A more extensive report is being prepared by the China National Rice Research Institute (CNRRI) which organized this workshop in Hangzhou. It was cosponsored by the China National Hybrid Rice Research and Development Center (CNHRRDC) and the Zhejiang Provincial Department of Agriculture, the province in which CNRRI is located.

Opening Session: The workshop chairman, Dr. Zhu Defeng (CNRRI), opened the event by thanking the participants from various research institutes, universities and extension departments for coming. He noted that SRI research has already demonstrated the system's validity in many different ecosystems in China and then reviewed the purpose and mission of the workshop.

The workshop was opened with several introductory talks, starting with Dr. Wang Qingli from the Ministry of Agriculture in Beijing. He discussed changes going on in the agricultural sector, particularly with China's entry into the WTO, and the Ministry's growing concern for impacts on the environment. He suggested that SRI would be very compatible with China's planned "adjustment of the agriculture structure and eco-agriculture." This strategy includes the use of best varieties and reductions in the use of fertilizer and water. He suggested that participants consider how to make SRI easier for farmers to learn. He said that the Ministry would continue supporting research and work on SRI according to its results.

Prof. Zhang Hongfang, deputy director of the Zhejiang Department of Agriculture, spoke next, discussing how the rice sector is changing in Zhejiang province. About 80% of the rice area is now single-cropped, with other crops being grown in alternation with rice. This is a favorable situation for SRI since such crop rotation is ideal for this system. SRI can give a higher yield from a single crop of rice in part because growing vegetables or pulses in between rice crops improves soil fertility for rice.

Dr. Zhu Deqi from the Zhejiang Department of Science and Technology emphasize that rice is the biggest crop in the province. SRI, he said, should be especially good for non-rich farmers, being a "simple" technology and one that should be easier to disseminate than more complicated, input-dependent technologies. He said that his department looked forward to getting information on SRI from other provinces and other countries from the workshop.

Prof. Liao Xiyuan, deputy director of CNRRI, standing in for the Institute's director-general who had to be in Beijing, observed that good land area for rice production is declining because of urban expansion and soil degradation, so it is important to raise yields. He said that increased quality of rice is also sought, adding that there is need for more support of research. He put in a plug for my presentation on SRI scheduled for the afternoon, saying it would "open our minds."

Dr. Ma Guohui, deputy director of CNHRRDC, then spoke on behalf of his Center, apologizing that its director, Prof. Yuan Longping, could not be present since he is attending the People's Consultative Congress in Beijing. He said Prof. Yuan is very happy about this workshop. This was an important endorsement for SRI since Prof. Yuan, regarded as "the father of hybrid rice," is China's best known rice scientist. Dr. Ma said that SRI is showing that the management of rice is as important as breeding.

Dr. Zhang Hongcheng from Yangzhou University also spoke in this opening session, but I had no translation of his comments. After his comments, we adjourned for a group picture.

Morning Technical Session: When the workshop regrouped, I made a brief comment to introduce the subject of SRI. I underscored that this system is not a single thing and should be seen as 'a work in progress.' I suggested a historical comparison between SRI development and the development of aviation, noting that Fr. de Laulanié's first SRI harvest in 1983-84 in Madagascar was like the Wright Brothers' flight at Kitty Hawk 100 years ago. This was the equivalent for rice of breaking the constraint of gravity by a heavier-than-air craft. After 20 years, SRI is now a serviceable system, reliable enough to be compared with the old DC3 aircraft that was the workhorse of the aviation industry for many years. But it was still a propeller-driven craft, as with current SRI methods we can get an average yield around 8 t/ha.

Airplanes were rapidly and variously developed during the first part of the 20th century, with improvements in wing design, engine power, navigation systems, etc. There were British planes, French planes, Russian planes, Chinese planes, etc., and different models within each country. We can see SRI similarly proliferating today, to everyone's advantage. No single model is needed or desirable.

The biggest advance in aviation during the past century came with the development of jet propulsion, to power larger planes faster. I suggested that this kind of transition is still to be made with SRI, though some of the "super-yields" attained with SRI show that much better performance is still possible. Better understanding and management of soil microbial activity could give SRI a large further boost, equivalent to jet propulsion. But this remains conjecture, not established scientifically. I said that I would discuss this subject more in the afternoon.

Dr. Zhu Defeng of CNRRI, who had taken the initiative to organize the workshop, then started the session off with a report on his institute's SRI research. He said that the rice sector should seek (a) to overcome the problem of declining yields, (b) to reduce the requirements for water and for fertilizer and chemicals (the latter adversely affecting water quality), (c) to improve the environment (including reduction in methane emissions from flooded paddies), (d) to increase resource efficiency in rice production. He briefly reviewed the history of SRI, starting from Fr. de Laulanié's experiments in 1983, saying that SRI had been introduced to China by Prof. Uphoff. Results so far have ranged from 10.6 to 16.0 t/ha, showing that SRI has great potential.

Dr. Zhu then focused on his own research on phyllochrons in rice, also referred to as leaf number. It was exciting to see his confirmation of Laulanié's work 15 years earlier, showing the different numbers and contributions of primary, secondary and tertiary tillers. He presented data showing different patterns of light penetration into the rice canopy with SRI vs. conventional spacing. It was clear that SRI practices give greater illumination to all levels of the canopy.

Dr. Zhu acknowledged that SRI requires more labor at the beginning, but this can be reduced with experience. Water control is another requirement, though this should not be a great problem in much of China, where reducing water off-takes for agriculture is becoming an imperative. There does need to be more knowledge for effective SRI, both on the part of farmers and researchers.

Improvements in SRI that he envisioned as being are particularly relevant for China are:

• use of plastic trays for planting and transplanting seedlings (minimizing root disturbance and saving labor);
• finding optimum combinations of chemical fertilizer and compost;
• developing raised-bed cultivation and furrow irrigation;
• identifying the best rice varieties with high tillering ability; and
• innovations in spacing, such as the triangular pattern already devised in Sichuan.

Dr. Ma Guohui of CHNRRDC then reported on his Center's evaluation of SRI used together with its hybrid varieties. The Center began evaluating SRI since 2000-2001. The triangular pattern of planting that Dr. Zhu had referred to was developed by a seed farm associated with the Hybrid Rice Center. It has given better results than single seedlings in a grid pattern in 7 out of 10 trials. The Center finds that using various seedbed media with plastic trays or paper containers gives good results. It has also experimented with a paper template when planting seedbeds to space single seeds more evenly. This offers some advantages. He showed a picture of a weeder that piles up a little soil around the culm (base) of the plant, and this appears to reduce senescence of leaves. Dr. Ma reported a similar range of yield results for SRI methods as did Dr. Zhu.

The improvements he suggested were also similar to Dr. Zhu's:

• use of trays for seedling planting and transplanting;
• combining compost and chemical fertilizer;
• combining herbicides with use of the mechanical hand weeder that aerates the soil;
• mechanization for transplanting and weeding; and
• finding ways to increase tillering, through breeding and/or soil and water management.

In the discussion that followed, one of the questions raised was whether SRI affects the quality of the grain. Someone responded that using organic fertilization improves rice quality, and someone else added that his evaluations have shown that SRI methods reduce chalkiness. A question was asked about possible changes in the harvest index with SRI. Dr. Ma said that no decrease had been found, and that with modified SRI, there are both more roots and more sink capacity. The morning session concluded promptly at noon, and we adjourned for a splendid lunch upstairs.

Afternoon Technical Session: Dr. Ma Jun from the Rice Research Institute of the Sichuan Agricultural University began with a report from this province which has the most practical experience with SRI so far. The Sichuan Academy of Agricultural Sciences is in its third year of evaluations. Its field trials in four locations during 2002 gave an average yield of 13 t/ha. Dr. Ma said that rice breeding efforts in China have produced and released many improved varieties, but there has been little impact on overall productivity from this.

In Sichuan, they have encountered some difficulties with SRI when 15-day-old seedlings are still too young for the colder temperatures. Also, sometimes it is difficult to get enough organic fertilizer to use instead of chemical fertilizer. They have been using weedicides with SRI because of labor constraints. He said that the seed-set rate with SRI has often been lower (though in my view this could be because chemical fertilizer and herbicides are being used). The various trials have shown a 6.4-16.3% increase with SRI methods, and in one case, a 7.7% decrease, compared with control plots.

Dr. Ma showed data confirming that younger seedlings give higher yield, other things being equal (not reflecting the synergies that are possible when using the full set of SRI practices). Also data showed better results with a combination of chemical and organic fertilization, though the best results reported came from using organic fertilizers together with microorganisms. He showed a picture of a farmer "swimming" in his rice field, with plants up to his shoulders. He commented that although biomass is greatly increased, there was no evident problem of lodging.

Dr. Ma then showed data on reduced chalkiness of rice grains with SRI methods and on increased length of panicle (147 vs. 114 cm), adding also that less disease was seen with SRI. There are larger root systems with SRI, and increased quality of root exudation. He concluded with some economic data from their demonstration areas. With SRI methods, the increase in profitability/hectare was between 1,135 and 3,567 yuan, i.e., between $150 and $450/hectare.

Dr. Ma summarized the "essence" of SRI as single plants, wide spacing, less water, and more attention to the root system. He said they have already formed an organization in Sichuan to promote work on SRI with 5 institutes cooperating. They want to focus on how to make it easy for farmers to learn and use SRI.

The next speaker was Dr. Yao Kemin from the Nanjing College of Meteorology, who had done an analysis of SRI's suitability for different regions within China according to temperature and other factors. His data did not show much or any advantage to using SRI methods. According to him, SRI should reduce harvest index and yield. His conclusion was that SRI will be suitable only in the south of China such as Hainan, Guangdong and Guizhou, where the climate is more like that in Madagascar. This was at variance with the report just made from Sichuan province, where SRI was being used successfully despite a colder climate. Indeed, the highest SRI yield so far achieved in China, 16 t/ha officially certified, is from Sichuan province, not Hainan.

Dr. Jin Xueyong, director of the Research Institute of North East Agricultural University (NEAU), reported on SRI experimentation and evaluation in Heilongjiang province. His pictures of vigorous SRI rice growing in the north also contradicted Dr. Yao's climatological assessment. At NEAU they have made, appropriately, some modifications in SRI practices because of the colder temperatures. Pictures were shown of rice seedbeds being started in plastic-covered greenhouses. The data presented were shown quickly and I did not have translation of their explanation, but they appeared to show yields in the range of 10.3-11.68 t/ha.

Dr. Tao Longxing from CNRRI made a presentation on the physiological characteristics of SRI, which he has been studying at the Institute. Like Dr. Zhu, he contributed a very interesting paper to the international SRI conference held in April 2002 at Sanya. It was interesting that Dr. Tao's data showed leaf area greater with SRI than the control, but leaf area index greater with the control than with SRI. Root dry weight was greater both at heading (13.2 vs. 8.2 g/hill) and at ripening (7.6 vs. 5.2 g/hill). Root penetration reached to 55 cm vs. 35 cm. The root pictures shown by Dr. Tao gave visual evidence of the large differences in root development with SRI.

There were big differences in tillers/hill: at the peak, 50 with SRI (single seedlings) vs. 20 with the control (multiple seedlings). The SRI plants showed greater height throughout the growth cycle. There were similarly large differences in size of shoots and panicles, with seed-set remaining high with SRI plants and falling to zero with the control. Dr. Tao's research did not evaluate yield differences, only measuring and documenting the physical differences in plant growth patterns that correlate closely with the yield results that others were reporting.

My presentation reviewed briefly the principles and practices of SRI, and then its history and spread beyond Madagascar. China was the first country outside of its country of origin where SRI was evaluated: at Nanjing Agricultural University in 1999. Then, in 2000-2001, the China National Hybrid Rice Center under its director Prof. Yuan tried out SRI methods and encouraged CNRRI, the Sichuan Academy of Agricultural Sciences and other institutions to do their own evaluations. This workshop has grown out of the encouraging results of these trials and reflects a increasing interest in SRI in China.

In the time I had been given, I focused on kinds of changes in soil microbiology that could be contributing to the higher yields being achieved with SRI methods. These increase root growth through the use of young seedlings, wider spacing and aerated soil. These conditions are in turn probably supporting soil microbial abundance and diversity through root exudation. I showed one data set from Madagascar that supported this tentative conclusion. I suggested that:

• Biological nitrogen fixation (BNF) can occur in rice roots and the rhizosphere, contrary to the idea that BNF occurs only with leguminous plants; there is research showing that BNF increases when aerobic and anaerobic soil horizons are mixed (Magdoff and Bouldin 1970);
• P solubilization increases in soil that is wetted, dried and rewetted, by very large percentages (Turner and Haygarth 2001);
• Mycorrhizal fungus "infections" of plant roots greatly increase the amount and variety of nutrients accessed by roots; this symbiotic association benefits 80-90% of plants, but not continuously irrigated rice because soil fungi cannot grow under hypoxic conditions;
• Rhizobia in the rhizosphere of rice increase yield and protein content per hectare, not through BNF, as these bacteria do in the nodules of legumes, but through production of auxins and other growth-promoting substances (Yanni et al. 2001);
• Protozoa that "graze" on bacteria living on the roots of plants exude N on the roots' surface because they require a lower C/N ratio than contained in the bacteria (Pinton et al. 2000).

Root exudation is probably a key factor in these various processes, as with SRI methods of plant, soil, water and nutrient management, canopies and root systems of rice are much larger. These should produce more exudate and increase rhizodeposition. There is little scientific research on these processes in rice, however, mainly because rice has been grown mostly under flooded conditions. In anaerobic soil, many of these processes are inhibited or suppressed. Consideration of soil microbiological contributions to increased rice production opens up some promising areas for research and for improved production practices. We think a better understanding of soil microbiological contributions to rice could also help improve the production of other crops.

Ms. Mao Guoquan from the Zhejiang Department of Agriculture spoke next. She said that 9 institutes in her province have been doing SRI experiments, with yield increases of 13-34%. I did not get all of her points because her remarks were not completely translated for me, but her conclusions were that SRI is very suitable for Zhejiang Province; it is very good for the environment, a growing concern; it gives good grain quality, and there is less cost of production. Finally, she commented that the system is "easy" for farmers to adopt. I could not tell whether this was a statement of experience in Zhejiang or a prediction based on what they know of SRI.

Dr. Zhang Hongcheng from Agricultural College of Yangzhou University reported on three years of trials. These have shown that one plant per hill is best, and that SRI methods produce more grains per panicle. I could not get all of the information he presented, but the yield of 674.7 kg/mu that he reported with SRI methods amounts to a yield of 9.7 t/ha, a 13.1% increase over the control shown. Another result reported with 20x30 cm spacing represents a yield of 10.5 t/ha.

Dr. Zhang Yuzu from the Hainan Rice Research Institute reported that with SRI methods in direct-seeded rice, weeds become a problem and need to be controlled by chemicals or by hand. Sparser planting (lower plant density) gives higher yield. He said that with the merits of SRI being clear, the next step is to develop improved and simplified methods of SRI. His institute recognizes that root growth is particularly important and is trying to promote this by SRI. It is now working on methods for raised beds and ridging.

Dr. Wang Xiehui from the Anqing Regional Agricultural Institute in Anhui spoke next. I did not get much from this presentation because I had no translation, but the Anhui presentation at the Sanya conference in April 2002 provided some of the strongest Chinese evidence of benefits from SRI practices.

To conclude the afternoon, Mr. Liu Zhibing, farm manager of the Meishan Seed Company in Sichuan province, reported on his experience with SRI. He has had yields of 15.996, 15.572 and 15.6 t/ha with seed-set rates around 90%, with panicles having 185-190 grains each. He discussed the "triangular" method of seed transplanting that he has devised, planting three seedlings per hill but having only half as many hills as in a standard SRI grid pattern because the hills are alternated. This gives 50% more tillers per hectare than with usual SRI practice of single seedlings per hill. His rice plants averaged 80 tiller at 60 DAT, half of which became fertile with an average of 250 grains per panicle. He rotates his rice crop with a potato crop, and he is now experimenting with inserting a crop of rape into the cycle as well. The photographs of his rice crops were very impressive, and his enthusiastic presentation held participants' attention even though, with so many reports, his report did not finish until after 5 o'clock.

With the workshop was well beyond its scheduled time to conclude, Dr. Ma Guohui gave a very clear and concise summary of the afternoon's presentations before the participants adjourned for the workshop's gala supper, with good food, drinks and camaraderie.

Planning Session: The next morning, the participants reconvened in two discussion groups, one focusing on research issues and priorities, and the other on extension strategies and methods. (Since I could not follow the discussions in Chinese, I went to Zhejiang University to make a presentation on SRI to students and staff in the College of Environmental and Natural Sciences.) In the afternoon, each group reported the results of its discussion in a plenary meeting, and Dr. Zhu summarized the workshop conclusions before closing. There were workshop participants from 11 provinces who contributed to the planning efforts.

CNRRI already has a web site for SRI. This will be further developed. There should be a lot more data on SRI experience generated during the present year 2003 to put on it. Various modifications and innovations in the basic SRI methodology will need to be made to suit different Chinese conditions. The web site will disseminate innovations and evaluations thereof.

There is agreement that reduced water application and good water control, young seedlings, wider spacing, and soil improvement with changes in fertilization to improve soil microorganisms will lead to greater and more sustainable rice production in China. One of the objectives of SRI development will be to reduce water pollution by reducing the use of chemical fertilizers and agrochemicals and to increase water use efficiency. There is need to devise a better rice agroecosystem that can be adapted to local conditions.

The creation of a national SRI network was agreed upon. It will be coordinated by Dr. Zhu Defeng of CNRRI and will have a steering committee headed by Dr. Zhu and Dr. Ma Guihui of CNHRRDC. Its membership includes:

• Prof. Ma Jun from Sichuan Agricultural University,
• Prof. Zhang Hongcheng, chairman of the College of Agriculture of Yangzhou University in Jiangsu Province,
• Prof. Jin Xueyong, director of the Rice Research Institute of the Northeast Agricultural University in Heilongjiang province, and
• Prof. Sun Jian, director of the Crop Management Division, Zhejiang Department of Agriculture.

The steering committee on behalf of the network will seek to mobilize increased financial support from government sources and donor agencies for research and extension work on SRI, will set up research priorities, and will organize meetings and other activities as needed.

It was agreed that all participants will undertake research and demonstration activities on SRI in different ecosystems and will develop adaptations of SRI suitable for their environments using whatever resources can be obtained at present, so that the understanding and practice of SRI in China begins to expand. Any increased support that can be obtained from government and other sources will accelerate this process and enable participating institutions to spread faster the benefits of SRI for people and for the environment.

POSTSCRIPT: After the Hangzhou workshop, I spent a week with the College of Humanities and Development of China Agricultural University in Beijing, since I serve on its external advisory committee. I was able to meet with Prof. Qu Dongyu, vice president of the Chinese Academy of Agricultural Sciences, who oversees the work of CNRRI and many other research institutions, to brief him on the workshop and tell him more about SRI. I was not able to meet the president of the Academy, Prof. Zhai Huqu, formerly president of Nanjing Agricultural University, and the first sponsor of SRI trials in China, because the Chinese People's Consultative Congress was in session. But I provided Prof. Zhai with a copy of the Sanya conference proceedings.

I also met with Prof. Jin Yimin, director in the Department of Rural and Social Development in the Ministry of Science and Technology, and Ms. Meng Jiandong, deputy chief of staff in the Ministry, whom I had told about SRI in August 2002 when she visited Cornell with a MOST delegation. They are now also interested in SRI. Dr. Zhu from CNRRI will follow up with both CAAS and MOST as well as the Ministry of Agriculture.

I was also able to speak at length by phone with Dr. Cao Weixing, senior administrator at Nanjing Agricultural University, who actually did the first SRI trials with Dr. Ding Yanfeng in 1999. Cao had planned to attend the Hangzhou workshop, but he is now a member of the People's Consultative Congress, as vice-president of the Jiangsu Provincial Consultative Congress. His presence was required in Beijing. He expressed his strong interest in continuing to work with SRI and with colleagues in the SRI network both in China and internationally. Dr. Ding is now back at NAU after doing a PhD program in Japan, and he will provide leadership for SRI research in Nanjing.

After the workshop, Mr. Liu, farm manager of the Meishan Seed Company in Sichuan, gave me a beautiful, full-color, plasticized sheet with pictures of his spectacular SRI plots, one of which was certified by the Provincial Department of Agriculture as having produced a yield of 16 t/ha. The sheet also had pictures of him with the trophy and certificate he received for his accomplishment, of him with Prof. Yuan who introduced SRI methods to him, of him with me during a visit to his farm last September, and of President Zheng Zimin addressing the International Rice Congress last year, talking about the need to increase rice productivity in China. Liu is promoting SRI within his province and beyond. With someone so motivated and talented publicizing SRI at his own initiative, and with results like his to publicize, the prospects for SRI look good in China, over and above what the new SRI network will be able to achieve.

* * * * *

Following the workshop, I spent a week in Beijing with the College of Humanities and Development at China Agricultural University, for which I serve as chair of its external advisory committee. During this time, I had an opportunity to meet with Dr. Qu Dongyu, vice-president of the China Academy of Agricultural Sciences (CAAS), who has oversight for CNRRI and a number of other such institutions. I briefed him on the workshop and said that Dr. Zhu would follow up with more information. Because the National People's Consultative Congress was in session, I was not able to meet with Dr. Zhai Huqu, president of SAAS and former president of Nanjing Agricultural University (NAU). He hosted the first seminar I was able to give on SRI at that university, and as a prominent rice breeder he took an interest in SRI, supporting the first trials in China. We hope to have his continuing interest.

I was also able to meet with Dr. Jin Yimin, director in the Department of Rural and Social Development of the Ministry of Science and Technology (MOST), and Ms. Meng Jiandong, deputy chief of staff in MOST, to report on the workshop. Madam Meng visited Cornell in August 2002 with a MOST delegation and took an interest in SRI. We hope that this ministry as well as the Ministry of Agriculture will give support to the expanding work on SRI in China.