4.1 Irrigated Rice Ecosystem
4.2 Rainfed Lowland Rice Ecosystem
4.3 Rainfed Upland Rice Ecosystem
4.4 Flood-Prone Rice Ecosystem
4.5 Cross Ecosystems Research Programme
4.6 Overall Assessment and Recommendations
IRRI's Research Programmes are organized by ecosystem (irrigated, rainfed lowland, upland, and flood-prone) plus a cross-ecosystems Programme which includes a diverse set of projects ranging from micro-molecular to macro-policy (see Chapter 7). The five Programmes are reviewed in terms of recent evolution and current focus, achievements and impact, and future strategy, followed by an assessment and any suggestions that have emerged. The chapter concludes with recommendations related to the organizational structure of IRRI's Research Programmes and the Panel's view on research staffing pattern and procedure.
In assessing the Programmes, the Panel defines an impact as something measurable that has favourably affected the livelihoods of people, such as the reduction in the real price of rice. An outcome is defined by the Panel as an effect or a consequence of research, such as the widespread adoption of a variety by farmers. Outputs are measures of scientific productivity (papers published, varieties released, etc.) which may, or may not. lead to a significant impact. In concluding that a Programme has had no impact, the Panel is not suggesting or implying anything about the quality of the science involved. We are simply saying that the Programme has not (yet) had an impact.
This review of Programmes posed difficulties for the Panel because of the near absence (only one) of Centre Commissioned External Reviews (CCERs) and the somewhat ad hoc nature of the documentation for some Research Programmes. The Panel also found difficulties in dealing with the same science appearing in several Ecosystem Programmes.
4.1.1 Recent Evolution and Current Focus
4.1.2 Achievements and Impact
4.1.3 Future Strategy
4.1.4 Assessment
Irrigated rice covers 55% of the world's rice-growing area and provides 75% of the production. IRRI continues to make its greatest contribution to the world's food supply by the production of high-yielding irrigated indica rice germplasm. Pressure on the programme remains high as demand is expected to increase by at least 60% over the next 30 years, a period during which the available rice-growing land is likely to decrease. IRRI's 1998 irrigated rice research programme is planned to cost US$ 8.9M, or 33% of the research budget, although the total relevant investment including the cross-ecosystems components will be much higher.
The 14 projects of the 1994-98 MTP have been consolidated into seven in the 1998-2000 MTP: (i) Breeding to break yield ceilings: A systems approach; (ii) Sustaining soil quality in intensive rice systems; (iii) Improving productivity and sustainability of rice-wheat systems; (iv) Increasing water-use efficiency; (v) Improving pest management; (vi) Coping with global climate change; and (vii) Irrigated Rice Research Consortium.
Over the review period, the annual increase in yield potential due to genetic improvement has remained at around 1% (Figure 4.1) alongside this continuing success by conventional pure-line breeding, which includes an effort to produce a new plant type (NPT), IRRI's F1 hybrid breeding programme has come of age. As always, constraints to progress are apparent, which need to be contained or overcome. Among the former are concerns about declining productivity in intensive systems and potential effects of global warming. Among the latter are improving the efficiency of inputs such as fertilizers and water, as well as the looming dilemma of how to handle intellectual property issues while maintaining a free supply of germplasm to NARS partners.
Figure 4.1 Improvement in Yield Potential of Rice Cultivars, 1965-1995
Rice production must increase to meet growth in demand, yet the land and water resources are diminishing in both quantity and quality (see Chapter 1). Also, as food habits change, there is increasing pressure to diversify the cropping system, making even less land available for rice. Fear is increasing of possible future climatic instability and of potential problems that may arise with the continuing intensification of production on irrigated lands. The availability and management of water for irrigated rice will become even more important as competition for water increases and as environmental concerns grow. A related concern will be improving nutrient use efficiency, perhaps by development of nutrient-efficient varieties. Movement of labour from agricultural to non-agricultural sectors and increases in wages have become important problems for rice farmers in many countries.
Crop intensification may influence pest problems. Crop protection research at IRRI has broadened its scope from only resistance breeding to the improvement of pest management based on an understanding of the ecological and epidemiological background behind pest development and relation of the latter to actual crop losses. The interpretation of the term "Integrated Pest Management (IPM)" in rice evolved during recent years from biological control of one particular pest to a concept of crop protection against pest complexes including animals, diseases, and weeds. It is based on a blend of biological control with cultural methods, healthy seed, varietal resistance, adaptation of the genotype to the environment, and encouragement of beneficial organisms present in the field by judicious use of pesticides with low impact on these antagonists. Nevertheless, emotive perceptions, rather than rational economics, determine the pest management decisions of many farmers and this has led to misuse of pesticides. IRRI has promoted the adoption of IPM concepts by millions of rice farmers in Asia through innovative socio-economic and educational programmes. The IPM Net in eight countries has evolved into an efficient channel for the transfer, through extension services to the farmer, of innovations in IPM technology developed through multi-institutional and multidisciplinary research partnerships. In order to strengthen further the output of this partnership, IRRI has integrated the pest and nutrient management research linked to the IPM Net and the "mega" project (Reversing Trends of Declining Productivity) into the Irrigated Rice Research Consortium.
To meet growing socioeconomic and environmental changes that affect irrigated rice, improved management and new technologies including direct seeding of rice, effective weed control, and increased mechanization in rice farming are required. Weeds constitute a major production threat in direct seeded fields. The Weed Management Peer Review (an internally commissioned external review) in April 1996 expressed concern over the lack of strategic research on rice-weed systems. Strategic research on weed science in irrigated and rainfed lowland rice systems was therefore included in the research agenda to increase knowledge needed in formulating integrated weed management strategies.
Studies of rice fields in the USA and Europe with high organic matter, high N use (200 kg/ha or more), and high yields (7 t/ha) indicated that rice paddies were an important source of methane. However, these estimates did not apply to the conditions of Asia under which much of rice is grown, in soils with lower organic matter, lower fertilizer use (0 to 150 kg/ha), and lower yields. IRRI has initiated research on nutrient and water management to reduce methane emission from rice fields, and on the effects of global warming on rice productivity and rice production systems, in collaboration with NARS and advanced institutions.
Despite a large increase in the global demand and production of rice, the price of rice in real terms has dropped by over 50% since the establishment of IRRI (Figure 4.2). This is the single best indicator of the long-term impact on the welfare of people. While the price decrease reflects an interaction with the price of other cereal grains, it does not reflect a lack of demand, and the contribution of IRRI's technology is clear and indisputable.
The analysis of long-term experiments (LTE) in rice monocropping systems at IRRI and some other experiment stations indicated yield stagnation or decline. Also, a decline in rice yield has been observed in rice-wheat rotations on some experiment stations in South Asia. In recognition of the threat posed by yield decline and decreasing factor productivity in intensively managed rice lands, the Fourth EPMR of IRRI in 1993 recommended a major research effort by IRRI to seek solutions for this complex of problems. To clarify the extent and possible causes for this, IRRI began a project on Reversing Trends of Declining Productivity (RTDP), usually called the "Mega project."
It is important to differentiate between the terms 'yield decline', 'decline in yield growth rate', and 'productivity decline'. Yield decline is defined by IRRI as "A decrease in grain yields per unit area over a period of at least several years". The IRRI definition would include both research stations and farmers' fields.
Decline in yield growth rate is defined by IRRI as "A slowdown in the (percentage) rate of increase of grain yield over time". IRRI gives the example of Indonesia where average national rice yield rose by 3.7% per year during 1967-81, by 1.6% per year during 1982-89, and by 0.5% per year during 1990-96. These figures indicate that Indonesia was experiencing a decline in yield growth rate during this time, but not a decline in yields per se. For example, although yield growth rate was 'only' 0.5% per year during 1990-96, rice yields increased from 4.25 t/ha in 1989 to 4.52 t/ha in 1996.
Productivity is defined as yield divided by the quantity of the factor affected, in appropriate units The IRRI definition of productivity decline is "A decline in total factor productivity (TFP) over time, where TFP is the productivity of all inputs taken together". Productivity decline is not the same as a decline in production or a decline in yields.
With the three definitions stated above, some statements can be made about the current situation. First, it can be stated that there is no evidence for yield decline in farmers' fields in Asia. Secondly, there is evidence that, except for South East Asia and parts of India, most of Asia experienced lower yield growth rates during 1990-96 than during 1982-89, The South East Asian countries where increased yield growth rates occurred during 1990-96 were Vietnam, Myanmar, Cambodia, and Lao PDR. Third, there is little or no evidence of productivity decline in intensively cropped rice-rice-(rice) systems, primarily because of a lack of data. If total factor productivity was declining, this would because for concern. This is the reason why IRRI is collecting such data in its irrigated rice research through the Irrigated Rice Research Consortium. Two other statements relating to the third point can also be made: (i) CIMMYT scientists believe productivity may be declining in wheat in certain parts of Pakistan and (ii) IRRI does have circumstantial evidence that partial or total factor productivity decline is occurring on experimental plots at IRRI, where yields declined over time while inputs remained constant.
Figure 4.2 Trends in World Rice Production and Price, 1961-96.
To ensure good measurements of the yield picture, IRRI has established more than 200 on-farm trials in 8 countries. Research results to date suggest: productivity decline observed in long-term trials at IRRI seems to be due to a decrease in the natural soil N supply and can be reversed by better N management; decreasing N supply from the soil is associated with changes in soil organic matter; depletion of soil K may be emerging as a factor; sheath blight may be involved in some cases; and micronutrients do not appear to be involved. Low fertilizer use efficiency is a continuing problem.
The IRRI breeding programme has continued to be remarkably effective. Some 750 new advanced lines are evaluated annually, and among the higher yielding lines advanced to multi-locational trials are genotypes with improved palatability and aroma and increased levels of resistance to green leafhopper. Pesticide use in many Asian countries has been reduced dramatically in recent years and it seems probable that IRRI's pest-resistant germplasms, combined with IPM, played a significant role in this outcome.
As a part of its conventional breeding programme, IRRI has been pursuing the development of a new plant type. Having developed a hypothesis about the factors limiting present day indica rice to around 10 t/ha, IRRI launched an initiative in 1989 to create a variety with fewer tillers to eliminate later unproductive growth, to increase grain number from 100 to 150 grains per panicle, to increase straw strength further to allow more height, and to incorporate thicker, more erect leaves for greater photosynthetic efficiency. NPT lines with 6-10, rather than the 25-27 productive tillers characteristic of, for example, IR72, have been produced. Potential grain number has been increased by a factor of 2-3. Some of these lines already have blast and bacterial leaf blight resistance incorporated.
The development of F1 hybrid rice has long been an IRRI aim and real progress has been made over the last few years. This is particularly appropriate as only recently have some NARS come into a position to exploit the potential benefits. Hybrid rice requires a more sophisticated agricultural and commercial infrastructure for viability than the conventional inbred lines.
The initial strategy chosen was the conventional cytoplasmic male-sterile (CMS), maintainer, and restorer line approach, and these genetic stocks have now been developed by both the Irrigated and Rainfed Lowland Programmes. Yield advantages of 15-20% have been achieved together with technologies for seed production that can yield 1-2 tons per ha by commercial growers. A system of on-farm seed production has also been developed which may allow sustainability without the usual commercial infrastructure. India (where it has been shown that some farmers will pay up to 10 times the cost of inbred seed), Vietnam, and the Philippines are the initial target countries for major adoption of hybrids. About 120,000 ha were grown in India in 1997 and IRRI expects this to rise to 2 million ha by 2000. Bangladesh and Indonesia are expected to adopt IRRI hybrids before the millennium.
Studies on the population dynamics of brown planthopper (BPH), using several genotypes and insecticide treatments, demonstrated that natural biological control combined with partial BPH resistance provides satisfactory limitation of BPH populations on irrigated rice. Neither BPH-resistant varieties nor sprays with Bacillus thuringiensis (Bt) were deleterious to the biological control of BPH and other planthoppers. This suggests that Bt-transformed rice would not affect the natural beneficial arthropod populations in the field. However, field trials will be necessary to provide a definite answer.
A methodology for assessing pest, disease, and weed management practices of rice farmers was developed and used through partnership with NARS scientists for surveys in 10 Asian countries. Widespread pesticide misuse, due to an erroneous perception of the pest problem by the farmer, was observed among 80% of farmers surveyed in the Philippines. On the basis of previous research on insect population dynamics, physiological effects of insect damage, and socio-economic analysis of farmers' decision making, an attempt was made to convince farmers in various areas, through farmer field schools or other means, that insecticide application for leaf folder in the first 40 days after sowing is not necessary. The media reached 97% of the farmers in the study sites in Long An province, Vietnam, and within 3 years, farmers' mean insecticide sprays dropped significantly from 3.4 sprays per farmer to 1.6, with no yield loss. The farmers accepted the IPM concept and the information spread to other districts with the result that 77% of the farmers stopped early season spraying. 1
1 Heong et al., 1998. Use of Communication Media in Changing Rice Farmers'' Pest Management in South Vietnam. Crop Protection, submitted.
Other achievements in the crop protection area included farmer-participatory research in The Philippines on tungro disease management which demonstrated the possibility to reduce disease losses by combining early planting of resistant varieties with crop sanitation to reduce inoculum carryover. Also, it was confirmed that rice yields may be improved up to 20% by cleaning rice seeds, which are often discoloured, unfilled, infected by pathogens and insects, and contaminated with weed seeds.
While yields at farm level are not declining in any of the major rice areas of Asia, indications of slower rates of yield growth have suggested the importance of research on prediction and maintenance of soil N supply, depletion of soil K, nutrient-disease interactions, and diversity and efficacy of microbes in the soil. Also, IRRI has initiated new long-term experiments with rice-rice systems at six sites in China, India, Vietnam, Thailand, and Indonesia. Preliminary results do not allow assessment of any trends, but it is thought that through intensive monitoring at the selected sites, a system for site-specific nutrient evaluation and management might be developed. Assessment of aeration during crop development or fallow periods, crop residue management, and other soil/nutrient management factors may help to determine their effects on soil properties, soil N availability, and microbial diversity.
In the area of breeding, much more still has to be done to optimize the traits of the new plant type and to combine these with more of the necessary conventional disease and insect resistance and preferred grain quality traits. Grain filling in particular is a problem but suitable parents with bolder grains have been identified and incorporated into the NPT programme.
For hybrid rice, work will continue on thermosensitive genetic male sterility (TGMS) as an alternative to CMS, both by introduction and mutation breeding. Some TGMS lines have already been shared with NARS for evaluation of performance, stability, outcrossing, and combining ability. Seed production in adequate quantities remains a problem in the tropics, relative to that achieved in China, and selection continues on the tropical material to increase seed yields.
The main goal in pest management improvement is to ensure continued production capacity by protecting the rice-growing environment and the farmers themselves in the face of an intensifying cropping system. Yield is optimized by maximising knowledge input in order to decrease pesticide input and extend the durability of resistant varieties. Research will thus proceed on the understanding of the interactions among the pests, the crop, and biological control agents under intensive irrigated rice-cropping systems in order to develop systems for sustainable IPM at the farm level.
Management schemes for transgenic rice with the Bt gene will be developed and evaluated for their long-term sustainability and impact on arthropod environments. The prospects for durable resistance to insect pests and diseases will be enhanced through testing location-specific recommendations for the deployment of germplasm with appropriate resistance characteristics. Sheath blight resistance and control will become a major target.
The database on pest incidence and rice-cropping systems, combined with GIS, will be used to identify pest risk zones in selected countries. Predictive models will be incorporated into district-level risk-zoning. An integrated system for regional decision making to manage blast using early detection and forecasting in two blast hot-spot areas in humid Asia will be evaluated. Diagnostic kits for blast and sheath blight will be integrated into disease management decision-making at the field level.
Rice field habitat biodiversity will be studied to understand and generate habitat manipulation favouring biological control of pests. The main target is the impact of crop residue/nutrient management on natural biological control of pests. Snail ecology and weed ecology working groups will be established within IRRI. For minimizing herbicide input, the population biology of Echinochloa and of wild and weedy red rices, particularly the weedy form of O. sativa, will be researched. Work on the environmental hazards of herbicides and herbicide resistance in weed populations will be initiated. Transgenic rice with resistance to herbicide will be evaluated for its benefits and risks to contribute to developing policies on how this technology might best be used.
IPM training will be modified to include women as a target audience and to inform them of the risks of unsafe use of pesticides. Farmers' seed health pest management will also be a target for improvement. The IPM Network, integrated into the Irrigated Rice Research Consortium (IRRC), will continue to play a vital role in adapting and promoting IPM at the national level in collaboration with NARS, various NGOs, and the FAO Intercountry Programme on Rice IPM. An Integrated Crop Management approach will be adopted, linking nutrient status with pest development and promoting nutrient- and pesticide-efficient technologies across irrigated rice systems in the tropics and subtropics. Responsibility of various NARS partners in the Philippines, Vietnam, Thailand, India, China, Malaysia, Indonesia, and Lao PDR will increase with the end of funding in 2000.
The hybrid programme has made impressive strides and adoption of hybrids is expected to expand, particularly in India. The partnerships forged with local seed companies will be important for the continued success of the programme. Work is planned to clarify the existence and nature of heterotic groups, possibly for the different ecosystems and environments within ecosystems, to facilitate the breeding of hybrid parents, rather than relying on the best products of the pure-line programme, seems appropriate and should be addressed in conjunction with molecular characterization of the germplasm collection (CE1). Application of QTL analysis to elucidate the genes involved in the heterotic response, specifically in IRRI hybrids, is also encouraged. The Panel is concerned that IRRI has allowed the press to quote the potential yield advantage of hybrids as 2 tons (which is achievable under only the most optimum experimental conditions), and that IRRI should encourage use of the 10-20% yield gain estimate, which would give farmers more reasonable expectations.
Apomixis research, which in theory will allow heterosis to be fixed, is continuing apace elsewhere and may become a realistic breeding adjunct in 10-20 years. In the Panel's view, IRRI should keep a close watching brief while not investing too heavily in this competitive area of fundamental research.
The Panel noted the continued progress with the development of the NPT. This strategy at present consumes about 30% of the resources of the irrigated rice breeding programme. Given the uncertainty that surrounds all plant breeding endeavours, it seems advisable to pursue the objectives of the NPT in the general context of the irrigated breeding programme. This effort was given too high a profile too soon and the Panel suggests that future references to this programme by Management should place it in an appropriate context, allowing the breeders the time and the options needed for the development of NPT cultivars.
The research investment of IRRI into understanding the relationship among pests, crops, and biological control agents has reached fruition in the development of improved pest management strategies transferred to farmers in 8 countries through the IPM Net. The combination of output from biological and socioeconomic research has proven very productive. The Panel commends IRRI for this achievement. The future programme builds on this successful strategy and will allow early identification of emerging problems and the initiation of research to provide the knowledge to manage them in an economically efficient and rational way. Priority setting, considering the lack of comparative advantage of IRRI for research on problems not occurring in the Philippines, will be a challenge. Further strengthening of NARS capacity and facilitation of international collaborations with ARIs in non rice-growing countries should be a part of the solution.
The responsiveness at the farmer level in relation to IPM should be most useful for channelling new information to farmers. This approach could also be valuable for other regions, but the very favourable conditions in the area chosen for the case studies should be kept in mind. Methods for efficient site-specific control of sheath blight, weeds, snails, and rats are in demand. A strategy to respond to these challenges should be developed. The sheath blight problem is also highlighted by the shift from insecticide to fungicide sprays by IPM farmers in Vietnam 2.
2 Do Kim Chung & Kim Thi Dung, 1996. Pest Management in Rice Production in Vietnam: A Socioeconomic Assessment.
The Panel welcomes the integration of the IPM Net and the Mega project into one consortium. This synergy will increase the value of the heavy investment in long-term trials and provide solid data on nutrient status of the plant, crop management, and pest development.
Strategic research on weeds in irrigated and lowland rainfed rice ecosystems is important. Research on environmental hazards from herbicide use, and survey and biology of herbicide-resistant weeds should continue. Research on weed and rice competition under different environments and cropping systems is essential to formulate site-specific weed management.
Herbicide-resistant rice is a controversial issue. Although IRRI has been non-committal with regard to the development of herbicide-resistant rice, IRRI has a responsibility to evaluate available transgenic rice for its benefits and risks, and to assess the feasibility for potential use in certain rice-cropping systems.
Managing intensive irrigated rice production will continue to be more complex Farmers want shorter duration varieties to save time in the field and to obtain better use of scarce land and water, by allowing double or triple cropping of rice or other rice + vegetable or legume crop systems. To this end, and to save labor, direct seeding is taking hold in many places. To be successful, direct seeding involves a complex of tillage, soil levelling, seeding methods, weed control, and water management. Strategic research is essential on the direct seeding/weed management/water management complex in almost all of the rice ecosystems, but it is especially needed in irrigated rice. Also, it can be expected that the higher-yielding hybrids and NPT lines will need to have strategic research on their management as well as the factors of production involved, e.g., soil, water, nutrient management, and pest management, including weeds. Such considerations were among the reasons the Panel was surprised to learn there was no general agronomist working in the irrigated rice programme.
Irrigated rice continues to be the IRRI flagship programme. The staff restructuring programme has bitten deeply into resources, particularly technical support for the breeders. Perpetuation of this situation can only be reflected in less spectacular outputs at the next review. The Panel also noted the need for more capacity in biotechnology to service the marker-assisted-selection needs of the breeders in the longer term. The Panel shared the breeders' concern about the continued diminution of resources for INGER.
4.2.1 Recent Evolution and Current Focus
4.2.2 Achievements and Impact
4.2.3 Future Strategy
4.2.4 Assessment
Rainfed lowland rice (RLR) is difficult to define precisely because of its central position in the continuum of rice ecosystems. In fact, it is usually defined by what it is not: (1) it is not irrigated; (2) it is not upland because it is flooded for part of the crop cycle; and (3) it is not deepwater (flood-prone) rice because the flooding depth is less than 50 cm. It is usually transplanted, and is grown in levelled, bunded fields that are shallowly flooded with rainwater. According to IRRI, it is grown on about 25% of the world's rice area (about 48 million ha) and contributes about 18% of the global rice supply. One-third of South and South East Asian ricelands lie within this ecosystem, which dominates rice areas of Bangladesh, Cambodia, Myanmar, Nepal, and Thailand, and is important in India, Indonesia, Laos, and Vietnam. Madagascar and Brazil also have significant areas in this system, and it is estimated that several million ha in Africa could support cultivation of RLR. Because the crop is rainfed, the water supply is variable, and both drought and flooding may occur in the same season. Soil fertility is low and problem soils are common in this ecosystem. Most of the farmers are resource poor.
RLR can be divided into four subsystems: favourable, drought-prone, submergence-prone, and drought- and submergence-prone. Because the favourable rainfed lowlands can benefit from irrigated rice technology, IRRI has chosen to focus on the other three subsystems, with highest priority being given to the two with greatest potential for increased yield, the drought-prone and submergence-prone subsystems.
The RLR Programme comprises five projects across the locations of the Rainfed Lowland Rice Research Consortium (RLRRC), which was established in 1991 and operates at seven locations in five countries (Bangladesh, India, Indonesia, Philippines, and Thailand). The research has four programmatic objectives: (i) identifying rainfed lowland domains with similar abiotic/biotic stress complexes, (ii) developing varieties for these domains that tolerate the predominant abiotic stresses, yet respond to applied inputs, or favourable conditions; (iii) developing nutrient and water management strategies that maximize resource availability for crop output while maintaining the resource base; and (iv) identifying cropping system alternatives and decision-making criteria that allow a farmer to adjust systems to local requirements. Within these four objectives and five projects there are ten research areas: (i) agroecological characterization, (ii) germplasm improvement, (iii) drought tolerance and water supply, (iv) submergence tolerance, (v) nutrients, (vi) blast, (vii) weeds, (viii) dry seeding, (ix) gender concerns, and (x) risk analysis. In addition, seven research areas are listed for the Consortium: (i) physiology and genetics of submergence tolerance; (ii) germplasm improvement for South and South East Asia; (iii) physiology and genetics of drought; (iv) crop establishment and intensification, including direct (dry) seeding, weeds, and on-farm storage; (v) nutrient management/use efficiency and interaction with drought and crop establishment; (vi) risk analyses and management of technology generation and adoption; and (vii) large-scale definition of sub-ecosystems applying remotely sensed data.
Steady progress has been made on characterizing the ecosystem utilizing leading-edge technologies. Members of the RLRRC have participated in this work and have gained skills in using the modem technologies. The germplasm improvement programme has been releasing cultivars through NARS, and developing or exploring new agricultural methods and techniques. Genotype by environment (G x E) work will allow progress toward understanding the physiology of tolerance to drought and submergence and improving screening procedures for both stresses. Characterization of blast populations using molecular markers has provided evidence for the role of parasexual and sexual recombinations as mechanisms for generating genetic variation and structuring field populations. This provides useful information to increase the efficiency of breeding for blast resistance. Strategies have been devised for more effective management of nutrients in the fluctuating water environment characteristic of rainfed lowlands. Various weed control strategies have been investigated including the competitiveness of cultivars. The agronomic aspects of dry seeding have been investigated with useful results. Gender concerns have been investigated with some useful findings and there has been an attempt to quantify the nature and magnitude of risk in rainfed environments.
The RLRRC has proved to be an active partnership between IRRI and committed national programmes, with each member taking on responsibilities for one or more research problems. The first six years have seen progress in defining the problems, understanding key constraints, and developing solutions. It is recognized that RLR experiences sharp transitions in nutritional status due to the rapid changes between anaerobic and aerobic conditions, due to wetting and drying of the soil. New tools have been used to characterize rainfed lowland areas as an aid to planning and setting priorities for collaborative research. Earlier planting and direct seeding of better adapted varieties can help reduce exposure of the crop to late-season drought, and permit a short-duration, second crop to follow, thereby increasing crop diversification and using residual soil moisture. Weeds are a major constraint in RLR areas, and for direct seeding to be successful, weed control is essential. Cropping patterns gaining in use are rice-chickpea in Bangladesh and India, and rice-pigeon pea in India. From available data it appears that yield gain in RLR in Asia averaged only 11 kg/ha/yr over an 18-year period (1969-87). During that same period in The Philippines, RLR yield gains were 47 kg/ha/yr, indicating perhaps that some improvements were made in that country during that period, including perhaps adoption of modem varieties developed for irrigated conditions.
Failing a revolutionary achievement by IRRI and the RLRRC, the impact horizon of this programme will be very long. Rainfed lowland farmers are risk-averse, and changes in cultivars and practices are made slowly. Varieties released more than 40 years ago, such as Mahsuri, are probably still gaining acceptance. Impact from the additional research, in terms of improved welfare of people, will occur slowly and may, even then, be difficult to measure. Nonetheless, even modest gains in productivity in RLR will help improve the lot of many resource-poor farmers and poor rice consumers, thus helping meet equity goals while raising rice production in general.
Programme scientists believe they have defined the issues that must be addressed during the next several years. They hope to determine the control of root growth and water extraction in the often alternating, difficult, anaerobic-aerobic conditions of rainfed lowlands, so that selection criteria for drought avoidance may be refined and molecular markers developed for this complex trait. They plan to continue marker-aided selection for submergence tolerance and hope to understand better the applicability of tolerance under different types of submergence. They see a need for research to improve the reliability of crop establishment in direct seeding, while at the same time ensuring that weed competition is effectively managed They hope to devise a means, through early seeding, to capture the early flush of nitrate at the beginning of the wet season. They plan to investigate the role of nutrient management in enhanced seedling vigor and the role of nutrients in buffering the adverse effects of drought. Longer term studies related to gender issues are aimed at understanding inter- and intra-household strategies in sustaining food security in rainfed environments and their implications for technology adoption.
It would be difficult to overstate the importance of achieving gains in productivity of RLR, not only because of the immense land area involved but also because millions of farmers in this environment face drought, floods, pests, weeds, and soil problems, often all present in the same field in a single season. Most farmers in the rainfed lowlands have few options. Even small gains in productivity of the system could affect many people. Some economic studies 3 have concluded that in Asian countries where RLR accounts for 40-50% of the area, research on this ecosystem would produce higher rates of return than research on the irrigated rice ecosystem. Indeed, production in the irrigated ecosystem is already at 70% of estimated yield potential while production in the RLR ecosystem is at only 45% of estimated potential.
3 Barker, R., R.W. Herdt and B. Rose, 1985. The Rice Economy of Asia. Resources for the Future, Washington, DC, USA.Fan, S. and P. Hazell, 1997. Should India Invest More; Less-favoured Areas. Environment and Production Technology Division Discussion Paper No. 22, IFPRI, Washington, DC, USA.
The science in this programme is strong and the Panel commends the good work being done. But the time horizon is long and, at this point, the outputs of this programme are normal and measurable impacts and outcomes are minimal.
More knowledge of key adaptive traits for RLR would also be desirable in defining future strategy. The potential exists to assimilate genes from the extensive germplasm collection held by IRRI to produce cultivars that could represent a revolutionary advance in drought and submergence tolerance, as well as blast resistance. Incorporation of both major race-specific and minor blast resistance genes in adapted varieties by using marker-aided selection would increase survival ability against the extreme diversity and plasticity of blast populations. While it is important to sustain the increments achieved with conventional breeding approach, albeit aided by shuttle breeding and MAS and the associated studies in management, gender, and risk assessment, IRRI should be aiming more of its efforts toward a major breakthrough.
Because of the potential (and need) for impact in the rainfed lowland ecology; because molecular marker-assisted breeding has real and realizable potential for significant outcomes in the form of improved drought and submergence tolerance as well as blast resistance; because this work is labour-intensive; and because research support personnel is the major limiting factor, the Panel suggests that even higher priority be given to these objectives.
4.3.1 Recent Evolution and Current Focus
4.3.2 Achievements and Impact
4.3.3 Future Strategy
4.3.4 Assessment
Upland rice is a staple crop for a large number of poor farmers in Asia, Africa, and Latin America. According to IRRI, it is grown on about 17 million ha, although the area used is larger because of rotation with fallow and other crops. It is reported to be grown on 9.2% of the rice area and accounts for about 3.8% of total rice production. It is almost always a subsistence crop with few purchased inputs. Yields are reported to have risen slowly over the last 30 years from 0.3%/yr in India to 1.6% in Indonesia. The source of this increase is not known and it seems most probable that yields have been static. While the area devoted to upland rice has remained steady or increased slightly in some countries, the total area in Asia has dropped 16% in the past decade, most dramatically in Thailand (-90%) and the Philippines (-70%) This drop is attributed to the production of vegetables and other cash crops for urban areas and to economic policy issues, such as the relative rice to maize prices and subsidies for terracing.
The low yields of upland rice have been attributed to drought, weeds, soil nutritional imbalances, poor cultural practices, diseases, insects, and a lack of suitable varieties. The stated objectives of IRRI's programme are: to develop understanding and technology to maximize productivity and sustainability of upland rice where it is grown; to help maximize returns for farmer effort; and to reduce the area needed to satisfy demands for upland rice. The research areas identified are (i) germplasm improvement to overcome the major abiotic and biotic stresses, moving away from traditional breeding and selection and using new technologies to target, characterize, and incorporate desired genes (possibly including perennial rice and allelopathy); (ii) strategic understanding of nutrient availability in upland soils, (iii) biotic constraints with a focus on weeds, nematodes, and blast; and (iv) study of the socioeconomics of the uplands designed to understand the dynamics of the systems and the impact of new technologies and policies. New statistical procedures have shown (to no one's surprise) that the upland ecosystem is quite variable, which has led to a strong initiative to decentralize the conventional breeding efforts.
Presently, the focus is on the following specific research topics:
(i) Germplasm improvement. Improvement of productivity and stability of varieties is sought through incorporation of(l) drought tolerance through adapted duration, good root systems, and osmotic adjustment, (2) durable blast resistance, and (3) improved interference with weeds through improvement of allelopathic potential and rice competitiveness. Areas under investigation include root ultrastructure, vulnerability of the water transport pathway, and direct effects of water deficits on panicle sterility. The ultimate aims are to transform upland rice - still basically an aquatic plant - into a crop fully adapted to aerobic soil conditions; improve the value of the crop through good grain quality; develop a perennial upland rice to help reduce erosion in areas with steep slopes and high rainfall, and develop a participatory plant breeding programme in Eastern India to analyze adoption rates of improved rainfed germplasm and determine if farmer participation at any stage of the breeding process improves adoption rates.(ii) Resource and crop management improvement. This research is concerned with the production potential of upland rice, long-term strategic studies on nutrients, nutrient x water interactions, weeds, and nematodes.
(iii) Economic and policy analysis. This research includes the examination of the livelihood strategies of upland farmers and changes in land use at the farm level, and the comparative analysis of the economics of upland rice systems over a range of biophysical and socioeconomic environments.
The Upland Rice Research Consortium (URRC), with collaborating institutions from Brazil, Germany, India, Indonesia, Laos, Philippines, Thailand, and Vietnam, provides a framework for much of this research. Numerous other countries and institutions are associated with the URRC.
Molecular technologies have been used to map genes controlling root morphology, to characterize blast pathogen populations, and to analyze blast resistance genes. Strategies (yet unproven) have been developed to improve the durability of blast resistance, and a genepool aimed at improving partial resistance to blast has been assembled. Crosses have been made to perennial types and mapping populations are being developed. Nematode resistance has been reported in O. longistaminata, one source of perenniality. Allelopathy in rice has been demonstrated under laboratory conditions but cannot be separated in the field - if it is manifested there - from competitive effects. Work has been conducted and papers produced on the productive capacity of upland soils and the production potential of upland rice, phosphorus deficiency in low-input upland systems, soil spatial variation in the uplands, subsoil acidity amelioration, fertilizer formulation and placement, nutrient dynamics, nutrient x water interactions, weed competition, allelopathy, nematode dynamics, importance of market access, comparative analysis of upland systems, and the analysis of soil conservation strategies.
Drought, weeds, nematodes, phosphorus deficiency, and blast are the dominant problems of the ecosystem. Genetic solutions will be pursued using marker-aided selection techniques resulting from recent advances in molecular biology. Research will be continued to expand the knowledge of genetic control of these traits using molecular markers. The feasibility of a perennial upland rice will be evaluated. Resource management work will focus on policy, nutrient management, dynamics of soil erosion, and integrated pest management (particularly weeds). The Upland Rice Research Consortium will provide a collaborative mechanism for work on (i) soil-plant-water relationships, (ii) socioeconomics, (iii) adapted germplasm, and (iv) human resources.
IRRI began working on upland rice more than a quarter-century ago, and although numerous outputs (varieties released, publications, etc.) have emerged, it seems fair to say that impact has not been demonstrated.
The quality of science in this programme is good, but is not appropriately directed when opportunity costs are considered. That is, the resource expended in this programme would provide higher returns if expended on the rainfed lowland ecology. Upland rice has represented a dilemma for IRRI for more than 20 years. The first strategic plan, formulated in the late 1970s, stated that investments in upland rice research were not justified. Because it is an important crop for so many small farmers, many donors have pressed for IRRI's continuing involvement even though the problem appeared, and continues to appear, intractable and the area in upland rice is declining.
Resource management research in the highly variable upland rice ecology populated by resource-poor farmers is a clear demonstration of the difference between a need (which it is) and an opportunity (which it is not). This observation is supported by common sense and the lack of any impact. The major need in the resource management arena is a solution to the problem of phosphorus availability. Without a solution to the phosphorus problem, other interventions (including genetic and policy interventions) will have little impact in the upland ecosystem.
IRRI has continued to accept resources directed toward upland rice research. Currently, of the 8.1 scientist-equivalent positions in the programme, only 1.8 receive salary support from core funds. About 80% of the operating budget comes in the form of restricted funds from specific donors. The NARS need upland rice research by IRRI and the donors support the programme. The Panel considers this situation acceptable because of the valuable spin-off toward the solution of problems common to other ecologies (phosphorus availability, drought tolerance, and blast resistance, for instance) but strongly suggests that, where possible, IRRI direct attention to the most tractable ecosystems, where impact could be expected 4, or where, even in the absence of impact, the intermediate outputs of this programme could be substantial and useful in relation to other rainfed ecologies.
4 IRRI Upland Rice Ecosystem Programme. The Future of Upland Rice. EPMR briefing document.
4.4.1 Recent Evolution and Current Focus
4.4.2 Achievements and Impact
4.4.3 Future Strategy
4.4.4 Assessment
Flood-prone rice is defined as that growing on inland areas where the sustained water depth exceeds 50 cm and tidal (coastal) wetland rices. Some flood-prone rice zones have undergone dramatic change with the availability of varieties and low-cost irrigation facilities that allow for the production of a high-yielding dry (boro) season crop. During the flooding period the land may be fallow or be used for alternative purposes, such as fish and shrimp farming. The traditional culture persists on about 70% of the original flood-prone area, covering approximately 12 million ha, mostly in Bangladesh, Cambodia, India, Myanmar, Thailand, and Vietnam and accounting for 4% of world rice production.
IRRI has adjusted its focus from emphasis on breeding deepwater rice varieties for flood-prone areas (work now devolved to NARS) to giving increased attention to strategic pre-breeding research and crop and resource management to improve productivity and sustainability of these lands. The objectives are to evaluate socioeconomic and biotic constraints to productivity of flood-prone rice, analyze the roles of mineral elements in flooding tolerance and productivity of rice cultivars, develop environmentally sound soil and water management strategies, and determine the potential effects of global warming and changing weather patterns on flood-prone ricelands. Coastal areas are the primary focus and there is an emphasis on partnerships between concerned NARS. Crop improvement activities now focus on semi-deep and coastal areas utilizing molecular marker-aided selection techniques for resistance/tolerance to soil-related stresses, submergence tolerance, novel genotypes for increasing yield and improving yield stability, and improving micronutrient content to enhance human nutrition.
To the extent that IRRI has been responsible for the cultivation of high-yielding, short-duration boro rice in the flood-prone zones, achievements and impact have been dramatic. However, it seems clear that the varieties that are utilized originated from irrigated breeding programmes, including that of IRRI, rather than the flood-prone programme, and it appears that their cultivation in the zone was a result of farmer initiative and was not based on research by IRRI, but rather by partner NARS. Thailand has now assumed regional responsibility for South East Asia at its own expense and Thai varieties have been released in Cambodia and Vietnam. To date, varietal improvement or modified cultural practices have resulted in few achievements, and the impact has been negligible. However, IRRI now focuses on the semi-deep coastal areas for which a new plant type has been developed incorporating need-based elongation and high-yielding plant type characteristics above water. Achievements of potentially major impact are the tagging of major genes and QTL identification for submergence tolerance, elongation ability, salinity tolerance, and phosphorous efficiency. The pyramiding of genes controlling different physiological mechanisms for salt tolerance has been accomplished, but field verification is needed to prove the usefulness of this approach.
A significant development in which IRRI has been involved is the pioneering reclamation of acid sulfate soils in Vietnam that are annually subject to severe flooding. Here a management system worked out by scientists of Vietnam, IRRI, the International Mekong Committee, and the University of Wageningen contributed to the conversion of hundreds of thousands of hectares of low-productive deepwater ricelands, many of them in the Plain of Reeds, into highly productive rice fields. Here double and triple cropping with high yields (4-5 t/ha/crop), even in the first year after reclamation, are being achieved. Some off-site adverse effects of this conversion have also been quantified and corrective measures applied. This development is exciting and the experience in the proper use of acid sulfate soils in flood-prone areas is valuable.
Using the approaches outlined in section 4.4.1 above, the strategy is to assist in the development of sustainable intensification/diversification systems to reduce poverty; to develop appropriate soil and water management techniques for acid, acid sulfate, and peaty soils that will prevent pollutants from leaching into waterways and affecting downstream flora and fauna; to produce novel genotypes tolerant of soil and water-related stresses and with improved nutritional properties; and to provide national policymakers with information on the socioeconomic and environmental impact of intensification as well as to increase the sustainable use of flood-prone ricelands.
The acid sulfate soil land reclamation effort in Vietnam has led to the settlement and conversion of flood-prone lands into highly productive ricelands, and the Panel commends the innovative collaborative NARS/IRRI /ARI research that made this impact possible. Otherwise, this programme has shown limited direct impact. The changes in cropping pattern in the flood-prone ecosystem would have happened without this programme. The quality of the science in this programme is good and a number of commendable outputs have been noted. While a number of exciting new efforts in the programme, such as the improvement of human nutrition through the genetic manipulation of micronutrient content, and the combining of salt-tolerance genes, are very worthy of pursuit, they are relevant to all ecosystems and do not justify the retention of the flood-prone research activities as a separate research programme.
4.5.1 Germplasm Characterization, Biotechnology, and ARBN
4.5.2 Exploiting Biodiversity for Sustainable Pest Management
4.5.3 Systems Approaches to Quantify Performance of Rice Ecosystems
4.5.4 Assessing Opportunities for Nitrogen Fixation in Rice
4.5.5 Constraints to Sustainable Development of Rice Ecosystems and Technology Impact and Policy Analysis
The Cross Ecosystems (CE) Programme was established as a vehicle for research comparing the different systems or developing tools and methods applicable to all ecosystems. The five projects in the Programme are extremely diverse, ranging from macro policy to molecular biology. In this review the Panel has separately addressed those projects involving:
(i) biotechnology tools and their application to breeding and germplasm assessment and exploitation, and the technology transfer to NARS (Projects CE-1,2 and 6);(ii) rice pest and disease biology (CE-3);
(iii) quantification of different ecosystems (CE-4);
(iv) nitrogen fixation in rice (CE-7); and
(v) socioeconomic aspects of the different ecosystems (CE-5).
4.5.1.1 Recent evolution and current focus
4.5.1.2 Achievements and impact
4.5.1.3 Future strategy
4.5.1.4 Assessment
IRRI holds the world's largest collection of cultivated and wild rices. It has long been recognised that the value of, and ease of access to, the collections is enhanced by description and characterization in terms of traits of interest. Characterization of the individual accessions with the new generation of molecular diagnostics can now add efficiency by identifying near-duplicates among the 80,000 accessions, and by identifying manageable subsets of the material - 'core collections' - which will enable access to the widest ranges of variation. Similarly, the wild relatives of rice, a vast untapped resource of potentially beneficial genes, are more accessible when their cytogenetics, genome structures, crossabilities, and taxonomy are better understood.
DNA markers, which first became available to plant geneticists 10 years ago, offered the possibility of 'breeding on the laboratory bench' for adaptive genes which are difficult or expensive to recognise and select for in the field, and to assemble them in combinations that would be extremely difficult by conventional breeding techniques. The first molecular maps of rice became available eight years ago and IRRI scientists were quick to recognise the opportunities afforded and have since been at the forefront of their application in breeding. Molecular biology also offered the possibility of transferring isolated genes directly into rice. Genetic transformation could be used to introduce beneficial variants of natural genes back into rice more quickly than by crossing and selection or to introduce natural or synthetic genes not normally available to the breeder to produce enhanced or novel phenotypes. Since 1992, IRRI has built up capability in the various techniques of transformation and is already producing transgenics containing genes with potential for insect, disease, and stress tolerance.
Preliminary research has demonstrated the power of molecular classification to identify duplicates and establish relationships among germplasm bank accessions of both cultivated and wild rice. More than 100 accessions representing the eight putative species in the AA genome group, which includes O. sativa, have been investigated by intercrossing and cytological analysis of the hybrids. The work has resulted in the reclassification of O. ridleyi and longiglumis as conspecific and elucidated crossability and chromosome pairing relationships among these species, which represent the first port of call for breeders screening in the secondary genepool.
Aided by molecular technology, the wide-cross programme, which has now succeeded in producing interspecific hybrids between all wild Oryza species except one, continues to provide both novel materials for use by breeders and fundamental knowledge of importance to rice taxonomy and cytogenetics. Notable successes of importance to IRRI's breeding programme include the identification and introduction of new genes from wild species including blast resistance, BPH resistance, and the broad spectrum bacterial blight resistance gene, Xa21, which has since been cloned elsewhere and returned to IRRI for use in the transgenic programme.
Among the achievements gained through application of markers was the alignment of the cytological and genetic maps, by locating the rice chromosome centromeres on the genetic map using unique cytogenetic stocks produced at IRRI. This work was significant in bringing together and standardising the major international rice genome maps, which are now shown using the IRRI alignment. The IRRI group has made major contributions in the genetic mapping of many adaptive genes, including resistance's to pests and diseases and traits relating to hybrid rice production. The same technologies have been exploited to produce novel resistance gene pyramids and transfers of beneficial genes discovered in wild relatives.
Progress has been made in developing transformation technology and, although efficiencies are still low, transformation is now possible with all rices, including the NPT. Potentially insect-resistant (Bt) and sheath blight-resistant transformants have been produced and many other constructs are in the process of being genetically engineered into rice.
Doubled haploids produced by tissue culture, an adjunct to transformation, have been used to accelerate breeding programmes, e.g., in the production of new salt-tolerant lines ready for release in the Philippines.
The Asian Rice Biotechnology Network was set up in 1993 with the aim of building biotechnology strengths in NARS institutes. An ARBN laboratory, overseen by IRRI staff, has been set up to receive trainees and collaborating researchers from the ten member countries. Some ten visiting scientists use the facility each year, and about five training courses, some lasting up to eight weeks, are run each year. Exciting collaborative research on bacterial blight and gall midge fingerprinting, and fine mapping and pyramiding of bacterial blight and blast resistance genes has been achieved.
Mass analysis of accessions with molecular markers is planned to move further towards establishing 'core' collections. The improved knowledge of the wild relatives of rice will be exploited to enhance methods of alien transfer from even more distant relatives, thus extending the available genepool in the search for novel genes, e.g., those controlling tungro, yellow stem borer, sheath blight resistance, and submergence and acid-soils tolerance. The mapping of wild-cultivated cross progenies has exposed unexpectedly small introgressed interstitial segments. The underlying mechanism will be further investigated.
The use of MAS will continue to be integrated as a common enabling technology for IRRI's breeding. Gene mapping activity will be directed increasingly at extended backcross analysis of the genes underlying quantitative traits (QTLs), such as yield, in both cultivated and wild rice.
Transformation research will continue, particularly to improve efficiencies from the current 0.1% and to enable the technology to work with all genotypes. Transformation targets include a host of newly isolated genes with potential to confer disease and pest resistance, submergence and drought tolerance, and grain size and quality obtained from collaborators, particularly in the RF Programme.
Overall, most of IRRI's work in the biotechnology area, both as housed in PBGB and EPPD, is of the highest quality. Moreover, the Panel believes that the strategic sub-projects being carried out alongside the more applied activities are important as the channel through which IRRI has access to, and can take advantage of, the new developments in molecular biology and the products of the extensive network of rice biotechnologists around the world.
The Panel endorses the research on germplasm characterization and taxonomy, which can improve still further the key resource represented by the collections. Moreover, the Panel endorses the new research role of GRC which, in earlier times, was assigned only a service activity. However, some overlap was noted between former CE-1 and CE-2 in taxonomy and genome characterization research and the Panel believes that greater coordination would be valuable. Identification of potential heterotic groups to facilitate hybrid breeding parents will become important (IR-1) and the Panel hopes that these goals will be aligned with the more general germplasm characterization (CE-1) in the future.
The proposals to extend research to the identification of key QTLs were considered appropriate. The Panel noted that the role of the Biotechnology Group no longer provided a MAS service for breeders, but now provided laboratories for the breeders' use. In view of the staff shortages which constrain use of the laboratory, the Panel suggests that IRRI investigate the potential and cost-effectiveness of some developmental work on the application of automation and robotics in MAS.
The work and future plans, in the tissue culture and transformation areas were judged to be appropriate and to be using external collaborations to best advantage. The Panel noted that a continued lack of a regulatory framework to enable field testing of transgenics in the Philippines would soon erode the advantages of IRRI over other partners as a potential collaborator, and suggests that senior management continue to do everything possible to hasten this process.
An area of science not addressed in the MTP is bioinformatics - the assembly, linkage, and manipulation of data relating to the collections, breeding, trials, genome studies, and DNA sequence data. In order to maximize utilization of the vast amount of information being generated at IRRI alongside that already collected in all the major cereals, the Panel suggests that IRRI establish a presence in this area while information exchange is still free (see Biometrics Unit in Chapter 5).
The Panel was impressed with both the concept and the operation of the ARBN. They noted a complementation between ARBN and the RF Network and saw the Project as an ideal means of partnership building and continuous technology transfer.
The excellence of IRRI's applied research and contributions notwithstanding, the recent radical developments in plant genomics, and rice genomics in particular, are good reason to revisit IRRI's future in the post-genomic age, which will begin as early as next year when the first rice genomic DNA sequence becomes available. IRRI should consider carefully how it can exploit its comparative advantages to maintain a central position and thereby continue to provide its NARS with the best advice, training, technology, and germplasm. The proposal presented to the Panel of establishing a major 'knockout' facility, using either transposons or deletion mutants, for application in gene function analysis may be extremely timely and should bear serious consideration (see Advances in Rice Genomics in Chapter 6).
4.5.2.1 Recent evolution and current focus
4.5.2.2 Achievements and impact
4.5.2.3 Future strategy
4.5.2.4 Assessment
This project evolved during the period under review from the previous Crop Ecology and Pest Science Sub-Programme. Its aim is understanding the key relationships between a spatial and temporal mix of organisms that maintain habitat diversity and that contribute toward the stability and resilience of rice ecosystems. This knowledge will assist in the development of strategies and on-farm practices that promote sustainable management of pests.
Many of the plant resistances incorporated in earlier years, especially those based on single genes, were not long lasting, when used on a large scale under certain pest pressures. Because durability of resistance to insect pests and diseases is a central issue in sustaining the productivity of rice across ecosystems, strategic research on the genetic diversity of pest populations as well as on the genetic bases for pathogenicity in the pest and for resistance in the host is required. This would allow the development of models for simulation of rice-pest coevolution and of strategies for improving and deploying resistance.
Predators and parasites of insect pests in the rice ecosystems are essential for containing harmful pest populations at a tolerable level. This has been demonstrated and utilised in developing IPM strategies. At the crop level, the diversity and dynamics of pathosystems and of pests' natural enemy communities must be analysed and linked to farming practices and the presence of habitats other than the rice crop within the rice ecosystems. These data will allow the development of biodiversity management strategies that ensure the resilience of the crop to pests.
Quantitative data on pest constraints, crop losses in varying production situations and environmental parameters at the plot and regional level are necessary for both systems analysis and simulation, and for a better understanding of the processes linking the pest genome, plant genome, crop, ecosystem, and the region. These simulations are used for predicting the durability of disease resistance genes, modelling pest development, assessing pest risk in emerging production systems, and developing location-specific pest control schemes.
Near isogenic lines and lines with gene pyramids were developed by IRRI to assay virulence of blast, bacterial blight, and tungro viruses, and for estimation of host resistance genes in germplasm. Mixed planting of rice varieties having different genes for resistance has demonstrated a positive effect on blast control in upland and irrigated rice.
Molecular tools have been developed and transferred through the ARBN to NARS for pathogen characterization and genetic diversity analysis of the blast, bacterial blight, and tungro pathogens. Effectiveness of Pi-2/Pil against blast and Xa7 genes against bacterial blight has been demonstrated in the Philippines. RFLP analysis has been applied on gall midge from six countries, and the results have been used in the specification of the screening sites for resistance and for possible resistance deployment strategies.
The efficacy of Bt toxins from eight lines of Bacillus thuringiensis has been tested against stem borer and leaf folder populations. Yellow stem borer populations collected in China were found to be much less susceptible to tested Bt toxins than populations from the Philippines, with resistance factors reaching around 70 against CryIIA and 160 against CryIC. This highlights the risk that plants with the gene for these toxins may not perform well for yellow stem borer control, as well as the need for undertaking in-depth studies of pest population susceptibility before embarking on rice transformation.
Pioneering studies have been undertaken on the importance of non-rice habitats for enhanced biological control of insect pests, identifying several predators and parasitoids from grassy areas. Biodiversity research tools have been developed and NARS partners trained in their use in various environments. Ecostatistical software has been written to elucidate the structure and dynamics of invertebrate communities across rice landscapes.
A crop-loss database has been developed by IRRI from surveys of more than 700 farmers' fields in South and South East Asia, and from experimental data from more than 400 plots. A survey portfolio has been produced for the characterization of rice pest constraints. Currently, the Experimental Rice Crop Loss Data cover a range of attainable yields from 1.7 to 11 tons/ha, and report damage in the range from 0 to 4.3 tons /ha (in absolute terms) or 80.6 % (in relative terms) in response to 11 different injuries alone or in various combinations. A simplified crop model has been developed to handle the database; and correspondence analysis techniques have been developed for improved pest profile characterization for IPM.
The project objective is to improve the use of and to develop new environment-friendly pest management tools. These tools will include deployment strategies, biocontrol agents, single- and multispecies models, and spatial designs of rice and non-rice habitats.
Gene deployment strategies will be developed according to both the type of resistance genes in rice germplasm characterized by rapid molecular diagnostic tools, and the durability of these resistance genes evaluated by predictive models taking into account possible fitness penalties associated with pathogen adaptation, through the frontier project 'Predicting durability of disease resistance: the gene to gene relationship'. Further, new genes will be identified from rice germplasm and non-rice sources for durable resistance to insects and diseases. Sheath blight resistance and control will be the main target of research.
Ecological statistics are being used to develop software that will determine spatio-temporal patterns of vegetation and invertebrate populations across paddy landscapes. Habitat diversity strategies for improved biological control of pests will be distilled into "simple rules" for farmers' participatory research. NARS scientists will be trained at a "lighthouse site" for site characterization, landscape ecology, invertebrate and microbial taxonomy, and varietal improvement strategies.
The new simplified crop growth model developed for characterizing pest damage will be used to check results of risk analyses and to extrapolate results for future scenarios. Crop-loss surveys will be carried out in other areas, enhancing the value of the crop-loss database. The epidemiological model for sheath blight will be further improved for use in evaluating sheath blight management options.
Very good innovative research on relevant topics for pest management improvement in various rice ecosystems is performed under this project. Some detailed portfolios for data collection and analysis have been prepared for use by trained NARS partners. The Panel commends the projects' scientists for these achievements.
The idea of combining the potentials of biotechnology for characterization of pathogen and host populations, with ecology and systems analysis of the rice ecosystems, in order to manage biodiversity for rice protection, is excellent. Cross-linkages between genome diversity and ecosystem diversity are identified in some of the research topics, and some staff members clearly articulated their views on ways to maximize these interactions. The project could increase its impact by focusing integrated team work with Ecosystems Programmes on carefully selected case studies.
The Panel is convinced that the assembly of quantitative information on crop losses due to pests under certain management practices and ecosystems and its use for systems analysis and modelling are vital to the development of sound plant protection management practices, as well as for making policy recommendations. The Panel was therefore pleased that a senior NRS scientist will assure the maintenance of the ecosystem database when the present projects' funding ends in 1999.
4.5.3.1 Recent evolution and current focus
4.5.3.2 Achievements and impact
4.5.3.3 Future strategy
4.5.3.4 Assessment
IRRI is involved in systems analysis to investigate processes and their interactions using mathematical models to describe and quantify individual processes and critical factors that limit yield. Individual processes are linked to crop performance by integration of models. Models so developed are used in specific ecosystem research programmes by providing a physiological basis for germplasm selection for specific environments. The project was funded in 1997 at US$ 659,000.
This work is carried out under Project CE-4, which includes activities in three areas: (i) relationship of response of physiological and soil-related processes to quantified environmental factors; (ii) process-based models that simulate rice production in various ecosystems developed for use in quantifying constraints to rice production in agroecological zones, and resolving specific ecosystem problems related to yield potential and optimization c water, nutrient, and pest management; and (iii) continued support for national programme teams trained in use of systems techniques for generating and synthesizing research knowledge.
The basis for a 15 t/ha yield of Shanyou 63 grown in the Yunnan Plateau in China was determined to be: high production of biomass, high sink size, high leaf area index and leaf are duration leading to high radiation use efficiency and low night respiration rates. Indica rices were found to have higher photosynthetic and respiration rates and higher biomass production than tropical japonica rices.
A rice crop physiological model, ORYZA, was developed to simulate rice growth and development. A model for N management was applied to estimate yields under different scenarios, including climate change. Crop models and pest modules are being coupled to estimate pest-induced yield loss. A simplified rice model was developed and calibrated to simulate injury mechanisms in different production situations.
Uptake of soil N was limited by root uptake properties and N diffusion to root surfaces. Completed physiological research indicated the importance of lowering panicle height to increase yield of irrigated rice. In modelling, the BLASTSIM.2 model was validated and calibrated in Thailand. A simple model identified the process that governs efficiency of grain filling. Simulation models were validated for rice-wheat systems in Bangladesh, India, and China, that gave improved understanding of variety, moisture, and nitrogen interactions. A rice blast polyclonal kit tested in northern Thailand indicated agreement between kit reflectometer readings and severity and incidence of panicle blast.
Fourteen teams from seven NARS were trained and equipped for systems research. National networks were developed in selected NARS for applications in crop and pest management.
This work is aimed at developing knowledge and/or methodologies to improve research by IRRI scientists and national collaborators, and its impact will be difficult to measure except in terms of improved research methods or output.
IRRI plans to acquire knowledge and tools by developing systems approaches and models that integrate knowledge on favourable and unfavourable environments and predict ecosystem behaviour in different situations. IRRI will work with NARS to establish joint research priorities for agroecological zones and rice ecosystems.
In physiology studies, IRRI plans to study causes of poor grain filling in the wet season, and evaluate the hypothesis that poor grain filling in the new plant type may be due to lack of apical dominance of spikelets.
IRRI has been involved in modelling work for more than a decade, and the work appears to be of good scientific quality. Developing tools and research methodologies, suggesting new approaches, improving the knowledge base concerning rice and its production, and helping NARS to improve research capacity are likely to be modelling's main contribution. The work on the physiological basis of very high yield in China is very interesting, as is the work on the physiological reasons for poor grain filling in NPT and in the wet season. Such work can provide essential information to help solve problems affecting rice productivity.
4.5.4.1 Recent evolution and current focus
4.5.4.2 Achievements and impact
4.5.4.3 Future strategy
4.5.4.4 Assessment
IRRI designated this effort as one of its new frontier projects in its 1994-98 MTP, and described it as follows: «Enabling rice plants to fix nitrogen to help reduce the dependence of farmers on chemical nitrogen and thereby achieve a more environmentally-friendly rice production system».
This work is done under Project CE-6, which includes research activities in four areas: (i) non-nodular associations, (ii) nodular associations, (iii) transfer of nif genes to rice, and (iv) CO2 fixation and N use efficiency. The project also includes efforts related to the Biological Nitrogen Fixation Working Group, which has prepared a proposal for a Frontier Project on N; Fixation.
Methods have been developed to identify nitrogen-fixing bacteria in pure culture and their presence in plant tissue. This method is used to help screen putative endophytic bacteria isolated from field-grown rice. Methods to isolate rice root endophytes have been developed and the endophytes were marked with a gene so their colonization could be studied. N; fixing bacteria have been found to be associated with rice seeds. In studies of possible nodulation, some rice varieties were found to have the ability to induce gene expression in Rhizobia.
The future impact of this project will depend on a major, and in the Panel's view unlikely, breakthrough.
Work will continue in the four research areas listed in 4.5.4.1 above.
This project is well-funded and was projected to receive US$ 0.869 M in 1997. In the absence of a CCER or peer review paper that had reviewed the project in depth as to focus, approach, and probability of success, it is difficult for the Panel to make judgments on its work. However, the possibility of success for nodulation in rice appears remote. The Panel is pleased that a peer review is scheduled to review the focus and approaches taken by the project on nitrogen fixation in rice. The Panel also strongly suggests that this review addresses the timeliness and likelihood of success of the project.
4.5.5.1 Evolution and current focus
4.5.5.2 Achievements and impact
4.5.5.3 Future strategy
4.5.5.4 Assessment
In order to facilitate the transfer of technology to farmers across various cross-ecosystems to enable them to augment their farm production, it is important to understand the socioeconomic setting in which they are operating and to have information on the numerous socioeconomic and biophysical constraints they may be facing in adopting the new technology. The main objective of this programme is to collect information on socioeconomic indicators and on characteristics of agroregional and rice ecosystems with a view to the planning and prioritization of research. This involves undertaking interdisciplinary research and coordination with various scientific and social science departments. It also involves setting up priorities for research that could both evaluate the nature and extent of new technology absorption and its limitation and its ex-post impact on different socio-economic groups, income distribution, and poverty alleviation. The programme also aims at monitoring developments in the rice sector regarding changing patterns of production and consumption and the impact of trade liberalization, input use, and pricing policies and their implications for rice research. The other objectives are: to project medium- and long-term demand and supply balances for rice for major rice-growing economies in the context of changes in socio-economic conditions and macro-economic policies; to assess the importance of rice in the rural household economy and analyze the relationship between productivity growth in rice cultivation and expansion of the rural non-farm sector; and to contribute to the strengthening of NARS to undertake socioeconomic research studies on their own or in collaboration with IRRI.
The main priority research areas are: (i) to maintain and update the socioeconomic database on rice ecosystems, monitor levels in rice output and yield and draw implications for policy; (ii) to develop methodologies for delineating homogeneous regions on the basis of both biophysical and socioeconomic factors and develop approaches to addressing location-specific problems; (iii) to study the effect of rice and non rice production systems on food security; (iv) to evaluate the impact of new improved technology on income distribution, poverty, and women's welfare; (v) to analyze priority research problem areas and evaluate rice research capacity of NARS; and (vi) to identify strategies for improving farmers' income through commercialization of farming.
In the comparative analysis of rice ecologies, IRRI has built a rice statistics database, conducted rice sector analysis, and collected data on rice research capacity in different NARS. Also, IRRI has completed a study identifying constraints and factors contributing to yield gaps in partner countries.
Two publications (World Rice Statistics, 1993-94; and Rice Almanac, Second edition, 1997) are being brought out regularly. Another statistical publication is Rice Area by Type of Cultivation. These, and other publications dealing with rice supply and demand at the regional level, will help policymakers and research scientists in determining research priorities and planning.
IRRI has completed, and published the results of, a project on "Projections and Policy Implications for Medium- and Long-Term Rice Supply and Demand", in collaboration with IFPRI and National Policy Research Institutes in seven major countries. IRRI has also undertaken various projects on identifying socioeconomic constraints in different environments.
Research has improved understanding of farmers' decision-making processes in different ecological settings in relation to changes in the rural household economy. Impact studies include evaluation of the effects of new technologies on equity, poverty, and sustainability. This work includes a survey, initiated in 1996, of households to assess the differential impact of technology changes across various categories of farmers and households.
National policymakers, donor communities, and research managers have been informed of technological progress in the rice sector and the constraints to removing the yield gap at the national and subnational levels.
The strategy is, using the approaches outlined above, to assist in the development of appropriate technological packages in different ecological regimes with a view to augmenting the production of rice, the main crop, and other crops grown with it. Attempts will be made to look at the household economy by including allied farm and non-farm activities.
The new MTP proposes studies on assessment of rural conditions in Myanmar, Eastern India, Madhya Pradesh, Bihar, and eastern Uttar Pradesh.
On gender, the Project aims to generate information on involvement of men, women, and children in economic activities and the household decision-making process and their separate contribution to household income, and future work will assess the impact of improved technologies on the well-being of women and children. IRRI will also evaluate the impact of migration of men on women's' involvement in farming and effect on management of farms and input use efficiency
Social science work at IRRI is quite strong. The creation of the Ricestat International and Country Sources database and its dissemination are an important development. The addition of GIS will also be a useful aid in delineating homogeneous regions. The numerous studies on demand - supply of rice, impact of new technology on poverty, and income distribution across different ecological regions are not only quite rich analytically but also provide useful guidelines for future planning and policy making.
However, the Panel observes some important gaps in the range of socioeconomic studies. Firstly, studies confined to rice only, without taking other crops into account, are only partially able to capture the impact of new technologies on household incomes and their living standards. Hence, there is a need to undertake studies of rice and other crops together with allied farm and off-farm activities. The second major limitation is that little information has been collected, and much less analyzed, on natural resource management and decisions taken by farmers. The Panel suggests that IRRI review its approaches regarding these areas.
The Panel was able to examine a considerable proportion of the IRRI research Programmes. It found that most of the work was scientifically exciting and of good quality.
Most of the important research objectives (drought tolerance, submergence tolerance, salt tolerance, improved human nutrition, etc.) cut across more than one of the three rainfed programmes and the breeding techniques needed to address these objectives are the same. A reduction in administrative effort and transaction costs is desirable and limited resources must be used in an efficient and flexible manner.
The Panel recommends that the Rainfed Lowland, Rainfed Upland, and Flood-prone Programmes be combined into a single Rainfed Rice Programme, in which related lines of work can be brought together, emphasizing those where prospects for success are greatest.
The Panel noted the fact that the Irrigated Rice Programme no longer has an agronomist. They also noted the need for more technical staff to support plant breeding, and in the longer term for more capacity in biotechnology to service the marker-assisted selection needs of the breeders. Molecular marker-assisted breeding promises important advances in the form of improved drought and submergence tolerance, improved pest resistance, and other important traits. This work is labour-intensive and research support personnel is the major limiting factor.
The Panel recommends that the research staffing in the Irrigated Rice Programme be reassessed with the aim of filling key positions, including an IRS agronomist with wide experience and certain skilled support staff in critical areas of work .
