The System of Rice Intensification (SRI) was developed as a set of insights and practices that beneficially change the management of the plants, soil, water and nutrients that are used for growing irrigated rice crops. These concepts and practices have been adapted also for growing rice that is not irrigated and for growing a number of other crops. So SRI is no longer just for growing irrigated rice. As discussed below, it is not understood as a usual kind of technology.

SRI methods, by promoting the growth of more productive and more robust plants, can give farmers higher yield (more kg of rice per hectare, or more tons per acre) with use of less seed and less water because the number of rice plants per m2 is greatly reduced, and paddy fields are not kept continuously flooded. SRI methods do not require farmers to purchase external inputs or new seeds, since chemical fertilizer and agrochemical protection are not necessary (although they can be used with the other SRI practices).

Practically all rice varieties have given higher yield with these methods, although some varieties respond better than others. 'Improved' varieties usually give the highest yields with SRI management; but yields of 'unimproved' varieties can also be greatly improved under SRI. After a summary discussion of this general question, more information is provided in response to a number of more specific questions that seek details.

• SRI methods are particularly accessible to and beneficial for the poor, who need to get the maximum benefit from their limited land, labor, water, and capital. However, SRI concepts and practices can be adapted and used with any scale of production, from small-scale to large-scale, and are amenable to considerable mechanization. So the methods can have broad applicability.
• In an unprecedented way, SRI methods raise the productivity of land, of labor, of water and of capital -- all at the same time. Because SRI's higher productivity can make more rice available in markets, it is hoped that over time the methods will lower food prices and will have widely distributed benefits. Everyone benefits if a country's need for staple food can be met with less of its land, labor, water and capital resources, which can then be devoted to meeting other needs.
• There is no secret and no magic with SRI. Its results are and must be explainable with solid and scientifically-validated knowledge.

From what we know so far, the management practices that are recommended under the rubric of SRI succeed in large part because they promote:

• Better growth and health of rice plant roots -- so that they grow larger and deeper and to not degenerate for lack of oxygen in the soil when rice fields are kept flooded, and
• The abundance, diversity and activity of beneficial soil organisms -- bacteria, fungi, earthworms and other soil biota -- that improve soil fertility and contribute to the growth and health of rice plants.

In practice, SRI involves some combination of the following simple changes in rice cultivation practices. These practices are reviewed in section 1.1 below, and the reasons why they are recommended are discussed in response to FAQ #6:

• Transplant seedlings at a very young age if establishing the rice crop by transplanting -- 8 to 12 days old, at most 15 days old -- instead of the usual seedling age of 3-4 weeks or more.
• Raise seedlings in unflooded nurseries -- not planted densely and well-supplied with organic matter. It is possible to establish an SRI crop with direct seeding, but transplanting is most common method for crop establishment.
• Transplant seedlings quickly, carefully and shallow -- taking care to have minimum trauma to roots, not inverting plant root tips upward, which delays resumption of growth.
• Transplant seedlings at wider distance and singly -- rather than in clumps of 3-4 plants -- and in a square pattern, usually 25x25 cm, giving roots and leaves more space to grow.
• No continuous flooding of the soil – since inundating the soil causes roots to degenerate and suppresses soil organisms that require oxygen. Either either apply small amounts of water daily to keep the soil moist but not saturated; or alternately flood and dry the soil a few days at a time.
• Use a simple mechanical hand weeder for weed control -- this aerates soil at the same time that it eliminates weeds, leading to better yield than with either hand-weeding or herbicides.
• Provide as much organic matter as possible to the soil – while chemical fertilizer gives positive results with SRI practices, the best yields have come with organic fertilization. This does more than just feed the plant -- it feeds the soil, so that the soil can feed the plant.

SRI methods were assembled in the early 1980s, as explained under FAQ #2 below. Only after 2000 did SRI begin to receive any attention around the world, and it was not very visible until after 2002. Thus, SRI has been on the world scene for only about a decade.

During this time, SRI methods have been validated in over 50 countries across a wide range of agroecosystems -- from equatorial to temperate climates, and from sea level to 2,700 meters asl. The number of countries where SRI merits have been demonstrated, and where it is expanding, is increasing year by year. Recent additions to 'the SRI club' include countries as diverse as Benin, Colombia, the Democratic People's Republic of Korea, the Dominican Republic, Taiwan, and Tanzania.

• The most rapid spread of SRI has been in Vietnam. Fewer than 10,000 Vietnamese farmers were using the methods in 2007 when the Ministry of Agriculture and Rural Development there issued a decree declaring SRI to be a 'technical advance.' Four years later, the Ministry of Agriculture reported that over 1 million farmers were using the alternative methods (http://qdnd.vn/qdndsite/en-US/75/72/182/156/189/164012/Default.aspx), and the number of farmers using SRI methods reached 1.3 million in 2012.
• The Indian state of Tamil Nadu now reports having over 1 million farmers using SRI methods there on 850,000 hectares. This spread has been supported by faculty at Tamil Nadu Agricultural University under the World Bank-funded IAMWARM project, with the state's extension service also promoting SRI utilization.
• In the Indian state of Bihar, where SRI was introduced in 2007 with 128 farmers on 30 hectares, the Ministry of Agriculture reported that SRI use had reached 335,000 hectares, with an average yield of 8.08 tons/ha, three times the usual paddy yields in Bihar.

SRI is considered as a methodology rather than as a technology because it consists of concepts and practices to be adapted by farmers to their local conditions to get best results. SRI is not a fixed set of things that farmers 'must' do. Using SRI methods requires no material inputs beyond what farmers already have; mostly it requires just a change in thinking and changes in practices.

• Although SRI is commonly referred to as a thing (a noun), the term is better used as a description (an adjective). SRI refers to the use of a number of practices that reflect key insights and principles for providing an optimum growing environment for rice plants.
• All plants, indeed all organisms, are the outcomes of interactive processes in which an initial genetic potential (genotype), through ongoing interaction with its environment, produces a unique creature or phenomenon (phenotype). This explains why even identical twins are never identical persons and personalities.
• By altering plants' growing environments, SRI practices create -- from any rice genotype -- specific phenotypes that are more productive. As noted above, there is no magic in SRI. All of the effects of SRI practices can be explained with well-established scientific knowledge, usually referring to very simple and elementary relationships, like not crowding plants together so that their root growth is not inhibited. SRI has been correctly called "just good agronomy." This is not a criticism or refutation of SRI, but rather a compliment.
• Because some varieties (genotypes) respond to SRI practices better than others do, it is clear that genetic potentials are important. Note that many traditional or 'unimproved' varieties give very good responses to SRI methods (6-10 tons/ha), and it can be more profitable for farmers to grow them because their market prices are high, being preferred by consumers. 'Modern' varieties are not necessary for farmers to cultivate rice productively and profitably.
• Because SRI results depend on the unfolding of biological processes and potentials, the results with SRI can be quite variable, more variable than if results were due to a fixed genetic 'blueprint,' or if crop outputs are primarily due to, and thus proportional to, external inputs such as fertilizer. This variability may be disconcerting to some persons, but it is this plasticity that offers great opportunities to farmers who can, through better management, learn to capitalize on these potentials.

SRI was developed in Madagascar through half-a-lifetime of effort of Fr. Henri de Laulanié, S.J. He spent the last 34 years of his life working with poor farmers in that country, to help them reduce their poverty and hunger through improved production of rice, the source of more than half of the calories that are consumed daily by typical Malagasies. He sought to rely on simple methods that would not require farmers to purchase external inputs, which were not affordable.

Henri Laulanié was born in France in 1920 and attended its leading agricultural college before the start of World War II. By the time he graduated in 1939, he decided to change careers, and he entered a Jesuit seminary in 1941. After graduation in 1945, he worked in France for 16 years, among other things teaching at the agricultural college at Angers. In 1961, he was sent by the Jesuit order to Madagascar as an agricultural missionary. Although he knew little about rice when he arrived, he was trained in agriculture in general, so he decided to focus on this crop.

Over the next two decades, he observed and experimented with various practices. Two SRI practices he learned from farmers who had made departures from traditional cultivation methods. A few farmers were transplanting single seedlings instead of clumps of 3-6 seedlings, and some other farmers did not keep rice fields continuously flooded -- only moist enough to meet their crops' needs. Laulanié himself took up the use of a rotating hoe that aerates the topsoil at the same time that it eliminates weeds. Weeds can become a big problem when farmers do not keep their rice fields always flooded.

As an original contribution to SRI, Laulanié introduced planting in a square pattern, 25x25cm (10x10 inches), so that he could use the mechanical weeder in perpendicular directions. This radically reduced the rice plant population, by 80-90%, giving all the plants ample room for the growth of their roots and above-ground parts, being better exposed to sunlight and air. Planting in a square pattern be permitting mechanical weeding in perpendicular directions enhanced both soil aeration and plant growth, while at the same time reducing pest and disease problems.

The biggest single step toward the development of SRI was the serendipitous discovery in 1983 that transplanting very young seedlings, just 15 days after rice seeds had been sown in a nursery, could greatly enhance yield (Laulanié, 1993, 2011). By using young seedlings, the plants' potential for prolific growth of roots and tillers is preserved, explained by an understanding of phyllochrons (FAQ #15).

SRI was developed using chemical fertilizer, but when the government removed its fertilizer subsidy in the late 1980s -- and small farmers could no longer afford to buy it -- Laulanié modified SRI practice to utilize compost. This proved even more beneficial for plant growth.

In 1990, together with several close Malagasy friends and colleagues, Laulanié formed a local NGO called Association Tefy Saina. The Malagasy name means 'improve the mind' rather than 'grow more rice' (1.2 above). The mission of Associastion Tefy Saina was to promote broad-based agricultural and rural development in Madagascar (Laulanié 2003).

In 1994, Tefy Saina began working with the Cornell International Institute for Food, Agriculture and Development (CIIFAD) on an integrated conservation and development project funded by USAID in and around Ranomafana National Park, intended to protect endangered rainforest ecosystems in the country's central-eastern escarpment.

Over the next three cropping seasons, farmers trained by Tefy Saina field staff achieved average yields of 8 tons/hectare, where previously they had averaged only 2 tons/hectare. Some farmers reached yields of 10, 12, even 14 tons, without changing varieties or relying on chemical fertilizer. In 1997, CIIFAD began trying to get colleagues in other countries to try out SRI methods and evaluated them for themselves.

By this time, sadly, Fr. Laulanié had died, in June 1995 at age 75, not knowing how successful his innovation could and would be. It fell to Tefy Saina and CIIFAD to carry on his work, building on his insights and trying to share more widely the opportunities that his lifetime of selfless, creative work had brought into existence. Because SRI was not promoted as a fixed technology, many farmers and other individuals as well as organizations around the world took 'ownership' of the innovation, which was extrapolated also to many other crops beyond rice.

SRI was purposefully developed to benefit households that are poor, resource-limited, and food-insecure -- ones needing to get the most productivity possible from the small amount of land that they have access to, and from their available household labor, if possible using less water, and without having to spend money to buy inputs (new seeds, fertilizer, agrochemicals) or to take loans to buy inputs that will push them (further) into debt.

By raising the productivity of the land, labor, water and capital invested in the production of rice without need to purchase certain inputs, SRI is unique among contemporary agricultural innovations. Poor households can take up SRI simply by changing their thinking and modifying familiar practices. SRI does not present the kind of barriers to adoption that have kept Green Revolution technologies from benefiting many of the world's poor households. The number of people in rice-producing areas of Asia, Africa and Latin America who are afflicted with chronic hunger has been previously estimated at 400 million (Surridge 2004), but this number has probably increased in recent years.

How SRI can improve the lives of poor households has been vividly seen in Cambodia:

• In 2002-03, the NGO ADRA persuaded 100 farmers in a village near Siem Riep, whose paddy yields averaged just 1 ton/ha, to try SRI. ADRA said it would compensate anyone whose yield fell below this average. According to Roland Bunch of World Neighbors, with SRI methods these farmers averaged 2.5 tons/ha -- and not a single farmer asked for any compensation (http://ciifad.cornell.edu/sri/countries/cambodia/cambadrepmay03.pdf).
• In 2006-07, a Family Food Production project of the NGO LDS Charities got 146 rainfed farmers in Kampong Chhnang Province who had been averaging 1.06 ton/ha to use SRI methods. With SRI practices, average yield was 4.02 tons/ha, and all exceeded their previous yield, with lower costs (Lyman et al. 2007). Such increases transform the life chances of the poor (http://ciifad.cornell.edu/sri/countries/cambodia/camldsrpt07.pdf).

In India, when the NGO PRADAN introduced SRI into the poverty-stricken district of Purulia in West Bengal state in 2003, just 4 farmers were willing to try the new methods. The next year, 150 farmers practiced SRI, and almost 4,000 by 2007.

• An evaluation team from the India Programme of the International Water Management Institute (IWMI) assessed SRI use and impacts in two villages in 2004, one of which had been hit by severe drought. With SRI, average increase in yield was 32% (50% in the village with normal rainfall; 11% in the drought-stricken one). Net income per hectare went up by 67% on average, with 8% less labor needed per hectare. One farmer had a yield of 15 tons/ha, as measured and weighed personally by the team leader. The IWMI report characterized SRI as a 'pro-poor' innovation (Sinha and Talati 2007).
• The Hindu has reported (Nov. 28, 2011) that by adopting SRI methods, whole villages in northern Madhya Pradesh state, with mostly poor tribal populations, are able to quadruple paddy yield without having to purchase any inputs, going from 2 ton/ha to 8.5 tons/ha with indigenous varieties (http://www.thehindu.com/sci-tech/agriculture/article2668286.ece).
• In northern Myanmar, the NGO Metta Development Foundation began introducing SRI in 2001 through farmer field schools (FFS) for mostly tribal farmers cultivating rainfed rice. Over four years, Metta trained 5,000 men and women, providing hands-on training which lasted one season. By the end of 2005, through FFS spread, about 20,000 households were using SRI methods.

On 1-acre (0.4 ha) FFS plots (N=30), SRI yields averaged 6.5 tons/ha, compared to farmers' usual yields of 2 tons/ha. On farmers' own fields after their training, yields averaged over 4 tons/ha even without their using all of the SRI methods. Because conventional rice growing has been little more than a break-even operation and farmers' costs of production with SRI did not increase, households' net income per hectare from rice went up by more than eight times -- from 296 kg/ha net production of rice to 2,585 kg/ha (Kabir and Uphoff 2007). The number of SRI users in the region is now estimated at 50,000.

SRI may be one of the few innovations that offers relatively greater benefits to poorer farmers. Such cultivators usually have access only to 'poorer' soils than those cultivated by richer farmers. A study evaluating 70 comparisons previously made in 15 countries found that yield increases with SRI methods were, on average, higher on low-fertility soils than on medium-fertility or high-fertility soils, giving farmers both absolute and relative gains (Turmel et al. 2011).

• This could be attributed, at least in part, to SRI's alternate wetting and drying of paddy soils as this can mobilize phosphorus from the soil's 'unavailable' reserves (Turner and Haygarth 2001). Soils with low fertility commonly have severe phosphorus limitations. That SRI yield gains are achieved in part through the methods' mobilization of otherwise unavailable P in soil that has little available P was confirmed by trials in Panama that compared the effects of SRI vs. conventional rice management (Turmel 2011).

SRI was developed to benefit smaller, poorer farmers, to counter their problems of poverty, hunger, and environmental degradation. It was not conceived of or presented as being in competition with what rice scientists commonly call 'best management practices' (BMPs). These are heavily dependent on the use of purchased inputs, so in practical terms they are generally beyond the reach of those rural households most needing greater productivity and income.

However, when SRI practices are used to best effect, their results can match or even surpass those associated with 'Green Revolution' technologies. This has ignited considerable controversy, as some rice scientists have defensively insisted that however beneficial SRI might be for poorer farmers, the results of using SRI's alternative practices are still inferior to what can be achieved by using the results of 'modern' plant breeding and using inorganic soil amendments.

Critics have been particularly dismissive of reports of occasional 'super-yields' with SRI methods. These are said to be above what some rice scientists have calculated as 'the biological maximum' that is attainable from current rice genetic potentials (Dobermann 2004; Sheehy et al. 2004; Sinclair and Cassman 2004).

Persons who have worked with SRI in the field and with farmers who have used the methods (which critics have not) are, however, satisfied that yields over 15 tons per hectare have been and can be attained with SRI management. But they do not attach much importance to super-yields. What count most, in their view, are three things: (a) differences in average yields; (b) comparisons with what farmers are producing with current methods; and (c) what can be achieved with methods that are neither costly for farmers nor harmful for the environment. In addition, they attach importance to (d) resistance to biotic and abiotic stresses because this gives farmers greater assurance of food security.

An article which claimed that SRI did not produce better results than 'best management practices' (McDonald et al. 2006) contained enough methodological and empirical flaws that it should not have been passed through peer review (these disqualifications are discussed in Uphoff et al. 2007). More important, the argument that SRI gives lower yields than BMP has been contradicted by a more extensive study employing a similar design with a more defensible data base (Turmel et al. 2011). The clearest contradiction is the new world record for paddy yield set by an SRI farmer in Nalanda district of Bihar state of India in the 2011 kharif season.

• The yield from Hemant Kumar's one-acre SRI plot was measured by government personnel who used standard estimation techniques: harvesting an area 10 meters x 5 meters in the middle of the field, threshing the grain collected from this area, and weighing it. This process was witnessed by hundreds of observers because everyone knew that this was the most productive paddy field ever seen.
• There is no basis for questioning how much grain was produced from this 50m2 area. The paddy yield was calculated to be 22.4 tons/ha, with a dried weight of 20.16 tons/ha. This was well above the previous world record of 19.2 tons/ha in China. The yield was at first denied by Indian rice scientists; but when the methods used were validated, the record was affirmed by the Indian Council for Agricultural Research (ICAR) and Ministry of Agriculture. See: http://www.thehindubusinessline.com/industry-and-economy/agri-biz/article 3016481.ece
• Four other farmers in the same village (Darveshpura) got similar super-yields that season of 19 or 20 tons/ha, so Hemant Kumar's yield was no chance occurrence. Details from an investigation by the government's Directorate of Rice Development and by PRADAN, the NGO which introduced SRI into Bihar state, have been published (Diwakar et al. 2012). These record yields were achieved with hybrid varieties and with 'integrated nutrient management,' applying some inorganic fertilizers (40 kg/ha of urea (N), 80 kg/ha DAP, 40 kg/ha potash) that complemented the larger amount of organic fertilization given: 6 tons/ha farmyard manure at time of land preparation, plus poultry manure (400 kg/ha), vermicompost (100 kg/ha), and an application including P-solubilizing bacteria (40 kg/ha).
• Most significant about the five farmers' SRI results was that their yields that season from the same hybrid varieties -- when using conventional crop management (older seedlings, closer spacing, flooding of the fields), and inorganic fertilization without the organic nutrients listed above, on the same soils and with the same climate -- were only 7 tons/ha. While this yield was about three times higher than the usual paddy yields in the area, but it was only about one-third as much as with SRI management.

The reasons why SRI methods produce different, more productive, more robust rice plants (better phenotypes) from a given variety (genotype) are still not fully understood. But earlier claims that SRI management methods do not improve rice plants' growth and yield through synergistic interactions (Dobermann 2004; Sheehy et al. 2004; McDonald et al. 2006) are no longer tenable (Thakur et al., 2010).

Some critics have dismissed SRI as 'just best management practices,' but this not a meaningful objection. Indeed, SRI methods are the 'best management practices,' but they are also more productive than the methods previously recommended by national and international rice scientists, and they have more theoretical justification. Previous work on BMPs has paid little attention to the growth and health of plant roots or to the abundance, diversity and activity of beneficial soil organisms. These are key factors which SRI management promotes, and they make SRI more than a just a pragmatic innovation.

What have been considered 'best management practices' previously assumed that to raise their yields, farmers need further improvements in crop genetic potentials (new varieties) and need to apply agrochemical inputs (fertilizers, pesticides, etc.). We recognize that having better genetic potentials can give some advantages for farmers, certainly more options; and to some extent and in some places, the application of certain agrochemical inputs can be beneficial for crops. But creating better growing environments for rice crops -- both above and below ground -- can have powerful effects for productivity, as seen with SRI experience. There should a burden of proof for use of inorganic inputs, that they will not impair naturally-occurring biological processes and potentials and will not have adverse long-term effects on soil health and productivity.

SRI does not require farmers to make any purchases or to take out loans, although SRI's impact is significantly enhanced by having access to and using inexpensive weeding implements that aerate the soil at the same time that they control weeds. This is the only equipment needed for SRI, and it can be substituted for by hand weeding, although this is less effective.

• Water control: The main requirement for SRI success is being able to apply small amounts of water regularly and reliably, or to be able to flood the field and then drain it after a few days, with assured water supply to re-flood it after a few more days.
  • In some places, farmers' access to water is too unreliable for them to be willing to try to operate with 'a minimum of water,' as recommended by Fr. Laulanié.
  • Or farmers may be cultivating rice where there is continuous inundation of fields from rainfall, such as in monsoon climates with poor drainage of fields.
  • Or they may have low-lying fields and heavy clay soils that make it difficult or impossible to evacuate water, and the fields are perpetually saturated.

  When the soil is saturated (anaerobic or hypoxic), the benefits from SRI practices will be reduced or even negated.

  The most favorable circumstance for practicing SRI is where fields are irrigated with water that is pumped from a groundwater source or from a river or reservoir. This gives farmers both:

  • Means to control and limit water, turning the pump on or off; and
  • Incentive to do this, saving money on pump operation, provided that the electricity or diesel supply for powering the pump is reliable.

  Where water control is difficult, such as in the middle of a large irrigation system where other farmers are practicing continuous flooding, digging drainage channels within the field and making raised beds can buffer the SRI plants from having too much water in their root zones. While precise water control is ideal for SRI, farmers can adapt their soil, plant and water management practices to benefit from its methods even under less-than-ideal water management conditions.

• Labor: A second requirement is that the farmers have enough labor and time to invest more of both while they are learning SRI methods. Any new practice requires some investment of time and effort for learning. Farmers starting SRI, especially those whose rice-growing practices have been more extensive than intensive, will find that they need 20-30% more labor per hectare at first. On the other hand, where farmers have been growing their rice intensively to get maximum yield from available land, SRI is usually labor-saving even in the first year, because there are so many fewer plants to raise, transplant and manage.

  Mechanical hand weeding may require more labor at first, but farmers report that they can offset this with labor saved from stopping chemical spraying, and there is considerable cost reduction. With practice they learn to do mechanical weeding more quickly and easily. One evaluation of women's labor in SRI found that the mechanical weeder reduced labor time by 72% (Mrulini and Ganesh 2008).

  As discussed next, more labor is required for making and applying compost compared to buying and spreading chemical fertilizer. But the amount of time required for this depends very much on availability of biomass, and there are usually large cost differentials.

  Farmers starting out with SRI need to be prepared to invest more labor initially. This can be a barrier to adoption for poor households which lead a hand-to-mouth existence, not being able to afford such an investment even if it is (and they know it is) more profitable for them to use SRI practices (Moser and Barrett 2003). For many farmers in countries like India and China, SRI labor-saving has become one of the main attractions of its methods.

• Biomass: A third requirement is supply of biomass if farmers want to rely heavily or entirely on organic sources for soil fertility enhancement, rather than continuing to use synthetic fertilizer from a bag. The compost that can substitute for fertilizer can be made from any available biomass (rice straw, weeds, loppings, possibly manure, etc.); but this is sometimes in short supply, especially for farmers whose rice fields are large so large amounts are needed.

  SRI does not require organic fertilization; but its yields are best when soil fertility and functioning are supported by organic inputs to the soil system. It is very desirable to have appropriate tools and implements for collecting, transporting, shredding, processing and applying decomposed biomass (compost). Many of these are currently of unimproved and very old design and are not very efficient means for managing and applying biomass in terms of their labor requirements.

  Making SRI more broadly accessible and productive will depend very much on developing better tools and implements to raise labor productivity in these processes, and on ways to grow or access more ample and convenient stocks of biomass. Little research and development have gone into tools and biomass supply to have efficient and reasonably attractive means for organic enhancement of soil fertility.

• Crop protection: Where both plants and grain production are increased, there is always a potential need for more or better crop protection measures to protect against pests and diseases. In general, farmers' reports and evaluations support the observation that with SRI crops, the incidence of and damage from pests and diseases is less (see data from Vietnam under FAQ #4). However, SRI farmers always need to be prepared to deal with pest and disease hazards.

  Since SRI is not necessarily 'organic' in its practice, chemical protection is an available and acceptable means, except for farmers who are committed for personal or economic reasons to organic production. SRI farmers in many countries use organic insecticides or pesticides and find these sufficient and preferable.

  The most difficult crop-protection problems encountered with SRI so far have been with vertebrate pests: rats, or in one case, snakes (Peru) or elephants (Aceh, Indonesia). Some farmers have reported rat damage to be less with SRI, as the wider spacing of plants discourages their entry into the field.

  In general, we recommend that, as much as possible, farmers utilize integrated pest management (IPM) practices and strategies. An interesting example comes from Tripura state of India where farmers place convenient perches for owls and raptors in their SRI fields to encourage natural control of rats, mice, etc. (see 5.1 below)

• Why start with young seedlings? Older seedlings, being larger, are easier to handle. However, once rice plants have started into their fourth phyllochron of growth -- generally about the 15th day after seeds have been sown in the nursery, as discussed for FAQ #14 -- they lose some of their potential for tillering and root growth. Older seedlings will retain less of their original potential for root and tiller growth. On the other hand, young seedlings when managed with the other SRI practices have more produce growth of roots and tillers. SRI plants can have up to 100 tillers or more, compared with the 5-10 or at most 20 tillers that rice plants have when grown from seedlings transplanted when 3-4 weeks old, or even older.
• Why change nursery and transplanting practices? Usual transplanting practices involve removing older seedlings from a nursery that has been continuously flooded. This means that the seedlings start out in a soil environment that lacks oxygen, and crowding inhibits their roots' growth. When removing seedlings from a conventional nursery, little care is taken to protect their roots, and they often lie in the open air and sunlight for hours or even days before being transplanted into the field, so they have desiccated roots that have dried out.

  Then, seedlings are pushed downward into flooded field soil that has little oxygen. Their root tips are invariably inverted upward when plunged down into the soil. Their root profile is like a J, and it takes days for the root tips to reorient themselves downward so that they can resume their growth. These practices result in what is called 'transplant shock,' a well-known effect that causes plants to languish for 7-10 days, or even longer. They often become yellowish for lack of nitrogen and lose their growth momentum at a particularly critical time.

  On the other hand, growing seedlings in well-oxygenated nursery soil enhances their performance (Mishra and Salokhe 2008). SRI seedlings are carefully removed from their garden-like nursery, with soil and seed sacs kept attached to their roots. They are transplanted quickly and gently into the main field with their roots not allowed to dry out in the sun. They are placed gently into aerobic soil, with roots laid in horizontally and shallow (1-2 cm), so that plant can result its growth almost immediately. The plants' profiles are more like the letter L or I than J. This gains plants 7 to14 days of vegetative growth before they flower, which adds disproportionately to their tiller number and root growth.

• Why use such wide spacing? Why reduce plant numbers so radically? When rice plants are crowded together, within hills of 3-6 plants packed together in a clump, with little space between hills, this reduces the room for roots to grow and maybe more important, the amount of sunlight that can reach the lower leaves.
  • Measurements made in 2002 by Dr. Anischan Gani at the Indonesian Rice Research Institute at Sukamandi found that with typical close spacing of rice plants, not enough sunlight reaches the lower leaves in the canopy to support photosynthesis. This means that these leaves, instead of contributing to the plant's pool of energy, were taking energy from the pool parasitically. Rice roots rely mostly on the lower leaves of the plant for their energy supply, to provide carbohydrates to support their metabolism (Tanaka 1958). Thus, crowding plants together impedes whole-plant photosynthesis and undermines the growth and functioning of root systems.
  
  
  

  When there are fewer plants per square meter, all of the leaves of the rice plant are active in photosynthesis, and their root systems are well supplied with photosynthates by the lower leaves. This makes the whole plant more productive. Each plant has more tillers, each with more grains, and the grains are themselves usually heavier. These changes in phenotype more than compensate for having fewer plants per unit area (Thakur et al. 2010).

• Why stop flooding rice paddy fields? Rice is an unusual plant in that it can survive under flooded conditions better than most plants can by having some of cortical cells in its roots disintegrate. The cortex is made up of cells around the column of vascular tissues in the center of the root that transport water and nutrients upward (the xylem), and transport carbohydrates and other synthesized compounds downward (the phloem). Disintegration of cortical cells creates air pockets (aerenchyma) that permit air, especially oxygen, to diffuse passively from the above-ground parts of the plant eventually to the root tips, which need oxygen as well as energy to continue their growth.

  Under flooded conditions, 30 to 40% of the cortex may disintegrate this way, impairing to some extent the transport of water, nutrients and food supply within the root (Kirk and Bouldin 1991). Although rice plants can adapt to hypoxic, oxygen-less conditions by creating aerenchyma in their roots, they do not necessarily perform at their best under such conditions. Under continuous flooding, about 3/4 of the root system degenerates by the time that the plant starts flowering and begins grain formation and grain filling (Kar et al. 1974).

  Nobody knows for sure when or why the continuous flooding of paddy fields began. Rice evolved as an upland plant in well-drained soils, but probably rice was grown in prehistoric times in low-lying fields with inundated, saturated soil because no other crop could be grown there. Over time, rice was planted in fields that were intentionally flooded, as flooding gave significant weed control because other plants, including most weeds, are less able to grow under hypoxic soil conditions.

  Growing rice in flooded fields requires less labor for weeding than does rice grown in upland, unirrigated fields where weed growth can be prolific. If weeds can be controlled by other means than flooding, this changes many calculations. SRI experience shows that much higher yields can come from rice plants grown in aerated soil, with small but reliable applications of water, where rice roots then grow larger, deeper, and function better instead of degenerating.

• Why use a mechanical hand weeder to control weeds? The answer to this question follows from the previous one. The use of rotating hoes or conoweeders to control weeds gives farmers a 'bonus' from active soil aeration. This condition enhances plants' health and crop yield. While weeds can be controlled or removed by hand weeding or by use of herbicides, this foregoes the benefits of soil aeration through 'weeding' that promotes root growth and enhances the abundance, diversity and activity of beneficial soil biota.
• Why use compost in preference to chemical fertilizer? The answer to this question derives from the foregoing discussion. SRI was developed by Fr. Laulanié using chemical fertilizer as the main source of supplementary soil nutrients. However, in the late 1980s when small farmers in Madagascar could no longer afford fertilizer, because the government removed its subsidy for fertilizer, he experimented with using compost. This gave even better results. Factorial trials have confirmed the advantages of compost (Randriamiharisoa and Uphoff 2002; Uphoff and Randriamiharisoa 2002). If not all of the SRI practices are used together, only a subset of SRI practices give higher yield with chemical fertilizer than with compost; but if all the SRI practices are being used, yields with compost surpass those from fertilizer.

  Compost is more than just an alternative source of nutrients, assessed in terms of the amount of nitrogen, phosphorus and potassium that it contains. This is less than chemical fertilizer provides. However, even if compost contains less macronutrients (N, P and K), it contains a host of micronutrients (iron, zinc, copper, molybdenum, etc.). These are important for helping plants to synthesize important enzymes that are essential for plants' metabolism. Compost serves as a more balanced and more complete source of nutrients for soil organisms as well as for the plant itself. Moreover, compost contributes to better structure and functioning of soil systems by better supporting soil organisms and the complex food web that operates underground in a healthy soil (Uphoff et al. 2006).

  Soil with good structure has more pore space, which means that both air and water can be well-distributed throughout the soil's volume. This porosity enhances the soil's capacity to absorb and hold water, so that rainfall does not just run off, carrying topsoil particles with it and eroding the amount and value of the soil. Soil biological activity supports the recycling of nutrients in the soil and the movement of nutrients from the 'unavailable' portion of the soil to become 'available' in the soil solution (Bonkowski 2004; Doebbelaere et al. 2003; Turner and Haygarth 2001; Thies and Grossman 2006).

• Why is agrochemical protection against pests and diseases less necessary? For a variety of reasons, rice plants grown with SRI methods are more resistant to pests and diseases, making it less profitable or not cost-effective for farmers to use agrochemical means of protection. Commonly with SRI management there is not enough pest and disease damage and loss to justify the expenditure and labor required for agrochemical applications.

  One possible explanation for SRI resistance to pests is that when plants are grown in unflooded soil, they will take up more silicon. This would account for the fact that rice plant tillers and leaves are tougher and stronger, resisting being blown over (lodged) by strong winds and rain. Chewing insects would also be deterred by tougher leaves and stems. Also, drier microclimatic conditions in the canopy are less favorable to a number of insects and other organisms.

  A theory called trophobiosis proposed by a French agricultural scientist (Chaboussou 2004) is consistent with what we observe with SRI, because SRI reduces or stops the use of chemical fertilizer as well as agrochemicals. According to Chaboussou's theory, plants' vulnerability to attacks by insects, bacteria, fungi, even viruses, is a consequence of imbalances or deficiencies in their nutrition. This adversely affects the plants' metabolism, which would otherwise (a) convert amino acids into more complex protein molecules, and (b) metabolize simple sugars into complex polysaccharides. These larger molecules would be less easily utilized by predatory insects, bacteria, fungi, even viruses.

  • When inorganic nitrogen is abundantly provided to plants through synthetic fertilizer, they take up more N and synthesize amino acids, the building blocks for proteins. However, with imbalanced nutrition, the plants do not quickly and effectively convert these amino acids into proteins. This leaves a surplus of amino acids in the plants' sap and their cells' cytoplasm, and these simple molecules are attractive to insects, pathogenic bacteria and fungi, and even viruses.
  • Similarly, with the application of pesticides, particularly chlorinated ones, plants' metabolism is interfered with so that the simple sugars that they create through photosynthesis do not get consolidated quickly and continuously into polysaccharides. The plants thus produce an abundance of simple sugars in the sap and cytoplasm which offer pests and pathogens an opportunity to feed easily and to expand their populations.

  Such 'surpluses' of amino acids and simple sugar make plants vulnerable to predation and disease. This explanation is supported by extensive research published in the peer-reviewed literature going back many decades. Chaboussou is not an advocate of 'organic' practices per se because he proposes that any nutrient deficiencies which can impede or unbalance plants' metabolism should be remedied -- by inorganic means if organic supplies are not available.

  More research should be done on this theory of trophobiosis, which has been largely overlooked or ignored by research institutions, and on the mechanisms for pest and disease resistance in SRI rice plants. These relationships and explanations warrant systematic and objective evaluation. Meanwhile, the phenomenon of SRI plants having resistance to pests and diseases, although not always observed, is confirmed by many farmers from their experience and by some scientific trials and evaluations (e.g., Gopal et al. 2010; Karthikeyan et al. 2010).

• Is transplanting necessary? Can rice crops be established by direct seeding? Although SRI was developed with and for farmers in Madagascar who were transplanting their rice crop, SRI does not require transplanting. The operative principle for SRI is that rice roots, which are essential to the plant's future growth, if transplanted should be treated very carefully and should be protected from any trauma and damage. Since Fr. Laulanié was recommending many changes in practices for SRI recommendations, he apparently considered that trying to changing also farmers' methods of crop establishment would have made gaining SRI acceptance more difficult.

  Some farmers who are not wedded to transplanting -- or who have labor shortages that make transplanting difficult to practice -- have been adapting SRI concepts and methods to direct-seeding methods of crop establishment, coupled with the other SRI practices. Their main objective is to reduce labor requirements. They want to achieve this goal even if it means that their paddy yield may be somewhat reduced because they are most concerned with favorable economics, not just agronomics.

  • One method developed by a Sri Lankan farmer, Ariyaratne Subasinghe, and evaluated by a rice scientist at Tamil Nadu Agricultural University in India, Dr. S. Ramasamy, is based on broadcasting germinated seed on a muddy, leveled field. Subasinghe uses five times more seed than if he were establishing his SRI crop with transplanted seedlings, i.e., he broadcasts germinated seed at a rate of about 25 kg per hectare, instead of establishing and transplanting a nursery with just 5 kg of seed per hectare.

    When the young plants in the field are 10-12 days old, Subasinghe 'weeds' his field as if he had transplanted it with seedlings at a spacing of 25x25 cm. This 'weeding' radically thins the stand of rice, eliminating about 80% of the young plants and leaving the remaining 20% of plants in a square, geometrical pattern. There is usually one plant left and the intersections of weeder transects; but sometimes there can be 2 or even 3 plants. Occasionally there is no plant within this intersected space; but the adjacent plants grow larger to fill in most of the open space.

    The goal is to have a sparse, evenly and widely-spaced plant population. This method reduces SRI labor requirements by 40%, according to Dr. Ramasamy, because there is no need to construct and manage a nursery, and it eliminates the task of transplanting (http://ciifad.cornell.edu/sri/countries/india/intnramasapster06.pdf). All the farmer has to do is broadcast his seed and then 'weed' the field just as he would have been done anyway if he had transplanted the crop. Subasinghe says that he is confident of getting a yield of at least 7.5 tons/hectare. While this is less than he could get from a more carefully-managed SRI field, he has many competing demands for his labor time. This gives him a profitable harvest with a much reduced expenditure of labor.

  This discussion underscores again the importance of flexibility and innovativeness to capture the benefits of SRI under specific local conditions. This Sri Lankan farmer had 2 hectares of rice land, and his operation was labor-constrained because his two children were still young. Economizing on labor requirements was a rational strategy for him. Farmers in various countries are trying out SRI concepts out innovations like raised beds, zero-tillage, intercropping with potatoes (gaining momentum in Vietnam), seedbed solarization, etc. These versions are not in conflict or competition with SRI, but are welcome adjuncts to the extent that they help farmers solve their problems and meet their goals.

(Note that most Environmental Benefits are discussed under #5 above)

SRI has been sometimes dismissed or ignored on the grounds that the results reported are 'too good to be true.' This, however, is a non-empirical and uninformed basis for rejection, relying on a priori reasoning and not on evidence. Use of SRI methods will not always produce all of the desired or reported benefits. However, the methodology offers a remarkable number of desirable outcomes that can be achieved at little or no incremental cost.

• Increases in yield: Grain yield per hectare is usually treated as a summary indicator of productivity. However, this measure of yield represents only the productivity of land, not the productivity of other factors of production: labor (earnings per day), water (crop per drop), or returns on capital (profitability). As a rule, farmers are interested in many considerations besides or beyond agronomic yield. But yield always attracts the most attention, especially if land is relatively the scarcest, most constraining factor of production.

  Increases in grain yield achieved on-farm with SRI methods usually range from about 20% to 200%, and sometimes even more, as seen in the data reported below from Madagascar, India and Cambodia. Gains in output that are achieved with reductions in one or more inputs (labor, water, capital) are more significant than if attained with a greater expenditure of these resources. Also, with SRI methods, there is usually an increase in the production of straw (biomass). This yield increase is very important for many small farmers, who derive much benefit from being able to use more straw for fodder, thatching and other purposes.

• Increases in factor productivity: Output per unit of input is the most important economic measure, more important than the simple agronomic measure of grain or straw yield, since achieving more output with greater expenditure on inputs is of uncertain benefit, whereas getting more from less is certainly advantageous. SRI is the only innovation that we know of where the productivities of all these factors of production -- land, labor, water, and capital -- are raised concurrently. This has made SRI 'suspect' because economists believe that this is impossible, insisting that there is 'no free lunch.' However, this premise applies to closed systems only, in which there must be diminishing returns, not necessarily for open systems.

  With SRI practices, energy and nutrient inputs come from biological activity. This is free, provided that certain conditions for organisms' growth and functioning are met. This is what makes possible the broad-based SRI gains in factor productivity. Total factor productivity is hard to measure and to report on in a summary manner because of the difficulty of combining land, labor, water and capital, even in monetary terms. But evaluations of each factor show gains, i.e., more output per unit of input.

  For many years, it was asserted that if rice plants had more tillers (panicles), they would necessarily have fewer grains per panicle, a manifestation of the rule of diminishing returns. This would make the profuse tillering observed with SRI management less desirable than if no such limitation existed. However, the assumption of an inverse relationship between number of panicles per plant and number of grains per panicle is empirically fallacious for rice plants whose root systems have not degenerated due to flooded soil conditions.

  With SRI management, we find that rice plants can have both more fertile tillers and more grains per panicle, making rice production a positive-sum proposition, rather than zero-sum or negative-sum. This is possible because when rice plants have larger, healthy root systems that do not die back or senesce due to hypoxic soil with continuous flooding. These systems enable them to function as open systems, rather than as closed systems (or worse, declining systems). This difference in plant physiology underlies SRI success (Thakur et al. 2010).

• Reductions in water requirements: Because SRI irrigated rice is grown without continuous flooding, less water is needed (25-50% less) for growing a crop that produces greater output. Water productivity can thus be doubled or tripled, even increased by six times (Ceesay et al. 2007). Under some soil conditions, it may not be possible to make such great gains in water saving; but under other conditions, as much as two-thirds of water has been saved.
• No need to rely on purchased inputs: Farmers can use SRI methods successfully and cost-effectively (a) without buying and using chemical fertilizer, (b) without purchasing new seeds, and (c) not needing pesticides, herbicides or other agrochemicals to control pests, weeds and diseases. While fertilizer can be used with positive results, farmers who have enough labor and access to biomass to make and apply compost can get quite high yields while improving their soil's fertility. Using a rotary weeder to control weeds not only eliminates weeds but also gives farmers the added benefit of stimulating more yield. This is thus a small capital investment well worth making Usually SRI plants are robust enough that chemical means of pest/disease control are not needed or are not worth the expenditure.

  Farmers should use the best available rice varieties that are suitable for their environment because starting with the best varieties for their conditions and objectives gives them the greatest return from the other resources that they invest in rice production. Since consumers will often pay more for local varieties of rice, because of their preferred qualities, it is often more profitable to grow traditional varieties than 'improved' ones, even if yield is not as high as from high-yielding varieties or hybrids. If 'organic rice' commands a higher market price, efforts to nourish plants and soil with organic inputs becomes more remunerative. For farmers, minimizing their monetary costs for rice production has many advantages, including avoiding indebtedness and the possible loss of their land if they have a crop failure.

  Some research has shown a combination of organic and inorganic nutrition (integrated nutrient management) giving higher yield than organic materials alone. Whether or not this will be more profitable for farmers will depend on the respective costs of buying fertilizer and of making compost. Rather than prescribe any particular practice or combination of inputs, with SRI, farmers should determine which will best meet their purposes considering the costs and constraints of the different inputs and also issues such as risk factors.

• Higher returns to labor, even labor-saving: Even when SRI methods require more labor inputs per hectare than conventional practice, the higher yields with SRI almost always give farmers higher production per hour/per day of labor, and this is what raises family income. As noted above, SRI farmers often find that once they have mastered SRI methods, they can grow their larger crop without additional labor or even with less labor. Whenever farmers can reduce labor as well as seeds, water and costs, this makes SRI very attractive.
• Reduced labor requirements for women: More research needs to be done on this, but a 2004 study with 100 farms in Tamil Nadu found that women's labor inputs per hectare were reduced by 25% because weeding, previously by hand, had been all done by women. When mechanical weeding was introduced, men took over this task as more appropriate to them, increasing their labor inputs to household rice production by 60%. Labor requirements were reduced by 11% per hectare overall. With net income per hectare going up from $242 to $519, men were well compensated for taking on this new operation.

  In Andhra Pradesh, where there is no cultural restriction on women working with machines, women have taken up mechanical weeding instead of doing weeding by hand. A study found that this greatly speeded up their weeding work with doubled yield. Women's time spent in weeding was reduced by 72%. The new technology was found also to improve women's postures for their weeding work and to alleviate muscle fatigue (Mrulini and Ganesh 2008).

• Increase in farmer net income and profitability: If farmers can achieve higher output with reduced cost of inputs, this raises incomes by even more than the increase in production. A report from Cuba on SRI rice production from 26 hectares on a cooperative farm showed a yield increase of only 15%, from 4 tons/ha to 4.6 tons/ha. But the net income received from the rice was 70% higher (747.15 vs. 439.34 pesos/ha) because: (a) seed costs were cut in half, (b) fertilizer use was reduced by 89% -- from 350 kg/ha to 37 kg/ha, (c) 40% less water was used, a significant saving because irrigation water had to be accessed by diesel pumps, and (d) the labor needed for transplanting was reduced from 16 persons to 5 persons (Socorro et al. 2008). A number of the economic considerations listed here were considered by farmers as more important than yield.
• Reductions in economic risk: The agronomic practices recommended for SRI -- use of very young seedlings, just one plant per hill, reduced plant population, no flooding of the field, and reliance on organic fertilization -- all appear risky. However, two evaluations that calculated actual risks based on data from large-scale random samples have shown otherwise.
  • In an evaluation done in Cambodia for GTZ, Germany's development agency, with 400 SRI farmers and 100 non-SRI farmers in the same villages randomly selected in five provinces, an economic risk assessment showed that farmers' risk of not achieving a target net income was significantly lower with SRI practice. With SRI, the probability of not reaching a target income of $US100/ha was calculated to be 17% vs. 42% with use of standard methods. Rice farmers were found to be 2.5 times more likely to lose money when using standard methods. The study concluded: "SRI is an economically very attractive methodology for rice cultivation with a lower economic risk compared to other cultivation practices" (Anthofer 2004).
  • An evaluation for the International Water Management Institute (IWMI) in Sri Lanka based on 120 farmers randomly selected in two districts, half of them using SRI methods and half not using SRI, calculated households' economic risks with SRI vs. conventional practice. Risk calculated according to three alternative wage levels, valuing labor inputs at (a) zero wage, assuming use of only family labor, (b) the prevailing agricultural wage, and (c) the prevailing non-agricultural wage, an opportunity-cost assessment. In the first calculation, the probability that a household would end the season with a net economic loss was 9 times greater when using conventional practices than when using SRI methods. At the prevailing agricultural wage, this probability was 8.4 times greater, and if labor inputs are valued at their non-agricultural opportunity cost, the probability of not losing money from rice cultivation was 6.4 times greater (Namara et al. 2004, 2008).
• Less susceptibility to pests and diseases: This was discussed under #3.2 above.
• Less vulnerability to adverse weather conditions, even resisting climate change effects: More evaluation needs to be done on this, but it has been often observed that SRI rice plants are:
  • More resistant to drought and water stress;
  • Better able to recover from flooding, provided they have gotten their root systems established before the flooding occurred;
  • More resistant to storm damage and lodging, to being knocked down (lodged) by strong rain and high winds (Chapagain and Yamaji 2009); and
  • Better able to tolerate extreme temperatures, provided their root systems have gotten well established in the soil.
• Resistance to storm damage and lodging is one of the most important improvements in crop performance that SRI management can achieve in a world with climate change, where severe storms with strong wind and rain are likely to become more frequent and more severe. The most extensive evaluation of SRI effects in this respect is by Dastan et al. (2013).

  Given the growing variability in weather patterns and the more frequent occurrence of 'extreme events,' it will become more and more important in the 21st century, especially for poor, resource-limited households, often cultivating in the most vulnerable regions, to have crops that can withstand as much as possible the adverse effects of climatic stress (Uphoff 2011).

• Shorter crop cycle, getting higher yield more quickly: SRI farmers in most countries report that their SRI crop reaches maturity 5-10 days sooner than do crops of the same variety grown with standard practice. When farmers can harvest their crop sooner, they can reduce the amount of water they need to grow their crops, and they reduce their crops' exposure to storms or other climatic hazards and to pest and disease damage which are more at the end of the season. These effects should be evaluated more broadly.
  • In 2007, 413 farmers in Morang district of Nepal who used SRI methods with eight different varieties had their crops mature on average 16 days sooner than these varieties normally reached maturity (Uphoff 2011: Table 7.2). Even with SRI management shortening the crop cycle by 11%, average SRI yields were more than double those achieved with farmers' usual practices, 6.3 tons per hectare compared with 3.1 tons.
• Higher milling outturn: It was noted already that farmers and millers are finding that unmilled SRI paddy rice gives greater output when milled, yielding more polished rice per bushel or per bag of paddy rice. This is attributable to there being less chaff because SRI panicles have fewer unfilled grains, and fewer broken grains because SRI grains being usually somewhat heavier and denser resist shattering during the milling process.
  • Already in 2002, grain millers in Sri Lanka were offering to pay farmers, even before they had harvested their SRI crop, a higher price per bushel of paddy (10% more) because they knew that they could get even more than 10% additional polished rice when they milled the SRI paddy and sold the final product.
  • An evaluation done at Sichuan Agricultural University in China in 2004 confirmed this, as the same variety of SRI paddy rice averaged 16.1% more total milled rice outturn, and 17.5% more head milled (whole grain) rice when milled (Ma 2004; He et al. 2004).
  • An informal survey of farmers in 12 villages in Tripura state of India in October 2007 found they got, on average, 18% more milled rice per bushel of unmilled SRI paddy due to less chaff and less breakage of grains.

  This information points to a 'bonus' of about 15% more edible rice on top of the increased harvest of paddy rice when SRI methods are used.

• Improvements in grain quality, and possibly in nutritional value: There are numerous reports that consumers consider the quality of rice grown with SRI methods to be higher, but we have no systematic evaluations on this.
  • The Sichuan Agricultural University evaluation noted above also measured chalkiness in SRI rice compared with the same variety grown conventionally. Chalkiness is considered an undesirable quality of rice, and it also contributes to more breakage during milling.
  • According to Ma (2004), SRI had 30.7% fewer chalky kernels, and 65.7% less chalkiness overall. Since there is less breakage of SRI paddy rice when it is milled, there could well be higher protein content, as resistance to breakage during milling is known to be associated with higher protein (Leesawatwong et al. 2004).
• There is also reason to think that there could be higher concentrations of micronutrients in SRI grain because the kernels are denser, i.e., heavier without being larger. This also could be contributing to less breakage during milling. Since SRI roots are larger and reach more deeply into the soil, they would have more capacity to acquire micronutrients. This would make the plants themselves healthier and better able to resist damage from pests and diseases. More uptake of micronutrients would give plants more of the needed building blocks for synthesizing enzymes needed for their metabolism.
• More systematic work should be done on this and other benefits. Some recent research has shown that growing rice under unflooded soil conditions enhances copper, zinc, magnesium and manganese in the grain (Xu et al. 2008). Other research in India has found significantly greater uptake of micronutrients (Fe, Zn, Mn, Cu) in SRI rice plants compared to plants of the same variety conventionally grown (Prasanna et al. 2012).
• Other health benefits: Apart from nutritional benefits still to be assessed, we know that stopping continuous flooding of paddy fields should reduce the incidence of mosquito-borne diseases like malaria and dengue fever. According to recent research, there is 10-15 times less uptake of arsenic by rice plants grown under unflooded soil conditions (Xu et al. 2008). But this too needs to be further investigated. There are also some health benefits specific for women that are reported in response to FAQ 10.

  Where the productivity gains of SRI permit farmers to reduce the amount of land they devote to rice production and to diversify their farming systems, they can produce more fish, more fruits, more vegetables, legumes and small livestock. This not only improves households' income but also their nutrition from having more diversified diets. For examples of such systems, see: http://ciifad.cornell.edu/sri/countries/cambodia/cambSidMPREng.pdf

So far, farmers have found that SRI methods improve performance of rice plants of practically all varieties, old and new, unimproved and improved, traditional and modern, local and hybrid – although not all respond equally well. Some varieties respond better to SRI management practices than do others. Farmers will understandably want to use whatever varieties of rice are the most productive under their own conditions. However, farmers evaluate productivity not only in agronomic terms, but are concerned also with their 'bottom line.' They will want to consider what prices they can get for their rice, and what will be their net income after deducting their costs.

The highest yields achieved so far with SRI methods have all been with modern high-yielding varieties or with hybrids, many of which have been bred for more tillering capacity, for example. The world-record paddy yields reported from Bihar state of India in 2011 were with hybrid varieties, as reported above.

At the same time we also know that many traditional local varieties give very good responses to SRI management: 6-10 tons per hectare, and even as high as 13 tons per hectare in Sri Lanka. Because consumers usually prefer the taste, texture and other qualities of 'unimproved' varieties, they are often willing to pay twice or even three times as much per kilogram for traditional rices.

Varieties can respond somewhat differently to SRI management compared to standard practice. In general for SRI, one wants varieties, old or new, that can tiller profusely. Significantly, Indian farmers have found that a favorite modern variety known as 'Swarna' (MTU 7029), which has had a reputation for being 'shy-tillering,' with SRI methods it tillers prolifically.

 

SRI methods can make rice production with hybrid varieties more profitable, since these usually give a very good yield response to SRI practices. SRI raises the economic returns from planting hybrids in part by reducing costs of production because it lowers seed requirements by 80-90%. For many farmers, the high cost of hybrid seed is one of the barriers to adoption of hybrids.

• In Indonesia, on the island of Bali in the 2006 dry season, 24 farmers who used SRI methods with Long-ping Chinese hybrids got an average yield of 13.3 tons/ha on 42 hectares, compared with a yield of 8.4 tons/ha when they cultivated the hybrids with their usual practices (Sato and Uphoff 2007). Thus, SRI can make hybrid varieties considerably more remunerative by raising their yield with lower costs of production.

However, traditional varieties grown with SRI methods can often be more profitable than are higher-yielding varieties because of a combination of yield increases and higher market prices, with low costs of production.

• An evaluation done in Orissa state of India of 99 traditional rice varieties by the NGO Sambhav found that under SRI management, three of them gave yields of 9 to 11 tons per hectare; 11 varieties gave yields in the 8-ton range; 15 (4 of them aromatic) gave 7-ton yields, while 36 varieties (5 aromatic) gave 6-ton yields, and 34 varieties (4 aromatic) gave 5-ton yields. With agroecological management, all of the traditional varieties evaluated gave quite respectable yields. They can also be quite remunerative since many of these varieties are preferred by consumers and have premium prices in the market.

Working with partners in a number of countries, CIIFAD has been trying to support production and marketing of indigenous rice varieties grown organically. It has worked with various partners to promote rice exports that can command premium prices for farmers and are produced in environmentally-benign ways, particularly reducing the demand for water.

In 2005, the SEED Initiative award (Supporting Entrepreneurship for Environment and Development) sponsored by IUCN, UNEP and UNDP was given to CIIFAD together with NGO and farmer-organization partners in in three countries (the Koloharena farmer organizations in Madagascar; CEDAC and the farmer groups that it works with in Cambodia; and Oxfam-Australia and farmer cooperatives that had been introduced to SRI in Sri Lanka).

While SRI methods can be used successfully with both 'new' and 'old' rice varieties, they may help to conserve rice biodiversity by making at least those local landraces that respond well to SRI management very profitable and thus competitive economically with hybrids and HYVs.

SRI is intended to give farmers more choices to improve their well-being and security. Those who want the highest possible yield will probably choose to use SRI methods with hybrids or modern varieties. Farmers thinking about their 'bottom line' may prefer to plant traditional varieties.

CIIFAD and several partners in other countries are working with Lotus Foods, a family company in San Francisco, California that imports specialty rices and markets them throughout the U.S. (http://www.lotusfoods.com/SRI/AboutSRI.aspx). Several product lines of organically-grown, 'traditional' SRI rices have already been developed with good packaging and quality control to gain market access and raise incomes for SRI farmers in Madagascar, Indonesia and Cambodia.

Initially, when farmers are starting to learn the new methods, it takes some time to use them quickly, confidently and well. Handling the tiny young seedlings -- when farmers are used to handling larger, older ones -- can be worrisome, and transplanting will go slowly at first. However, once farmers get accustomed to the new methods, they can carry out the practices more quickly. Depending on the difficulty of controlling water, applying small amounts of water regularly can take more time than simply keeping fields always flooded. But in general, we find that as farmers become comfortable with and more confident in SRI practices, this alternative system is labor-neutral, i.e., it does not increase labor, or SRI can it even become labor-saving, reducing farmers' labor requirements.

• An evaluation of SRI by Moser and Barrett (2003) raised the issue of greater labor-intensity with SRI management. It was surely correct for the five villages surveyed in Madagascar. However, a later evaluation by these authors plus two other colleagues (Barrett et al. 2004), analyzing a data base of 108 farmers having different lengths of SRI experience, found that although SRI required more labor at first, by the 4th year, labor inputs per hectare were 4% less, and by the 5th year, 10% less. In countries beyond Madagascar, the achievement of labor-saving has usually been more rapid and more extensive than in Madagascar, for reasons not yet understood.
• Evaluations of SRI in Cambodia for the German aid agency GTZ (Anthofer 2004) and in Indonesia by a Nippon Koei technical assistance team (Sato and Uphoff 2007) found that on average, SRI methods were labor-neutral. New SRI farmers required more labor to grow their SRI crop, but more experienced SRI farmers need less labor; so on average there was no change. Evaluations in China (Li et al. 2005) and in India (Sinha and Talati 2007) have shown average reductions in labor requirements, even from the first year. Reports on SRI use in China and India have noted that SRI is considered by farmers and officials to be generally labor-saving (Uphoff 2007b; The Hindu, January 1, 2008).
• In India, a detailed evaluation in 2004 by researchers from Tamil Nadu Agricultural University of side-by-side plots on 100 farms in the Tambiraparani river basin managed respectively with SRI and conventional methods found that labor inputs per hectare were 11% lower with SRI (Thiyagarajan 2004). Disaggregation showed that men's labor inputs per hectare went up by almost 60%, because men considered SRI weeding being done with mechanical weeders to be 'men's work.' Accordingly, they took over this activity, and this reduced the amount of labor that women put into rice production by 25%.

A more important consideration is that farmers frequently find ways to reduce their labor inputs with SRI by improving the design and operation of weeders, by finding ways to reduce their labor requirements for nursery management and transplanting, or by doing alternate wetting and drying of paddy fields rather than more labor-demanding daily water management. So, the labor needs with SRI are not something fixed. Both farmers' skills and the techniques or practices that they use with SRI are undergoing continuing change. Farmers should always be informed that SRI will usually require more labor when starting out, but there are reasons to expect that this increase will be transitory or transitional, rather than necessary and invariant.

How the introduction of SRI affects women's situation will depend on what is the prevailing gender division of labor in rice production in the local situation. Most reports have said that women's labor burden has been reduced when SRI is introduced and the work itself is less onerous or unhealthy.

• Rice transplanting becomes quicker once the new methods are learned because plant populations are reduced by 80-90%.
• Men often take over the task of weeding the fields (often culturally classified as women's work) when mechanical weeding is introduced (this is considered as 'men's work').

The first systematic evaluation of gender impact was done in Tamil Nadu in 2004, looking at the labor inputs for 100 farmers in the Tambiraparani river basis, cultivation one-acre lots of SRI and conventional rice side-by-side. As reported for the preceding FAQ, overall labor inputs were reduced by 11%, but there was a stark difference in labor investments according to gender. Men took over the (mechanical) weeding because this was considered to be 'men's work,' so their labor inputs increased by 60% (Thiyagarajan 2004).

• In these first-year comparisons, women's labor for transplanting went up by 36%, but usually women report a reduction in the time for transplanting once they acquire skill and confident in handling very young seedlings. These are only 10-20% of the previous number and much lighter and easier to handle. Rice women farmers often have greater labor burdens these days as a result of the 'feminization' of agriculture as men migrate to employment in the cities, leaving women with more agricultural work overall. But when this question was studied in Cambodia, the consensus among women was that SRI lightened their burdens (Resurrection et al. 2008).

There are also some health benefits reported from the Philippines and India. This subject deserves much more study, but so far the reported results from SRI adoption have been favorable to enhancing the status, health and comfort of women.

• Women's Health is an NGO in the Philippines that is trying to minimize women's health problems in Isabela Province in Northern Luzon. It has reported that because with SRI, transplanting and weeding are no longer done in standing, stagnant water, the incidence of women's urinary tract and vaginal infections has been reduced, a benefit that is not normally spoken about. But this NGO made it a point of evaluating this effect.
• In Orissa state of India, there is a study underway for a PhD dissertation from Wageningen University which examines the effects of SRI practice on women's physical well-being. Results of surveys and measurements conducted in 47 villages across 10 districts are still being evaluated, but there is agreement among women that SRI methods are less strenuous, because of less discomforting postures and less total amount of work.

Also, curtailed used of pesticides and chemical fertilizers is reported to reduce skin and other ailments. Indirectly there is the benefit reported of more time to cook and take care of housework. An innovative effort is being made to quantify physical effects through what is called a Rapid Comparative Pain Assessment, using participatory methods and a body map for identifying loci and severity of pain. There is wide agreement among women that adoption of SRI methods reduces discomfort and pain (Sabarmatee 2012).

At the same time, some women in the focus groups reported less enjoyment as the field work is done more quickly with less and socializing. This can be considered as a 'cost' for SRI introduction, but it was not enough to be a deterrent to the women's using SRI practices.

When SRI was being developed during the 1980s, Fr. Laulanié used chemical fertilizer because at the time this was thought to be the best, indeed the only way to achieve higher yields. When the government of Madagascar stopped subsidizing fertilizer in the late 1980s, the farmers who were working with Laulanié could no longer afford to purchase fertilizer. So he and they began to enhance soil fertility with compost, out of necessity. Happily, they learned that compost when used in conjunction with the other SRI practices could improve rice yields more, and certainly more cheaply, than could fertilizer on farmer's 'poor' soils. The compost was made mostly from rice straw and any other available biomass (weeds, grass, shrubs, leaf litter, etc.), decomposed for 30 to 60 days. Little or no farmyard manure was used since few of these farmers were rich enough to own cattle.

Chemical fertilizer used in conjunction with other SRI practices can certainly enhance SRI yield. But factorial trials have indicated that organic fertilization can increase grain yield by more than does inorganic fertilizer when all the other SRI methods are used with it. Association Tefy Saina does not regard compost as a requirement for SRI, but only as an accelerator or booster. Some combination of the two sources of nutrients can be optimizing, provided that chemical fertilizer does not inhibit or adversely affect the soil organisms that contribute to rice plant productivity.

In Indonesia, two versions of SRI are encouraged by the Nippon Koei technical assistance team: (a) basic SRI, in which farmers' reliance on chemical fertilizer is reduced by 50% as they increase the amount of compost added to their soils to support crop growth; and (b) organic SRI, in which only organic inputs are used. Under some marketing regimes, the latter qualifies for a premium price in the marketplace. Some SRI proponents have a strong preference for 'organic SRI,' but SRI can be either organic or only partially organic.

• Very often, when farmers convert their cropping systems, previously dependent on chemical fertilizer, to fully 'organic' production methods, they have to go through a transition period when their yields decline for a year or two, as the soil systems adjust to the new nutrient regime. Their soil systems need time to adjust to functioning without a supply of inorganic nutrients and to build up the soil biota, often depressed or unbalanced by the use of synthetic fertilizers.
• When starting to use SRI, on the other hand, we seldom see a 'transition' phase. Usually, farmers who adopt SRI practices with greater or total reliance on organic fertilization get an increase in yield already in their first year. There is a 'windfall' from converting their soil from anaerobic status, without air, to aerobic conditions, well supplied with oxygen.

With petroleum prices rising and with prices for chemical fertilizer and other agrochemicals increasing, there will be more and more interest in production systems that do not depend on these purchased inputs. Also, there is growing concern for the quality of soil and water resources, and for soil health and human health. We can anticipate more and more demand from consumers for organic food products, i.e., food produced without agrochemical nutrients or crop protection. There will also be a demand from citizens to reduce the build-up of nitrates in groundwater and surface water supplies, and to stop the accumulation of toxic chemicals in their soil and water. Thus, we anticipate that organic versions of SRI will gain in popularity -- and in productivity.

So far, benefits from SRI practices have been seen in a wide variety of environments. But there must be some limitations on any biological process. There will certainly not be equal advantage from the new methods in every context. But it is remarkable is that SRI methods have been seen to improve productivity in a wide range of circumstances.

• In Madagascar, two large-scale factorial trials were conducted in 2000 and 2001 using standard scientific methods that validated the merits of SRI practices, respectively and collectively. Six variables (age of seedling, spacing, water management, etc.) were evaluated with random-block-design trials in two contrasting environments: in Morondava (N=288), at sea level, on poor sandy soils, and with tropical climate; and in Anjomakely (N=240) at 1200 meters above sea level, on better clay and loam soils, with temperate climate. The patterns of yield response to SRI practices, evaluated respectively and in all different combinations, were essentially the same under both sets of trials, despite large differences in agroclimatic conditions (Randriamiharisoa and Uphoff 2002; Uphoff and Randriamiharisoa 2002).
• In Nepal, SRI methods have been used successfully in the southern terai, which has a sub-tropical environment a few hundred meters above sea level, and up to 2,700 meters elevation, which presents much colder and limiting climatic conditions.
• In Africa, we have seen good SRI results in the Gambia, a low-lying tropical environment (Ceesay et al. 2006), and in Mali in the Timbuktu region on the edge of the Sahara desert (Styger et al. 2011).

One of the remarkable things about SRI is thus its versatility, although one should never assume that its methods will be successful everywhere. It is always necessary to try out the methods under specific circumstances to see how they perform. There are some contexts where we do not expect SRI practices to be successful, or very successful.

• Cold climate: Young seedlings are vulnerable to cold temperatures, so these can be a limiting factor. Rice plants' biological age is reckoned by how far they have developed -- by their number of leaves (leaf number), rather than by the numbers of calendar days since sowing. In colder climates, one should start with seedlings 15-25 days old since these are likely to be physiologically equivalent to seedlings 8-12 days old grown under warmer temperatures. Reporting seedling age in terms of number of days after sowing or transplanting, age is more precisely and meaningfully assessed in terms of 'leaf stage.' For purposes of SRI management, seedlings in the 2-3 leaf stage should be considered 'young.'
  • A rice production system known as 3-S, devised by the late Prof. Jin Xueyong of the Northeastern Agricultural University in Haerbin, not far from Manchuria, has many similarities with SRI: single seedlings, wide spacing, reduced water, more organic matter. 3-S uses 45-day-old seedlings grown in nurseries under plastic tents, started when there is still a foot of snow on the ground. Yields of 8-9 tons/ha have helped spread 3-S in northern China (Jin et al. 2005; http://ciifad.cornell.edu/sri/countries/china/cn3ssys.html). Cold temperatures can require some modifications of practice, particularly regarding seedling age, but SRI principles can usually be adapted to them.
• Water control: If the soil cannot be kept intermittently moist and mostly aerobic, less or maybe no benefit can be derived from SRI practices. Low-lying soils that are continuously waterlogged are not suitable for SRI cultivation because aerobic soil organisms cannot prosper in an environment which lacks oxygen. In such areas, it may be possible to install drainage facilities to remove excess water, however. The increased economic returns achievable with SRI methods could make such investments financially feasible.
  • In Indonesia, farmers cultivating rice in the middle of large-scale irrigation systems with field-to-field irrigation flows have little control over their water supply. Accordingly, they have devised in-field soil management practices that enable them to use SRI methods quite effectively -- raised beds within the field coupled with drainage channels around the inside edge of the field that help to get rid of unneeded water more quickly.
  • In India, rainfall in the state of Tripura averages 2500 mm/year. It has been seen there that by putting small drainage channels across the length of SRI fields every 8 or 9 rows, the soil can be kept well-drained enough to use SRI methods successfully. This sacrifices 11-12% of the field area for drainage purposes, not planting, but the yield increase on the remaining 88-89% of the area more than compensates for reconfiguring the field.

  These examples show how adaptations and innovations can be introduced to deal with the constraint of excessive water. The numerous advantages of SRI can provide sufficient economic justification for investing in improving water control structures -- gates, channels, canals, and drainage facilities. The multiple benefits also give farmers incentive -- once they are agreed on the advantages of using SRI practices -- to form water user associations which can take the necessary steps to manage their water supply appropriately so as to gain from SRI opportunities.

• Soils: Favorable soil characteristics have significant impacts on crop productivity in general; so also with SRI. Most of the good results with SRI methods have been seen on soils that are acidic (pH less than 6.0); generally there has been less improvement, even poor performance, on alkaline soils. Initial evaluations of SRI in the Punjab state of India showed that its methods gave a 30% lower yield on soil affected by salinity, while on a heavy clay ('sticky') soil, SRI yield was 70% higher; on three other categories of soils considered more typical, the SRI yield increases averaged 62% (reported by Dr. Amrik Singh, ATMA-Gurdaspur).

  Research in India has shown that the application of compost or other organic matter on saline soils can alter their pH and make them more fertile, probably also by improving soil structure to make root growth easier (Rangarajan et al. 2002). It may be possible with applications of organic matter to make saline soils amenable to SRI production methods. Research on SRI methods used with saline soils in Mozambique gave some but not uniformly positive results (Menete et al. 2008).

  It has been proposed that SRI will augment yield particularly on, and possibly only on, soils with high content of iron (Doberman 2004). The conclusion that SRI benefits are limited to iron-rich soils is contradicted by trials conducted in 2003 across all 22 districts of Andhra Pradesh state of India. The staff of the state agricultural university (ANGRAU) evaluated SRI on a wide range of soil types and under diverse agroecological conditions. These trials showed SRI methods giving, on average, 2.5 tons/ha more yield across the state. All of the districts showed yield improvements while using less water and less fertilizer. The largest average increases (4.8 tons/ha) were on the interior, lighter, well-drained soils, while heavier, low-lying coastal/delta soils had average increases of 1.8 tons/ha (Satyanarayana et al. 2006).

  Research in Panama by Turmel on the effects of soil type on SRI performance has showed that yield increases with SRI management can be greater on 'poor' soils than on 'better' soils, particularly on soils with low available phosphorus (Turmel 2011). Her analysis of 70 data sets from across 15 countries showed SRI methods increasing yields more, in both absolute and relative terms, on soils classified as 'poor,' according to FAO criteria, than on soils that can be classified as 'good' (Turmel et al. 2011). Poor rural household usually must cultivate their crops on soils that are poorer than those cultivated by richer farmers. This suggests that SRI is an unusual innovation also for its being more productive where resource endowments are poorer.

This is an important question because much of the world's poverty is found in rural areas where there are no irrigation facilities. If SRI is beneficial only under irrigated conditions, it could not redress the worsening of income distribution in the world. However, NGOs working with farmers in the Philippines, Myanmar and India have found that with suitable adaptations, SRI concepts and practices can give substantial increases in rice production under upland or rainfed conditions, even averaging 7 tons per hectare in some places. Most rice farmers with irrigation facilities would be very happy to achieve such yields.

• Philippines: In 2003, the NGO Broader Initiatives for Negros Development (BIND) in Negros Occidental Province did 20 on-farm replicated trials, with a popular local variety Azucaena, to evaluate rainfed SRI results with five different spacings (4 trials each) on a total area of 4,000 m2. Three or four seeds were sown in hills spaced at 15x40 cm, 20x40cm, 25x40cm, 30x40cm, or 35x40 cm. Then 12-15 days after sowing, the hills were thinned to one plant each, leaving the most vigorous young plant. The soil between the hills was mulched with leaves and branches of a leguminous shrub (Gliricidia) to conserve soil moisture, suppress weeds, and lower soil temperatures so that soil organisms would become more abundant. The highest yield was 7.7 tons/ha with 20x40cm spacing. The average grain yield for all 20 trials was 7.2 tons/ha, compared with usual rainfed yields in the area of 1.5 tons/ha. Organic fertilization was used (chicken manure and a seaweed foliar spray). Quite remarkably, virtually all of the tillers that the plants produced were fertile, for almost 100% effective tillering. See: http://ciifad.cornell.edu/sri/countries/philippines/binuprst.pdf
• Myanmar: Starting in 2001, Metta Development Foundation, a Burmese NGO, began introducing SRI methods through farmer field schools in the ethnic-minority states of Kachin and Shan in the north. SRI practices were adapted to rainfed conditions because farmers in the region have no irrigation facilities. Average rainfed yields in the area are 2 tons/ha. On farmer field school demonstration plots, where SRI methods were used as expected, average yields were over 6 tons/ha. However, on farmers' fields, paddy yields averaged over 4 tons/ha without full use of SRI methods, and have increased year to year (Kabir and Uphoff 2007).
• India: In 2003, PRADAN, an NGO working in impoverished districts of eastern India, introduced SRI with 4 farmers in Purulia district of West Bengal state. The number expanded to 150 farmers the next year. An evaluation team from the India Program of the International Water Management Institute (IWMI) found that farmers who used all of the recommended methods, adapted for rainfed production, achieved 9 tons/ha, and one reached 15 tons/ha (pers. comm., S.K. Sinha, team leader; see Sinha and Talati 2005). SRI use has kept on expanding in Purulia district, and a PRADAN report from Purulia in 2007 showed 3,793 households using SRI methods; 54% had yields in the 6-8 tons/ha range, and 28% were over 8 tons/ha , where rainfed yields have usually been 2-3 tons/ha. The 5-year average SRI yield in Purulia was reported as 7.4 tons/ha. PRADAN has been introducing SRI in rainfed areas of Bihar, Orissa, Jharkhand, Madhya Pradesh and Chhattisgarh, also with good results.

There is therefore reason to expect that the benefits of SRI, practiced with appropriate adaptations, can enhance food security and incomes for households in regions without irrigation. Farmers need to learn not to hoard rainwater in their fields when the rains come, as this will lead to degeneration of their rice plants' root systems. If they transplant, rather than do direct seeding, they should establish several nurseries, not just one, started in sequence about 2 weeks apart. Farmers should being willing to sacrifice all but the one of these nurseries, using just the one that has young seedlings with the best age (under 15 days) when the rains come. This can be a 'hard sell,' but this decision can lead to greatly augmented yield. Instead of making seed savings of 90%, farmers planting three nurseries make only a 70% saving – but they can get yields several tons/ha more than if they do not have 'young' seedlings ready to transplant when the rains begin.

SRI is not a technology with prescribed, fixed practices; rather it is a set of agronomic principles and practices that, among other things, promote greater root growth and more abundant, diverse soil biota. Beneficial effects for resulting plant growth and performance can be anticipated if SRI methods are suitably adapted for the production of other crops. We have seen such extrapolations and extensions of SRI methodology particularly in India, but also in Ethiopia and Nepal. In Bihar state of India, the acronym SRI now stands for System of Root Intensification. We are hopeful that applications of these principles can be extended elsewhere in Africa and in many other parts of the world.

• Wheat: This major cereal crop, like rice, belongs to the grass (gramineae) botanical family, so it could be expected to be responsive to SRI management practices. When the People's Science Institute (PSI), an NGO based in northern India, tried SRI methods with two varieties of wheat in on-station trials in 2006, it recorded 28% and 40% increases in yield, plus an 18% increase in straw (important for farmers in this region as cattle fodder). In 2007, the yield increases for 25 farmers making comparisons were 95% with irrigated wheat and 63% without irrigation. See: http://business.outlookindia.com/inner.aspx?articleid=2162&edition id=58&catgid=2&subcatgid=973

  PRADAN and other NGOs have been expanding the System of Wheat Intensification (SWI) in the state of Bihar. From 415 farmers in 2008-09, the number expanded to 48,521 farmers two years later. SWI yields have averaged 4 tons/ha compared to 1.6 tons/ha with usual methods. In 2011, one farmer achieved a yield of 13.58 tons/ha without use of any chemical fertilizers, setting a record for organically-grown wheat (http://www.newstrackindia.com/ newsdetails/2012/04/12/321--Bihar-farmer-sets-organic-wheat-production-record-.html). In 2012, wheat area under SWI management was 183,000 ha, with an average yield 5.1 tons/ha.

  In Ethiopia, on small plots in Tigray province, farmers have gotten SWI yields of 9-10 tons/ha. In 2009, farmers applying SWI practices to durum wheat in Gembichu province had yields ranging from 1.25 tons/ha (the national average) to 8.5 tons., Also, farmers in Mali and Nepal have also experimented successfully with SRI methods for their wheat. So we have reason to believe that potentials for higher yield with modified management of wheat are similar as with rice (http://sri.ciifad.cornell.edu/aboutsri/othercrops/wheat/index.html).

• Finger millet: Two Indian NGOs have been working with this cereal crop which is very important for millions of poor households: the Green Foundation in Karnataka state, and PRADAN in Jharkhand and other eastern Indian states. Farmers have achieved yield increases of 100-200% by adapting SRI concepts and methods to this crop: young seedlings, wide spacing, soil aeration, increased organic matter, etc. People's Science Institute had a 33% increase in finger millet yield with SRI methods for 5 farmers in 2007, and a 60% increase for 43 farmers in 2007. As so often, first-year results can be improved upon as farmers gain insight, confidence, experience, and skill.

  In Tigray province of Ethiopia, an elderly woman farmer independently using practices very close to SRI for her finger millet crop got a yield of 7.5 tons/ha, many times the usual yield in her region (report from Institute for Sustainable Development, Addis Ababa). Since then, adaptations of SRI ideas to finger millet in Tigray have been spreading, with yields usually of 3.5 to 4 tons/ha, so that this methodology has become standard practice in the Axum region. We are hopeful that finger millet too, which is so critical a crop for so many poor people, can be enhanced with SRI methods.

• Sugar cane: A number of SRI farmers in Andhra Pradesh and Karnataka states began adapting SRI methods to sugar cane production by 2004, getting yields as much as 100 tons per hectare where before they got 30 to 60 tons. This encouraged the Worldwide Fund for Nature (WWF) together with the International Crop Research Centre for the Semi-Arid Tropics (ICRISAT) to prepare and release a manual for an application of SRI concepts to sugar cane, the Sustainable Sugarcane Initiative (SSI) in 2009. This offers at least 20% improvement in yield, and often more, with 30% less water and 25% less application of fertilizer and agrochemicals: http://www.indiawaterportal.org/sites/indiawaterportal.org/ files/SSI%20Training%20Manual_WWF_ICRISAT_2009.pdf

  A company known as AgSRI, based in Hyderabad, has been established to promote SSI as well as SRI on a larger scale, focusing on Maharashtra: http://www.agsri.com/ssi.html. Farmer experiences with SSI are reported in an AgSRI publication: http://www.agsri.com/ images/documents/ssi/AgSri_SSI %20casestudy%20book%2015-03-12.pdf AgSri has begun advising on the introduction of SSI in Cuba, Belize and Tanzania.

• Beans and pulses: In 2006, the People's Science Institute in northern India reported five farmers got an average yield increase of 43% when they adapted SRI methods for rajma, the local name for kidney beans. With 113 using these methods in 2007, learning from the first year's experience, the average yield difference was 67%. This helped launch a number of similar adaptations as farmers working with PSI have extended their experimentation and evaluation to still other crops such as soyabeans, maize, peas, lentils, and sesame in Uttarakhand and Himachal Pradesh states. In Karnataka state, the Agriculture-Man-Environment Foundation (AMED) has worked with farmers applying SRI methods to red gram: http://sri.ciifad.cornell.edu/aboutsri/othercrops/otherSCI/InKarnSCI RedGram _AME2011.pdf A book on SRI applications to green gram has been published in Germany: http://www.amazon.com/System-Crop-Intensification-Greengram-Innovative/dp/ 3847372769
• Tef: The typical yields of this cereal grown and eaten in Ethiopia are around 1 ton/ha, as the crop is grown 'extensively,' i.e., broadcast, with little management. When tef is growth with adapted SRI practices (transplanting young seedlings 20x20 cm, using more organic fertilization, applying small but regular water, etc.), yields are 3 to 5 tons/ha, or even higher. After several years of successful demonstration, a new government department, the Agency for Transformation of Agriculture (ATA) is working to popularize what is called STI; see reports on SRI website: http://sri.ciifad.cornell.edu/aboutsri/othercrops/teff/index.html. In 2012-13, as many 7,000 farmers are using full STI methods with transplanted seedlings, while over 100,000 are using a less-demanding, direct-seeded version of SRI.
• Mustard: This crop also known as rape, rapeseed or canola is also very responsive to SRI management, with yields around 3 tons/ha but some even higher, rather than 1 ton/ha which is usual (http://sri.ciifad.cornell.edu/aboutsri/othercrops/otherSCI/index.html#mustard).

These kinds of adaptations can be grouped under the heading of SCI, the System of Crop Intensification, or as noted above, these are called versions of the System of Root Intensification. In Bihar state of India, SCI ideas have been extended to tomatoes, chillies and eggplant: http://sdtt-sri.org/wp-content/themes/SDTT-SRI/Document/output.pdf Some farmers in Tamil Nadu state have developed a 'system of turmeric intensification' (STI) applying SRI ideas, doubling their net income per hectare of this spice crop grown from rhizomes.

In Ethiopia, the application of SRI concepts and practices to a variety of rainfed crops, including sorghum, maize, barley, lettuce and cabbage, is referred to as Planting with Space. This name is easier to explain than 'SRI.' See manual produced by the Institute for Sustainable Development: http://www.isd.org.et/Publications/Planting%20with%20space.pdf

Even greater imagination has been shown by some farmers as they have extended their SRI experience and thinking beyond plant management, to raising the collections of lac, an entomological crop, in Jharkhand state of India, and to raising chickens in Cambodia. Better management of fewer birds yields households more meat and eggs from a smaller flock. Getting 'more from less' they see as a manifestation of SRI principles.

All of these management strategies aim to capitalize on biological processes and potentials that are forgone with the kind of 'industrial' thinking that shapes current recommendations for improving agriculture. Farmers in various countries are finding productive, quick and inexpensive ways to improve their agriculture by reflecting on their SRI rice experience and extrapolating ideas to enhance their production and income from other crops.

The profuse tillering of SRI plants can be explained in part by an understanding of phyllochrons. When Fr. Laulanié learned, almost by accident, that transplanting very young seedlings can give much more robust and productive mature rice plants, it seemed somewhat mysterious. But four years later he happened to read a book on rice science by Didier Moreau (1986) and learned about the research on phyllochrons done by a Japanese crop scientist T. Katayama during the 1920s and 1930s. Unfortunately, his findings were not published until after World War II (Katayama 1951) and still have not been translated into English.

Katayama discovered a periodicity in the emergence of tillers (stalks) by studying rice, wheat and barley. This arises from the way that all grass-family (gramineae or poaceae) species grow. Understanding phyllochrons can account for why rice seedlings, if they are transplanted before 15 days of age, can give a different and greater growth response than do seedlings transplanted at an older age, after the start of the 4th phyllochron.

The most detailed discussion in English of phyllochrons is by Nemoto et al. (1995), but for a short presentation, see Stoop et al. (2002). Surprisingly, there is no mention of phyllochrons in the Oxford dictionary on plant science with over 6,000 entries (Allaby 1998). There is a 4-page entry on phyllochrons in the English translation of a three-volume Japanese encyclopedia of rice science (Matsuo et al. 1993).

The term 'phyllochron' comes from combining two Greek words meaning leaf + time. The word denotes a period of time during which a leaf (or more than one leaf) together with its associated root and tiller (or multiple associated roots and tillers, beyond the start of the 4th phyllochron) emerge from the same meristematic tissue, which is located at the soil surface between the plant's root system and its above-ground canopy.

A functional unit of leaf, tiller and root, collectively referred to as a phytomer, grows both upward and downward from the plant's meristem in the crown of the plant. As plants' shoots (i.e., their leaves and tillers) grow upward, their roots which emanate from the same cell-division process as do the leaves and tillers grow downward into the soil.

The length of a phyllochron for rice can vary considerably, from:

• 8-10 days if the growing conditions are unfavorable, with many stresses on the plant, to
• 4 days if conditions are ideal, and the plant encounters no stresses that slow its growth.

If growing conditions are good (favorable temperatures, enough water and sunlight, adequate soil nuitrient availability, good space and friable soil for root growth, etc.), a phyllochron can be 5 to 6 days in length, and the plant can go through 10, 11 or 12 phyllochron periods before it reaches the end of its period of vegetative growth. At this point, the plant switches into its reproductive period, with a progression through several stages, from panicle formation to flowering, then to grain filling, ripening and maturation, at which time the grains can be harvested.

• During rice plants' 1st phyllochron, i.e., the first cycle of tiller and root emergence, the first phytomer (functional unit of leaf/tiller/root) is produced from the seed. The tap root grows downward, and the main tiller (with flag leaf) grows upward.
• During the 2nd and the 3rd phyllochrons, the plant does not produce any additional phytomers (units of roots, leaves and tillers) from the apical meristematic tissue at its base from which the plant's above-ground and below-ground structures (organs) emerge. This is a time of apparent dormancy. If growing conditions are reasonably good and the phyllochron length is 5 days, this period extends from about the 5th day to about the 15th.
• During the 4th phyllochron, a second tiller, in association with a leaf and a root, emerges from the base of the original main tiller. This is called the first primary tiller.
• During the 5th phyllochron, another phytomer containing the second primary tiller emerges. bringing the total number of tillers to three.
• During the 6th phyllochron, an acceleration of growth begins, as now two more phytomers are produced from the meristematic tissue in the culm: a third primary tiller from the base of the main tiller, and a first secondary tiller from the base of the first primary tiller.
• During the 7th phyllochron, there are three more phytomers emerging: a 4th primary tiller, along with two more secondary tillers from the base of the 1st and 2nd primary tillers, respectively.
• During the 8th phyllochron, there are now five more phytomers emerging: one more primary tiller (the 5th), secondary tillers from the 2nd, 3rd and 4th primary tillers, and a first tertiary tiller from the 1st secondary tiller branching off from the 1st primary tiller. This sounds complicated, but it can be understood visually by looking at the diagram which Fr. Laulanié prepared from studying the work of Katayama.

The structure and logic of this diagram becomes clearer by studying the numbers shown in the following table. At present, few rice plants currently complete a full 12 cycles (phyllochrons) of growth before they start to flower and begin forming spikelets (potential grains) which if fertilized and properly nourished grow into grains. When phyllochrons are longer than 5 days, e.g., 7 or 8 days, because less than fully favorable environmental conditions prevail, the plant's growth is slower. Fewer phyllochrons of growth, with their predetermined numbers of tiller and roots emerging, will be completed before the rice plant switches from its vegetative growth period to its period reproduction (grain formation and filling).

Whether a plant can maintain a rapid rate of growth and can complete 10, 11 or 12 phyllochron cycles before flowering, with 33, 53 or 84 tillers respectively, depends on the conditions under which the plant is growing. If plants are crowded together, they receive less sunlight and are competing for soil nutrients; so the kind of semi-exponential growth seen in the figure and table above is not possible. Also, if the soil is kept flooded, the plant's roots degenerate for lack of oxygen and they cannot support rapid growth.

Some readers will notice that the pattern of tillering shown in the table corresponds to what is known as a Fibonacci series. In such a series, the number emerging in each period is the sum of the previous two periods: 1+1=2, 1+2=3, 2+3+5, 3+5=8, … (Such a series or sequence was made famous by the 2003 best-selling book by Dan Brown, The DaVinci Code.) The number of tillers produced in each period is approximately 2/3 more than emerged in the preceding period.

A rice plant that is able to complete 12 phyllochrons (cycles) of growth before the end of its vegetative growth phase -- when it moves into its reproductive phase, which includes panicle initiation, flowering and grain-filling -- could have as many as 84 tillers. It would also have a comparably profuse associated root system below-ground because its roots emanate from the same meristem cells that go through cell division to create the tillers and leaves.

If rice plants are transplanted during the 4th phyllochron or in even later phyllochrons, it is seen that the production of phytomers is decelerated and diminished. Accordingly, the rice plants when they begin their reproduction will have fewer tillers and fewer leaves, and also fewer roots.

What is described here is a mechanistic presentation of a biological process. Specific plants do not necessarily grow as mathematically as the model above indicates; the length of phyllochrons is not always uniform, for one thing. Rice tillers increase according to a modified Fibonacci series, not a perfect one (12 periods should hypothetically produce 89 tillers, rather than 84), in part because physical congestion in the culm (base) of the plant keeps the emergence of tillers (and roots) from the meristematic tissue from reaching the theoretical maximum.

Transplanting rice seedlings during their 2nd or 3rd phyllochron of growth -- roughly between the 5th and 15th days -- represents a window of opportunity for best management of rice plants. Roots will be less traumatized if transplanted during this relatively dormant period, and the plants when they resume their growth after transplanting will produce more phytomers in an accelerated way.

When transplanting occurs later than about the 15th day (the exact date is affected by the length of the phyllochrons, which is a variable length), rice plants will not experience as much growth or as rapid growth. The factors that shorten -- or lengthen -- phyllochrons have been discussed in Nemoto et al. (1995). The table below presents them in an analytical way. The concept underlying the analysis is that the plants' grow according to a kind of 'biological clock' that runs faster or more slowly, depending on the totality of favorable or unfavorable growth conditions:

 
 
 
 
 

More research remains to be done on phyllochrons and their effects on growth. There has been considerable research on phyllochrons in wheat, e.g., special issue of Crop Science (1995, 35:1), and on forage grasses, especially in Australia. However, there has been little consideration of rice phyllochrons except by rice scientists in Japan and China, where they are well-known. In the English-reading world, phyllochrons do not figure much in plant science considerations.

Research has been done along similar lines in terms of 'degree-days,' but these are not linked to an understanding of plant physiology and morphology as closely is an analysis done in terms of phyllochrons. For SRI, understanding phyllochrons helps explain why the use of young seedlings has such a strong positive effect, validated empirically (Uphoff and Randriamiharisoa 2002). The rapid tillering and root growth which is possible when the full set of SRI practices are used together is not seen when older seedlings are used and when rice plants are grown under continuously flooded conditions with degenerating plant roots, which lengthens phyllochrons. We hope that this area will become the focus of extensive research.

One of the first published reports on SRI in Madagascar (Moser and Barnett 2003) reported that there was a high rate of disadoption of SRI methods, as much as 40%, after farmers had tried out the methods, mostly because of the greater requirement of labor inputs for SRI. It was reported that especially very poor farm households, which need continuous income flow to survive, said that they could not afford to practice SRI, even though they knew that the methods would give higher yield. This article, based on well-designed and well-conducted field research, established the idea that many readers accepted SRI is 'too labor intensive' and is often given up by farmers after they try it.

As seen above, however, there is considerable evidence that SRI is, in fact, labor-neutral or labor-saving, rather than being necessarily labor-intensive (FAQ #9). Also, there has been other evidence from Madagascar which suggests that disadoption is not a big problem in that country. This comes from an evaluation of a large French-funded irrigation project on the High Plateau (Hirsch 2000) that was implemented in the same period as the USAID project in Ranomafana that got CIIFAD involved with Association Tefy Saina and with SRI.

Around Ranomafana Park, farmers using SRI methods had average paddy yields of 8 tons/ha, compared to their usual 2 tons/ha on the same fields with the same varieties when using conventional methods. As seen below, the French project reported similar results as those in Ranomafana. Comparisons were made with farmer practice and with SRA, the Système de Riziculture Ameliorée (SRA, the System of Improved Rice Production). The latter involves use of inorganic fertilizer, row planting, continuous flooding, the 'modern' practices recommended by government scientists. There is little indication of disadoption of SRI in these data, even though SRI extension was not being promoted by a formal program as in Ranomafana, mostly spreading farmer-to-farmer.

Disadoption has not been reported as a noticeable problem for SRI elsewhere except in Andhra Pradesh state of India. While there are no systematic data on disadoption in the state, the most common reason reported for farmers stopping their use of SRI methods has been unreliability of electricity supply (Adusumilli and Laxmi 2009). This is a deterrent to adopting SRI and to continuing with it in Andhra Pradesh, since water control is important for SRI and electricity is the main source of power for the pumps that provide fields with irrigation water. This is not a shortcoming of SRI itself, but of the infrastructure that serves rice farmers. It reflects not on the agronomics of SRI, but on the pragmatics of using its methods.

• SRI use has dramatically increased in the Indian state of Tripura, as seen below. After several years of on-farm evaluation, adjusting SRI practices to local conditions, the state government decided in 2005 to support its dissemination. See: http://www.indiatogether.org/2008/jan/agr-sritrip.htm The head of the SRI extension effort in Tripura, Dr. Baharul Majumdar, a senior rice agronomist with the Department of Agriculture, says that he has not heard of any Tripura farmers disadopting SRI methods once these have been tried. The table below shows the uptake of SRI in Tripura. As often happens, the yield improvements have declined somewhat with the rapid expansion and lower quality of training. But large increases are being attained, mostly under rainfed condition. With reductions in farmers' costs, their net incomes increased by more than their higher yields.

• In Cambodia, the introduction of SRI started in 2000 with 28 farmers, encouraged and supported by the NGO CEDAC. By 2011, this number had increased to over 190,000 farmers. In 2007, CEDAC conducted an in-depth evaluation of SRI adoption and non-adoption in a sample of 21 villages in three districts where SRI had been available for at least five years. The survey team interviewed 348 adopters and 292 non-adopters in these villages. The study determined that 46% of the households in these villages were using SRI methods, with demonstrably good results.

  Of relevance here, the survey determined that the average number of disadopting households, ones which had tried SRI methods and had given them up, was only 1 per village -- out of 200 households (CEDAC 2008). So in Cambodia, where there is extensive experience with SRI, the rate of disadoption in rice-growing areas has been negligible, about 0.5%.

  The CEDAC evaluation was useful in identifying which methods within the SRI system of management have been less-adopted or were disadopted because farmers found them difficult, such as frequent soil-aerating weeding. The study has helped CEDAC make further improvements in its extension strategy. The study showed that only some Khmer farmers are using all of the SRI methods recommended, which helps to explain why the SRI yields achieved in Cambodia have been generally lower than have been obtained with SRI methods on farmers' fields elsewhere. The main reason for lower average SRI yields in Cambodia is that over 80% of the farmers have no irrigation facilities and rely on rainfall. Regarding disadoption, the study confirmed the observations from other countries that once SRI practices have been learned and used by farmers, not many give up this approach to rice production.

• In India, the National Consortium for SRI (NCS) has commissioned a three-year study on reasons for disadoption, being carried out by PRADAN, which will survey 1500 farmers across several states. The first year of data gathering (2011-12) covered 210 farmers in two districts in Bihar and Orissa states. The rate of disadoption was determined to be 12%; but none of the 25 disadoptions were voluntary. In most cases, farmers had been deterred by severe drought, and in other cases family illness reducing the labor force was reported (Barah 2012). The full study when completed should give a thorough understanding of the extent of and reasons for disadoption. That SRI practices might not be suitable for every farmer does not invalidate the benefits they can bring to large numbers of farmers.

SRI is a civil society innovation that did not originate through the usual channels of agricultural scientific research. This may account, at least in part, for the resistance among many scientists to being accepted. SRI was assembled through several decades of work with farmers by Fr. Laulanié, who sought simple, low-cost, accessible ways for them to increase the productivity of their land, labor, water and capital when growing irrigated rice. The insights that Laulanié gained from working with rice plants and farmers can be extended to unirrigated rice production and to growing other crops, as discussed above. Clearly, SRI is not a typical 'technology' that can or should be extended in the same way that most agricultural innovations have been spread in recent decades.

There was initially little interest in SRI from government agriculturalists in Madagascar or from the International Rice Research Institute (IRRI) representatives in that country; so responsibility for further evaluation of SRI and its dissemination fell first to the NGO that Fr. Laulanié and Malagasy colleagues established in 1990, Association Tefy Saina (http://www.tefysaina.org/). This NGO had links with a variety of other NGOs, church groups, and interested individuals. Since 2008, there has been a Groupement SRI Madagascar (G-SRI) which has grown to over 260 partners and affiliates with a trilingual website (http://groupementsrimada.org/), supported with grants and technical assistance from Jim Carrey's Better U Foundation of Los Angeles.

Starting in 1994, Tefy Saina and the Cornell International Institute for Food, Agriculture and Development (CIIFAD) (http://ciifad.cornell.edu) started working together on the evaluation and demonstration of SRI methods. In 1998 they began working with faculty and students at the University of Antananarivo. Then in 1999, a small grant was received from the Rockefeller Foundation through CIIFAD for SRI evaluation by a consortium involving Tefy Saina and its university partners with a researcher from FOFIFA, the government's agricultural research organization. Such multi-sectoral collaboration has been typical of the way that SRI has been spread ever since.

CIIFAD in collaboration with Tefy Saina established and maintains an SRI website that has served to disseminate information on SRI around the world (http://sri.ciifad.cornell.edu/). NGOs that support low-input sustainable agriculture such as LEISA and ECHO have assisted this process by publishing articles on SRI, e.g., Rabenandrasana (1999) and Berkelaar (2001).

Representatives of Tefy Saina and CIIFAD and, increasingly, partners in the various countries have spoken on SRI at international and national forums and seminars such as:

• Southeast Asian regional conference on Sustainable Agriculture and Natural Resource management in Chiangmai, Thailand, organized by the University of Hohenheim (2002);
• International Rice Congresses in Beijing (2002), New Delhi (2006), and Hanoi (2010), organized by International Rice Research Institute;
• Latin American regional rice meetings convened in Cuba (2002 and 2008), and a regional organic agriculture conference (2003);
• World Rice Research Congress at Tsukuba, Japan (2004), organized by IRRI and the Japanese Ministry of Agriculture, Forestry and Fisheries (MAFF);
• International Farming Systems Association conferences in Orlando, FL (2004) and in Rome (2005);
• 21st International Rice Commission meeting in Chiclayo, Peru, invited by FAO (2006);
• 13th meeting of the U.N. Commission on Sustainable Development in NYC (2008) and a UNCSD inter-sessional meeting in Namibia and a preparatory meeting in NYC (2009);
• 8th and 10th meetings of the Paddy and Water Environment Engineering Society (PAWEES) in Bogor, Indonesia (2009), and Taipei, Taiwan (2011);
• ECHO agricultural annual conference, Ft. Myers, FL (2010); and regional agricultural conferences in Chiangmai, Thailand (2009) and Ouagadougou, Burkina Faso (2011).
• Biovision International Conferences at the Biblioteca Alexandrina in Alexandria, Egypt (2010, 2012);
• Soil and Water Conservation Society annual meetings, St. Louis, MO (2010) and Washington, DC (2011);
• International Symposium on Abolishing Hunger, College de France, Paris (2011); and
• 2nd Global Conference on Agricultural Research for Development in Punta del Este, Uruguay (2012).

Country by country, persons have come forward with an interest in raising productivity and incomes for rice farmers and in doing this with protection/enhancement of the environment.

The backgrounds of these persons who have given national leadership for SRI have been diverse: an agricultural technician working with an international NGO in northern Afghanistan; a retired agricultural economics professor in Bangladesh; a agronomy-PhD leader of a national NGO in Cambodia; a Canadian agronomy PhD student doing thesis research in Panama; a number of senior rice scientists in China; a rice farmer in Costa Rica; a retired animal nutritionist in Cuba; an agricultural NGO founder in Ecuador; an agricultural research station director in The Gambia; the director of extension for a state agricultural university in India, as well as a business school professor, a retired agricultural-economics research director, and many NGO personnel; a senior rice researcher in Iraq, and a more junior rice researcher in Iran; the team leader for a Japanese private-sector consulting firm team leader working in Indonesia; university professors in Kenya, Indonesia, Malaysia and Korea; an agricultural-extension specialist at district level in Nepal; a farmer/businessman/inventor/ philanthropist in Pakistan; a private agricultural consultant in Peru; an electrical engineer heading an environmental NGO in the Philippines; a senior civil servant, a farmer-environmental activist, and a deputy minister of agriculture in Sri Lanka; an NGO agricultural activist in Zambia; and so it goes. Leadership for SRI has been heterogeneous by disciplines, roles, statuses and age, but with shared interest in finding low-cost, widely accessible, environmentally-friendly ways to raise agricultural production, especially for food-insecure households.

A number of SRI colleagues have travelled to other countries to help transfer SRI knowledge: from Sri Lanka to India; from India and Bangladesh to Afghanistan; from China to North Korea; from Madagascar to Rwanda; from Indonesia to Malaysia and the Solomon Islands; from Cambodia to Vietnam; from India to Morocco and Tanzania; and so forth.

Colleagues in a number of countries have hosted visitors from other countries to share SRI knowledge: Madagascar (Indonesia and Sierra Leone); Sri Lanka (India and Pakistan); India (Bangladesh); Cambodia (Vietnam). Norman Uphoff, former director of CIIFAD, who has more opportunities for travel than most persons have, has made presentations on SRI in >40 countries.

Leadership for SRI evaluation and dissemination in any country, state or district can come from any sector: from government agencies, NGOs, universities, research institutes, private firms, or farmer organizations. In each country, there usually has emerged a cooperative network of like-minded persons from these different sectors. Persons and institution contributes respectively to the advancement of SRI understanding and practice according to their respective comparative advantages.

• In Kenya, starting from two professionals, respectively at the Jomo Kenyatta University of Agriculture and Technology and in the World Bank office in Nairobi, persons from a number of other organizations came together to launch an SRI campaign in that country: the National Irrigation Board (NIB); the African Institute for Capacity and Development (AICAD), an NGO; the NIB's Mwea Irrigation Scheme and its Mwea Irrigation Agricultural Development Centre (MIAD); the Improving Management of Agricultural Water in Eastern and Southern Africa, at the time a network of ICRISAT funded by IFAD; the Ministry of Water and Irrigation (MWI); the Ministry of Agriculture (MoA); the Central Kenya Dry Areas Project; private consultants; the World Bank Institute (WBI) in Washington, DC; Cornell University in USA; and progressive farmers from the Mwea Irrigation Scheme. This broad coalition has enabled rapid expansion in Kenya: http://sri.ciifad.cornell.edu/countries/kenya/index.html

Once SRI has been introduced and demonstrated, various donor agencies, foundations and international NGOs have begun to support SRI extension:

• GTZ in Cambodia;
• World Vision in Sierra Leone;
• ADRA in Madagascar and Indonesia;
• Catholic Relief Services in Madagascar;
• Oxfam-America in Cambodia and Vietnam; also supporting SCI (other crops) evaluation and publications in Ethiopia;
• Oxfam-Australia in Laos and Sri Lanka; Oxfam-Great Britain in Bangladesh and later Philippines; and Oxfam-Quebec in Vietnam;
• ActionAid in Bangladesh;
• The Rockefeller Foundation and the World Bank helped to fund participation in the first international SRI conference, held in China in 2002;
• The World Bank has funded SRI expansion through large projects in Tamil Nadu state of India and then in Bihar state;
• The World Bank Institute has supported SRI extension in India, Kenya and Malawi and in Southeast Asia through distance learning media;
• The Medco Foundation in Indonesia;
• The Better U Foundation in Madagascar, Mali and Haiti;
• The Asia Foundation and American Friends Service Committee for DPRK;
• The Bill and Melinda Gates Foundation for spread of STI (tef) in Ethiopia;
• A few private individuals have also given financial support for SRI's spread as 'angels.'

Jim Carrey's Better U Foundation has been the only foundation to give direct support for SRI extension around the world, through a gift made to CIIFAD in 2010 to help establish the SRI International Network and Resources Center (SRI-Rice) at Cornell University. SRI-Rice, among other things, maintains the international SRI website: http://sri.ciifad.cornell.edu

In a number of countries, farmers have emerged as SRI spokespersons and trainers, giving of their own time and money to promote SRI practices to dozens and even hundreds of fellow farmers; a few have even reached more than a thousand farmers at their own expense.

The extension of SRI has thus had extremely diverse support, notable for its voluntary nature. Cornell's support for the overall communication and exchanges has come from CIIFAD and from individual contributions of time and money.

Relative to the limited amount of institutional and financial support that this far-flung, voluntaristic SRI campaign has received, the amount of impact that it has already achieved is unprecedented. But then, SRI is itself unprecedented as an innovation. It is possible that this SRI experience will open up new styles, strategies, roles and relationships for agricultural extension and development.

With SRI, farmers not regarded as adopters or as recipients of scientific and technological advice. Rather, they are partners and innovators in the process of SRI adaptation and further evolution. Farmers, by themselves or in cooperation with NGO, government or other partners, have been making substantial improvements in SRI, particularly to reduce its labor requirements. This stems from regarding SRI not as a technology that is finished and fixed, but instead seeing it as a work in progress.

Usually in the 20th century, technological advances followed from improvements first made in the domain of scientific knowledge. In the case of SRI, on the other hand -- as with the invention of the airplane -- it preceded the science that could explain it. Most scientists and technologists seem to have had some difficulty in accepting that SRI was not like other technologies; indeed, SRI proponents avoided using the designation 'technology' as discussed above (1.2). Many persons tried to conceive and evaluate SRI as they would a new variety or a new input, not appreciating that it was more mental than material, that its results were more organically than mechanically produced, more a matter of ideas than of inputs. It did not fit aptly in the usual categories used for scientific thinking, so SRI encountered some resistance and opposition from the start.

Early on, SRI received endorsements from two of the world's most eminent rice scientists: Dr. M.S. Swaminanathan, often called as 'the father of the Green Revolution in India' and a former director-general of IRRI; and Prof. Yuan Long-ping, known as 'the father of hybrid rice' in China and around the world (Yuan 2002). But for some time, 'mainstream' rice scientists still withheld their approval, and this affected the willingness of governments and foundations generally to support, or even evaluate, SRI effects.

It was thus left to civil society, broadly defined, to investigate and disseminate SRI practices. Also, as noted below, a number of agricultural scientists in different countries who understood the ideas behind SRI did their own assessments. Once they were satisfied that SRI methods enhance crop productivity, they often began cooperating with others engaged with SRI – farmers, extensionists, NGO workers, private sector agents. In country after county, policy makers have begun to respond to the new opportunities as they see benefits available for farmers, consumers and environment, immediately and at low cost.

When Fr. Laulanié first introduced SRI to scientists at the University of Antananarivo and in the Ministry of Agriculture in 1990, we are told that there was disbelief and derision. The IRRI representative in Madagascar was not interested in SRI when Prof. Uphoff of CIIFAD asked him about it in December 1993. Uphoff found no more interest among colleagues at Cornell when they were told about SRI in 1996, after two years of results were in hand. In 1998, Uphoff discussed SRI with the newly-appointed Director-General of IRRI, Dr. Ron Cantrell. The conversation was cordial, but after Uphoff gave a seminar on SRI at IRRI in February, 1999, there was no evident interest in cooperating in evaluating the methods. After Uphoff gave another seminar on SRI at IRRI in March 2003, several scientists working at or associated with IRRI began publishing critiques of SRI (see Dobermann 2004; Sheehy et al. 2004; Sinclair 2004; Sinclair and Cassman 2004; for a review of this controversy, see Uphoff 2012).

Fortunately, scientists in several national agricultural research institutions were more receptive to SRI's new ideas. Their willingness to evaluate the new methods helped to establish the scientific foundations for SRI. This process started in 1999 with evaluations done at Nanjing Agricultural University in China and at the Agency for Agricultural Research and Development rice research station in Indonesia. Further evaluations were done by scientists at the China National Hybrid Rice Research and Development Center (CNHRRDC) and the China National Rice Research Institute (CNRRI), and then in India, Thailand, and Indonesia. A project set up by Wageningen University researchers with Dutch government support in 2000 got scientists in China, India, Indonesia and Madagascar started on SRI evaluations.

Already starting in 1998, top students in the University of Antananarivo's Faculty of Agriculture (ESSA) in Madagascar began doing baccalaureate thesis research projects under its director of research, the late Prof. Robert Randriamiharisoa. Factorial trials clearly showed the merits of SRI practices both respectively and collectively (Randrimiharisoa and Uphoff 2002; Uphoff and Randriamiharisoa 2002).

In India, evaluations of SRI began at Tamil Nadu Agricultural University in 2000, and expanded evaluation began in Andhra Pradesh (AP) state in 2003. With support from the World Wide Fund for Nature (WWF) through a joint program with ICRISAT on food, water and environment, the evaluation in AP expanded further, involving scientists from ANGRAU, the AP state agricultural university; from the Directorate of Rice Research (DRR) of the Indian Council for Agricultural Research (ICAR); and from ICRISAT, all based in Hyderabad. The Ministry of Agriculture's Directorate of Rice Development (DRD) in Patna also undertook its own evaluations in this period and became supportive of SRI based on results. In 2007, the Government of India allocated $40 million for dissemination of SRI practices in >130 food-insecure districts under its National Food Security Mission, with the endorsement of the Ministry of Agriculture and ICAR.

The rice research programs in Iraq and Iran, at Najaf and Amol respectively, began their own evaluations in 2005 and demonstrated the benefits of SRI practices to the satisfaction of their scientists (see Iraq and Iran country pages on SRI website: http://sri.ciifad.cornell.edu/). There was, however, resistance encountered from rice scientists in some countries, such as Sri Lanka, Bangladesh and Cambodia who were influenced by the criticisms from international rice scientists. Over time, even this scientific opinion has become more favorable toward SRI, given that the national rice research systems in China, India, Indonesia and Vietnam, where over 2/3 of the world's rice is produced, became supportive of SRI based on their assessments of evidence.

A proposal put together in 2008 for a joint evaluation of SRI planned by researchers from IRRI, Cornell University and Wageningen University, which would follow agreed-upon protocols in at least three countries of Asia with differing agroecological conditions, was submitted to the Bill and Melinda Gates Foundation in 2009; but this was not ultimately funded. Instead, a grant was made to Wageningen for a review of SRI extension experience. Wageningen also received funding from the Dutch Government for a four-year evaluation of SRI experience in India, including PhD training programs for four Indian professionals, which is still in progress.

Meanwhile, more and more Master's and PhD theses are being done on SRI at universities around the world, and more and more articles are being published on SRI in peer-reviewed journals. These established a better understanding of how and why SRI's innovative practices produce the effects reported here (e.g., Horie et al. 2005: Mishra et al. 2006; Mishra and Salokhe 2008). SRI represents a paradigm shift, from making agricultural advances based on genetic improvements and the application of external inputs, to increasing productivity through modified management practices that promote root growth and stimulate beneficial activity from soil biota.

Paradigm shifts seldom come smoothly, even (especially?) in the realm of science. In the case of SRI, the shift could come more quickly because the validity of the new ways of thinking will not be decided just by inner and outer circles of scientists, but more importantly by the actions and assessments of millions of farmers, eventually hundreds of millions. When they give SRI a vote of confidence by their use and continuation of the methods, any dismissals and disparagements in the peer-reviewed literature carry little weight, even among most scientists. It is an open question is whether the paradigm shift will be limited to rice production, or will have ramifications for agricultural science and practice more generally. It could forge some better pathways to food security and prosperity in the 21st century. This presents many challenges of land and water limitations, rising costs of energy, climate change, environmental degradation, and immense unmet human needs.

The earlier limited involvement of scientists in SRI's creation and development is changing, and the 'controversy' has abated as evidence and peer-reviewed journal articles keep accumulating. Before 2002 there were few published articles; in 2011-12, the number published annually was over 60; and the total number is now nearing 400, most of which can be accessed from the SRI-Rice website (http://sri.ciifad.cornell.edu/research/JournalArticles.html).

A strength and a weakness of SRI is that it is that it is not a fixed or a single thing. It continues to evolve, change and expand as more experience is gained and as more and different persons engage with its ideas and effects. We refer to SRI as a methodology rather than as a technology, and Fr. Laulanié preferred to call it a system. While SRI has biological/biophysical bases and manifestations, at the same time it is a social/socioeconomic reality. It is referred to by some as a movement, encompassing hundreds and now thousands of persons from many disciplines and from even more countries. This aggregation of like-minded people complements, activates and learns from the agronomic factors and relationships that are manifested in SRI in material ways.

The most summary and inclusive term to use for SRI is probably phenomenon. In any case, there is something real and existing that is covered by the acronym, even as it morphs into things like the System of Root Innovation, or the System of Crop Intensification. Nobody can say what SRI will look like or refer to 10 years from now, but there are definite bases for some conjecture.

• Broader agroecological development: SRI is related to a broad set of management systems that can be grouped under the term agroecology. This includes agroforestry, integrated pest management, and particularly conservation agriculture (CA). CA is gaining acceptance in most countries around the world, with over 125 million hectares now under such cultivation. CA is a progressive evolution of what started decades ago as no-till agriculture, now linked with maintaining permanent ground cover (mulch, cover crops, etc.) and optimizing crop rotations, getting away from monocropping as well as from repetitive disruption of the soil (Kassam et al. 2009).

  SRI was developed without making any changes in land preparation methods, trying not to require too much change of farmer practices at one time. But farmers in several countries have already begun experimenting with SRI on permanent raised beds; and in Pakistan, there has been a merger among SRI, CA and organic agriculture, saving both water and labor through mechanization. The system is referred to by Sharif (2011) as 'paradoxical agriculture' because it produces more output with less input. There are substantial benefits from not continually ploughing up the topsoil, including enhancement of soil biodiversity and soil fertility. SRI practitioners are learning how to tap these benefits, letting soil organisms like earthworms carry out the active soil aeration that is now accomplished by using mechanical hand weeders.

• Broader applications within agriculture and horticulture: We are already observing the extension of SRI concepts and methods to a wide range of other crops, as seen in #13 above, and not just for cereal crops or gramineae species. So far we have not seen SRI principles applied to fruits, but they have shown they can raise the productivity of various vegetables as well as pulses and legumes. The application of SRI to sugarcane (SSI) is proceeding very well in India, and has now started in Cuba. So, what began from inductive work devoted to irrigated rice is now spreading throughout agriculture and horticulture.
• Linkages with soil biology: With the spread of SRI to more crops, we should see more links forming between agricultural science and microbiology generally. We are now learning how symbiotic endophytes -- microorganisms that normally reside in the soil but that can enter and live in plant tissues and cells in mutualistic relationships above-ground – can contribute to plant growth and health (Uphoff et al. 2012). Agriculture has been heavily dependent for its increases in output on direct inputs of agrochemicals. As such, SRI does not promote 'organic' agriculture, which is often conceptualized simply as 'anti-chemical.' Rather, SRI experience justifies a strategy for agricultural production that is 'pro-biological' -- being pragmatically 'organic' rather than doctrinally so.
• New emphasis in genetic analysis: SRI does not make all further improvements in crop genetic potentials unnecessary. The great production increases that are possible by better exploiting existing potentials do not mean that no more improvements in genetic potential should be sought and made. Rather, SRI experience shows how important it is to advance knowledge and practice with regard to the expression of genetic potentials, not just continuing to increase those potentials. These are beneficial only to the extent that they can and will be expressed.

  The budding scientific field of epigenetics should be getting much more attention among agronomists and plant breeders that it presently receives from them, tending to be fixated on genes per se. SRI plant phenotypes show how huge can be the effect of altering plant growth environments, above and below ground – with impacts greater than the effects achieved from plant breeding as such. There is so much that needs to be better understood about managing plant environments to get desired results express in phenotypes.

  • Recall that in the community in Bihar state of India -- where five SRI farmers matched or exceeded the previous world record for paddy yield (19 tons/ha) in the 2011 kharif season using new hybrid varieties – when conventional management methods (older seedlings, close spacing, flooding of paddies, etc.) were used with these same hybrid varieties on the same farms, with the same soil and climatic conditions, the yields were only one-third as high (7 tons/ha).

  SRI offers plant breeders and physiologists a great opportunity to advance the science of epigenetics by studying what happens, where and why, in the expression of rice plants' genetic potentials when their growing environments are changed. Chi et al. (2010) have shown that symbiotic endophytes (soil rhizobia) when living in rice plant leaves and sheaths can affect the up-regulation and down-regulation of specific genes which produce proteins that affect photosynthesis and protection against pathogens. This is really important new knowledge.

• One of the most important impacts from the introduction and spread of SRI may well turn out to be changes in the way that agricultural research and development are conducted. In most of the 20th century, agricultural R&D followed what has been characterized as a linear model of organization. This regarded agricultural progress as derived primarily from agricultural research in formal institutions, which produced operational knowledge to improve productivity that would be passed on to extension agents and through them to farmers as end-users. While feedback loops were provided for in theory, in practice these were non-existent or weak. 'Technology transfer' was idealized and usually unidirectional.

  In recent decades, support has grown for what is called farmer-centered research and extension and for participatory technology development, recognizing the weaknesses of the established one-way systems set up for 'transfer of technology.' But except for the farmer field school (FFS) movement developed under FAO auspices for extending IPM knowledge and practice, there has been little real change in the thinking that guides agricultural R&D and in the way that research and extension institutions operate.

  The SRI experience originated from civil society rather than from research centers, and it advanced in the initial years without support from formal institutions for research and extension (except in Sichuan and Zhejiang provinces of China, Tripura and Tamil Nadu states of India, and Vietnam). There are many examples of how farmers have themselves made many improvements in the initial SRI recommendations, with labor-saving practices, new implements, extrapolations to other crops, etc. In countries where farmers have become operational partners with NGOs and state extension services in improving and disseminating SRI, it is unlikely that the formal R&D institutions will ever be quite the same again, because of the confidence built up in farmers in their own capabilities and because of the respect engendered among more-educated professionals for what farmers can contribute to their own and others' development.

• SRI experience is providing a foundation for creating more participatory and more egalitarian modes of agricultural R&D for the future, appropriate for the 21st century. This is expected to be more democratic than the preceding century, not least because farming communities around the world are now more educated and better informed than their predecessors. It is appropriate that the name which Fr. Laulaniè gave to the NGO that he established in 1990 with Malagasy friends (Association Tefy Saina) means 'improve the mind' (or mentality) rather than 'grow more rice.'

  For our further development, the human factor is ultimately the most important. Providing more abundant, healthier and cheaper food for people will surely make a significant contribution to expanding human capabilities and empowerment. But ultimately there need to be advances made in people's mentality, as constraints and injustices need to be dealt with in many realms, not just agriculture. We look forward to the contributions that SRI experience, expanded across much of the agricultural sector, can make to human confidence, imagination and solidarity.

  These qualities are needed to put humankind in a better position to deal with the many other serious problems that will need to be tackled for us to make it through this current century -- beyond hunger and poverty, which SRI can redress. The odds of success given climate change, failing states, inequality, and other adverse trends are not good. But the bundle of ideas, friendships and aspirations that have been aggregated through the dynamics, impacts and opportunities of SRI can, we hope, improve the odds. This discussion thus ends on a more philosophical note than an agronomic one.