Previous PageTable Of ContentsNext Page

Strategies for bridging the yield gap in rice:
a regional perspective for Asia

R.C. Chaudhary

Chairman, Participatory Rural Development Foundation, Shivpur-Shahbazganj, Gorakhpur 273014, India


Rice in the Asia and Pacific region, where 90 percent of it is produced and consumed, will remain the lifeline of the people. Demand for rice is expected to grow faster than production in most countries (Swaminathan, 1998), so much so that by 2025, 800 million tonnes of it will be needed annually. How an additional 300 million tonnes per year will be produced (Hossain, 1997) with less land, water, pesticide and labour, is the crucial question for the next millennium. The reduced resource base and the gradually eroding quality of land, water and the environment pose threats of another magnitude. The currency crisis in most Asian countries has an added dimension in importing and exporting countries through its effect on the cost of rice and investment in research or production. The yield deceleration, stagnation and decline observed in high-yielding environments is another danger signal (Papademetriou, 1998). Rice area has already started to decline in China, Malaysia, Bangladesh, the Red River delta in Viet Nam, and elsewhere (Table 1).


Asia and the Pacific rice production, yield and area (1997) and their growth rates, 1987-1997


Production (P) (`000 tonnes)

Area (A) (`000 ha)

Yield (Y) (kg/ha)

1987-1997 growth rate (%)






1 352


8 244





28 183

10 177

2 769







1 667





3 390

1 950

1 771





198 471

31 348

6 331







2 246





123 012

42 200

2 915





50 632

11 100

4 449




Iran, Islamic Rep.

2 600


4 240





12 531

1 953

6 416




Korea, Dem. People's Rep.

2 347


3 841




Korea, Rep.

7 100

1 049

6 794




Lao People's Dem. Rep.

1 414


2 902





1 970


3 008





189 000

6 070

3 064





3 711

1 511

2 455





6 546

2 316

2 827




Papua New Guinea



3 023





11 269

3 840

2 933




Sri Lanka

2 610


3 954





21 280

9 932

2 143




Viet Nam

26 397

7 021

3 760





523 784

133 696

3 918




Rest of world

49 479

16 115

3 070




World total

573 263

149 811

3 827




Source: FAO-RAP Publication 1998/21.

Approaches to bridging the gap between projected demand and the current level of production include expanding the rice area (horizontal expansion), increasing the yield (vertical expansion), bridging the yield gap and reducing yield losses. Yield gap bridging, i.e. filling the gap between the best experimental yields and those that farmers can achieve, is promising. But, because of its complexity, there are different views concerning its value as a tool for increasing rice production (Duwary, Tran and Nguyen, 1998). Pingali, Hossain and Gerpacio (1997) argue that, in favourable rice ecologies, yield gaps are not significant for exploitation or increasing rice production.

In such situations, further increase in yield is possible only through the deployment of new technologies, such as hybrid rice. The causes of the large gap between available and adopted technology have also to be understood (Fujisaka, 1994). In the meantime, pre- and post-harvest yield losses of as much as 49 percent continue to be recorded.


The narrowing of the yield gap in rice requires integrated and holistic approaches, including appropriate concepts and policy intervention (Cromwell, 1996). If any one of these components is missing or weak, the yield gap in that area cannot be narrowed (Tran, 1996). Narrowing the yield gap (Figure) aims not only at increasing rice yields and production but also at improving the efficiency of land and labour use, thereby reducing the cost of production and increasing sustainability. Exploitable yield gaps of rice are caused by various factors, including physical, biological, socio-economic and institutional constraints, which can be improved effectively through participatory and holistic approaches to action and government intervention (IRRI, 1998b). An integrated programme approach is required. The narrowing of the yield gap is not static but dynamic, adapting to technological developments in rice production. Gaps tend to increase when the yield potential of rice varieties is improved.

Benefits of yield gap bridging

While efforts are made to raise the yield ceiling, there is an even more pressing need to address the yield gap (Duwari, Tran and Nguyen, 1998). A yield gap reduction can be seen as the local solution to a global problem. It can lead to increased production with the additional incentives of cost reduction, poverty alleviation, social justice and equity. While no major breakthrough is expected immediately, reducing the yield gap alone could supply 60 percent of the increased annual rice demand by the year 2025.

Yield gaps among theoretical potential, experiment station yield,
potential farm yield and actual farm yield

Source: modified from Gomez, 1977.

Lessons from selected countries

In Indonesia, through various schemes, the national rice yield increased by about 4.9 percent per year during the 1967 to 1977 period and by 4.3 percent between 1977 and 1987. Indonesia's experience illustrates the importance of establishing strong test-transfer centres for technology (Chaudhary, 1998).

Viet Nam became the fourth largest exporter (Le, 1998), highlighting ownership policy reform - from government to family holding. Various methods and models of technology transfer have been tried since the green revolution began in India (Chaudhary et al., 1998; Taimni and Verma, 1998). Compact block frontline demonstration (FLD) was started in 1990 in a focused demonstration. Egyptian rice yield increased from 5.8 tonnes/ha in 1987 to 8.5 tonnes/ha in 1997 (Table 2), owing to the adoption of new technology which reduced yield gaps (Badawi, 1998). In Australia, the Ricecheck concept and use of a very efficient system of technology transfer reversed a yield decline of 2.4 percent per year in 1967, to the current level of 10 tonnes/ha (Lacy, 1998).

In the United States, the yield gap is small (David Mackill, personal communication) while the national yield of Greece, where there is no yield gap, reached 7.6 tonnes/ha in 1996, setting a good example for others to follow.


Egyptian rice yields in demonstration plots, national averages and yield gaps, 1988-1997 (tonnes/ha)


Demonstration plot yield

National average

Yield gap













































Source: Badawi, 1998.


Stable performing varieties

Available superior-yielding varieties can take farmers' yield to 8 tonnes/ha if grown properly (Chaudhary, 1996), although their performance is variable owing to a higher proportion of genotype X environment interaction (Gauch, 1992; Chaudhary, 1996). The genetic causes of stable performance may be difficult to understand, and there is a need to identify and release stable- yielding varieties that are appropriate to a specific area, rather than releasing relatively less stable varieties over a wider area. There are strong genotypic differences among varieties for this interaction, and methods are available for selecting the more stable ones (Gauch, 1992; McLaren and Chaudhary, 1998). The new rice plant type (called Super Rice), which has been developed by the International Rice Research Institute (IRRI), may raise the present yield potential by 25 to 30 percent (Khush, 1995). Rice biotechnology has recently made considerable progress and may also provide an opportunity to increase rice yield in a more effective and sustainable manner. Hybrid rice has been adopted in China, India, Viet Nam, Myanmar and the Philippines, and it too may reduce yield gaps.


Weeds reduce rice yield by competing for space, nutrients, light and water, by serving as hosts for pests and diseases, and as a result of allelopathy. Under farmers' conditions, yield losses, and thus gaps, are high owing to improper or untimely operations.

Biotic and abiotic stresses

Rice has been cultivated for thousands of years. As a result, it has served as a host for a number of diseases and insect pests: 54 in temperate zones and about 500 in tropical countries. Of the major diseases and pests, 45 are fungal, ten bacterial, 15 viral and 75 are insect pests and nematodes. In view of the economic losses caused by diseases and pests, efforts have been directed to understanding the genetic basis of resistance. Host-plant interaction and other control measures reduce losses in proportion to their use, and this is one way of reducing yield gaps.

Soil problems

Rice is grown from the equator to 45oS and 50oN, and from below sea level to 2 500 m above. Rice soils vary from sand to clay, acidity from 5 to 10 pH, organic matter from 1 to 50 percent and salt content from 0.1 to 1 percent. As well as on normal soils, rice is also grown on marginal and problem soils, where plants face nutrient deficiency or even toxicity. Inappropriate varieties and improper management cause heavy yield losses and resultant yield gaps. Technologies are developing fast to ameliorate such situations.

Soil fertility and fertilizer

Soil degradation and quality deterioration limit crop yields on many intensive farms in Asia. Major indicators of deteriorating soil quality include changes in organic matter and capacity to supply nutrients, nutrient imbalance and multinutrient deficiency, waterlogging and iron toxicity, soil salinity and alkalinity, and the development of hard pan at shallow depths. Many yield gaps can be attributed to a lack of knowledge (Balasubramanian et al., 1998). Rice suffers from a mismatch between its nitrogen (N) demand and the N supplied as fertilizer, resulting in a 50 to 70 percent loss of applied N fertilizer. Two basic approaches may be used to solve this problem: regulating the timing of N applications, based on the plant's needs, thus increasing the efficiency of the plant's use of the applied N; and increasing the ability of the rice system to fix its own N through nodulin genes and bacterium genes.

Water and irrigation

Most studies on constraints to high rice yield indicate the availability of water as the main cause of yield gaps and yield variability between experiment stations and farms. A recent study conducted by the International Water Management Institute (IWMI) estimates that, by the year 2020, one-third of the Asian population will face a water shortage. Wars may be fought over water (Gleick, 1993). The growth rate in the development of irrigation has already declined (Barker et al., 1998).

Integrated crop management

Based on extensive and critical testing of rice varieties and crop management technology, it is possible to develop a rice farming prescription for an individual farmer and a specific situation. The concept was tested on a limited scale in Indonesia during 1996/97. Integrated weed management practices and integrated pest management (IPM) increase yields and decrease the cost of pesticides, the cost of production and the risk to health and the environment. Other knowledge-based techniques will also help (Price and Balasubramanian, 1998). Narrowing yield gaps by improving the crop management practices of small farmers in developing countries is often not an easy task. It is therefore essential that practices should not be applied in isolation but within integrated crop management packages (ICMPs) that have flexibility to adjust to prevailing environmental, socio-economic and market factors.


The introduction of more efficient technologies for handling the drying, storage and milling of rice at the village level is essential for the reduction of post-production losses. The post-production phase is labour-intensive, as the operations involve hand harvesting, sun drying, threshing by trampling, and wind winnowing. Physical losses are higher in wet season harvests, and arise from problems with drying and the use of an-tiquated mills. This results in poor-quality milled rice, with problems that include grain discoloration. People living in communities whose livelihood is affected are expected to motivate themselves to make the changes that will ensure increased benefits. It is also believed that local farmers and entrepreneurs should be given the opportunity to define their own post-production needs and should be consulted during the selection of appropriate technologies.


Risk, cost and return

Rice is a very risk-prone crop in ecosystems that are rainfed, upland or flood-prone. Even in irrigated ecosystems, rice is at risk from pests, diseases and flood. These dangers are a problem for farmers, particularly when they are applying costly inputs such as seed and fertilizer. Economists, as well as farmers, advocate diversification and maximization of income from sources that are not closely connected to rice farming. Farmers are confident about using inputs and enthusiastic about reducing yield gaps only when stress-resistant rice varieties are readily available.


The difficulty in obtaining sufficient, timely credit for the development of infrastructure, capital costs and crop loans is a major constraint. As well as the lack of collateral, the interest rate is a threat to credit availability, and this affects farmers' input-applying capacity and profit margins. Government policy and the institutional framework affect credit availability to farmers (Cromwell, 1996). A policy intervention must be implemented in order to make farmers creditworthy and enable them to acquire enough timely credit to buy annual inputs. Crop insurance is a suitable policy support in India.

Tradition and attitude

The current level of factor productivity, which is still declining, cannot satisfy the food requirements of a growing population. Productivity must be increased in a sustainable manner. The major causes of instability in productivity are weather aberrations, crop management practices and the severity of pest outbreaks. Although varieties that are tolerant to environmental stresses, more stable and more resistant to pests have helped to stabilize production, there is still room for improvement in these aspects. The main issue in establishing stable productivity is to ensure that there is no degradation of the resource base in terms of soil, water and the biodiversity of rice types.

Input availability

Fertilizers, especially N, play an important role in rice production and productivity. Farmers need adequate amounts of fertilizer at the right time if they are to obtain high yields. The supply of fertilizers needs to be decentralized to village markets, and the quality of fertilizers should be assured. Small farmers are usually unable to buy sufficient quantities in time for application; the provision of village credit could greatly help. The Bangladesh Grameen Bank is an interesting example of rural credit provision to landless and resource-poor farmers. Loan proposals are received by the bank on a group basis only (involving at least five persons) and focused on technology loans, housing loans, joint loans and general loans (Dadhich, 1995). The use of quality seed is the most effective way of realizing the yield potential of the recommended technology. High-quality pure seed ensures correct germination, good crop stand and freedom from weeds and seed-borne diseases. In general, it is recognized that the use of high-quality seed ensures yields that are 10 to 15 percent greater than they would have been under the same crop management practices.


The adoption rate of knowledge-intensive technology is affected by the availability of agricultural credit, the supply of inputs (seeds, fertilizers, pesticides), the availability and quality of contract services and machinery for different farm operations, and repair and maintenance services in rural areas (Price and Balasubramanian, 1998). The government and private institutions associated with credit, inputs and pricing have a direct influence on the adoption and level of use, and thereby on the yield level.


Centres of the Consultative Group on International Agricultural Research (CGIAR), such as IRRI, the International Centre for Tropical Agriculture (CIAT) and the West Africa Rice Development Association (WARDA), are starting to listen to farmers (IRRI, 1996) and plan farmer-relevant research with farmers' cooperation. Future projects could involve collaborating institutions, the training of local researchers on the basics of yield gap bridging, strategic planning with local counterparts, the development of operational plans, the introduction of selected technologies, and extension success cases. A model framework for such projects is urgently needed.

As a first step, rapid appraisal studies are undertaken, when such information does not already exist. Projects have the following components:

Survey, analysis and prescription

Yield gap analysis must take into consideration the uniqueness of the individual situation. Another interesting area of inquiry would be a study of the relative magnitude of yield gaps in different yield-potential groups of rice, say within tolerant varieties (TVs), high-yielding varieties (HYVs), hybrid rice and new plant type (NPT) rice. The current hypothesis is that yield gaps should be lowest with hybrid rice, as the seed cost is high and farmers take relatively better care of crops, including fertilizer application, weeding and irrigation. This is where the methodology proposed by De Datta et al. (1978) needs to be examined critically and modified to satisfy the following criteria:

Technology resourcing

Technology resourcing has to be carried out on an area- and situation-specific basis to support prescription farming and comprises:

Farmer outreach: models of technology transfer

Communication issues are significant to development agencies when technologies are transferred to farmers. Methods for communicating knowledge-intensive technology to farmers must be found. The aim is to increase the impact of appropriate high-potential rice production and utilization interventions, and promote their adoption through development and the application of improved dissemination strategies. The following are some of the institutional arrangements that have been devised.

Government organizations (GOs). The positive aspects of government extension services lie in early successes with simple seed-based technologies, promotion of State objectives, and the fact that they are relatively well resourced and staffed. Among their negative points are staff's responsibility for non-extension activities, slow bureaucratic responses, weak links with research, a tendency to favour more advantaged farmers, and a top-down focus.

Non-governmental organizations (NGOs). The positive aspects of NGOs are that they are strongly decentralized, less bureaucratic and more flexible, people- and poor-centred, capable of articulating the needs of disadvantaged groups, community-focused, effective in community mobilization and motivation, and active where the State and market are absent. Their negative points are their restricted impact, distance from policy decisions, professional and technical limitations, and limited infrastructure, coordination, accountability and resources.

Private organizations (POs). The positive aspects of private-sector organizations are that they have strong, centralized systems and good resources, management skills and technical competence; they are less bureaucratic, more flexible and innovative; and they have clear objectives. There are, however, also a number of significant negative points which focus on POs' putting profits before development, their manipulation of markets and adoption of questionable technologies, and their tendency to favour large and resource-advantaged farmers.

Participatory (GO-NGO-PO). GOs, NGOs, POs and farmers may collaborate in, for example, the resource-test-demonstrate-adopt model. When there are appropriate roles and decision-making processes, this may be the most effective, broad-based and sustainable solution. Recently, IRRI has implemented a number of such proposals (IRRI 1998a; 1998b). Partnerships among NGOs, POs and GOs are likely to be the agents of change for the future, although the most appropriate combination for a given situation will vary depending on the existing circumstances and technology.

The general trend away from the traditional GO approach towards new agents of change (POs and NGOs) is recognized, but partnership is itself an experiment in development and technology transfer. Several issues arise: Are these new agents of change more effective in promoting technology transfer? In what kind of circumstances and for what kinds of technologies are the new agents most effective? and What are the most appropriate roles, responsibilities and linkage mechanisms?

Another set of questions must be taken into account when comparative advantages and maximization of resources are being assessed in order to achieve common objectives: Is the partnership feasible? What can be done to encourage partnership formation? In what kinds of circumstances and for what kinds of technologies is partnership appropriate? and What are the most appropriate roles, responsibilities and linkage mechanisms within such partnerships?

Next, weaknesses within the system need to be addressed and strengths capitalized on, and the relative strengths of the agencies pulled together, be they GO-NGO-PO partnerships, research-NGO-PO partnerships, hybrid diffusion systems, or unitized new tools.

Farmer-to-farmer. The experiences of successful farmers illustrate the relative importance of biophysical and socio-economic constraints. The hypothesis in this case is that farmers with small yield gaps will have used sets of technologies and knowledge that are adapted to their locations within a dynamic decision-making process that is different from those followed by farmers with high yield gaps. Tapping the knowledge of successful farmers is considered an effective way of resourcing and spreading location-specific, improved technologies. Appropriate national agencies must learn the gamut of successful technologies, knowledge and experience in order to transfer them to other farmers.

Policy support

Government policies provide the environment for research investment, improve productivity, alleviate poverty, ensure the sustainability of systems, protect the environment and provide food security. There is a need for various types of policy interventions that influence investments in research, ensure the availability of inputs and credit, provide crop insurance policy to encourage the use of inputs, and reduce risks. It is therefore imperative that appropriate policies and socio-economic adjustments be affected in terms of input-output pricing, institutional support and addressing the needs of rice farmers so as to complement technological gains. There are examples of success, such as that from Indonesia; when firm and consistent policy support is provided, the impact on bridging yield gaps becomes noticeable. Experience in this area also emphasizes the need for consistency.


Linkages within and outside the country are important aspects of the project for which the following list may be considered:


Stratégies visant à réduire les écarts de rendements du riz:
perspectives régionales pour l'Asie

On enregistre d'importants écarts de rendements, de 10 à 60 pour cent, entre les diverses zones rizicoles de l'Asie et du Pacifique. Différents éléments sont imputables selon les pays, et notamment des facteurs biophysiques (conditions météorologiques, sols, eau, incidence des ravageurs et des plantes adventices); techniques (tallage, variétés/semences choisies, eau, éléments fertilisants, plantes adventices, ravageurs et gestion des récoltes); socioéconomiques (situation économique, traditions et connaissances des agriculteurs, taille des familles et revenus des ménages); et institutionnels/politiques (action gouvernementale, prix du riz, crédits, fourniture d'intrants, régimes fonciers et marchés). Les limitations liées aux transferts de technologies concernent la compétence et l'équipement du personnel de vulgarisation; l'intégration de la recherche, du développement et de la vulgarisation; les préjugés des agriculteurs, les connaissances et les qualifications; les contacts réduits entre le secteur public et le secteur privé; les ONG et le personnel de vulgarisation. Le modèle conventionnel d'écarts de rendements a été modifié pour dégager l'écart de la recherche.

La réduction des écarts de rendements dans le secteur du riz nécessite des efforts concertés de toutes les parties concernées à la fois aux niveaux national et international (gouvernements, ONG, organismes privés, centres internationaux de recherche agronomique et organismes du système des Nations Unies). L'analyse des contraintes a fait l'objet de nombreuses études contrairement à la réduction des écarts de rendements. Cependant, les technologies et les variétés spécifiques à une zone, comme les méthodes de gestion intégrée des cultures «Rice Check System» utilisées en Australie, devraient être adaptées dans divers pays en «Tropical Rice Check». Des activités intensives de transfert des technologies devraient être utilisées, en appliquant, pour favoriser les technologies visant l'accroissement des rendements, des modèles ayant donné de bons résultats, comme les expériences de première ligne. Des caractères de stabilisation des rendements devraient être incorporés dans les variétés en utilisant à la fois des méthodes traditionnelles et innovatrices. Les raisons génétiques de la stabilité des résultats sont parfois difficiles à comprendre, mais des variétés à rendements plus stables et plus élevés sont nécessaires même pour des zones spécifiques. Il faut comprendre les limitations des écosystèmes les plus productifs pour pouvoir étudier les questions du déclin de la productivité des sols et de la mauvaise utilisation ou du déséquilibre de l'utilisation des éléments fertilisants, l'accroissement du rendement de l'eau, l'adaptation des diverses variétés, le recul des profits et le soutien inadapté fourni aux agriculteurs par la recherche et la vulgarisation. Un projet régional de réduction des écarts de rendements devrait être entrepris en procédant à des levés de terrain, au renouvellement des technologies, et en favorisant des accords institutionnels (liaisons entre les organisations internationales, les gouvernements, les ONG et les organismes privés), la communication et la recherche en vue de résoudre les blocages technologiques.

Estrategias para colmar las disparidades de rendimiento del arroz:
perspectiva regional para Asia

Las ecologías arroceras de la región de Asia y el Pacífico presentan notables disparidades de rendimiento, que van del 10 al 60 por ciento. Entre los diversos factores que normalmente contribuyen a esas disparidades de rendimiento que se dan en los distintos países figuran factores biofísicos (tiempo, suelos, agua, presión de las plagas, malas hierbas); factores técnicos (cultivo, selección de variedades y semillas, agua, nutrientes, malas hierbas, plagas y ordenación postcosecha), factores socioeconómicos (situación económica, tradiciones y conocimientos de los agricultores, tamaño de la familia, ingresos familiares), factores institucionales y políticos (política gubernamental, precio del arroz, crédito, suministro de insumos, tenencia de tierras, comercio). Entre las limitaciones de transferencia de tecnología figuran la competencia y el equipo de extensionistas, la insuficiente integración de la investigación, el desarrollo y la investigación, los «bloques cognoscitivos», los fragmentarios conocimientos y pericia de los agricultores, unas conexiones flojas entre el personal público, el privado, el de las ONG y los extensionistas. Se alteró el modelo tradicional sobre disparidades de rendimiento para esclarecer el déficit de investigación.

Para reducir el déficit de rendimiento del arroz hacen falta esfuerzos concertados de todas las partes interesadas tanto de ámbito nacional como internacional (gobierno, ONG, organismos privados, organismos de las Naciones Unidas y centros internacionales de investigación agrícola). Existen abundantes estudios sobre análisis de limitaciones pero son muy pocos los que tratan sobre cómo colmar esas disparidades de rendimiento. No obstante, en varios países habrán de desarrollarse como controles del arroz tropical las variedades y tecnologías propias de cada lugar, es decir, un sistema integrado de ordenación de cultivos semejante al sistema de control del arroz empleado en Australia. Hace falta intensificar las actividades de transferencia de tecnologías empleando modelos satisfactorios como demostraciones de vanguardia para fomentar tecnologías que potencien los rendimientos. Habrá que incorporar a las variedades mediante sistemas tanto tradicionales como innovadores rasgos estabilizadores del rendimiento. Aunque tal vez resulte difícil comprender las razones genéticas de estabilidad del rendimiento, sin embargo hacen falta variedades de rendimiento superiores y estables, incluso en zonas específicas. Hace falta comprender las limitaciones de los ecosistemas más productivos para contrarrestar el descenso de la productividad de los suelos y la utilización inadecuada o desequilibrada de nutrientes; aumentar la eficiencia del empleo del agua; conseguir la adecuación de las variedades, y hacer frente a la reducción de beneficios, a la falta de investigación y de apoyo otorgado a los agricultores. Habrá que lanzar un proyecto regional que colme las disparidades de rendimiento y que contenga elementos de reconocimiento y prescripción, provisión de recursos para tecnologías, arreglos institucionales (conexiones entre organizaciones internacionales, gobiernos, ONG y organismos privados), comunicación e investigación específica para resolver tecnologías de atasco.

Previous PageTop Of PageNext Page