathak 1995), West Bengal (Chatterjee 1995), and in of rice ecosystems, but they differ with respect to water small areas in Bihar (Thakur 1995). In Bihar, cultivation is depth and other abiotic stresses. Consequently, the varietal mainly confined to favorable areas adjoining West Bengal,
22 Thakur et al the traditional boro region, and has not spread beyond this while restructuring the new breeding projects designed for the area mainly because of very low temperatures in so-called rainfed ecosystems, where different breeding programs were nontraditional boro areas. developed and executed for each of the major ecosystems. A Boro rice produces more grain yield than wet-season clear understanding of the major challenges in these ecosys- rice in the same ecology (Singh et al 2003). As previously tems became the basis for revising the breeding objectives. mentioned, some varieties such as Gautam recorded yields Rainfed lowland. Broadly, rainfed lowlands have two between 8 and 10 t ha–1, even in farmers’ fields (Thakur et al distinct subecosystems: favorable and unfavorable. Unlike 1994). The productivity of boro rice is high mainly because the situation in the unfavorable ecosystem, several HYVs of higher solar radiation, lower night temperature throughout were available for the favorable ecosystem. Consequently, the crop growth period in winter, and favorable temperature local cultivars were predominant in the unfavorable areas. As during ripening (Singh et al 2003). a breeding strategy, pureline selections that were sufficiently Major constraints: Low temperature during winter tested were promoted in areas with extreme adverse condi- prevents the expansion of boro rice in nontraditional low- tions based on practical experience and the poor performance lying areas. This is because low temperature leads to of the released modern varieties (Thakur 1995). Local germ- l Poor and slow germination of seeds; plasm was thus included in the multilocation trials conducted l Stunted seedling growth and death in some cases for the rainfed lowland, along with improved breeding lines. when temperature becomes very low; The following are examples of varieties common in these l Nonsynchronized tillering and flowering; areas: l Delayed flowering and maturity, which increases l Mahshuri. This cultivar was once very popular in irrigation cost and exposes the crop to higher tem- the lowlands. However, it virtually failed when peratures during summer. planted late (more than two consecutive years) due Among the biotic stresses commonly encountered dur- to delayed monsoon rains. It also became highly ing the boro season are insect pests such as brown planthop- susceptible to BPH and drought. The breeding pro- pers (BPH), stem borers, and leaffolders and diseases such gram then focused on developing a variety that is as brown spot and sheath blight. Drought and western rains more suited to these growing conditions. As stated during maturity are also encountered. Currently, boro rice earlier, experimental trials were usually conducted receives greater attention and higher priority in the state’s under normal sowing and planting conditions. research agenda because of its very high yield potential and However, through the shuttle breeding program, the less vulnerability. It is an excellent option to increase and performance of new entries under delayed planting sustain the productivity of deepwater rice land. was assessed for the first time, along with normal planting, and this has been very successful. Three Collaboration with IRRI varieties were subsequently developed based on the work of the shuttle breeding network (Thakur et al The Rajendra Agricultural University has a very long partner- 1994,1998). ship with IRRI through shuttle breeding, farmers’ participa- l Vaidehi. This pureline selection from a local cultivar tory breeding, and deepwater rice breeding projects. Bihar is tall and has a thick and sturdy stem and dark green also participated in the IFAD-funded project on validation foliage. It is highly resistant to drought and tolerates and delivery of new technologies for flood-prone rice lands. water depths of up to 100 cm. Photoperiod-sensitive Through these projects, Pusa was selected as the main center and suitable for delayed planting, it is recommended for testing and distribution of improved deepwater rice in for growing in relatively more adverse conditions. eastern India and Bangladesh. l Satyam. This variety was developed from a cross between floating rice cultivar Desaria and an im- Progress in breeding (1991-2000) proved deepwater variety, RD19. It is semitall, with 140-d duration, and is highly adapted to delayed Significant achievements in breeding varieties for rainfed and transplanting. It tolerates water depths of up to 90 boro rice production systems were noted during this period. cm and is resistant to BPH, whitebacked planthop- The hybridization program was reorganized, and evaluation pers (WBPH), and bacterial leaf blight (BLB). The of new entries was modified to suit growing conditions. entry is exclusively tested in eastern India under the These efforts and adjustments resulted in the development shuttle breeding program for 3 years. of several popular varieties, replacing most of the traditional l Kishori. Developed from IR8 and Barogar (a deep- cultivars. Earlier, selection of segregating populations was water local cultivar), this variety underwent exten- made using nurseries grown under good management condi- sive testing through the shuttle breeding network in tions on research stations. Cultivars identified through this eastern India. It is semitall with 145-d maturity. It is approach were released as varieties with record yields, but highly adapted to delayed planting conditions and these were not adopted widely in the ecosystems for which is resistant to BPH, WBPH, and BLB. they were intended. This aspect was carefully considered
Breeding for rainfed lowland, deepwater, and boro land in Bihar, India: achievements and challenges 23 Bihar is also an active partner in the farmers’ partici- however, did not extend beyond this region, and the most patory breeding network. The project was mandated to seek important constraint was the extremely low temperature dur- farmers’ views and opinions when selecting varieties and ing winter. Varieties grown in the traditional boro region did fixed breeding lines that would be grown along with check not perform well elsewhere because of the low temperature. varieties popular in the target region. In the past, farmers’ If cold-tolerant varieties were made available, then it would views were seldom taken. This project practically docu- be possible to expand boro rice in chaur land to make this mented the criteria used by farmers in accepting or rejecting deepwater ecosystem more productive. a cultivar. Many times, a cultivar was rejected by breeders Research to develop cold-tolerant varieties for the boro on the basis of such criteria as yield, duration, and reaction season began in the early 1990s at Pusa (hot spot for cold to diseases and pests. However, farmers have their own (and screening), where a large number of improved cultivars, even different) opinions. For example, in spite of the release including those grown in traditional boro regions, were evalu- of several improved varieties for a specific ecology, farmers ated. One of the EMS mutants of Rasi yielded as high as 8.8 continue to rely on their own traditional varieties. A few years t ha–1. Named Gautam, it had an average yield of more than earlier, they participated in testing new varieties. Brown Gora, 7.5 t ha–1 in multilocation yield trials (Thakur et al 1994). a traditional cultivar for the uplands, is predominantly grown Gautam has dwarf habit, profuse tillering, dark green leaves, in Chotanagpur Plateau. Several improved varieties have been and high tolerance for low temperature at the seedling stage released for this system, but this cultivar remained popular and for high temperature at anthesis. Its grains are medium because some inherent qualities in it are lacking in the new fine. The success of Gautam in deepwater chaur land in- varieties. We came across an interesting situation in which tensified research efforts to develop varieties with different several improved cultivars were included in a demonstration durations and superior grain quality. Richharia, Dhanlaxmi, trial both in farmers’ fields and on the research station. Those and Saroj were developed in subsequent years (Thakur et al with poor ratings were rejected and new cultures were added. 2002a,b). A survey was then conducted to find out the composition l Richharia: Developed from IET7464/Pusa 33. It of varieties under rainfed lowland in the village where the is a semidwarf and an early-maturing variety with demonstration trial was held. Cultivar RAU1306 was found long grain. Released in 2000, it has excellent syn- to be grown on a large scale, replacing local landraces. This chronized flowering in the boro season. line was selected from the demonstration trial in farmers’ l Dhanlaxmi: Developed from ES1-2-3 (very early)/ fields; ironically, it was dropped from the research station IR36. It has medium duration during the boro season trials because of relatively poor yield. Subsequently, this and good grain quality. It is tolerant of Zn deficiency line was formally included in multilocation trials and was and resistant to bacterial leaf blight, brown plan- released as Santosh in 2001 (Thakur et al 2003). Santosh thopper, and bacterial smut. It was also released in was developed from a cross between Pankaj and Br8. It is 2000. tall, with 135−140-d duration, and phenotypically resembles l Saroj: Developed from a cross between Gautam and Mahshuri. It has very good grain and cooking quality. It type 3 (Basmati rice). It was basically developed for performs well under low-input conditions and is adapted to the kharif (wet) season but has now been widely late sowing and planting, traits not found in Mahshuri. adopted for the boro season. Its grains are long and slender. It was released in 2001. Deep water These varieties became popular in selected nontra- ditional and traditional boro areas in low-lying deepwater To date, no significant achievements in breeding for this chaur lands. However, there is still a demand for aromatic ecosystem have been made, in Bihar or elsewhere. Still, tra- boro varieties. A program started and several cultivars and ditional varieties predominate (Thakur 2004). Excess water breeding lines were evaluated during the boro season. One is a major impediment, where it sometimes remains in the of the breeding lines, RAU1397-19-3-7-9-4-2 derived from deepwater chaur lands up to March and April and sowing of IR36/type 3, has been identified. It has a yield potential of 7 dry-season wheat becomes impossible. An alternate technol- t ha–1 with long cylindrical grains and fine aroma. This will ogy is therefore critical. Cultivation of boro rice in such lands be the first aromatic variety for the boro season. is feasible from October-November to April-May. Currently, Vaidehi, which is released for the rainfed lowland, is also Identification of cold-tolerant breeding lines grown in these deepwater areas. Boro rice in West Bengal and Assam is grown on a large Cold-tolerant varieties are needed to expand boro rice area in scale during this period. In fact, it has long been grown in the vast waterlogged chaur lands in Bihar and Uttar Pradesh. some districts of Bihar near West Bengal. It first started in Normally, this is not required in traditional boro regions, but chaur lands, then moved on to irrigated medium lands. The in 2000-03, a severe cold during winter damaged most of the climatic conditions (temperature, soil type) in these Bihari boro seedlings, especially in Assam. This triggered a large- districts are similar to those in West Bengal and so this area scale screening of breeding lines developed locally through is considered a traditional boro region. Boro rice cultivation, hybridization and of those received from the Directorate of
24 Thakur et al Table 1. Cold-tolerant breeding lines for the boro season identified in 2002-03 at Pusa-Samastipur, Bihar.
Designation Pedigree Survival (%)
RAU1406-1-3-2-2-5 Turant Dhan/Rasi 90.4 RAU1406-2-1-4-2-3 Turant Dhan/Rasi 95.2 RAU1401-28 Rajshree/Gautam 92.0 RAU1400-5-7-CT Gautam/Type 3 96.2 RAU1395-2-7-3 Vaidehi/Sita 97.0 RAU1429-3-9-10-11 IR68888A/Saroj 97.0 IET17257 HKR 239/Basmati 370 98.0 IET16958 IR38699/IR 19431 92.5 IET17296 Indrajani/SYE 3-43-57 98.5 IET17243 – 99.0 IET17256 HKR239/Basmati 370 96.0 IET17457 CO 1 064-5/IR9129-320-3-3-3 96.5 IET17269 IR54 95.6 IET17519 IR38699/IR19431 98.3 IET16935 JLD4/ADT39 98.4
Rice Research, Hyderabad. About 80 entries were tested Vaidehi was initially recommended for the rainfed low- during 2001-02 and 210 entries were evaluated in 2002-03. land, but it was also found suitable for the mixed-cropping Thirty entries were identified as highly cold-tolerant. The 15 scheme with mungbean/sesame in medium deepwater areas. best entries are listed in Table 1. This variety has an erect growth habit and is not affected when grown with nonrice crops. Mixed cropping, practiced Evaluation of screening methods during March-April, currently covers areas in Samastipur, Darbhanga, and Muzaffarpur districts. Satyam has become The evaluation of new improved germplasm was done popular in BPH-endemic regions, especially in south Bihar. through an on-farm research trial. It comprised 8-10 entries It covers more areas in Madhubani and Darbhanga districts. that differ in height, duration, and grain quality, along with Kishori, because of its bold grains, is getting popular in areas locally adopted checks. The trials were conducted jointly by where special rice is in demand. Santosh has spread in all farmers and scientists at several sites following farmers’ prac- zones and has a wide coverage in Zone I in north Bihar. This tices. The same set of lines was also grown at the research variety performs very well under low fertilizer input and farm. The good performers across all trials were then included adverse conditions, particularly in the irrigated Zone IIIB. in the on-farm trials using two to three entries per farmer During 2002-03, the monsoon season was delayed, and the and a large number of sites (10−15). The data from all sites majority of rice varieties were planted late. This resulted in were compiled and analyzed; the entries that spread across nonsynchronized flowering of most varieties such as Nata the whole village were selected and considered accepted by Mahshuri and Sita but not Santosh, which made this variety farmers. This method was developed during the evaluation more popular in these areas. of Santosh; the same protocols were applied for testing and The most visible impact of new varieties was seen with release of Saroj and Prabhat for the boro season and Satyam the new boro lines. When Gautam was released, it spread so and Kishori for the wet season in the rainfed lowland. fast that, in the first year of its release, it covered about 100 ha of chaur land in Zone II in Madhubani District. It also became Progress in deploying new varieties popular in the traditional boro region of Saharsa District. Within just a few years, all new boro areas were covered by The Agriculture Department in Bihar is well established to this variety alone. However, with large-scale cultivation, this disseminate new technologies developed by the university variety started to show high susceptibility to BPH, suggest- to farmers. Joint meetings of both agencies are periodically ing the need for a new resistant variety. This variety is also held and the seed production program was set on the basis late maturing; the water stored in chaur land was sometimes of target demand. Seed production of varieties favored by not sufficient to irrigate larger areas. Richharia, Dhanlaxmi, farmers is done on a limited scale as these varieties normally and Prabhat (varieties recommended for the upland) started spread relatively faster among farmers. Some difficulties are to cover wider areas in traditional as well as nontraditional encountered in disseminating the new varieties recommended boro areas. At present, Prabhat is most popular in Madhubani for adverse conditions; it takes a long time for these varie- District and has replaced Gautam. ties to become accepted. In a situation like this, university The government of Bihar launched a project to popular- and Agriculture Department staff discuss ways to enhance ize boro cultivation in the nontraditional boro regions, with delivery mechanisms to end users. the hope that a wide coverage of these areas will lead to
Breeding for rainfed lowland, deepwater, and boro land in Bihar, India: achievements and challenges 25 enhanced productivity of deepwater land. Selected varieties season as well as in upland areas during the wet season, not from Bihar, West Bengal, Assam, and Uttar Pradesh were only in Bihar but also in other states. jointly introduced in on-farm trials over the past several Thus, high yield as the primary target in breeding years. Gautam in West Bengal, Prabhat in Uttar Pradesh, and may not be sufficient; other features such as good grain Saroj in Bihar and Assam were recognized as the top yield- quality and attractiveness should be considered. For yield ers (Chaudary et al 2003). Saroj performed exceedingly well to stabilize at higher levels, new varieties must be tolerant in the traditional regions of Bihar and was highly preferred of common stresses associated with the ecosystem—pests, for its grain quality. This variety also became popular in the diseases, drought, and submergence. Several stresses are as- irrigated midland plains of south Bihar. sociated with specific unfavorable ecosystems and regions as evidenced by the popularity of specific cultivars in a par- Breeding objectives and challenges ticular region. Since local cultivars are usually grown under in the next 5–10 years adverse conditions, the new varieties must have such distinct advantages as higher grain yield, but they at least must retain There is now a subtle change in farmers’ attitude toward new similar grain and cooking qualities. varieties and yield is no longer the sole criterion that ensures Several varieties need to be developed to account for adoption of a particular variety. Besides yield, grain quality the wide range of variability in these ecosystems. This great receives prime consideration. An example can be cited to elu- challenge for breeders could be overcome if conventional cidate the point. We developed a very early-maturing variety breeding approaches were supplemented by biotechnology (75 d total duration), Turanta Dhan, for upland conditions tools. Yield and quality do not seem to be well related, and and also for double rice cropping. It has a yield potential of there might be some barriers for their combination. This 4 t ha–1. When we asked a resource-poor farmer who had problem could probably be resolved through molecular breed- successfully grown this variety, he said he did not retain its ing. That this could be achievable is reflected in the efforts seeds (he sold all of them in the market) in spite of its high to develop Basmati rice for these ecosystems. Eventually, yield. His reason is that it had very poor eating quality. Even a there is a need to devise special and novel breeding strategies farmer of this category now needs better quality rice. Though for the adverse rainfed lowlands and deepwater ecologies as good grain quality is preferred, greater attention should be breeding endeavors have so far not been very successful. As given to the cooking and eating qualities. such, it is better to use local cultivars in hybridization and Disease-resistant varieties are needed to replace cur- improve their deficiencies. For deepwater areas, boro rice is rent susceptible ones. The university recently released a probably an excellent choice for Bihar and Uttar Pradesh, BPH-resistant, high-yielding variety, Kanak, for the irrigated but varieties that are highly tolerant of cold stress are also midland and favorable lowland areas. Its yield in frontline needed. demonstration trials in a 8-ha field was 7.5 t ha–1. A serious extension campaign was carried out to disseminate it; the aim References was to replace susceptible variety Sita. These efforts failed, however, and Sita remained popular. The main reason again Catling HD. 1992. Rice in deepwater. Manila (Philippines): Interna- was Kanak’s poor grain and cooking quality. tional Rice Research Institute. 542 p. While deciding on the objectives of a breeding program, Chatterjee SD. 1995. Status of boro rice in West Bengal. In: Report of one emerging aspect is the focus on phenotypic appearance of an INGER boro rice monitoring visit to India, Bangladesh and the genotype, which seems to help in the spread of a variety. Myanmar. Manila (Philippines): International Rice Research Sita was released in Bihar in the early 1970s and became very Institute. popular in the irrigated areas. However, it was later found to Chaudhary VK, Thakur R, Dwivedi J, Maity S, Deo Chaudhary PK, be highly susceptible to BLB. A BLB-resistant, high-yield- Singh NK, Nillanjay. 2003. On-farm evaluation of improved ing variety, Rajendra Dhan, was subsequently released to boro rice germplasm in eastern India. In: Singh RK, Hossain replace Sita, and it outyielded Sita in yield trials. Again, the M, Thakur R, editors.. Boro rice. IRRI India Office, NASC, variety was not accepted mainly because its look was not Pusa New Delhi. p 221-232. that appealing. Leaves were erect, virtually covering the Morshimo-Okino H. 1964. Report of observations on floating rice in panicles. Besides, its grains were not as long as Sita’s. Sita, Thailand. 9 p. (unpubl.) in contrast, has an attractive look; erect, dark green leaves; Pathak AK. 1995. Problems and prospects of boro rice in Assam. In: and long drooping panicles. When this variety is grown by Report of an INGER boro rice monitoring visit to India, Bang- a farmer in a village, other farmers are attracted to grow it ladesh, and Myanmar. Manila (Philippines): International Rice solely because of its attractive look. This variety has been Research Institute. cultivated in the last two decades. Another example is upland Singh RK, Thakur R, Chatterjee SD. 2003. Harnessing boro rice poten- variety Prabhat. At the dough stage, all leaves of this variety tial for increasing rice production in deepwater areas of Eastern are droopy, except for the flag leaf, and only the grains are India. In: Singh RK, Hossain M, Thakur R, editors. Boro rice. visible. This look has helped it gain popularity in the boro IRRI India Office, NASC, Pusa, New Delhi. p 51-72.
26 Thakur et al Thakur R. 2004. Genetic improvement of flood-prone rice: Where are Thakur R, Singh AK, Singh RS, Singh NK, Mishra SB, Singh UK, we today and what are the future prospects? In: Bhuiyan SI, Chaudhary VK. 2002b. Saroj an early maturing variety released Abedin MZ, Singh VP, Hardy B, editors. Manila (Philippines): in Bihar, India. Int. Rice Res. Newsl. 27(2):18. International Rice Research Institute. Thakur R, Singh NK, Chaudhary VK. 2003. Recent advances of boro Thakur R. 1995. Status of boro rice in Bihar, India. In: Report of an rice research in Bihar. In: Singh RK, Hossain M, Thakur R, edi- INGER boro rice monitoring visit to India, Bangladesh, and tors. Boro rice. IRRI India Office, NASC, Pusa New Delhi. Myanmar. Manila (Philippines): International Rice Research Thakur R, Singh NK, Mishra SB,. Singh AK, Singh KK, Singh RK. Institute. 2003. Santosh: a high-yielding variety for rainfed lowland Thakur R. Sahi SP, Mishra SB, Singh UK, Mishra M, Rai JN. 1994. developed through participatory breeding for Bihar, India. Int. Gautam: an improved rice variety for winter (boro) season in Rice Res. Notes 28(2):25-26. Bihar, India. Int. Rice Res. Newsl. 19(2):19. Thakur R, Singh AK, Singh RS, Mishra SB, Singh NK, Rai JN. 1998. Notes Satyam and Kishori, two high-yielding varieties developed fro the rainfed lowlands of Bihar, India. Int. Rice Res. Notes Authors’ address: Department of Plant Breeding, Rajendra Agricultural 23(3):20-21. University, Pusa, Samastipur 848125, India. Thakur R, Singh AK, Singh RS, Singh NK, Mishra SB, Mishra M, Singh UK, Rai JN. 2002a. Dhanlaxmi and Richharia, very early varieties released in Bihar, India. Int. Rice Res. Newsl. 27(1):31-32.
Breeding for rainfed lowland, deepwater, and boro land in Bihar, India: achievements and challenges 27 Breeding rice for deepwater and flood-prone areas of Thailand
Wilailak Sommut, Kalaya Kupkanchanakul, Prayote Charoendham, Udompan Promnart, and Suthep Nuchsawasdi
Rainfed rice in Thailand accounted for approximately 88% of the total cultivated rice area and 79% of total production in 2004. Flood-prone rice (FPR), one of the rainfed rice ecosystems, accounts for 4.7% of total cultivated area, 5.4% of rainfed rice area, and 1.4 million tons of rice production annually. The cultivation system has changed somewhat from monocropping of FPR varieties in the wet season to double cropping, primarily due to the expansion of irrigation infrastructure. However, monocropping of FPR still predominates. Breeding for FPR in Thailand was strengthened with the establishment of collaboration with the International Rice Re- search Institute (IRRI) during 1991 to 2004. With the Prachinburi Rice Research Center (PRRC) and Huntra Rice Experiment Station (HTA) as key sites, collaborative activities involved exchange of germplasm and capacity building through degree and nondegree training and workshops. Breeding strategies followed included enhancing genetic variation by effective collection, evaluation, and use of rice germplasm, hybridization and selection, screening for biotic and abiotic stresses, the use of rapid generation advance (RGA) to shorten progeny selection time, screening of F2-F3 populations under natural flooding in farm- ers’ fields prior to replicated yield testing, and farmer participation in multilocation yield trials. More than 1,000 crosses were made and hundreds of breeding lines were tested by collaborators. Some lines were elevated to replicated yield trials and a number have been tested under acid-sulfate soil conditions, with a few lines in the process of being released. Three Thai varieties have also been released in Cambodia and five varieties, including two new plant type FPR, were released in Thailand. Training activities and workshops were held at PRRC and three scientists received doctoral degrees from the University of the Philippines Los Baños under IRRI’s sponsorship.
Thailand lies between 5o and 21oN latitude and between 97o try’s most important export crop. Administratively as well as and 106oE longitude. This peninsular Southeast Asian country geographically, Thailand is divided into four regions: central, shares boundaries with Myanmar in the west, Lao PDR and north, northeast, and south. Each region has different rice- Cambodia in the northeast, and Malaysia in the south. The growing environments and yield potential (Table 1). South China Sea touches the east coast, while the Indian Almost one-third of the land area of Thailand is located Ocean and Andaman Sea border the west coast. Thailand has in the northeast region, which has more than one-half of the 51 million ha of land area, of which one-third is cultivated rice land but accounts for only 37% of total rice production. with annual crops and about 7% is under cultivation with The average rice farm size is smaller than in other regions perennial crops. and soil erosion and drought during the dry season are acute Thailand belongs to the warm subhumid tropics. Four problems. seasons are recognized: southwest monsoon from May The central region, an intensively cultivated alluvial through September, a transition period from the southwest area, accounts for about one-fifth of the total cultivated rice to the northeast monsoon during October, the northeast mon- land. Average farm size is large, and most farms have irriga- soon from November through February, and a premonsoon tion facilities, thus allowing two rice crops per year. Almost hot season during March and April. Temperatures in the 75% of the dry-season rice grown under irrigated conditions Central Plain during the rainy season (May to November) is located in this region, which produces up to 31% of total average 27 ºC, with only an 8–10 ºC difference between the production. Farm operations are almost entirely mechanized daily minimum and maximum temperatures. There is a brief and farmers use the direct-seeding method of crop establish- cool period (December and January) when temperatures are ment to save labor. as low as 2–3 ºC in the northern highlands. The northern region has almost one-third of the land area of Thailand. Upland rice is grown in the lower altitudes Rice environments and production of high hills and in upland areas. Lowland rice is grown mainly in lower valleys and on some terraced fields where Rice is not only a traditional staple food but is also linked water is available. This region has about 24% of the total rice with life and culture in Thailand, aside from being the coun- land and accounts for 28% of total production.
28 Sommut et al Table 1. Rice area, production, and yield by region in Thailand, 2003-04.
Planted area Production Yielda
Region (million ha) Total (million tons) (t ha–1)
Major rice Second Major rice Second Major rice Second rice rice rice
North 2.30 0.46 4.96 1.85 2.64 4.09 Northeast 6.04 0.10 8.54 0.30 1.72 3.14 Central 1.67 0.66 4.70 2.89 3.04 4.60 South 0.50 0.04 0.85 0.11 2.13 2.90 Total 10.51 1.26 19.05 5.15 2.38 4.24
aBased on harvested area Source: Center for Agricultural Statistics (2004).
The southern region constitutes about 14% of the to- Table 2. Flood-prone rice areas in Thailand based on a tal area of the country. The region has only 5% of the total water depth, crop year 1992-93. rice area and accounts for only 4% of total rice production, Cultivated area (ha) mainly for local consumption. Among the four regions, the Province Total Water depth Water depth north and central are more productive in terms of rice yield <100 cm >100 cm during both the major and second rice crops. Thailand has two main rice ecosystems—rainfed and Ayutthaya 33,328 64,123 97,451 irrigated. The rainfed ecosystem can be further divided into Nakorn Sawan 59,527 11,593 71,120 Pichit 52,136 13,859 65,995 four subecosystems: rainfed lowland accounting for 80% Nakorn Nayok 17,686 25,875 43,561 and distributed mainly in the northeast and north, deepwater, Ang Thong 12,326 29,037 41,363 floating (also called flood-prone), and upland. This paper Lop Buri 12,793 21,325 34,118 covers the flood-prone rice (FPR) ecosystem, which includes Pitsanulok 21,049 8,598 29,647 floating rice (FR) and deepwater rice (DWR). Prachin Buri 5,796 16,248 22,039 Singh Buri 9,856 8,917 18,773 Ratcha Buri 9,989 3,943 13,932 Importance of flood-prone rice Suphan Buri 7,778 2,456 10,234 Sara Buri 4,336 5,278 9,614 The large proportion of FPR is mostly grown along the Chao Chainat 4,114 1,917 6,031 Phraya River in the Central Plain of Thailand, the most pro- Nakorn Pathom 3,403 1,233 4,636 Uthai Thani 3,369 383 3,752 ductive rice area (Table 2). The FPR area declined from 0.85 Petcha Buri 1,661 1,001 2,662 million hectares in 1982 to 0.50 million hectares in 1992, Nongkai 492 3,072 3,564 approximately 5.8% of the cultivated rice area (Charoendham Khon Kaen 327 2,846 3,173 et al 1995). However, it has also been reported that 21.9% of Chaiyaphoom 2,887 18,986 21,873 Thailand’s total rice area is devoted to FPR (Huke and Huke Udorn Thani 10 622 632 1997). Nevertheless, FPR provides around one million tons Total 262,863 241,306 504,169 of rice each year. Two types of FPR rice are grown: the deep- aEstimated area = no. of families × average cultivated area/ water rice (DWR) type that grows at average water depths family. of 50–100 cm for up to one month or more and the floating Source: Charoendham et al (1995). rice (FR) type that is grown at a water depth of 1–5 m with a flooding period of 3–4 months. Submergence tolerance is the most essential characteristic for DWR while good elongating time of recession of the flood is essential for the selection of and kneeing ability are desirable for FR. appropriate FPR varieties. FPR has been grown as a monocrop in the past using traditional varieties that, while popular, yield only 1.2–1.8 t Special characteristics of FPR ha–1. Since 2000, however, double cropping has been prac- ticed using modern varieties (Table 3; Sommut et al 2004). Many types of FPR can be found in a region and different Short-duration pre- or postflood rice crops have also replaced maximum water depths require distinct FPR types. Farmers FPR in some areas of the Central Plain of Thailand (Table 3). may use several FPR varieties where water depths change Classification of areas by time of arrival of floodwater, rate with topography over a short distance: slowly elongating (2–3 of rise of water level, range of maximum water depth, and cm day–1) varieties for 50–80-cm water depth and fast-elon-
Breeding rice for deepwater and flood-prone areas of Thailand 29 Table 3. Current cropping patterns (% of land) in the flood-prone ecosystem of Thailand, 2000-01. to the soil; hence, the common name “floating rice.” Knee- ing keeps the canopy and panicles erect and above the water Cropping patterna Lowland Lowland Deepwater Deepwater level. It maintains grain quality by preventing submergence of rainfed irrigated rainfed irrigated panicles in water and protects grain from damage by aquatic Permanent fallow 1.0 3.0 0.7 0.3 fauna (Vergara et al 1977, Vergara 1985). Traditional FPR Upland crops-fallow 0.0 2.7 1.5 0.8 cultivars have good kneeing ability but some modern varie- TV rice-fallow 55.4 13.8 58.2 42.8 ties lack this trait. ITV rice-fallow 25.3 7.0 22.9 8.4 MV rice-fallow 13.2 46.3 8.5 16.1 Submergence tolerance MV-TV rice 0.0 4.2 4.9 12.8 MV rice-MV rice 5.1 20.8 0.7 15.2 Submergence tolerance and elongation ability are distinct Other crop-rice 0.0 0.2 8.3 16.4 plant responses that represent opposite mechanisms for flood Total 100.0 100.0 100.0 100.0 adaptation. Attempts to combine these two mechanisms in Cropping intensity 105.1 127.4 105.8 131.9 a single rice variety have failed. DWR may be completely aTV = traditional rice varieties, ITV = improved traditional rice varieties through submerged if floodwaters rise rapidly, but elongation is usu- pure-line selection and/or mutation breeding, MV = modern high-yielding ally sufficient to raise part of the foliage above the water varieties. level. Photosynthesis can then continue and starch and sugars Source: Sommut et al (2004). in emergent leaves and in plant parts under the water are maintained (Setter et al 1987) and further elongation can take place. Submergence tolerance is most useful for nonelongat- gating (15–20 cm day–1) varieties for water deeper than 150 ing rice that may be submerged for a few days by rapidly cm. The rice in the first category has the greatest potential to increasing and then decreasing water levels. increase yield. FPR has three special adaptation mechanisms: Puckridge et al (1994) surveyed 87 farmers’ fields in (1) elongation of stems and leaves; (2) kneeing ability, which the Central Plain in 1988-89 and found 37 DWR varieties is the upward bending of the terminal parts of the plant; and grown. Yield ranged from 0.65 to 4.87 t ha–1, with a mean of (3) the ability to develop nodal tillers and roots from upper 2.13 t ha–1. This compared well with the estimated 2.0 t ha–1 nodes in the water. Kneeing keeps the reproductive parts average yield for DWR (Catling et al 1982). In another survey above water as the floods subside. The first two traits are conducted during the 1992-93 cropping season involving 889 accentuated in floating rice such as Leb Mue Nahng 111, Pin farmers from 184 villages in 20 provinces with major DWR Gaew 56, Plai Ngahm, Prachinburi, Leuang Yai, Khao Luang, areas in north, central, and northeast Thailand, 86 varieties and Tewada. Nodal tillers arising during the flood period can were recorded (Charoendham et al 1995). Only 18% of the sometimes compensate for sparse stands. farmers grew the five recommended varieties. Local varie- ties were grown by 75% of the farmers, who considered the Elongation ability recommended varieties to be unsuitable to their areas and to The ability to elongate is an escape mechanism for survival have uncertain or low yields of 2.2 t ha–1. Farmers sold 80% from partial or total submergence. Total plant elongation, of their produce, kept 8% for seed, and used 8% and 4% for including an increase in lengths of leaf blades, leaf sheaths, family consumption and land rent, respectively. The most and stems, may be as much as 20–25 cm in 24 h during initial desired FPR characteristics were high yield, good elongation, flooding (Choudhary and Zaman 1970). The stems produce and drought tolerance. The varieties grown matured from new nodes where leaves are attached and the internode early December to mid-January and were 150–200 cm tall, can increase in length when submerged. Elongation of leaf had high tillering, long panicles with good exsertion, good blades and leaf sheaths is important for seedling survival kneeing ability, medium to nonshattering panicles, drought but internode elongation is the most important mechanism and submergence tolerance, long-slender grain with light for increasing plant length in very deep water (Vergara et al hull color, and aroma and good cooking quality, and were 1975). Internode elongation commenced between 4 and 6 nonlodging. Some 98% of the farmers kept their own seeds, weeks of age for most of 100 lines of FPR growing in rising with only 7% doing seed multiplication. More recently, Som- water (Puckridge et al 1990). The elongation induced after mut et al (2004) found 120 rice varieties grown in the wet panicle initiation affects the final 4–5 internodes and is not season, but only 18 in the dry season. influenced by water depth (Morishima 1975, Bekhasut et al 1990). Elongation stops after flowering in all varieties. Major constraints of FPR Kneeing ability The major constraints for FPR cultivation are drought, flood- Kneeing is the bending upward of the upper parts of the culms ing, problem soils, and diseases and insects. Drought is the (stems) as water levels fall and rice plants lodge during the major problem in areas where rainfall is bimodal, with a dry recession of floodwater. When culms lie on the water surface spell occurring between the two peaks (Figs. 1, 2). Damage and the upper leaves are held vertical above the water, the from drought occurs mainly at the seedling stage. Although plants appear to float even though their base is still attached FPR can tolerate flooding, either through the mechanism
30 Sommut et al Rainfall (mm) 1994 Water depth (mm) Rainfall (mm) 1995 Water depth (mm) 150 250 150 250 120 Rainfall Flowering 200 120 200 Water depth Flowering 90 Sowing 150 90 150 60 100 60 100 30 50 30 50 0 0 0 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Rainfall (mm) 1996 Water depth (mm) Rainfall (mm) 1997 Water depth (mm) 150 250 150 250 120 200 120 Flowering 200 90 150 90 Sowing 150 60 100 60 Sowing 100 30 50 30 50 0 0 0 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Rainfall (mm) 1998 Water depth (mm) Rainfall (mm) 1999 Water depth (mm) 150 250 150 250 120 Rainfall Flowering 200 120 Flowering 200 Water depth 90 90 150 Sowing 150 Sowing 60 100 60 100 30 50 30 50 0 0 0 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Rainfall (mm) 2000 Water depth (mm) Rainfall (mm) 2001 Water depth (mm) 150 250 150 250 120 200 120 200 Sowing Flowering 90 Sowing 150 90 150 60 100 60 100 30 50 30 50 0 0 0 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Fig. 1. Daily rainfall distribution and water depth at Prachinburi Rice Research Center, 1994-2001. of submergence tolerance or elongating ability, it can oc- Cambodia, Myanmar, and Indonesia, and other varieties have casionally be damaged if flash floods occur at the seedling been exchanged between countries. Many new FPR varieties stage. The minimum growth period of FPR must be 30–45 have performed better than the local varieties in Cambodia, days after emergence. Somehow, prolonged flooding at ma- Vietnam, Myanmar, Indonesia, and Africa, and emphasis on turity may cause damage. Acid-sulfate soils are a problem testing in the appropriate environments should speed up the in FPR areas mostly in the central-east areas of the country, development of better yielding varieties. especially at the early growth stage. The combined effect of drought and acid-sulfate soils makes rice yields in this area Collaborative project on FPR breeding very low. Important diseases for FPR are blast (caused by Pyricularia grisea) occuring at the seedling stage as well as Varietal development bacterial blight (caused by Xanthomonas oryzae) occurring Varietal improvement of FPR in Thailand dates back to 1916 at the flowering stage. Insect pests such as stem borers and when the first rice experiment station, Klong Rangsit Rice brown planthopper are also constraints. Experimental Farm, was established. The first step of rice Other considerations under the FPR ecosystem include improvement emphasized head selections for grain quality the lack of adapted varieties since FPR varieties are usually and yield in a large number of collected local varieties. Rice location-specific and there is a lack of an efficient extension breeding continued until a formal linkage between Thailand system to disseminate existing technologies, an insufficiency and the International Rice Research Institute (IRRI) in 1960 of good seed of new varieties and high price of good seeds, a that was further emphasized through the Thai-IRRI Deepwa- lack of attractive rice policies to encourage farmers to grow ter Rice Collaborative Project on varietal improvement and FPR, and a lack of young FPR scientists, especially breeders, training started in 1986. During that period, up to 1990, three to work on FPR. FPR varieties—RD 17, RD 19, and Huntra 60—suitable for Despite these constraints, the breeding programs in 50–100-cm water depth and incorporating some desirable Thailand have many FPR lines under development, with traits from IRRI lines and Thai traditional varieties were re- different maturities and elongation ability, and with resist- leased. Unfortunately, these varieties were not widely adopted ance to pests and diseases. Breeders working in collaboration by farmers due to the photoperiod insensitivity of RD 17 and with IRRI have released new varieties in Thailand, India, high chalky grain of RD 19 and Huntra 60. However, RD
Breeding rice for deepwater and flood-prone areas of Thailand 31 Variation among years Variation among locations Water depth (cm) 1996 Water depth (cm) 1998 300 300 250 Prachantakam Flowering time 250 Kabinburi Flowering time 200 200 150 150 100 100 50 50
0 0
l
t
t
t
t t
v
t
v
v
v
c
c
p
p
p
g
p
g
ov
ov
ug
N
No
N
A
De
Se
1 Oc 1
7 Oc 7
7 Oc 7
4 No 4
4 No 4
5 No 5
2
20 Ju
9 Se 9
2 De 2
3 Au
2
21 Oc
28 Oc 14 Oc
11
2
18
18
17 Au
23
31
30 Se 30 23 Se 23
1997 1998 300 300 250 Prachantakam Flowering time 250 Banna Flowering time 200 200 150 150 100 100 50 50
0 0
l
l
t
t
t
t
v
v
c
v
v
p
g
p
n
p
p
p
g
g
g
un
Oc
Ju J
Oc
Se No
Au
Se
No
Ju
De
Se
8 Ju
5
1 No 1
5 5
7 No 7
2
6 6
5 Au
22
10 Oc
28
10
24 Oc
14
24
15 Au
2
30 2
19
26 Se 26
1
12 Se 12
29 Au 11
1998 1998 300 300 250 Prachantakam Flowering time 250 Inburi 200 200 Flowering time 150 150 100 100 50 50
0 0
t
t
t
t
t
t
t
t
t
v
v
v
v
v
v
v
p
v
p
p
p
p
p
g
p
g
p
g
e
S
Oc
Oc
Se
No
Se
1 Oc 1
2 Oc 2
4
9 Oc 9
8 Oc 8
4 No 4
4 Se 4
0 Au
3 Se 3
1
5 No 5 3 Au
8 No 8
6
7 No 7
8 No 8
25
1
30 Oc
23 Oc
16 Oc
1
1
11 Se 11
21 No 21
20
3
1
18 Se 18
2 2
2
15 No 15
27
25 Se 25 28 Au
1999 1998 300 Flowering time 300 250 Prachantakam 250 Nakhonnayok 200 200 Flowering time 150 150 100 100
50 50
l l
l 0
l
t
t
t
t
v
v n
0 p
p
p
p
n
g
g
g
g
ov
ov
Oc
N
Se
N
No
Au
Au
7 Oc 7
6 Oc 6
9 Ju 9
29 Ju
8 Se 8
3
15 Ju 15
4
9
7
28 Ju
14 Ju
30 Ju
20 Oc
21
18 No 18
25 Au
23 Se 23
22 Se 22
1
26
11 Au 12
2000 1998 300 300 Flowering time 250 Prachantakam Flowering time 250 Prachantakam 200 200 150 150 100 100 50 50
0
t
t
t
v
v
p
g
t
g
p
t
t t
0 t
v
v
p
p
p
p
g
ec
ov
ov
e
D
Oc
Oc
Oc
Oc
Oc
N
No
No
N
Se
Se
No
S
Au
Oc
Oc
No
Se
Se Au
2 Oc 2
7 Au 7
9 9
4 Se 4
2 2
0
4 4
7 7
30
16
27
13
18
11
21
16
23
3
28
14
21
18
11 28 25
Fig. 2. Distribution of flooding patterns in farmers’ fields of flood-prone rice ecosystem of Thailand during 1996-2000.
19 is the most appropriate prototype for a high-yielding FPR at Prachinburi (PRRC) was made responsible for improving cultivar and is still being used in our breeding program. FPR in the Central Plain of Thailand. PRRC is well staffed The FPR collaborative project was strengthened further and equipped, with a satellite station at Huntra (HTA) that after the signing of a memorandum of understanding (MOU) has other screening facilities and a different flooding pattern between the Ministry of Agriculture and Cooperatives and from that of PRRC. A summary of key elements and activities IRRI on the occasion of the visit to IRRI of Her Royal High- undertaken under the collaboration is given in Figure 3 and ness Princess Maha Chakri Sirindhorn on 29 August 1991. Table 4. It called for Thai and IRRI scientists to work together in IRRI conducted prebreeding research and provided improving FPR by developing new rice varieties and farming assistance to collaborating NARES. Prebreeding research practices for farmers in Thailand and other countries of the focused on submergence tolerance and elongation under region. As a result, a major part of the IRRI FPR breeding pro- slow- and fast-rising water. This included development/re- gram was transferred to Thailand in 1992. IRRI and Thailand finement of screening methods, providing donors, studying jointly developed and distributed plant breeding materials the genetics and mechanisms of tolerance, and mapping and associated technologies for improving production in FPR and tagging of genes. Sturdy stems were deemed essential areas of South and Southeast Asia. The Rice Research Center to FPR, particularly for medium water depths. IRRI aimed
32 Sommut et al FPR ecosystem Research on flood and soil constraints to rice production
Prebreeding research at IRRI and other laboratories Zn and P Al toxicity Stagnant flood Fe toxicity Submergence Salinity efficiency tolerance tolerance tolerance tolerance tolerance
Germplasm improvement research Thailand and India
Exchange breeding materials with NARES
Irrigated rice
In-country testing (NARES)
Socioeconomic potential
Deepwater rice Boro rice Coastal rice
Integration Soil and water management
Rainfed lowland rice Nutrient and carbohydrate research
Fig. 3. Key elements in the research program for Thailand’s FPR and linkages to the ir- rigated and rainfed lowland rice ecosystems.
at incorporating sturdy stems of some rainfed lowland and Training, workshops, and visits upland types into FPR. In order to provide knowledge on FPR breeding and to gain PRRC intensified its evaluation and selection of breed- useful experience for improving rice yield, training courses ing materials. IRRI assisted in dry-season rapid generation were held regularly from 1993 to 1997 at PRRC. The training advance, characterization of elite lines, multilocation testing courses had two main parts, one during crop establishment of elite lines through the International Network for the Ge- and the other during the maturity-harvesting period. For each netic Evaluation of Rice (INGER), and producing F1s that part, both lectures from experts and practical work on-sta- were difficult to handle at PRRC. Though flooding patterns tion and in farmers’ fields were conducted. Each part lasted were similar, soil and climatic (temperature and daylength) about 3 weeks. Before conducting the FPR training course, differences between FPR areas of South and Southeast Asian PRRC staff attended a training course on Basic Training regions posed a selection problem at PRRC. To alleviate Skills Development provided by IRRI and held at PRRC in this, F2 populations evaluated in Thailand were selected for March 1993. A training course titled Principles of Research in advancement at IRRI. Deepwater and Other FPR Ecosystems that was attended by Rigid selection is practiced for grain quality traits in scientists from 11 countries of South and Southeast Asia and Thailand but this procedure tends to discard very high yield- Africa was also held in May-June 1993. Until 1997, training ing genotypes. Selection for grain quality was thus relaxed courses and annual meetings were held among breeders who with two objectives: (1) to isolate lines that could demonstrate came to PRRC and HTA to select FPR populations suitable the yield potential of FPR and (2) to distribute materials to for their own countries as well as to present results from their other countries where grain quality standards were not so own work. Two rice breeders and a physiologist from PRRC high. The overlap between the rainfed lowland and FPR were also able to pursue Ph.D. degrees at IRRI. Two Thai rice breeding programs at IRRI for the two ecosystems at water breeders also visited Vietnam, Cambodia, Myanmar, India, depths of around 50 cm was recognized and addressed. and Bangladesh to evaluate and select breeding lines that were sent from Thailand. Besides high yield, maturity was
Breeding rice for deepwater and flood-prone areas of Thailand 33 Table 4. FPR breeding: collaborative activities during 1992-97.
Responsibilitya Activity Thailand India IRRI
1. Submergence and elongation (fast & slow) a) Screening methods (lab/greenhouse) b) Screening (for donors) — lab/greenhouse — field +++ c) Donor characterization ++ d) Donor distribution e) Genetic mechanisms f) Gene tagging 2. Hybridization +++ +++ ++ 3. Evaluation a) Early bulks +++ +++ b) RGA +++ c) Adv = advanced pedigree +++ +++ d) Observation nurseries +++ +++ 4. Farm field tests +++ +++ 5. Elite line characterization (pests, diseases, +++ submergence, elongation, grain quality) 6. Multilocation national tests +++ +++ 7. INGER evaluation +++ 8. Seed increase +++ +++ 9. Release +++ +++
a+++ = high, ++ = complementary.
the most essential character for selection. Ranges of flowering Table 5. List of promising lines from yield trials in Day time from 1-30 November and 1-10 December were suitable Eth, Cambodia, and Vietnam, 1996. for Indochina, and Assam, India, whereas the early-flowering Designation Height (cm) Flowering Yield group (October) was good for Uttar Pradesh, India. date (t ha–1)
Exchange of germplasm Day Eth, Cambodia HTAFR84021-7-1-3 170 12-XII 3.0 Though many breeding materials from the collaborative HTAFR87050-1-9 156 7-XII 3.3 project have been entered in national testing programs, re- Mali Tawng 154 5-XII 3.0 leases of FPR varieties have been few. Some were tested for PCR87006-8-7 154 5-XII 2.6 more than 15 years before being released because (1) FPR PCRFR84007-43-B 155 6-XII 2.6 plants require a long growing period of 6–8 months, of which PCR89118-8B 154 5-XII 2.7 Vietnam 2–3 months must be before flooding; (2) different flooding DWCT82-B-107-7 170 4-XII 3.3 regimes require different times of maturity in each location; HTAFR84035-B-18-7 167 19-XI 2.5 (3) FPR varieties fall into two maturity groups according to HTAFR94096-B-4-8 171 10-XII 3.0 latitude (Bangladesh and India 23–28oN; Myanmar, Cambo- HTAFR85035-B-20-5 167 10-XII 3.3 dia, Thailand, and Vietnam 9–17oN) and the transfer of rice varieties between zones needs a long period of selection for maturity; (4) photoperiod sensitivity ensures flowering at the right time with respect to flooding, but only one generation 13 leading DWR and RLR lines/varieties from Vietnam were can be harvested each year; and (5) field tests in the target introduced and used in the Thai DWR breeding program; (5) environments have been very limited. lines possessing acid-sulfate tolerance from Vietnam were Despite constraints, in 1996, the program exchanged used in the acid-sulfate screening and breeding program of germplasm as follows: (1) early generations of F2–F4 popula- Thailand: Ca Dung Phen, Lua Thong, Nang Coi, Soc Nan, and tions (84) were sent to Ghagharahghat, India; (2) fixed and Trang Hon Bihn; (6) 1,000 early-generation breeding lines advanced lines were sent to Vietnam (18) and Cambodia (79) were grown at PRRC and HTA for evaluation and selection with the characteristics of promising lines summarized in by rice scientists from counterpart countries for their own Table 5; (3) 11 leading DWR Indian lines were evaluated and target areas; (7) an acid-sulfate screening experiment was used in the FPR breeding program at PRRC but were found conducted in a farmer’s field at Prachinburi under soil pH unsuitable for FPR areas of Thailand because of earliness; (4) 3.5 with 200 entries evaluated and the output used by col-
34 Sommut et al laborators and other countries encountering acid-sulfate soil year, a group of FPR breeding lines, include F2–F6 popula- problems; and (8) valuable germplasm exchanged through tions, stable lines, elite lines, and local germplasm, were INGER in the International Deepwater Rice Observation screened in ponds under target water depth of 90, 110, and Nursery (IDRON), International Blast Nursery (IRBN), and 150 cm using 35-day-old seedlings at HTA. Many Thai breed- International Brown Planthopper Nursery (IRBPH) was sent ing materials, especially local varieties, had good elongating to PRRC from 1992 to 1998. Out of 316 IDRON lines, one stems above water at 150-cm depth. line (WAR87-7-R-10-2-3B-1-2B) was selected for testing in replicated yield trials. A few good lines resistant to BPH Photoperiod sensitivity and blast such as IRAT2440, IRAT300, BR11, IAC47, and Photoperiod sensitivity is an essential characteristic in all RT1031-69 (Acc. 15092) were used in our breeding program. FPR areas in Thailand, mostly involving three groups of PRRC and HTA regularly nominated a set of breeding lines maturity. First, for FPR in central-east areas, flowering dates to the three nurseries, particularly IDRON. must be in early November. Second, in deeper low-lying areas Apart from germplasm exchange, FPR breeding lines of central-east Thailand, flowering should be at the end of from PRRC were sent to rainfed lowland areas of Thailand November, synchronized with water recession. Third, for the in order to strengthen and make a more efficient linkage of late-maturity group suited for very deep low-lying central rice improvement between the FPR and rainfed lowland areas of Ayuthaya, Ang Thong, Singh Buri, and Suphan Buri, rice ecosystems. The materials sent included 9 glutinous the appropriate flowering time should be early December. and nonglutinous entries to the north in 2004, 200 glutinous However, photoperiod-insensitive varieties are also desired and nonglutinous entries to the northeast between 1998 and for favorable areas needing short-duration high-yielding 2004, and 246 nonglutinous entries to the south in 2003. varieties (less than 100 days) for the recent conversion of traditional monocropping of FPR into double cropping of FPR breeding thrusts at PRRC modern varieties in FPR areas. Submergence tolerance Type of endosperm and grain quality A simple mass screening method for submergence tolerance Nonglutinous endosperm type with acceptable grain charac- using a controlled water level in ponds was developed in teristics (long, slender, clear grain) and eating quality similar Thailand (Boonwite et al 1977, Supapoj et al 1979). Mass to those of Khao Dawk Mali 105 (soft and sticky after cook- screening in ponds is now the most common method for ing), Khao Tah Haeng 17, or Leuang Pratew 123 (rather hard testing submergence tolerance. Seedlings 30 days old are and fluffy after cooking) are essential. Currently, due to the transplanted in the pond and, 30 days after transplanting, linkage of the rainfed lowland and flood-prone ecosystems, the pond is filled to 150-cm water depth for 10 days or until glutinous varieties are also needed for northern and northeast the death of a susceptible check. Submergence tolerance Thailand. However, the greatest demand will still be for the scoring is done immediately after draining water and 14 nonglutinous type. days later. As far as our work is concerned, all FPR breeding lines, including local varieties, lack the trait of tolerance of Other biotic and abiotic stresses complete submergence of 10 days, but the recovery abil- Drought resistance at the vegetative stage is essential as well ity of some lines is good, including that for the donor line as adaptability to acid-sulfate soil conditions in central-east BKNFR76106-16-0-1-B. Thailand. To improve FPR yield, resistance to major diseases and insect pests should be incorporated into the varieties for Elongation ability release. Elongation ability is a stable and heritable trait. The number of internodes and length of culm elongation in FPR appear to FPR breeding activities be controlled by different genes (Morishima 1975). Supapoj et al (1977) reported that populations from crosses between Germplasm collection, evaluation, and use floating and nonfloating varieties segregated with only a small Thailand is believed to be situated in the center of origin portion of true floating types and more of intermediate types. for rice; hence, high genetic diversity is visible. The col- Hamamura and Kupkanchanakul (1979) identified partial laborative effort initiated in 1982 between Thailand and dominance and the multigenic nature of floating ability. IRRI with support from the Japanese government collected Some researchers suggested multigenic control but were not 23,903 accessions of local varieties and other rice. During certain of its degree of dominance. In some crosses, partial 1995-2000, a total of 557 accessions of local varieties were dominance was indicated, but in some cases dominance was evaluated and characterized at PRRC under natural flood- found. They also suggested that hybrid populations should ing. Table 6 summarizes some quantitative and qualitative be flooded for a short period only and that the parents, F1, characteristics. Collection of local varieties was repeated in and F2 be grown in the same environment to avoid variation. 1992 and 2000. Tripathi and Balakrishna Rao (1985) reported that a single dominant gene controlled early nodal differentiation. Each
Breeding rice for deepwater and flood-prone areas of Thailand 35 Table 6. Characteristics of 557 local rice varieties evaluated at Prachinburi Rice Research Center in wet season during 1995-2000.
Quantitative characteristics Qualitative characteristics
Characteristic Mean Range Characteristic Description
Height at maturity 174 ± 18.88 79–246 Culm habit Erect, semierect, (cm) open Flowering date 2 Oct.-4 Jan. Spreading, prostrate Days to 50% 167 ± 16.85 92–156 Has awn Absent floweringa Number of tillers 9 ± 2.39 4–21 Leaf blade Green, dark green per hill color Number of panicles 8 ± 1.84 3–19 Purple on tip per hill Number of spikelets 230 ± 56.75 27–353 Node color Green, purple per panicle Green with purple line Panicle length (cm) 26.3 ± 2.23 19.0–33.3 Internode color Green Spikelet fertility (%) 89.2 ± 4.72 76.4–95.8 Ligule color Green 100-grain weight 2.7 ± 0.36 1.3–4.3 Auricle color Green, purple (g) Grain size Green with purple line Paddy rice Hull color Straw black, yellow, brown Length (mm) 9.9 ± 0.67 7.4–11.5 Straw with brown line Width (mm) 2.7 ± 0.22 2.2–3.8 Straw with brown spot Thickness (mm) 2.1 ± 0.10 1.4–2.8 Straw with purple spot Brown rice Brown rice White, red color Length (mm) 7.3 ± 0.55 5.4–8.8 Dark purple, light brown Width (mm) 2.3 ± 0.15 1.8–2.9 Spikelet tip Pink, red, straw, color purple Thickness (mm) 1.8 ± 0.10 1.1–2.4 Stigma tip color White, light purple Amylose content 27.52 ± 5.05 5.56–31.63 Dark purple, yellow (%)
aCounted from when seeded.
Varietal development flood-prone areas of Prachinburi, Nakon Nayok, Ayuthaya, Genetic variation in FPR populations was created by hy- Lopburi, Pitsanulok, and Pichit provinces. Maximum water bridization and induced mutation. Realizing the importance depth was 70–190 cm. Some lines were tested in more than of photoperiod sensitivity, elongation ability, as well as sub- 30 trials and two lines, HTA88095-5B-4 and Pahn Tawng-67, mergence tolerance, exotic and domestic donor parents were outyielded the check by 18% (Table 7). An irradiated mutant used. Most of the exotic breeding materials were sent through of FR Plai Ngahm Prachinburi with grains more translucent IRRI. Segregating populations were subjected to abiotic than those of the original variety was obtained. and biotic stress screening and then advanced to yield trials. The highlights from breeding FPR tolerant of acid-sul- Screening of segregating populations under natural flooding fate conditions are shown in Table 8. DWCT82-2-2, a com- in farmers’ fields increased confidence in crop survival before posite cross line, and pure-line-selected Khao Bahnnah-432 entering materials in yield trials. In addition, other supportive had more translucent grain and better yield than the check information, for example, fertilizer response, reactions to varieties. These two lines will be released soon to a target diseases and insect pests, together with grain physical and area of about 5,000 ha. Approximate grain yield and some chemical properties, was recorded. The participation of farm- agronomic characteristics together with reactions to major ers in selection and conducting yield trials was successfully diseases and insects pests of some advanced elite lines from enlisted (Fig. 4) and has been used since 1998. From 1998 to various sources are given in Table 9. The line WAR87-7-R- 2004, yield trials were conducted at 6–8 sites every year in 10-2-3B-1-2B was selected from the IDRON of INGER.
36 Sommut et al Table 7. Mean grain yield and some agronomic characteristics of flood-prone rice from farmers’ fields in yield trials during 1998-2004, Thailand.
Designation No. of trials Flowering date Height (cm) Grain yield Grain length Chalkinessa Amyloseb (%) (t ha–1) (mm)
HTA88026-5B-38 31 28 Nov 212 2.2 6.76 1.38 26.91 HTA88026-5B-39 12 27 Nov 192 2.3 7.60 2.00 27.79 HTA88037-5B-21 12 30 Nov 214 2.3 7.90 2.20 28.16 HTA88095-5B-4 31 24 Nov 223 2.6 7.57 1.64 28.21 HTA88110-5B-20 25 24 Nov 234 2.2 7.37 1.51 27.22 Khao Luang-28 16 16 Nov 209 2.1 7.49 1.72 23.54 Pahn Tawng-67 31 29 Nov 218 2.6 7.04 1.61 26.85 PCR90155-3B-10-1 25 22 Nov 225 1.9 7.48 2.16 26.93 PCR91038-8-3-3B 25 29 Nov 215 2.1 7.40 1.38 27.73 PCR92058-R-1 25 23 Nov 218 2.4 7.60 2.12 27.46 PCRC93018-48 8 23 Nov 273 2.4 7.35 0.86 26.78 PCRFR84004-17-B 6 27 Nov 262 2.3 7.66 2.22 28.38 PCRFR84007-48-B 6 27 Nov 255 2.3 8.03 1.07 27.87 PCRFR84009-40-B 6 27 Nov 225 2.3 7.68 1.56 28.57 PCRFR84013-43-B 39 26 Nov 242 2.3 7.95 2.38 29.04 PCRFR84019-21-B 39 25 Nov 232 2.4 7.82 3.84 28.43 60 PNG’93G2 Co-17-7 8 4 Dec 296 2.1 7.62 1.49 29.56 60 PNG’93G2 Co-62-30-B 8 28 Nov 288 2.2 7.64 1.09 29.94 Plai Ngahm Prachinburi 39 28 Nov 265 2.2 7.62 2.13 27.11 (check)
aScale of 0–9: 0 = none, 1 = small (less than 10% of kernel area), 5 = medium (10–20% of kernel area), 9 = large (more than 20% of kernel area). bAmylose content (%): low = less than 20%, intermediate = 20–25%, high = more than 25%.
Table 8. Yield, flowering date, plant height, and other characteristics of floating rice in farmers’ fields in yield trials under acid-sulfate soil conditions conducted by Prachinburi Rice Research Center during 1998-2004 wet season.
Designation Flowering date Height (cm) Grain yield (t ha–1) Grain length (mm) Chalkinessa Amyloseb (%)
BKNFR82033-2-10-2-5 19 Nov 220 3.2 7.57 3.90 29.00 DWCT82-2-2 19 Nov 214 3.1 7.56 2.52 28.29 Khao Bahnna-432 12 Nov 223 3.1 7.70 2.60 29.41 Khao Bahnna-184 19 Nov 226 2.7 7.46 4.36 28.77 Khao Tah Petch-6 16 Nov 225 2.7 7.73 4.35 29.38 BKNFR82033-2-10-2- 21 Nov 225 2.6 7.43 3.86 28.81 316 BKNFR82033-2-10-2- 18 Nov 241 2.5 7.54 3.98 28.92 3-6 BKNFR82033-2-10-2 21 Nov 222 2.5 7.47 2.51 29.00 Khao Bahnna (check) 11 Nov 223 2.9 7.77 3.76 28.66 Khao Tah Petch (check) 11 Nov 227 2.7 7.59 4.01 29.98 aScale of 0–9: 0 = none, 1 = small (less than 10% of kernel area), 5 = medium (10–20% of kernel area), 9 = large (more than 20% of kernel area). bAmylose content (%): low = less than 20%, intermediate = 20–25%, high = more than 25%.
Varietal release by agricultural extension officers and researchers from HTA. Varietal improvement of FPR in Thailand began many years The process of on-station and farmers’ field trials was used before the establishment of the Huntra Rice Experiment Sta- to evaluate yielding ability. Promising lines with high yield tion (HTA) in 1941. Simple breeding methods were used to and good grain quality were proposed for release to farmers. collect and purify local varieties. Five recommended varieties In 1959, the FR varieties Nahng Chalong (glutinous rice), were released: Mali Awng, Mali Tawng, Jampah 133, Kawd Chek Choey, Ta Pao Kaew 161, Leb Mue Nahng 111, and Pin Pawm, and Khao Med Lek. The FPR varietal improvement Gaew 56 were released, followed by Khao Nahng Noey 11 program was strengthened in 1950 by using the pure-line and Khao Puang. BKN6986-66-2 and BKN6986-147-2 were selection method. Collecting rice germplasm was handled released as RD 17 and RD 19, respectively, in 1979. RD 17
Breeding rice for deepwater and flood-prone areas of Thailand 37 Table 9. Approximate grain yield and some agronomic characteristics of flood-prone rice lines currently tested in replicated yield trials at Prachinburi Rice Research Center, Thailand, 2005 wet season.
Flowering Height Yield Reactions to Designation date (cm) (t ha–1) Chalkinessb BPHa GLH Bl/PCR Bl/KHS
GS.21054 21 Oct 178 3.5 S MR VS VS 1 PCRC01146 21 Oct 184 4.5 S R VS VS 1–2 PCRC01085 23 Oct 177 3.1 MS MR MR R 1 GS.4937 24 Oct 181 3.9 MS S VS VS 1 GS.4903 25 Oct 174 3.2 MS MS VS VS 1 PCR93011-77-10-B 26 Oct 176 3.6 MR MR VS VS 0–1 PCR93128-1-2B-1-2B 26 Oct 167 3.3 MR MR VS VS 0–1 PCRC01037 28 Oct 166 3.4 MS MS VS MS 2 PCR92082-B-3-3-B 29 Oct 172 3.1 MR MS MS VS 2 PCR91052-4-B-1-B-1-3B 31 Oct 179 4.0 S MS MS MS 1–2 PCR93093-55-1-2-2-2 1 Nov 182 3.3 S MR R S 1–2 PCR93186-2-2-1-4B 5 Nov 172 3.6 MS MR VS VS 1–2 PCRC01055 5 Nov 166 3.5 S R VS VS 1–2 PCR93151-15-2B-2-B 7 Nov 182 3.3 MR R MS MS 1–2 PCR92093-14-2-2B-290 10 Nov 187 3.4 S MS VS VS 1 PCR93098-7-B-1 10 Nov 180 3.6 MR MS R R 1–2 PCR94020-1-2-B 12 Nov 184 3.4 MR MS R MS 1–2 SPR’76 Com4-100-1-3-21 12 Nov 181 3.2 MR R R R 1 WAR87-7-R-10-2-3B-1-2B 12 Nov 197 3.5 MS MR R MR 1–2 PCRC01109 13 Nov 176 3.1 S MS VS VS 0–1 GS.20933 16 Nov 185 3.2 S S VS VS 1 PCR92093-45-1-2B 16 Nov 179 3.2 S MS R R 1 SPR’76 Com4-100-1-3-24 16 Nov 181 3.1 MS R MR MR 1–2 HTA60’93G160Co-67-7-250 21 Nov 183 3.3 S S VS VS 0 PCRC01001-13 23 Nov 201 3.5 MS S VS S 0–1 PCRC01159 23 Nov 171 3.7 MS MR VS VS 1 PCRC01001-39 24 Nov 370 3.1 S S S S 0–1 GS.93018-48 25 Nov 185 3.3 MR MS VS VS 0–1 GS.12386 26 Nov 183 3.7 MS S VS S 0–1 GS.15927 28 Nov 178 3.2 S MS VS VS 0–1 GS.9368 28 Nov 164 3.1 R S VS VS 0–1 Krating Daeng 28 Nov 181 3.3 MR MS VS VS 1 PNG’93G160Co-73-45 1 Dec 204 3.6 MS S MS VS 2 PNG93G160Co-73-49 5 Dec 180 3.3 MR MS S S 2
aBPH = brown planthopper, Bl = blast disease, Bl/PCR = blast disease tested at Prachinburi Rice Research Center, Bl/KHS = blast disease tested at Royal Study Development Project, GLH = green leafhopper. R = resistant, MR = moderately resistant, S = susceptible, MS = moderately susceptible, VS = very susceptible. bScale of 0–9: 0 = none, 1 = small (less than 10% of kernel area), 5 = medium (10–20% of kernel area), 9 = large (more than 20% of kernel area).
and RD 19 yielded 3.0 and 3.9 t ha–1 and outyielded Pin Gaew 100–150 tons of foundation seed of the five FPR varieties 56 by 50% and 100%, respectively. In 1987, SPR7270-18 are produced annually due to budget limitations. Therefore, (Khao Nahng Noey 11/C4-63) was recommended as Huntra seed insufficiency occurs every year, which might cause seed 60 for deepwater areas of 50–100-cm water depth, where mixtures in farmers’ fields. Farmers prefer to buy seeds from it yielded as high as RD 19 but had better grain quality. the Rice Research Center/Station rather than from the Center The FPR varietal improvement program during 1992-2004 of Seed Multiplication. However, they often complain that released five varieties through the Thai-IRRI Collaborative seed prices are too high. FPR Project (Table 10). Achievements Seed production and distribution During 1992 to 2005, more than 200 Thai breeding lines The Rice Research Institute has the mandate to produce (with domestic and exotic parentages) were tested in India, breeder and foundation seeds of newly released varieties as Bangladesh, Vietnam, Cambodia, and Indonesia for adapt- well as anticipated breeder seeds of a proposed variety in ability. Three lines were officially released for and became Thailand. Foundation seeds are sold to the Department of popular in the deep-flooded areas of Cambodia—Don Agricultural Extension’s Center of Seed Multiplication. Only (HTAFR77022-45-3-2), Tewada, and Khao Tah Petch. Two
38 Sommut et al Table 10. Description of characteristics of released varieties.
Variety name Parenta Year of Flowering Yield Endosperm Target areas Distinguishable release date (t ha–1) typeb charactersitics
Plai Ngahm Prachinburi Mass selection from 1994 28 Nov. 2.4 NG Water depth more l Photoperiod sensitive (PNG Prachinburi) local variety than 100 cm l Excellent elongating type l Slender grain with rather high chalkiness l Big panicle, good exsertion l Seed dormancy of 7 weeks l Moderately blast resistant l Good for rice products, e.g., noodles Prachinburi 1 (PCR 1) Composite 1998 25 Nov. 3.4 NG Water depth l Photoperiod sensitive
crosses of 29 F2 50–120 cm l Intermediate grain size populations with l Sturdy culm, big panicle, IR lines good exsertion l Moderately tolerant of drought, submergence, acid-sulfate soil l Good weed competition at early stage Prachinburi 2 (PCR 2) BKNFR80086*/ 2002 18-23 Shallow NG Water depth l A new plant type, high- (new plant type) HTAFR80038** Nov. water: 5.3 50–100-cm yield, flood-prone rice Deepwater: l Photoperiod sensitive 3.7 l Slender grain with high chalkiness l Moderately tolerant of submergence, acid-sulfate soil l Resistant to blast l Good for rice products Ayuthaya/(AY 1) (new U-Ta Pao/ Khao 2004 6-10 Shallow NG 50–100-cm l A new plant type, high- plant type) Dawk Mali 105 Nov. water: 5.3 water depth yield, flood-prone rice Deepwater: l Photoperiod sensitive 3.4 l Earlier than PCR 1 and PCR 2 l Resistant to BPH l Moderately tolerant of submergence, acid-sulfate soil l Good for rice products l Photoperiod insensitive Bang Taen SPR60/IR60//IR64 2004 90–95 4.4 (wet NG l Flood-prone- days season) areas where l Sturdy, short stature, (Nov. 5.2 (dry converted to 100–110 cm to Mar. season) double-crop of l Long, slender, and clear sown), MV pre- and grain 110–115 postflood l High yield l High amylose content days l With (sown supplemen- l 15–20 days earlier than May. to tary irrigation current popular MV Aug.) a* = derived from Rayada No. 14. ** = derived from IR8234-OT-20. bNG = nonglutinous endosperm. continued on next page
Breeding rice for deepwater and flood-prone areas of Thailand 39 Table 10 continued.
Variety name Parent Year of Flowering Yield (t ha–1) Endosperm Target areas Distinguishable release date type characteristics
l Resistant to blast, BPH, and GLH in areas of Prachinburi Nakon Nayok, and Chachoengsazo provinces l Tolerant of acid-sulfate soil
PCR89151-27-9-155 IR46/Hawm Nai Pon Expected 25-30 3.3 (rainfed NG 50–100-cm l Photoperiod sensitive in 2005 Oct lowland), 2.8 water depth l Good-quality, flood-prone (deepwater rice area) l Long, slender, clear grain with aroma l Low amylose content l Sturdy culm
varieties of new plant type DWR were also developed in in the cultivation of other types of rice, research on the FPR Thailand, with one released as Prachinburi 2 in 2002 and ecosystem should be strengthened to retain Thailand’s posi- another as Ayuthaya 1 in 2004. These two varieties performed tion as the world’s best quality rice producer and leading very well in the flood-prone areas of the eastern plain of Thai- rice exporter.Therefore, total productivity-enhancing initia- land, with yields of about 3.6 t ha–1 in deepwater areas and tives should focus on a systems approach rather on the FPR 5.0 t ha–1 in shallow-water areas. Another composite cross plant. variety, Prachinburi 1, with excellent weed competitiveness For shallow-flooded areas (50 to 100 cm), activities in the early stage as well as good yield at 3.75 t ha–1, was should include (1) development of medium-high-yielding also released in 1998 to replace local varieties in deepwater varieties with some essential agronomic traits—grain quality, areas. Plai Ngahm Prachinburi, an FPR variety, was released intermediate height, new plant type DWR traits, submergence in 1994. Moreover, a photoperiod-insensitive variety, Bang tolerance, and low fertilizer responsiveness; (2) improvement Taen, was released in 2004 to facilitate the change in cropping of submergence tolerance in glutinous rice for the northeast; intensity in flood-prone areas. This variety has the shortest and (3) continuation of pure-line selection using traditional duration among modern varieties and is also suitable for varieties, especially for quality and specialty rice (for nutri- fully irrigated areas. tion and processing). For deep-flooded areas (over 100 cm), research priori- Future prospects ties can include (1) development of shorter duration (95–110 The national plan for rice production during 2004-08 envi- days) varieties with good grain quality and resistance to sions improving productivity and quality while maintaining brown planthopper to facilitate the shift to double cropping; planted area at 9.2 million hectares for the major crop and (2) development of appropriate technologies for good crop 1.44 million hectares for the second crop in order to support establishment of the major and second rice crops; (3) con- domestic consumption and keep the export market at about tinuation of pure-line selection of traditional varieties with 10 million tons of milled rice annually. Therefore, average the desirable traits of elongation ability, kneeing ability, as rice yields must be increased by 30%, from 2.2 t ha–1 in 2003 well as good grain quality, for both direct consumption and to 2.8 t ha–1. In addition, yield of the second crop must also industrial use; and (4) proper water management. be increased by 18%, from 4.2 t ha–1 to 5.0 t ha–1. To reach these targets, rice research in the FPR ecosystem should (1) References strengthen research by a consortium approach and (2) clearly define its research priorities. Bekhasut P, Puckridge DW, Wiengweera A, Kupkanchanakul T. 1990. Forming a consortium would benefit farmers in the 12 Sequential elongation of internodes of deepwater rice at different million hectares of FPR area in South and Southeast Asia. water depths. Field Crops Res. 24:195-209. Despite constraints, this area has high potential for improv- Boonwite C, Setabutara C, Jackson BR, Prechachart C, Anugul P. ing rice productivity if scientists work closely together. The 1977. Screening for submergence tolerance using a deepwater exchange of germplasm and technologies across ecosystems pond. In: Proceedings of the 1976 Deepwater Rice Workshop, and within and across national borders should enhance tech- Bangkok, Thailand. Manila, (Philippines): International Rice nology adoption and productivity of farmers. Although there Research Institute. p 101-108. is a declining trend in the growing of FPR and an increase
40 Sommut et al Catling HD, Puckridge DW, Boonyaviwatana S, Pattrasudhi R, Kupkan- Supapoj N, Prechachat C, Kupkanchanakul K. 1979. Screening for flood chanakul T, Tanasate M, Supapornhaemin P. 1982. Deepwater tolerance in the field. In: Proceedings of the 1978 International rice crop-cuts for 1981. In: Report of Deepwater Rice Planning Deepwater Rice Workshop, Calcutta, India. Manila (Philip- Meeting, March 1982. Department of Agriculture, Bangkok, pines): International Rice Research Institute. p 135-138. Thailand. p 39-40. Tripathi RS, Balakrishna Rao M.J. 1985. Inheritance studies of charac- Charoendham P, Palaklang W, Niyomvit L, Prommanart P, Charoen- ters associated with floating habit and their linkage relationship silp N. 1995. Survey of deepwater rice cultivation in Thailand, in rice. Euphytica 34:875-881. 1992/1993. Prachinburi Rice Research Center. (In mimeo.) Vergara BS, Visperas R, Peralta J, Shuwisitkul E, Karin S, Sophonsakul- Choudhary MA, Zaman SMH. 1970. Deepwater rice of East Pakistan. kaew S. 1977. Screening for kneeing ability. In: Proceedings of Paper presented at the International Rice Commission Working the 1976 Deepwater Rice Workshop, Bangkok, Thailand. Manila Party on Rice Production and Protection, 13th Session, Teheran, (Philippines): International Rice Research Institute. p 123-124 Iran. 19 p. (In mimeo.) Vergara BS, Mazaredo A, De Datta SK, Abilay W. 1975. Plant age Hamamura K, Kupkanchanakul T. 1979. Inheritance of floating ability and internode elongation in floating rice varieties. In: Proceed- in rice. Jpn. J. Breed. 29(3):211-216. ings of the International Seminar on Deepwater Rice, August Huke RE, Huke EH. 1997. Rice area by type of culture: South and 1974. Bangladesh Rice Research Institute, Dacca, Bangladesh. Southeast Asia. Los Baños (Philippines): International Rice p 178-183. Research Institute. Vergara BS. 1985. Growth and development of the deepwater rice plant. Ministry of Agriculture and Cooperatives. 2004. Agricultural statistics IRRI Research Paper Series, Number 103. Manila, (Philippines): of Thailand, crop year 2003/04. Bangkok, Thailand. International Rice Research Institute. 30 p. Morishima H. 1975. Floating ability as an adaptive character of the rice and its measuring method. In: Proceedings of the International Notes Seminar on Deepwater Rice, August 1974. Bangladesh Rice Research Institute, Joydebpur, Dacca, Bangladesh. p 109-114. Authors’ address: Senior rice breeder, former senior rice breeder, Puckridge DW, Kongchum M, Thongbai P, Sattarasart A, Sinoupakarn former director, research assistant, and director, Prachinburi S. 1994. Nitrogen uptake and yield of deepwater rice in the Rice Research Center, Thailand 25150. E-mail: w_sommut@ Central Plain of Thailand. Field Crops Res. 37:193-204. hotmail.com. Puckridge DW, Kupkanchanakul K, Peasaschar K, Sripongpankul S, Acknowledgments: Grateful acknowledgement is due to the govern- Runtan S. 1990. Influence of time of flooding on development of ment of Thailand and the International Rice Research Institute basal internodes. In: Report of Deepwater Rice Planning Meeting for useful collaboration on rice improvement. We owe much to 1990, Department of Agriculture, Bangkok, Thailand. the former program leader, Dr. D.W. Puckridge, former project Setter TL, Waters I, Greenway H, Atwell BJ, Kupkanchanakul T. 1987. leader, Dr. Drrk HelleRisLambers, and the late Dr. D. Senadhira Carbohydrate status of terrestrial plants during flooding. In: for their kind effort and supportive collaboration. Our respec- Crawford RMM, editor. Plant life in aquatic and amphibious tive thanks go to Dr. M. Hossain for his kind guidance on the habitats. British Ecological Society Symposium No. 5. Oxford socioeconomic aspects as well as Dr. V.P. Singh for his kind (UK): Blackwell Scientific Publications. p 411-433. support. Special thanks also go to all scientists involved in the Sommut W, Bose LM, Singh VP, Hossain M. 2004. Agriculture and FPR collaborative project on germplasm exchange for South livelihoods in the flood-prone ecosystem in Thailand. Kasetsart and Southeast Asia and to the staff of PRRC and HTA for their J. (Soc. Sci.) 25:69-89. helpful assistance. Supapoj N, Setabutara C, Kupkanchanakul K, Shuwisiku E. 1977. Segregation for elongation ability in two crosses of floating rice with ordinary lowland rice. 1. Estimation of heritability and implication of selection efficiency. In: Proceedings of the 1976 Deepwater Rice Workshop, Bangkok, Thailand. Manila, (Philip- pines): International Rice Research Institute. p 29-36.
Breeding rice for deepwater and flood-prone areas of Thailand 41 42 Sommut et al Shallow-flooded (submergence- prone) aman areas of South Asia
Breeding rice for submergence-prone and aman areas of India 43 44 Mallik et al Breeding rice for submergence-prone and aman areas of India
S. Mallik, J. Ahmed, S.K. Bardhan Roy, J.N. Reddy, and G. Atlin
The varietal development program for rainfed lowland rice, which occupies 17 million hectares in India, has lately received increased attention. Efforts were intensified only during the early 1990s through the launching of different projects, comple- mentary to the national and state programs, to enhance the productivity of this ecosystem. A shuttle-breeding program, a collaborative project between the Indian Council of Agricultural Research and the International Rice Research Institute, provided an opportunity for enhancing the flow of breeding materials of diverse origin among the eastern Indian states to strengthen the breeding program. The materials developed through this project also served as inputs to other breeding programs for this ecosystem. Many promising cultures having a diverse genetic base (15 to 20 parents from different countries), better plant type, and improved sink components have been developed. They have attained 20% higher yield than the best checks and, we hope, will increase the productivity of this ecosystem.
Rice is the most important food crop in India, occupying 44.5 Table 1. Area, production, and productivity of rice in India million ha, with a total production of 85.5 million tons and in different seasons during 1974-75 to 2000. –1 average productivity of 1.9 t ha (Tables 1 and 2). The area Year/season Area Productivity Production under rainfed lowland and deepwater conditions is 17 million (million ha) (t ha–1) (million t) ha (Table 2), of which 10 million ha face submergence at different growth stages. These areas are mainly concentrated Kharif (wet season) in eastern India (Table 3). 1974-75 35.96 1.0 35.93 Rainfed lowland rice, accounting for nearly 40% of the 1999-2000 40.67 1.9 76.71 rice area in India (Khush and Baenziger 1998), received less Rabi (dry season) 1974-75 1.93 1.9 3.65 attention from the rice research community than it deserved 1999-2000 4.30 3.0 12.77 compared with irrigated rice. The first coordinated trial for Total rainfed lowlands was constituted in 1976 with some pure-line 1950-51 30.81 0.7 20.58 selections from Uttar Pradesh (UP), Bihar, and West Bengal 1999-2000 44.97 2.0 89.48 (WB). However, in the annual rice workshop of the All India Source: Ratho (2004). Coordinated Rice Improvement Project (AICRIP), Hydera- bad, renamed as the Directorate of Rice Research (DRR) under the Indian Council of Agricultural Research (ICAR) in the deepwater ecosystem. In the classification used by DRR, April 1979, it was decided to undertake a systematic multi- shallow water represents a situation in which standing water locational coordinated program for lowland conditions, after generally rises up to 40 cm, for the semideepwater 41–75 cm, 14 years’ dominance of the semidwarf breeding program for and for deep water more than 75 cm. the irrigated ecosystem. Initially, for the first seven years, The reasons for the slow progress in varietal develop- the Rice Research Station (RRS), Chinsurah, was entrusted ment for the rainfed lowland ecosystem are numerous (Mallik with the responsibility of coordinating and monitoring the 1995, Mallik et al 1995): rainfed lowland rice trials, but the program was taken over l Harsh, heterogeneous, and unpredictable environ- by the DRR in 1986 (Mallik 2000). ment. The DRR program has aided actively in the devel- l Wide variability, both between locations and within opment and release of 720 rice varieties in the country so locations over years, with large G × E interaction. far (DRR 2001a,b, 2002, 2003a,b, 2004a,b, 2005, Prasad l The crop is generally grown once a year, and most et al 2001), of which 277 (38.5%) have been for rainfed varieties are long-duration/photoperiod-sensitive, ecosystems (Table 4). The number of released varieties for thus limiting the advancement of generations to once the rainfed lowland ecosystem is only 175 (24.3%), among in a year. which 129 (17.9%) are for the shallow ecosystem, 31 (4.3%) l Advancing the generations (F2 onward) did not use for the semideepwater ecosystem, and only 15 (2.1%) for in situ testing in many of the locations.
Breeding rice for submergence-prone and aman areas of India 45 Table 2. Rice ecologies, area, production, and productivity in India (2000-01). cm of water) and deepwater (>75 cm of water)—and the average productivity of this ecosystem is relatively low (<1 Area Production t ha–1) compared with that of other rice-growing ecosystems Rice ecologies Yield (Singh 2000). Out of 277 varieties for the rainfed ecosystem –1 Million % Million % (t ha ) released in India, only 46 (6.4% of the total varieties and ha tons 16.2% of the varieties released for the rainfed ecosystem) have been released for submergence-prone areas. Four of Irrigated 20.5 46 60.0 70 2.9 Wet season 16.5 36 46.0 50 2.8 these, Nalini, Amulya, Purnendu, and Jitendra, were released Dry season 4.0 9 14.0 20 3.5 by the Central Variety Release Committee (CVRC) of ICAR, Upland 6.0 14 5.5 6 0.9 while the remaining 42 were released by the State Variety Favorable 2.0 5 3.0 4 1.5 Release Committees (SVRC) of different states. During Drought-prone 4.0 9 2.5 2 0.6 2000-04, a total of 88 varieties were released in India, of Rainfed lowlands (0–50 cm) 13.0 29 16.0 19 1.2 Drought-prone 4.0 9 6.0 7 1.5 which only two are for the submergence-prone ecosystem. Favorable 3.0 7 6.0 7 2.0 Of 46 varieties released so far, 19 have been developed from Medium deep, waterlogged 3.0 7 2.5 3 0.8 single crosses, 5 from three-way crosses, 2 through mutation Submergence-/flood-prone 3.0 7 1.5 2 0.5 breeding, and the remaining 20 through pure-line selection Deepwater (>50 cm) 4.0 9 3.0 4 0.8 from landraces. The varieties thus far developed, except for Deepwater 3.0 6 2.5 3 0.8 Floating rice 1.0 2 0.5 1 0.5 Bhudeb (Mallik et al 2003), for this ecology have a narrow Coastal wetlands 1.0 2 1.0 1 1.0 genetic base as a limited number of parents (2–3, mostly of Total 44.5 100 85.5 100 1.9 Indian origin) were used in their breeding ancestry. Conse-
Source: Singh (2002). quently, these varieties have a limited area of adaptability and are not providing higher yields in this harsh and variable growing environment. l Inadequate generation of new materials with a wide This paper summarizes the varietal improvement efforts genetic base. for this ecosystem. l Inadequate donors for desirable traits, required for this ecosystem. Programs for enhancing the productivity of rainfed lowland l Lack of proper screening methodologies. rice in eastern India l Lack of a proper testing facility at many of the re- search stations. Besides the state and national programs, the following l Lack of proper supporting environmental data such projects are either in operation or completed with different as water depth, duration, and depth of submergence objectives for increasing the productivity of rainfed lowland when presenting trial results by researchers—such rice. as inadequate environmental characterization. l Fewer researchers involved than with the irrigated The Rainfed Lowland Consortium ecosystem. This collaborative project between the International Rice l Less involvement of NGOs, especially in varietal Research Institute (IRRI) and the National Agricultural Re- improvement, unlike the hybrid rice development search Programme (ICAR) started in 1991. The objectives program. were to expand and strengthen the research base for rainfed Out of 17 million ha of rainfed lowlands, 10 million ha rice ecosystems. The emphasis was on strategic and applied are submergence-prone, where improvement has been mini- research to generate new technologies that will have both mal. There are two major subecosystems—semideep (41–75 local and regional application.
Table 3. Rainfed lowland rice area (million ha) in different subecosystems in east- ern India.
Subecosystem Assam Bihar Orissa Madhya Uttar West Total Pradesh Pradesh Bengal
Lowland (0–30 cm) 0.9 1.7 1.7 2.7 1.9 1.7 10.6 Lowland (30–50 cm) 0.5 0.5 0.5 – 0.3 0.5 2.3 Deepwater (>50 cm) 0.4 0.4 0.4 – 0.2 0.4 1.8 Floating (>100 cm) 0.1 0.7 0.1 – 0.5 0.7 2.1 Total (lowland and 1.9 3.3 2.7 2.7 2.9 3.3 16.8 deepwater) Source: Singh (2000).
46 Mallik et al Table 4. Varieties released in India for different ecosystems up to 2004.
Ecosystem No. of varieties Percentage National checks
Rainfed Upland 95 13.2 Heera, Vandana, Annada, PS Dhan1 Shallow 129 17.9 Savithri, Pooja, Salivahana Semideep 31 4.3 Sabita, Purnendu Deep 15 2.1 Jalmagna, Dinesh Hill 7 1.0 VLDhan221, CH988 Total 277 38.5 Irrigated Early 139 19.3 Heera, Vandana, Annada, PS Dhan1 Medium-early 43 6.0 Sasyasree, IR64, RH204 Medium 178 24.7 Jaya, KRH2 Saline-alkaline 21 2.9 CSR13, CSR27, Jaya Scented 30 4.2 P. Basmati1, T. Basmati Hill 32 4.4 VL Dhan 61, RP2421 Total 443 61.5 Grand total 720 100
Initially, Masodha under NDUAT, Faizabad, UP, and l A plow with a seeding attachment has been found Polba-Chinsurah under the Department of Agriculture, useful by farmers because it increases the plant government of WB, were the two key sites in India among population and reduces the cost of weeding. eight locations in five countries (IRRI 1992). The Central l Opportunities for scientists to participate in meet- Rice Research Institute (CRRI), Cuttack, was later included ings, workshops, conferences, and symposia for as another site for the consortium. The main research focus better understanding of the rainfed lowland ecosys- for Masodha, Polba-Chinsurah, and CRRI was flash flood, tem. stagnant flooding, and mechanization, respectively. Informa- l Human resource development through training on tion regarding environmental characterization, mechanisms experimental design and data analysis, G × E analy- of submergence tolerance, and germplasm improvement sis, etc. was generated for the ecosystem (below 50 cm water depth) l Infrastructure development. through this project (Mallik et al 1995). Activities and ac- complishments included Eastern India Rainfed Shuttle Breeding Program l Mechanisms of submergence tolerance character- A significant change in the breeding approach in the early ized. 1990s was the use of a shuttle-breeding program. To improve l Characterization of floodwater in rice fields. the genetic yield potential of rainfed lowland rice varieties of l Testing of early-generation breeding lines in farmers’ eastern India, an ICAR-IRRI Collaborative Shuttle Breeding fields and participation of farmers in varietal devel- Programme was begun in 1992 mainly for shallow-water opment and selection through farmer participatory (up to 50 cm water depth) conditions. Different cooperating breeding. centers are Titabar, North Lakhimpur, and Gerua in Assam; l Identification of three photoperiod-sensitive groups Pusa and Patna in Bihar; Motto and Bhabanipatna in Orissa; for eastern India (second week of October for Chat- Raipur in Chhattisgarh; Masodha in UP; Chinsurah in WB; tishgarh, third week of October for Bihar, eastern UP, and CRRI, Cuttack, being the coordinating center. The main and Assam, and first week of November for Orissa objectives of the project are as follows (CRRI 2001d): and WB). l To make available diversified donors/improved l Donors such as Dhulia, Gangasuili, Kusuma, Ra- breeding lines suitable for rainfed lowlands for each vana, Luni-form, etc., from local varieties of Orissa cooperating center. and other improved lines such as IR38784-15-19, l To provide segregating populations with a broad IR67709-AC57-4, TCA 48 (Vaidehi), and CN genetic background to all centers for effective selec- 718-21-10-8 (Sudhir), for submergence tolerance, tion as per location-specific requirements. comparable to FR13A. l To evaluate elite breeding lines, developed by the l Suitable genotypes such as CR 683-123 (Durga), cooperating centers and IRRI, especially for sub- CR 780-1937, IR49745-CPA-42-B-1-5-1-CR1-1, mergence tolerance, photoperiod sensitivity, yield IR54112-B2-1-6-2-2-2-CR2-1, OR 1334-8, OR potential, and adaptability in eastern India. 1537-6, PSR 1119-13-3 (Kishori), NDR 96002, l To organize a breeders’ workshop for eastern India to NDR 960014, NDR SB 9730020, and NDR SB evaluate and select breeding materials at key sites. 9730018, etc., for delayed transplanting.
Breeding rice for submergence-prone and aman areas of India 47 Parents are selected through field and artificial screening (Single, double, or three-way cross) ×
P1 P2
Shuttle-breeding program State and national trials
Sent to IRRI and other Grown in shallow water breeders-collaborators F (5–20 cm) to prevent of shuttle-breeding 1 loss of genotypes program in eastern India Direct-seeded (40–70 cm Received from IRRI deep). Panicle selection on the basis of plant type, and also sent F2 F2 to IRRI and duration, better grain type, etc. collaborators
F3–F5, progeny row, direct- seeded (40–70 cm), selection for abiotic and biotic stresses
F3/F4
Bulk-transplanted and yield Sent to collaborating trial, selection against F centers and also receive 6 diseases and pests along with from them selection for other desirable characters
Yield trials
Shuttle-breeding Nomination National trial (DRR), testing network (2–3) AP Cess Fund Scheme
Multiplication and On-farm Promising entry state adaptive trials
Release
Field/on-farm
Fig. 1. Linkage between station and shuttle-breeding program at Rice Research Station, Chinsurah.
l To conduct on-farm evaluation of promising culti- standing until harvest. As a result, breeding materials during vars to study their adaptability and acceptability to generation advance have been exposed to several abiotic farmers. stresses such as drought at early vegetative stage, flooding RRS, Chinsurah, is one of the founding cooperators of of varying depth and duration at different growth stages, and this project. Initially, there were only five centers. The number biotic stresses such as aquatic weeds, insects, and diseases. of locations increased as the program became stronger. Gerua Preliminary yield trials are conducted using F6 under trans- in Assam was included as one of the cooperating centers planted conditions before replicated yield trials in F7. during kharif 2001. Figure 1 depicts the linkage between the A segregating population, mainly F2s received from station and shuttle-breeding program at RRS, Chinsurah. IRRI, is also assessed and advanced following the procedure Parents are selected through either artificial or field mentioned above. RRS, Chinsurah, shares its own F1s and F2s screening. Natural calamities such as drought in 1979 and with IRRI and other collaborators to facilitate the exchange unprecedented floods in 1978 and 2000 in WB also helped of breeding materials. It is important to mention that locally to identify donors. The F1s are grown in shallow water. From adapted traditional and long-duration cultivars were used in F2 onward, dry seeds are sown in the field during late April IRRI’s breeding program. For example, Sabita, a predominant to early May, prior to premonsoon rain. Drought commonly variety for the rainfed lowland ecosystem in WB, has been occurrs for about a month or so as the monsoon generally used in more than 50 crosses by IRRI (Mallik et al 1999). sets in during mid-June. Water starts accumulating in the field Exchange of breeding materials (F3/F4) among breeders also from mid-July and increases gradually, depending upon rain- takes place at breeders’ workshops at different sites, which fall and surface runoff from neighboring fields and remains
48 Mallik et al Motto, Orissa
CRRI, RRS, Cuttack, Chinsurah, Orissa West Bengal F3/F4
North Lakhimpur, OUAT, F /F F /F Assam 3 4 3 4 Orissa
F2 from IRRI, Philippines
AAU, NDUAT, Titabar, Masodha, Assam Uttar Pradesh
F3/F4
RAU, RAU, Patna, Pusa, Bihar Bihar Raipur, Chhattisgarh
Fig. 2. Flow of breeding materials to different sites of eastern India through shuttle-breeding program. facilitates the selection of breeding materials that are suited the project are as follows (CRRI 2001a,b,c,d): to local conditions (Fig. 2 and Table 5). l Characterization of the target environment. Promising advanced breeding lines, identified after l Collection and evaluation of germplasm for use in replicated yield trials, are nominated to national trials (DRR), breeding programs. the A.P. Cess Fund Scheme trial (described below), and shut- l Genetical, morphological, and physiological studies tle-breeding trial to consider suitability and the merits of the on submergence tolerance and adaptation to low material. Some 194 materials were nominated to the national light conditions. program for the rainfed lowland ecosystem during 1995 to l Multilocational evaluation of promising lines in the 2004 by three centers (Table 6). respective location under the target environment. Through a shuttle-breeding program, several rice l Reorientation of breeding objectives to develop plant varieties have either been released or are in the prerelease types adapted to local conditions. stage in different states (Table 7). Kishori and Satyam were Some 153 breeding lines of different generations (F3 to released in Bihar and Bhudeb in WB in 2002. OR 1206-25-1 F9), contributed by the cooperating centers, were evaluated in OUAT, Orissa; NDR 96005(IR66363-10-NDR-1-1-1-1), in six locations in five states—Bhubaneswar, OUAT and NDR 8002 (IR67493-M-2), and NDR 96006 (IR67440-15- CRRI, Cuttack; in Orissa; Pusa in Bihar; North Lakhimpur NDR-1-1-1-1) in UP; Prafulla (TTB 238-3-38-3) in Assam; in Assam; Masodha in UP; and Chinsurah in WB—through R 650-1817 and IR42342 in Chhattisgarh; and IR54112-B- the above project, and 2,995 materials were selected during 1-1-6-2-2-CR-1 and IR49745-CPA-42-B-1-5-1, developed the wet season (kharif) of 2000. Out of 153 breeding mate- at CRRI in Orissa, are in the prerelease stage (Mohanty et al rials, 142 nominations were from RRS, Chinsurah (CRRI 2000). 2001b). The number of selections made by the individual location was 1,180 at OUAT, 273 at CRRI, 452 at Pusa, 692 A.P. Cess Fund Research Scheme at North Lakhimpur, 268 at Masodha, and 130 at Chinsurah. ICAR sanctioned a scheme titled “Enhancement of genetic The maximum water depth was 160 cm at Chinsurah, 85 cm yield potential of rainfed lowland rice with emphasis on semi- at North Lakhimpur, and, for the rest of the locations, it was deepwater ecology” starting in 1997. CRRI is the coordinat- below 50 cm. The maximum number of selections (45) was ing center with six other centers, one each in Assam (North made from two crosses: IR67431-CN 7-1 (Biraj/IR53479- Lakhimpur, AAU), Bihar (Pusa, RAU), Orissa (Bhubaneswar, B-45-3-2-3) and IR72646-CN 6-1-16 (Banla Phdao/CN OUAT), eastern UP (Masodha, NDUAT), Chhattisgarh (Am- 646-6-6//IR40931-33-2-3-2), and these were in the F7 and F5 bikapur), and WB (RRS, Chinsurah). The main objectives of stage, respectively, during kharif 2000. Biraj and CN 646-6-6,
Breeding rice for submergence-prone and aman areas of India 49 Table 6. Number of nominations to different lowland trials in the DRR program. Sec 4,938 4,281 1,829 3,006 2,282 1,462 3,763 4,156 25,717 Trial Masodha CRRI Chinsurah Total Total AVT-ShW/AVT-L 6 – – 6 Evtd 3,971 7,920 7,356 4,901 3,687 4,257 2,896 3,656
38,644 IVT-ShW/IVT-L 8 – 11 19
– – – AVT-SDW – – 2 2 Sec 168 264 182 209 160 983 IVT-SDW 9 11 30 50 NSDWSN 2 24 59 85 – – lines AVT-DW – – 12 12 597 842 531 669 579 219 Advanced Evtd
3,437 IVT-DW – – 20 20 Total 25 35 134 194 2
34 57 15 Period 1998-2002 1997-2004 1995-2004 641 111 104 185 Sec
1,149 a
7 AVT = advanced variety trial; IVT = initial variety trial; NSDWSN = national F semideepwater screening nursery; ShW = shallow; SDW = semideepwater; 10 36
893 462 162 181 217 235 DW = deepwater; L = late. Evtd 2,196 544 539 187 128 186 158 118
Sec Table. 7. Promising materials identified from shut- 2,975 1,115
6 tle-breeding program in different states. F
684 902 444 273 500 281 267 State Variety/cultivar Evtd 1,259 4,610
Assam Nandang (PSR 1119), Gakapani, 767 682 221 341 421 320 145 322
Sec and Prafulla (TTB 238-3-38-3) 3,219
5 Bihar Kishori, Satyam, Shakuntala, Rajen- F dra Mahsuri-1, and Rajendra Sweta 890 572 574 855 592 299 454
Evtd Orissa Durga, Jagabandhu (OR 1206-25- 1,888 6,124 1), CRLC 899 (IR54112- B2-1- 6-2-2-2-CR 2-1), CR 2003-2 593 581 419 751 435 279 595 568
Sec (IR67638-CR-15-1-6-1-4), and OR 4,221
4 1898-8-21 F Uttar Pradesh NDR 8002 (IR67493-M-2), NDR 817 962 732 994 427 872 867
Evtd 96005 (IR66363-10-NDR-1-1-1- 1,707 7,378 1), and NDR 96006 (IR67440-15- NDR-1-1-1-1) 534 654 446 695 387 419 996
Sec West Bengal Bhudev (CN 1035-61), CN 1230- 1,033 5,164
3 12-2, IR 72035-CN 32-8, and CN F 1231-11-7 889 798 800 802 Evtd 1,956 3,106 1,186 1,185 10,722 –
522 812 532 Pusa; RAU (Bihar); and Ghagraghat, NDUAT (UP). The Sec 1,927 1,184 1,222 1,807 8,006
2 main objectives are as follows (CRRI 2001c): F – a l To test promising rice varieties in multilocation trials
748 730 229 260 106 429 with farmers’ participation. Evtd 1,675 4,177 l To analyze farmers’ perceptions on adaptability and acceptability of rice varieties. Year
Evtd = evaluated; Sec selected. l 1997 1998 1999 2000 2001 2002 2003 2004 Total a Table 5. Breeding materials Table evaluated and selections made under shuttle-breeding program from five centers—Chinsurah, Cuttack, Raipur, and Titabar—during 1997-2004. Patna, To expedite interstate flow of promising rice varie- ties. l To verify and refine location-specific technology packages for improvement of rice yield. developed from RRS, Chinsurah, were one of the parents in l To exploit production potential employing improved the two breeding lines. technology for realizing maximum yield. recently released rice varieties for the rainfed lowland On-farm evaluation of deepwater (up to 100 cm water ecosystem, from the five states, Sarala and Durga (Orissa); depth) rice varieties and production technologies in Barhaborodhi and Jallahari (UP); Rajshree and Vaidehi (Bi- the rainfed ecosystem of eastern India har); Ranjit, Bahadur, Padmapani, and Panindra (Assam); and Starting in kharif 2001, under the National Agricultural Bhudeb, Mahananda, Ambika, and Hanseshwari (WB), are Technology Project (NATP), ICAR has sanctioned a project being tested in farmers’ fields on a large scale through NATP with five centers in five eastern Indian states: CRRI, Cuttack starting in kharif 2001. Five hundred farmers in each state (Orissa); RRS, Chinsurah (WB); Titabar, AAU (Assam), (100 in one district) have been provided with 5 kg of seed
50 Mallik et al Table. 8. Promising varieties identified through varietal evaluation trials during 2001-03.
State 2001 2002 2003
Orissa Durga, Sarala, Rajshree, Durga, Sarala, Bhudeb, Durga, Ambika, and Bhudeb, and Ranjit and Rajshree Sabita Assam Ranjit, Bahadur, Bhudeb, Ranjit, Padmapani, Panin- KDML105, Panindra, and and Barh Avarodhi dra, and Bahadur Padmapani Bihar Rajshree, Mahananda, Rajshree, Mahananda, Mahananda, Ambika, and Durga, and Sarala Durga, and Vaidehi Vaidehi West Bengal Rajshree, Durga, Ranjit, Ambika, Ranjit, Bhudeb, Ambika, Hanseshwari, and Bhudeb and Rajshree and Bhudeb Uttar Pradesh Jal Lahari, Barh Ava- Barh Avarodhi, Ranjit, Jal Barh Avarodhi, Jalpriya, rodhi, Rajshree, and Lahari, and Rajshree and Jal Lahari Bhudeb
of new varieties to expedite the interstate flow of promising Table. 9. Performance of agro-management trial in farmers’ fields in rice varieties. different states during kharif (wet) season, 2003.a During 2001 to 2003, several varieties were identified as promising in different states (Table 8) through evaluation Yield (t ha–1) Increase (%) over and they were higher yielding than farmers’ varieties by State Location 10–20%. Varieties Rajshree, Ranjit, Durga, Sarala, Ambika, T1 T2 T3 T4 T1 T2 T3 and Bhudeb performed well in more than one state and Orissa 17 2.05 2.21 2.48 2.66 29.75 20.55 7.38 were accepted by farmers. One interesting finding was that Assam 24 2.85 3.50 3.37 4.01 40.70 14.57 18.99 Mahananda, a variety developed from WB, was accepted by Bihar – 1.49 1.82 1.98 2.25 51.00 23.63 13.64 the farmers of only Bihar and was found to be consistently WB 9 2.44 2.67 2.64 2.88 18.03 7.86 9.09 good in all three years. UP 40 2.26 2.59 2.77 3.02 22.70 16.37 19.38 Experiments were also conducted in farmers’ fields to Mean 2.22 2.56 2.65 2.96 32.44 16.60 13.70 a improve productivity through improved agronomic practices. T1 = local variety with traditional practice, T2 = local variety with improved practice, Improved management, which included management of the T3 = improved variety with traditional practice, and T4 = improved variety with improved practice. seedbed (seeding time, seeding rate, application of fertilizer, and seedling age) and maintaining an optimum plant popula- tion in the main field, increased production by 7.4–19.4% in different states (Table 9). Application of fertilizer (N, P, and that more productive varieties will be released for this harsh K) also increased production by 5–20% (Fig. 3) in all the and unpredictable ecosystem within the next five years, as states (CRRI 2002, 2003, 2004). they are now in an advanced stage (F7 to F9) at different centers of eastern India. The national program The materials that were nominated by different cent- ers for a semideepwater and deepwater ecosystem and were The initial effort for varietal improvement for the rainfed identified as promising through the national testing program lowland ecosystem was mainly confined to the purification during the last five years are presented in Tables 12 and 13, of landraces through pure-line selection. Most of the varieties respectively. The average yield of the promising entries in the developed up to the mid-1980s were pure-line selections. semideepwater ecology was more than 3 t ha–1 and they were Due to increased emphasis for this ecosystem by breeders superior to national, regional, and local checks in most cases during the mid-1980s, 60 varieties, 42 for shallow, 10 for (Table 12). The traditional varieties grown in this ecosystem semideep water, and 8 for deep water, were released during have mostly short bold grains with red kernels, whereas the 1991-2000 (Table 10) and 10 varieties were released during new promising materials have long to medium grains. In the 2001-04 (Table 11). deepwater ecosystem, the average yield of the promising The operation of the shuttle-breeding program in the entries was more than 2 t ha–1 though a few entries, such as early 1990s was a significant change in breeding approach. CN1234-7-34, CN1230-12-2, and CN1165-6-3, yielded more Materials with a broad genetic base were available. Exchange than 3 t ha–1 (Table 13). They were also superior to national, of breeding materials among breeders in eastern India pro- regional, and local checks. vided a better opportunity for selection of genotypes suited to local conditions. Normally, 10–15 years are needed to The state program release a variety for this ecosystem starting from the year of crossing. The shuttle-breeding program helped to reduce Considering the magnitude of the problem and potentialities, the time for varietal development by 3–4 years as breeders efforts for germplasm improvement for the deepwater eco- received F2 or even F3/F4 populations (Fig. 3). It is expected system started in the early 1970s at the Rice Research Station
Breeding rice for submergence-prone and aman areas of India 51 Yield (t ha–1) 3.5 Orissa N4O PO KO 3.0 NO PO KO N40 P2O KO N4O PO KO N40 P2O K2O N40 P2O KO West Bengal 2.5 N4O P20 K2O NO PO KO Assam N4O P2O K2O NO PO KO Uttar Pradesh 2.0 N4O PO KO NO PO KO N40 P20 KO N4O PO KO 1.5 N40 P2O K2O N4O P2O KO Bihar N4O P2O K2O NO PO KO 1.0
0.5
0 Orissa Assam Bihar West Uttar Bengal Pradesh Fig. 3. Effect of fertilizer application in deepwater ecosystem, 2003.
Table. 10. Number of varieties released (RRS), Chinsurah, and were subsequently intensified during during 1991-2000 for the lowland eco- the early 1990s through the launching of different projects. system. Crosses were made to generate new breeding materials with a broad genetic base, and also under a shuttle-breeding pro- State ShWa SDW DW Total gram, the F2s/F3s received from IRRI were exposed to real Assam 11 – 1 12 growing conditions for generation advancement, assessment, Bihar 1 2 – 3 and selection, adopting the method suggested by Mallik et al Orissa 10 – – 10 (2002). the segregating generations from F2s were exposed Uttar Pradesh – 2 2 4 to several abiotic stresses, such as drought, at the vegetative West Bengal 3 6 3 12 stage, flooding of varying depth and duration at different Other 17 – 2 19 Total 42 10 8 60 growth stages, and biotic stresses such as aquatic weeds, insect pests, and diseases in situ. One of the parents used in aShW = shallow, SDW = semideepwater, DW = deepwater. generating such F2s was generally a locally adapted variety of
Table 11. Varieties released for lowland ecosystem during 2001-04.a
Name/designation IET no. Yield range Ecosystem Resistance Grain (t ha–1) type
Hemavati 13943 2.5–3.0 DW R to BL MS ADT-44 14099 5.5–6.0 ShW R to BL, GLH, ShB SB SKL-8 – 5.0–5.5 ShW R to GM, ShB, MR to BL LS C 11-A-41 15358 5.5–6.5 ShW R to BL, GM, LF; SB MR to BLB, ShR, ShB Bapatla Sannalu 16305 5.0–5.6 ShW R to BLB MS Tholakari 16672 5.5–6.0 ShW MR to BLB, BPH MS Godavari 16673 5.5–6.0 ShW MR to BLB, BPH MS Giri 14400 4.5–5.0 ShW R to GM LS Bhudeb 14496 4.5–5.5 SDW R to BPH, GM, MS BLB; MR to ShB, ShR, Rajendra Mahsuri – 5.5–6.0 ShW MR to BPH, LF, MS BLB, ShB, BS, ShR
aShW = shallow, SDW = semideepwater, DW = deepwater, R = resistant, MR = moderately resis- tant, BL = blast, GLH = green leafhopper, ShB = sheath blight, BLB = bacterial leaf blight, BPH = brown planthopper, GM = gall midge, ShR = sheath rot, MS = medium slender, SB = short bold, LS = long slender.
52 Mallik et al Table 12. Promising entries in semideepwater (SDW) ecosystem in national trials dur- ing 1999-2004.
Yield Increase (%) overb Grain Year Designation Triala IET no. (t ha–1) typec NC RC LC
1999 PLA750 AVT1 (SDW) 16209 3.4 11.9 16.5 12.0 MS PLA10934 15514 3.1 4.3 8.6 4.6 LS 2000 OR1535-3 16473 2.7 53.6 32.0 44.5 MB OR1551-6-2 16472 2.4 39.1 19.6 30.9 MB 2001 IR53487-141-3- AVT1 (SDW) 16958 3.3 7.0 5.7 – LS 3-2-1 NDR40053-3-1 16955 3.2 2.5 1.3 – LB CRLC899 AVT2 (SDW) 16481 3.3 3.4 11.5 – LS PLA787 16488 3.1 – 6.1 – LS 2002 OR1234-12-1 AVT1 (SDW) 17318 4.3 32.5 34.5 – SB NDRSB9830121 16913 3.2 – – – LB 2003 OR1234-12-1 17318 3.5 82.6 62.8 20.0 SB 2004 NDR9930029 AVT1 (SDW) 17786 2.6 12.1 34.4 24.0 LS OR1898-8-21 17807 2.5 7.5 28.9 18.9 MS CR978-8-2 17402 2.4 5.5 26.6 16.8 MS Mean 3.1 29.1 22.2 21.4 1999 NDR4209 IVT (SDW) 16477 3.2 48.0 57.0 28.0 LS NDR40075 16237 3.2 48.0 56.0 27.0 LB 2000 NDR40053-3-1 16955 2.6 30.3 24.1 22.9 LB CN1057-114 16952 2.1 8.9 3.8 2.8 LS 2001 OR1234-12-1 17318 3.1 31.9 29.2 23.5 MB NDRSB9830121 16913 2.9 25.5 22.8 17.4 LB 2002 CN1231-11-7 17792 5.0 19.6 30.4 14.2 LS OR1550-10-2 17804 4.9 18.5 29.1 13.1 MS 2004 CN1414-1 IVT (SDW) 18130 2.5 17.2 40.8 17.9 LS IR53945-CN35-8-3 18193 2.5 15.2 38.4 15.9 MB CN1230-20-1-16-1 2.3 5.4 26.7 6.1 MS Mean 3.1 24.4 32.6 17.2 aAVT = advanced variety trial, IVT = initial variety trial, bNC = national check (Sabita), RC = regional check (Purnendu), and LC = local check (recently released best adapted variety of a particular location). cMS = medium slender, LS = long slender, LB = long bold, MB = medium bold.
Table. 13. Promising entries in deepwater (DW) ecosystem in national trials during 1999-2003.
Yield Increase (%) overb Grain Year Designation Triala IET no. (t ha–1) typea NC RC LC
1999 CN53 AVT (DW) 16115 2.6 15.4 3.5 – LB DWR4107 13493 2.5 11.0 – – LS 2000 CN1414 16847 1.2 36.0 – – LS NDGR448 16850 1.0 6.9 – – SB 2002 CN1234-7-34 17683 3.6 42.1 61.2 28.1 LS CN1230-12-2 17684 3.1 22.3 38.8 10.3 LS CN1165-6-3 17674 3.1 22.0 38.4 10.1 LS IR72035-CN32-8 17686 3.0 18.1 34.0 6.6 LS 2003 IR72035-CN32-8 17686 2.2 58.4 14.8 12.5 LS CN1230-12-2 17684 2.1 54.0 11.7 9.4 LS Mean 2.4 28.6 28.9 12.8
aAVT = advanced variety trial, bNC = national check (Jalmagna), RC = regional check (Dinesh), LC = local check (recently released best adapted variety of a particular location). cLB = long bold, SB = short bold.
Breeding rice for submergence-prone and aman areas of India 53 eastern India such as Sabita, TCA 48, Rajshree, and Mashuri. program for testing across diverse growing conditions. The breeding materials, including F1s, thus developed were They have a diverse genetic base involving 15 to 20 shared with other research centers in eastern India, including different parents from various countries, such as India, Thai- IRRI. IRRI used these materials as one of the parents in a land, Sri Lanka, Cambodia, Vietnam, and Myanmar. They breeding program as in IR76210 (Cn1205/TCA48), IR76219 also possess resistance genes inherited from Oryza nivara (CN1217/IR67632-4-1-1), IR76220 (CN1217/IR68087-55- and many useful genes for biotic and abiotic stresses and 1), and IR76221 (CN1217/TCA48). The F1 seeds of CN1205 superior grain quality from different landraces. (CRK2-6/CN1035-36) and CN1217 (CRK2-6/CN1035-15) were provided by RRS, Chinsurah. Several promising lines Conclusions having better plant type and improved sink components have been developed. They were used as parents or nominated to On-farm testing serves as a venue for visual assessment by the national program. farmers and for selecting location-specific varieties as well as To estimate yield-attributing traits and sink components, a source of widespread dissemination of seeds, which is one seven promising lines—CN1231-10-7-6-1, CN1231-16-3-1- of the major constraints to enhancing the productivity of this 1 (IR73232: IR57519-PMI-4-1-1-3-1/CN846-6-6//IR58910- ecosystem. Results have shown that new varieties coupled 202-1-3-2-2), CN1233-31-5-1-1, CN1233-12-1-1 (IR73236: with improved management and fertilizer application can IR58895-PMI-5-1-3-3/Sabita//Rajshree), IR67624-CN6-1- improve productivity by 10–20% or even more. 1-1 (IR40931-33-1-3-2/IR51089-65-1-1-3//IR43506-UBN- IR36 is the most adapted, stable, and successful variety 520-1-3-1-1), IR70242-CN36-10-3-2-1 (Kong Phlouk/ developed at IRRI for the irrigated ecosystem and it is grown IR52555-UBN-3-2-1//Sabita), and IR70418-CN11-26-2-1 widely in many rice-growing countries. Therefore, it is likely (IR66295-71-2/IR67016-45-6-3)—along with three checks, that the promising cultivars identified during this investiga- Sabita, Purnendu, and Jitendra, were sown during May 2001 tion with inherent high yield potential due to increased sink and 2002 at RRS, Chinsurah. capacity, and a diverse and broad genetic base, are expected Sabita and Purnendu were the national checks for to express greater adaptability and stability of performance the semideepwater ecosystem, while Jitendra was the only over the heterogeneous, harsh, and unpredictable environ- deepwater (>75 cm of water) variety released by CVRC. The ments of the submergence-prone ecology. two lines, CN 1231 and CN 1233, were selected from IRRI A future breeding program should emphasize the fol- F2s—IR73232 and IR73236, respectively. The mean number lowing objectives: of filled grains per panicle was more than 300 in CN1231- l Identification of suitable donors 16-3-1-1 and IR70418-CN11-26-2-1 compared with 137 in l A broad and diverse genetic base Sabita and 217 in Purnendu (Table 14). All the test cultures l Submergence tolerance had more primary (PB) and secondary (SB) branches in the l Late planting panicles to support more grains than the checks. Though l Photoperiod sensitivity the fertility percentage was higher in CN1233-31-5-1-1 l Thermo-insensitivity (92.4%), the number of filled grains per panicle was high- l Strong seed dormancy est in CN1231-16-3-1-1 (370), which is more important for l Semitall and stiff straw higher yield. The mean fertility percentage on PB was higher l Higher panicle weight, with higher grain number (85.8) than that of SB (74.2). (200) The test weight (1,000-grain weight) of the cultivars l Drought tolerance at early and flowering stages varied from 19.8 g for CN1231-16-3-1-1 to 32.9 g for Sabita. l Tolerance of major pests and diseases Out of the seven test cultivars, four have long slender grains l Higher grain yield with an L/B ratio of kernel more than 3 and the other three have medium slender grains. There is an increasing demand for varieties with improved grain quality, preferably hav- References ing medium slender to long slender grains, which can fetch premium prices in the market. Therefore, the identified elite CRRI (Central Rice Research Institute). 2001a. Annual progress report lines combining high yield with better grain type are likely of the A.P. Cess Fund research scheme for the year 2000-01. to be accepted quickly by farmers and consumers. CRRI, Cuttack, Orissa, India. p 1-54. The new cultivars identified through an evaluation test CRRI (Central Rice Research Institute). 2001b. Annual progress re- are semitall (140–145 cm), with moderate tillering ability port of the A.P. Cess Fund research scheme for 2000 for CRRI, (3–4 effective tillers per hill), possess stiff straw, and have Cuttack, Pusa, Masodha, North Lakhimpur, Chinsurah, and high yield potential (3.75–4.68 t ha–1), while current deep- Bhubaneswar, India. water varieties such as Sabita, Purnendu, and Jitendra are tall CRRI (Central Rice Research Institute). 2001c. Research proposal: (170–180 cm), susceptible to lodging, and have low yield on-farm evaluations of deep water rice varieties and produc- potential (2.21–3.38 t ha–1) with inferior sink capacity. The tion technologies in rainfed ecosystem of eastern India. CRRI, cultivars are being multiplied for nomination to the national Cuttack, Orissa, India. p 1-41.
54 Mallik et al – – CD 0.38 0.21 9.49 4.32 0.67 0.74 1.11 8.90 1.99 8.16 6.76 68.03 63.02 (0.05) 19.5 – – 2.23 3.11 3.89 4.2 6.91 Mean 242.5 177.3 77.04 14 65.20 85.76 62.7 74.20 27.23 2.09 2.84 5.62 3.25 6.0 217 158 61.82 14 59 81.94 70 56.63 20.00 1:5 1:2.68 (check) Purnendu 55 3.26 7.60 2.33 2.21 5.0 109 90.83 11 54 93.10 22 88.71 32.50 (check) 1:2 1:1.02 Jitendra 79 2.24 3.56 3.38 6.0 7.98 137 80.61 32.90 85.09 12 58 92.06 31 Sabita (check) 1:2.58 1:1.36 2.90 6.11 2.11 4.41 3.0 2-1 234 313 79.64 16 79 90.80 75 76.47 20.50 1:4.69 1:2.96 IR70418- CN11-26- 3.10 7.32 2.36 4.68 3.0 220 292 15 72 85.71 73 71.90 30.90 74.87 3-2-1 1:4.87 1:3.05 IR70242- CN36-10- 2.34 3.01 7.05 4.05 4.0 148 210 50.51 29.10 56.30 15 62 77.50 71 1:4.73 1:2.39 IR67624- CN6-1-1-1 2.40 2.94 7.05 4.37 4.0 297 239 66.29 16 58 66.67 89 66.20 29.40 1:5.56 1:4.12 12-1-1 CN1233- 3.33 2.24 4.43 4.0 7.46 219 140 90.32 31.40 92.40 14 79 96.34 43 1:3.07 1:1.77 CN1233- 31-5-1-1 2.12 2.75 3.75 5.83 3.0 302 370 82.79 19.80 84.28 15 68 90.67 83 1:5.53 1:4.44 CN1231- 16-3-1-1 2.08 3.38 7.04 4.36 4.0 261 198 77.64 25.80 78.85 15 63 82.89 70 1:4.67 1:3.14 CN1231- 10-7-6-1 a ) –1 Characters (no.) plant (no.) panicle (no.) (no.) PB = primary branch, SB secondary L length, and B breadth. a Kernel breadth (mm) Kernel L/B ratio Filled grains on SB (%) of SB Fertility PB:SB Filled grains on PB:SB weight (g) Test length (mm) Kernel Yield (t ha Effective tillers per Filled grains per (%) Fertility PB per panicle (no.) Filled grains on PB (%) of PB Fertility SB per panicle (no.) Table 14. Yield, yield components, and panicle characters of newly developed entries for the deepwater ecosystem. Table
CRRI (Central Rice Research Institute). 2001d. Annual progress report CRRI (Central Rice Research Institute). 2003. Annual progress report for eastern India rainfed lowland shuttle breeding programme for the project “On-farm evaluations of deep-water rice variet- for 2000-01. CRRI, Cuttack, Orissa, India. p 1-55. ies and production technologies in rainfed ecosystem of eastern CRRI (Central Rice Research Institute). 2002. Annual progress report India,” 2002. CRRI, Cuttack, Orissa, India. for the project “On-farm evaluations of deep-water rice variet- CRRI (Central Rice Research Institute). 2004. Annual progress report ies and production technologies in rainfed ecosystem of eastern for the project “On-farm evaluations of deep-water rice variet- India, 2001. CRRI, Cuttack, Orissa, India. ies and production technologies in rainfed ecosystem of eastern India,” 2003. CRRI, Cuttack, Orissa, India.
Breeding rice for submergence-prone and aman areas of India 55 DRR (Directorate of Rice Research). 2001a. High yielding rice va- Mallik S, Banerji C, Kundu C, Nayak DK, Chatterjee SD, Nanda rieties of India—2000. Bulletin 2001-1. DRR, Rajendranagar, PK, Ingram KT, Setter TL. 1995. Rice germplasm evaluation Hyderabad, India. p 1-102. and improvement for stagnant flooding. In: Ingram KT, editor. DRR (Directorate of Rice Research). 2001b. Draft proceedings—36th Rainfed lowland rice: agricultural research for high-risk envi- annual rice group meetings. All India Coordinated Rice Im- ronments. Manila (Philippines): International Rice Research provement Programme, Indian Council of Agricultural Research Institute. p 97-109. (ICAR), 9-12 April 2001. Mallik S, Kundu C, Mandal BK, Chatterjee SD, Sen SN, Maiti PK, DRR (Directorate of Rice Research). 2002. Draft proceedings—37th Bose S. 2003. Bhudeb, a new variety for the rainfed lowland annual rice group meetings. All India Coordinated Rice Improve- ecosystem in eastern India. Int. Rice Res. Notes 28(1):35-36. ment Programme, Indian Council of Agricultural Research, Mallik S, Mandal BK, Sen SN, Sarkarung S. 1999. Sabita, a donor for 8-10 April 2002. rainfed lowland ecosystem. Oryza 36:149-150. DRR (Directorate of Rice Research). 2003a. Draft proceedings—38th Mallik S, Mandal BK, Sen SN, Sarkarung S. 2002. Shuttle breeding: an annual rice group meetings. All India Coordinated Rice Improve- effective tool for rice varietal improvement in rainfed lowland ment Programme, Indian Council of Agricultural Research, ecosystem in eastern India. Curr. Sci. 83(9):1097-1102. 7-10 April 2003. Mohanty HK, Mallik S, Grover Anil. 2000. Curr. Sci. 78:132-137. DRR (Directorate of Rice Research). 2003b. Progress report, 2002. Vol. Prasad GSV, Prasadarao U, Sobha Rani N, Rao LVS, Pasalu IC, Mu- 1. Varietal improvement. All India Coordinated Rice Improve- ralidharan K. 2001. Indian rice varieties released in countries ment Programme (ICAR). around the world. Curr. Sci. 80(12):1508-1511. DRR (Directorate of Rice Research). 2004a. Draft proceedings—39th Ratho SN. 2004. In: Sharma SD, Prasada Rao U, editors. Ge- annual rice group meetings. All India Coordinated Rice Improve- netic improvement of rice varieties of India. New ment Programme, Indian Council of Agricultural Research, Delhi (India): Today & Tomorrow Printers and Publ., 11-14 April 2004. Part I. p 39-99. DRR (Directorate of Rice Research). 2004b. Progress report, 2003. Vol. Singh BN. 2002. High yielding rice varieties in India. Rice India 1. Varietal improvement. All India Coordinated Rice Improve- (March issue):5-6. ment Programme (ICAR). Singh VP. 2000. Sustaining rice farming in eastern India. in: Singh VP, DRR (Directorate of Rice Research). 2005. Draft proceedings—40th Singh RK, editors. Rainfed rice: a sourcebook of best practices annual rice group meetings. All India Coordinated Rice Improve- and strategies in eastern India. International Rice Research ment Programme, Indian Council of Agricultural Research, Institute, Philippines. p 41-45. 8-11 April 2005. IRRI (International Rice Research Institute). 1992. Steering committee Notes meeting–rainfed lowland rice research consortium–reports from sites for 1991. International Rice Research Institute, Manila, Authors’ addresses: S. Mallik, J. Ahmed, and S.K. Bardhan Roy, Rice Philippines. Research Station, Chinsurah, 712 102; J.N. Reddy, Central Khush GS, Baenziger PS. 1998. Crop improvement: emerging trends Rice Research Institute, Cuttack, 753 006, India; G. Atlin, in rice and wheat. In: Chopra VL, Singh RB, Varma A, editors. International Rice Research Institute, DAPO Box 7777, Metro Crop productivity and sustainability–shaping the future. New Manila, Philippines; e-mail: [email protected]; annada@ Delhi (India): Oxford and IBH Publishers. p 113-125. vsnl.net. Mallik S, Kundu C, Banerji C, Chatterjee SD, Ingram KT, Setter TL. Acknowledgment: The authors gratefully acknowledge the facilities 1995. Seedling vigor: screening, physiology, and relationship provided by the principal secretary and director of agriculture to submergence tolerance. In: Ingram KT, editor. Rainfed & ex-officio secretary, Department of Agriculture, government lowland rice: agricultural research for high-risk environments. of West Bengal. They also thank ICAR and IRRI for finan- Manila (Philippines): International Rice Research Institute. p cial support and the collaborators for their cooperation. The 111-118. help rendered by Dr. S. Islam, assistant entomologist, RRS, Mallik S.1995. Recent efforts in genotype improvement for rainfed Chinsurah, during preparation of the manuscript is also duly lowland. In: Deb DL, editor. Sustaining crop and animal produc- acknowledged. tivity–the challenge of the decade. New Delhi (India): Associated publishing co. p 37-46. Mallik S. 2000. Rainfed lowland rice research in India: perspectives and future projections. In: Sharma RD, Gahlot P, Gahlot M, editors. Advances in agricultural research in India. Dehradun (India): International Book Distributors. Vol. XIII. p 1-32.
56 Mallik et al Breeding rice for submergence-prone and aman areas of Bangladesh
M.A. Salam
Composite cross, backcross, and shuttle-breeding programs have been carried out to improve rice varieties for the submer- gence-prone and aman areas of Bangladesh. BRRI dhan33 was released in 1997 and BRRI dhan39 in 1999 as replacements for the early photoperiod-sensitive landraces. BR5226-3-2, a strongly photoperiod-sensitive line, has been isolated for delayed planting up to 15 September coinciding with the recession of floodwater. Also, BR6110-10-1-2 was found to be adaptable to tidal wetland areas up to 60-cm tidal depth. These two lines are being placed in the variety release pipeline. Under a shuttle-breeding program, IR70175-54-1-1-2-3-HR2 showed an advantage of earliness by 10 days and lodging tolerance over the standard varieties BR11 and BRRI dhan32. Also, slow-elongating IR lines have been evaluated in shallow flooded areas, but they showed poor adaptability. The evaluation of IR lines for flash-flood submergence isolated tolerant genotypes but these were prone to lodging. Moreover, 53 landraces were collected from the submergence-prone areas of the country for characterization and identification of new gene sources for submergence. Several breeding lines have been developed through composite crosses having fine grain, improved plant type, and insensitivity to photoperiod. Short-duration near-isogenic lines have been developed using the recurrent parents BR11 and Swarna under a backcross-breeding program. Another backcross program has been progressing in applying the Sub1 gene to develop varieties for flash-flood submergence.
Aman refers to two types of rice in Bangladesh: (1) trans- RLRRC collaboration planted aman (T. aman) and (2) broadcast aman (B. aman). T. aman is a synonym for rainfed lowland rice (RLR) and Five distinct breeding strategies have been used to improve B. aman for deepwater rice (DWR) or floating rice. Submer- rice varieties for the drought-prone and flood-prone RLR (T. gence is a common phenomenon in both production systems aman) growing systems of northern Bangladesh. from flash flood to complete flooding. Also, RLR suffers from drought stress. This paper therefore intends to review Strongly photoperiod-sensitive rice progress in breeding for the drought-prone and flood-prone The advantage of this type of rice is its flexibility of seeding aman rice production environments of Bangladesh. from June to August and transplanting with aged seedlings The development of varieties for unfavorable rice- (> 60 days) to cope with early-season drought. The use of growing environments such as drought and flood has been modern photoperiod-sensitive genotypes did not show a an important part of the national program of the Bangladesh yield advantage in these conditions due to drought at the Rice Research Institute (BRRI). BRRI joined the Rainfed reproductive phase. Moreover, strong photoperiod sensitiv- Lowland Rice Research Consortium (RLRRC) and the ity is a requirement for flood-prone RLR areas where aged International Fund for Agricultural Development (IFAD) seedlings are transplanted coinciding with the recession of program targeting drought-prone and flood-prone environ- floodwater from late August. Transplanting is recommended ments, respectively. Bangladesh is now a member of the up to mid-September in this environment such that flowering Consortium for Unfavorable Rice Environments (CURE) is completed in or before mid-November to avoid the pre- Working Group 2 (WG 2, flood-prone) and Bundesminister vailing low temperature (<18 °C) and to avoid late-season fur Wirtschastliche Zusammenarbeit/Federal Ministry for moisture stress. BR5226-3-2 has been identified as a superior Economic Cooperation (BMZ) for submergence tolerance. genotype over the standard variety BR22 and landraces (Table These collaborations have strengthened BRRI’s role in the 1). It is now in the pipeline for release as a variety. national programs of Bangladesh. Early photoperiod-sensitive rice These varieties are usually transplanted in early August and the crop flowers in early October before the onset of drought. Short-duration photoperiod-sensitive genotypes were evalu-
Breeding rice for submergence-prone and aman areas of Bangladesh 57 Table 1. Breeding lines in the pipeline for release as varieties for de- layed planting up to mid-September in flood-prone T. aman areas.
Genotype Height Date of Yield Remarks (cm) flowering (t ha–1)
BR5226-3-2 105 14 Nov. 4.7 Medium-bold grain BR22 (standard check) 105 13 Nov. 4.3 Good grain Talosh (local check) 115 16 Nov. 3.5 Prone to lodging
Table 2. Short-duration varieties released for the drought-prone RLR (T. aman) ecosystem.
Line Origin Growth duration Yield Remarks (days) (t ha–1)
BG850-2 Sri Lanka 118 4.5 Released in 1997 as BRRI dhan33 BR5969-3-2 Bangladesh 122 4.5 Released in 1999 as BRRI dhan39 IR33380-7-2-1-3 IRRI 122 4.5 Not released by National Seed Board
ity in the growing conditions of Bangladesh. Nonetheless, IR71075-54-1-1-2-3-HR2 was selected for evaluation in on-farm trials, the last step of the variety release pipeline (Table 3).
Population improvement Both aromatic and nonaromatic fine-grain rice varieties are popular in the RLR ecosystem of northern Bangladesh. To develop high-yielding fine-grain quality rice, a population improvement tool or composite cross was designed involving BR, IR, Basmati 370, KDML 105, and popular aromatic and nonaromatic fine-grain rice varieties of northern Bangladesh. Seven breeding populations consisting of 5–7 parental crosses were developed. The major problem was the heterogeneity of the breeding populations even in the F10 generation. Above all, two breeding populations, BR6817 and BR6818, showed promising results and several fixed lines were isolated with grain type similar to that of the popular aromatic variety Fig. 1. National Seed Board of Bangladesh released BR5969-3-2 as Kataribhog, along with improved plant type and insensitivity BRRI dhan39. to photoperiod (Table 4, Fig. 2).
Backcross breeding BR11 has been the most popular variety of the RLR eco- ated and three improved lines were successfully isolated system throughout Bangladesh. On the other hand, Swarna (Table 2). Among these lines, two were released (as BRRI is popular in the highlands of Rajshahi region. The need is dhan33 and BRRI dhan39) for the drought-prone RLR areas to develop varieties similar to BR11 and Swarna with short (Table 2, Fig. 1). growth duration. Because of long growth duration, the crop suffers from drought at the heading stage and seeding of Shuttle breeding winter crops is delayed. To develop short-duration BR11 A resource scientist of the RLRRC for Bangladesh shared 635 and Swarna, backcross breeding was employed involving breeding lines from Thailand. Screening and evaluation of four donors and four recurrent parents. This program has these lines were done for several years in the drought-prone advanced to the BC5 and near-isogenic lines (NILs) similar environment. The major disadvantages of these lines were to recurrent parents BR11, Swarna, and Dadkhani have been very broad flag leaf, long growth duration, and susceptibility successfully developed with the same grain types along with to leaf diseases. Only a few genotypes showed adaptabil- 7–10 days earlier growth duration (Table 5).
58 Salam Table 3. Advanced lines isolated from the shuttle-breeding program of RLRRC for including in the variety release pipeline, T. aman, 2003-04.
Plant Growth Yield Breeding line height (cm) duration (t ha–1) Remarks (days)
IR70175-54-1-1-2-3-HR2 95 130 4.7 Lodging tolerant, good grains (long slender) BR11 (standard check) 105 145 4.5 Popular variety of T. aman, medium-bold grain BR32 (standard check) 113 130 4.5 Lodging susceptible, medium-bold grain
Table 4. Breeding lines developed through composite crosses, T. aman, 2003-04.
Breeding population Cross combination No. of promising lines
BR6817 Basmati370/BR1867-20-1//Kataribhog/ 5 IR46288-22-6-5-8///Kataribhog/IR46288- 22-2-6-5-8// Horinandi/IR33380-7-2-1-3 BR6818 BR4970-42-1-2/Basmati 370//IR46288- 10 22-6-5-8/Basmati 370///Kataribhog/ IR46288-22-2-6-5-8//Horinandi/ IR33380-7-2-1-3
Fig. 2. Advanced lines developed with fine-quality grain through composite cross: A—BR6818-25-3-2-3, B—BR6817-5-4-5-2.
IFAD collaboration it showed yield potential of 4.30 to 4.46 t ha–1. This genotype is now in the pipeline to be released as a variety. Breeding programs concentrated on two types of flood-prone environments: (1) tidal submergence and (2) shallow flooded Shallow flooded DWR DWR. Long periods of submergence happen to occur due to a usual increase in flood level day by day in the DWR ecosystem. Tidal submergence Flood depths and patterns are highly variable within a loca- This program was launched for the nonsaline tidal wetlands tion in succeeding years. This causes vulnerability of the of southern Bangladesh. Landraces have been grown at crop and hampers the proper evaluation of breeding materi- flooding depths from 30 to 100 cm. Evaluation of breeding als. However, the infrastructure, in particular embankments, lines at 30–60-cm tidal depths showed the adaptability of roads, and road-cum-embankments, has been stabilized at several genotypes. Furthermore, BR6110-10-2-1 has been water depth within 1.0 m in some DWR-growing areas. This identified as superior to a popularly grown landrace (Table shallow flooded area of the country was targeted to evaluate 6). Its adaptability was verified up to 60-cm tidal depth and 50 improved slow-elongating lines from IRRI. The top five
Breeding rice for submergence-prone and aman areas of Bangladesh 59 Table 5. Generation of NILs after selfing of BC5F1, T. aman, 2004.
Original cross Recurrent Characters parent
Swarna/BR6389-12-2 Swarna Swarna type but short duration BR11/Minikit BR11 BR11 type but short duration Rajshahi1042/BR6398-12-2 BR6398-12-2 Short duration with long slender grain Dadkhani/BR6398-12-2 BR6398-12-2 Improved plant type but dadkhani grain
Table 6. Yield and growth duration of BR6110-10-1-2 in variety trials in nonsaline tidal wetlands, T. aman, 2002-03.
Genotype Yield (t ha–1)
Pirozpura Barisal Bakerganj Jhalakathi Mean
BR6110-10-1-2 5.7 5.5 5.8 4.0 5.3 (148) (147) (144) (148) (147) BR11 (check) 3.6 4.6 4.0 3.4 3.9 (146) (145) (144) (145) (145) Moulata (local check) 2.24 2.42 2.19 2.40 2.31 (160) (160) (160) (160) (160)
aNumbers within parentheses are days to maturity.
Flooding depth (cm) 70
60 Maligram Gangabordi Shingria Suterkandi 50 Mazikanda Barokota Vennabari Nagbari 40
30
20
10
0
1 1 1 1 1 1 1 1
01 01 01 01 01 01 01 01
-0 -0 -0 -0
X- X- X- X- V- V-
XI-01 XI-01 XI-01 XI-01 XI-01
VI-0 VI-0 VI-0 VI-0
VII-01 VII- VII-
VIII-01 VIII-01 VIII-01 VIII-01 VIII-01 VIII-01 VIII-01
5- 9-
9- 7-IX 5- 9-
5-
5- 12- 22- 14- 29-
1- 6- 9-
17-IX 22-IX 25-IX 19- 24-
11- 17- 25-
10- 15-
26- 11- 15- 22-
Date of flooding Fig. 3. Flooding patterns in experimental fields, IFAD-DWR trial sites, 2001. lines were evaluated in an on-farm trial at 10 locations of BMZ and CURE collaboration the shallow flooded areas. The flooding patterns at trial sites showed a wide variation in depth and longevity of the flood- This program centered on breeding rice for flash-flood sub- ing period (Fig. 3). These lines showed some yield advantage mergence and medium stagnant water environments. over landraces and had maturity duration similar to that of the local varieties (Table 7). The adaptability of these lines Shuttle breeding is still in question because of their inability to cope with sub- A total of 700 F7 lines and 24 advanced lines from IRRI were mergence at high rates of increasing flood depth due to slow evaluated in an on-station trial in T. aman 2003. Thirty-day- elongation capacity. Therefore, introgression of submergence old seedlings were transplanted and the crop was submerged tolerance into these lines has been suggested. for 10 days after 3 days of transplanting. The lines that re- covered from submergence were tall in stature and prone to
60 Salam Table 7. Yield potential of slow-elongating lines at IFAD sites, DWR (B. aman), 2001-02.a
Genotypes Yield (t ha–1)
L1 L2 L3 L4 L5 L6 L7 L8 L9 L10 Mean
IR64588-47-3-2-2B- 2.3 2.2 2.1 2.4 2.5 2.4 2.8 4.5 1.7 1.1 2.5 12-1-2-3 IR64588-47-3-2-2B- 2.2 3.0 1.9 1.8 2.7 2.0 2.7 2.5 1.6 1.5 2.3 9-2-2-3 IR60436-B-65-2 1.5 2.2 1.2 1.2 2.0 2.2 2.8 1.5 1.2 1.5 1.7 PCR89114-B-R-2- 2.0 2.5 1.9 1.0 2.3 2.4 2.7 1.9 1.3 1.5 2.0 2-2-1 IR62653-8-3-3 1.6 1.8 1.4 0.9 2.1 1.9 2.5 1.8 1.4 0.9 1.6 Farmers’ variety 1.8a 2.3b 1.8c 1.2d 2.5e 2.7f 1.8g 1.5h 1.2i 1.0j 1.8
aL1 (Maligram), L2 (Shingria), and L3 (Mazikanda) under Faridpur District; L4 (Vennabari), L5 (Gangabordi), and L6 (Suterkandi) under Madaripur District; L7 (Duepara-1) and L8 (Duepara-2) under Tangail District; L9 (Nagbari) and L10 (Barokota) under Comilla District.
a = Laxmidigha, b = Laxmidigha, c = Laldigha, d = Karticksail, e = Gabura, f = Karticksail, g = Chamara, h = Chamara, i = Chamara, and j = Hijoli khama.
Table 8. List of selected genotypes tolerant of flash-flood submergence, T. aman, 2004.
Seedling height (cm)
Designation Before After Phenotypic score submergence submergence Elongation Survival at flowering (30-day-old (15-days’ (%) seedlings) recovery)
IR66036-3B-12-2-B 32 55 23 94 2 IR66036-3B-13-2-B 28 48 20 94 2 IR7518-B-11-2-B 31 47 16 97 3 IR7518-B-11-3-B 31 54 23 96 3 IR7518-B1-3-B 34 54 20 97 3 IR75407-R-R-R-R-5 34 47 13 94 2 IR75407-R-R-R-R-7 34 47 13 93 2 IR75407-R-R-R-R-8 28 54 26 88 3 IR75407-R-R-R-R-10 33 47 14 94 3 IR75407-R-R-R-R-11 33 50 17 94 3 FR13A (check) 41 58 17 99 2
Table 9. List of BC1F1s to generate BC2F2s for marker-assisted selection, breeding for submergence tolerance, T. aman, 2005.
Original cross Recurrent parent No. of Target no. of seeds
BC1F1 seeds for BC2F1
IR67518-B-1-2-B/BR11 BR11 52 300 IR67520-B-14-3-4/BR11 BR11 149 300 lodging at maturity. However, 27 lines were selected with germplasm presently grown in the submergence-prone and relatively short plant height. These lines were evaluated under medium stagnant water environments of the country. In total, controlled submergence in an on-station trial (Table 8). The 53 landraces were collected from 10 districts. best lines will be evaluated for yield potential. Application of marker-assisted selection Collection of landraces Two advanced lines were selected from IRRI lines to gener- A mission was launched in cooperation with the Depart- ate BC2F2 populations for the application of marker-assisted ment of Agricultural Extension (DAE) to collect the rice selection (MAS) for the Sub1 gene. IR67518-B-1-2-B and
Breeding rice for submergence-prone and aman areas of Bangladesh 61 IR67520-B-14-3-4 were used, involving Godaheenati and Kurkaruppan-1 as sources of Sub1, respectively. These lines were crossed with BR11, the most popular T. aman variety in Bangladesh. BC1F1s were developed using BR11 as a recurrent parent (Table 9). The development of flash-flood submergence-tolerant rice varieties has been a research priority at BRRI. Efforts of conventional and marker-assisted breeding have been emphasized. The program for introgression of Sub1 into mega-varieties is strengthened to increase the productivity of submergence-prone areas in Bangladesh.
Notes Author’s address: Chief scientific officer, Plant Breeding Division, BRRI, Bangladesh.
62 Salam Tidal wetlands/ problem-soil ecosystems
Rice breeding for the tidal wetlands of Indonesia 63 64 Sulaiman et al Rice breeding for the tidal wetlands of Indonesia
S. Sulaiman, I. Khairullah, and T. Alihamsyah
Tidal wetlands or tidal swamplands are becoming important swamplands, because it is cheaper and easier to be adopted land resources for agricultural production and employment by farmers once suitable varieties are developed. This paper opportunity in Indonesia because of the recent increase in presents an overview of the rice-growing environment and human population and the increasing conversion of fertile cultivation systems, the rice breeding strategy and programs, lands to other nonagricultural purposes. Of the approximately progress made so far through rice breeding, and the current 9 million hectares of tidal swamplands considered suitable challenges and future directions for rice breeding for the tidal for agricultural use, about 4,200,000 hectares have already swamplands of Indonesia. been reclaimed (Widjaya Adhi et al 1992, Direktorat Bina Rehabilitasi dan Pengembangan Lahan 1995). Some tidal Rice-growing environments and cropping systems swampy areas have been cultivated for many years with rice and a range of upland crops as well as coconut, coffee, and Rice-growing environments citrus. Single cropping of rice using traditional photoperiod- Rice cultivation in tidal swampland areas is mainly along sensitive varieties (6–8 months) is commonly practiced by coastal areas where the sea tides fluctuate through rivers and farmers in these areas; however, rice grain yield is very low, canals in the fields during part or all of the growing season. usually varying from 1 to 2.5 tons per hectare. Moreover, the The high tides normally cover the highest part of the land. adoption of improved high-yielding, early-maturing, and pho- These areas are characterized by shallow inundation in the toinsensitive varieties is still very low in this ecosystem. wet season, caused mainly by stagnant rainwater. Tidal Farmers in tidal swamplands face various problems swampland of Indonesia occurs mostly along the coasts or in developing an efficient rice production system. These the large rivers of Sumatra, Kalimantan, Sulawesi, and Irian problems include poor water control facilities, environmental Jaya. The water regime in these areas is dominated by daily problems (low soil pH, iron and aluminum toxicity, peat soil, tidal fluctuation of the rivers and water depth is influenced salinity, deepwater flooding, fluctuating water regimes), and by the tide and rainfall. Based on the influence of the tides, socioeconomic problems such as a lack of labor and capital. tidal swamplands can be divided into four types (Noorsyamsi To support agricultural development in tidal swampland ar- et al 1984, Widjaya Adhi 1986): (1) the area with tidal type eas, the government of Indonesia has made various efforts, A is flooded by tidal water during spring and neap tides, (2) including developing drainage facilities and other physical the area with tidal type B is flooded by tidal water only dur- infrastructure, and funding some research activities (Aliham- ing spring tide, (3) the area with tidal type C is not flooded syah 2004). Several research institutions worked to develop by tidal water but its groundwater table is less than 50 cm effective management options for various aspects of this deep, and (4) the area with tidal type D is also not affected ecosystem such as proper soil, water, and crop management by tidal water but its groundwater table is more than 50 cm strategies as well as rice varietal improvement. Some research deep. These tidal types are illustrated in Figure 1. outcomes have also been tested and are being adopted by Brown and Sulaiman (1984) divided tidal swampland farmers in these areas. into four major groups based on prevailing abiotic stresses: Rice production in tidal swampland areas can be in- acid-sulfate land, non-acid-sulfate land, peat land, and saline creased by practicing proper land reclamation techniques land. The majority of acid-sulfate lands in Indonesia fall into and growing improved tolerant high-yielding varieties. the soil family Sulfic Tropo Fluaquent and they can be Sul- However, the application of land reclamation techniques in faquents for the potential acid-sulfate soils and Sulfaquepts tidal swampland is costly and needs various inputs such as for the actual acid-sulfate soils. These soils are also character- soil amelioration material and fertilizer, because of the low ized by pH of less than 4.5, high iron and aluminum contents, soil quality and lack of water control. Therefore, the adop- and low base saturation, which could cause problems for rice. tion of land reclamation techniques by farmers is progress- Non-acid-sulfate lands occur in several soil groups, including ing very slowly. This condition makes the use of improved extensively cultivated Troposulfaquents, Tropo Fluaquents, stress-tolerant and high-yielding varieties a very effective and and ferrolyzed soils, with soil pH of 4.5 or higher. These soils important way to increase and sustain rice production in tidal normally suffer from phosphorus deficiency and low nutri-
Rice breeding for the tidal wetlands of Indonesia 65 II C B A I B C D 50 cm Tidal maximum ---
50 cm
Flufiatil
Peat
Minimum
Marine
Fig. 1. Illustration of tidal types in swampy areas. ent status, which could hinder rice production. Peat soils are Disease incidence is normally not serious under the characterized by high organic matter or a peat layer of more existing single rice-cropping pattern using traditional local than 20 cm thickness. In soil taxonomy, this soil is mainly rice varieties. However, rats are considered to be one of classified as Tropaquepts. It could be peat land (peat layer the most serious pests for rice in many tidal swampy areas. around 50 cm thick) and peaty land (peat layer <50 cm thick). The most important diseases observed in farmers’ fields The major problems of these soils are low base saturation, are brown spot, Cercospora leaf spot, bacterial leaf blight, organic acid toxicities, and poor root anchorage. Organic sheath blight, and blast. Tungro virus occasionally causes layers are often underlain by potential acid-sulfate soils, moderate damage at a few sites in South Kalimantan. Grassy which may also contribute to iron and aluminum toxicities. stunt and ragged stunt viruses are less common than tungro Saline lands occur mostly in coastal areas that are subject to under single rice-cropping, but they should be recognized sea-water intrusion. as potentially serious under a more intensive two-rice-crop Similar to the above classification and for practical pur- production system. Stem borer, rice bug, green leafhopper, poses, Widjaya Adhi (1986) grouped tidal swamplands into and brown planthopper are considered potential problems in four land types: deep acid-sulfate lands, shallow acid-sulfate these areas. lands, peat lands, and saline lands. Deep acid-sulfate lands are The socioeconomic environment in the tidal swamps ones having a pyrite (FeS2) layer with a concentration of less also affects the breeding program. Farmers’ incomes are than 2% at a depth of more than 50 cm. Shallow acid-sulfate generally low and farmers use little or no external inputs lands are ones having a pyrite layer with a concentration of in their rice production systems. Available farm labor and greater than 2% at a depth of less than 50 cm. Peat lands are capital are also limited and seasonal. To raise farmers’ income lands having a peat layer in their surface horizon with more by introducing improved rice varieties will require that the than 20 cm thickness, whereas saline lands are lands subject increased yields of the improved varieties be more than the to salt-water intrusion for more than 3 consecutive months added costs of fertilizer, pesticide, and labor. in a year. Based on land conditions, the main problems for de- Cropping systems veloping rice production systems in tidal swampland areas Single cropping of rice using long-duration varieties is include high soil acidity and toxic element concentrations, commonly practiced. By using photoperiod-sensitive rice especially iron and aluminum, fluctuating water regimes and a varieties, farmers can expand the period of land preparation lack of water control facilities, water flooding or submergence or transplanting activity, which helps them to use their own during the rainy season, and drought during the dry season. labor resources and reduce the need for hired labor. In addi- The main problems for practicing double rice-cropping are tion, local rice varieties do not require high external inputs rat and bird damage for the first crop, poor water control and have good grain cooking and eating quality with a high facilities, and limited farmer capital and labor. Research is price and low risk of failure. When the water level is not too still needed for developing better varieties and management high, early-maturing improved rice varieties can be grown practices for rice double-cropping systems in tidal swamp- as the first crop, followed by a second planting using local lands. varieties. Only a few farmers practice double- or triple- cropping systems because of poor water control facilities in
66 Sulaiman et al most tidal swampland areas and limited labor and capital. short-maturing tolerant rice variety and a local photoperiod- Inorganic fertilizers or insecticides are rarely applied. On sensitive rice variety. Seeds of both varieties are usually sown the other hand, incorporating grassy weeds in the soil during on dry seedbeds at the beginning of the rainy season (around land preparation provides some organic nutrients. Weeds are October). Twenty to 25 days after sowing the improved vari- not a major problem since farmers grow traditional varieties, ety and 40 days after sowing the local variety, the seedlings which are tall and have droopy leaves, besides the effect of are transplanted in fields that were previously well prepared flooding, which helps in weed control. using a hoe or a tractor. The area covered by the improved According to Alihamsyah (2004), farmers practice rice variety is about 80% of the rice field and that of the local four rice cultivation systems in tidal swampland areas: tradi- variety about 20%. After the improved variety is harvested, tional transplanting, improved transplanting, a combination usually in February or March, seedlings of the local variety between traditional and improved transplanting, and direct are re-transplanted over the whole field. The local rice variety seeding. The traditional transplanting system is practiced is usually harvested in July or August. Therefore, the crop- by the majority of farmers in tidal swampland areas, espe- ping intensity in this rice cultivation system is 180%. This cially local farmers such as in Banjarese and Bugise. This system allows crop intensification without greatly altering transplanting system uses local photoperiod-sensitive rice the traditional system, so that it could be easily adopted by varieties, which are mostly adapted to local conditions. A farmers. high water level during the early vegetative stage requires The direct-seeded rice system is usually practiced by seedlings that can be transplanted up to 3 times at intervals some farmers using improved rice varieties in areas having of 40 to 50 days to produce seedlings that are tall and strong a good water management system or in areas with C and D enough to withstand the high water after the third transplant- tidal type (Fig. 1) as well as in shallow freshwater swamp- ing in the main field. lands. In this rice cultivation system, rice seeds are sown Rice seeds at the rate of 5 kg of seed per 150 m2 of land in well-prepared wet or dry soils. In wetland conditions, are usually sown on dry seedbeds (locally called teradakan) the rice seeds are broadcast by hand in the field, whereas at the beginning of the wet season (October-November). After in dryland sowing, the seeds are sown in a hole made by a 30–40 days, the seedlings are transplanted to the lower parts wooden stick. Rice seeds at 25–30 kg ha–1 are usually sown of the rice field (locally called ampakan). Forty days later, at the beginning of the rainy season (around October) and the seedlings are transplanted again on larger areas in the the crop is harvested in February or March. Fields are kept field (locally called lacakan), which still cover only about flooded for up to 4 weeks before harvest, after which the crop one-third of the total field to be planted. The time for final is harvested. Farmers usually use an improved transplanting transplanting depends on the water level in the main rice system so that they can grow two rice crops in a year. field but this is mostly done around 40 days after the second transplanting. Most farmers do not apply any kind of chemical Rice breeding for tidal wetlands fertilizer and this traditional rice cultivation system seems to be suitable for these tidal swampland situations. The system Rice breeding strategy and priorities also has some advantages, such as preventing the oxidation The rice breeding program for tidal swamplands should be of pyrite, alleviating iron toxicity problems, escaping from intensified in order to support rice production in these highly seedling submergence, reducing fertilizer application, and unexploited lands. Based on the specific local soil problems spreading the labor requirements for farming activities. Rice and farmers’ preferences and constraints, a rice varietal im- yields for this cultivation system are reported to be about provement program should focus on (1) developing improved 1.5–3 t ha–1. rice varieties to replace the current long-duration local rice The improved rice transplanting system is mostly prac- varieties and (2) developing improved rice varieties with ticed by some trans-immigrant farmers in several swampland better performance than the existing high-yielding early-ma- areas after reclaiming their agricultural lands and applying turity rice varieties, with particular focus on their adaptability proper water management techniques, besides using im- to local conditions and acceptability to farmers. The varietal proved tolerant rice varieties and cultural practices. Rice improvement program for tidal swamplands should focus on seeds at 25–30 kg ha–1 are usually sown on dry seedbeds at the developing high-yielding varieties that are more adapted to beginning of the rainy season (around October). After 20–25 fluctuating water regimes and transient submergence, and the days, seedlings are transplanted in rice fields previously well various soil problems, especially high soil acidity and iron prepared using hand hoes or tractors. The improved tolerant and aluminum toxicity. rice varieties usually mature within 120 to 140 days, with An effective breeding program for these diverse ecosys- a grain yield of 4–6 t ha–1. By using this rice transplanting tems needs to be target-specific to develop varieties that can system, the farmers can grow two rice crops each year. The tolerate the existing abiotic stresses and meet local farmers’ first planting period is October-February or March and the needs. Of particular importance are crop duration to fit in the second planting period is April-August. existing cropping patterns and local farmers’ preferences. The combination of traditional and improved trans- Target ecosystems are grouped by soil type rather than by planting systems (locally called sawitdupa) uses an improved water regime because water problems are common across all
Rice breeding for the tidal wetlands of Indonesia 67 sites. There are four target ecosystems: saline soils, acid-sul- l Good eating and cooking quality with high amylose fate soils, non-acid soils, and organic soils. The first priority content similar to that of local varieties to suit the should be given to acid-sulfate soils and organic soils because taste of local people, and intermediate amylose of their relatively large area and their occupation mostly by content for transmigration farmers. transmigration farmers. In term of duration, priority should be given to early (105–115 days) and intermediate (115–135 Achievements in rice breeding days) maturing varieties. These varieties could then be used in double or triple rice-cropping systems, whereas the improved Released varieties high-yielding photoperiod-sensitive varieties can replace Good progress has been achieved in developing varieties the existing local rice varieties for the single rice-cropping adapted to this ecosystem. Between 1981 and 2003, the system. government of Indonesia, through the Department of Agri- culture, released 18 improved rice varieties adapted to tidal Rice breeding program and implementation swampland conditions (Table 1). However, most of these Developing a suitable variety is an important component for varieties are only tolerant of certain problems encountered enhancing the productivity of the swampland rice farming in these areas, such as high soil acidity and iron toxicity system. The varietal improvement program for tidal swamp- under good management practices. Most of the improved land is being handled by breeders at the Indonesian Center rice varieties also have low or medium texture and cooking for Rice Research (ICRR) in Sukamandi, West Java, and is quality, with large grain size, and are suited only for certain mostly focused on developing high-yielding and early- or farmers or consumers. Some of those rice varieties, such intermediate-maturity rice varieties with resistance to pests as Kapuas, Lematang, Lalan, Batanghari, and Banyuasin, as well as iron toxicity and soil acidity. Breeding of early- have been grown by farmers but in limited areas. The slow maturity rice varieties with higher yield and better adaptation adoption of those improved rice varieties was attributed to to tidal swampland conditions has been handled by research- insufficient seed availability, higher input and labor require- ers at the Research Institute for Food Crops in Swamp ments, and inadequate quality to meet farmers’ preferences. Areas since 1993. The breeding program aimed to develop For example, Banjarese farmers, especially in Central and acceptable rice varieties that meet the need of local farmers South Kalimantan, prefer a tall plant with long and slender and consumers, especially in Central and South Kalimantan. grain and medium rice texture or cooking quality. Therefore, the adoption of those rice varieties by farmers in Margasari and Martapura are two rice varieties devel- tidal swamp areas is expected to be faster. oped from crosses between the local rice variety Siam Unus Breeding varieties for tidal swamplands is being with Cisokan and Siam Unus with Dodokan, respectively, conducted under the Indonesian Agency for Agricultural in order to meet the needs of local farmers and consumers, Research and Development and this involves two institu- especially in Central and South Kalimantan. The major tions, the Research Institute for Food Crops in Swamp Areas characteristics of those rice varieties (Table 2) are similar to (Banjarbaru, South Kalimantan) and the Indonesian Center those of local rice varieties, such as slender and long grain for Rice Research (Sukamandi, West Java). Some research with high amylose content (27–28%) or medium texture, activities are being handled in collaboration with the Inter- moderately tall plants (90–130 cm), moderately leafy plants, national Rice Research Institute, especially in the exchange and tolerance of iron toxicity. These characters seem to match of germplasm for testing in target areas. The following traits the requirements of the local farmers and consumers. These should be considered in developing rice varieties for tidal varieties are expected to be easily adopted by farmers, with a swampland rice-farming systems: yield potential of 3–4 t ha–1, and shorter duration of 120–125 l Moderately tall varieties (90–130 cm) that are leafy days compared with the yield of local varieties of 1–3 t ha–1 to reduce the labor required for weeding. and a much longer duration of 6 months. In 2002, the local l Perform well under low fertilizer input. extension personnel together with farmers successfully tested l Easily threshed, without awns and with good panicle a double-cropping system using Margasari and Martapura as exsertion. the first crop on about 400 ha of rice fields in Barito Kuala l Have some amount of submergence tolerance be- District, South Kalimantan. The farmers obtained a rice yield cause the crop is occasionally submerged, especially of about 4 t ha–1. Tidal swampland areas planted with those at the seedling stage. varieties are progressively increasing, especially in the south l Resistance to common rice diseases and insects. and central parts of Kalimantan. It is now estimated that more l Tolerance of drought for varieties planted as a second than 1,000 ha of swampland area in Central and South Ka- crop in the dry season. limantan are being planted with Margasari and Martapura. l Tolerance of adverse soil conditions and locally adapted. This may be sought in crosses with tradi- Promising rice lines for adverse soil conditions tional local rice varieties. As mentioned in an earlier section, the adverse soil chemi- l Have small slender grain (local variety type) or cal characteristics that constrain rice cultivation in tidal medium slender grain (IR66 type). swamplands include a high concentration of soluble iron
68 Sulaiman et al Table 1. Improved rice varieties adapted to swampland released during 1981–2001.a
Year Maturity Yield Rice Fe Resistant to Name released (days) (t ha–1) texture toxicity BPH SB BS BL BB PB
Barito 1981 140–145 3 Loose – R-1 MR – – MR – Mahakam 1 1983 135–140 3–4 Loose – S-123 MR – – MR – Kapuas 1 1984 127 4–5 Medium R R-1 MR – – MR – Musi 1988 135–140 4–5 Loose – R-2 R – R R – SeiLilin 1991 115–125 4–6 Loose – MR-2 – – MS – – Lematang 1991 125–130 4–6 Loose – R-1 – – MR – – Lalan 1997 125–130 4–6 Loose – R-123 – – R – – Banyuasin 1997 115–120 4–6 Sticky MR R-3 – R R – – Batanghari 1999 125 4–6 Loose R R-12 R – R R – Dendang 1999 125 3–5 Sticky R R-12 – MR MR – – Indragiri 2000 117 4.5– Medium R R-2 R – R R – 5.5 Punggur 2000 117 4.5–5 Medium R R-23 – – R – – Margasari 1 2000 120–125 3–4 Medium R MR-2 – – R – R Martapura 1 2000 120–125 3–4 Medium R MS-2 – – R – MR Air Tenggulang 2001 125 5 Loose R R-123 R – R R – Siak Raya 2001 125 5 Loose R R-2 – R R – – Lambur 2001 120 4 Sticky R R-2 – – R – – Mendawak 2001 115 4 Sticky R MR–3 – – MR – –
aResults by Research Institute for Food Crops in Swamp Areas, Banjarbaru. bR = resistant, MR = moderately resistant, MS = moderately susceptible, S = susceptible, BPH = brown planthopper, SB = sheath blight, BL = blast, BS = brown spot, BB = bacterial blight, PB = panicle blast. Source: Lesmana et al (2002).
Table 2. Some characteristics of Margasari and Martapura. swamplands of South Sumatera, three promising rice lines were developed with yield significantly higher than that Characteristic Margasari Martapura of Lematang (Tumarlan et al 2000): IR48948-B-2-Mr-1, IR51471-2B-2-2B-Mr-1, and B-655f-KA-141-6, with mean Age at maturity 120–125 days 120–125 days –1 Plant type Moderately leafy Moderately leafy grain yield of 4.5, 4.0, and 3.8 t ha , respectively, whereas –1 Plant height 120–130 cm 120–130 cm Lematang yielded only 3.0 t ha . Tillering ability Medium Medium Rice breeding activities at RIFSA or RISA (Banjarbaru) Brown rice shape Slender-small Slender-small mostly focus on developing photosensitive rice varieties Panicle threshability Easy Easy for acid-sulfate and peat tidal swamplands. From a series Panicle exsertion Well exserted Well exserted Lodging resistance Moderately resistant Moderately resistant of breeding activities in Kalimantan, breeders at RIFSA or 1,000-grain weight 21 g 22 g RISA (Banjarbaru) identified some promising rice breeding Amylose content 27% 28% lines adapted to tidal swampland conditions, particularly with Cooking-rice structure Medium Medium tolerance of soil acidity and iron toxicity. Eight promising Eating quality Good Good rice lines were identified: GH47, GH137, GH149, GH173, Grain yield 3–4 t ha–1 3–4 t ha–1 Resistance to GH460, GH493, GH505, and GH591. However, these Bph2 Moderately resistant Moderately resistant eight lines are susceptible to brown planthopper biotype 2, Sheath blight Moderately susceptible Moderately resistant moderately resistant to blast disease (except GH460), and Panicle blast Resistant Moderately resistant moderately susceptible to moderately resistant to sheath Fe toxicity Tolerant Tolerant blight disease (Table 3). They have intermediate plant height, with long slender grains, high amylose content, medium rice texture, and tolerance of Fe toxicity, so they are considered suitable for the local conditions of South and Central Kali- and aluminum, salinity, and soil acidity. New rice varieties mantan. should therefore have tolerance of those adverse soil char- For the shallow acid-sulfate soils of Belandean and acteristics. From a series of breeding activities conducted at Oanda Jaya, most of the 15 lines tested were found to the Bogor Research Institute for Food Crops (BORIF), some perform well, with tolerance of iron toxicity. However, promising rice lines adapted to tidal swampland conditions of these lines, only four produced higher grain yield than were identified, particularly with tolerance of soil acidity Margasari and IR64 (Table 4). Those are GH173, IR53709- and iron and aluminum toxicity. For the deep acid-sulfate 36-10-2, Tox3118-B-E2-3-2, and B10179B-MR-1-4-1. In
Rice breeding for the tidal wetlands of Indonesia 69 Table 3. Characteristics of promising rice lines selected for tidal swamplands. Data are from RIFSA.
Code of Plant height Maturity Brown rice Amylose content Resistancea to pests lines (cm) (days) (%) Fe toxicity Length L/W ratio Bph2 SB PB Tungro
GH47 126 120 M Slender 28 T – – – MS GH137 112 118 M Slender 28 T S MR R S GH149 116 123 M Slender 27 T S MS MR S GH173 124 122 M Slender 27 T S MR MR S GH460 114 123 M Slender 27 T S MR S S GH493 110 124 M Slender 29 T S MS MR MS GH505 113 123 M Slender 29 T S MS R S GH591 116 121 M Slender 28 T S MR S S
aR = resistant, MR = moderately resistant, MS = moderately susceptible, S = susceptible, T = tolerant, M = medium. Source: RISA (2002).
Table 4. Yield and Fe toxicity scores of rice lines in shallow acid-sul- long slender grains and medium rice texture, which match a fate soils of Belandean and Oanda Jaya, South Kalimantan, 2002. local farmers’ preferences. Data on genotype by environment interaction from trials Belandean Oanda Jaya Line conducted in some tidal swampland sites of South Sumatera Yield Fe toxicity Yield Fe toxicity and Central and South Kalimantan showed a significant inter- (t ha–1) score (t ha–1) score action between rice genotype and environment. This indicated that some of the promising lines were well adapted only in GH47 (KaI9407d-Bj-18-2) 3.0 1 3.0 cde 1 certain locations or soil conditions. Soil pH at Kalimantan GH137 (KaI9408d-Bj-28-4) 3.1 2 3.8 def 1 GH149 (KaI9408d-Bj-38-1) 3.0 2 3.1 abc 2 locations is lower (pH 3.3–3.8) than at Sumatera locations GH173 (KaI9408d-Bj-70-4) 3.8* 1 4.0 efg 1 (4.6–4.8), and this affects crop performance. Based on the GH460 (KaI9414d-Bj-110- 3.3 2 3.1 abc 3 average yield of the six locations, five promising rice lines 1) produced grain yield at least 10% higher than the yield of GH493 (KaI9420d-Bj-6-2) 2.9 2 3.6 cde 3 Margasari (Table 6). Those lines are GH47, Tox3118B-E- GH505 (KaI9420d-Bj-14-1) 3.4 2 3.5 bc 3 GH591 (KaI9420d-Bj-11 2.8 3 3.4 bc 2 2-3-2, IR61242-3B-B-2, IR58511-4B-4, and B9852E-35- 0-2) KA-66. IR61242-3B-B-2 was the only line that yielded more IR53709-36-10-2 3.5 3 4.4 fg 3 than 4 t ha–1. However, GH47 was considered to be the most BW307-6 3.3 3 3.3 bcd 3 stable high-yielding rice line. B10278B-MR-3-3-1 2.7 3 3.4 bc 3 Based on farmers’ preferences, GH47 was selected by B10179B-MR-1-4-2 2.6 5 2.9 ab 3 B 1 0277B-MR-1-4-3 2.6 5 2.6 a 3 farmers at three locations in Kalimantan and one location TOX3118B-E-2-3-2 3.1 3 5.0 h 2 in Sumatera, while GH137 was selected by farmers at three B10179B-MR-1-4-1 3.2 3 4.4 fg 2 locations in Kalimantan and two locations in Sumatera. Margasari (check) 3.0 1 3.6 cde 3 IR61242-3B-B-2 was selected at three locations (Sumatera). IR64 (check) 2.3* 5 2.9 ab 4 Tox3118B-E2-3-2 and IR58511-4B-4 were selected by Kapuas (check) 2.8 1 4.4 9 3 farmers at only one location in Sumatera. GH47 and GH137 ans = nonsignificant, * = significant at the P = 0.05 level. Means followed by have intermediate plant height and small-slender grain type, the same letters are not significantly different according to Duncan’s multiple range test. while Tox3118-B-E2-3-2, IR61242-3B-B-2, and IR58511- Source: RISA (2003). 4B-4 are semidwarf with long-slender grain type. GH137 was selected at five locations because of its high yield but lower stability. This indicates that farmers’ preferences for different lines vary across locations. deep acid-sulfate soils of Handil Manarap, no iron toxicity Promising rice lines with submergence tolerance was observed, and comparable rice yields were given by Flooding or submergence occurs in tidal swamplands with GH47, GH149, GH460, BW307-6, and B10277B-MR-1-4-3. tidal type A and B along the rivers. Until 2003, breeding Results of field trials in three tidal swampland sites showed to enhance submergence tolerance for tidal swamplands of that only four rice lines have high tolerance of acid-sulfate Indonesia was very limited, and not much information was soils and grain yield of more than 3 t ha–1 (Table 5). Those available. A submergence trial was conducted in an artificial lines are GH47, GH137, TOX3118B-E-2-3-2, and IR58511- pond at RISA, Banjarbaru, during the 2003-04 rainy season. 4B-4. Rice lines GH47 and GH137 are not only tolerant of Twenty-one-day-old seedlings were submerged for three high soil iron concentration and acidity, but they also have different durations of 1, 2, and 3 weeks with 125-cm water
70 Sulaiman et al Table 5. The performance of rice genotypes in shallow acid-sulfate land of Bari- to Kuala, South Kalimantan, 2003.
Belandean Terantang Oanda Jaya Rice genotypes Yield Fe toxicity Yield Fe toxicity Yield Fe toxicity (t ha–1) score (t ha–1) score (t ha–1) score
GH47 4.38 3 3.64 2 3.67 2 GH137 3.70 2 2.88 2 3.58 2 GH173 2.45 4 3.22 3 4.21 3 GH460 2.95 4 4.08 4 3.60 4 TOX3118B-E-2-3-2 3.55 4 3.12 2 3.69 4 IR58511-4B-4 4.11 4 3.15 3 4.50 4 Mendawak 3.72 4 4.70 2 4.47 4 Margasari 2.95 3 3.07 2 3.73 2 CV (%) 4.7 16.0 9.9 Soil pH 3.52 3.57 3.93 Fe content (ppm) 297 206 153
Source: RISA (2004).
Table 6. Yield of rice breeding lines in acid-sulfate soils of Belandean and Unit Tatas, in peat soils of Pinang Ha- bang, Kalimantan, in acid land of Jambi and Karang Agung I, and in peat soils of Karang Agung II, Sumatera.
Yield (t ha–1) Genotype Belandean U. Tatas P. Habang Jambi K. Agung K. Agung II Mean W2a
GH47 (KaI9407d-Bj-18-2) 3.16 abc 3.05 bcd 3.14 4.09 b 4.37 ab 3.06 cd 3.48 0.0755 GH137(KaI9408d-Bj-28-4) 3.13 abc 3.36 cd 3.17 2.59 a 4.22 ab 3.18 cd 3.28 2.3777 GH460(KaI9414d-Bj-110-1) 3.30 be 3.12 bcd 3.02 2.72 a 3.90 a 1.78 a 2.98 1.8158 GH505(KaI9420d-Bj-14-1) 3.27abc 3.13 bcd 2.81 2.92 a 4.11 ab 2.63 be 3.15 1.0341 BW307-6 3.17 abc 3.32 cd 2.88 4.62 bcd 4.33 ab 2.36 ab 3.41 0.5772 Tox3118b-E-2-3-2 3.05 ab 3.17 bcd 3.16 5.34 d 4.14 ab 2.56 be 3.50 1.6008 B10179b-Mr-1-4-1 2.81 a 2.99 be 3.10 4.15 be 4.21 ab 2.17 ab 3.24 0.3474 IR61242-3B-B-2 3.62 3.57 d 3.44 5.15 d 5.20 d 3.59 d 4.09 0.4131 IR58511-4B-4 3.55 c 2.32 a 3.21 5.24 d 4.56 bc 3.43 d 3.72 1.9711 B9852E-35-KA-66 3.53 be 2.76 ab 3.23 4.84 cd 4.84 cd 3.61 d 3.79 0.8795 Margasari (control) 3.18 abc 3.11 bcd 2.85 3.79 b 2.09 ab 3.19 3.19 0.3270 Mendawak (control) 3.21 abc 3.02 bc 3.37 4.16 be 3.15 cd 3.57 3.57 0.1322 CV (%) 7.6 9.2 11.7 11.8 13.3
Soil pH (H20) 3.33 3.50 3.8 4.69 4.48 Fe content (ppm) 527.0 1098.98 153.79 42.50 1157.33
aW2 = stability index of “Wricke’s ecovalence.” Means followed by the same letters are not significantly different according to Duncan’s multiple range test. Source: RISA (2003). depth. A total of 70 rice genotypes were tested in this trial, clear water and during the 2004-05 rainy season using turbid of which 40 were from IRRI, 24 from RISA, and 5 from water. Twenty-one-day-old seedlings of 16 rice lines from improved varieties. The results showed that after 1 week IRRI, including IR66 and Tapus as check varieties, were of submergence, 24 genotypes showed 100% survival, 20 transplanted in rows with 28 plants per entry. One week after genotypes showed 92% survival, and 14 genotypes showed transplanting, all plants were completely submerged with 85% survival. However, after 2 weeks of submergence, only 3 a water depth of about 125 cm for durations of 1, 2, and 3 rice genotypes showed 100% survival, 10 genotypes showed weeks in clear water and 6, 12, and 18 days in turbid water. 92% survival, and another 10 showed 85% survival. The three Survival was counted 1 week after submergence. After 1 week rice genotypes are IR66036-3B-13-2-B, IR70215-2-CPA-2- of submergence in clear water, 11 rice genotypes showed 1-B-1-2, and IR73047-6-1-1-1-B-2-B. All rice seedlings died 100% survival, but, after 2 weeks of submergence, only 3 after 3 weeks of submergence. rice genotypes had 100% survival (Table 7). These three rice Two similar trials were also conducted in an artificial genotypes are IR73047-6-1-1-1-B-2-B, IR70215-2-CPA-2- pond at RISA, Banjarbaru, during the 2004 dry season using 1-B-1-2, and Tapus. After 6 days of submergence in turbid
Rice breeding for the tidal wetlands of Indonesia 71 Table 7. Percentage survival of rice genotypes after submergence in clear water (DS 2004) and in turbid water for different periods (WS 2004-05).a
Clear-water Muddy-water survival (%) Genotype survival (%)
1 week 2 weeks 6 days 12 days 18 days
IR69502-6-SRN-3-UBN-1-B-1-2 85 (5) 46 (9) 46.4 (9) 0 0 IR69502-6-SRN-3-UBN-1-B-1-3 100 (1) 85 (5) 35.7 (9) 3.6 0 IR70181-5-PMI-1-2-B-1 100 (1) 62 (7) 85.7 (5) 35.7 3.6 IR70213-9-CPA-12-UBN-2-1-3-1 100 (1) 77 (5) 92.9 (5) 67.9 17.9 IR70213-10-CPA-2-3-2-1 100 (1) 69 (7) 100 (1) 85.9 85.7 IR68835-44-8-B-B-4-1 100 (1) 62 (7) 92.9 (5) 53.6 7.1 IR70181-32-PMI-1-1-5-1 100 (1) 69 (7) 96.4 (1) 82.1 78.6 IR66036-3B-13-2-B 100 (1) 85 (5) 96.4 (1) 75.0 39.3 IR70213-10-CPA-4-2-1-1-3 100 (1) 92 (5) 96.4 (1) 92.9 82.1 IR70215-2-CPA-2-1-B-1-2 100 (1) 100 (1) 100 (1) 100 39.3 WAR115-1-2-1-3-6-B-B-2 85 (5) 8 (9) 92.9 (5) 96.4 0 WAR115-1-2-4-2-4-B-B-4 62 (7) 15 (9) 64.3 (5) 64.3 0 IR73047-6-1-1-1-B-2-B 100 (1) 100 (1) 96.4 (1) 46.4 3.6 GH137 62 (7) 46 (9) – – – TOX3118B-E-2-3-2 69 (7) 46 (9) – – – IR66 92 (5) 8 (9) – – – Tapus (check) 100 (1) 100 (1) 82.1 3.6 0
aNumbers in parentheses = % comparative survival score (CS) = (% survival of entry/% survival of con- trol) × 100%. CS 100 = 1; CS 75–94 = 5; CS 50–74 = 7; CS 0–49 = 9. Source: RISA (2005).
Table 8. Insect and disease resistance of some local rice varieties collected from tidal swamplands in South Kalimantan.a
Variety Location Brown Sheath Leaf Neck Stem Bph1 spot blight blast blast borer Runtai Hulu Sungai MR MS MS HS S MS Bayar Pahit Sei Tabuk MR MS MS S S MS Siam Unus Sei Tabuk MR MS MR MR S MS Siam Arjan Kalumpang R S R MS S MS Siam Sanah Sungai Raya MR S MR S S MS Ketan Siam Sei Tabuk S MS S R S MS Runut Simpur S MS S MR S MS Siam Bam- Simpur MS MS R MS S S ban Siam Cinta Simpur S MS R MS S MS
aR = resistant, MR = moderately resistant, HS = highly susceptible, MS = moderately sus- ceptible, S = susceptible. Sources: Balittan Banjarbaru (1995) and Mukhlis and Imberan (1998).
water, only 2 rice genotypes (IR70213-10CPA-2-3-2-1 and Kalimantan, collected, characterized, and conserved a large IR70215-2-CPA-2-1-B-1-2) had 100% survival, and, after number of local rice cultivars. By 2001, about 239 lines 12 days of submergence, only IR70215-2CPA-2-1-B-1-2 were collected, out of which 107 cultivars were selected, showed 100% survival. Genotypes IR70213-10-CPA-4-2- consisting of 26 cultivars for deep acid-sulfate lands, 63 1-1-3, IR66036-3B-13-2-B, and IR70213-9-CPA-12-UBN- for shallow acid-sulfate lands, 6 for peat lands, and 12 for 2-1-3-1 were considered moderately tolerant. freshwater swamplands (RIFSA 2001). Screening of 58 lo- cal rice varieties was done for brown planthopper resistance Collection of rice cultivars (Bph1), stem borer, brown spot, sheath blight, leaf blast, and To support breeding research activities and conserve rice neck blast. Most of them were susceptible to these insects cultivars and genetic resources in tidal swamplands, re- and diseases (Table 8). Prayudi et al (2002) reported ten local searchers, especially at RIFSA or RISA, Banjarbaru, South rice varieties in tidal swamplands of South Kalimantan that
72 Sulaiman et al are susceptible to tungro disease: Siam Unus, Pandak, Siam become the key entry point for developing new manage- Raden, Lemo, Bayar Pahit, Siam Perak, Siam Teladan, Siam ment practices. The eating and cooking qualities of the new Adil, Kandangdukuh, and Siam Putih. rice varieties should also be given a high priority. Since In 2001 and 2002, researchers at RISA collected an research facilities and funds are limited, well-coordinated additional 182 local rice cultivars from the swamplands in and collaborative research activities, involving both regional South and Central Kalimantan as well as from Lampung and international networks, are needed to optimize the use and South Sumatera. Most of the local rice varieties are of available resources. Collaboration and coordination be- photoperiod-sensitive, with yields varying from 1 to 4 t ha–1, tween breeders and other scientists, extension workers, and and they require 4 to 9 months to reach maturity. The local policymakers are needed to develop a cirtical mass for such rice varieties have well-exserted panicles, are awnless, and concerted efforts. have intermediate plant height, 7–17 panicles per hill, 17–23 Some of the future challenges for varietal improvement grams per 1,000 grains, slender grain type, and medium to for tidal swamplands include short brown rice length. 1. Developing better varieties for the major cropping Seventy-one rice cultivars had been characterized for systems and environments. their agronomic and growth characters (RISA 2003). Some 2. Continuously evaluating and improving the breeding of them, such as Bayar Palas, Pandak Putih, Siam Unus, and objectives, priorities, and strategies. Lemo Putih, were found to be relatively resistant to lodging, 3. Forming stronger linkages with other institutions whereas Siam Puntal, Siam Adus, Siam Suruk, and Siam and agencies in Indonesia and abroad to improve the Lantik were resistant to tungro. Six rice varieties were found testing and extension of improved tidal swampland to have iron content of more than 66 ppm, and only Siam rice varieties. Pandak and Siam Wol had higher iron content of 83 and 70 4. Improved facilities and well-trained breeders to ppm, respectively. Fifteen rice cultivars had zinc content of handle the expanding research activities. more than 72 ppm; one of them, Siam Panangah, had zinc content of 108 ppm. Future prospects for rice breeding for tidal wetlands Challenges and future prospects for rice breeding Rice breeding programs for tidal swamplands should be intensified in order to support the increasing demand for Challenges for rice breeding for tidal wetlands rice production in these areas, and to resolve the existing Developing adapted high-yielding varieties is an important challenges. Future rice breeding programs for these areas component for enhancing and sustaining productivity in must be directed not only to solve the existing problems of swampland rice-farming systems. Eighteen improved rice rice production systems but also to fulfill farmers’ require- varieties for tidal swamplands had been released and some ments and preferences. Therefore, the aim of rice varietal promising rice lines had been identified, but their adop- improvement programs for tidal swamplands should be to (1) tion by farmers was still very slow. This is probably due to develop high-yielding photoperiod-sensitive rice varieties to several reasons, including high external input requirements, replace the existing long-season local rice varieties and (2) low farmers’ or consumers’ preference, limited adaptation improve existing early-maturity high-yielding rice varieties, to only one or a few existing problems, and greater demand particularly for their adaptability or tolerance of specific local for labor, land, and water management. On the other hand, problems, yield potential, and grain quality. local rice varieties commonly grown in tidal swampland areas With limited research facilities and resources, more have wider adaptation to these conditions and still perform coordinated and collaborative efforts are needed, at both the well with low external inputs, but they have low yield and national and international level. Moreover, careful analysis long duration. These situations could be challenging for rice and prioritization are needed to develop more focused and breeding programs to support the increasing demand for rice efficient research plans to tackle these complex issues. Mod- production in tidal swamplands. ern rice breeding techniques such as biotechnology, mutation Farmers in tidal swamplands benefited very little from breeding, and molecular breeding should be explored together modern agricultural systems, including improved rice varie- with conventional methods in order to streamline breeding ties. Most of these farmers still grow traditional local rice programs and speed up the development of suitable new rice varieties using traditional production systems. Strong and varieties. sustained efforts are needed to improve the rice production systems and increase rice production in tidal swampland References areas. These efforts include considerable investments in research to develop adapted high-yielding rice varieties with Alihamsyah L. 2004. Agricultural research and development on acceptable grain quality, together with proper management swampland or flood-prone ecosystem in Indonesia. Paper pre- practices. In most instances, highly site-specific management sented at the International Workshop on Flood-Prone Rice in practices are required because of the specificity of these Asia. Cantho, Vietnam, 8-9 February 2004. stresses, and suitable varieties for such specific situations
Rice breeding for the tidal wetlands of Indonesia 73 Balittan Banjarbaru. 1995. Laporan Hasil Penelitian. Balai Penelitian RISA (Research Institute for Swampland Agriculture). 2002. Annual Tanaman Pangan, Banjarbaru, Indonesia 1994/95. 554 p. (In report for 2001. Banjarbaru, South Kalimantan, Indonesia. Indonesian.) RISA (Research Institute for Swampland Agriculture). 2003. Annual Brown KD, Sulaiman S. 1984. Present state and future challenges of report for 2002. Banjarbaru, South Kalimantan, Indonesia. the BARIF Indonesia swamp rice breeding. I. Tidal swamps. RISA (Research Institute for Swampland Agriculture). 2004. Annual Indonesia Agric. Res. Dev. J. 6(1 & 2):1-9. report for 2003. Banjarbaru, South Kalimantan, Indonesia. Direktorat Bina Rehabilitasi dan Pengembangan Lahan. 1995. Luas RISA (Research Institute for Swampland Agriculture). 2005. Annual penggunaan lahan rawa pasang surut, lebak, polder dan rawa report for 2004. Banjarbaru, South Kalimantan, Indonesia. lainnya di tujuh propinsi. Dirjen Tanaman Pangan dan Hortikul- Tumarlan IW, Suastika, Nasution B. 2000. Uji daya hasil padi di lahan tura. Departemen Pertanian, Jakarta. (In Indonesian.) pasang surut Karang Agung Ulu, Sumatera Selatan. Proceedings Lesmana OS, Toha HM, Las I. 2002. Deskripsi varietas unggul baru of National Seminar on Agricultural Research and Development padi. Research Institute for Rice (RIR), Sukamandi, Indone- in Swampland. Cipayung, 25-27 July 2000, Indonesia. sia. Widjaya Adhi IPG. 1986. Pengelolaan lahan pasang surut dan lebak. Mukhlis M, Imberan. 1998. Pengujian ketahanan varietas padi lokal Jurnal Litbang Pertanian V(1), Januari 1986. Badan Litbang terhadap penyakit bias. Kalimantan Agrikultural, IImiah Fak. Pertanian. Pertanian Lambung Mangkurat 5(1):12-15. Widjaya Adhi IPG, Nugroho K, Ardi DS, Karamadan AS. 1992. Noorsyamsi H, Anwarhan H, Sulaiman S, Beachell HM. 1984. Rice Sumber daya Lahan Pasang Surut, Rawa, dan Pantai : Potensi, cultivation of the tidal swamps of Kalimantan. In: Workshop on Keterbatasan dan Pemanfaatan. Oalam proceedings Pertemuan Research Priorities in Tidal Swamp Rice. Los Baños (Philippi- Nasional Pengembangan Lahan Pertanian Pasang Surut dan nes): International Rice Research Institute. p 17-28. Rawa. Cisarua, 3-4 Maret 1992. Prayudi B, Widiarto IN, Sulaiman S. 2002. Ketahanan genotipe padi rawa pasang surut terhadap penyakit tungro. Oalam Prosiding Notes Seminar Nasional Pertanian Lahan Kering dan Lahan Rawa. Balai Penelitian Pertanian Lahan Rawa. p 41-48. (In Indone- Authors’ address: S. Sulaiman, I. Khairullah, and T. Alihamsyah, Plant sian.) breeder, assistant plant breeder, and director, respectively, of RIFSA (Research Institute for Food Crops in Swamp Area). 2001. Em- Research Institute for Upland Agriculture, Banjarbaru, South pat Puluh Tahun Balittra: Perkembangan dan Program Penelitian Kalimantan, Indonesia. Ke Depan. Banjarbaru, South Kalimantan, Indonesia.
74 Sulaiman et al Rice breeding for acid-sulfate soils in Vietnam
Bui Chi Buu and Nguyen Thi Lang
To feed the current population (81 million) of Vietnam, our Three main problems in acid-sulfate soils have been farmers cultivate nearly 4.3 million ha of rice (total agri- targeted: (1) aluminum toxicity, (2) iron toxicity, and (3) cultural area of 7.4–7.6 million ha). Rice land is declining phosphorus deficiency. The Cuu Long Rice Research Institute because of urbanization, industrialization, and other reasons (CLRRI) has focused its research on the first two problems. to about 4.0 million ha. In addition, water shortage and salt For iron-toxicity tolerance, we initiated a crossing program intrusion will become serious problems and large challenges to develop breeding populations for QTL mapping and selec- for human food security, even though Vietnam is currently tion. a net rice exporter. Rice is still the most important crop as it The problems and constraints encountered in these soils supplies about 67% of the calorie intake in the country. Rice vary across ecosystems, which require different breeding and production was about 15.7 million tons in 1985 but doubled management strategies to solve particular problems. Cur- to about 35.9 million tons in 2004. Current average yield is rently, suitable water management and agronomic practices about 4.87 t ha–1 (MARD 2005). Acid-sulfate soils account have been recommended and rice varietal improvement is for 40.8% and 5.6% and alluvial soils for 30.1% and 48.5%, also considered a key approach for enhancing yield on these respectively, in the Mekong Delta and the Red River Delta. soils. Clearly, acid-sulfate soils have become the main constraint to rice production in the Mekong Delta (Table 1). Tolerance of aluminum toxicity The exploitation of genes from wild rice relatives has been Table 1. Problem soils in rice cultivation successful in introducing many useful genes into the culti- areas in the Mekong Delta (MD) and Red River Delta (RRD). vated gene pool. A good example is the development of the high-yielding variety AS996 (IR64/Oryza rufipogon), which Area (million Percentage is tolerant of aluminum toxicity, is of short duration, and is Soil type ha) of total area adapted to acid-sulfate soils. Major QTLs on chromosome 3 were identified that are found to be associated with aluminum MD RRD MD RRD toxicity (Fig. 1). This is achieved through the development and analysis of a recombinant inbred line (RIL) population Acid-sulfate 1.60 0.11 40.8 6.0 developed from the cross of IR64/O. rufipogon. QTLs as- Salinity 0.74 0.09 18.9 5.2 Peat soils 0.02 0.002 20.0 0.1 sociated with control root length (CRL), stress root length Gray soils 0.13 0.15 3.4 8.0 (SRL), and relative root length (RRL) were mapped (Nguyen et al 2002).
Chr 3 Chr 3 Chr 4H Chr 4DL Chr 4R RZ448 RZ448 WG114
111.9 104.8 140.4
RG191 RG745 13.1 12.7 CDO1395 BCD117 1.1 AltB 0.0 Alp 2.1 BCD1230 BCD1230 11.9 0.6 RG391 15.7 2.1 Alp 10.2 CDO1395 ACC-CTG2 CDO1395 CDO1395 CDO836
Rice Rice Barley Wheat Rye Fig. 1. Comparison of the QTL identified on chromosome 3 in rice with that of wheat, barley, and rye (Nguyen et al 2002).
Rice breeding for acid-sulfate soils in Vietnam 75 Table 2. Phenotyping for Al toxicity tolerance us- Table 4. Rice varieties adapted to acid-sulfate soils in the Me- ing Yoshida solution and two concentrations of kong Delta in 2003 and 2004. aluminum, 0 and 30 ppm. Area covered (ha) Material Relative root length Designation Origin 2003 2004 Oryza rufipogon (acc. 106412) 1.158 AS996 1.133 OM1490 OM606/IR44592-62-1-1-3 213,157 344,263 OM1490 1.058 VND95-20 IR64 mutant 153,994 185,201 OM1314 0.899 AS996 IR64/O. rufipogon 21,206 23,743 Ca Dung Do (tolerant check) 0.843 OM3536a TD8/OM1738 37,449 88,693 IR29 (sensitive check) 0.514 IR64 (check) IR18398-36-3-3 101,211 98,747 LSD 0.230 0.05 aOM3536 is intended to replace AS996 due to its better grain quality in less favorable areas.
Table 3. Area covered by AS996 during 2000 to 2004. later released as a national variety in 2002 for cultivation in less favorable acid-sulfate soils (Tables 3 and 4). Year Area (ha)
2000 38,064 Tolerance of phosphorus 2001 93,666 2002 125,320 Genotypes with promise for tolerance of phosphorus 2003 21,206 deficiency came from the following crosses: Kasalath 47/ 2004 23,743 OM4495 (BC2F2), AS996/OM2395 (RILs F6), M23/AS996, M362/AS996, M379/AS996, M382/AS996, and Kasalath 47/AS996. OM4498 is a backcross line developed from Three populations of O. rufipogon were collected IR64/OMCS2000//IR64 using marker-assisted selection to from Tram Chim (see Table 2) by Duncan Vaughan and Bui incorporate two QTLs for salinity tolerance, one each on Chi Buu in 1989. The soil of Tram Chim (bird sanctuary) is chromosomes 1 and 8 (with markers RM315 on chromosome strongly acid-sulfate, with pH from 2.8 to 3.2. Later, three 1 and RM223 on chromosome 8). This breeding line has been crosses were developed (D.S. Brar) as follows: extensively tested and will be released soon due to its large- scale adoption in large areas in less favorable regions. AS1007 IR64/O. rufipogon (acc. 106412) Four traits related to P-deficiency tolerance were AS833 IR64/O. rufipogon (acc. 106424) considered (Table 5). AS996 and IR26 had a higher relative AS996 IR64/O. rufipogon (acc. 106412) index of shoot length, root length, shoot dry weight, and root dry weight than OM2395 and IR36. The genetic control All the O. rufipogon accessions were collected from of some characters related to P-deficiency tolerance was Tram Chim. The F2 progenies were sent to Vietnam for test- studied through diallel analysis. Suitable materials chosen ing in the target areas. AS996 was subsequently selected and were OM723-11, OM850, IR64, IR50404, OM997, and
Table 5. Phenotyping for P-deficiency tolerance in Yoshida solution using two tests
of 5 and 10 ppm of P2O5.
Shoot length (cm) Root length (cm) Shoot dry weight Root dry weight Line (mg) (mg)
5 ppm 10 ppm 5 ppm 10 ppm 5 ppm 10 ppm 5 ppm 10 ppm
AS996 29.9 32.7 9.14 9.70 19.4 20.6 2.5 2.8 0.91a 0.94 0.94 0.89 OM2395 28.7 31.9 8.40 11.10 20.7 26.8 2.3 2.9 0.90 0.76 0.77 0.79 IR26 27.4 29.2 9.76 10.72 18.2 22.6 2.4 2.6 0.94 0.91 0.81 0.92 IR36 27.8 32.9 8.04 10.06 15.3 29.2 2.1 2.5 0.84 0.80 0.52 0.84
aRelative index = value at 5 ppm/value at 10 ppm × 100. Data are means of 5 replications.
76 Bui Chi Buu and Nguyen Thi Lang Table 6. Inheritance of some growth and yield traits under P-deficiency stress.
1/2 2 2 2 Trait (H1/D) 2 σ gca/(2 σ gca H ns (%) (narrow-sense + σ2sca) heritability)
Tilling capacity 1.94 0.16 19.70 Growth duration 0.98 0.56 33.90 Filled grains per panicle 5.80 0.01 3.10 Root dry weight 0.81 0.03 20.90
IR59606. Tillering ability was considered a good indicator of develop prebreeding materials. The integration of biotech- performance under P deficiency and was used as a selection nology tools with conventional breeding methods offers new criterion. Maximum tiller numbers were scored at 45 days opportunities to increase rice productivity and sustainability after transplanting in the hybrids and the parents, constituting and develop better varieties with higher tolerance of acid- a 6 × 6 diallel set. From this analysis, it was observed that sulfate toxicity. The potential of the existing genetic diversity shoot dry weight was the most sensitive growth parameter for enhancing the productivity of acid-sulfate soils has not yet affected by P deficiency, followed by root dry weight and then been adequately explored. We need further collaboration to number of tillers. The proportion of dominant and recessive make better use of the latest biotechnological developments to genes in the parent (KD/KR = 1.6) was more than one unit. be employed in conjunction with conventional rice breeding This means that dominant gene actions are more important programs for germplasm improvement. under P-deficiency stress. The effect of positive alleles was clear (H2/4H1 = 0.37), showing that the higher the root dry References weight, the better tolerance of P deficiency. The variance ratio 2 σ2gca/(2 σ2gca + σ2sca) was computed from the expected MARD (Ministry of Agriculture and Rural Development). 2005. Annual components of the mean square assuming a fixed model to report of 2004. Hanoi (Vietnam): MARD. assess the relative importance of additive and nonadditive Nguyen DB, Brar DS, Buu BC, Tao NV, Luong PN, Nguyen HT. 2002. gene effects in predicting progeny performance. Data in Table Identification and mapping of the QTL for aluminum tolerance 6 demonstrate the effect of nonadditive gene action in the introgressed from wild sources, Oryza rufipogon Griff., into in- inheritance of the characters, except for growth duration. dica rice (Oryza sativa L.). Theor. Appl. Genet. 106:583-593.
Work plans Notes
The following objectives were set: Authors’ address: Cuu Long Rice Research Institute, Vietnam. l To perform QTL analysis for P-deficiency toler- ance. l To select promising lines from the current crosses at CLRRI’s experimental field and target areas (Long An, Tien Giang, Dong Thap, Kien Giang) to go into observation yield nurseries, preliminary yield trials (PYTs), and advanced yield trials (AYTs). l To create new crosses based on phenotyping and a DNA survey among gene pool materials. l To do G × E interaction analysis (9–15 sites). l To analyze iron deficiency through both phenotyping and genotyping. Rice production in Vietnam was successfully increased because of the effective collaboration with IRRI and others in various research areas. Rice breeding for acid-sulfate soils will be considered as a key activity in the coming years to help narrow the yield gap in these less favorable areas. Priorities considered will be the development and use of marker-assisted selection combined with the advantages of conventional breeding methods. Vietnam needs to increase and strengthen its capacity in biotechnology for further rice improvement and will need further assistance from IRRI to
Rice breeding for acid-sulfate soils in Vietnam 77 Breeding rice for salt-affected areas of India
R.K. Singh, B. Mishra, A.M. Ismail, and G.B. Gregorio
Breeding of rice varieties for salt-affected areas remained and place of occurrence, the salt-affected soils of India are elusive for a long time because of the poor understanding grouped into three major problem areas: sodic soils, inland of the problem and the complexities associated with it in saline soils, and coastal saline soils, with estimated areas real farmers’ fields, where multiple stresses are common. of 3.4, 3.1, and 2.1 million ha, respectively (CSSRI 1997). Recently, considerable progress has been made by different With the introduction of large-scale irrigation canal networks organizations in developing suitable salt-tolerant rice vari- throughout the country, a majority of the salt-related problems eties for target-specific areas after considerable efforts and are emerging under the canal command areas (Table 1), most investments were made on characterization of these problem of which still remain potential sources of secondary saliniza- soils. To some extent, almost all soils contain some salts but tion. Although both salinity and sodicity can be found in a they are not designated as salt affected until the concentra- canal command area as well as outside of it, most of these tion of specific harmful salt becomes too high and hinders command areas are saline (inland). A majority of sodic soils plant growth and development. Sodium-based salts usually are located in Uttar Pradesh (1.1 million ha), Punjab, (0.35 predominate in salt-affected areas. Before addressing the million ha), Haryana (0.18 million ha), and some parts of needs and strategies for breeding rice varieties suitable for Bihar, Rajasthan, and Karnataka, whereas inland saline ar- salt-affected ecologies of India, we will provide an overview eas are predominant in Rajasthan, Gujarat, Andhra Pradesh, of the type and extent of salt-affected areas. Karnataka, Punjab, and Haryana. Coastal salinity is found along the 5,734-km stretch of What are salt-affected soils? coastline of the Indian mainland, which can be divided into four coastal regions: Gujarat coast plains, West coast plains, Salt-affected soils are classified broadly into two categories: East coast plain, and Midnapur coast plain. The ecology of saline and sodic (or alkali) soils. However, another category the coastal region is very unusual because of several com- of salt-affected soils is also sporadically found and is referred pounded problems such as salt-water intrusion, salt sprays, to as saline-sodic soils. Sodium is the dominant cation in all bank erosion, impeded drainage, and other climate-induced salt-affected soils, with soluble chloride and sulfates as the factors, which eventually have a drastic effect on agricultural corresponding anions in saline soils, resulting in an electrical production. Soils of the coastal regions vary considerably, conductivity (EC) of the soil solution of greater than 4 dS but generally are medium to heavy textured with NaCl as a m–1. In sodic soils, however, sodium also predominates over dominant soluble salt, except in Kerala and some areas of other cations but rather more in the exchange complex than West Bengal, where soils also contain large quantities of in soil solution, together with high concentrations of carbon- Na2SO4. Soil salinity status varies by season and is at its ate/bicarbonate anions. Such soils have high exchangeable peak in May. It decreases with the onset of the monsoon and sodium percentage (ESP > 15) and pH (>8.5 and sometimes is generally the lowest during August-September. Depth of up to 10.7) and poor soil structure. Saline-sodic soils, also groundwater and its salinity also vary by season. called saline-alkali soils, have both high ESP (>15) and EC –1 (>4 dS m ). Projected losses in salt-affected soils Extent of salt-affected soils in India India incurs about a 12% loss in its production system because of degraded lands, which includes salt-affected soils (Singh Abrol and Bhumbla (1971) estimated the country’s salt-af- et al 2003). This loss is estimated at about US$291 million fected soils as 7 million ha in 1971. This was followed by per annum in salt-affected areas alone. The states with larger various estimates by different scientists and agencies, and, salt-affected areas share most of the losses such as Gujarat, after critical evaluation of these varying estimates, the Central Rajasthan, Uttar Pradesh, Andhra Pradesh, and West Bengal, Soil Salinity Research Institute (CSSRI), Karnal, endorsed when compared with other states with fewer affected areas. the figure of 8.6 million ha from Singh (1992) as the latest Overall losses nationally are projected as reliable figure. Broadly based on chemical characteristics
78 Singh et al Table 1. Extent and distribution of salt-affected soils in India (000 ha).
Salt-affected areas State Canal command Outside canal Coastal Total area command area
Andhra Pradesh 139.4 390.6 283.3 813.3 Bihar 224.0 176.0 0.0 400.0 Gujarat 540.0 372.1 302.3 1,214.0 Haryana 455.0 na 0.0 455.0 Karnataka 51.4 266.6 86.0 404.0 Kerala na na 26.0 26.0 Madhya Pradesh 220.0 22.0 0.0 242.0 Maharashtra & Goa 446.0 na 88.0 534.0 Orissa na na 400.0 400.0 Punjab 392.6 126.9 0.0 519.5 Rajasthan 138.2 983.8 0.0 1,122.0 Tamil Nadu 256.5 na 83.5 340.0 Uttar Pradesh 606.0 689.0 0.0 1,295.0 West Bengal 0.0 na 800.0 800.0 Total 3,469.1 3,027.0 2,069.1 8,562.2
Source: Singh (1992), na - not available.
Country’s productivity at 2002 prices: on the quantity and quality of irrigation water, if available. Rs. 12,288 per ha ($273 per ha) Therefore, the only ecology that is totally rainfed in salt-af- Losses due to degraded lands: fected soils is predominantly in coastal saline areas during Rs. 1,521 per ha ($34 per ha) the wet season. However, dry-season cultivation in these Value of productivity losses due to salt-affected soils: areas invariably depends upon ensured irrigation for a major 8.6 × 106 × 1,521 portion of the season. Rs. 13.1 billion per annum $291 million per annum Why is rice the major crop in salt-affected areas? The rainfed rice ecosystem in salt-affected Almost the entire coastal belt of the Indian mainland is prone areas of the country to inundation by seawater during high tides, which results in salinity intrusion and occasional submergence. Thus, Limited freshwater availability is a serious constraint along only salt- and submergence-tolerant crop varieties, such as the coastal belt of India. Excess waterlogging and stagnant rice, are economically viable in these areas. Rice is the only partial flooding due to high rainfall within a short span in the economic crop that can grow well in waterlogged environ- monsoon season, coupled with poor drainage, particularly ments and it also tolerates salinity to some extent. In sodic along the eastern coast, create serious problems for rice soils, low infiltration rate due to poor hydraulic conductivity growth and yield. Water shortage, on the other hand, caused and poor soil physical and chemical conditions (high ESP) by scanty and uncertain rainfall limits cropping intensity and causes stagnation of water on the soil surface, a condition productivity during the postmonsoon season. Consequently, that cannot be tolerated by any other crop except rice. Thus, coastal area is mostly monocropped with rice, except in some rice is recommended as the first crop to start with during regions of the Gujarat plains. About 82–85% of the area is reclamation of sodic soils. For these reasons, rice remains covered by rice during the wet (kharif) season under rainfed the most preferred crop in salt-affected areas. conditions, whereas summer/boro (rabi/dry season) paddy cultivation is restricted to areas where sufficient fresh water Challenges and constraints facing rice production is available for irrigation. in salt-affected areas Cropping in inland sodic or saline areas mostly depends on the availability of irrigation facilities, and most of the sodic Despite the large area under rice in the coastal belt, the and saline-sodic areas where these facilities are not available average yield is far lower (1.2–1.4 t ha–1) than the national remain barren. People try to bring them back into cultiva- average (approx. 2.0 t ha–1). Possible causes for such low tion using various reclamation strategies in which relatively productivity are good-quality irrigation water is a prerequisite. Similarly, in 1. Occasional flooding, cropping in medium lands/ inland saline areas where either rice- or non-rice-based crop- lowlands, deep waterlogged areas, and floodplains ping systems are being practiced, farming depends mostly under the rainfed rice ecosystem.
Breeding rice for salt-affected areas of India 79 2. Frequent water-deficit stress coupled with salinity. areas. Concerted efforts are needed to bridge the produc- 3. Continued cultivation of traditional low-yielding tion gap in these areas given that they are highly populated local rice varieties and landraces. with impoverished communities. To achieve the production Under these circumstances, losses of crop yield are targeted for 2020, the following important areas of research 10–80% and sometimes even 100% when rainfall is erratic. need further attention: During the dry (rabi) season, farmers are compelled to keep l Steadily raising the genetic yield ceiling in rice land fallow because of high salinity, lack of good-quality varieties. irrigation water, and, above all, lack of suitable crops and l Reducing the gap between potential and actual varieties with high salinity tolerance. Current well-adapted yield. rice varieties grown in coastal areas are mostly traditional There is an immediate need to devote greater efforts photoperiod-sensitive, fertilizer-non-responsive, tall (prone and resources to develop target-specific, locally adapted to lodging) varieties with low yield potential (1.5–2.5 t ha–1) high-yielding varieties for the diverse rainfed ecologies and poor grain type. (Paroda 1998). The gap between potential and actual yield The major hurdle for the reclamation of sodic and in- in these coastal areas is immense, and is widening because of land saline soils is the high initial cost. Most such areas are the development of many improved salt-tolerant, fertilizer- either owned by the local government or by poor farmers who responsive, and intermediate stature high-yielding varieties. do not have enough resources and risk-bearing capacity to However, the potential of these varieties is not yet being reclaim or amend these soils; hence, such lands remain bar- realized in farmers’ fields due to a lack of knowledge and ren, particularly during the dry season. However, once these seeds of these varieties in target areas. Therefore, for major soils are reclaimed, they remain very productive, reflecting yield breakthroughs in the rainfed coastal ecologies, both rice a vast scope for bringing back such lands into the produc- varietal improvement and ensured availability of quality seed tion chain, provided resources and technical help become along with technological awareness are needed. available. Improved rice varieties with a reasonable amount of salinity tolerance and adaptation to various ecologies are Linkage established between IRRI and network sites of India becoming available, but their spread to farmers’ fields is not yet adequate because of the lack of sufficient seeds and No single institute is capable of undertaking all the complex knowledge. A lack of adequate communication facilities and tasks associated with problem soils. Therefore, to develop poor marketing infrastructure in general further add to the a collaborative research program for the improvement of slow spread of these varieties. rice yields in problem soils, the International Rice Research With the recent expanding urbanization and population Institute (IRRI) convened a workshop in November 1986, growth, a minimum of 2.5% annual growth in food grain attended by participants from South and Southeast Asian production will be needed to meet the increasing demand. rice-growing countries. After reviewing the status of research India’s rice production target for 2020 is 140 million tons and future needs, it was decided that concerted efforts should (Paroda 1998). Rice occupies 42 million ha under diverse be devoted to expedite progress through a network program. ecologies and has large untapped yield potential even with A cooperative was established to share virtual resources and current available modern production technologies. However, responsibilities in research on problem soils. IRRI agreed the growth rate in rice production, which reached 4.4% in the to undertake the prebreeding research and to strengthen the 1980s, declined to only 2.3% in the 1990s. Therefore, if the national program for each type of problem soil, so it could country has to maintain self-reliance in food production, it assume regional responsibilities. India was also selected as would have to strive to achieve and sustain a minimum an- a partner for cooperative network research on salinity. nual growth of 2.5% in rice production. However, a growth Before the formal establishment of IRRI’s Consortium rate of at least 3.1% in production needs to be regained in for Unfavorable Rice Environments (CURE), the ICAR-IRRI order to have sufficient rice for both home consumption and collaborative research program on “Germplasm Improvement exports (Swaminathan 1999). for Saline Soils” began in 1987. This project aimed to develop salt-tolerant varieties suited to target environments in the Overcoming the challenges country through shuttle breeding. The Central Soil Salinity Research Institute, Karnal, was identified as the lead center Crop productivity from salt-affected areas in general and for collaboration. Based on the variability and complexity of from coastal areas in particular is lagging behind most other adverse soils, which are invariably compounded by climatic inland areas. Serious thought should be given to increasing hazards, and nutritional toxicity or deficiency, and mostly the share of coastal areas for the country’s overall economic influenced by the interactions among these factors, target growth as the contribution of coastal areas to rice production network sites were identified and characterized in the country. is the lowest. Out of a total of about 92 million tons of rice The major theme of the network approach is the sharing of being produced in India, only about one-fourth is from the resources (in terms of their in situ stress conditions) to iden- long coastal belt of the country. The production growth rate tify the best adapted material for target areas. The selected in coastal states is much lower than that of irrigated inland materials were advanced and used in breeding programs at
80 Singh et al Table 2. Varieties developed at CSSRI-Karnal through ICAR-IRRI collaborative research.
Stress adaptation Duration Grain Variety (days) typea Remarks pH1:2 ECe (dSm–1)
CSR21 9.8–10.0 < 9.0 125–130 LB Recommended by the 30th AIRW for sodic soils CSR22 9.8– 9.9 < 10.0 135–140 MS Released by state variety release commit- tee for salt-affected soils of Karnataka CSR23 9.8–10.0 < 10.0 135–140 MS Released by CVRC for alkaline and coastal saline soils CSR26 9.8–10.0 < 9.0 125–130 LS Recommended by the 33rd AIRW for alkaline and coastal saline soils CSR27 9.6– 9.9 < 10.0 125–130 LS Released by CVRC for alkaline and coastal saline soils CSR28 9.8–10.0 < 10.0 130–135 SB Recommended by the 31st AIRW for sodic and inland saline soils CSR29 9.8–10.0 < 10.0 115–120 LS Recommended by the 31st AIRW for sodic and inland saline soils
aLB = long bold, MS = medium slender, LS = long slender, SB = short bold. all the network centers for the development of target-specific Site and plant type characterization as well as widely adapted genotypes. The objectives of the network were set as follows: One of the most important reasons for the early and consider- l To develop high-yielding salt-tolerant rice varieties able success of the project was the initial site characteriza- for enhancing productivity for sustainable agricul- tion undertaken at the beginning of the project. The soil and ture on salt-affected soils. water quality and plant type requirements were worked out l To collect and evaluate the germplasm available to supply the best genotypes to the specific network site for from traditional areas. further testing of their suitability and adaptability. l To generate breeding materials as per the need of Hot spots were identified all over the country (Table 3, target sites. Fig. 1), and then a shuttle-breeding approach was adopted to l To enhance the amount of salt tolerance in rice evaluate the advanced breeding lines and populations. Materi- genotypes. als were selected, advanced, and used for breeding programs l To strengthen the research capacity of the scientists at all the network centers to select genotypes suitable for their involved in the research network. particular target sites and plant-type preferences (Table 4). Germplasm exchange was the main component of the Major accomplishments collaborative research program of this network. IRRI supplied 978 advanced breeding lines distributed to all collaborators The progress made under the ICAR-IRRI collaborative re- of the country after multiplication. The year-wise distribution search program, considering the low budget of the project, is appears in Table 5. quite remarkable in terms of the progress made in developing This advanced rice germplasm received from IRRI improved genotypes. Out of nine varieties (CSR21 to CSR29) under the project was in addition to the regular IRSSTON developed by CSSRI using material exchanged through this trial received through INGER (the International Network for network, six (CSR21, CSR23, CSR26, CSR27, CSR28, and the Genetic Evaluation of Rice). The breeding materials were CSR29) were recommended by the All India Rice Workshop multiplied, evaluated, selected, advanced, and distributed to as promising genotypes for various salt-affected ecologies. the network centers. Promising lines were nominated to the Out of these varieties, CSR27 and CSR23 have already been national nursery and trials by the different target centers as released by the Central Varietal Release Committee (Table 2). per their plant-type requirements. The number of selected Similarly, other collaborating centers also screened large sets entries nominated by CSSRI, Karnal, for the national nursery of germplasm at specific target sites over the years and used (NSASN) and trial (SATVT) is presented in Table 6. the advanced material for direct release or indirectly in their breeding. Recently, CSR22 was released as a high-yielding The development of rice varieties for salt-affected areas rice variety for salt-affected areas of Karnataka. in India—a look back The development of suitable salt-tolerant rice varieties for salt-affected soils in India dates back to the pre-Independ- ence era. A majority of such rice varieties were from the
Breeding rice for salt-affected areas of India 81 Table 3. Test sites of the ICAR-IRRI national network for the development of salt-tolerant improved rice genotypes.
Site (state) Coordinating research institute Ecosystem, type of salt stress, and other stressesa
Kaithal (Haryana) Central Soil Salinity Research Irrigated, sodic (high RSC) Institute Karnal (Haryana) Central Soil Salinity Research Irrigated, nonstress for yield com- Institute parison Lucknow (U.P.) Central Soil Salinity Research Irrigated, high sodicity Institute Canning Town (W. Bengal) Central Soil Salinity Research Rainfed, coastal saline, waterlog- Institute ging Anand Central Soil Salinity Research Irrigated, inland saline Institute Kanpur (U.P.) C.S. Azad University for Agri- Irrigated, sodic culture & Technology Faizabad (U.P.) Narendra Deva University for Irrigated, sodic Agriculture & Technology Vyttila (Kerala) Kerala Agricultural University Rainfed, coastal deepwater, acid saline soils Panvel (Maharashtra) Konkan Krishi Vidyapeeth Rainfed, coastal saline drought- prone Machilipatnam (A.P.) Andhra Pradesh Agricultural Rainfed, coastal saline University Tiruchirapally (Tamil Nadu) Tamil Nadu Agricultural Irrigated, sodic University Motto (Orissa) Orissa University of Agriculture Rainfed, coastal saline and Technology Nawagam (Gujarat) Gujarat Agricultural University Irrigated, inland salinity Karaikal (Pondichery) Pandit Jawaharlal Nehru Sodic (high RSC) College of Agriculture and Research Institute
aRSC = relative sodium content.
coastal belt due to the perpetual nature of salinity problems, yielded some results when rice researchers attempted a few where farmers do not have many options except to grow rice. crosses before the 1970s, involving salt-tolerant landraces. Local rice germplasm grown in saline and sodic soils was But, unfortunately, the rice varieties developed were not very poor in yield and quality. Regional rice stations were accepted largely due to insufficient salt tolerance and poor established in different rice-growing states with the help of and unacceptable grain quality. The genetic drag of the local ICAR to tackle these regional problems. Scientific efforts traditional types was large, resulting in tall genotypes with in different ecosystems resulted in the development of salt- poor grain quality that were nonresponsive to fertilizers tolerant rice varieties, though most of them were selections (Table 8). from the locally adapted landraces prevailing for a long time in the target environments. Details on these local genotypes Post-1970 breeding progress—the development from different parts of the country and the early selections of new rice varieties are given in Table 7. Some of these selections were released by the states, whereas most of them remained as selections In the post-1970 era, the Indian Council of Agricultural in very specific and small isolated areas. Research (ICAR) entrusted this responsibility to national Lately, efforts from some regional rice research sta- institutes such as CSSRI, Karnal, with a mandate to work on tions resulted in the development of a few salt-tolerant rice salt-affected areas at the national level. Similarly, some of the varieties. The varietal development program for salt-affected state-owned research stations also devoted some resources to areas progressed very slowly for numerous reasons: (1) the develop target-specific salt-tolerant rice varieties for various complex nature of this ecosystem with high temporal and environments. The concerted efforts to develop target-specific spatial variability, (2) the complex nature of inheritance of the as well as widely adapted salt-tolerant rice varieties resulted tolerance trait, (3) a lack of sufficient resources and research in the release of numerous varieties. Initially, most of the re- capacity, and (4) a lack of precise and repeatable screening leased varieties were from different states for target areas, that techniques. Lately, more attention has been devoted, which is, specific soil-stress conditions such as CSR 5 and Usar 1
82 Singh et al 2. CSSRI, Karnal
3. Kanpur 1. IRRI (Philippines) 4. Lucknow 5. Faizabad
Index Collaborating centers 1. IRRI–Philippines 15. Nawagam INDIA 2. CSSRI–Karnal
14. Anand 6. Canning Network sites Town 3. CSAU Kanpur (U.P.) 7. Motto 4. CSSRI-RRS Lucknow (U.P.) 13. Panvel 5. NDAUT Faizabad (U.P.) 6. CSSRI–RRS Canning Town (W.B.) 8. Machilipatnam 7. NARP, Motto (Orissa) 8. ARS, Machilipatnam (A.P.) 9. Pajancoa, Karaikal 10. CARI, Port Blair 11. SSRC, Tiruchirapalli 12. RRS, Vyttila (Kerala) 10. Port Blair 13. KRS, Panvel (Maharashtra) 14. CSSRI–RRS, Anand (Gujarat) 9. Karaikal 15. GAU, Nawagam 12. Vyttila 11. Tiruchirapalli
Fig. 1. ICAR-IRRI collaborative research network of India. for the sodic soils of Haryana and Uttar Pradesh, respectively; dividends in terms of the release of many target-specific varie- Vyttila 1 and Panvel 1 for the coastal saline soils of Kerala ties by the state governments for specific niches. State-wise and Maharashtra, respectively; and Co 43 for the saline so- details of these varieties are given in Table 9. dic soils of Tamil Nadu. Many target-specific improved rice varieties were also developed by different research stations Infrastructure and screening facilities and released by their respective states. The ICAR-IRRI collaborative research project was mainly National releases of rice varieties committed to developing and distributing the germplasm to various collaborating centers due to limited resources. Considering its national mandate and responsibility, CSSRI Subsequently, various centers were able to develop their own took the lead by developing the first salt-tolerant rice variety facilities by capitalizing on the knowledge and experiences of for sodic and inland saline soils at the national level that was other partners. Most of the collaborators used in situ field tri- released in 1989. So far, only 12 salt-tolerant rice varieties als to screen germplasm due to the paucity of funds. However, were released nationally through CVRC, of which six were CSSRI, Karnal, and a few other centers were able to upgrade developed by CSSRI at Karnal, four by the CSSRI Regional and expand their existing screening facilities with the help of Research Station at Canning Town, and one each by the some other projects. Screening facilities and procedures for Central Rice Research Institute, Cuttack, and Directorate of salt-tolerance studies are enumerated as follows. Rice Research, Hyderabad. Some characteristics of these rice varieties are given in Table 9. In situ field evaluation Screening of genotypes under an actual field-stress environ- State releases of rice varieties ment in 2 to 5 long rows of 8 to 20 meters to minimize the effects of spatial variability was followed at most of the During the post-1970 era, various states started giving im- target sites, including CSSRI, depending on the availability petus to the development of salt-tolerant crop varieties for of suitable fields. Replicated trials are always preferred for their specific problem areas. These intensified efforts gave advanced stabilized material. Field screening is usually the
Breeding rice for salt-affected areas of India 83 Table 4. Survey of specific plant-type requirements of the target Table 5. Breeding lines evaluated through the network during sites. 1987 to 2000.
Target site Plant-type characterization Year Germplasm Number
1. Vyttila (Kerala) l Plant height not less than 125 cm 1987 Advanced breeding lines 50 l Total duration not more than 125 1987 Anther culture derivatives 40 days 1988 Advanced bulk populations 131 l Tolerance of flooding (early seedling 1988 Anther culture derivatives 7 vigor) 1989 Advanced bulk populations 106 l Tolerance of salinity 1990 Advanced bulk populations 38 l Tolerance of acidity 1991 Advanced bulk populations 40 2. Machilipatnam (A.P.) and l Semidwarf to medium tall 1992 Advanced bulk populations 51 Motto (Orissa) l Total duration more than 140 days 1993 Advanced bulk populations 29 l Tolerance of salinity 1994 Advanced bulk populations 60 l Fertilizer responsive and high yield- 1995 Advanced bulk populations 16 ing (traditional varieties have disap- 1996 Advanced bulk populations 28 peared from the problem areas) 1997 Advanced bulk populations 85 3. Panvel (Maharashtra) l Dwarf and semidwarf 1997 Recombinant Inbred lines (IR29 × Pokkali) 89 l Fine grain 1998 Advanced bulk populations 54 l Salinity tolerant 2000 Advanced bulk populations 70 l Early maturing (less than 125 days) 2000 Advanced breeding lines 84 4. Canning Town, Chinsurah l Medium tall Total germplasm received under the project 978 (W. Bengal), and Port Blair l Tolerant of salinity l Tolerant of waterlogging (early seedling vigor) Table 6. Genotypes and vari- l Weakly photosensitive eties nominated for the na- l Medium growth duration tional and nursery trials from For upland and well-drained condi- 1991 to 2005. tions: l Early maturing Year No. of entries l Height 60 to 90 cm l Fertilizer responsive and photo-in- National Saline Alkaline sensitive genotypes (Tolerant) Screening Nursery 5. Haryana, Punjab, and Uttar l Dwarf/semidwarf (NSASN) Pradesh l High yielding 1991 5 l Early maturing (125 days) 1992 10 l Fertilizer responsive 1993 14 l Slender to medium coarse grain 1994 15 l High tolerance of sodicity and 1995 10 salinity 1996 10 6. Anand and Nawagam l Semidwarf 1997 – (Gujarat) l High yielding 1998 9 l Medium growth duration 1999 4 l Fertilizer responsive 2000 4 l Slender to medium coarse grain 2001 1 l High tolerance of salinity 2002 3 7. Karaikal (Pondicherry) l Semidwarf 2003 3 l High yielding 2004 – l Early maturing (125 days) 2005 1 l Fertilizer responsive Total 89 l Medium coarse grain l High tolerance of sodicity and Saline Alkaline Tolerant Vari- salinity etal Trial (SATVT) 1993 4 1994 8 1995 8 1996 4 1997 7 1998 4 1999 5 2000 3 2001 – 2002 – 2003 2 Total 45
84 Singh et al Table 7. Traditional landraces and subsequent selections adapted to salt-affected soils in India.
Landrace Selection State Area of adaptation (local variety)
Jhona Jhona 349 Punjab Sodic soils of Punjab and Haryana Chaul local T 21 Uttar Pradesh Sodic soils of Uttar Pradesh Kashi Lakra or T 22A Uttar Pradesh Sodic soils of Uttar Pradesh Bejhari Uttar Pradesh Sodic soils of Uttar Pradesh Kalambank SR 26B Orissa Coastal saline soils of east coast Local collection SR 8 Orissa Coastal saline soils of Orissa Geti Orissa Coastal saline soils of Orissa Raspanjore Orissa Coastal saline soils of Orissa Lunpatali Orissa Coastal saline soils of Orissa Choudusannalu Andhra Pradesh Saline soils of Andhra Pradesh Budda Mologolakulu MCM 2 Andhra Pradesh Coastal saline soils of Andhra Pradesh Tellathokavadlu Andhra Pradesh Coastal saline and sodic soils of Tamil Nadu Kalar samba Tamil Nadu Coastal saline and sodic soils of Tamil Nadu Chuvada samba Tamil Nadu Coastal saline and sodic soils of Tamil Nadu Kullakar Tamil Nadu Coastal saline and sodic soils of Tamil Nadu Vellaikattai Tamil Nadu Coastal saline and sodic soils of Tamil Nadu Kallimadyan Tamil Nadu Coastal saline and sodic soils of Tamil Nadu Local Rata Kala Rata 1-24 Maharashtra Coastal saline soils of Maharasthra Local Rata Bhura Rata 4-10 Maharashtra Coastal saline soils of Maharasthra Morchuku Maharashtra Coastal saline soils of Maharasthra Dodka Maharashtra Coastal saline soils of Maharasthra Khare Bhat Maharashtra Coastal saline soils of Maharasthra Mahadi Rata Maharashtra Coastal saline soils of Maharasthra Patnai Patnai 23 West Bengal Coastal saline soils of West Bengal Patnai Patnai 298 West Bengal Coastal saline soils of West Bengal Damodar CSR1 West Bengal Coastal saline soils of West Bengal Dasal CSR2 West Bengal Coastal saline soils of West Bengal Getu CSR3 West Bengal Coastal saline soils of West Bengal Benisail Matla West Bengal Coastal saline soils of West Bengal Nona Bokra Hamilton West Bengal Coastal saline soils of West Bengal Velki (Bhaluki) West Bengal Coastal saline soils of West Bengal and Orissa Rupsal West Bengal Coastal saline soils of West Bengal Kumargore West Bengal Coastal saline soils of West Bengal Nonasail CSR 6 West Bengal Coastal saline soils of Orissa and West Bengal Arya Arya 33 Karnataka Saline and sodic soils of Karnataka Karekagga Ankola Karnataka Saline and sodic soils of Karnataka Bilekagga Karnataka Saline and sodic soils of Karnataka Bali Kerala Coastal saline soils of Kerala Oarkzhama Kerala Coastal saline soils of Kerala Pokkali Kerala Coastal saline soils of Kerala Oarpandy Kerala Coastal saline soils of Kerala Ormundakan Kerala Coastal saline soils of Kerala Odacheera Kerala Coastal saline soils of Kerala Chettivirippu Mo 1 Kerala Coastal saline soils of Kerala Kalladachampavu Mo 2 Kerala Coastal saline soils of Kerala Kunjathikkara Mo 3 Kerala Coastal saline soils of Kerala Chottupokkali Vytilla 1 Kerala Coastal saline soils of Kerala Korgut Goa Coastal saline soils of Goa Azgo Goa Coastal saline soils of Goa
Breeding rice for salt-affected areas of India 85 Table 8. Salt-tolerant rice varieties evolved through recombina- tion—past efforts (before 1970). a depth of about 0.8 to 1 m. Artificially prepared soil of dif- ferent EC (maintained by irrigating with saline water) or pH Variety Parentage Institution is maintained in these microplots in a manner that very much reflects field conditions except for the better control of soil PVR 1 MTU 1/SR 26B Rice Research Station, Peravoorani, heterogeneity. The plot size is usually very small (a single Tamil Nadu row) but, because of good control over micro environment, MCM 1 CO 18/Kuthir Rice Research Station, Machlipatnam, Andhra Pradesh the screening is highly reliable and repetitive. The microplots MK47-22 Malkudai/KR 1-24 Kharland Research Station, Panvel, are being used to screen mostly early segregating and stable Maharashtra populations besides genetic studies. SR 3-9 KR 1-4/Zinnya 149 Kharland Research Station, Panvel, Maharashtra Screening in pots MR 18 SR 26B/Wannar-1 Rice Research Station, Mandya, Karnataka Pot screening is mostly conducted for precise, controlled studies of individual plant response under a constant stress. The edaphic environment in pots is more or less uniform throughout the plant growth period with respect to the extent best way to select adapted genotypes and do final evaluation. of stress. Spatial variability is determined by extensive soil testing of the trial fields. Screening in solution culture Salinity. Solution culture techniques are being used in two Screening in microplots ways, one for screening during the seedling stage and the Soil heterogeneity and spatial variability are the major factors second for screening up to maturity. In the first category, that sometimes limit the reliability of field-testing data. This 7-day-old seedlings grown on nonstress modified Yoshida attracted the development of mini-field environments with culture solution (Yoshida et al 1976) are transferred to the varying controlled salinity and sodicity in plots built out of desired amount of stress either in a bread box with a perfo- brick-mortar-concrete materials, measuring 6 m by 3 m, with rated lid or perforated styrofoam with a bottom mesh. This modified culture solution has KH2PO4 and K2HPO4 in place
Table 9. Rice varieties released for salt tolerance in India.
Stress tolerance limit Stress adaptation Year of Days to 50% Name of variety Parentage (type of soil) release flowering Grain typea
pH2 ECe (dS m–1)
Notified and released by center CSR 10 M40-431-24-114 / Jaya 10.2 11 Alkaline and 1989 100 SB inland saline CST 7-1 Damodar/IR24 – 8 Coastal saline 1991 115 LB Lunishree Mutant of Nonasal – 8 Coastal saline 1992 115 LS CSR 13 CSR 1/Bas. 370//CSR 5 9.9 8 Alkaline and 1999 115 LS inland saline CSR 27 (derived from Nona Bokra/IR5657-33-2 9.9 8 Alkaline and 1999 95 LS IR51471) coastal saline CSR 30 (Yamini) BR 4-10 / Pak. Basmati 9.5 – Normal and 2001 120 LS (Basmati) Alkaline CSR 23 (derived IR64//IR4630-22-2-5-1- 10.0 8 Alkaline and 2004 105 MS from IR52713) 3/IR9764-45-2-2 coastal saline Sumati Pankaj/NC 678 – 8 Coastal saline 2004 105 SB Bhutnath SR26B/Pankaj – 8 Coastal saline 2004 Photo-senst. SB Jarava (B90-15) B32-Sel-4/Spontanea 4 – 8 Coastal saline 2005 Photo senst. SB (IET 15420) //B296 CSR36 (Naina) CSR 13/Panvel 2//IR36 9.9 – Alkaline 2005 110 LS (IET 17340) CSRC(S)7-1-4 Pankaj/SR26B – 9 Coastal saline 2008 115 LB (IET 14199/ 18250)
Continued on next page
86 Singh et al Table 9 continued.
Stress tolerance limit Stress adaptation Year of Days to 50% Name of variety Parentage (type of soil) release flowering Grain typea
pH2 ECe (dS m–1)
Developed/released by state Haryana/Punjab CSR 5 (Vikas) TKM6/IR8 9.7 7 Alkaline 1979 95 MS Kerala Vyttila 1 Selection from Chotupok- 4.5–9.7 9 Coastal saline – 90 MB kali Vyttila 2 Selection from Cheruvir- 4.5–9.7 8 Coastal saline 1980 95 LB uppu Vyttila 3 Vyttila 1/TN 1 4.5–9.9 9 Coastal saline 1987 85 MB Vyttila 4 Chethivirippan/IR4630- 4.5–9.7 10 Coastal saline 1993 95 LB 22-2-17 Vyttila 5 Mutant of Mahsuri 4.5–9.5 8 Coastal saline 1996 95 MB Vyttila 6 Cheruviruppu/IR 5//Jaya 4.5–9.7 9 Coastal saline 2004 85 MB
Maharasthra Panvel 1 IR8/BR4-10 9.7 8 Coastal saline 1984 95 SB Panvel 2 BR4-10/IR8 9.8 9 Coastal saline 1988 90 LS Panvel 3 Damodar/Pankaj 9.8 9 Coastal saline 2000 100 LB
Andhra Pradesh MCM 1 CO 18/Kuthir – 8 Coastal saline 100 SB West Bengal Mohan (CSR 4) Mutant of IR8 – 7 Coastal saline 1981 115 SB
Karnataka CSR 22 IR64//IR 4630-22-2-5-1- 9.9 8 Inland saline 2006 105 MS 3/IR 9764-45-2-2 Orissa SR 26B Selection from Kalambank 9.7 8 Coastal saline 1988 115 LB TamilNadu PVR-1 MTU-1/SR 26B 9.4 4 Coastal saline 1964 115 SB CO 43 IR20/Dasal 9.7 8 Alkaline and 1982 110 MS coastal saline ASD 16 ADT 31/CO 39 9.6 7 Coastal saline 1986 95 SB TRY 1 IR578-172-2-2/BR 1-2- 10.0 10 Alkaline and 1995 110 SB B-19 coastal saline TRY 2 IET 6238/IR36 9.9 8 Alkaline and 2001 85 LS coastal saline
Gujarat SLR51214 Vijay/PTB-21 – 8 Inland and coastal 1982 100 LS saline Dandi (IET 14906) Panvel 2/IET 8320 – 9 Inland and coastal 2003 95 LS saline
Uttar Pradesh Usar 1 Jaya/Getu 9.9 8 Alkaline 1985 100 SB Narendra Usar 2 Selection from the IR 9.8 8 Alkaline 1998 95 LB 2058 derivative Narendra Usar 3 Selection from IR 46330 9.9 8 Alkaline 2000 100 LS derivative aSB = short bold, LB = long bold, LS = long slender, MS = medium slender, MB = medium bold.
Breeding rice for salt-affected areas of India 87 of NaH2PO4 as sodium salt within culture solution and may Mutation breeding was also used before, as in the increase the Na ion concentration (T.J. Flowers, personal development of the first salt-tolerant variety, CSR 10. The communication). However, for screening for tolerance during female parent of CSR 10, M40-431-24-114, was derived from the reproductive stage, 2–3-week-old seedlings are trans- the γ-irradiated F1 seed of the cross CSR 1/IR8. Similarly, planted in pots filled with ½ kg of soil and irrigated through an Lunishree and CSR 4 were also developed through mutation automated circulating irrigation system. After establishment breeding. Shuttle breeding was adopted in the development of of the seedlings for 3–5 days, the salinity is raised to 50 mM salt-tolerant varieties such as CSR 23 and CSR 27 under the in culture solution stored in an underneath tank using NaCl. ICAR-IRRI collaborative project for salt tolerance (Mishra This is further raised to 70 mM salinity after 10–15 days. The 1994). automated circulatory irrigation system maintains the desired exact salinity in the rhizosphere. Plants are irrigated from the Nonconventional breeding lower tank through a timer-controlled pump. Anther culture derivatives generated at IRRI and evaluated Sodicity. A TRIS buffer-based system has been devised under the ICAR-IRRI collaborative network in India by to screen rice genotypes, in which a certain pH is maintained CSSRI led to the identification of promising breeding lines in solution and micronutrients are supplied through foliar IR51500-AC-17, IR51485-AC-1, and AC6534-4 for salin- sprays. Genotypic performance is evaluated from the change ity; AC6533-3 for sodicity; and AC6534-1 for dual tolerance in seedling weight and phenotypic performance. A TRIS (Singh et al 1992, Singh and Mishra 1995). IR51500-AC- (Tris hydroxymethyl aminomethane) concentration of 4 mM 17 and AC6534-1 were later named as CSR 21 and CSR (pH 8.6) provides a means for evaluating genotypic differ- 28, and were recommended by the All India Coordinated ences in response to sodicity. Details are given in Singh et Rice Improvement Program for release. Rice breeding line al (2002). IR51500-AC-17, selected in India and identified as CSR Screening in trays. For large-scale screening of varie- 21, was released as PSBRc50 or Bicol in the Philippines ties at germination and seedling stage, salt-affected soil filled (Senadhira et al 2002). in shallow-depth wooden or metal germination trays with a polythene sheet lining on the inner face is used. This is quite Impact of salt-tolerant rice varieties: geographical convenient for control of salinity, sodicity, and moisture. distribution, area covered, and obstacles in uptake and seed production Breeding methods used for developing salt-tolerant rice varieties Salt-tolerant rice varieties are widely grown in the salt-af- fected soils of different states and are in great demand by Conventional breeding farmers. It is difficult to realize the direct impact of these Mostly conventional breeding techniques such as introduc- varieties but their adoption, popularity, and impact could be tion, selection, hybridization, mutation, and a shuttle-breed- measured through their breeder and certified seed production ing approach were used for developing salt-tolerant rice and distribution chains. To quantify the impact of salt-tolerant varieties. Varieties such as Damodar (CSR 1), Dasal (CSR rice varieties, seed distributed by CSSRI, Karnal, to various 2), and Getu (CSR 3) are pure-line selections of the local salt- seed-multiplying agencies was taken as a case study. Until tolerant traditional cultivars that prevailed in the Sunderban 2004, CSSRI supplied 124.42 quintals of breeder seed and areas in West Bengal. These varieties were adapted to salinity 2,387 quintals of certified seed of all the released salt-tolerant stress but, when they were first introduced in sodic areas at rice varieties to various seed multiplication agencies and land Karnal, they were found to be tolerant of sodicity as well. development corporations of the different states. Almost all other released varieties of rice were developed The seed multiplication ratio of breeder seed to certified using one or more of the following techniques: seed is taken as 1:150 in rice (when the breeder to founda- l Pedigree method tion and foundation to certified seed chain is not included in l Modified bulk pedigree method: Individual selected this case study; otherwise, it would be many thousand-fold). F2 plants are bulked up to the F4 generation, followed If we extrapolate the multiplication ratio of rice, then 124 by panicle selection and then advancing the popula- quintals of breeder seed would have produced a minimum of tion as in the pedigree method. 18,600 quintals of certified seed in one production season. In l Shuttle breeding addition to this, CSSRI supplied 2,387 quintals of certified The segregating materials are grown in long rows under seed, so the total certified seed quantity would be 20,987 salt stress, with space planting, particularly in F2. Selection quintals within one season. The seed produced for sowing pressure is gradually increased with generation advance- is always much more expensive than that used for consump- ment simultaneously under moderate stress and high stress tion. Even with conservative estimates, if we consider only of sodicity or salinity. At the F5 or F6 generation, a part of 20,000 quintals of total certified seed (instead of 20,887 q) the material is screened at high pH of up to 9.9–10.0 and in the production chain, which might have covered more salinity of about 10–12 dS m–1. than 50,000 ha of area, this would have produced more than 1.5 million quintals of paddy (taking average production of
88 Singh et al only 30 q ha–1). This additional food grain from salt-affected Example of the UP Land Development Corporation (UPLDC) soils must be worth several million rupees, besides providing in disseminating salt-tolerant varieties employment opportunities. Indian farmers usually do not replace their seed every season and keep it for a few years Land development programs in Uttar Pradesh were started before seed replacement. The above estimate is only a one- in the 1990s. A land development corporation oversees the season scenario; therefore, we could expect an even greater overall availability of farm inputs, including seeds of salt- impact after a few seasons. Besides CSSRI, Karnal, many tolerant varieties, and with back-up support of technological seed multiplication agencies are continuously engaged in know-how. In 1993, the UPLDC started an ambitious project seed supply to the land development corporation of different of reclaiming 47,000 ha of sodic land in phase I with help states, other agencies, and directly to farmers. Because of this from the World Bank and European Economic Community scenario, the actual impact of salt-tolerant rice varieties will (EEC). Through concerted efforts, UPLDC successfully be much greater than what was predicted above. doubled the targets by reclaiming 97,056 ha of sodic land Another notable impact is bringing back unproductive within 1993-2000 (UPBSN Status Report, April 1999, Am- barren soils into rice farming or enhancing the yield on soils bekar 2004). UPLDC targeted 150,000 ha of sodic land for with very low productivity. Two scenarios are possible. In the reclamation in its phase II program (2000-05). Up to March first, a new high-yielding variety is replacing another old va- 2002, it had already reclaimed 126,861 ha, which is also riety and increasing yield from 5 to 6 t ha–1 with an advantage ahead of its scheduled targets. of 1 t ha–1 only. In the second scenario, unproductive barren In both phases, the major cause of this great success land is cultivated using a salt-tolerant rice variety with no or was the concerted efforts and timely availability of resources. a small amount of gypsum, which gives yield of about 3.0 t As per the UPLDC Status Report (April 1995), rice variety ha–1 in the first year and probably more in subsequent years. CSR 10 produced its maximum yield under C class soils in Comparing both scenarios, though productivity is more in comparison with other rice varieties. C class soils are those the first scenario, the yield advantage is about 3 times more where no crop has been grown before during both the wet in the second scenario in addition to more land being put and dry seasons. Until now, UPBSN, Lucknow, is the bulk under cultivation. Clearly, the development and deployment seed purchaser of salt-tolerant rice varieties from CSSRI and of salt-tolerant rice varieties could allow both horizontal and other seed-producing agencies for its sodic soil reclamation vertical expansion in food production. program. Use of these salt-tolerant varieties is an integral component of its reclamation technology in farmers’ fields. Seed production and availability problems Breeding objectives and challenges for Despite the urgent need of seed of salt-tolerant rice varie- the coming 5–10 years—opportunities ties, seed-producing agencies are not effective in producing for new and enhanced strategies sufficient seed to meet local demand except in a few areas. The major bottlenecks are If we compare the progress in breeding for salt tolerance with l Risk aversion by farmers. Because of uncertainty of that for irrigated ecosystems, it is apparent that the first pro- production in these unfavorable areas, farmers do gram is still far behind. The immediate concern is to raise the not want to invest in costly inputs, including seed. average yield of salt-affected areas from the current 1.0–1.2 t l Low demand for seed as landowners of salt-affected ha–1 to about 2.0 t ha–1. As such, rice production will increase areas are mostly resource poor. by more than 8.5 million tons. Salinity, sodicity, or any other l Problem areas need mostly site-specific varieties. salt-related stress seldom happens in isolation and more ef- Seed-multiplying agencies do not want to invest in forts are needed to ensure site-specific evaluation of breeding producing a relatively small amount of seed of many lines at target sites. Sharing of resources through germplasm target-specific varieties. exchange networks is the first step toward a proper diagnosis l Salt-tolerant varieties adapted to multiple stresses of problems specific to each site and also for the develop- such as salinity and submergence or sodicity and ment of adapted varieties. As the latest technologies such tolerance of P and Zn deficiency are not yet avail- as marker-assisted selection become available to breeders, able. more target-specific varieties with enhanced tolerance of salt These problems could be solved to some extent if the stress as well as other prevailing stresses will be developed. state government-owned land development corporations With the tagging of agronomically important QTLs such as could ensure the procurement of seed of salt-tolerant rice Saltol, Sub1, and Pup1, the complexities of breeding for these varieties, and help resource-poor farmers with substantial environments could be eased and future programs could focus backup of technological know-how and access to other re- on developing marker systems and strategies to introgress sources needed for reclamation or as an input. these multiple major QTLs and genes into agronomically superior backgrounds that meet the quality requirements of farmers in any target area.
Breeding rice for salt-affected areas of India 89 References Singh RK, Mishra B, Senadhira D. 1992. Promising salt tolerant
F1 anther culture derivatives (ACDs). Int Rice Res Newsl. Abrol IP, Bhumbla DR. 1971. Saline and alkali soils in India – their 17(1):17. occurrence and management. FAO World Soil Resources Rep. Singh RK, Mishra B, Chauhan MS, Yeo AR, Flowers SA, Flowers TJ. p 41:42-51. 2002. Solution culture for screening rice varieties for sodicity Ambekar VW. 2004. Sodic land reclamation in Uttar Pradesh: some tolerance. J. Agric. Sci. Cambridge (UK) 139:327-333. policy issues. In: Report of International Conference on Sustain- Singh RP, Vasisht AK, Mathur VC. 2003. Quantitative assessment of able Management of Sodic Soils held at Lucknow, India, 9-14 economic losses of degraded land in India. Division of Agri- February 2004. p 14-18. cultural Economics, IARI, and Advance Publishing Concept, CSSRI. 1997. Vision 2020: CSSRI Perspective Plan. CSSRI, Karnal. New Delhi. 51 p. 95 p. Swaminathan MS.1999. Predict: century of hope. Towards an era of Directorate of Rice Research. 2000. High-yielding rice varieties of harmony with nature and freedom from hunger. East West Books India—2000. DRR, Hyderabad. 102 p. (Madras) Pvt. Ltd. p 155-199. Mishra B. 1994. Breeding for salt tolerance in crops. In: Rao et al, edi- UPBSN. 1999. Status report: UP sodic lands reclamation project. UP tors. Salinity management for sustainable agriculture—25 years Bhumi Sudhar Nigam (erstwhile UPLDC), Lucknow. 53 p. of research at CSSRI. Central Soil Salinity Research Institute, UPLDC. 1995. Status Report: UP sodic lands reclamation project. Part Karnal, India. p 226-259. I. UP Land Development Corporation, Lucknow. 78 p. Paroda RS. 1998. Sustaining the Green Revolution: new paradigms. Yoshida S, Forna, DA. Cock JH, Gomez KA. 1976. Laboratory manual Dr. B.P. Pal Memorial Lecture. In: Proceedings of the 2nd In- for physiological studies of rice. Manila (Philippines): Interna- ternational Crop Science Congress. New Delhi (India): Oxford tional Rice Research Institute. 34 p. & IBH. p 79-110. Senadhira D, Zapata FJ, Gregorio GB, Alezar MS, de la Cruz HC, Notes Padolina TF, Galvez AM. 2002. Development of the first salt- tolerant rice cultivar through indica/indica anther culture. Field Authors’ addresses: R.K. Singh, A.M. Ismail, and G.B. Gregorio, Crops Res. 76:103-110. International Rice Research Institute, Los Baños, Philippines; Singh NT. 1992. Land degradation and remedial measures with B. Mishra, project director, Directorate of Rice Research, Hy- reference to salinity, alkalinity, waterlogging and acidity. In: derabad, India. Deb DL, editor. Natural resources management for sustain- able agriculture and environment. New Delhi (India): Angkor Publications. 442 p.
Singh RK, Mishra B. 1995. Screening F1 anther culture derivatives of rice for salt tolerance. In: Sharma B et al, editors. Genetic research and education: current trends and the next fifty years. New Delhi (India): Indian Society of Genetics and Plant Breed- ing. p 509-513.
90 Singh et al Upland ecosystems
Breeding rice for the sloping uplands of Yunan 91 92 Tao et al Breeding rice for the sloping uplands of Yunnan
D. Tao, F. Hu, G.N. Atlin, S. Pandey, P. Xu, J. Zhou, J. Li, and X. Deng
Farmers have been growing upland rice for more than 4,000 years in Yunnan Province of China using traditional varieties under shifting cultivation systems that are based on slash-and-burn practices. Even now, upland rice production is still pri- marily a subsistence-oriented activity in southern Yunnan, with farmers using a large share of their resources to meet food needs. Our efforts over the years have been to develop japonica cultivars with strong seedling vigor, drought tolerance, blast resistance, adaptation to the intermediate elevation of southern Yunnan, and weed competitiveness. In the meantime, germ- plasm introduction from outside Yunnan has been kept through active domestic and international cooperation since 1989. These endeavors have led to the release of varieties IRAT104, B6144F-MR-6, Yunlu 29, and Yunlu 52 in Yunnan. The goal of hybridization breeding in Yunnan is to combine the good adaptation of Yunnan traditional upland rice cultivars with good plant type, blast resistance, and the drought tolerance of foreign upland rice accessions. The keys in our inbred and hybrid breeding programs are the identification of donors for intermediate stature and blast resistance, of recombinants of aerobic adaptation and high yield, and of promising crosses. A traditional upland cultivar, Mengwanggu, usually yields about 2 t ha–1 with optimized fertilizer management on terraces and favorable lower slopes, whereas improved cultivars Yunlu 29, Yunlu 52, and B6144F-MR-6 yield 4 t ha–1. The yield advantage of the improved cultivars results mainly from the significant increase in tiller and panicle size. In 2002, breeding research shifted from breeding rice for sloping uplands to developing aerobic rice with high yield, blast resistance, and fine grain quality under favorable environments and good management. Indica type, because of its input response, became our major target type. One strategy for improving the yield potential of aerobic rice cultivars is to transfer their aerobic adaptation to lowland high-yielding varieties (HYVs), which partition up to 50% of total dry matter to grain, but do not perform well under upland conditions.
Upland areas in Yunnan are characterized by rugged terrain, In this region, upland rice planting area has been stable poor access to markets, environmental degradation, and a at 100,000 hectares, with an average yield of about 2.5 t ha–1. high incidence of poverty. These are also border areas inhab- As upland rice cultivation is related to the food security of ited by ethnic minorities. Farmers have been growing upland the people based in hilly regions, it is not possible to replace rice for more than 4,000 years using traditional varieties under upland rice by other crops in a short time (Zhou 1987). How- shifting cultivation systems that are based on slash-and-burn ever, sloping fields can be changed into terraces to control practices (Zhou 1982). Despite increasing commercialization water and soil erosion, fertilize soil, use agricultural imple- of agriculture, upland rice production is still primarily a sub- ments, use improved input-responsive cultivars, and finally to sistence-oriented activity in southern Yunnan, with farmers increase yield per ha. Among the numerous factors leading to using a large share of their resources (primarily land and increased upland rice yield, breeding for improved cultivars labor) to meet food needs. Self-sufficiency in food production and their cultivation have an important role to play. However, is a dominant objective of upland farmers. Income-generating the Yunnan Seed Management Station (1992) reported that commercial activities and eco-reservation activities thus need around 129 accessions are the most common upland rice to be based on the improvement of household food security, cultivars grown by farmers in upland areas, out of which 119 which often means that households are able to produce a bulk are japonicas and 10 are indicas. None of these cultivars has of their own food requirements. a planting area of more than 6,700 ha. Since 1980, breeding In five prefectures (Wenshan, Honghe, Puer, Xishuang- and the introduction of upland rice cultivars have been one bana, and Lincang) of southern Yunnan, upland rice is one of of the leading research projects at the Yunnan Academy of the three staple food crops (irrigated rice, maize, and upland Agricultural Sciences (YAAS). Our manuscript summarizes rice). The government is paying more and more attention to the achievements and experiences of 25 years of research on poverty alleviation for people located in these regions and the upland rice improvement at YAAS. development of upland rice is one of the integrated measures for becoming self-sufficient in grain production.
Breeding rice for the sloping uplands of Yunan 93 Target environments and breeding objectives Development Association (WARDA), Centro Internacional de Agricultura Tropical (CIAT), Institute for Research in ο Upland rice cultivation in Yunnan extends from 21 to 28 N Tropical Agriculture (IRAT, now CIRAD), and Brazil. and from 73.2 to 2,200 m above sea level. Among 129 rice- Upland rice breeders from IRRI, CIAT, CIRAD, IITA, IRD growing counties in Yunnan, upland rice cultivation is prac- (Institut de recherche pour le développement), WARDA, and ticed in 50 to 60 counties, accounting for 13% of the total Southeast Asia visited Yunnan, and upland rice breeders from rice area (Xu 1993). Of 179,866 ha of upland rice in 1982 YAAS visited IRRI, CIAT, CIRAD, IRD, WARDA, Brazil, in Yunnan, 87% was distributed in five prefectures (Puer, Costa Rica, Côte d’Ivoire, Madagascar, Thailand, Indone- Xishuangbana, Wenshan, Lincang, and Honghe) in south sia, and Vietnam. Seven young scientists from YAAS were ο and southwest Yunnan between 21 and 24 N, and 80% was trained at international agricultural research centers. Active, grown in areas 800 to 1,600 m above sea level. These areas are strong cooperation and exchange between YAAS and foreign mountainous regions inhabited by minor ethnic groups, and research organizations have been set up. they are the target population of environments of our upland As a result, IRAT104, bred by IRAT/CIRAD and breeding programs (Tao 1995, Tao et al 1993, 1996). introduced from IRRI in 1991, was released by the Agri- The predominate germplasm for uplands in Yunnan is cultural Department of Yunnan Province formally in 1995, japonica. However, indica hybrids for the irrigated ecosystem and it became the first improved upland rice variety released have been successfully introduced to uplands below 1,300 m formally by authorities in Yunnan (Hu et al 1997, Tao et al (Tao et al 1993, 1998, 1999). The constraints to upland rice 1997). B6144F-MR-6 was released in Yunnan in 2000, and production in Yunnan are poor soil fertility, weeds, drought, has now become one of the most popular cultivars for uplands and blast disease. Therefore, selection for improved seedling in areas below 1,200 m. vigor, drought tolerance, blast resistance, and weed competi- tiveness is considered in our upland rice breeding programs. Hybridization breeding Our preliminary research revealed that the biomass of tra- ditional upland rice cultivars was about the same as that of The strategy of hybridization breeding in Yunnan is to com- irrigated rice, but their harvest index and spikelets m–2 were bine the advantage of good adaptation and cold tolerance significantly lower than those of irrigated rice. Thus, the ap- of Yunnan traditional upland rice cultivars and good plant proach we apply to increasing yield potential is to decrease type, blast resistance, and drought tolerance of upland rice the plant height of traditional upland rice cultivars to avoid accessions (Tao 1995, Tao et al 1996, 1998, 1999) introduced lodging, while increasing tillering ability and panicle size from outside. (Tao et al 1996). Since 1989, 2,032 combinations of F1 have been made, and 207 advanced breeding lines were developed (Tao et al Approaches 1998). Besides some extension measures such as changing sloping Donors for intermediate stature fields into terraces, reforming cropping systems, applying From 207 advanced breeding lines bred, upland rice acces- chemical fertilizer and herbicides, using irrigated indica sions IRAT104, IRAT216, CT8402-27-M-4-2-3M, CT6947- hybrids for upland planting, and using better traditional varie- 7-1-1-1-7, IR47719-2-2-1-2, CT7979-4-M-7-5-3M, CT9278- ties, higher priority is always given to cultivar improvement. 13-8-9-3-M, and Yunnan irrigated japonica Yunjing 136 were We have had a policy of simultaneous introduction from promising accessions of intermediate stature, especially abroad and breeding by ourselves since 1989 (Hu et al 1996, IRAT104 and IRAT216, which were really effective in reduc- 1997, Tao et al 1993, 1997, 1998). ing the height of Yunnan traditional upland rice (Table 1).
Germplasm introduction from abroad Donors for blast resistance Screening for resistance against blast has been one of the YAAS has been an active member of INGER (International major objectives of our breeding program. Several years of Network for Genetic Evaluation of Rice) and the URRC screening at our breeding site in Puer led to the identification (Upland Rice Research Consortium), and this has provided of nine introduced accessions and 16 Yunnan local upland rice a good opportunity for exchange of information, breed- cultivars with resistance to blast (Table 2). Ancestor analysis ing materials, and donors; scientific visits; and training of of 207 advanced breeding lines indicated that IRAT104, scientists from YAAS at the International Rice Research IRAT216, Bayuenou, Boyiegu, Dahongguxuan-3, Huanpigu, Institute (IRRI), Philippines, and at other institutes. This Yuanjingdao, and Zaxima were important donors of blast has been an important component of our breeding strategy, resistance (Table 3), and improvement in blast resistance is and has strongly improved upland rice breeding in Yunnan. an important area of our breeding program (Table 4). Until 2004, 3,158 accessions of upland rice were introduced from IRRI (mainly via INGER or URRC), the International Institute of Tropical Agriculture (IITA), the West Africa Rice
94 Tao et al Table 1. Frequency of intermediate status ances- tors in 207 advanced breeding lines. Breeding site selection (hot spot) A breeding site should closely resemble the targeted area Ecotype Cultivar Frequency (%) of cultivation, and constraints such as drought, blast, and problem soil should appear every year. Another important Upland rice IRAT104 29.5 issue is the convenience of operations. Puer is an ideal site IRAT216 19.8 for all of this. CT8402-27-M-4-2-3M 8.7 CT6947-7-1-1-1-7 6.3 IR47719-2-2-1-2 6.3 Cross slection of F1 generation CT7979-4-M-7-5-3M 5.8 Genetic research indicated that general combining ability CT9278-13-8-9-3-M 5.3 (gca) was important for the inheritance of heading date, plant height, panicle number per plant, and spikelet number per Irrigated rice Yunjing 136 16.4 IR28 1.9 panicle (Yang et al 1997). Meanwhile, our breeding experi- ence also indicated that cross selection from the F1 generation was simple, convenient, and effective (Fig. 1).
Generation acceleration Table 2. Upland rice resources with blast resistance (scored accord- In Yunnan, upland rice could finish one generation within ing to Standard Evaluation System for Rice).a a year. To accelerate breeding progress, in October after Cultivar 1991 1992 1993 1994 1995 harvest from Puer, seeds can be sown in Sanya, Hainan, and harvested again in March of the next year. In Puer, normal BL PB BL PB BL PB BL PB BL PB sowing time is in April. In this way, two generations can be completed within a year (Yang et al 1995). OS6 3 3 3 5 – – 1 1 – – IRAT8 – 5 – 5 – – – – – – IRAT104 5 1 1 5 5 – 1 3 3 3 Adaptation test IRAT216 3 3 1 5 1 – 1 1 3 3 Since G × E is obvious and important for upland rice, we CT9993-5-10- 3 3 1 5 – – 3 1 3 – began multisite testing in Yunnan (5–10 sites) when a fixed 1-M line was bred to make selection for adaptation. CT10035-32-4- 5 3 3 5 – – 1 1 3 – 4-M CT10035-33-6- 3 1 1 5 – – 3 1 – – Performance of improved cultivars 2-M CT9278-11-1-1- 1 3 3 5 – – 3 7 3 – The results of about 25 years of practice indicated that we 2-M have partially reached our breeding objectives. Yunlu 29 CT9278-11-5-8- 1 1 3 5 – – – 5 3 – and Yunlu 52 were released in 1999 and 2004, respectively. 2-M Haobingxuan – 1 – 5 – – – 5 – – Both became important improved cultivars used by farmers Epian – 1 – 5 – – – – – – in uplands above 1,200 m. Bayuenou – 1 – 5 – – – 0 3 3 On terraces and favorable lower slopes, and with Xiaobaigu – 1 – 5 – – – 1 3 3 optimized fertility management, Mengwanggu (traditional Zibo – 5 – 5 – – – 1 1 1 upland cultivar) usually has about 2 t ha–1 yield, whereas Ela – 1 – 5 – – – 1 3 1 –1 Huanpigu – 1 – 5 – – – 1 3 1 Yunlu 29, Yunlu 52, and B6144F-MR-6 usually yield 4 t ha Zaxima – 3 – 5 – – – 1 3 1 (Table 5). Improved upland rice cultivars (aerobic rice) have Biangyidabaigu – 3 – 5 – – – – 3 3 been shown to outyield traditional varieties by more than Haohai – 1 – 5 – – – 0 1 1 1 t ha–1. This difference is mainly because of the increase in Dahongguxuan-3 – 3 – 5 – – – 0 – – panicles m–2 and spikelets panicle–1 (Atlin et al 2006, Tao Hongxuan No.4 – 1 – 5 – – – 0 – – et al 1999). Boyiegu – 3 – 5 – – – – – – Sanbang 70 Lo – 1 – 5 – – – 1 – – Yuanjingdao – 3 – 5 – – – 0 3 1 New challenges Sanlicun – 3 – 5 – – – – – – aBL = blast, PB = panicle blast. According to Yunnan government policy, in 2005, upland rice area was to be 667,000 ha instead of the 100,000 ha in 2001. However, all rice was to be planted on terraces or flat fields with new cultivars, suitable inputs, and appropriate manage- ment. Average yield was projected to be 3 t ha–1. Traditionally, upland rice breeding was for sloping uplands, emphasizing yield stability, blast resistance, and tolerance for low soil fertility. Food security of farmers de- pends mainly on planting area and the application of chemical
Breeding rice for the sloping uplands of Yunan 95 Table 3. Frequency of blast resis- tance ancestors in 207 advanced Improved cultivars cover about 30% of the upland rice area breeding lines. in Yunnan. In recent years, we have paid special attention to Donor Frequency (%) returning slope fields to forestry and grasses, constructing terraces, and alleviating poverty via the extension of science IRAT104 29.5 and technology. IRAT216 19.8 Bayuenou 13.2 Slash-and-burn systems for upland rice have been Boyiegu 9.2 replaced by intensive management with new cultivars and Dahongguxuan-3 6.3 new technologies on terraces or flat fields. The food security Huanpigu 3.9 of upland rice farmers has greatly improved. Thus, farmers Yuanjingdao 3.9 could have surplus resources for planting cash crops and do- Zaxima 2.4 ing livestock farming. The economic situation in rural areas has improved greatly because of the improvement in food security (Kam 2003). But, eco-environment deterioration Table 4. Progress of blast resistance has not been thoroughly reversed. The adoption of a package improvement (scored according to Standard Evaluation System for for upland rice development is still low, and the nature of Rice). diversification in Yunnan is little understood. On terraces and favorable lower slopes, and with op- Year Generation Planted Selected timized fertility management, improved (aerobic) cultivars have been shown to outyield traditional varieties by more –1 1992 F2 7.59 7.27 than 1 t ha . But, currently, yields of improved rice cultivars F3 7.04 6.29 have reached a plateau of 4 to 5 t ha–1 in farmers’ fields (Atlin F 6.73 6.68 4 et al 2006). To consistently obtain higher maximum yields, however, the current plant type of aerobic rice cultivars 1994 F2 5.48 3.76 F3 4.27 3.17 will have to be further improved. Currently used varieties F4 5.00 5.00 B6144-MR-6-0-0 and Yunlu 52 are relatively tall, leafy, and F5 3.66 3.00 low-tillering, rarely achieving a harvest index of 0.4 or a F6 3.88 3.66 2 F 3.67 2.30 panicle number above 250 per m . One strategy for improv- 7 ing the yield potential of aerobic rice cultivars is to transfer their aerobic adaptation to lowland high-yielding varieties, which partition up to 50% of total dry matter to grain, but fertilizers and herbicides. Now, farmers have to improve their which do not perform well under upland conditions. food security from less land, but with high yield potential, input-responsive cultivars, and good resource management References because of high population pressure and rural development. Thus, starting in 2002, all trials had to be conducted on Atlin GN, Lafitte HR, Tao D, Laza M, Amante M, Courtois B. 2006. terraces, and the breeding objective shifted to high yield Developing rice cultivars for high-fertility upland systems in (nitrogen-responsive, lodging-resistant), fine grain quality, the Asian tropics. Field Crops Res. 97:43-52. and blast resistance. Of course, the indica type, because of Hu F, Tao D, Yang G, Yang J. 1996. Preliminary studies on the intro- its high input responsiveness, became our main target type. duction of upland rice from abroad. Southwest China J. Agric. Within a package for upland rice development in Yun- Sci. 9(2):8-11. nan, improved cultivars play a basic and irreplaceable role.
1990 Summer 1990 1991 1991 1992 1992 1993 1994 1995 Winter Summer Winter Summer Winter Summer Summer Summer
37 F2 32 F3 16 F4 15 F5 12 F6 10 F7 8 F8 7 F9 (Better)
765 F1
168 F2 6 F3 6 F4 4 F5 4 F6 4 F7 3 F8 (OK)
Fig. 1. Effect of F1 selection. (In 1990 in Simao, 37 crosses with PAcp scale 1–3 were called a better cross, and 168 combinations with PAcp scale 5 were called ok crosses and are shown here. The other 560 crosses with PAcp 7–9 were discarded from 765 F1 cross combinations.)
96 Tao et al Table 5. Performance of improved upland rice cultivars in Yunnan (summarized from reports of Yunnan Province Regional Trial for Upland Rice, 1993-2002).
Days to Plant ht. Panicles Spikelets Fertility 1,000-grain Grain yield (t ha–1) Cultivar heading (cm) (no. m–2) (no. panicle–1) (%) weight (g) Short-season Long-season
Mengwanggu (check) 103 121 168 119 77 30 2.57 ± 0.10 2.92 ± 0.12 Yunlu 29 103 111 245 123 80 27 3.34 ± 0.20 3.29 ± 0.24 Yunlu 52 109 114 191 172 78 27 2.86 ± 0.23 4.00 ± 0.28 B6144F-MR-6 105 89 266 120 71 23 1.70 ± 0.17 4.13 ± 0.20
Hu F, Tao D, Yang G, Yang J. 1997. Genealogical analysis of IRAT Xu P. 1993. Eco-characterization of upland rice in Yunnan. J. Shanxi upland rice cultivars. In: Poisson C, Rakotoarisoa J, editors. Normal Univ. (Nat. Sci. Ed.) 21(Sup.):8-12. Riziculture d’altitude, Actes du Seminaire riziculture d’altitude. Yang G, Tao D, Hu F. 1995. A research on the change of the heading Montpellier (France): CIRAD-CA. p 181-184. period and plant height of various paddy rice and upland rice in Kam SP. 2003. Integrated natural resource management for rice produc- Sanya and Simao. J. Southwest Agric. Univ. 17(4):338-343. tion. Int. Rice Res. Notes 28(2):12-18. Yang G, Tao D, Hu F, Yang J. 1997. Studies on combining ability of Tao D, Hu F, Neng N, He Y. 1993. Introduction of upland rice from the main economic characters in upland rice. Chinese J. Rice abroad. J. Shanxi Normal Univ. (Nat. Sci. Ed.) 21(Sup.):32- Sci. 11(2):77-82. 38. Yunnan Seed Management Station. 1992. Records of rice cultivars in Tao D. 1995. Upland rice breeding. In: Jiang Z, editor. Rice in Yun- Yunnan.. Yunnan Seed Management Station, Kunming, China. nan. Kunming (China): Yunnan Science and Technology Press. p 179-220. p 221-230. Zhou J. 1982. History and presence of upland rice in Yunnan. Yunnan Tao D, Hu F, Yang G, Yang J. 1996. Upland rice research in Yunnan, Agric. Sci. Technol. 5:22-26. China. In: Piggin C, Courtois B, Schmit V, editors. Upland rice Zhou J. 1987. Regrowing up of upland rice planting. Yunnan Agric. research in partnership. IRRI Discussion Paper Series No. 16. Sci. Technol. 4:22-24. Manila (Philippines): International Rice Research Institute. p 96-102. Notes Tao D, Hu F, Yang G, Yang J. 1997. Introduction and utilization of IRAT/CIRAD upland rice cultivars. In: Poisson C, Rakotoarisoa Authors’ addresses: D. Tao, F. Hu,, P. Xu, J. Zhou, J. Li, and X. Deng, J, editors. Riziculture d’altitude, Actes du Seminaire riziculture Food Crops Research Institute, Yunnan Academy of Agricultural d’altitude. Montpellier (France): CIRAD-CA. p 35-38. Sciences, Kunming 650205, China; G.N. Atlin, S. Pandey, Inter- Tao D, Hu F, Yang J, Cheng H. 1998. Upland rice breeding strategy national Rice Research Institute, Los Baños, Philippines. in Yunnan, China. Presented at the VI National Rice Research Conference and the Third Upland Rice Breeders Workshop, Goiânia, Goiás, Brazil, 9-13 March 1998. Tao D, Hu F, Yang Y. 1999. New upland rice cultivars and their culti- vation. Reported in China/Myanmar Alternative Development Cooperation Meeting, 5-9 April 1999, Simao, Yunnan, China.
Breeding rice for the sloping uplands of Yunan 97 Progress of upland rice breeding in Indonesia since 1991
Suwarno, E. Lubis, and B. Kustiano
The major constraints to high yield of upland rice are blast disease and aluminum toxicity in humid regions, drought and short duration of the wet season in arid regions, and shading in interplanting systems. Breeding programs for upland rice to develop improved varieties with high yield and resistance to or tolerance of the respective major constraints have been conducted continuously in Indonesia. Traditional local varieties and introduced varieties from IRRI and other countries were used as genetic sources for the desirable characteristics. Several improved varieties have been released with major improvements in yield, maturity, blast resistance, and tolerance of Al toxicity, drought, and shading. Some promising lines containing two blast resistance genes and tolerance of Al toxicity have been developed and breeding lines with different blast resistance have also been selected. Different sources for blast resistance, including monogenic lines and traditional varieties, were used to diversify the genetic base for blast resistance of the improved varieties. Cultivation of the improved varieties with proper crop management to achieve high yield was demonstrated. Further breeding programs will emphasize blast resistance, Al-toxicity tolerance, drought tolerance, very early maturity, and grain quality. Diversification of blast resistance has been put into action in the breeding program.
Indonesia has an upland area of about 5 million hectares, the production of upland rice could therefore be achieved which can potentially be used for producing food crops. The through an intensification program using existing cultivation upland rice cultivation area of the country, however, has not technologies to increase yield and an intensification program developed rapidly and has remained almost constant at 1.2 for increasing upland rice cultivation area. million hectares. Farmers in these areas cultivate traditional varieties under low-input conditions and obtain low yields, A breeding program averaging 2.3 t ha–1. With this condition, upland rice con- tributes only 5% to national rice production. Despite this The objective of a breeding program for upland rice in low contribution, however, upland rice is very important in Indonesia is to develop improved varieties with high yield providing food and as a source of income for people living potential, good grain quality, early maturity, and resistance in dryland areas. to or tolerance of biotic and abiotic stresses that are present The cultivation area of upland rice is distributed over a in the targeted ecosystems. The breeding program for the wide range of agroclimatological zones—from humid areas arid region, for example, has emphasized drought tolerance with eight or more wet months to arid areas with four or fewer and early maturity; for the humid region, blast resistance and wet months. In the humid area of West Java, Sumatra, and tolerance of problems associated with acid soil such as Al Kalimantan, blast and brown spot diseases, soil acidity, Al toxicity and P deficiency have been the focus; and for inter- toxicity, and P deficiency are the major constraints to high planting, emphasis has been on tolerance of shading. yield. In contrast, in the arid areas of Central and East Java, All of the programs are conducted continuously at the NTB (Western Nusa Tenggara Islands), and NTT (Eastern Indonesian Institute for Rice Research (IIRR); however, Nusa Tenggara Islands) provinces, drought is predominant. the prioritization of objectives has changed over the years. Breeding programs for upland rice to develop improved Resistance to blast disease has always had a high priority in varieties with high yield and resistance to or tolerance of the the upland rice breeding program. Very early maturity and corresponding major constraints have been conducted con- tolerance of drought had high priority during 1991-94, toler- tinuously over a long period of time in Indonesia. Improved ance of shading during 1991-96, and tolerance of Al toxicity varieties have been released and several promising breeding from 1991 to 2000. Rice research on suboptimal ecosystems, lines are being further evaluated to select better varieties. including breeding for upland rice, was intensified recently. Cultivation of improved varieties with proper manage- The previous breeding objectives were continued but breed- ment has been demonstrated in some locations to achieve high ing for blast resistance, Al toxicity, drought tolerance, and yields of more than 5 t ha–1. This indicated a high yield gap good to premium grain quality were accorded high prior- between research and farm levels for upland rice. Increasing ity.
98 Suwarno et al Table 1. Genetic sources for some desirable characteristics used in the breeding program for upland rice in Indonesia.
Desirable characteristic Varieties
Drought tolerance Salumpikit, Cabacu, Gajah Mungkur, Kalimutu, ICOXI- B-66, Lagos Blast resistance Carreon, Tetep, Tadukan, Klemas, Jambu, Mat Embun, Cabacu, Cuil, Danautempe, C101LAC, C105 A51, Sayap, Bonti, Jadah, Perak, Raden Intan, Sejang Ungu, Grogol, Dupa Tolerance of low pH and Seratus Malam, Hawara Bunar, Ketombol, Lawean, low pH and Al toxicity Meulaboh, Dupa, Simariti, IRAT352, IRAT379, Gro- gol, Simedan Grain quality Cabacu, Gajah Mungkur, Membramo, CT6510-24, Ketan Tuban, Seratus Malam Aromatic rice Simariti, Dupa, Mesir, Bengawan Solo
Table 2. Rice varieties used as genetic donors for blast resistance Released varieties in the breeding program for upland rice in Indonesia.
Name Type of Name Type of Thirteen improved varieties of upland rice have been of- variety variety ficially released in Indonesia since 1991 (Table 3). All of these varieties had resistance to blast disease and other de- Bonti Traditional Sirendah Traditional Bulan Sabit Traditional Sirendah Pulen Traditional sirable characteristics related to the targeted environments. Cabacu Introduced Tetep Traditional The approach taken in breeding for blast resistance was to Dupa Traditional Asahan Improved pyramid resistance genes. Danau Tempe, which was resistant Grendel Traditional Limboto Improved to six races of blast pathogen, was released in 1991. Varie- Grogol Traditional Wayrarem Improved ties with a wider spectrum of blast resistance were released Jambu Traditional Batu tugi Improved Ketombol Traditional IRBL 8 Monogenic (Pik-h) later, namely, Way Rarem, Jatiluhur, and Batutegi, which Klemas Traditional IRBL 10 Monogenic (Piz-5) were resistant to 10 races of the pathogen. The cooked-rice Lampung Arak Traditional IRBL 19 Monogenic (Pi3) texture of the varieties differs in accordance with the diverse Lampung Kuning Traditional IRBL 23 Monogenic (Pi12(t)) preference of consumers. Lampung Putih Traditional IUF5 1 Monogenic (Pii) Silugonggo was selected from a breeding line in- Malio Traditional IUF5 70 Monogenic (Pi7(t)) Simacan Traditional troduced from IRRI: IR39357-71-1-2-2. This variety had resistance to blast and bacterial leaf blight and had a very early maturity of 90 days. It could be adapted to upland areas mainly in Java, where Al toxicity is not prominent. Some programs to intensify upland rice cultivation by Most of the breeding programs were conducted follow- introducing improved varieties and cultivation technologies ing a modified bulk method in which populations of early have been conducted. The introduction of newly released up- F2–F5 generations were grown with closed spacing. Panicle land rice with drought tolerance and early maturity was done selection was conducted on F5 populations followed by pedi- in arid regions of East Java in 1994. Programs to introduce gree selection and observation yield trials, preliminary yield the improved varieties Way Rarem and Jatiluhur, which were trials, and advanced yield trials. To speed up the breeding tolerant of blast, Al toxicity, and shading, were implemented process, the bulk populations were grown under irrigated in the newly developed area for estate crops and Imperata lowland conditions during the dry season. sp. involving idle lands with a total area of 100,000 ha and Various genetic sources have been used in breeding 215,000 ha during the planting seasons of 1994 and 1995, for important characteristics of upland rice (Table 1). The respectively. The program succeeded and high yields of 3–4 approach of breeding for blast resistance was modified from t ha–1 fresh rice were reported (Directorate of Estate Crop developing varieties with resistance to many blast races or Production 1995). Most of the programs, however, were not pyramided resistance genes to varieties with different resist- sustained and farmers in upland areas still cultivate traditional ance genes, that is, diversification of blast resistance. Differ- varieties with low inputs, leading to low yields. These low ent sources for blast resistance, including those with identified yields might also be caused by a breakdown in disease resist- genes and traditional varieties, were used for crossing (Table ance (Amir and Nasution 1995) or a decrease in soil fertility 2). due to a reduction in soil organic matter content (Adiningsih and Mulyadi 1993, Hidayat et al 2000).
Progress of upland rice breeding in Indonesia since 1991 99 Table 3. Improved varieties of upland rice officially released in Indonesia since 1991.
Variety Year of Maturity Cooked-rice Yield Resistance and release (days) texture (t ha–1) tolerancea
Danautempe 1991 115 Hard 2.5–3.5 Bl, Al Gajah Mungkur 1994 95 Medium 2.5–3.5 Bl, Dr Way Rarem 1994 105 Hard 3.0–4.0 Bl, Al, Fe Kalimutu 1994 95 Medium 2.5–3.5 Bl, Dr Jatiluhur 1994 115 Hard 2.5–4.5 Bl, Sh Cirata 1996 120 Medium 3.0–5.0 Bl, BPH, Sh Limboto 1999 105 Hard 3.0–5.0 Bl, Al, Dr Towuti 1999 120 Soft 3.0–5.0 Bl, BPH, BLB Batutugi 2001 116 Medium 3.0–5.0 Bl, Al, Dr Danau Gaung 2001 113 Medium 3.4–5.0 Bl, BS, Al, Fe Silugonggo 2001 90 Medium 4.5–5.5 Bl, BLB Batang Gadis 2002 122 Medium 3.6–5.6 Bl Situ Bagendit 2002 105 Soft 3.0–5.0 Bl, BLB
aBl = blast, BS = brown spot, BPH = brown planthopper, Dr = drought, Al = aluminum toxicity, Fe = iron toxicity, Sh = shading.
Table 4. Average yield of improved upland rice va- Promising lines and other breeding materials rieties cultivated with integrated crop management technology in Seputih Rahman, Lampung, WS 2003- 04. Breeding for upland rice is conducted continuously and pro- duces promising lines and other materials at different stages in Yield (t ha–1) the breeding pipeline. All of the promising lines tested in yield Name of trials at Tamanbogo and Sukadana, Lampung (Table 5), were farmer Batutegi Limboto Situ Average Patenggang resistant to blast, with some of them containing the two resist- ance genes, Pi-1 and Pi-2, and others also having tolerance of Wira 6.24 5.81 5.71 5.92 Al toxicity. Eight promising lines had high yield above 4.0 t Era 6.37 5.25 5.67 5.76 ha–1, which was comparable with that of the improved check Eri 6.20 5.72 5.29 5.74 variety Limboto and higher than the yield of traditional va- Pt. Sukasi 5.98 5.54 6.02 5.85 Ayu 6.29 5.93 4.47 5.56 riety Sirendah. Participatory varietal selection that began in Oka 6.31 5.95 5.20 5.82 the wet season of 2001-02, in which farmers were allowed to Budiarte 6.25 6.52 6.33 6.37 select their preferred promising lines (Suwarno et al 2002), Agus P 5.96 7.02 5.52 6.17 was also applied to the promising lines. Average 6.20 5.97 5.52 5.90 Most of the released varieties have cooked-rice texture of medium to hard. The preferences of rice consumers in Indonesia differ across regions. In Java and Bali, consumers Technologies for rice cultivation as a component of prefer soft rice, while in Sumatera and Kalimantan they prefer a farming system aimed at sustaining or improving soil hard rice. Many farmers living in Sumatera and Kalimantan productivity are available (Toha and Fagi 1995). Upland came from Java and Bali and thus prefer soft rice. Breeding rice cultivation based on integrated crop management with lines with good phenotypic performance and amylose content improved varieties has also been tested in farmers’ fields, associated with soft to medium rice texture were selected obtaining high yields of 3.88–4.74 t ha–1 in the wet season (Table 6). These lines were resistant to both leaf and neck (WS) of 2002-03, which increased to 5.52–6.20 t ha–1 in the blast and tolerant of Al toxicity. The lines had medium grain WS of 2003-04 (Table 4) (Toha, H.M., personal communica- shape, medium to long grain length, and small to medium tion). chalkiness. The most popular variety in Indonesia, IR64, Blast resistance of varieties broke down after they were had long grain with medium shape and medium chalkiness. cultivated in a wide area for several consecutive years (Amir The premium quality rice Cianjur, on the other hand, had and Nasution 1995). The development of improved varieties intermediate length, shape, and chalkiness. with different blast resistance genes could be an appropriate To develop improved varieties with different blast approach for a breeding program on upland rice. Thus, such resistance, the existing breeding lines were screened by a program was started and some selected breeding lines have inoculating them with combinations of two blast races. A been obtained. total of 479 breeding lines derived from different crosses
100 Suwarno et al Table 5. Promising lines and check varieties of upland rice tested in yield trials in Tamanbogo and Sukadana, Lampung, during WS were obtained (Table 7). More systematic crosses involving 2004-05. monogenic lines to develop improved varieties, each contain- ing two resistance genes, were undertaken. Line/variety Maturity Yield Resistance/tolerance trait (days) (t ha–1) Exchange of breeding materials Bio528B-TB-12-1-1 102 4.76 R to blast; Pi-1 and Pi-2 genes The exchange of breeding materials was conducted mainly Bio511B-61-2-3-1 110 4.07 R to blast; Al tox.; drought through activities or networks coordinated by IRRI, includ- Bio511B-5-12-5-1 111 3.63 R to blast; Al tox. ing the International Network for Genetic Evaluation of Rice B9071F-B-7 103 4.64 R to blast; Al tox. (INGER), the Asian Rice Biotechnology Network (ARBN), Bio511B-61-2-4-1 109 2.98 R to blast; Al tox. the Upland Rice Research Consortium (URRC), and the Con- Bio512-MR-1-4-PN-26 109 3.33 R to blast; Al tox. Bio530B-39-3-6 108 3.08 R to blast; Pi-1 and Pi-2 sortium for Unfavorable Rice Environments (CURE). Many genes varieties and breeding lines exchanged were evaluated and Bio530B-5-6-5-4 109 4.51 R to blast; Pi-1 and Pi-2 used in Indonesia. For example, improved variety Silugonggo genes was selected from a breeding material obtained from IRRI. Bio530A-5-14-2-2-8 107 4.01 R to blast; Pi-1 and Pi-2 Many other varieties and lines received from IRRI were also genes TB360B-TB-26-1 103 3.86 R to blast; Al tox. used as genetic donors in the breeding program for upland TB393B-TB-17-1 108 4.73 R to blast; Al tox. rice in Indonesia. Some of these were Cabacu, ICOXI-B-66, TB396B-TB-14 111 4.40 R to blast; Al tox. and Lagos for drought tolerance; IR60080-23, IRAT352, and TB437B-TB-1 109 2.76 R to blast; Al tox. IRAT379 for Al-toxicity tolerance; and Cabacu and CT6510- TB356B-TB-18-3 103 4.30 R to blast; Al tox. 24 for grain quality. Limboto 105 4.42 Sirendah 116 3.33
Table 6. Selected breeding lines resistant to leaf and neck blast and tolerant of Al toxicity en- tered in preliminary yield trials during WS 2005.
Plant Maturity Resistancea Grain characterb Line height (cm) (days) LB NB Al Amylose Length Shape Chalkiness (%)
TB364C-TB-12-2 121.4 100 R R R 21.4 M M M TB393C-TB-2-2 109.0 102 R R R 21.4 M M M IR55423-01 103 104 R R R 16.8 M M M TB356B-TB-20-3-1 106 98 R R R 21.2 M M M TB356B-TB-47-1-1 127 102 R R R 21.2 M M M TB356B-TB-47-3-2 98 96 R R R 19.2 M M M TB361B-TB-17-2-2 14 100 R R R 21.6 M M S TB361B-TB-30-6-2 109 100 R R R 24.1 M M M TB366B-TB-2-1-3 124 104 R R R 21.3 M M M TB425B-TB-6-4-3 119 101 R R R 23.0 M M M BP1966B-12-5-TB-1 110 102 R R R 21.5 M M S BP1978B-24-5-TB-1 114 103 R R R 21.5 L M S TB59-TB-16-1-1 111 103 R R R 24.9 L M S BP241D-TB-18-6-1 110 97 R R R 20.1 M M M BP1351D-1-2-PK-3 102 102 R R R 20.1 M M M TB356B-TB-20-4 101 100 R R R 18.2 M M M TB356B-TB-47-3 94 104 R R R 19.2 M M M TB356B-TB-52-2 103 104 R R R 20.7 M M M TB360B-TB-26-1 93 102 R R R 20.7 M M M TB396B-TB-14 94 110 R R R 19.1 M M M BP1976B-23-7-TB-1 100 102 R R R 20.8 M M M TB437B-TB-5 80 106 R R R 20.1 M M M TB457B-TB-2 133 102 R R R 17.8 M M S IR64 L M M Cianjur rice M M M
aLB = leaf blight, NB = neck blast, Al = aluminum. bM = medium, L = long, S = small.
Progress of upland rice breeding in Indonesia since 1991 101 Tabel 7. Number of F6 selected breeding lines resistant to the respective combination of P. grisea races and their major parental varieties, WS 2003-04.
Race combination Number of Major parental varietiesa selected lines
123 and 173 182 TB154E-TB-2/IRAT144//IRAT379 (18); IR60080-23/BP303 (17); Memberamo/TB154E-TB-2 (17); G.M./Cabacu//B. Sabit/Memberamo (16) 001 and 031 36 Asahan/Dupa//GM/Cabacu (6); G.M./Cabacu//B. Sabit/ Mem- beramo (5); Bonti/G.M. (4) 021 and 033 120 G.M./Cabacu//B. Sabit/Memberamo (18) Memberamo/ IR60080-23 (16); TB154E-TB-2/ IRAT144//IRAT379 (12) 023 and 100 141 Bonti/Malio//G.Mungkur/IRAT144 (16 galur) Memberamo/ TB154E-TB-2 (15) B8503E/IAC25 (14); Asahan/Dupa// GM/Cabacu (14)
aNumber in parentheses indicates the number of lines.
Monogenic lines for blast resistance obtained from more expensive than improved varieties. The incorporation IRRI were also evaluated for their reaction to different blast of good grain quality should be included in the development races (Table 8). These monogenic lines were useful not only of improved upland rice varieties for both humid and arid as genetic donors in the upland rice breeding program but regions. also in the characterization of blast pathogen populations in Moreover, farmers’ preferences could be different Indonesia. not only in terms of grain quality but also in terms of the performance of an improved variety itself. To meet farm- A future breeding program ers’ preferences, participatory varietal selection that already started will be continued. The constraints of blast disease and acid soil–related prob- lems are still prominent in the existing and potential areas for References upland rice cultivation in humid regions. Thus, the breeding program for upland rice in Indonesia has put great emphasis Adiningsih SJ, Mulyadi. 1993. Alternatif teknik rehabilitasi dan pe- on blast resistance and Al-toxicity tolerance. manfaatan lahan alabg-alang. Prosiding Pemanfaatan Lahan The blast pathogen has many races that are capable of Alang-Alang untuk Usahatani Berkelanjutan. Pusat Penelitian breaking down the resistance of improved rice varieties after Tanah dan Agroklimat, Bogor. Hal. 29-49. cultivating them for several consecutive planting seasons. Amir M, Nasution A. 1995. Status dan pengendalian blas di Indonesia. Diversification of blast resistance through the development of Hal. 583-592. In: Dalam Syam M et al, editors. Kinerja Penelitian improved varieties containing two resistance genes originat- Tanaman Pangan. Buku Puslitbangtan. Badan Penelitian dan ing from selected monogenic lines and varieties with blast Pengembangan Pertanian. resistance originating from different traditional varieties will Amir M, Nasution A. 1997. Pemanfaatan galur-galur monogenik padi be continued. (Oryza sativa L.) untuk pengendalian penyalit blas (Pyricularia In the arid region, blast disease is not severe but drought grisea). In: Proceedings of the Indonesian Phytopathological So- and the short duration of the rainy season are major problems. ciety Congress and National Seminar, Palembang, 27-29 October To provide better and improved varieties for this region, 1997. Sriwijaya University, Palembang. p 171-177 the breeding program for drought tolerance and very early Hidayat A, Hikmatullah, Santoso D. 2000. Potensi dan pengelolaan la- maturity will be intensified. han kering dataran rendah. In: Adimihardja, editor. Sumberdaya One of the reasons that high-yielding and improved Lahan Indonesia dan Pengelolaanya. Pusat Penelitian Tanah dan varieties could not be adopted by farmers in upland areas is Agroklimat, Badan Litbang Pertanian. Hal. 197-225. their grain quality, which does not satisfy farmers’ require- ments. Traditional varieties cultivated by farmers have good grain quality, are preferred by consumers, and are frequently
102 Suwarno et al Table 8. Reaction of monogenic lines to 20 races of the blast pathogen.
P. grisea racea Line 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
1. C101 A51 (p-1) R R R R R R R R R R R R R R R R R R R R 2. C102 A51 (p-2) R S R R R R R R R R R R R R S R R R S S 3. C103 A51 (p-3) S S R R R R S R R R R R R R R R R R R R 4. C104 A51 (p-4) S S R R S R R R R R R R R R R R R R R R 5. C105 A51 (p-5) R R R R S R R R R R R R R R S R R R S R 6. C 101 LAC (p-6) R R S R R R R R R R R R R R R R R R R R 7. C104 LAC (p-6) R R S R R R R R R R R R R R R S R R R S 8. C101 PKR (p-8) R R R R R R R R S R R R R R R S S S S S 9. C102 PKR (p-9) R R R R S S R R R R R R R R R R S S S S 10. C104 PKR (p-10) R S S R S S R S R R R R S S R R R S S S 11. C. 101 RRP-1 (p-11) R S S S R R R R S R S R S S S S S S S S 12. C101 RRP-2 (p-12) R S S R R S R R R R R R S R R S S S R S 13. C101 RRP-3 (p-13) S S S R S S R S S R S R S S S S S S S S 14. C101 RRP-4 (p-14) S S S R S R R R S R R R R S S S S S S S 15. C101 RRP-6 (p-15) S S S R S R R S R R R R R R R S R R S S 16. C101 RRP-6 (p-16) R S S R R R R R R R R R S R S S S S R S 17. C103 RRP (p-17) R R R R R R R R R R R R R R S R R R R R 18. C105 RRP -1 (p-18) S R S R S R R R S R R R R R R S S S S S 19. C105 RRP-1 (-19L9) R R S R S R R R S R R R R R R S S S S S 20. C105 RRP-2 (p-19L23) S S S S S S R S R R R R R S R S R R S S 21. C105 RRP-4 (p-20L6) R R S R R R R S R R R R R R R S S S S R 22. C101 RRP-4 (p-20L6) S S S S S R S S R R S R R R R S S S S S 23. Asahan R R R R R R R R R R R R R R R R R R R S 24. Kencana Bali S S S S S S S S S S S S S S S S S S S S
a1. ID-14 (15) Lampung, 2. IG-2 (24) Sukabumi, 3. IG-1 (260) Bandung, 4. IC-15 (39) U. Pandang, 5. ID-13 (60) Cianjur, 6. 001 Sukabumi, 7. 003 Sukabumi, 8. 013 Sitiung, 9. 021 Sukabumi, 10. 023 Sukabumi, 11. 031 Sitiung, 12. 033 Sitiung, 13. 100 Kalimantan, 14. 103 Kali- mantan, 15. 121 Kalimantan, 16. 123 Sukabumi, 17. 133 Kalimantan, 18. 161 Sukabumi, 19. 173 Kalimantan, 20. Sitiung. Source: Amir and Nasution (1997).
Suwarno, Kustianto B, Arjasa WS, Atlin GN. 2002. Participatory selection on upland rice in Sumatera. In: Witcombe JR, Parr LB, Atlin GN, editors. Breeding rainfed rice for drought-prone environments: integrating conventional and participatory plant breeding in South and Southeast Asia. Los Baños (Philippines): International Rice Research Institute. p 61-63. Toha HM, Fagi AM. 1995. Budidaya tanaman pangan dan sistem usahatani konservasi di DAS Jratunseluna bagian hulu. Kinerja Penelitian Tanaman Pangan. Badan Litbang Pertanian. Buku 3:810-824.
Notes
Authors’ address: Indonesian Center for Rice Research (ICRR), Suka- mandi, West Java, Indonesia.
Progress of upland rice breeding in Indonesia since 1991 103 Breeding rice for the Indian plateau uplands
P.K. Sinha, M. Variar, and N.P. Mandal
Rice area in India has fluctuated between 40.2 million ha trient management, and blast. Subsequently, in the second (2002-03) and 45.1 million ha (an all-time high in 1999-2000) and third phases of the consortium, the scope was widened since 1994 (DES 2004), while production varied between to incorporate research lessons from other consortium sites 72.65 million t in 2002-03 and 93.34 million t in 2001-02. to develop drought-tolerant breeding lines as well as crop Variation in rice production was closely related to total annual resource management techniques. rainfall. Nearly half of the total rice area is rainfed, located mostly in the eastern Indian states of Assam, Chhattisgarh, Drought occurrence Jharkhand, Madhya Pradesh, and Orissa; altogether, they have less than 25% of the area under irrigation. Rice produc- Spatial heterogeneity and seasonal variation in rainfall pat- tion dropped by more than 20 million t in the drought year terns profoundly affect the growth of upland rice, especially of 2002-03, when deficiency in rainfall was about 19% of in the plateau regions of Jharkhand, Orissa, and Chhattisgarh. the normal. Of this, about 14 million t were lost in the pre- One of the activities in the first phase of the collaborative dominantly rainfed eastern Indian states. Uplands constitute program was, therefore, site characterization, especially about 15% of the rice grown under rainfed conditions and quantification of the magnitude of drought at various phe- the rest is lowland (30%) and flood-prone (7%). Rice yields nological phases of upland rice at Hazaribag. The monsoon in the rainfed ecosystems are low and fluctuate greatly from months of June-October (22–43 metro-wk) were divided year to year. Yield in regions with irrigation is 3.2 t ha–1; the into initial (7 wk), intermediate (8 wk), and terminal (7 wk) range is from 0.6 to 1.5 t ha–1 in uplands, 0.9 to 2.4 t ha–1 in phases and the frequency of occurrence of drought (number rainfed lowlands, and 0.9 to 2.0 t ha–1 in deepwater areas. of weeks in which no rainfall was recorded) in each phase was Average rice yields in eastern India vary from 0.8 to 1.96 t analyzed covering 1972 to 1991. The probability of drought ha–1 (Shobha Rani et al 2002). spells of 2−4-wk duration in the initial phase and of 3−6-wk Upland rice is grown on flat lands in coastal Orissa, duration in the terminal phase was high. Drought spells of Assam, and eastern Uttar Pradesh; on gently rolling lands 7-wk duration occurred rarely (5% probability) in the initial (up to 8% slope) in Chhattisgarh and Madhya Pradesh; and or terminal phase. on sloping lands (>30% slope) in Jharkhand, western Orissa, Drought spells were infrequent in the intermediate Meghalaya, and Uttaranchal hills. Around 70% of the upland phase, with low variability, whereas drought spells coincid- rice area is drought-prone, with the other 30% in coastal ing with germination and establishment of the rice crop and Orissa, Himachal Pradesh, Uttaranchal, and Bihar having at the reproductive stage were frequent and highly variable. access to irrigation. In northeastern India, around 550,000 At 50% probability level, in this agro-climatic region, 14 d of ha are under shifting cultivation (Singh 2002). Intermittent maximum drought spell were generally expected during the drought spells, weed problems, declining soil fertility and latter part of the crop-growing period. Though the number of productivity, and inability of farmers to invest in inputs are rainless days or the percentage of below mean rainfall described the major constraints to productivity in the upland ecosys- only part of the complex phenomenon of water balance, it was tem. These technical constraints are important not only in nevertheless an important criterion in the development of a India but also in other upland rice-growing regions of South crop management strategy. and Southeast Asia. The commonality of constraints across regions, in rainfed uplands as well as in lowlands, was Drought-tolerant genotypes the raison d’être for the establishment of the upland and lowland rice research consortia (URRC and LRRC) by the Assessment of genetic variability for drought tolerance in Philippines-based International Rice Research Institute. The upland rice and the identification of traits associated with Central Rainfed Upland Rice Research Station in Hazaribag drought tolerance were taken up initially with selected aus, is the lead center for consortium activities on upland rice indica, and japonica genotypes with known drought toler- in India, with drought as the major mandate for research. ance and later with mapping populations for marker-aided Other consortium partners focused on weed problems, nu- selection of quantitative traits associated with drought tol-
104 Sinha et al Table 1. Drought-resistant breeding lines of intermediate stature.
Designation Plant Duration Grain Drought height (cm) (d) typea reactionb
IR58662-05 97 120 MS R B2997c-TB-60-3-(ACI-2)-C 81 116 LS MR IR57918(ACI-6)-C 85 112 LS MR IR57920(AC25-2)-C 87 114 LS MR CT6510-24-1-2-19 73 114 LB MR IR60080-32 87 112 LB R IR55435-05 70 112 LB MR IR55419-04 86 108 LB MR Kalinga III (high-yielding check) 108 92 LS S Brown Gora (traditional check) 118 95 LB MR
aLB = long, bold, LS = long, slender, MS = medium, slender. bR = resistant, MR = moderately resistant, S = susceptible.
erance. A repeatable drought condition in a defined growth Sathi 34-36) exceeding 1.5 mm thickness, 5 cm below the stage was difficult to achieve in outdoor screening nurseries first node. during the crop season. Delayed sowing to coincide with The ability of tissues to withstand dehydration (dehy- terminal drought was compounded by other stresses—for dration tolerance) was measured in several genotypes by ex- example, leaf and neck blast interfered with the evaluation posing detached leaves to desiccation. Leaf water content in of leaf drying under drought stress. The success of selecting excised leaves, measured as the difference in weight between tolerant genotypes, therefore, depended on the incidence and turgid and rolled leaves, was also correlated with drought intensity of drought during the season. Among the selection scores of the genotypes. Japonica and drought-tolerant indica parameters, delayed rolling/absence of rolling and wilting and genotypes maintained high leaf water content up to 145 min drying under prolonged drought were the most discernible of excision. The rate of water loss decreased after comple- features easily scored in the field. Several upland indica geno- tion of leaf rolling. A low rate of percentage water loss was types (CR143-2-2, RR18-3) did not wilt and had low grain recorded in moderately drought-tolerant genotype Azucena sterility under drought, comparable with known drought- and drought-tolerant genotypes Annada and IRAT112. An- tolerant tropical japonicas (IAC25, IAC1131, Kinandang nada (indica genotype) transpired small amounts of water, Patong, Salumpikit). Several breeding lines from IRRI also whereas japonica varieties Azucena, IRAT112, and IRAT216 had good tolerance for drought and had the same yields as transpired a greater quantity of water during the period of local checks, but they were longer in duration (Table 1). drought stress. Courtois and Lafitte (1999) compared earlier reports of root system development and osmotic adjustment in Drought-tolerance traits a broad sample of varieties and concluded that japonica types generally have a deep, thick root system and good penetra- Root parameters obviously were the most difficult to measure tion but poor dehydration tolerance and osmotic adjustment. in the field. Selected upland rice genotypes were grown in Indica types have shallow, thin root systems but they show polyethylene tubes (14 cm diameter, 1 m length) filled with better dehydration tolerance and osmotic adjustment. Even compacted field soil and placed in similarly sized holes. though the evaluation at Hazaribag was limited to a few Two sets were maintained: drought was induced in one set genotypes, the results confirmed the findings of Courtois by withholding watering 35 d after sowing; the other set was and Lafitte earlier. Since dehydration tolerance is readily watered regularly. Root length was measured after the stress measured in the laboratory, mass screening of genotypes is period by washing off the soil from the plants. Root and shoot feasible. volume were also recorded in the stressed and nonstressed plants. In general, drought stimulated root growth in tolerant Weed competitiveness genotypes; Vandana recorded a 15% increase in root length under stress, whereas susceptible check Co 13 and traditional Weeds are next only to drought as a limiting factor in upland upland drought-tolerant cultivar Brown Gora had less root rice production. Consortium activities therefore emphasized length under stress. Root volume was highest in IRAT112 the identification of weed-competitive genotypes and traits (not stressed) and Salumpikit (stressed), mostly distributed in associated with competitiveness. Aus genotypes were more the upper 30-cm segment. Root thickness in 60-d-old plants, weed-competitive (Brown Gora, Black Gora, Birsa dhan measured with a magnifier, also differed significantly with 102, Saita, Kalakeri, Sathi 34-36, VHC1253, RR174-1) and some drought-tolerant genotypes (OS6, VHC1253, IRAT112, tropical japonicas were the least competitive when they were screened under natural populations of weeds in unweeded
Breeding rice for the Indian plateau uplands 105 upland plots. A few indica genotypes also showed weed com- indirect screening technique to evaluate genotypes for their petitiveness (Salumpikit, Vandana). Experiments conducted ability to translocate carbohydrates from the stem to the in 1991-92 also revealed that, besides Brown Gora, improved grains was examined. Chemical and mechanical defoliation genotypes Vandana and RR51-1 were also weed-competitive at 5 d after anthesis in 15 selected upland rice genotypes in terms of early emergence of plumule and initial higher leaf revealed highly significant differences among the genotypes area index, tiller number, number of roots per plant, and root in terms of reduced grain number, test weight, and yield. In depth. However, no genotype had a weed-suppressing effect general, aus genotypes were more efficient in translocating in terms of a reduction in weed biomass. The traits contribut- stem-reserve carbohydrates to grain in upland genotypes ing to weed competitiveness could not be identified under than indica or japonica ones. The reduction in grain yield unweeded conditions, but, under once-weeded conditions, (Fig. 1A) and thousand-grain weight (Fig. 1B), as a result weed-competitive genotypes had early seedling vigor, greater of defoliation treatment, was lowest in Kalakeri and Sathi plant height, and higher biomass. 34-36, respectively. The magnitude of the percentage reduc- Differences in weed biomass at different crop growth tion in yield due to defoliation was greater for yield than for stages in competitive (Brown Gora) and noncompetitive thousand-grain weight. This may indicate that defoliation (Heera) cultivars showed that the former had better weed- treatment causes complete sterility of the florets of late- suppressing ability at all stages. The slope of the weed sup- formed tillers or induced floret abortion in the higher-order pression curve, in the case of the weed-competitive cultivar florets. The highly significant genotype × treatment interac- at 35, 50, and 65 DAS, was significantly higher when seed tion indicates that this technique can be used to discriminate density increased from 100 to 300 seeds m–2. There was a genotypes based on their response to post-anthesis drought. corresponding increase in rice biomass at these seed rates, Both techniques appeared effective in revealing genetic with a steep slope. Genotypes with similar weed suppression variation in post-anthesis stress tolerance. Refinement and curves at normal seed rates would be expected to have good precision in the methodology were essential in the case of rice competitiveness. In the case of the noncompetitive cul- chemical defoliation. Mechanical defoliation was, on the tivar, however, the total weed biomass was higher than that other hand, labor-intensive and less likely to be effective for of rice biomass at all seed densities of rice. An analysis of mass-screening populations. traits contributing to rice biomass indicated that, in the case of the weed-competitive cultivar, a maximum reduction in Marker-aided selection for drought and blast plant height occurred at normal seed rates, whereas, at higher seed rates (400–500 seeds m–2), plant height increased, espe- Work with two crosses (IR64/Azucena and Co 39/Morob- cially at the later stages of crop growth. Plant height generally erekan) developed at IRRI began at Hazaribag (Table 2) in increased in the case of the noncompetitive cultivar at higher 1995. This site had natural drought occurrence and its soil seed densities. The reduction in tiller number was highest at type differed from that in the Philippines. Drought scores, leaf lower plant densities and minimum at 500 seeds m–2. Changes rolling, and root thickness were recorded in these populations. in leaf number and area also showed a similar trend. Eight markers were identified for root thickness, measured at 30, 40, and 60 DAS. Regression on flanking markers Techniques for mass screening resolved six segments, two on chromosome 4 and one each for rooting depth under drought on chromosomes 1, 3, 6, and 12. Four of these segments overlapped with those identified for drought scores and leaf Reliable field-screening techniques for mass screening of rolling. Alleles coming from the best parent (Moroberekan) populations for drought are yet to be developed. A herbicide exerted a positive effect in all cases. Champoux et al (1995) injection technique to measure rooting depth was evaluated. reported 26 markers associated with root morphology in Co Here, herbicides were injected at specific depths (50, 75, and 39/Moroberekan in greenhouse experiments at IRRI. At 30, 100 cm) and symptoms of injury on the canopy were scored, 40, and/or 60 DAS, most of the markers associated with root an indication that roots have reached those depths. Based thickness were commonly identified in our investigation. on symptom appearance at 50 and 75 cm, Moroberekan, QTLs associated with root thickness were located on the same Vandana, CR143-2-2, Sathi 34-36, Saita, Black Gora, and chromosomal region as leaf rolling and drought avoidance, Kalakeri were identified as deep-rooted genotypes. Minor suggesting a clustering of different QTLs. modifications in the methodology were attempted to improve Major genes and QTLs associated with blast resistance, repeatability, but the technique was not standardized and did effective under rainfed upland conditions at Hazaribag, were not appear feasible for screening large populations. first detected in Co 39/Moroberekan recombinant inbred Depletion of water supply in the reproductive phase lines (Variar et al 2002). The stability of these QTLs was due to drought spells can reduce reproductive growth and studied in multilocation testing in eastern India. Further, a cause severe yield reduction. It is reported that, under such Vandana/Moroberekan advanced backcross population was situations, transfer of pre-anthesis assimilates to the grain developed to improve the blast tolerance of Vandana and increases two- to threefold and that variation for this character identify QTLs associated with blast resistance. This population might exist in a broadly based gene pool (Turner 1982). An was phenotyped in several locations in India and also at IRRI,
106 Sinha et al Table 2. Mapping populations studied at CRURRS, Hazaribag, for drought- and blast-toler- ance QTLs.
Designation Population Characters
IR64/Azucena Doubled-haploid lines Components of drought tolerance Co 39/Moroberekan Recombinant inbred lines Components of drought tolerance and blast resistance Azucena/Bala Recombinant inbred lines Blast resistance
Vandana/Moroberekan Advanced backcross (BC2F4 Blast resistance, identification of superior and BC3F4) transgressed lines with breeding value. Components of drought tolerance
Table 3. Donors that confer resistance to abiotic and biotic stresses (selected from breeding network, 1999-2000).
Duration Plant Drought. Brown Blast score Variety Origin (d) height score spot score (0−9) (cm) (0−9) (0−9)
CT10006-7-2-M-5-1P-3-M Colombia 82 68.8 6 4 0 CT11891-3-3-3-M Colombia 72 65.4 4 2 0 CT13370-2-1-M Colombia 82 72.8 6 2 0 CT13366-8-2-M Colombia 74 64.8 4 6 0 CT13370-8-M Colombia 87 70.0 4 3 0 SMGC89001-6 Thailand 88 89.6 5 5 3 TRI 8409178 Thailand 89 95.7 6 4 3 IRAT144 Côte d’Ivoire 76 74.2 3 5 0 PCT-4\AV \0\0>IR11-1-1 Philippines 74 79.3 3 7 0 PCT-4\AV \0\0>IR4-1-1 Philippines 73 74.9 2.5 6 2 RR166-645 India 81 79.3 6 3 0 RR20-5 India 75 50.8 3.5 2 4 RR36-141 India 68 83.2 2.5 3 4 Vandana India 73 77.5 3.5 5 6
Philippines, for drought and blast. Field-testing of BC3F4 lines to this ecosystem. Breeding priorities for subecosystems are identified several lines with partial resistance to blast and good therefore different and germplasm exchange is favored only agronomic acceptability. Some of the lines showing partial for the identification of specific traits and their introgres- resistance and carrying different introgression region genes sion into locally adapted cultivars at each site. Tolerance for were intercrossed to accumulate the introgression region in drought and blast are traits that are needed in every subsystem a common background. Multilocation testing of the derived and evaluation of a common set of genotypes at different lines in blast-endemic locations in India and assessment of consortium sites led to the identification of several donors drought and grain quality characteristics at IRRI in 2004 led with the desired characteristics. Table 3 lists the genotypes to the selection of several agronomically acceptable lines with identified for improved drought and blast resistance. Since blast resistance. the preferred breeding strategy was to introgress these traits into locally adapted traditional or yield-improved cultivars, Germplasm exchange new crosses were attempted to exploit the drought and blast tolerance traits of these donors. Breeding in the target zones The upland ecosystem is structured into three zones climati- with farmer participatory selection was emphasized for these cally—subhumid mainland Southeast Asia, dry plateaus of populations to improve the adoption rate of high-yielding eastern India and Bangladesh, and equatorial humid areas of genotypes among farmers in eastern India. Indonesia, South Vietnam, and southern Philippines. Upland rice varieties grown in the semihumid subecosystem are Varieties developed and released generally traditional tropical japonicas (panicle-weight type, glutinous rice), whereas, in the eastern Indian plateaus, tra- The development of varieties using drought-tolerant aus ditional varieties are aus and the few high-yielding types are and high-yielding indica genotypes was simultaneouly un- indica (short-duration, panicle-number type). In the equato- dertaken at CRURRS, Hazaribag. Vandana, which is highly rial humid subecosystem, farmers grow indicas and also some drought-tolerant and adaptable to upland conditions, was improved japonicas, but tropical japonicas are not adapted developed from one such cross (C22/Kalakeri). C22 is a
Breeding rice for the Indian plateau uplands 107 Table 4. Drought-tolerant rice varieties bred for Jharkhand.
Variety Parentage Duration Salient features (d)
Vandana C22/Kalakeri 90 Tall, long bold grains, deep root system, moderately resistant to leaf blast and brown spot, suit- able for tanr land and don 3 Anjali Sneha/RR149-1149 95 Semitall, short bold grains, broad erect leaves, resistant to brown spot and gall midge biotype 5, moderately resistant to leaf blast and sheath rot, suitable for tanr land and don 3 Birsadhan 101 Fine Gora/IET 2832 85 Semidwarf, long bold grains, resistant to blast, suitable for intercropping with pigeon pea and soybean Birsa Gora 102 Selection from Brown Gora 100 Tall, long bold grains, red kernel, suitable for tanr land Birsadhan 105 Fine Gora/IET 2832 90 Semidwarf, short bold grains, resistant to blast, brown spot, bacterial leaf blight, and gall midge Birsadhan 106 Bala/Black Gora//OS36/CH1039 95 Semidwarf, short bold grains, resistant to blast, brown spot, bacterial leaf blight, sheath blight, and gall midge Birsadhan 107 Gora mutant/ IAC25 95 Semidwarf, compact and long panicles, short bold grains, resistant to blast and bacterial leaf blight, moderately resistant to brown spot Sadabahar BRRI SAIL/IR10181-58-3-1 105 Semitall (90–105 cm) variety, stiff straw, very good early and late vigor, long bold grains, suitable for don 3 and 2 Hazaridhan IR42/IR5853-118-5 120 Semidwarf, erect leaves, stiff straw, long slender grains, white kernel, good eating quality, blast resistant, suitable for don 2 semitall, medium-duration variety from the Philippines and tall and matures in 95 d when direct-seeded. Anjali is resist- Kalakeri is a tall, drought-tolerant, traditional cultivar from ant to brown spot and gall midge biotype 5 and moderately Orissa. Vandana was released by the Bihar State Variety resistant to leaf blast and sheath rot. RR345-2 was another Release Committee in 1992 for the plateau region (Sinha et highly promising, high-yielding, drought-tolerant breeding al 1994) and later for uplands of Orissa in 2002. Vandana is line developed at this center, identified for release in the All tall, matures in 95 d, and has moderately acceptable grain India Coordinated Varietal Trials. Several other breeding lines quality. It has a very good (deep) root system and is highly developed at Hazaribag—RR18-3, RR165-1160, RR348-1, tolerant of drought. Vandana is moderately resistant to leaf RR348-6, and RR354-1—were found to have good drought blast and brown spot and has moderate tolerance for major tolerance. insect pests. The average grain yield of Vandana is 2.5−3.0 A sound breeding program is under way to develop t ha–1, but it can yield up to 5.0 t ha–1 when conditions are drought-tolerant, productive genotypes at BAU, Ranchi. favorable. The breeding program at Hazaribag resulted in the Under this program, besides drought tolerance, early vigor, development of another variety, Anjali, in 2002. Anjali was plant height, and earliness were emphasized during selection. identified through national coordinated trials (DRR) for the These efforts have led to the development of several promis- uplands of Jharkhand, Bihar, Orissa, Assam, and Tripura and ing varieties. A list of drought-tolerant varieties developed is was released by CVRC in 2002. It is moderately drought- given in Table 4. tolerant with thick, broad, erect leaves and high vegetative vigor. Anjali has short bold grains; its eating quality is highly acceptable to farmers (Singh et al 2000). This variety is semi-
108 Sinha et al Participatory research blocks in the same field at the respective sites in the 1999 wet season (WS). The farmers and breeders continued their Though several new varieties have been released for the selection of materials and places and, during the 2000 WS, rainfed uplands, their adoption has not been to the level ex- four sets of final (bulk) selections were made—breeders’ pected. The reasons for the low adoption are a lack of access selection on-station, farmers’ selection on-station, breeders’ to seeds and, more important, the unsuitability of the new selection on-farm, and farmers’ selection on-farm. These four varieties to specific local conditions and farmers’ needs. In sets were pooled together along with three checks (Brown addition, a range of socioeconomic constraints may have also Gora, Kalinga III, and Vandana) and evaluated both on-farm influenced farmers’ ability to adopt the new varieties. and on-station during the 2001 WS using an alpha-lattice de- While the first phase of the consortium focused on sign with two replications. The PPB trial data were analyzed strategic research to tackle problems of common impor- using a mixed model, with selector and selection environ- tance across regions, the second phase emphasized farmer ments taken as fixed variables, and lines within selection participatory approaches. These involved selecting and environment × selector combinations considered random. testing new genotypes (participatory variety selection, The analysis was conducted with the REML algorithm of PVS) and populations (participatory plant breeding) derived SAS PROC MIXED. from parental lines identified in the first and second phases. The coefficient of concordance among farmers was Yielding ability of drought-tolerant genotypes under stress highly significant in all the trials conducted in all 3 years. in drought-prone environments was also evaluated in the This indicated that farmers’ rankings were not randomly second phase to validate the strength of the tolerance traits attributed and there was a good agreement among them. hitherto identified. Three villages in Hazaribag District were The concordance among breeders’ rankings was also high selected for this project on the basis of the representativeness but often not significant because of the small number of of the environments targeted in the breeding work, diversity breeders. There was a good agreement between farmers and range of ecological conditions, involvement of women and breeders in genotype choice. The farmers’ or breeders’ in farm activities, availability of previous survey data, and rankings at maturity were not always highly correlated with easy access to the site. Ten farmers from each village were yield, especially the farmers’ ranking in low-yielding trials. selected for the PVS trials based on their primary occupation This may be because farmers considered criteria other than (rice farming, especially upland rice), landholding, and some grain yield when selecting varieties for their farms. degree of literacy. There was no correlation between farmers’ and breed- ers’ rankings and duration or plant height, with a few excep- Participatory varietal selection tions. The variance component analysis for the on-farm trials PVS trials were conducted for three consecutive years from showed large and significant differences among cultivars in 1997 to 1999, with 15−16 upland rice genotypes (including grain yield under farmer management. The grain yield data local check Brown Gora). The participating farmers of a averaged over the four on-farm sites and 3 years showed that given village and the breeders ranked the genotypes in the one elite line (RR348-5) significantly outyielded Vandana and village and also at the research station at two phenological recorded about 100% more yield than Brown Gora (Table stages (vegetative and reproductive), from “most liked” to 5). “least liked,” on the basis of their selection criteria. In addi- No variety × location or variety × year interaction was tion, breeders also recorded duration, height, yield and yield detected. The three-way interaction, variety × location × year, components, and reaction to diseases and pests in all trials. was large (0.246) but cannot be separated from the within- To compare the ranks given by farmers and breeders, the trial error in this analysis as the trials were unreplicated. This Kendall coefficient of concordance (W) was used (Siegel result does not support the hypothesis that varieties exhibit 1956). Farmers’ and breeders’ rankings were compared fol- specific adaptation to particular subenvironments within the lowing the Spearman rank coefficient of correlation (Courtois target region of the research station. The repeatability or et al 2001). broad-sense heritability of different on-farm and on-station trials was estimated to judge the precision of the trial. It was Participatory plant breeding (PPB) found that the repeatability of the on-station trial was poorer Participatory plant breeding started 1 year later after farm- than that of the on-farm trial. This may be because of the low ers were given some basic training on the objectives and yield of the trials on-station as upland rice is being grown methodology of single-plant/line selection in the segregating continuously. This resulted in poor soil fertility. population. One hundred segregating lines from 12 crosses in The regression of on-farm performance on on-station the F5 generation were grown at one on-farm site (Khorahar) performance was not significant. This meant that on-station and at the research station. Lines were scored by farmers and testing is not able to predict on-farm performance of the geno- breeders at both sites at different growth stages. At maturity, type and there is a need for integration of on-farm testing at single-plant selections were made separately by farmers and an early stage of the breeding program. At the end of 3 years breeders at both sites. The plants selected by farmers and of PVS trials, farmers selected five genotypes—RR354-1, breeders on-station and at Khorahar were grown in separate RR347-166, RR151-3, RR166-645, and RR 51-1—us-
Breeding rice for the Indian plateau uplands 109 Table 5. Mean grain yield (t ha–1) of Table 6. Mean grain yield (t ha–1) of upland cultivars and breed- upland cultivars and breeding lines ing lines evaluated under farmer management in three villages evaluated under farmer manage- and on-station in Hazaribag, Jharkhand, 2001 WS. ment in four villages near Hazaribag, Jharkhand, over three wet seasons. Cultivar Chichi Khorahar Peto Mean over CRURRS villages (on-station) Cultivar Trials Mean over years (no.) and locations RR354-1 1.32 1.47 1.70 1.49 2.67 RR347-1 1.46 1.53 1.30 1.43 2.92 Brown Gora 12 1.12 RR347-166 1.61 1.55 1.10 1.42 3.25 Vandana 11 1.74 RR348-5 0.74 1.65 1.50 1.30 2.78 RR347-166 12 1.92 RR345-2 1.47 1.38 1.00 1.29 2.30 RR348-5 12 2.25 Kalinga III 1.13 1.30 0.80 1.08 1.83 RR151-4 9 1.11 RR361-1 1.03 1.28 0.90 1.07 2.35 RR203-16 9 1.47 CR876-6 1.29 1.20 0.70 1.06 1.95 RR354-1 12 2.12 RR51-1 1.44 0.87 0.80 1.04 2.67 RR51-1 12 1.78 Brown Gora 1.26 1.22 0.60 1.03 1.85 RR151-3 12 1.68 RR151-3 1.09 1.10 0.50 0.90 2.10 RR166-645 12 1.49 Vandana 1.01 0.95 0.70 0.89 2.43 RR50-5 9 1.63 RR166-645 1.24 0.87 0.40 0.84 2.25 RR139-1 12 1.25 RR363-737 0.90 0.92 0.60 0.81 2.38
LSD0.05 for means 0.47 RR139-1 0.55 1.32 0.35 0.74 1.87 over nine trials RR361-783 0.51 0.90 0.33 0.58 1.15 LSD0.05 0.70 0.70 0.70 0.42 0.76
ing their own selection criteria. The highest-yielding line, Table 7. Predicted broad-sense heri- RR348-5, was not selected by the farmers. Farmers did not tability (H) of grain yield for cultivars consider yield as the only important trait. Long slender grains evaluated under farmer manage- ment in three villages near Hazari- (RR166-645), high tillering (RR51-1), tall stature, and good bag, Jharkhand, 2001 WS. cooking quality (RR354-1, RR347-166) appealed to most farmers. These varieties, given to farmers in the following Sites (no.) Replicates (no.) H year (2000), performed better than the local check (Brown Gora) during that drought year, except for RR51-1, which 1 1 0.23 failed as it is susceptible to drought. Drought tolerance was 1 3 0.44 not taken into account by farmers as previous seasons were 3 1 0.50 favorable. 3 3 0.70 Mother-baby trials began during the 2001 WS. The analysis of mother-trial data at three on-farm sites showed large and significant differences among cultivars for grain farmers in explaining the trial failure. In a few cases, drought yield under farmer management (Table 5). Averaged over the damaged the crop at key phenological stages. three sites, RR347-1, RR354-1, and RR 347-166 significantly Farmers concentrated on fewer crosses than breeders. outyielded Vandana (Table 6). These lines also performed They rejected crosses that did not produce plants that met well during 1999-2000. They yielded about 40% more than their criteria. In general, farmers preferred breeding lines with Brown Gora and about 30% more than Kalinga III in 2001 and tall stature, high tillering ability, and long panicles. Plants have consistently outperformed these lines in previous years. generated from certain crosses (VHC1253/Sathi 34-36, N22/ Variance component analysis was conducted to estimate the RR20-5, Annada/RR151-3, and RR139-1/IR57893-08) were precision of on-farm trials through estimation of repeatability preferred by both farmers and breeders. The results indicate or broad-sense heritability (H). The H of grain yield measured that, on average, farmer selections significantly outperformed in on-farm trials was relatively high, as in previous years. breeder selections, and that selection on-station was superior The results showed that evaluation of genotypes at three sites, to selection on-farm (Table 8). Farmer selections yielded, on with three replicates per site, would give adequate precision average, almost twice as much as breeder selections (Table in detecting cultivar differences (Table 7). 9). Of the five highest-yielding lines, four were selected by In baby trials, three varieties, along with a local variety, farmers on the research station (Table 10). These were also were given to each farmer to grow. Farmers’ perceptions of superior to check varieties. the different varieties were recorded. At Chichi, two farmers Farmer agreement on ranking varieties was highly preferred RR354-1 and one farmer each preferred RR347-1, significant, although differences in opinion occurred. The RR363-737, Vandana, and Kalinga III, respectively. In most relatively high agreement contradicted our initial assumptions of the trials, severe weed infestation was the reason given by that farmers’ preferences will vary because of their diverse
110 Sinha et al Table 8. F tests for effects of selection by farmers ver- Table 10. Means and selection history of the checks sus breeders and selection on-farm versus on-station on and five highest-yielding lines. grain yield of selected lines evaluated under farmer man- agement at Khorahar, Jharkhand, 2001. Line Selection history Grain yield (g plot–1) Source F value Pr >F RR356-77 Farmer selected on-station 831 Selector (farmer vs. breeder) 16.6 0.0002 RR356-72 Farmer selected on-station 825 Selection environment (station vs. farm) 20.4 0.0001 RR356-74 Farmer selected on-station 769 Selector × selection environment 5.5 0.0246 RR356-71 Farmer selected on-station 569 RR356-51 Breeder selected on-station 519 Kalinga III 413 Vandana 413 a Table 9. Effect of selection by farm- Brown Gora 256 ers versus breeders and selection on- LSD0.05 412 farm versus on-station on grain yield (g plot–1) of selected lines evaluated under farmer management at Khorahar, Jharkhand, 2001. for the targeted region. But, the precision of on-farm trials Selection Selector environment is no less than that of on-station trials and 3 years of on-sta- Farmer Breeder Mean tion trials failed to predict on-farm performance. This clearly indicates that PVS should be integrated at an early stage of On-farm 243 162 202 testing to better predict varietal performance. On-station 566 264 415 The farmers’ intention of voluntarily testing new varie- Mean 405 213 ties in their own fields indicates that access to new varieties aF tests of main effects of selector and selection or information is a major constraint as the local seed system environment were significant (Pr >F 0.0002 and 0.0001, respectively). is not very effective. Out of five varieties farmers tested during the 2000 WS, two did not perform well and were rejected by farmers. At the early stage, farmers may adopt more varieties, but some will be dropped as testing is done for a number of years. Once a variety is adopted by farmers, socioeconomic backgrounds. This may also be due to the it can be disseminated to the neighboring villages (Joshi et limited number of farmers involved in the project or to the al 2001). low diversity in wealth, caste, and ethnicity of the sample In PPB, in which both farmers and breeders make selec- farmers. The high agreement among breeders indicated tion from segregating populations, farmers selected progenies similarity in selection. The agreement between farmers and from fewer crosses. Farmers immediately rejected crosses breeders was good in most cases. Participation will bring that did not produce progenies according to their set criteria. little improvement when there is close agreement between Because breeders know the background of these crosses, they farmers and breeders. expected desirable segregants in later generations. The degree of agreement is highly influenced by the The significant superiority of farmers’ selection over materials used for selection. For example, in the uplands of breeders’ selection is quite unexpected. It has established eastern India, the variety has to be tall and has to mature that farmers are also able to identify high-yielding entries within 100 d. Less variability within the tested genotypes (Ceccarelli et al. 2001). The better performance of on-station perhaps influenced the farmers’ and breeders’ preferences selection over on-farm selection shows that selections at the for variety or trait. The rankings of farmers and breeders research station can serve the target region. This may be due were correlated with yield in only a few cases, as farm- to the consistently low yield at the research station. ers considered other factors also in their decision. Though The PVS program has given breeders a systematic way duration and stature had little association with yield, these to approach farmers. The interaction with farmers and social are important characters in varietal choice. Again, because scientists involved in the project helped all stakeholders gain of the low variability in the tested material, these were not a better understanding of the complexity of the problem. As identified as important. Therefore, the ranking of varieties the local seed system is not functioning, access to new varie- must be combined with the survey data on farmers’ selection ties or information about the new technology seemed to be criteria to get the actual information (Courtois et al 2001). major constraints. Proper allocation of resources for varietal The results of the 3-year on-farm testing indicated that testing is needed as on-station evaluation alone cannot pre- varieties tended to be ranked similarly across farms. There- dict cultivar performance in the target region. PPB work has fore, genotype × environment interaction is not a significant produced very interesting results. Though it is still early to factor in influencing varietal performance. This indicates that make conclusions, we can surmise that farmers can handle an on-station breeding program can serve specific purposes advanced-stage segregating populations. Farmers were able
Breeding rice for the Indian plateau uplands 111 to select superior progenies from segregating generations Government of India. 1984. Area, production and productivity of rice and it is thus possible to expose breeding materials at the in India. Directorate of economics and statistics. Ministry of advanced segregating stage to farmers. Agriculture and Cooperation, New Delhi, India. Joshi et al. 2001. How narrowly adapted are the products of decentral- The Upland Rice Shuttle Breeding Network ized breeding? The spread of rice varieties from a participatory plant breeding programme in Nepal. Euphytica 122:589-597. To strengthen the exchange of genetic materials among dif- Siegel S. 1956. Non-parametric statistics for the behavirol sciences. ferent research institutions, the Upland Rice Shuttle Breeding New York (USA): MacGraw Hill. 312 p. Network was established under the auspices of the ICAR-IRRI Singh BN. 2002. High yielding rice varieties in India, Rice India collaborative program. This network greatly facilitated the (March), p 5-6. generation of information on varietal performance in a range Singh RK, Prasad K, Mandal NP, Singh RK, Courtois B, Singh VP. of soil and climatic conditions, with better precision. A major 2001. Sensory evaluation of upland rice varieties with farmers: goal was to address location specificity in the deployment an experience in eastern India. In: An exchange of experiences of genotypes by developing and evaluating breeding lines from South and Southeast Asia. Proceedings of the International under naturally occurring stress in the target environment. Symposium on Participatory Plant Breeding and Participatory The breeding lines identified at Hazaribag and the adapted Plant Genetic Resources Enhancement, Pokhara, Nepal, 1-5 germplasm from individual centers were pooled and evalu- May 2000. Cali (Colombia): Centro Internacional de Agricultura ated under the network as preliminary and advanced multi- Tropical. p 319-327. environment trials (2002). In subsequent years (2003, 2004), Sinha PK, Variar M, Singh CV, Prasad K, Singh RK.1994. A new the centers also received a set of fixed lines originating from upland rice variety ‘Vandana’ for Bihar plateau. Indian Farm- IRRI, which were selected along with the center’s own ing 44(3):1-3. breeding lines. Seed increase was done at Hazaribag in the Variar M, Sinha PK, Mandal NP, Maiti D, Shukla VD, Sridhar R, Dash dry season and seeds were redistributed for further selection. AB, Bhat JC, Sengar RBS, Veracruz CM, Atlin G, Courtois B. Promising entries from the advanced multilocation trials were 2002. Linking QTL detection with varietal development. Paper contributed to the PVS trials and to AICRIP testing. presented at National Symposium on Upland Rice Production Syetems, 26-28 September 2002, CRURRS, Hazaribag, India. References Notes Ceccarelli et al. 2001. Farmer participation in barley breeding in Syria, Morocco and Tunisia. Euphytica 122:521-536. Authors’ address: Central Rainfed Upland Rice Research Station, Champoux MC, Wang G, Sarkarung S, Mackill DJ, O’Toole JC, Hazaribag, India. Huang N, McCouch S. 1995. Locating genes associated with root morphology and drought avoidance. Theor. Appl. Genet. 90:969-981. Courtois B, Lafitte R. 1999. Improving rice for drought-prone environments. In: Ito O, O’Toole J, Hardy B, editors. Ge- netic improvement of rice for water-limited environments. Los Baños (Philippines): International Rice Research Institute. p 35-56. Courtois et al. 2001. Comparing farmers’ and breeders’ rankings in varietal selection for low-input environment: a case study of rainfed rice in eastern India. Euphytica 122:537-550.
112 Sinha et al