A Wikiblog E-Book by Norman Uphoff with many others
Chapter 19: MAKING SRI MORE ATTRACTIVE AND
LESS LABORIOUS THROUGH MECHANIZATION
One of the main constraints to getting SRI methods taken up more widely has been that SRI requires, at least initially, more hand labor, as discussed in Chapter 7. When starting to use any new methods, expending more time and effort should be expected because learning and mastering new techniques invariably involves some investment. However, unless learning curves are flat, this should be a transitory problem, lasting only a few weeks or even a few days. Most evaluations have shown that SRI methods reduce rather than increase labor requirements once the methods have been learned.
Still, any increase in the amount of labor required to grow rice can be a disincentive for taking up SRI, even if this is a transient cost. And needing to invest more labor in rice-growing will be a deterrent wherever farmers’ current cultivation practices are already somewhat mechanized, or where farmers face limited availability of agricultural labor, or where hiring labor is costly because of its scarcity and high wages. The ability to mechanize SRI field operations with appropriate tools, implements and machines that reduce labor requirements is desirable under these conditions, and indeed under most conditions.
Growing rice is difficult and never very pleasant work. It involves hundreds of hours of hard labor per hectare, usually working in the hot sun, slogging through mud, exposed to diseases and injury. Spending much of a lifetime in rice-growing operations can become physically debilitating from the months and years of working in stressful, unnatural postures with day-long repetitive motions. In general, it is desirable to mechanize operations such as transplanting, weeding, and harvesting as much and as soon as possible.
When SRI ideas and practices were put together in the 1980s, equipment was not of much concern. The rural households whom SRI was intended to benefit usually had very small landholdings and little money with which to purchase implements. The new methodology was designed to use make more productive the resource that poor households had in relative abundance, labor, so that the productivity of their limiting resource, land, could be enhanced.
The production increase that SRI methods achieved was great enough so that the productivity of households’ other resources was also increased, producing more rice per unit of labor, of water, of seed, and of capital that was invested in growing rice. There was one simple implement that was strongly recommended to get the most benefit from SRI’s set of practices, however: a manual push-weeder made from metal, or possibly wood.
As seen already several times in this memoire, mechanical weeders offer farmers many advantages, reducing their labor time, lessening discomfort and disability, while also enhancing yield (Chapter 11). However, this simple mechanization was not required. Manual weeding rice paddies by hand has remained an option for farmers who cannot afford to buy a weeder or who have no access to one. It was unfortunate that any farmer would have to forgo the yield enhancement of active soil-aeration with SRI, but he or she could achieve yield improvement without any special equipment. Mechanization, thus, did not figure into initial SRI thinking and practice.
During subsequent decades, however, manual and motorized equipment have become more important as labor availability and labor costs have become more of a constraint for many if not all farmers, and as farmers with larger landholdings began to consider taking up the new methods.
While some equipment for rice production already existed in most countries, most of these implements were not suitable for SRI purposes. The main operations that created bottlenecks for the adoption of SRI were marking of the fields’ surfaces for proper SRI transplanting, either transplanting or seeding, especially weeding, and finally harvesting and threshing. These last two operations were made challenging by the larger yields achieved with SRI methods, but this was a welcome problem, and these operations are the easiest to mechanize.
The kinds of equipment most needed to reduce labor and accelerate the adoption of SRI were:
Weeders, the most immediate need to support SRI spread because weed growth increases when water is not kept standing on rice fields;
Transplanters, available mechanical transplanters were not designed to handle small, single seedlings and to place them in a square grid;
Markers, that mark a square grid on the soils’ surface to speed up manual transplanting;
Seeders, where direct-seeding is preferred to transplanting for establishing the crop, being able to place just one or at most two seeds in a hill with wide and precise spacing to form a grid pattern.
SRI farmers and potential SRI adopters who wanted to mechanize some or all of their operations have had either to reconfigure or repurpose existing equipment or to design their own prototypes where suitable equipment was not available. The optimization of equipment performance changes with altered planting geometry, seedling age and number, soil type, weed pressure, water control, cost considerations, and users’ gender. So, there were many questions for which there were no definite answers as mechanization of SRI operations began. There was little available information to assist farmers, program managers, or project staff in figuring out how to integrate mechanized operations into their practice of SRI.
This chapter reviews how bringing equipment and implements into SRI operations has evolved since the original version of SRI which required no machinery and relied on skilled, motivated labor to achieve its results. Inventing or adapting implements that are suitable for SRI operations is a challenge that has been taken up by farmers and others in a number of countries. The introduction of better implements and machines into rice production represents a big change in most rice cropping systems.
We consider first the mechanization of harvesting operations for rice production because this presents no special challenges and becomes very important when harvests are much increased. Then we look at improvements made for weeding and at several different ways in which the establishment of SRI rice crops has been mechanized. These operations need considerable modification for SRI cropping. Sourcing of equipment is discussed briefly at the end of this chapter.
Collecting SRI rice from the field can be done with any implement that mechanically harvests rice no matter how it is grown, possibly with a few adjustments. At harvest time, there is no difference between SRI and conventionally-grown paddy rice except that there is more grain to be brought in from an SRI field, which is always welcome.
As a rule, rice grains in SRI fields ripen more evenly than with conventionally-managed rice. This is an advantage as is little or no lodging of the stalks and panicles. That SRI plants have grown singly rather than in clumps makes for somewhat more efficient mechanical harvesting of their panicles because there is little need to employ human judgment about which panicles are ripe enough to be cut and which are not.
The removal of rice stalks and their panicles from an SRI field is thus a simple although usually larger task, quite suitable for mechanical operations. When SRI harvesting is done by hand, reapers frequently comment that it is not harder for them to bring in even twice as much grain per day because SRI plants grow more uniformly, and SRI panicles are mostly at the same height and similarly mature.
Below we see the stalks and panicles of rice that were hand-harvested from Sumant Kumar’s SRI field in the 2011 kharif season in Darveshpura village, Bihar state in India (Chapter 10). Such a crop would have been easier to reap and then thresh mechanically. When rice has been hand-harvested, it is desirable to thresh it mechanically if possible, removing grains from the stalks by machine. With larger panicles, there is every reason to use appropriate machines if available for threshing the SRI panicles to accumulate the grain in bags for sale or storage.
Below is a mechanically-transplanted SRI field in Nepal being harvested with a combine-harvester in Belbari municipality in Morang district, where SRI mechanization has been promoted by Rajendra Uprety (Chapter 42). With mechanical reapers, Nepali farmers can reduce their costs of harvesting by up to 20%; and with combine-harvesters that both reap and thresh, their costs of harvesting are lowered even more, by 30-50%. It has been found in Nepal that mechanization can reduce labor requirements for harvesting and threshing operations by more than 40 laborer-days per hectare. Of course, where farmers’ rice fields are very small or are difficult for machines to access, there will be no alternative to harvesting by hand.
Many kinds of implements have been developed to carry out this essential operation in ways that save farmers’ time and energy and achieve better soil aeration than is possible with weeding by hand, or by using herbicides. When rice plants are more widely spaced, this gives weeds more opportunity to grow. Various push-weeders that have been designed and made can speed up weeding and reduce the discomfort that results from weeding by hand. Manual weed control requires stooping over in paddy fields for many hours on end, a task usually assigned to women. Both men and women benefit from mechanization of this arduous work.
Manual mechanical weeders
Below are some of the simple, inexpensive weeders that have been developed for SRI cultivation. Their respective suitability depends on the type of soil being cultivated (how heavy and sticky, or how light and granular); on the upper body strength, height, and gender of the person using the weeder; on the height and width of the rice plants, on how closed is the canopy that they spread over the soil, etc.
Below is one of the earliest mechanical weeder designs, displayed by a Sri Lankan farmer who called it a ‘Japanese weeder’ because of its association with rice cultivation methods in Japan such as row-planting. Gamini Batuwitage who was traveling with me said that in China, it is called a ‘wolf fang’ weeder, the name referring to its teeth which churn up the soil and bury weeds. Its main limitation is that in heavy soil, the teeth can get quickly clogged with mud and weeds, so it takes considerable and repeated effort to keep the weeder free of detritus.
One of the most popular designs has been the cono-weeder, seen below, which became used more widely after IRRI engineers improved its design. Its cones disturb the soil without accumulating mud and weeds. However, it is heavier than the weeder above, and this weight has been considered a liability, especially by women farmers.
Below on the left is an improved weeder designed by Premaratna in Sri Lanka (Chapter 26). This is lighter and easier to handle than usual cono-weeders and has been fabricated locally in Sri Lanka for as little as US$10 in bulk orders. This weeder can be kept from getting clogged up with weeds and mud by keeping more weight on the front pad than on the turning wheels. On the right is a ‘hybrid’ weeder built by a farmer in Timor Leste combining features from several different designs.
An improvement in design that has spread fairly widely among SRI farmers in India was the Mandava weeder seen below, developed by Kishan Rao in Andhra Pradesh state. This simple, lightweight, and sturdy weeder, was named for Kishan’s village. Women rice farmers have found it particularly easy to manage.
A more complicated design was developed in Madagascar by an engineer who was interested in SRI. This two-row cono-weeder could be adjusted for weeding between rows of different widths. Although ingenious and adaptable, it was too heavy to be practical, however, especially in heavy soils.
On the right below is a more serviceable and lighter four-row cono-weeder developed by Gopal Swaminathan in Tamil Nadu state of India. He reported that this did not require more upper body strength than is possessed by most women farmers. However, it did not become as popular as the Mandava weeder. For strong farmers, such a four-row weeder could reduce weeding time and effort by 75% compared to a single-row design.
In my SRI travels, I have come across a number of farmer innovations for very simple hand weeders that were remarkably cheap and save considerable labor time. The most original and aesthetically-pleasing was a weeder designed and built by Nong Souvann in Cambodia, shown below. He drove large nails into a wooden axle, bending them a bit before he mounted the axle within a metal frame that he had welded together. A handle was attached pushing the implement. After his first season of using this weeder, he got the idea of attaching a fork below the handle to break up and aerate the soil as he went up and down the rows.
Souvann told me that with this weeder he needed only half a day to weed his small rice plot (0.05 hectares). The labor cost for this was just 2500 riels (60 cents). With SRI methods, his yield was 50% greater than he had been able to produce before, so he calculated that this weeder, which cost him less than US$ 3 to make, helped him earn US$ 20 more net income. Souvann’s investment in materials and labor was thus very cost-effective, and he said that weeding with this implement was much easier on his body than hand weeding.
An even cheaper and more cost-effective weeder was conceived and made by Govinda Dhakal in ward 6 of Indrapur village in Morang district. When Govinda first tried using SRI methods, mechanical weeders were not yet available in his district, and his first SRI rice crop was overrun with weeds, giving him such a poor yield that his neighbors thought that he would never try the new methods again.
But Govinda, being persistent, made himself the simple weeder shown below, modeled after a push broom. He hammered some nails into the bottom of a block of wood 20 cm long and attached a stick handle to the block. Then he ‘swept’ between the rows of rice, destroying weeds while at the same time surface-aerating the soil. The materials for this weeder cost him only a few rupees, he said, about 20 US cents.
When Govinda found that the handle of his first weeder was too long and the nails were too short, he made another weeder with a shorter handle and used longer nails. This tool was so cheap that he could make new and different ones as needed. Previously, he said, weeding one hectare of rice land had required 25-30 laborers for a day. With this weeder, now 10 laborers could complete the task in one day. This not only cut his costs of production, but made the work less arduous, he added.
Govinda reported also that with SRI management, his preferred variety of rice, which usually took 155 days to mature, was ready to harvest after 123 to 125 days. Shortening the time for growing rice by a month meant that he could plant a follow-on crop, tomatoes or chillies, much sooner, getting both a better yield and a higher price in the market because his products reached it earlier. Govinda and his fellow farmers agreed that most rice farmers in the area were now using SRI methods, with their yields doubled or more.
Other hand-weeder designs have been developed by farmers in Nepal. Initially, SRI farmers used a locally-fabricated double-wheel weeder, below left, which was rather heavy and required considerable strength to operate. Subsequently, a lighter single-wheel weeder became popular among SRI farmers. Below right is one of these weeders being used in an SRI basmati rice field in Sundarharincha municipality in Morang district.
The cono weeder, Mandava weeder, and other push-weeders all speed up the task of weeding and reduce physical discomfort compared to weeding by hand. It is certainly desirable to continue developing ergonomically more-efficient designs for manual push-weeders, but there is now a quest to develop similarly-effective motorized weeders for SRI use. Powered implements can weed several rows, even three or four, at the same time. Such implements, although more costly, greatly reduce the time and effort needed to weed a field.
When applying mechanical power in place of human energy, these weeders can break up and aerate the soil around rice plants more thoroughly, which is an advantage. SRI-Rice at Cornell has developed an international network of SRI colleagues who are cooperating and sharing information to develop better weeding equipment, whether powered by human or inanimate energy sources, with a particular interest in motorization.
The first effort that I encountered for motorizing weeding for SRI was by Subasinghe Ariyaratne, a farmer in System H of the Mahaweli irrigation scheme. He is seen below kneeling beside a weeder that he had built himself by mounting an imported Chinese motor on a metal frame that he constructed by welding. He then painted it a bright yellow. The total cost for fabricating this weeder was about US$750, he said, most of this spent for the motor.
While this implement was fairly expensive, Ariyaratne said that it was a good investment. He had 2 hectares of paddy land to cultivate, and his two children were still too young to help with field work. With this implement and his innovative method for crop establishment, discussed below, he was able to cultivate his medium-sized rice farm all by himself, not having to make any expenditure for labor.
Subsinghe told me that with this weeder and his planting method, he was reliably producing each season a yield of 7.5 tonnes per hectare, which was 3 tonnes more than his neighbors produced. The extra yield from just one of his two hectares can return him the cost of his weeder in a single season. So, this was indeed a good investment. Subsinghe’s weeder was designed for inter-cultivating single rows. I never learned whether he was able to design a two-row weeder as he intended, which would cut by half the labor needed for weeding.
Below is another motorized weeder developed subsequently in Sri Lanka, costing about 100 US$. Its cutting action for weeding between and across rows resembled a mowing machine more than the common design inspired by a plough or harrow. The mowing action was adapted so that it would disturb the topsoil much like other mechanical weeders.
One of the early criticisms of SRI, discussed in Chapter 7, was that farmers would not adopt it because its weeding requirements require too much labor. When SRI rice plants are spaced farther apart, this gives weeds more opportunity to grow so more weeding is necessary. But with SRI cultivation, mechanical weeding gives farmers two benefits from a single expenditure: weed control and higher yield (Chapter 11). Both manual push-weeders and motorized weeders raise crop yield by controlling weeds and by actively aerating the soil to stimulate the growth of plant roots and beneficial soil organisms.
Skeptics about SRI have underestimated the ingenuity of farmers for dealing with the problem of weed control so that they could capitalize on SRI opportunities. Motorized weeders will surely be needed for expanding SRI use in the future, as labor in the agricultural sector becomes scarcer and more costly. But for some time to come, there will also be need and demand for efficient, low-cost hand weeders of many designs. There will probably be more innovations like the ones discussed and shown above, and like the innovative bicycle-weeder shown below.
Manufacture of weeders
Getting weeders manufactured locally to make them available more readily and cheaply, and matching local conditions and preferences, has been a long-standing concern. Not having access to appropriate equipment that is effective, inexpensive and durable has been one of the main constraints on SRI use and expansion. One of the first experiences with local fabrication of weeders was in Sri Lanka, where Gamini Batuwitage and W.H. Premaratne promoted the making of the red push weeder shown above. Here is a case study from Kenya.
SRI methods were introduced here in the Mwea irrigation scheme in 2009. At the time, the only mechanization of rice cultivation was tillage of the fields with tractors. These were owned by the National Irrigation Board (NIB) which provided tillage services to farmers on a fee-for-service basis. Leveling of the paddy fields was done and continues to be done using animal-drawn wooden boards. All other rice farming activities were by manual labour, i.e., nursery seeding, transplanting, weeding, harvesting, and threshing.
In 2009, there were perhaps three rotary weeders in the whole Mwea scheme. These had been brought in by researchers for various studies. It was not clear which design of weeders would be most suitable for the Mwea area and for its soils. So, in 2010 a study was conducted in Mwea led by Prof. Patrick Home of JKUAT which assessed seven weeders from various countries, most from Japan or India, but also a wooden model procured from Tanzania. This was hands-on practical research in which groups of farmers tried their hand with each of the seven types of weeders and shared their assessments among themselves and with researchers.
On the left below, we see Moses Kareithi, the first farmer in Mwea to plant and validate SRI (whom the NIB appointed as an SRI extension agent), and Bancy Mati, a professor at Jomo Kenyatta University of Agriculture and Technology who spearheaded SRI work in Kenya, both looking at weeders. On the right are the farmers who did the testing and evaluation of the weeders. They settled on three weeders that they considered most effective and easiest to operate, among which their favorite was a Japanese model.
The next year, the three preferred models were taken to three foundries based in the Wanguru market in Mwea and to a youth training centre (Mucii wa Urata Youth Polytechnic). This step of the process was managed by Moses. The youth training centre made an error and produced weeders with tines that were parallel, so that the drum in the back cut the soil in the same line as the drum in front, which made for inefficient weeding. Efforts were therefore concentrated on the commercial foundries in town, finally settling on the one that had the requisite metal forging lathes and machines. The foundries initially made just a few weeders a day, mostly for projects that distributed them to farmers free.
In 2015-16 when implementing an SRI extension project that served three irrigation schemes in Western Kenya , the making of weeders was introduced to a foundry in Ahero town. With support from the Indian social enterprise AgSRI (Chapter 34), this foundry initially made 30 weeders which were distributed to farmers. Below are the weeders after production and after being distributed in the field.
It is now over 10 years since SRI was introduced in Kenya. As SRI adoption has spread, the demand for weeders has grown, and individual farmers have started ordering weeders directly from local fabricators. Rotary weeding has become the preferred mode for weeding in Mwea, Ahero, and West Kano irrigation schemes.
The weeders sell for between Ksh.2,000 and Ksh.3,000 (US$ 20-30), and farmers are able and willing to buy these locally-made weeders without further project support. Local manufacture of weeders is a win-win development, creating employment opportunities in the area while saving labor and reducing the drudgery of paddy rice production.
This operation which starts the crop cycle is critical for SRI success. It is usually done through transplanting when rice is being grown with irrigation. But this operation, being laborious, can be a bottleneck for the expansion of SRI in its original form. The best SRI results depend on the seedlings being transplanted while still very young, before the start of their fourth phyllochron of growth (Chapter 6). These small, delicate seedlings, 10-15 days old, should be transplanted individually and very carefully to minimize trauma to their roots. Also, they should be planted quite shallow, just 1-2 cm, not buried deep in the soil as this inhibits their growth because of hypoxia.
It has been seen that SRI rice crops can be established not only by transplanting seedlings, but also by direct-seeding with wide and square spacing of plants, provided that a satisfactory rate of germination can be assured. Also, transplanting with unirrigated SRI is seldom feasible or not economical. Direct-seeding, on the other hand, is usually suitable for rainfed SRI, but it is also an option for irrigated SRI. Direct-seeding makes it unnecessary for farmers to establish and manage a nursery and then transplant seedlings. This makes for significant savings of time, about 40%.
There have been efforts to adapt existing transplanting machines developed for conventional rice cultivation so that they can establish a rice crop according to SRI precepts. This has not been easy, but it is possible. At the same, equipment for direct-seeding has also been being developed or adapted to give SRI farmers an alternative to transplanting.
We report here on two different examples of mechanized transplanting and on two methods for mechanized direct-seeding. It may seem like a large modification of SRI if the rice crop is established by direct-seeding rather than by manual transplanting, a ‘signature’ practice for SRI at the outset. But farmers weigh a tradeoff between higher yield and less need for labor.
So far, the highest SRI yields have usually come with careful hand transplanting, a kind of artisanal rice production. However, often it can be more profitable to adopt mechanized techniques even if they do not give the highest yield because mechanization reduces farmers’ costs of labor as well as time. Higher net return is farmers’ goal rather than highest possible yield.
The first effort that we know of to mechanize transplanting according to SRI principles was made in Latin America rather than in Asia, where most of the world’s rice is grown. Interest in mechanizing SRI production in Latin America has been spurred by the limited availability of agricultural labor in most rural areas of the region and by the relatively high cost of agricultural labor. This has been a major constraint on SRI uptake in Latin America, and previously it has kept most farmers there from even considering the use of SRI. The story of mechanized SRI transplanting began in Central America.
A progressive Costa Rican farmer in Guanacaste province who had not grown rice before, Oscar Montero, first learned about SRI in 2004 from a Colombian visitor who was interested in rice-transplanting machinery. To try out SRI methods for himself, Oscar acquired a small modern Japanese rice-transplanting machine, shown below, which he modified so that it could transplant just one seedling per hill (or at most two) with spacing of 22-24 cm by 30 cm. This mechanical transplanter thus established 12-15 plants per square meter, a density slightly below the density usually recommended for SRI, 16 m⁻².
Oscar grew his rice seedlings in nursery trays 28 × 58 × 3 cm, shown above on the right, filled with burned rice hulls as a planting material. To this he added some mycorrhizal fungi in a commercial formulation known as EcoMic, developed in Cuba. By 10 days after sowing, the seedlings reached a height of 15 cm. The material in the trays had become coherent enough, its particles held together by fungal hyphae and the seedlings’ roots, so it could be rolled over like a rug and transported without harming the young seedlings inside the roll, as seen above. The motor-driven transplanter on left was guided up and down the field by a single operator, placing single seedlings (or sometimes two) in the soil at regular, rectangular intervals.
To inhibit the growth of weeds, Oscar initially kept more water on the field than is recommended with usual SRI, but doing intermittent draining to achieve some soil aeration. He used organic methods to control pests and diseases, applying entomopathogens, and he also inter-planted rows of garlics and chili peppers to repel noxious insects.
In his first year, Oscar achieved a yield of 8 tonnes per hectare, almost double the national average production, 4.2 tonnes per hectare, with lower costs of production. This was the first demonstration at field scale that mechanical transplanting could be used successfully with the other SRI methods. Oscar’s SRI field being mechanically harvested is seen below.
In the past dozen years, there have been other attempts to adjust and adapt mechanical rice transplanters for SRI use in Asian countries. Most mechanical rice transplanters already require the use of fairly young seedlings because older seedlings are larger and bulkier. However, it is not easy to separate young plants mechanically so that the single seedlings can be transplanted without damage to their roots. Thus, mechanical transplanters often transplant two or even seedlings in a hill. But with adjustments, planting four to six plants in a clump is no longer the norm.
Turning multi-row transplanters around at the end of rows in small rice paddies takes considerable space and is often awkward. So, mechanical transplanter designs still need some improvement. However, serious engineering efforts are underway. As mentioned above, the SRI-Rice Center supports an on-line network of persons, mostly in Asia, who share their knowledge and experience regarding the design and use of SRI equipment.
Transplanting by machine can be made easier, with less trauma to the roots, by starting seeds in plastic trays, resembling egg cartons, that have rows of separated pockets of soil into which single rice seeds can be placed to germinate. The single seedlings with clumps of soil around their roots can then be handled mechanically with little root disturbance. Machines can stick these plugs of soil, each with a single seedling growing upward and root going downward, into muddy soil with very little trauma to the roots. This practice for growing seedlings can also make transplanting by either hand or machine both easier and quicker.
This technique was shown to me by farmers in Karnataka state of India during a visit to their village in 2006. The farmers actually borrowed this idea from their growing of vegetable seedlings. The plastic trays used are both cheap and reusable. This method is feasible for rice-growing because with SRI many few plants are needed to establish the field compared to a conventionally-managed field, requiring 8-10x more seedlings. There are a number of variants of this methodology such as I was introduced to on a visit to Cuba.
One adaptation of growing young seedlings in containers (or balls) has led to what is called the ‘parachute’ method in several countries, e.g., Iraq and Thailand. With this method, clumps of dirt each with a single rice seedling growing in it are tossed onto the muddy soil of a flooded rice paddy to establish themselves wherever they land. However, with this practice of crop establishment, it is not possible to achieve any regular and linear spacing between plants, so soil-aerating weeding with a mechanical implement is not possible.
In Sri Lanka, according to Gamini Batuwitage, when the Department of Agriculture began to promote growing of single rice seedlings for ‘parachute’ transplanting, SRI farmers who wanted to get the benefits of wide spacing and square planting used this seedling-growing method for their mechanical transplanting which established regularly-spaced single seedlings in a square pattern.
The ‘parachute’ method of crop establishment pushes the limits of what is considered as ‘SRI’ since active soil-aeration around the rice plants is a key element, and square planting is necessary for mechanical weeding operations. The idea of establishing single young seedlings in a block of soil so that there is no root disturbance when transplanting is consistent with SRI practice. In Sierra Leone, Gerald Aruna has developed a method for growing single seedlings in balls of dirt, then planting them by hand with SRI spacing under upland, rainfed conditions, which can extend the scope for SRI in his country.
When my wife Marguerite and I visited Sichuan province of China in 2013, our host in the Sichuan Academy of Agricultural Sciences, Zhang Jiaguo, showed us a simple ‘automated’ method that he had developed for growing SRI seedlings with less labor that could be used in mechanical transplanting. The plastic tray on the bottom had 400 small cups (20 × 20) that were each filled with potting soil. He had designed another plastic tray that could be set on top of the lower one, with indentations that were aligned with the cups below.
When all of the indentations in the upper tray had a single rice seed in them, by pressing a lever that slightly shifted the top tray, Zhang could drop individual seeds simultaneously into each of the cups in the tray below. These trays are shown below. After 10-12 days, the seedlings grown in their respective cups would be loaded on a mechanical transplanter and put into the field with wide, controlled spacing. Many more such innovations will surely be devised to reduce the labor required for establishing SRI crops.
In this country, a highly-mechanized version of SRI has been developed in Punjab province, the epicenter of the Green Revolution, by a successful farmer-businessman with strong environmental and philanthropic interests. Asif Sharif, who was an importer and seller of farm machinery for almost 30 years (Chapter 25), has his own machine-shop facilities in which he can manufacture a variety of farm equipment, or at least prototypes thereof. That he is an innovative farmer is indicated by his being the first farmer in Punjab to do laser-leveling of his fields, to make achieve water savings and to practice precision agriculture.
Having promoted and engaged in Green Revolution agriculture for several decades, Asif became dismayed by the damage that heavy tilling and continuous flooding of fields was doing to the health and fertility of Punjab soils. In part also because of the overuse of chemical fertilizers and agrochemicals, these soils were no longer among the world’s most productive.
Asif learned about SRI from the SRI-Rice website in 2008. The principles and reports of SRI made sense to him, and about the same time he learned also about Conservation Agriculture from Amir Kassam, who has figured in the SRI story in a number of important roles. Conservation Agriculture is discussed in the next chapter.
Asif set out to create a synthesis of SRI and Conservation Agriculture methods, guided also by an understanding of the theory and practice of organic agriculture. When these ideas were combined, he designated this as ‘paradoxical agriculture,’ paradoxical because it enabled farmers to produce more output with fewer inputs.
Crop establishment with Asif’s methodology is not 100% mechanized, as explained and seen below. But compared to the usual methods of rice production in Punjab, his methodology can reduce the labor inputs required per hectare by 70%, and also reduces the amount of water required to irrigate the rice crop by 70%. His resulting paddy rice yield of 12 tonnes per hectare was more than triple the usual paddy yield in Punjab province.
The test plot that Asif used for his first trial of mechanized SRI, having fabricated or adapted the requisite machinery in his own workshop, was 17.5 hectares, many times larger than the size of typical test plots. The field seen below had been laser-leveled so that water could be applied very sparingly through furrow irrigation, rather than by flood irrigation. With irrigation water supplied to the field by siphons from the main canal, there was no need for any pumping, so this also reduced energy costs substantially. On the field as shown below, raised beds 105 cm wide and as much as 600 m long were made by a specially-designed machine at the same time that it created 20-cm furrows between the beds. On each bed, five rows of rice seedlings were transplanted, spaced 22.5 cm apart.
When the whole field had been configured with these beds, a composite machine shown below drove along the beds’ furrows with laborers riding on it, dropping 10-day-old rice seedlings into holes that the machine had punched into the soil. The rows and holes were all 22.5 cm apart. The seedlings for this operation were grown on organic mats similar to those prepared by Oscar Montero in Costa Rica.
The machine injected a small amount of fertilizer and compost into each hole as it passed over at a rate per hectare of 500 kg of NPK fertilizer and 875 kg of compost. As the machine moved along and after the seedlings and fertilizer were both in place, each hole was filled with water from a tank mounted on top of the machine. This water slowly seeped into the dry soil around the seedling to keep them from desiccation.
When the whole field had been transplanted in this way, it was flooded with water up to a level 2.5 cm above the top of the beds. This gave the whole field a one-time soaking to help the plant roots start growing. Thereafter, only intermittent flows of irrigation water were siphoned from the canal into the furrows which the water run down. Because the soil at the bottom of the furrows had been firmly compacted by the machines, the water flowed mostly laterally into the soil of the raised beds, providing water to meet the plants’ needs and those of the soil biota.
Several times during the season, the rows were mechanically weeded by a tractor-weeder shown below. Because the rice plants were being grown on raised beds, no perpendicular cross-weeding was possible without destroying the beds, but this was not necessary. The tractor-weeder was able to do a very thorough job of inter-cultivation between the rows. Given the precision planting, several one-way passes with this machine were enough to both control weeds and achieve good soil aeration.
Because the beds and rows were laid out so precisely, the tractor could be operated by radio control without a driver, being guided mechanically to follow along the furrows between the raised beds. Packing down the soil at the bottom of the furrows with what is called controlled tillage helped to facilitate lateral diffusion of water when it ran along the furrows.
This was mostly a proof-of-concept demonstration because it was very capital-intensive. Given the low market price for rice, this version of SRI has not spread much in Punjab province. But it showed that a doubling of yields or more could be achieved by mechanized SRI with large reductions in both labor and water, both by 70%, integrating SRI with Conservation Agriculture and organic agriculture. This was a creative extension of the original ideas of Fr. Laulanié. As seen in the next chapter, the concepts and adaptations of Asif’s ‘paradoxical agriculture’ have been productively and profitably extended to wheat and a number of other crops beyond rice.
A less capital-intensive system of mechanical transplanting to establish SRI crops has been developed here, starting in 2009 at the instigation of Rajendra Uprety. Initially, farmers used four-row walk-behind transplanting machines with 10 to 12-day seedlings. This made work easier, less time-consuming, and less costly. The precision transplanting also facilitated mechanical weeding of the crop.
After 2015, farmers began using eight-row riding-transplanters as shown below. In several places, women’s groups are operating all the machines, representing quite a socio-cultural change. Farmers have found it advantageous to establish community nurseries for growing the young seedlings on plastic trays, as also shown below.
Below is a machine-transplanted SRI field in Morang district in 2018. This fine-grained, 135-day variety gave a yield of 8.4 tonnes per hectare. The net profit from the crop was 72,780 NR (US$ 650) per hectare, compared to 18,050 NR (US$ 160) using standard methods of rice cultivation. While this result was better than the average, it showed what this version of SRI can attain. Mechanization by itself was calculated as raising farmers’ profit per hectare by 36%, while mechanization together with SRI practices raised their returns by 58%. With mechanized transplanting and weeding, farmers’ labor requirements per hectare were reduced by 60%. Such results will accelerate the uptake of mechanical transplanting in the region. Farmers are also experimenting with mechanized direct-seeding, using a drum-seeding implement like the ones discussed next.
After three years of introducing SRI in Chitoor district of Andhra Pradesh state, Bala Hussain Reddy and colleagues at the KVK agricultural extension center in Tirupati concluded that local labor constraints, especially for transplanting, were slowing the uptake of the productive new methods. So, they began experimenting with specially-designed equipment to establish SRI rice crops by doing direct-seeding rather than by transplanting, still following SRI principles as much as possible.
Sprouted seeds, 15 kg per hectare, were put into four perforated plastic drums that were mounted on a long axle as seen below. As the implement was pulled along, the drums dropped the seeds into a leveled puddled field in eight rows at a time, with the seeds falling 5-8 cm apart in the row. While this seed rate and resulting plant density were more than double that of transplanted SRI, they were less than half of what occurred with farmers’ usual practices.
With this method, some weedicide was applied initially, and mechanical weeding was done in one direction, not two, following the rows of plants that had been established by the seeder. This produced an approximation of usual SRI plant establishment, with yield less than could have been attained with fuller use of the new methods. But this methodology was profitable for farmers, and as seen below, the drum seeder was transportable by motorbike from one field to the next. This was a first step for mechanized direct-seeding utilizing SRI ideas.
The labor-saving with this implement is substantial. Reddy reported that one farmer could establish a hectare of direct-seeded SRI rice in 5-6 hours. Conventional transplanting required as many as 30-40 laborers for 3-4 days to transplant one hectare by hand. Total labor cost of crop establishment with this direct-seeding method was 4,025 rupees (US$ 53) per hectare, compared with 11,875 rupees with usual farmer practices (US$ 100 more). Farmers’ net returns were 35,020 rupees per hectare (US$ 460) vs. 19,890 rupees (US$ 260) with farmer methods, an increase of 75%. The critical requirement for this methodology to succeed was getting a high and predictable germination rate. This is achievable by good seed selection and by having an optimal soil-water ratio when planting.
Given labor shortages in this country’s industrializing economy, direct seeding is an attractive alternative to transplanting. This method for crop establishment was promoted under a program funded by Australian Aid (AusAid) and managed by the Dutch NGO SNV. The program with SRI introduction and dissemination was intended not just to raise the productivity of Vietnamese rice farming, but also to reduce the greenhouse gas emissions that irrigated rice production currently contributes to global warming.
The direct-seeding implement used is very similar to the one that was developed in India, discussed above. Possibly it was learned from the SRI-Rice website where Reddy’s reports from India were posted for the purpose of disseminating SRI mechanization practices. The equipment in Vietnam uses six drums on one axle, so it can establish 12 rows of rice on each pass across the field.
A study in Kenya published in 2019 studied the reduction in labor inputs and income impacts when drum-seeding as in India and Vietnam was combined with mechanical weeding. Replicated trials showed that mechanized direct-seeding reduced the labor requirements for crop establishment by 97% per hectare compared with establishing a nursery and then having to transplant seedlings.
Mechanical push-weeders reduced the labor required for weeding by 28% compared to weeding by hand. Using all of the SRI practices including transplanting gave 46% more yield than was produced by farmer methods. Drum-seeding used with the other SRI practices yielded 27% less paddy rice than did full SRI management. However, drum-seeding used with SRI methods gave 10% more yield than from farmer practice.
Considering the respective costs of production and yield results associated with the different methods, calculations of net income per hectare for the alternative methods of crop establishment and weeding showed full use of SRI giving 94% more income than farmer practice. Drum-seeding used with the other SRI practices, given reduced labor costs, yielded 39% more income than farmer practice. Gaining 40% more income with much less requirement of labor could make direct-seeding an attractive option for SRI in Kenya.
When Subasinghe Ariyaratne in System H of the Mahaweli scheme developed the motorized yellow weeder shown above, he also modified his method of SRI crop establishment. Because his family was small and his children were still young, his farming was labor-constrained. So, he adapted a practice that was being used by some Sri Lankan (and other) farmers: broadcasting seed, and then radically thinning the plants after they emerged. His innovation was to do the thinning of plants in a grid-like manner consistent with SRI precepts.
When he established his SRI crop by hand transplanting, Ariyaratne needed only about 5 kg of seed per hectare. With his alternative strategy, after he had leveled and puddled his rice paddies, Ariyaratne took 25 kg of pre-germinated seed, five times more seed than used for transplanting, and broadcasted it by hand across his puddled field. This meant that he did not need to make and manage a nursery and he had no need to transplant. His broadcasting operation took only a few hours, not days.
When the emergent rice plants were 10 days old, he took his home-made motorized weeder and ‘weeded’ the field as if he had transplanted it with standard SRI spacing of 25 × 25 cm. His weeder churned about 80% of the young plants into the soil, usually leaving just one plant (although possibly 2 or 3 plants) growing at the intersections where his perpendicular passes of the weeder across the field left small square areas in a geometrical grid-like pattern. If there was more than one seedling growing in a particular space, he did not worry about this. Their roots still had more room to spread than if he had transplanted seedlings in clumps of three or more, as was done traditionally.
Ariyaratne’s operations with the motorized weeder were repeated several times during the season, accomplishing weeding and soil aeration at the same time after the initial mechanized thinning. While Ariyaratne’s field lacked the spatial precision of a neatly-transplanted SRI plot, his labor-saving with this method was great, and his yield results were quite satisfactory, as reported above. When researchers at Tamil Nadu Agricultural University in India did an evaluation of this methodology, they reported a 37% reduction in the labor required without any significant sacrifice of yield.
SRI EQUIPMENT RESOURCES AND DISSEMINATION
There are some agricultural equipment companies that sell affordable manual and motorized paddy weeders suitable for SRI use, most of them located in India and China. Somewhat more expensive but well-designed motorized equipment is also available in Japan and Korea, a number of these accessible through the Alibaba website.
The Indian NGO WASSAN undertook a farmer-participatory evaluation of SRI weedings in 2006 that is still the most systematic consideration of push-weeder options. More recently, the California-based NGO, Earth Links has begun publishing a series of open-source computer-assisted design (CAD) drawings of popular manual weeders such as the cono-weeder, the Mandava weeder, and the rice dragon weeder, which Oxfam America has been popularizing for SRI in Cambodia as especially suitable for women farmers’ capacities. Earth Links’ drawings enable blacksmiths or fabricators anywhere in the world to have technical information on how to make the respective weeder parts and to assemble them.
SRI-Rice has collected numerous videos of SRI-adapted equipment, and Lucy Fisher supports an on-line SRI equipment innovation group on Facebook. In 2014, she convened an international workshop on SRI equipment at the Asian Institute of Technology in Bangkok, following the 4th International Rice Congress there. This workshop, co-sponsored with AIT’s Asian Center of Innovation on Sustainable Agricultural Intensification (SCISAI), brought together colleagues from 12 countries with SRI equipment expertise to work on designs, adaptations, and access issues.
However, there remain many impedances to making affordable, locally-appropriate equipment accessible to farmers throughout much of Asia, Africa and Latin America. Several national Asian SRI networks, most notably SRI-Mas in Malaysia and SRI-Pilipinas in the Philippines, have helped farmers to assess and access equipment, but their information is not yet easily accessible outside the country.
The development and adaptation of equipment for SRI purposes has progressed unevenly, but it has progressed through a variety of dispersed in initiatives.
In India, WASSAN has held workshops periodically to evaluate and redesign weeders. It was responsible for producing and making available the popular Mandava weeder, shown above, which is lighter and easier to use than the classic cono-weeder.
In Cambodia, an Oxfam-sponsored workshop collected a variety of weeders that were then vetted by women rice farmers, with the final product, called the Rice Dragon weeder, marketed as a women-friendly weeder in many parts of the country.
In Kenya, as described above, colleagues from the Jomo Kenyatta University of Agriculture and Technology and the National Irrigation Board, went through a systematic process of farmer-evaluation of different weeders, to arrive at a farmer-preferred model, which was then manufactured locally.
Staff at the Tamil Nadu Agricultural University have spent over a decade designing and testing weeders and transplanters, working with programs and manufacturers to make them available to rice farmer in Tamil Nadu state of India.
In Latin America, the Inter-American Institute for Cooperation on Agriculture (IICA) and Fontagro  have been working on acquiring and testing both motorized weeders and motorized transplanters for accelerating the spread of SRI, being interested also in motorized seeders. Latin American rice farmers have little interest in manual equipment because of their constraints of labor availability and cost, so motorized implements are the focus of innovation in this region.
The variety of crop-establishment methods reported in this chapter is indicative of the ways in which SRI has been a dynamic enterprise, with innovative farmers and others in different countries devising equipment that can make SRI cropping more efficient and labor-saving under their particular local and personal conditions. The various practices and implements that have been developed have given farmers greater access to the productivity of SRI insights that were presented in Chapters 4, 5 and 6. While SRI began without any mechanical or motorized assistance 30 years ago, it has not remained a necessarily manual methodology.
That SRI principles could be utilized without much or any capital expenditure made it accessible to the poorest and neediest households, a highly desirable characteristic. But as an innovation based on mobilizing biological potentials in the plants and soil, SRI is not a scale-limited innovation.
Over the past 20 years, there has been an extension and extrapolation of SRI ideas not just to other crops, as seen in Chapter 12, but also to larger units of operation, using machinery and non-human sources of power that can make SRI’s ideas and methods applicable to medium and large-size farms are well as to smaller ones. This is an important part of the evolution of SRI that has been on-going in India, Nepal, Malaysia, Philippines, Pakistan, Kenya, and other countries, especially now in Latin America.
NOTES AND REFERENCES
 See the discomforting video produced by Flooded Cellar Productions of how rice-field labor has disfigured and impaired the body of Phumani Kuji, a woman rice farmer in India.
 One qualification to this generalization is that there can be some adverse impacts on persons and families who rely on performing such agricultural labor for their sustenance and who may have few other opportunities to earn income. See evaluation of SRI in Andhra Pradesh state of India by Alfie Gathorne-Hardy and co-authors, ‘System of Rice Intensification provides environmental and economic gains, but at the expense of social sustainability - A multidisciplinary analysis in India,’ Agricultural Systems, 143:159-168 (2016).
The data gathered by Gathorne-Hardy et al. showed that SRI reduced labor requirements enough to that wage employment for agricultural laborers was cut by 50%. This was the only negative effect that they found from farmers’ adoption of SRI. With mechanization, there would be further labor displacement. The study showed SRI methods generating so many economic and environmental benefits that it would be unfortunate to forgo these just to avoid labor displacement.
The study’s economic evaluation of SRI methods showed that farmer’s net profit per hectare was increased so much that compensating landless households and particularly poor women for losses in employment would leave SRI farmers much better off economically than before; and the society would gain from environmental benefits.
 Association Tefy Saina designed the super-simple and very cheap weeder shown below. Made of wood and small metal blades for the poorest farmers in Madagascar, this weeder costs about US$ 2. Its main drawback was that it took two persons to operate it, one person pulling it from in front with a rope, and the other guiding it to accomplish good weed removal and surface-soil aeration. With no moving parts, this weeder might not be considered as mechanization, but it was designed for effective, inexpensive weed control.
 Land preparation is not considered here because preparation of rice paddies for SRI cultivation is not much different from conventional practice, apart from using less water for flooding during ploughing, puddling, and leveling. Leveling should be done more carefully, to be able to use irrigation water more efficiently, but this is a one-time operation.
Where land preparation can be done with a tractor, large or small, this applies for SRI or usual crop management. As discussed in the next chapter, there is a convergence in process between SRI and Conservation Agriculture (CA) which will reduce or eliminate tillage. Where raised beds are used, their conformation of the soil is also a one-time operation, possibly or preferably mechanized.
 If mechanically transplanting tiny, individual and delicate rice seedlings seems hard to visualize, a powerpoint on this process has been prepared by B.J. Pandian, Tamil Nadu Agricultural University, available on the web: The Development of SRI Transplanter: TN-IAMWARM Experience (2014).
 Markers are not discussed in this chapter because these implements do not fit most people’s understanding of mechanization. However, they are considered in this long endnotel because the transplanting of seedlings is greatly speeded up by using some simple implement to mark a square grid on the surface of rice paddy soil prior to transplanting. This makes easier the transplanting of single seedlings with the desired spacing, usually 25 × 25 cm, but spacing can be adjusted to be wider or closer.
The original method used in Fr. Laulaniè’s time was certainly not mechanized. Farmers were instructed to stretch across the paddy field strings (or ropes) tied to sticks on both sides of the field that were 25 cm apart. This gave farmers parallel lines along which to plant seedlings. Knots were tied in the strings (or ropes) at 25 cm intervals, which showed where the seedlings should be put carefully into the soil to have precise spacing that would permit a mechanical weeder to be pushed across the planted field in perpendicular passes several times during the plants’ vegetative growth phase. This would keep weeds under control until the rice plants’ leaves formed a canopy over the soil and could thwart further growth of weeds by shading them.
In Tripura state of India, my wife Marguerite and I saw in 2007 about the simplest method for marking lines on the surface of a muddied field before transplanting. Farmers crouching on opposite sides of the rice field stretched between them across the field an elastic rope held at surface level. They pulled it taut so that a third farmer could lift up the rope and let it snap back onto the muddy soil, imprinting on this a straight line.
By moving the rope across the field and setting it down repeatedly at intervals of 25 cm and then doing the same from the other two opposite sides of the field, a grid could be demarcated on the soil surface very quickly to guide precise transplanting. A slightly more complicated way to speed up transplanting with accurate spacing was to use a lattice like the one shown below, made from bamboo by farmers in Madhya Pradesh state of India. Both practices would hardly be considered by most persons as mechanization.
A more advanced and quicker way to demarcate grids on fields was the development of simple markers that emerged in a number of countries rather spontaneously. The first markers were simple rakes made from wood with tines spaced 25 cm apart. When pulled across a muddy field in perpendicular directions, these rakes marked it with a square grid. Examples of such rakes made by farmers in the state of Assam in India and in Indonesia are shown below. Most people would still not consider these tools as mechanization.
More mechanical are the roller-markers made from metal tubing or other material that farmers started devising in many countries. These could be pulled across the field to imprint a square grid rather easily, quickly and exactly. Examples from Cambodia and Punjab state in India are shown below. This equipment saved farmers considerable labor time, although probably too simple to be considered as mechanization by most people.
To see a roller-marker in operation, watch this short video(0:31), showing the use of a well-designed roller-marker developed in Thailand by Supachai Pitiwut, an SRI colleague who established the Weekend Farmers Network in his country.
 Information provided by Rajendra Uprety from farmers’ field operations.
 See trip report from 2005, pages 9-10.
 See trip report from 2006, pages 6-9.
 An informal SRI Equipment Network operates on Facebook under the group name SRI Equipment Innovators Exchange. This service was created and is curated by Lucy Fisher. As of April 2020, there were over 300 subscribers who exchange information on design, adaptation, and access to rice production equipment to be used with SRI methods.
 What was learned from Ariyaratne during a visit to Sri Lanka in 2003 is written up in a trip report on the SRI-Rice website.
 This bicycle-weeder was developed by a farmer in Maharashtra state of India, Gopal Bhise. He told a journalist that “Economic frustration at not being able to even buy a pair of bullocks led me to develop this device.” Bhaise could see that a bicycle is cheaper than a pair of oxen, and has no maintenance or feeding cost. So, he set about creating a weeder from bicycle parts that can also serve as a plough or harrow with different attachments. The implement sells for US$ 25 and can weed 0.8 hectare (2 acres) in an hour. “‘People laughed at me in the beginning, but I never gave up. Perseverance finally paid off, and today my Krishiraja [meaning farmer-king, the name that he gave the weeder] is received well in the local market,” says the poor farmer proudly.” M.J. Prabu, ‘Bicycle-inspired plough and weeder gains popularity,’ The Hindu, April 29 (2010).
 This has been the main reason for the slow uptake of SRI in the Latin American and Caribbean region, and it is a main focus for the Inter-American Institute for Cooperation on Agriculture (IICA) in its efforts to bring SRI opportunities to the region (Chapter 7).
 How Oscar learned about SRI (SICA in Spanish) and how he adapted its ideas for mechanized production has been reported by him on the SRI-Rice website in Spanish and in English, so only a short summary is given here. Although Oscar was not a rice farmer himself, there was some old rice-transplanting machinery on his farm which a visitor, Diego Uribe, had heard about and was interested in seeing and rehabilitating. Diego had learned about SRI from our colleague in Cuba, Rena Pérez.
 Ecomic is discussed by R. Rivera and F. Fernandez, ‘Inoculation and management of mycorrhizal fungi within tropical agroecosystems,’ in N. Uphoff et al., eds., Biological Approaches to Sustainable Soil Systems, 486-487, CRC Press, Boca Raton, FL (2006).
 For a good review of technical issues, with pictures and drawings, see conference paper by M.S. Imran, M.S. Abdul Manan, A.N.M. Khalil, M.K.M. Naim and R.N. Ahmad, ‘Design of transplanting mechanism for system of rice intensification (SRI) transplanter in Kedah, Malaysia’ (2017).
 As noted in endnote 10 above, this network operates through Facebook with support from Lucy Fisher and Steve Leinau.
 See page 21 of trip report on a field visit to Dypasandra village in Karnataka state in 2006.
 A variant on this idea I refer to as ‘the chocolate cake method,’ described on pages 4-5 of my trip report from a 2008 visit to Cuba. A young agronomist still working on her PhD, Yoannis Martin Enrique had learned about SRI from our colleague in Japan, Shuichi Sato, while she was doing a six-month training course on rice technology in that country.
To facilitate the transplanting of young SRI seedlings without disturbing their roots, Yoannis blended a mixture of soil and compost and pressed it into a flat mold, like a square cake pan about 3 cm high, to form a block of soil and compost that resembled a flat, moist cake of deep brown color.
The top of this ‘cake’ was scored with parallel and perpendicular lines, about 10 cm apart, to make a grid. Into each demarcated square, Yoannis placed one rice seed, and she then covered this ‘cake’ with a dark cloth for the seeds to sprout and grow in the medium that she had prepared.
When the seedlings were 8 to 12 days old, the ‘cake’ was cut into square pieces along the grid lines, and these ‘pieces of cake’ were put into the soil at 30 × 30 cm spacing, making the transplanting operation quick and easy, Yoannis reported. Her research was published in Cultivos Tropicales: Yoannis Martin, F. Soto, Y.E. Rodriguez and R. Morejón, ‘El Sistemo Intensivo de Cultivo del Arroz (SICA) disminuye de cantidad de semillas para la siembras, aumenta los rendimentos agricolas y ahorra el agua de riego,’ 31: 70-73 (2010).
 This methodology is shown in a Flooded Cellar video on Gerald Aruna’s work in Mendesora village, starting about minute 10.
 This is explained in Asif Sharif’s article, ‘Technical adaptations for mechanized SRI production to achieve water saving and increased profitability in Punjab, Pakistan,’ Paddy and Water Environment, 9: 111-119.
 Asif found it hard to believe that 25 cm spacing between rice plants would be optimal, so he designed the machinery for between-plant distances of 22.5 cm. When the crop was mature, he agreed that 25 cm would have been better spacing because the plants in the middle three rows were not as productive as the plants in the outer two rows.
 Because the soil had previously been managed with chemical inputs, Asif assumed that the abundance and diversity of soil organisms would probably not be sufficient to cultivate the field purely organically, so he phased in organic production, relying still on some inorganic fertilization when starting the conversion to organic agriculture.
 Information from paper by Uprety, ‘Meshing mechanization with SRI methods for rice cultivation in Nepal,’ presented at 5th International Rice Congress in Singapore 2018).
 See A Report on Direct-Seeding Technology Using SRI Concepts in Rice Using Drum Seeder in Chitoor District of Andhra Pradesh, India by Hussain Reddy et al. (2012) and their discussion of how to deal with problems: Field Problems in Direct-Seeded Rice Using Drum-Seeder – and Solutions (2014).
 A team of researchers at the Indian Agricultural Research Institute in New Delhi (one of whom, Anchal Dass, has done extensive research and publishing on SRI), have developed an effective implement for direct-seeding of SWI wheat, useful for smallholding farmers. H.L. Kushwaha et al., ‘ Design, development and performance evaluation of manual planter for system of wheat intensification,’ Indian Journal of Agricultural Sciences 89: 678–87 (2019).
 See a video on this methodology which shows the protocol for direct-seeded SRI. See also the SNV website on the project which reports that greenhouse gas emissions are reduced by 22%.
 N.K. Kathia, B. Mati, J. Ndiiri and R. Wanjogu, ‘Integrated mechanical weeding and planting to reduce labour input under the System of Rice Intensification (SRI),’ Agricultural Sciences 10: 121-130 (2019).
 The practice of sowing germinated seed by broadcasting and then thinning out the plants that grow has also been reported in eastern India, but not with the geometric precision and wide spacing of SRI.
 S. Ramasamy, C. Susheela and K. Sathyamoorthi, ‘Direct-Planting System – Energy Saving, High-Output Rice Establishment Technique for Lowland,’ Tamil Nadu Agricultural University, poster at 2nd International Rice Congress, New Delhi (2006).
 This website has quite a variety of agricultural machinery, but interspersed with many other products.
 System of Rice Intensification Weeders: A Reference Compendium, WASSAN, Hyderabad (2006). This document reports farmers’ evaluations of the advantages and disadvantages of different weeder models, on a broader scale than the Kenya evaluation reported above.
 Simple but complete technical drawings of SRI equipment can be downloaded free from the Earth Links website.
 See first column of this website page on videos. This is the link to the Facebook equipment group.
 The reports presented at the workshop are available on the SRI-Rice website.
 Here is a powerpoint presentation by Ravindra A. on WASSAN activities, design criteria, and products that was shared at the 2014 international equipment workshop.
 See website on the Rice Dragon weeder for more information.
 Presentations are available on TNAU work on weeders and transplanters by Dr. K.V. Ravichandran and Dr. B.J. Pandian, respectively.
 Here is a report on this collaborative effort.
 In 2006, Alipati Satyanarayana took me to visit a large farmer in Andhra Pradesh state, N.V.R.C. Raju, who through good training, supervision and motivation of his hired labor force was able to use standard labor-intensive SRI methods on a rice farm of 44 hectares (110 acres). He got an average harvested yield of 7 tonnes per hectare, not calculated by sampling. See pages 24-25 of trip report. This was reported in D.J. Raju (no relation), ‘Rice farmers’ participatory research has played a key role in implementing the System of Rice Intensification,’ in K. Toriyama et al., editors, Rice Is Life: Scientific Perspectives for the 21st Century, Proceedings of World Rice Research Congress, Tsukuba, November 2004, pp. 212-214, IRRI/JIRCAS (2005).
PICTURE CREDITS: Anil Verma (PRAN, India) (2); Rajendra Uprety (Nepal); Gamini Batuwitage (Sri Lanka) (2); T.M. Thiyagarajan (India) (2); Norman Uphoff (Cornell); Oxfam/New Zealand publication; WASSAN (Hyderabad, India) (2); Glenn Lines (CIIFAD); Gopal Swaminathan (India); Norman Uphoff (Cornell) (3); Rajendra Uprety (Nepal) (2); Gamini Batuwitage (Sri Lanka) (2); M.J. Prabu (The Hindu, Chennai); Bancy Mati (Kenya) (6); Oscar Montero (Costa Rica) (3); Norman Uphoff (Cornell); Asif Sharif (Pakistan) (3); Rajendra Uprety (Nepal) (3); Bala Hussain Reddy (India) (3); SNV publication (Netherlands);
Endnotes: Tefy Saina (Madagascar) (2); Anoop Tiwari (Madhya Pradesh, India); Bhaskar Mahanta (Assam, India); Shuichi Sato (Nippon Koei, Japan); Y.S. Koma (CEDAC, Cambodia); Amrik Singh (ATMA, Punjab).