The Three-S System
(from SRI Trip Report to China - February, 2004)
Norman Uphoff
The Three-S System: This is the name given to the system that Prof. Jin Xueyong introduced in Heilongjiong, the northernmost province of China, with low temperatures and a short growing season. (Prof. Jin is now unfortunately deceased. See the presentation assembled from pictures and data that he presented on 3-S at a national rice conference in Haerbin in August 2004.)
Jin started working on this system experimentally in 1994 at about the same time that CIIFAD started working with Tefy Saina in Madagascar on evaluating and spreading SRI there. Jin did not start introducing his methods to farmers until 1999, the same year that the first SRI trials outside of Madagascar were conducted (at Nanjing Agricultural University in China) confirming the merits of SRI methods.
SRI had been synthesized by Father de Laulanié in Madagascar ten years before Prof. Jin started to develop his alternative methods. But it did not become known in China until 15 years later, by which time Jin had worked out his system. So clearly, his was an independent creation, a kind of 'rediscovery' based on the same genetic potentials that have always existed in the rice genome.
Both SRI and Three-S change certain practices that have inhibited or suppressed these potentials for centuries. Having two separately developed alternative systems demonstrating that these potentials exist makes both more significant. I told Prof. Jin that just as we are learning that plant roots 'cooperate' rather than 'compete' when in close proximity, both systems can and should cooperate, with each informing and improving the other. The three elements of Three-S are:
(1) Spacing: Rows are spaced 33 cm apart with 16 cm between plants, though on better soil the spacing between rows can be greater -- 36 cm, and even up to 40 or 43 cm. In principle, as with SRI, single seedlings are planted, though sometimes 2 seedlings (or seeds) are planted together. The plant density with Three-S is 13-17 per square meter, which is essentially the same as with SRI -- 16 plants per square meter (25x25 cm spacing in a square pattern). Although with Three-S there is row rather than square planting, the principle and practice is virtually the same.
(2) Super-Rice Varieties: Farmers are recommended to choose whatever hybrid or improved variety gives best response to the growing conditions of Heilongjiong and also the other elements of the Three-S system. This differs from SRI in that we suggest farmers continue using whatever variety that they are already using, since SRI methods can enhance the yield for any cultivar. However, it makes sense to select a variety that is most responsive to the methods under local conditions so that farmers can get the best possible return from their investment of land, labor, capital and water. We should be giving more emphasis to determining and suggesting the best available varieties for SRI use.
(3) Sustainability: With Three-S, farmers should prepare and add compost or other organic material to enrich the soil, though chemical fertilizer is also added to increase soil nutrient availability. (If I understood correctly, they are even experimenting with bear manure; I didn't know there were enough bears to have any significant supply.) There does not appear to have been any consideration of how inorganic fertilization may inhibit soil microbes, so the net gain from such amendments may not be as cost-effective as when trying to maximize/optimize the contribution from soil biological processes.
In China, inorganic fertilizer has largely replaced the use of organic-source nutrients, so it is hard to find anyone willing to rely entirely on organic fertilization even in experimentation. It is possible that given low soil temperatures and lowered microbial activity in Heilongjiong, there is more need to employ inorganic fertilization than in warmer growing environments. This is a question to be further examined. The Three-S promotion of adding organic matter to the soil is innovative under present conditions -- though not historically, as Chinese farmers have for millennia put organic matter back into their soil.
(4) Water Management: This is essentially the same with Three-S as with SRI. Three-S fields are not kept continuously flooded. Water is put onto the field for 3-5 days, and then it is drained and left unflooded for 5-7 days. Land preparation is done by tractors without flooding. When in-season fertilizer applications are made, fields are flooded before this, and weeding is also done.
With Three-S, the reduction in water use is 50 to 60%, according to Prof. Jin. Water shortages are a major problem in much of China. In Heilongjiong, a maximum acreage has been set for rice, 1.5 million ha, because the water resources in the province are not sufficient to grow more rice than this without sacrificing other crops (see discussion of this below). Three-S (and/or SRI) methods should be able to relax this constraint on rice production.
(5) Weed Control: With no continuous flooding, there is the same need to control weeds as with SRI. In Three-S, they use the same kind of 'rotating hoe' that is recommended with SRI. There is some recognition that this aerates the soil, but that has not been emphasized. Now that they know about SRI and understand the explanations that we propose for its success, NEAU faculty will probably look more closely at effects of soil aeration.
(6) Crop Establishment: Some differences between Three-S and SRI are necessary in this respect because of the different and very severe climatic conditions in Heilongjiong. I was shown pictures of the plastic-covered 'greenhouses' in which rice seedlings are started, with a foot or more of snow around them outside. This is far from the tropics.
Seedlings are grown in individual soil compartments on plastic trays (shaped like in egg cartons) with single seeds or two seeds put each into a separate 'plug' of soil that can be later removed from tray and put directly into the field with desired spacing. (Two seeds are planted when there is some doubt about whether single seeds will all germinate.) This technique is similar to the way that tree seedlings are often grown in plastic 'sleeves.' The roots are hardly disturbed by the transplanting process, and root tips remain pointed downward in a natural position, not inverted upward as happens with standard method of transplanting. With this planting method for rice, the soil medium is kept well-drained, never saturated as happens in flooded nurseries.
The question was raised whether with this system they are transplanting 'too late' according to our SRI understanding. If the transplanted seedlings experience no disturbance or trauma, it probably does not make much difference when the transplanting occurs, as long as it is before the seed root outgrows the separate space provided before it. Right now, transplanting with the Three-S system is usually done at the 6-leaf stage, when the plant has three tillers (about the fifth phyllochron). Whether productivity would be higher by transplanting earlier is an empirical question. Transplanting is now done between May 20 and June 2. Transplanting earlier than this would probably be harmful to the young plants given the cold temperatures in Heilongjiong.
The SRI principle is not: one should always transplant when seedlings are very young, with only 2 leaves, before the 4th phyllochron. Rather it is: when transplanting, there should be minimum or no trauma to the young plant and especially not to its root system. The use of containers for starting seedlings may make 'age of transplanting' irrelevant for SRI, since when the soil around the root is kept intact, there is no trauma, no inversion of root tips, and no loss of seed sacs.
Three-S transplanting is presently done with strings put across the field in parallel lines and spaced usually 33 cm apart. Knots every 16 cm apart indicate to transplanters where to place the seedling. The 'rake' method developed in Madagascar for marking a grid on the field and now used in other countries because it requires less time than stretching and moving strings may not be as much benefit in Heilongjiong as elsewhere, because spacing between plants within rows need not be so exact. However, the 'roller' that farmers in Andhra Pradesh have developed for SRI may be useful here, if Three-S fields are only muddy, not flooded, when transplanting is done. Right now they are transplanting into fields with more water than recommended for SRI.
Rice production has been greatly increased and improved in Heilongjiong since the Chinese Revolution. In the 1950s, the area under rice was 110,000-340,000 ha, with an average yield of 2.3 t/ha. In the 1960s, the area averaged around 200,000 ha, with a yield of 2.1 t/ha. The period 1970-83 saw yields begin to rise, to 2.9 t/ha, while area was 170,000-280,000 ha. Between 1984 and 1996, yield more than doubled, to 6.2 t/h, and rice area expanded to 780,000-1,100,000 ha. From 1997-2000, rice yield went up further, to 6.8 t/ha, with area reaching 1,700,000 ha. But this was more area than can be served by available water supplies, without reducing other, more valuable crops. So as noted above, rice area has now been officially limited to 1.5 million ha.
The Three-S system is expanded to about 650,000 mu (over 40,000 ha) in Heilongjiong, still a relatively small share of its total rice area. But the benefits are definite: an average yield of 8.5 t/ha, with a 60-60% reduction in water requirements. The goal of Prof. Jin's program is to reduce costs of production by 10% while raising yield by 20%, with higher quality grain produced. The Three-S method is being extended also in Liaoning Province and Inner Mongolia.
