Rice - Fish Culture in China (IDRC, 1995, 240 p.) Part II: Patterns and Technology Different Methods of Rice-Fish Farming New Techniques for Raising Fish in Flooded Ricefields Methods of Rice-Fish Culture and their Ecological Efficiency Ridge-Cultured Rice Integrated with Fish Farming in Trenches, Anhui Province Development of Rice-Fish Culture with Fish Pits Techniques Adopted in the Rice-Azolla-Fish System with Ridge Culture Semisubmerged Cropping in Rice-Fish Culture in Jiangxi Province Rice-Azolla-Fish Symbiosis Economic and Ecological Benefits of Rice-Fish Culture Cultivating Different Breeds of Fish in Ricefields Rice-Fish Culture in Ricefield Ditchponds Rice-Fish Culture in China: Zero-Tillage Ricefields Demonstration of High-Yield Fish Farming in Ricefields Rice-Azolla-Fish in Ricefields

Rice - Fish Culture in China (IDRC, 1995, 240 p.)

Part II: Patterns and Technology

Different Methods of Rice-Fish Farming

Ni Dashu and Wang Jianguo

Many methods of rice-fish farming have been developed in China. Although they involve various production systems, these different methods are inseparable and interlinked. The common aim is to boost rice production by eliminating weeds and pests. Many different types of rotation are practiced.

Rice-Fish Mutualism

Early, middle, and late rice are planted continuously without interruption. Two kinds of fry (fingerlings and summer fry) are released directly into the flooded ricefields. Specific practices include raising fingerlings in flooded ricefields, raising fish in ricefields and in nearby ponds, planting rice on the ridges while raising fish in the furrows, and raising fish in ricefields in which channels have been dug.

Breeding Fry in Ricefields

To reduce the cost of summer fry, a model has been devised that involves releasing fry directly into early flooded ricefields. Grass carp (Ctenopharyngodon idella) fry are generally used, and because feed is not needed, the method is economical.

After middle rice is planted, 1000 fry, 3.3-5 cm in length, can be harvested from the early ricefields. Because costs are kept to a minimum, it is easy to popularize the method in areas with large expanses of water. The early rice planting season (late April) in Hunan, Jiangxi, Anhui, Jiangsu, and Zhejiang coincides with the production of common carp (Cyprinus carpio) fry. Therefore, after the rice seedlings have been transplanted, the ditches dug, and the screens installed, C. carpio fry can be released into the ricefield. The fry are too small to uproot the seedlings and because this is the peak period for plankton, fry growth is enhanced. It is best to release the fry as early as possible to take full advantage of this peak in plankton growth. If the artificial hatching of fry is delayed, the application of base manure and the transplanting of early rice should also be delayed to maximize the mutual benefits that can be achieved.

At present, large fish-fry breeding farms have advanced the season of fry hatching to late April and the rice growers on these farms delay the transplanting of rice seedlings. But the practice has not gained much popularity. Additional effort is needed to disseminate the idea and launch demonstration projects. For every hectare of ricefield, 45 000 artificially hatched fry are needed. Along the Yangtze and Zhujiang Rivers, where people catch river fry, it is better to put these fry into early ricefields because it makes it easier to regenerate fish of the same family. Stocking fry, which have just begun to eat, into early ricefields 3-4 days after the rice is transplanted offers many advantages. It eliminates the need to buy summer fry, means that ponds are not needed for summer fry, and maximizes the mutual benefits of growing rice and fish together. It is an economical, practical method that yields better and faster results.

Ctenopharyngodon idella, with its ability to eliminate weeds and worms, can help increase rice output and reduce the need for labour. However, in areas with few ponds or banks, other fish species (e.g., C. carpio, crucian carp, and tilapia) can be raised. Even in ricefields overgrown with weeds, it is feasible to grow C. idella along with some C. carpio and C. auratus.

When this method is adopted, the banks of the fields must be raised 50-70 cm and strengthened before the fry are released into the field. Lime is applied (375-750 kg/ha) to kill leeches, eels, and other natural enemies of the fish. Six to eight days later, water is channelled into the field and base manure is applied. The field is raked level, and the rice seedlings are transplanted. Fish canals and ditches (30 cm wide and 30 cm deep). Where the canals cross, a fish ditch 100 cm long, 70 cm wide, and 80-100 cm deep is dug. Rice seedlings in the canals should be transplanted to the edges of the field to from a fence. Screens, each 100 cm wide and 80-90 cm tall, should be installed in the water inlet and outlet. Each screen should be arch-shaped with thin bamboo strips placed 0.2 cm apart. Fry may then be released into the field. Field management should be strengthened. Before the rice ripens and all the weeds are eaten by the fish, the canals and ditches are opened and the water is drained slowly to force the fish to gather in the canals. The fish are then driven into the ditch where they are netted.

This is the best method for raising fingerlings. In Sanming City, Fujian Province, the area of ricefields for raising fingerlings increased in 1982-84, from 270 ha to more than 670 ha, and the number of fingerlings raised increased from 2 to 8 million (62% of the fingerlings raised in the entire city). In addition, rice output increased by 6-17%.

The catch of adult C. idella from fishponds has remained low. Because they are unable to adapt to the environment in fishponds, the fish easily become ill. Usually only 20-30% survive. In ricefields, on the other hand, the ecological environment is suitable for C. idella, and few, if any, become ill. This is why the output of freshwater fish can be doubled.

Rice, Fish, and Azolla

In this method, the raising of C. idella or tilapia in ricefields is organically combined with the growing of azolla. Rice is grown in the field, azolla on the surface of the water, and fish in the water. Fish feed on the azolla, and the field is fertilized by fish excrement. Melons and beans can also be planted on the banks of the field to form a vertical cultivation system. Instead of the conventional equal-distance planting method, rice growers use wide and narrow rows. They raise azolla and fish in the wide rows and plant rice in the narrow rows. This keeps the field well ventilated and maximizes the use of sunshine and the effects of edge rows. As a result, it ensures stable and high yields of rice and fish, good economic returns, social benefits, and ecological efficiency.

This research was sponsored in recent years by Liu Zhongzhu, President of the Fujian Academy of Agricultural Sciences. After 3 years of experimentation, the method is now widely applied in Jianning County, Fujian Province. In 1986, the county devoted 6670 ha or 46% of its ricefields to this method of farming. The method generates additional income of CNY 2250-2700/ha of rice planted, and rice output can be increased by about 7%. Because there are fewer weeds and pests in the fields, there is less need to apply chemical fertilizers and pesticides, which helps reduce costs by about CNY 150/ha. The county reported a total fish catch of 1.15 million kg from ricefields in 1985 and 1.5 million kg in 1986. Most of the fish were sold in the market, which added, on average, CNY 20 of income per household.

Raising Fish in Ricefields with Wide Ditches

This method is used to raise winter fingerlings. Ditches, about 1-m wide and 1-m deep, are dug on the water inlet side and inside of the field bank. The total area of the ditches is about 5-10% of the area of the ricefield. The ditch ridge is raised 25 cm above field level. A 24-cm opening every 3-5 m links the ditches with the field and allows the fish to move freely from the ditches to the field. Long before the rice-transplanting season, winter fingerlings are put in the ditches so that they can enter the ricefield for food as soon as the early rice seedlings turn green. Jiangxi Province devoted 6670-9330 ha of ricefields to this method in 1985-86 and reported a 20-50% increase in rice output.

Ricefield Plus Fish Farming in a Pond

In rice-fish farming, there is a time difference of about 1 month between the early rice and the hatching of summer fingerlings. The rice plants need sunlight, fertilizer, and pesticides. These conditions are not favourable for fish farming. In areas where a double rice crop is planted, the fingerlings and early rice must be harvested between the two rice crops, the field must be worked, and the late rice must be transplanted. At the same time, the ditches must be redug and fry released. Therefore, there is a need for more labour than is available. If rice-fish culture is combined with pond culture these contradictions can be eased. The method is easy to popularize.

One basic condition is that there be ditches or ponds around the ricefield. The pond should be 10-30 m² and about 1.5 m deep. The pond can be dug in advance and should be linked by a bank to the ricefield. It can also be used to hatch the fry. After the early rice is transplanted and the fish canal dug, the pond and ricefield are linked to let the fish in the pond swim across into the ricefield. Just before the early rice is harvested, the fish are driven back into the pond. After the field is reworked, the second rice crop is transplanted, a ditch is dug, and the fish in the pond are allowed back into the ricefield.

In 1983 at Lingshan Village in Meichuan District, Guangji County, Hubei Province, a rice farmer named Hu Maoyu used this method on a 0.17-ha ricefield linked to a 0.02-ha natural pond. He raised fish in the ricefield for 348 days, including 117 days when rice and fish lived together (61 days for early rice and 56 days for late rice). He put in 2143 fry and netted 1770 fish that had a net weight increase of 216.2 kg and a harvest rate of 82.6% (Table 1). The output was 5431 kg/ha for early rice and 4073 kg/ha for late rice, or 5.81% more than the output from fields in which fish were not raised. Average net income was CNY 2156/ha. This method is gradually gaining popularity.

Rice-on-Ridges and Fish-in-Furrows

Ridges are built in the ricefield for the rice and fish are raised in the furrows. This method was developed on the basis of a semidry cultivation method advocated by Hou Guangjun. This method improves low-yielding ricefields because it makes multiple uses of available resources. It helps increase the contact of soil and air; balances water, air, and heat to raise soil temperature; and reduces the formation of toxic matter. Soil, water, microclimate, and heat are therefore stabilized at an appropriate level. This stimulates rice to grow roots, which absorb water and nutrients and changes gravitational water in the ricefield into lateral water that rises through capillaries to moisten the rice roots. Movement of fish in the furrows moves the water in the lower strata, stimulates the solution of nutrients, and increases soil fertility. The deep furrows increase the volume of water stored in the ricefield and create more room for fish activity. Fertilizers applied in the furrows make the water fertile and increase natural feed for fish.

In 1986, 16 counties of the Southeast Miao and Tong Autonomous Prefecture of Guizhou Province popularized this farming method over 688 ha. To ensure its success, the prefecture and the country earmarked CNY 100 000 for the project. Thirty six persons went on a study tour to Sichuan Province and 83 persons from the Departments of Aquatic Products and of Soil and Agricultural Technique Popularization were sent to the fields. Various districts, townships, and villages ran 21 training courses for 1000 people. In 1985, a 4.5-ha experiment area yielded more than 10 350 kg/ha of rice and 472 kg/ha of fish.

Specifically, the method involves digging a ditch, 50-cm wide and 67-cm deep, and building ridges 70-cm wide (enough to plant 4-6 rows of rice). Mud from the ditch is spread onto the ridge and rice is transplanted without working the soil. In a 0.07-ha field, 300 5-cm fingerlings [100 C. idella, 75 silver carp (Hypophthalmichthys molitirix), 50 bighead carp (Aristichthys nobilis), and 75 C. carpio and C. auratus] are released. During the growing season, green grass is put into the ditch to feed the C. idella, but no other feed is provided for the other species.

Research using this high-yielding, high-efficiency semidry cultivation method in Chongqing City showed that yields of 6750-7450 kg/ha of rice and 705-765 kg/ha of fish could be achieved. This method has been popularized in the Mianyang and Huangbo Counties of Hubei Province, and in Hunan and Jiangxi Provinces where conditions are suitable. Good economic returns have been reported.

Rotating Rice and Fish

In this method, rice and fish are alternatively raised in one ricefield. In 1 year, only one rice crop is planted, the rest of the time is devoted to fish farming. First, rice and fish are farmed in one field. When the rice is ripe, the rice and fish are harvested and the straw is left in the field to rot. Adult fish are then released into the harvested ricefield. The method can also be applied in double-cropping areas, but the fish are only raised in winter.

Rotating Rice and Fish in Low-Lying Land

In 1982, this method was adopted on 1.3 ha of low-lying land farmed by the Luopitang Production Brigade of the Huaqiao People’s Commune in Guangji County, Hubei Province. This piece of low-lying land previously grew only one late rice crop a year and remained fallow for the rest of the year. On 2 July 1982, fish ditches (50 cm wide and 27 cm deep) were dug, and the next day, rice seedlings (Gu-154) were transplanted at a distance of 11.5 x 17 cm. The field was not weeded during the entire rice-growing season and no pesticides were applied. Only 300 kg of sodium bicarbonate (232.5 g/ha) and 140 kg of urea as a top dressing (109 kg/ha) were used. The rice output was 5530 kg, 10% more than the expected 5000 kg, with a per-unit output of 4298 kg/ha. On 23 July, 19 690 fingerlings [84% C. idella, 5% black carp (Mylopharyngodon piceus), 10% H. molitirix, and 1% A. nobilis] were introduced at a rate of 115 300/ha. The fish were grown for 64 days without feed and on 24-25 September, 10 094 fish, weighing a total of 229.5 kg (176.5 kg/ha) were collected. Ten percent of the fish were 10 cm in length, 70% were 10.1-20 cm, and 20% were over 20.1 cm.

During the second rotation season, 10787 fingerlings (vaccinated for C. idella bleeding) were introduced (8385/ha). The total weight was 279 kg and the average size of the fingerlings was 15.6 cm. A small amount of fertilizer was applied after January 1983 and the rate was increased after April. During the entire rotation season, 40 kg of urea, 1450 kg of night soil, 600 kg of vegetable cake, 3508 kg of azolla, and 1830 kg of green grass were applied. Because 5500 kg of rice straw were left in the field, the total amount of fertilizer and feed was 12 928 kg.

On 26-27 June 1983, 1689 kg of fish (average 1300 kg/ha) were harvested. Excluding the fingerlings, the net catch was 1095 kg/ha. The net income from fish alone was CNY 2519, or an additional CNY 1957/ha.

Raising Fish in Winter Ricefields

This rotation method makes full use of the ricefields after the late rice harvest until the middle rice is planted the following summer or the next late rice crop is planted. In some areas, fingerlings are released right after late rice is transplanted, and the fish are harvested either before the spring festival in January or February or before the next early rice crop is transplanted. This method yields a high output of fish, mostly as food. During the winter season, most ricefields store water that is overgrown with plankton and bottom organisms, especially in East Sichuan Province. This water is very suitable for fish.

In the winter of 1983, Cheng Jinghong of the Freshwater Products Institute of Fujian Province reared fish in three pieces of land covering 0.25 ha at the Andou Fry Farm, Jinjiang County, Fujian Province. On 20 November 1983, he released 57.5 kg of fingerlings (C. idella, C. carpio, H. molitrix, and A. nobilis). On 28 March 1984, 128 days later, he collected 85 kg of fish, a net increase of 27.5 kg. The weight of C. carpio increased 5-8 fold (average 0.2 g/day), and the survival rate was 89.3%. After the costs for fingerlings and feed were deducted, the net profit was CNY 92 (CNY 5520/ha).

Cheng Jinghong raised the field bank by 50 cm and packed it firm after harvesting the late rice. He dug a ditch 30-cm wide and 30-cm deep along the field banks (1 m from the banks) and dug two fish ditches or pits (each covering 1 m²) near the water inlet. He installed screens at the inlet and outlet of the field, stored water in the field, and released the fingerlings. Fish feed consisted of peanut cake, rice husk, wheat bran, and fish powder mixed in a ratio of 8:6:5:1 with water. The mixture was spread in the ditches or put on a food platform at 14:00-15:00 each day. The total amount of feed used was 2-3% of the total weight of the fingerlings, depending on the weather and how well the fish fed.

Nie Dashu and Wang Jianguo are with the Institute of Hydrobiology, Academia Sinica, Wuhan, Hubei Province.

New Techniques for Raising Fish in Flooded Ricefields

Wan Banghuai and Zhang Qianlong

The traditional model of raising fish in ricefields has been practiced for a long time. It does not include digging ditches or pits; therefore, the field remains flat. With this model, fish raising and rice growing do not interfere with each other. This method proved that fish and rice could coexist and elucidated the relationship between fish and rice. It inspired the theory of rice-fish mutualism in ricefields. This traditional model suited cultivation systems in which high-stalk strains of rice were planted in ricefields with relatively high water levels. The fields were not directly exposed to the sun, weeding was not performed, and chemical fertilizers and pesticides were not applied. The model helped improve skills in fish raising in ricefields and promoted reforms in the cultivation system. The fact that fish raising in flat ricefields has been in practice for over 2000 years clearly demonstrates the viability of this technology.

However, the traditional model had disadvantages. The technology was not fully developed, management was poor, production was on a small-scale and done spontaneously by farmers, and the method was limited to hilly, mountainous areas and areas with sufficient water. The fish that were raised were usually a single species, small in size and in quantity. Yield was low (75-150 kg/ha) because of extensive raising and poor management. Moreover, production was subject to the constraints of the cultivation system and natural disasters. Therefore, development was slow and unstable. The history of fish-raising in flat ricefields in Jiangxi Province illustrates this slow development. The amount of land devoted to rice-fish culture was 3300 ha in 1956, 2800 ha in 1957, 3470 ha in 1958, 14 620 ha in 1959, 20 640 ha in 1960, 10 570 ha in 1961, and 12 520 ha in 1962. There was a steady decline until 1976, when fish raising in ricefields became virtually extinct. The area gradually increased again and by 1983 reached 18 700 ha.

Emergence of the New Technology

Fish culture in flat ricefields developed to its height in the 1950s and 1960s in China when it covered a total area of 0.7 million ha. By 1986, the area had grown to 1 million ha; however, the following year it declined to 0.73 million ha. Fish yield was less than 150 kg/ha. The cultivation system has changed, but fish raising in ricefields has not completely returned to its peak level, and even in areas where rice-fish production has been restored or developed it is still normally based on flat ricefields.

However, in some areas, fish raising in ricefields has been more highly developed. For example, in 1984-1986 demonstration areas (66 000 ha of ricefields) were cooperatively developed in 18 provinces, municipalities and autonomous regions all over the country. Fish yields reached 300-750 kg/ha (maximum 1500 kg/ha), and rice output increased by about 10%.

New models were developed in the 1960s. These reforms in the cultivation system led to increased rice output, but also sharpened the conflicts in fish raising in ricefields. Because the old model was no longer suitable, efforts were made to find solutions to these conflicts. During 1959-60, people in Jihe Village, Yaoxia Township, Suichuan County, Jiangxi Province, raised fish in double-cropping ricefields by digging ditches and pits (more or less like the current ditch-pit models). Fish yields reached 375 kg/ha. However, this technique remained on a small scale of only 0.67 ha and yields remained constant for over 20 years. It was not until the beginning of the 1980s that progress was made.

In 1982, scientists in Jiangxi Province began to systematically spread the new technology over the entire province and to promote the theory of the supportive coexistence of rice and fish. Extensive research was conducted on several rice-fish cultivation models: flat ricefield, ditch-pit, wide-ditch, zigzag ditch, small pit, field pond, semidry ridge and ditch, and ridge-ditch.

In 1984 and 1986, cooperative extension efforts were undertaken in 18 provinces, municipalities, and autonomous regions. In Jiangxi Province, the Provincial Aquatic Products Department sponsored and entrusted the cities and prefectures of Yichun, Wanzhai, Shangyou, Suichuan, Puyang, Fengxin, Shanggao, and Anyi to conduct technical extension and demonstrations of ditch-pit, ditch-pond, and ridge-ditch models. In Jiangxi Province, this marked the beginning of the expansion of the new technology.

Advantages and Disadvantages

The new technology is based on the theory that raising fish in ricefields and growing rice with fish improves the harvest of both crops. Rice-fish culture improves the ecological cycling of material and energy in the ricefields and therefore enhances the growth of rice and fish.

Fish raising in ricefields can be classified into various categories based on cultivation system: simultaneous cropping of rice and fish, rotation cropping of rice and fish, and intercropping of rice and fish. It can also be classified according to crop system and raising method into fish raising in double-cropped ricefields, fallow winter fields, lotus fields, and wild-rice fields. From the point of view of engineering, technology, and technique, rice-fish culture can be divided into flat-field, ditch-pit, and ditch-pond (inclusive of small pond, wide and semidry ridge, and ditch-in-the-middle sided by rice). All these models are now in use.

Ditch-pit model. In Jiangxi Province, this is the primary model for the new technology. Fish are raised in ricefields in open ditches and pits. Each ditch is as wide as two rows of rice and is 24 cm deep (or as deep as the hard layer of soil). The pits are 50-70 cm deep and 1 m² in area. One or two pits are placed at each inlet and outlet for water at the corner(s) of the field. Rice seedlings are planted along the sides of each ditch and along three sides of each pit to serve as a fence.

This model does not require much work or a large investment, it is easy to operate, and can be used in most fields. The conflicts between rice and fish are mitigated, rice output can be increased by 10%, and fish yields can be doubled or tripled compared with flat ricefields. Farmers readily adopt this model; however, because the ditches are shallow and the pits small, the fish are grown for a limited time and are small. These fish can be used as fish fry or as food by the farmers, but cannot be sold in the market. The yield is low.

Ditch-ridge model. A semidry rice planted on raised ridges is combined with fish culture in the ditches. At the same time, lotus, wild rice, and azolla are grown in the ditches. One advantage of this model is that there are many ditches filled with water where azolla can be grown as fish feed. This solves the conflict between rice and fish. The fish can be raised for a longer time and be given supplemental feed. Fish yields easily reach 750 kg/ha. This method improves the yield from ricefields and creates economic benefits.

The disadvantage of this model is its limited adaptability, particularly in Jiangxi Province, which is dominated by double-cropped rice. As well, the method requires a lot more work that must be repeated each year. Therefore, this model is not well accepted by farmers, and extension efforts have only been successful in establishing this model in 0.5% of the areas involved in rice-fish farming in the province.

Ditch and pond model. This is the most widely used model in the province because it can be practiced in different ways to suit local field conditions. A small pond is dug at one end of the field, or shallow pond(s) between the ricefields can be used. The ponds are 1-m deep and occupy only 6-8% of the total field area. The ditches are 30-50 cm deep and cover about one-third of the total pond area.

This model provides an optimized environment by using improved engineering and by extending and controlling the time available to raise the fish. Different varieties of commercial-size fish can be grown yields are 750-1500 kg/ha. Rice production is also increased by over 5%. This is an ideal model for raising fish in double-cropped ricefields. The disadvantages of this model are that a lot of labour is involved, the engineering work must be done each year, and the fish cannot be grown during the winter.

Principles and Economic Benefits

The new technology can create an artificial ecosystem that is similar to the natural ecosystem. The new models attempt to develop cash crops and livestock that are suitable for socioeconomic development. These models incorporate the advantages of pond raising. They produce high yields and fully use the ecological conditions of the ricefield. The fish have sufficient natural feed and this can be supplemented with artificial food. Therefore, conflicts between rice and fish are resolved and a balanced ecosystem is developed in which it is possible to increase the output of both rice and fish.

Empty spaces and small patches of land can be used by putting up shelters that not only provide refuge, but also provide suitable conditions for crops such as melons. Vegetables and beans can be planted on the ridges, and lotus, wild rice, taro, and azolla can be raised in the water. This comprehensive use of the land makes it possible to obtain good harvests of rice, fish, and vegetables. Because the model makes good use of land, water, biological, and nonbiological resources in the ricefields, it is an ideal model of production for fish-rice culture. Substantial ecological, economic, and social benefits are produced.

Biological Benefits

Weeding and fertilizer. Grasses that grow with rice absorb available fertilizer and compete for nitrogen. When 500-600 fish that feed on grass are raised, the grass is kept under control and the fertilizer is reserved for the exclusive use of the rice. Experiments have shown that every 1000 g of grass-eating fish fry can eat as much as 40-60 g of weeds/day. Experiments in 1988 in the Shang-gao area showed that a 500-g grass carp (Ctenopharyngodon idella) can eat 3.5 g/day of weeds in a ricefield and therefore turn weeds into fish protein. Fish excrement, in turn, fertilizes the rice, increases organic matter in the soil, and improves the fertility of the ricefield.

Pests and plant diseases. The insects (and their eggs) that harm rice are good food for the fish. When insects move about in the water, or when their eggs fall into the water, they are eaten by the fish. In this way, fish raising benefits the rice because it reduces plant diseases and eliminates pests. In experiments in the Shang-Gao area, a 150-g common carp (Cyprinus carpio) was found to eat 1.3 g of insect pests each day. One study compared the index of harm caused by pests (e.g., leaf folders and stem borers) and found that the leaf folding rate of every 100 clusters of rice was 90 cases in ricefields with fish and 210 in ricefields without fish. The rate of blight was zero in ricefields with fish and 0.014% in ricefields without fish. Fish also help rid rice plants of surplus tillers and disease-ridden leaves. This increases the penetration of light and reduces the occurrence of plant diseases and pests.

In addition, the movements of fish stir the water and soil, which increases oxygen in the water, speeds the release of fertilizer in the soil, and improves the development of the root system of the rice. Loosening the soil, eliminating weeds, and adding fertilizer raises rice production in many ways. The need to weed by hand is eliminated and reduced amounts of chemical fertilizer and insecticide are needed. These benefits produce savings in labour and investments and help improve the rural environment and the health of the people.

Economic Benefits

Rice-fish culture using the new technology makes multiple uses of available land. It is an intensive farming model that requires a small investment, is highly efficient, and produces quick benefits. The results of the new technology are illustrated by demonstration work organized by the Jiangxi Provincial Aquatic Products Department from 1984 to 1986. The lowest ratio of investment to added value was 1:4.8; the highest was 1:11.8 (Table 1). An investment of CNY 36 produced an increased value of CNY 175. At the high end, an investment of CNY 18 produced an increased value of CNY 214. These figures do not include the extra income generated from melons, vegetables, and beans.

Social Benefits

Increased employment. Surplus labour is employed. This is particularly true in poor mountainous areas where only a limited amount of arable land is available and there are few opportunities for sideline occupations. In Suixiang Township, Yichun Prefecture, for example, although many people were engaged in sideline occupations in the past, many others often visited relatives or friends because they had a lot of leisure time. Now, in over 95% of the nearly 46.7 ha of ricefields, fish are raised. The people are happy: fish raising in the ricefields gives people things to do and that they live contentedly.

More fish to market. Fish are a well-liked source of high-quality animal protein. The average per capita consumption of fish in China is 10 kg/year; in Jiangxi, it is only 6-7 kg/year. It is difficult to supply fish to hilly and mountainous areas. However, the extensive development of rice-fish cultivation and efforts to increase fish catch will help activate the urban and rural markets and supply more fish for the table.

Increase of income. Fish raising and the associated intensive farming are attractive to farmers because of the increased yields of rice and fish and the income derived from other cash crops. The extension of the new technology will promote reforms in the cultivation system, further enrich rice-fish farming, and further develop agriculture and increase national income.

Conclusion

Research and extension using the new technology are similar in Jiangxi Province to other places in China. These models can be further improved if they are adapted to local conditions and if rice output is increased. The new technology will be extended through the process of application, and perfected through extension.

For thousands of years, fish raising in ricefields in China has been more advanced than in other countries. However, considering China’s favourable natural conditions, the wisdom of the people, and the fact that China is basically an agricultural nation, development has been slow and unbalanced, and the amount of land devoted to rice-fish farming is limited. Moreover, the catch is still low (about 150 kg/ha on a national scale over a million ha). The new technology and its different models are still in the primary stages of development.

Administrative and professional leadership is needed. An agricultural extension program must be established to encourage fish raising in ricefields. This is the only way that the technology can be introduced to transform flat-type agriculture into three-dimensional agriculture and to produce increased yields and generate more income.

Wan Banghuai is with the Aquatic Products Department, Jiangxi Provincial Bureau of Agriculture Animal Husbandry and Fishery, Nanchang, Jiangxi Province; Zhang Qianlong is with the Animal Husbandry and Fishery Department, Yichun Prefecture, Yichun, Jiangxi Province.

Methods of Rice-Fish Culture and their Ecological Efficiency

Wu Langhu

Rice-fish culture is the best model of an artificial ecological system. Rice is predominant, but weeds, plankton, and saprophytic and photosynthetic bacteria compete with the rice plants for nutrients and diminish the environment. The introduction of fish into the ricefield creates a new link in the food chain that uses energy that would otherwise be lost and improves the function of the system and its economic efficiency. Appropriate techniques make bumper harvests of rice and fish possible.

Models of Rice-Fish Culture

Intercropping of Rice and Fish

This model is suitable for most types of ricefields. It requires ample water resources. Before the fish are introduced, ridges (50-70 cm high) and fish ditches are constructed. The number of ditches is determined by the size of the field. Each ditch is 33-50 cm wide and 25-30 cm deep. A pit (100 cm long, 50-70 cm wide, and 80-100 cm deep) is dug where the ditches cross. The distance between the ridges and the ditches is 40-60 cm. Rice plants on the ditches and the pit are transplanted near the ridge to form a fence or marginal row. All water inlets and outlets are fitted with screens to prevent the fish from escaping. About 19 500-225 000 summer fingerlings of summer grass carp (Ctenopharyngodon idella) are introduced per hectare.

Rotation of Rice and Fish

This model is suitable for cold-water fields. Its pattern is one season rice, one season fish. Ridges (100 cm high and 50 cm wide) are built and then 1000-1500 summer fingerlings of C. idella are introduced per hectare. When the rice matures, the grain is harvested but the straw is left in the water. Another 5000-7000 fingerlings of silver carp (Hypophthalmichthys molitrix) and variegated carp (C. carpio) are then introduced per hectare.

Rice-Fish with Ridge

This model is suitable for marshy lowlands or deep, cold, water-logged land. Rice is planted on ridges and fish are reared in ditches. The ricefield is reconstructed in two steps. First, ditches (50 cm deep and 20 cm wide) are constructed to cover 45-50% of the field, and ridges (20-130 cm wide) are built with soil from the ditches. Second, the ridges are levelled with mud from the ditches. Rice seedlings are transplanted close together on the ridge surface. Close planting compensates for the space taken up by the ditches. About 1500-3000 winter fish fingerlings and 12 000-18 000 variegated carp fingerlings are introduced per hectare.

Rice-Fish with Fish Pit

The fish-pit model is devised according to the principles of fish-pond culture. The technique provides 7500 kg of rice and 750 kg of fish (the Chinese slogan was “thousand jin rice grains, hundred jin fish”) and solves the problems of drying and of applying fertilizer and insecticide. A pit is dug on one side of the field and covers about 8-10% of the area of the field. The pit should be 2-3 m wide and 1.5-2 m deep. Its length depends on the length of the field. The pit can be used to hatch carp fingerlings, breed summer or winter fingerlings, and raise adult fish. About 3000-5000 summer or 300-500 winter fingerlings are introduced per hectare.

Rice-Fish with Wide Ditches

In addition to the ridges and the ditches used in the mutualism model, this model requires that a wide ditch (1-2 m deep and 1-2 m wide) be dug along the side of the water entrance of the field. The wide ditch usually takes 7-10% of the area of the ricefield and has an inner ridge (26 cm high and 23 cm wide). Between the wide ditches, at intervals of 3-5 m, are passageways. About 4500-7500 winter fingerlings are introduced per hectare. The wide ditches can also be used to hatch fingerlings.

Economic and Ecological Efficiency of Rice-Fish Culture

Economic Indicators

An experiment was conducted between 12 May and 15 July 1983 in three neighbouring ricefields (0.15 ha, 0.07 ha, and 0.01 ha). The ricefields produced good harvests of early maturing rice under both drought and water-logged conditions. The 0.01-ha field was used as the control. By the end of the experimental period, fingerlings in the control ricefield reached an average length of 8 cm, and 936 C. idella were harvested. The rate of recovery was 50%. Rice yields in the two experimental fields were 7584 and 7992 kg, an increase of 1209 kg/ha (19%) and 1617 kg/ha (25%) over the control (Table 1).

In October 1984, similar results were obtained in another experiment with late maturing rice. Four plots (each 141 m²) were established in a 0.08-ha field. Three plots were used as replicates, one as the control. The rate of increase in rice yields was 10.2-20% (Table 2). Tables 1 and 2 show that rice-fish culture can increase effective tillering, improve the grain fertility rate of the rice, and increase (or at least maintain) the fertility of the field.

Effect on Weeds

On 29 July 1982, the amount of weeds in two fields were measured. In the field with 100.8 kg of weeds, fish were introduced. In the field with 44.2 kg of weeds, no fish were introduced. On 13 October, there were 20.2 kg of weeds with the fish and 273.0 kg without the fish. Similar results were obtained in 1984-1985. In experimental plots in October 1984, there were 2.0 kg (148.5 kg/ha) of weeds in the field with fish, and 29.8 kg (2100 kg/ha) in the field without fish. The field with fish had 1951.5 kg/ha less weeds. On 2 May 1985, before fingerlings were introduced, the amount of weeds averaged 1504 kg/ha. On 27 July, after fish had been raised for 2 months, there were 17.1 kg of weeds in the treatment fields and 263.1 kg in the control fields. About 246 kg of weeds had been eaten by the fish.

Increase in Soil Porosity

Porosity directly influences the ability of the soil to retain water, the aeration of the soil, and the movement of water. It also indirectly affects the activity of aerobic and anaerobic bacteria and therefore influences the decomposition rate of organic substances and the capacity of the soil to provide nutrients. The activity of the fish can lessen the unit weight of the soil, increase its porosity, and therefore improve ventilation.

This increase in soil porosity has been experimentally verified. Experiments in October 1982 showed that soil porosity of a field with fish was 59%; whereas, porosity in a field without fish was 53%. In other experiments in 1984, soil samples were taken twice. The first samples showed no difference in soil porosity. The second samples, taken after 2 months of rice-fish culture, showed that soil porosity in the field with fish was 8.1-14.3% higher than that in the field without fish (Table 3).

Fertilization of Fields

Rice plants obtain two thirds of their nutrients from the soil and one third from fertilizer applied during growth. However, weeds, plankton, and saprophytic and photosynthetic bacteria compete with rice for nutrients. Weeds can reduce rice yields by 10-30%.

Rice-fish culture can eliminate or inhibit weeds and help retain soil fertility. Only about 30% of the weeds and plankton eaten by fish are digested and absorbed, about 70% are excreted, which increases the organic matter content and fertility of the soil. This was verified by testing the organic and alkaline-nitrogen content of the soil in ricefields with and without fish (Table 4). The ricefield with fish showed increases of 0.114% in organic nitrogen, 6.95 ppm in alkaline nitrogen, and 0.0044% in total nitrogen, compared with ricefields without fish.

Insect Control

Fish eat harmful insects and their larva, especially planthoppers. In an experiment in September 1982, three fish from a ricefield were dissected. Leafhoppers and planthoppers were found in the bellies of common carp and silver carp. In 1984, the population density of snout-moth larva and leafhopper were investigated. Fish were dissected to determine their capacity to eat insects. In the ricefield without fish, there were 820 snout moths and 11 275 leafhoppers; in the ricefield with fish, there were only 692 snout moths and 10 127 leafhoppers. When the fish were dissected 3 of the 10 C. idella had snout moths in their digestive organs, and all 10 contained 2-3 leafhoppers. More leafhoppers were found in C. idella that were more than 7 cm long.

In 1985, a detailed investigation was done on planthoppers (Table 5). The number of planthoppers per 100 clumps of rice in ricefields with fish was 17-28% less than that in fields without fish. By the fourth generation of planthoppers, there were 820 per 100 clumps of rice in the ricefield without fish, which was enough to warrant the use of an insecticide. The ricefield with fish had only 590 planthoppers per 100 clumps of rice and no insecticide was needed. Rice-fish culture can also help control mosquitoes, which significantly improves the health of people in rural areas.

Wu Langhu is with the Hubei Aquatic Science Research Institute, Wuhan, Hubei Province.

Ridge-Cultured Rice Integrated with Fish Farming in Trenches, Anhui Province

Yan Dejuan, Jiang Ping, Zhu Wenliang, Zhang Chuanlu, and Wang Yingduo

The earliest record of rice-fish farming in Anhui is from the Ming Dynasty. Traditional techniques produce low yields; however, improved methods have been introduced in recent years in many parts of the country. Changes include mixed culture of fish instead of monoculture; release of adult fish instead of fry; provision of feed for the fish; and disease prevention and control. Research on biology and ecology have attempted to find ways to increase the yields of rice and fish. Engineering facilities for rice-fish farming have been continuously modified. Some new ecological techniques include digging trenches that surround the field, horizontal trenches, and ditches, semidry cultivation, and ridge-cultured rice integrated with fish farming in trenches (high ridge and deep trench).

Advanced Cultivation Techniques

In 1987, the Provincial Aquatic Product Technique Extension Station in Fengtai County, Anhui Province, conducted experiments and established a project to demonstrate advanced techniques of ridge-cultured rice integrated with fish farming in trenches that obtained bumper harvests of rice and fish.

Selection of Experimental Plot

A 7.1-ha ricefield managed by 31 households at Xinji Village, Chengbei Township, Fengtai County, was selected. The plot was smooth and had medium soil fertility and a convenient irrigation and drainage system. The cropping pattern was wheat-rice. Field preparation began in early June after the wheat harvest. Trenches were dug and ridges constructed. Ditches (1-1.7 m deep and 1.5-2 m wide) were dug in 3-7% of the ricefield. Three types of ditches were used:

· Ridge type. Ridges 50 cm wide with two rows of rice seedlings spaced 25 x 12 cm; 165 000-168 000 plants/ha with each hole containing 1-2 transplanted rice plants; trenches 50 cm wide and 40 cm deep.

· Wide ridge type. Ridges 1 m wide with six rows of transplanted rice seedlings, rows and plants spaced 20 x 20 cm; 180 000-186 000 plants/ha; trenches 50 cm wide and 40 cm deep.

· Bed type. Bed 2 m wide with 12 rows in the bed, rows and plants spaced 20 x 22 cm; 195 000-201 000 plants/ha; trenches 60 cm wide and 40 cm deep.

Plot Management

Before the rice seedlings were transplanted, 600-900 kg/ha ammonium sulphate and 450-600 kg/ha calcium superphosphate and manure were applied. From 15 to 20 June, seedlings of hybrid rice varieties (Xianyou 3 and Xianyou 6) were transplanted. A week later, fish fingerlings were released. For adult fish culture, 2700-3000 fingerlings (10-20 cm in length) were released. The major varieties for fish culture were grass carp (Ctenopharyngodon idella) and common carp (Cyprinus carpio) mixed with a few variegated carp (C. carpio), silver carp (Hypophthalmichthys molitrix), and crucian carp (Carassius carassius). For fingerling production, about 225 000-300 000/ha C. idella and C. carpio fingerlings 3-5 cm long were released into the ricefield.

To accommodate the water requirements of both the rice and fish, the depth of the irrigated water was controlled according to the needs of the different growing stages of the rice. A week after the rice seedlings were transplanted, the ridge was flooded to a depth of 3-6 cm. After the seedlings turned green, the field was irrigated frequently. The trench was kept full of water to saturate the ridge and promote the growth of the plant-root system. The time from booting to milking is the peak period when both rice and fish need water; therefore, the ridge was flooded to a depth of 7-10 cm with irrigation water. After the rice had reached the milking stage, the trench was filled with water to continuously saturate the field and ensure full development of the rice grains. During the growth period for the rice plants, 7-10 kg of urea applied as one or two top dressings were used according to the soil fertility of each plot. Two or three applications (one or two times less than that applied in ricefields without fish) of pesticide were used depending on the rate of insect-pest infestation.

In early September 1987, the fields surrounding the experimental plot at Chengbei Township were attacked by rice planthoppers. The farmers shook the rice plants with sticks to knock the planthoppers into the water. This method of controlling insect pests did not use chemicals that pollute the environment and provided fish with additional food.

Fish management included control and prevention of disease, prevention of fish escape, the timed supply of feed in the ricefield, and the culture of adult C. idella mixed with secondary fish varieties. The feed consisted of fodder grass (10 500-12 000 kg/ha) and concentrated feed (450 kg/ha). In ricefields in which grass carp were the main fish variety, concentrated feed and green feed were used; in ricefields with fingerlings 450 kg/ha of concentrated feed and some green feed were applied.

Rice and Fish Yields

Rice yields. On 10 September, before the rice was harvested, a sample was taken of ridge-cultured rice integrated with fish farming. There were 15.5 panicles/bunch, 175 kernels/panicle, and 19.2% empty grains. With the other types of rice-fish farming, there were 12 panicles/hill, 166.8 kernels/panicle, and 19.3% empty grains. An estimate of yield was made by selecting a representative plot from each experimental model. The yield of each experimental plot was independently calculated after the harvest (Table 1). The rice yield from the ridge-cultured rice integrated with fish farming was 1-5% higher than from conventional rice-fish farming (Table 1). The ridge-cultured rice-fish system produced 12-16% more rice than ricefields without fish.

Edge effect. Experiments on wide-ridge rice-fish farming in Huo Shan County in 1987 indicated that the number of panicles per bunch, grains per panicle, and weight per thousand grains were higher in outside crop rows than in the rows at in the centre (Table 2). However, the ridge and wide-ridge systems require excessive labour to dig the trenches. Field preparation and transplanting is done during the busy season; therefore, it is difficult to popularize this method.

Fish yield. Before the rice harvest, adult fish with commercial value were marketed;. the remaining fish were counted, weighed, and put into the trenches and fishpond. The fingerlings were 10-20 cm in length, the adult C. idella weighed 0.5-1.5 kg, and H. molitrix weighed about 0.5 kg. The total harvest of fish from the wide-ridge treatment was over 800 kg of adult fish and 312 kg of fingerlings per hectare (Table 3).

Economic Efficiency Analysis

Production value and cost accounting showed that income from the ridge and wide-ridge fish farming systems was much higher than from conventional rice-fish culture. Income from the ricefields with adult fish culture was also higher than the income from the field with fingerlings. Net income was 2-3 times greater than that from ricefields without fish culture (Table 4).

Discussion and Conclusion

Ridge-cultured rice integrated with fish farming in trenches has several advantages. It is suited to lowland ricefields, cold waterlogged fields, and level ricefields. The optimum sizes for the ridges and trenches are being studied in different parts of the country. The economic efficiency of the ridge-based system is higher than for conventional rice-fish farming. The rice plants grow vigorously and have many large panicles and full grains. The wide-ridge system requires less labour than the ridge system and is therefore easier to popularize.

The wide-ridge system and especially the ridge system are more economical and efficient than conventional rice-fish culture because:

· Frequent irrigation with shallow water is the most appropriate environment for rice growth and development. The model of ridge-cultured rice integrated with fish farming in trenches is suited to irrigation with shallow water.

· Hybrid rice, the high-yielding varieties, need wider spaces between rows and narrow spaces between plants. Ridge-cultured rice integrated with fish farming in trenches fulfils these requirements and provides suitable growing conditions for high-yielding varieties.

· The ridge and wide-ridge systems can alleviate the conflicts between the water requirements of rice and fish. These systems meet the need of rice for water depth at different growing stages, provide a good environment for fish, and enlarge the holding capacity for fish. They also make full use of edge effects for the rice by improving ventilation and light penetration, which enhance photosynthesis, reduce diseases and pests of rice, and deepen the symbiosis of rice and fish to increase the yields of both crops.

Based on previous experiments, experiments have been initiated using zero tillage in high ridges, deep trenches, and wheat-rice cropping patterns. This system could reduce the need for field preparation and trench digging, improve conditions of water, fertility, atmosphere, and heat, and prevent damage to the soil structure. In addition, mechanical diggers must be designed to replace manual labour. If successful, this system could play an important part in improving economic efficiency.

Yan Dejuan and Jiang Ping are with the Anhui Aquatic Product Technique Extension Station, Hefei,; Zhu Wenliang is with the Fengtai Animal Husbandry and Anhui Bureau of Aquatic Products, Fengtai and Zhang Chuanlu and Wang Yingduo are with the Chengbei Township, Gengtai County, Anhui Province.

Development of Rice-Fish Culture with Fish Pits

Feng Kaimao

The model for rice-fish culture with fish pits was developed as an improvement to traditional rice-fish culture. It has now become the main type of rice-fish culture and, in some regions, the major way for farmers to increase their incomes.

China has a long history of rice-fish culture. The traditional method of rice-fish farming in flat fields faces many conflicts between rice and fish and is easily upset by changes in natural conditions. Farmers often have to sacrifice the fish to save the rice, which has diminished the role of rice-fish farming and hindered its development. Before 1980, rice-fish farming in Dazu County yielded only 22.5-52.5 kg of fish per hectare. In 1981, rice-fish farming in flat fields advanced to some extent, but fish yields were still very low (Table 1).

Before the 1980s, aquaculture scientists in China experimented with fish troughs combined with fish trenches, which had been adopted in ricefields in the southern part of Jiangsu Province. However, during the midsummer droughts in Dazu County, the shallow troughs and small trenches did not provide sufficient water for the rice and fish. As well, the fish were not able to adapt to the high temperatures experienced during the drought.

Researchers in Dazu County studied the factors that influenced fish growth, e.g., the relationship between temperature and depth of the different water layers, the upper and the lower limits of temperature that suit fish growth, the relationship between the appropriate temperature range and the environment of the ricefield, the quantities of dissolved oxygen produced and consumed during the day and at night, and the form of oxygen molecules moving in the water. Based on their research, they developed a new approach to solve the conflicts between fish and rice. They dug 1-m deep fish pits that covered about 6-8% of the ricefield and connected the pits to trenches. Field experiments were conducted in many areas between 1980 and 1983. This new model of rice-fish farming was verified and accepted by farmers and the county government.

Development of the Method

At the beginning of the trials in 1981, the area for rice-fish farming with fish pits was 0.2 ha. In 1982, the area reached 1.1 ha. Multiple-plot trials totalled 14 ha in 1983. By October 1984, the area had expanded to 3080 ha, and by the end of 1985, 3570 ha had been developed. Farmers had discovered that the economic benefits from the new model were 3-8 times greater than from flat field rice-fish culture. Fish yields as high as 3195 kg/ha were obtained. From 21 July to 31 August 1985, Dazu County experienced a midsummer drought and 5800 ha of flat-field rice-fish farming (58.8% of the total area devoted to rice-fish farming) were damaged. However, good harvests of both rice and fish were obtained from rice-fish farming with fish pits. This convinced the farmers of the value of using the new model and the technique became popular. By the end of August 1986, the total area of rice-fish culture with fish pits reached 41 474 ha. It is expected that new developments will move toward combining fish pits with shallow ponds.

Results

Aquaculture production in Dazu County from all types of water areas has increased (Table 2). Rice-fish farming has developed most quickly. Rice-fish farming includes rice-fish farming with fish-pits and flat field rice-fish farming. In the past few years, production changes from both types of farming have been remarkable (Table 3). The fish-pit method shows greater resistance to natural disasters compared with the flat-field style. Although adult fish yields from the fish-pit method increased remarkably in 1984-1985, the unit yield of fish from flat fields decreased, apparently because of the midsummer drought. Since then, the area devoted to the fish-pit method has increased each year. Fish production from fish-pits varies depending on the progress of experiments, demonstrations, and extension efforts. Actual production levels in 1985 are presented in Table 4.

Although the environment in the fish-pit method is superior to the environment in the flat-field style, production of adult fish was directly influenced by management level (Table 5). Large-scale experiments, demonstrations, and extension were carried out in 1984-1985. Yields decreased as the level of researcher involvement decreased. Yields were highest in experimental areas and lowest in the extension areas.

These large-scale areas where used to obtain economic data. The ratios of input and output in 1984 were 1:2.86 for the flat fields, 1:3.59 for the fish-pit method, 1:3.35 for demonstrations, and 1:3.43 for extension. In 1985, the ratios were 1:2.61, 1:3.98, 1:3.16, and 1:3.23, respectively. Average profits from raising fish in flat ricefields was CNY 1833/ha, with 92.5% of the profit from rice and 7.5% from fish in 1984. In 1985, profit was CNY 1600/ha with 90.5% from rice and 5.5% from fish. Because of the midsummer drought in 1985, profits decreased by 2%. In ricefields with fish pits, profits (CNY 3204/ha with 58% from rice and 42% from fish) were highest from the experimental district. The income from ricefields with fish pits was 6.3-9.8 times greater than from the flat fields in 1984 and 17.7 times greater in 1985. Practice has proven the remarkable beneficial results of the new model, and additional developments of this model are expected.

Feng Kaimao is with the Agriculture, Animal Husbandry and Fish Bureau, Dazu County, Dazu, Sichuan Province.

Techniques Adopted in the Rice-Azolla-Fish System with Ridge Culture

Yang Guangli, Xiao Qingyuan, and He Tiecheng

The rice-azolla-fish system features an ecological three-dimensional approach to agriculture. Rice (the main crop), azolla, and fish are combined in a symbiotic complex. Rice is planted on the ridge, azolla grown on the water surface, and fish raised in the water. This new farming system was studied between 1984 and 1987.

Experiments were conducted on two plots of land, each with an area of 0.2 ha and a 6-m2 fish pit. Double-cropping rice was planted in one plot, single-cropping rice in the other. Each plot was randomly arranged with three replications and planted with local high-yielding varieties or hybrid combinations of rice and azolla (Azolla filiculoides or A. caroliniana) (6000 kg/ha). Fish species selected for mixed raising were grass carp (Ctenopharyngodon idella), tilapia (Oreochromis nilotica), lotus carp, dull carp, and Hunan crucian carp (Carassius auratus). Growth of rice, azolla, and fish was recorded. The nutrient content of azolla and fish dung was determined by standard methods of analysis.

Results and Discussion

Proportion of Ridges to Ditches

Ridge width had a bearing on the yields of rice, azolla, and fish (Table 1). The yield of double-cropped rice on the 106-cm wide ridge (13 834.5 kg/ha) was 4.1% higher than in the control (conventional planting), 5.1% higher than on the 53-cm ridge, and 2.5% higher than on the 80-cm ridge. The yield of azolla grown on the water surface between the 53-cm ridges (72 930 kg/ha), was 70.3% higher than in control, 13.4% higher than between the 80-cm ridges, and 55.6% higher than between the 106-cm ridges.

Yields of fresh fish increased as ridge width decreased. The yield of fresh fish with the 53-cm ridges (841.5 kg/ha) was 89.3% higher than the control, 12.2% higher than with the 80-cm ridges, and 22.2% higher than with the 106-cm ridges. Therefore, wide ridges favoured rice yields, and the narrow ridge favoured growth of azolla and fish.

When ridge width was uniform and ditch width was varied, rice yields differed. With a 40-cm ditch, rice yield was 10 462.5 kg/ha, the same as in conventional planting, 3.5% more than with 46-cm ditches, and 11.4% more than with 53-cm ditches. Yields of azolla and fish were directly related to increases in ditch width (Table 2).

Production practices have proven that to determine the proportion of ridges to ditches for to obtain high rice yields, it is necessary to consider soil fertility, the characteristics of the rice varieties, whether the proportions favour growth of azolla and fish, and whether rice, azolla, and fish are well coordinated. In a ricefield in which rice is planted on the ridges, azolla is grown on the water surface, and fish are raised in fish pits, the water area of the fish pit should be 5-10% of the total area of the ricefield.

Best results are obtained if the ridge width is 53-106 cm and the ditch width is 40 cm. On each ridge, 4-8 rows of rice are planted 13-16.5 cm apart to obtain 300 000-375 000 hills of seedlings per hectare. If hybrid rice varieties are used, plant spacing should be 16.5-20 cm with 3-7 rows on each ridge to obtain 225 000-300 000 hills of seedling per hectare.

Planting Rice During Different Seasons

When ordinary rice varieties were planted in the early crop season and hybrid rice varieties were planted in the late crop season, yearly rice yield was 12925.5 kg/ha or 1405.5 kg less than that obtained from two crops of hybrid rice (Table 3). However, there were no differences in the yields of azolla and fish between ordinary rice + hybrid rice and hybrid rice + hybrid rice. To increase rice yields, it is important to determine the proper time for planting hybrid rice varieties in the early and late crop seasons.

When early- or medium-maturing rice varieties (growth duration 100-110 days) are used as the early crop, the rice varieties used in the late crop season should be late-maturing (growth duration about 120 days). If late-maturing rice varieties (growth duration 115-120 days) are used as the early crop, the rice varieties used in the late crop season should be early or medium-maturing (growth duration 100-110 days). When two crops of hybrid rice are grown in a year, the hybrid combinations used in the early and late seasons should be all early or medium-maturing (growth duration about 110 days). If a single crop of medium rice is planted, ordinary rice varieties or hybrid combinations that are late maturing (growth duration 130-135 days) should be used. Conditions of location, cultivation practice, and soil fertility should also be considered.

Yields of Fresh Azolla

Annual yields of fresh azolla grown in ridged ricefields that use the rice-azolla-fish system can reach 113 877-132 235 kg/ha. Yields of different azolla species vary. The yield of A. filiculoides (132 235 kg/ha) was 15.5% higher than that of A. caroliniana. In a mixed culture of A. filiculoides and A. caroliniana, yield (124 417 kg/ha) was between the yield of pure cultures of the two species.

Yields of azolla also varied with different locations (Table 4). In winter (8 November-16 March), A. filiculoides propagated more rapidly in the hilly region of Hunan and the yield of fresh azolla reached 47 010 kg/ha. On average, yields doubled in 18.8 days. Yield of A. filiculoides were 59.7% greater than yield of A. caroliniana. Mixed cultures of these two azolla species, on average, doubled yields in 19.1 days.

During the spring propagation stage in Changsha, in the hilly region of Hunan, A. filiculoides also propagated faster than A. caroliniana, and the propagation rate of mixed cultures was between the rate for pure cultures. In the region along Dongting Lake in northern Hunan, the yield of A. filiculoides was higher than A. caroliniana, and the yield of the mixed culture was lowest. Differences in yield were related to the slow rise in temperature in the spring. In the mountainous region of southern Hunan, the propagation rate of different azolla species was the same as in middle Hunan because the temperature rose rapidly in early spring.

When azolla was grown between ridges, yields of fresh azolla in the mountainous region of southern Hunan was highest, followed by middle Hunan and northern Hunan. The rapid temperature rise between May-June in northern Hunan affected the propagation rate of azolla. In southern Hunan, the day temperature was high and the difference in temperature between day and night was great, which favoured growth of azolla.

When azolla was cultured on the ridge to over-summer, no matter where it was grown (in middle, northern, or southern Hunan), the propagation rate of A. caroliniana was highest, A. filiculoides was lowest, and the mixed culture was intermediate. These results reflect the fact that A. filiculoides is unable to tolerate high temperatures. Therefore, in winter or spring, it is better to grow A. filiculoides, which can tolerate low temperatures; and in summer or autumn, it is better to grow A. caroliniana, which can tolerate high temperatures. To compensate for deficiencies in each species, a mixed culture is recommended.

Yields of Fish

Effect on azolla. Four fish species (grass carp, tilapia, crucian carp, and lotus carp) were raised with azolla for 110-112 days. The fish species best suited to the rice-azolla-fish system are grass carp and tilapia (Table 5). Both like to eat azolla and adapt easily to the ricefield environment. The omnivorous crucian carp and lotus carp (which are benthic and planktivorous feeders) can be raised in the ricefield in lower numbers (Table 5). Grass carp and tilapia eat over 60% of their body weight in azolla each day; whereas, crucian carp and lotus carp eat about 8% of their body weight in azolla.

Of the four species, grass carp and tilapia had the highest growth rate. Their body weights increased 4.2 and 6.2 times, respectively. The body weights of crucian carp tended to decrease when they grew to a certain stage, and when lotus carp were fed only azolla, their body weight decreased. The digestibility of A. filiculoides (18.1%) was higher than A. caroliniana (12.9%) for grass carp and tilapia, respectively. Therefore, A. filiculoides is a better food for fish than A. caroliniana.

If it is assumed that fish yield was 750 kg/ha and that the amount of azolla eaten per gram of fish daily was 0.605 g/day, the amount of dung excreted by the fish would be 52.9% and it would contained 2.7% N. Therefore, when fish are raised in a ricefield for 100 days, they provide the ricefield with 85.5-103.5 kg N per hectare.

Effect of different fish species. When fingerlings bred in spring were raised, yields of rice and azolla were correlated negatively with the proportions of grass carp and tilapia, but yields of fish were correlated positively with the proportions of grass carp and tilapia. To fully use the azolla and the natural resources of the ricefield, the proportion of fish raised should be 60-70% grass-eating grass carp and tilapia and 30-40% omnivorous crucian carp and lotus carp (Table 6).

Effect of stocking density. When fingerlings bred in spring were raised, the yields of fish were correlated positively with stocking density. With 30 000 fingerlings/ha, yield of fish was 630 kg, which was 15% higher than the yield obtained from 15 000 fingerlings/ha and 5.6% higher than from 22 500 fingerlings/ha. Therefore, stocking density has an important bearing on fish yields.

Survival rate of fingerlings (e.g., grass carp) was correlated negatively with stocking density. At a density of 15 000 fingerlings/ha, survival rate was 3% higher than at 30 000 fingerlings/ha. Increases in body weight of fish followed a similar trend. When stocking density was 15 000 fingerlings/ha, body weight was 11.4 g heavier than at 22 500 fingerlings/ha and 20.6 g heavier than at 30 000 fingerlings/ha (Table 7).

Effects of pesticides. In general, pesticides are applied in the rice-azolla-fish system to control rice pests and diseases. The routine doses of pesticides such as methamidophos, dimethoat, dichlorphos, chlordimeform, trichlorfon, MIPC, and kasugamysin did not harm grass carp, tilapia, crucian carp, and lotus carp. Malathion and EBP were safe for crucian carp and dull carp, but had lethal effects on tilapia. Phenthoate was harmful to all the fish tested. The toxicity of various pesticides to fish was in the order: kasugamucin < methamidophos < trichlorfon < dimethoat < chlordimeform < tetra chlorvinphos < dichlorphos < malathion < phenthoate (Table 8).

Conclusion

In a rice-azolla-fish system that includes ridge culture, the ricefield should be constructed with: ridges 53-106 cm wide, ridge ditches 40 cm wide and 20-25 cm deep, a main ditch (50 cm wide and 50 cm deep) in the centre of the field, and deep ditch (50 cm wide and 50 cm deep) surrounding the ricefield. A fish pit (80-100 cm deep) should occupy 3-10% of the total area of the ricefield. Of the fingerlings raised in the pit, 3-5% are bred in spring and 8-10% are overwintering fingerlings. Soybean and melon can also be planted around the pit.

High-yielding rice varieties are used. Two crops of hybrid rice are planted with wide row spacing and narrow plant spacing to increase the border effect and enhance the use of light. In a ricefield of ordinary rice, 300 000-375 000 hills of seedlings per hectare (5-7 seedlings/hill) are planted; if hybrid rice is used, 225 000-300 000 hills of seedlings per hectare (1-2 seedlings/hill) are planted. More P and K fertilizers and less N fertilizer are applied in deep placement. Late rice seedlings can be transplanted without tillage. Attention must be paid to field management. The water level on the ridge surface should be regulated according to the growth stage of the rice.

Low-temperature tolerant A. filiculoides and high-temperature-tolerant A. caroliniana can be grown. In general, 300-500 kg of A. filiculoides are planted in the field in early or middle March, and 200-400 kg of A. caroliniana are planted 7 days after transplanting early or medium rice. Mixed culture of these two azolla species is possible.

Suitable fish species are grass carp, tilapia, crucian carp, and lotus carp. All four species like to eat azolla, grow rapidly, and adapt themselves to the ricefield environment. The fingerlings (grass carp and tilapia 60-70%; crucian carp and lotus carp 30-40%) are released into the ricefield in early May. Generally, 6000-12 000 overwintered fingerlings or 30 000-45 000 fingerlings bred in the spring are raised per hectare.

Routine dosages of common pesticides are safe to these four fish species. However, malathion and EBP are lethal to tilapia, and phenthoate is harmful to all of the fish.

Yang Guangli, Xiao Qingyuan, and He Tiecheng are with the Soil and Fertilizer Institute, Hunan Academy of Agricultural Sciences, Changao, Hunan Province.

Semisubmerged Cropping in Rice-Fish Culture in Jiangxi Province

Liu Kaishu, Zhang Ningzhen, Zeng Heng, Shi Guoan, and Wu Haixiang

Most of the 0.5 million ha of croplands in Jiangxi Province are pit fields, ridged fields, alluvial fields, and low grounds near the lakeside. These areas make up about 20% of the total ricefields in Jiangxi, but they are mostly cold, waterlogged, middle- or low-yield plots. The high water table is mainly responsible for the poor drainage of accumulated water. Constant water saturation has turned the soil to gley, which is cold, infertile, acidic, poisonous and lacks oxygen. Under these conditions, water, nutrients, air, and temperature are unfavourable to the growth and development of rice. Most regions yield one crop of rice a year, i.e., middle-late rice, but yield is low. To transform these low-yielding lands and increase crop production, a semiarid rice production initiated by Professor Hou Guangjiong was introduced in 1986 into the mountainous area of Gannan Prefecture, Jiangxi Province. This method of semisubmerged cropping in rice-fish culture has been improved to suit local conditions and has increased crop yields.

Main Principles

This method makes drastic changes to the system of rice cropping: planting on mounds instead of in furrows, putting ridges and ditches side by side to change the conditions in the field and add an active layer, planting rice on dikes, and culturing azolla and fish in the ditches. The physical changes raise the temperature of the soil and water, speed the catabolism of organic matter and the release of nutrients, and decrease the effect of toxic substances. As a result, seedlings revive sooner after transplanting, grow quickly, and have more white roots.

This method of rice-fish culture can turn single-crop agriculture into a double or multiple-harvest system and the slack winter season into a busy time. It is an excellent model of ecological agriculture that is applicable in all districts.

Continuous Nontillage

After the topsoil is ploughed for the first time, ridges and ditches are constructed side by side in the fields. Rice is planted on the ridges and fish are cultured in the ditches. Thereafter, the topsoil is not ploughed or harrowed to ensure that the active top layer is not destroyed. The soil does not become a caked mass; it grows softer with time. If the topsoil is ploughed, pockets of air and water capillaries in the soil are blocked. This decreases the percolation ratio, destroys the soil structure and the balance of water, nutrients, air, and temperature, and, as a result, reduces crop yields.

Continuous Ridge Tillage

Ridge tillage raises both the temperature of the water and soil and the oxidation-reduction potential of the soil. It activates soil nutrients and reduces toxic substances. This stabilizes the water, nutrients, air, and temperature and makes conditions more suitable for the growth and development of rice.

Continuous Infiltration

Capillary water in the soil is the only form of water that contains available nutrients and can flow freely, aerate the soil, and conduct heat. The key to semisubmerged cropping is improving the hydrological system of the soil. The continuous infiltration of capillary water aerates the soil, conveys nutrients, and prevents the soil from becoming a caked mass. The water level must be controlled according to the growth of the rice and the needs of the fish.

Demonstration and Application

No Tillage, Rice on the Ridge, Fish in the Ditch

Experiment were carried out for the first time in 1986 in Longhui Village, Nankang County, Ganzhou Prefecture, in cold, waterlogged mountain fields with an area of 0.3 ha. In 1987, the area was increased to 2 ha. Experiments were also carried out in lateritic low-yield plots at Luoding Village, Xingjiang County, and in a waterlogged lowland area near the lakeside at the Dongfeng Branch Farm of the Hongxing State Reclamation Farm. The total experimental area in these two areas was about 1.3 ha. In 1988, the method was popularized in over 1330 ha in several counties (Ruijing, Nankang, Shicheng, Xingfeng, and Shangyou). In Ruijing County alone, there were 667 ha. Rice and fish were equally emphasized. In a few of the experimental plots, azolla was also cultured. The method was extended by the Agriculture, Animal Husbandry and Fish Department to an area of 200 ha in the counties of Fuzhou Prefecture.

Rice on the Bed, Fish in the Ditch

In 1987, this method was demonstrated on 2.8 ha in Shangyon County, Ganzhou Prefecture. The method features a wide ridge (0.8-1.2 m) that is constructed after the topsoil is ploughed. Rice is planted 13-17 cm apart on beds in rows that are 20 cm apart; fish and azolla are cultured in the ditch.

Rice on the Bed, Fish in the Ditch

In 1987, this method was demonstrated on 2.8 ha in Shangyon County, Ganzhou Prefecture. The method features a wide ridge (0.8-1.2 m) that is constructed after the topsoil is ploughed. Rice is planted 13-17 cm apart on beds in rows that are 20 cm apart; fish and azolla are cultured in the ditch.

Benefit Analysis

Because management of agricultural production is presently carried out by individual households, the experiments, demonstrations, and applications were arranged at the household level. During the entire production period, technicians were sent to the areas to provide technical advice, conduct quality inspections, and observe and record results. The results are analyzed and compared in Tables 1-3.

Compared with conventional flat cropping, there were considerable increases in rice output and income from fish whether the rice-on-ridge or the rice-on-bed method was used. For example, 11 households used the rice-on-ridge method. Their average rice output increased by 2223 kg/ha (range 420-7140 kg/ha) and average net income from fish was CNY 2010/ha (range CNY 450-4935/ha). When the value of the fish and the increased amount of rice were both counted, the total rate of increase in value was 36.5-216.9% (average 86.3%). In another four households that used the rice-on-bed method, the increase in rice production was 680 kg/ha (range 530-990 kg/ha) and the net value of the fish was CNY 1540/ha (range CNY 531-3170/ha). The total net value of fish and rice increased by 76% (Tables 1 and 2).

The rice-on-ridge method is superior to the rice-on-bed method because it improves the ecological environment of the farmland (and enhances the growth of rice) and because it has a larger area of water, which is favourable for fish breeding. Higher economic benefits were obtained from the rice-on-ridge method when mixed species of fish, instead of a single species, were raised (Table 3).

Soil Improvement

Professor Hou Guangjiong has reported many improvements in the soil using the rice-on-ridge, fish-in-trench, no-tillage method of semiarid rice cultivation. Preliminary observations, suggest that unit weight of the soil decreases, temperature increases, and that the soil contains more organic matter, total nitrogen, available nitrogen, phosphorus, and potassium (in some cases, there was a tendency toward less total phosphorus compared with conventional flat cropping) (Tables 4-6).

Remaining Problems

In experiments and demonstrations, the semiarid rice-on-ridge, fish-in-trench method has remarkably increased production and income and improved soil conditions. In 1989, the Prefectural Department of Ganzhou planned to apply the new method on 33 330 ha. Farmers who had become aware of the benefits of the method were happy. But, to extensively disseminate any new technique, potential problems should be examined and addressed.

Farmers do not believe that rice can be grown without ploughing the field. For thousands of years, rice has been planted in water using the flat basin irrigation cropping method. Ricefields with ridges are new. Most farmers doubt that the rice plants can absorb water and nutrients when they are planted on the ridges. Some farmers also complain that the ridges make it difficult for them to put their threshing tubs and machines in the fields at harvest time. More demonstration and extension efforts are needed to overcome these problems.

Many places have no previous experience with raising fish in ricefields. There are also some social problems. Fish in the fields are often stolen, especially in cold, waterlogged fields that are usually located in remote mountain areas. Farmers worry about this. Local governments must strictly enforce the law to protect farm production from theft. The villagers could also develop some protective measures.

The farming activities in the new method (e.g., digging trenches, forming ridges, clearing mud from the trenches, and applying fertilizer) require much labour. In Longhui Township, Nankang County, a small iron spade that was light and handy for clearing mud from trenches was popularized. Such labour-saving tools or devices for trench digging, ridge forming, and row fertilization need to be developed.

When a new technique is applied in a large area, farmers, because of their different levels of understanding, sometimes fail to follow the technical requirements for certain farming activities. Because of this, not only should extension and guidance be stressed, but input supplies, such as chemical fertilizers and pesticides, must also be made available.

Liu Kaishu, Zhang Ningzhen, and Zeng Heng are with the Jiangxi Agricultural University, Nanchang; Shi Guoan is with the Agricultural Administration Bureau, Ruijin County; and Wu Haixiang is with the Land Administration Bureau, Shangyou County, Jiangxi Province.

Rice-Azolla-Fish Symbiosis

Wang Zaide, Wang Pu, and Jie Zengshun

Rearing fish and Azolla spp. in ricefields is an important component of traditional organic farming in China. Rice-azolla-fish symbiosis is a new development in ecological agriculture. The results of a 2-year field experiment indicate that rearing fish and azolla in ricefields increases the output of rice, azolla, and fish.

The highest output of rice was obtained from the rice-azolla-fish system (7096.2 kg/ha). This was an increase of 9.3% compared with the control. The rice-azolla system had the next highest yield, followed by rice-fish culture and the control. The coefficient of use of light energy was highest for the rice-azolla system, followed by rice-azolla-fish, rice-fish, and the control (Table 1).

Output of fish was highest in the rice-African catfish fry field (717.0 kg/ha), followed by food fish reared in the rice-azolla-fish field (536.3 kg/ha). The lowest fish output (265.4 kg/ha) was obtained from food fish (Table 2).

The cost of inputs of seed, fertilizer, labour, and fry were compared with the outputs of rice, straw, and fry. Total net income was highest for the rice-azolla-fish field (CNY 8814.5/ha); an increase of CNY 4521.8/ha over the control. The incomes from the other systems were: rice-fish (CNY 7968.3/ha), rice-azolla (CNY 4637.9/ha), and the control rice plot (CNY 4292.7/ha) (Table 3).

The rice-azolla-fish system had obvious economic benefits. In the rice-azolla-fish system, rice output and net income were higher than the control, and there were good ecological effects as well. Soil organic matter was highest in the rice-azolla-fish system (1.4), followed by rice-azolla and rice-fish (1.25) and the control (1.09). The pH was highest in the control (8.66) and lowest in the other fields (8.03-8.02). Nitrogen content was the highest in the rice-azolla field (0.081), followed by rice-azolla-fish, rice-fish, and the control. The rice-azolla-fish, rice-azolla, and rice-fish systems improved the physical and chemical properties of the soil (Table 4), reduced cost, diseases, pests, and weeds, and avoided environmental pollution.

Wang Zaide, Wang Pu, and Jie Zengshun are at the Beijing Agricultural University, Beijing.

Economic and Ecological Benefits of Rice-Fish Culture

Li Xieping, Wu Huaixun, and Zhang Yongtai

Research on the economic and ecological benefits of rice-fish culture was conducted in 1985-87. The objectives were to more fully exploit agricultural resources in rural areas and to improve field productivity. A 0.1-ha experimental field at the Li-Xia-He Regional Agricultural Institute was used. In total, there are more than half a million hectares in Li-Xia-He where the cropping system is wheat and rice. In general, the soil is a heavy clay that is low in elevation and has good water retention. Conditions for rice-fish are good.

Normal hybrid rice (Xian-you 63) and a hybrid (Xian-you 63P, developed by the Yangzhou Regional Agricultural Institute) were planted under single factorial designs for 3 years. The variable factors were density, planting pattern, fertilization strategy, and amount of nitrogen. Experiments were carried out simultaneously in 0.001-ha plots with three replicates, and in 0.04-ha test fields with two replicates. Two types of cultivation were used, one for fry, the other for mature fish. Each method of cultivation was subjected to a number of treatments to test the effects of different proportions of fish species, stocking size of fish, and density. A ricefield without fish was used as the control.

The main fish species used were Fu-shou fish (Oreochromis mossambicus x O. niloticus F1), grass carp (Ctenopharyngodon idella), and common carp (Cyprinus carpio) with a small proportion of bighead carp (Aristichthys nobilis) and white crucian carp (Carassius cuvieri).

After the fields were ploughed, fish trenches, 0.33-m wide and 0.40-m deep, were dug in an X pattern. At the cross points of the trenches, two fishponds (each 2.5-m long, 1-m wide, and 1-m deep) were constructed. The area of the trenches and ponds was 3.5% of the total area of the field. Fingerlings were stocked in the ricefield 10-17 days after the rice was planted and were fed mainly azolla and fresh grass and a small proportion of grain products. For 30 days after stocking, the fish were fed 4.5-6.0 kg/ha of wheat flour with small amounts of azolla each day. After 30 days, 7.5-15 kg of grain products and 75-150 g/ha of fresh grass were fed each day. The amount of feed was increased as the fish grew. The water in the field was normally kept at a depth of 3-10 cm, but the level was lowered for several days before differentiation of the rice panicles to dry the field slightly. The field was drained twice during the filling stage. The control plot was under regular water management.

The number of rice stems and tillers, the number of weeds, plant and soil nutrients, and the amount of sunlight penetrating to the bottom of the rice plants were determined. After harvesting, the ears and characteristics of the rice plants were measured. The number of fish harvested, their individual weights, and the yields of rice and fish were determined.

Economic Benefits

Yield of Rice

Yields of rice each year were 9054 kg/ha (1985), 7929 kg/ha (1986), and 7848 kg/ha (1987) in the control field and 8662, 7884, and 7997 kg/ha in the ricefield with fish. Rice yields in the field with fish were lower by 4.3% in 1985, but in 1986-87, the yields were almost the same. Because the area of fish trenches and ponds occupied 3.5% of the total area of the field, and the actual planting area had been reduced by 7.0%, it can be concluded that the yields of individual rice plants in the rice-fish field increased in 1986-87. This increase offset losses from the reduction in planting area; therefore, yields of rice with or without fish in the field were about the same.

Yield and Value of Fish

In 1985, the average yield of fish from two ricefields was 524 kg/ha valued at CNY 1730/ha. In 1986, despite predation by snakehead fish (Ophiocephalus argus), the average yield of four plots was 442 kg/ha valued at CNY 1750/ha. The highest yield was 615 kg. In 1987, there was a shortage of fingerlings and appropriate species, and a large number of fish escaped. The average yield of four plots was 208 kg/ha valued at CNY 989/ha. The highest yield was 300 kg. The average value of fish for the 3 years was CNY 1490/ha with a net income of CNY 991/ha.

Net Income

The average investment in rice-fish culture was CNY 1009/ha. This was CNY 406 more than the control and an increase of 67.3%. The reduced use of pesticide and chemical fertilizer in the rice-fish field lowered costs by 16-25% (1986-87). When the increase in yield and the reduced costs are considered, the net income for the rice-fish field each year was CNY 1090/ha (1985), CNY 1222/ha (1986), and CNY 796/ha (1987) more than in the control. The average was CNY 1036/ha, which is 40% of the total income from the control (Table 1). These results indicate that the cultivation of fish in ricefields does not significantly reduce rice yields and that cost reductions and increases in net income are significant.

Ecological Effects

Weed Reduction

In 1985, the ricefields had four species of weeds [wild arrowhead (Sagittaria sp.) was the main species]. Before the fingerlings were stocked, there were 408.6 weeds/m² in the control and 519.3 weeds/m² in the rice-fish field. Twenty days after the fingerlings were stocked, there were 131.4 weeds/m² in the rice-fish field (a reduction of 75%). Although the number of weeds in the control plot (289.8/m²) also decreased, there were still twice as many as in the rice-fish field. The weight of the fresh weeds in the control plot was 2268 kg/ha, which was 5.5 times more than in the rice-fish field. The number of weed species in the control increased to more than ten.

When the rice plants in the rice-fish field reached the stages of booting and filling, the lower part of the plants were clean and the surface of the field had only 9 weeds/m². There were 248.4 weeds/m² in the control with a fresh weight of 244.8 kg, which was 17 times more than in the field with fish. Before stocking, the number of weeds in the control increased from 408.6 in 1985 to 501.3 in 1987. In the field with fish, the number of weeds decreased from 519.03 in 1985 to 176.4 in 1987. Apparently, the cultivation of fish in a ricefield reduces weeds during that year, and also has long-term effects.

Soil Fertility

The fish in a ricefield can transform insoluble nitrogen in the soil into a soluble state, which increases soil fertility. Although the control plot used less fertilizer, the fertility in the rice-fish field improved significantly from 1985 to 1987. The percentage of organic material in the soil increased from 2.0 to 2.4%, total nitrogen from 0.14 to 0.16%, soluble phosphorus increased by 76.1%, and potassium by 20.69%. These results are similar to those reported by the Sanming Agricultural Research Institute in Fujian Province.

Physical Characteristics of Soil

The stirring movements of the fish aerate the soil and improve its structure. Vertical sections of the soil at depths of 0-20 cm were tested in 1985 and 1986. In rice-fish fields, the unit weight of topsoil was 2.1% lower than in the control, and porosity was greater by 1.2%. This effect was more obvious for subsoil. At depths of 10-20 cm, the unit weights of soil in rice-fish fields were 3.2 and 4.3% lower than in the control, and porosity was greater by 2.0 and 2.4% (Table 2).

Losses from Insects

Fish eat insects that float on the water surface and mosquito larvae in the water. Fu-shou fish and carp jump to catch rice lice (Sogata furcifera) in the lower stems of rice plants. In early September (the season during which rice lice emerge), the number of rice lice per 100 holes of rice was 234 in the control and 424 in the rice-fish plots. Ten days later, the rice-fish plot had only 42 rice lice, and the control had 138.1% more lice although insecticide was used in the control. Therefore, less insecticide can be used in rice-fish fields.

Nitrogen in Rice

The cultivation of fish in the ricefield increased the amount of nitrogen in the soil and the amount of nitrogen absorbed by the rice plants. In 1987, the total nitrogen content of rice plants during the booting stage was higher by 35.9% in the field with fish than in the control field, although the control plot had used 22.7% more fertilizer. Therefore, in the rice-fish field, more nitrogen is transported to the rice grains. The nitrogen content of rice from the rice-fish field was 11.2% in 1986 and 10.9% in 1987 (or 0.8% and 1.0% higher than the control). The quality of the rice in the fields with fish was better.

Factors Affecting Yields of Rice in Rice-Fish Fields

Planting Density

The yield of rice for a specific variety usually correlates positively with planting density. In practice, the upper limit of density is usually used. However, the same variety planted in a rice-fish field, would produce a different yield. In an experiment in 1985, densities within the range of 150 000-375 000 holes/ha were used for hybrid rice. The yield of the control plot increased with density,; whereas, yield in the rice-fish field decreased. With densities of 300 000 holes/ha and 375 000 holes/ha, yields from rice-fish fields were lower than the control by 3.9% and 7.5%; however, the decreases in yield were not statistically significant. Similar results obtained in 1986 and 1987 (Table 3) indicated that densities of 300 000 holes/ha or below did not affect the yield of hybrid rice.

The main reason for this effect on yield is the type of water management used in the rice-fish field. Taller plants, larger leaves, and longer nodes are produced, and at high densities, the population becomes too large, which decreases ventilation and illumination at the bottom of the rice plants. The development of the rice plant is stunted, which makes it vulnerable to insect pests, diseases, and lodging during bad weather. In contrast, at relatively lower densities, rice plants grow fast, individuals are strong, and plants are more resistant to lodging.

Therefore, the appropriate planting density for rice-fish fields is the lower limit for the variety (usually 10-20% lower than the density used under regular cultivation). For hybrid rice (Xian-you 63) the following parameters are suggested: distance between rows (26-33 cm), distance between holes (11-13 cm) with 270 000 holes/ha and 1 200 000-1 350 000 stems and tillers per hectare. As the rice grows, the highest number of stems and tillers will reach 3900000/ha and there will be 2 400 000-2 700 000 ears/ha. This planting density produces a good population structure, well-developed individuals, and high yields.

Fertilizer

If the amount of fertilizer used for regular cultivation is used in rice-fish fields, it will upset the carbon-nitrogen ratio and reduce soil fertility, weight per 1000 grains, and yield. For example, in 1985, the same amount of fertilizer was used in both the control and the rice-fish field. Grain-seed percentage decreased by 2.3%, weight per 1000 grains by 0.7 g, and yield by 4.3%.

In 1986 and 1987, the amount of fertilizer applied in the rice-fish field was 19.2% and 18.5% less than in the control. The grain-seed percentage, average weight per 1000 grains and yield were about the same as in the control. However, the level of nitrogen was 22.8% higher in rice from the rice-fish field. To achieve high and stable yields of rice from rice-fish, 20% less fertilizer should be applied (Table 4).

The strategy for rice-fish culture is low planting density, less fertilizer, a small population, and strong individuals. Improved ecological conditions (ventilation and illumination) prevent lodging and help produce larger ears, heavier grains, and high and stable yields.

Factors Affecting Fish Yield

Species

In 1985, the effects of mature fish and fry were compared. Mature fish grew quickly and individual weights at the time of harvest reached 79.3 g for grass carp, 60.0 g for common carp, and 76.9 g for Fu-shou fish. The largest individual weighed 150 g. White crucian carp, however, grew slowly (individual weight only 28.4 g) although the same size fingerlings were used for all species.

The fry were small and the densities high, but results were about the same as for mature fish. Average individual weights of grass carp, common carp, and Fu-shou fish were 3-5 times greater than white crucian carp and bighead carp (Table 5). Grass carp, common carp, and Fu-shou fish are the ideal species for rice-fish culture because they adapt easily and produce high yields. However, bighead carp and white crucian carp grow slowly and should only be used in small proportions.

Stocking Density

Yield correlates positively with stocking (Table 6). In 1986, two parts of the rice-fish field (A and B in Table 6) were used. There were 56.9% more fish in A than in B; 26.3% more fish were harvested from A, and yield increased by 37.6%. There were 65.9% fry in A than in B; 4.1% more fry were harvested from A and yield increased by 45%. Results were about the same in 1987 except that percentage yield increased. For densities of between 4500 and 18 000 fish/ha, yields increased with density (regression equation of y = 0.370 + 0.0312 x; where y = yield of fish per hectare and x = number of fish harvested per hectare; r = 0.9778). Higher densities inhibits the development of individual fish.

Stocking Size

Stocking size has a significant effect on individual weight gain during one season. Experiments using different densities and species, showed that the larger the fingerlings, the higher the individual weight gain (Table 7). When the stocking size of fish was increased from 4.0-5.0 cm to 8.3 cm in 1985 and from 6.6 cm to 8.3 cm in 1986, average individual weight increased by 131.4% (1985) and 18.8% (1986). In 1987, stocking size was 4.0 cm and individual weight was only 34 g, or 40% of the weight of the 8.3-cm individual fish stocked in the previous two years. The performance of grass carp and common carp were similar to that of Fu-shou fish. Increases in stocking size increase individual weight and improve the final yield of fish.

Discussion

Rice-fish culture is a feasible and efficient way to improve the use of agricultural resources. It improves soil fertility, reduces damage from weeds and insects (and therefore reduces costs for insecticides and chemical fertilizers), and improves the quality of rice. With a stocking rate of 15 000-30 000 fingerlings/ha, the rice yield is 7500 kg, fertilizer can be reduced by 37.5-75 kg/ha N, and 7.5 kg/ha less herbicide can be used. As a result, net income can be increased by over CNY 750/ha.

To ensure good rice yields from rice-fish fields, planting density should be lowered and less fertilizer should be applied. This helps develop a plant population with strong individuals and minimizes shading and lodging. Strong seedlings should be plated at a density that is 10-20% lower than the density used in a regular fields. Nitrogen fertilizer should be reduced by 20%; 85-90% should be applied early in the growing season, and the rest applied during later stages. The fish species should be adaptable, resistant, high-yielding, and be able to tolerate heavy doses of fertilizer.

The difficulties in rice-fish culture are the limited growth period, the small amount of water, changes in water levels, and unstable ecological conditions. To achieve high yields, it is important to choose the appropriate fish species and to use the proper stocking density and size. Grass carp, common carp, and Fu-shou fish are ideal species for rice-fish culture. They should be used as the main species are be combined with small proportions of other species. To produce mature fish, 70-80% Fu-shou fish and 20-30% grass carp and common carp should be used. For fry, 80% grass carp and common carp and 20% Fu-shou fish are recommended.

To produce 750 kg fish/ha, the stocking density should be 12 000-15 000 fingerlings for mature fish and 30 000-37 500 fingerlings for fry. It is best to use large fingerlings. To produce mature fish, the stocking size should be greater than 6.3 cm for Fu-shou fish and 4.3-6.3 cm for grass carp and common carp.

Although productivity of the individual rice plants was enhanced when fish were raised in the ricefield, the yield per unit area remained about the same. The improvement in the quality of the individual rice plants compensated for the reduced planting area. Results from this study show demonstrate that rice-fish culture maintains rice yields at the same level as regular cultivation. In contrast to other reports, this study did not find a tendency for rice-fish culture to increase yield.

Li Xieping, Wu Huaixun, and Zhang Yongtai are at the Li-Xia-He Regional Agricultural Research Institute, Yangzhou, Jiangsu Province.

Cultivating Different Breeds of Fish in Ricefields

Wan Banghuai and Zhang Qianlong

Experiments were conducted in Shanggao, Shangyou, and Shuichuan Counties in 1985-1986 to study the adaptability of certain breeds of fish to ricefields. Experiments were carried out simultaneously in double-cropped ricefields in three separate villages. Soil fertility was poor in one village, average in one, and good in another. At each site, the experiment was replicated. The method of rice-fish culture with trenches and ponds was used at all three sites. The trenches and ponds took up 4-10% of the total area of the ricefield.

In 1985, each site had 30 ricefield plots and a total area of 1.3 ha. In 10 plots (0.05-ha each), the fish breeds were cultured separately (monoculture) and given no supplemental food. In 10 plots (0.003-ha each), the fish were cultured separately and given supplemental feed. In the final 10 plots (0.06-ha each), different polyculture mixtures of fish were given supplemental food (single replicate per trial).

The polyculture mixtures contained nile tilapia (Oreochromis niloticus), grass carp (Ctenopharyngodon idella), silver crucian carp (Carassius auratus), local carp, and six other local breeds. In one trial, equal quantities of each breed was used (200 fish per breed). In the unequal mixed cultures, the main species (one of nile tilapia, grass carp, silver crucian carp, or local carp) made up 50% (1000 fish) of the total number of fish raised. Chub and variegated carp made up 2.3% (45 fish) each, and the other seven fish species made up 6.5% (130 fish).

Culture of Different Fish Breeds

The breeds chosen for culture were fish that could grow to 3-4 cm in length in that year. Ten breeds of seven fishes were selected: nile tilapia, grass carp, silver crucian carp, local breeds of red carp (Xingguo red carp, pouch red carp, and glass red carp), shortnose catfish, chub carp, and variegated carp. The three red carps, the silver crucian carp, and the shortnose catfish were distributed by the local government; the other breeds were produced on-site. The fish were stocked before the end of May, except for nile tilapia, which were put into the ponds from late May to early June.

Lime was used to sterilize the ricefield before the fish were stocked. The main fish fed was natural food in the ricefield; however, concentrated feeds, such as fine chaff, wheat bran, and rapeseed cake, were added as needed.

The rice plants were grown and managed in the same way as rice in fields without fish. Although the area for growing rice was reduced because of the fish trenches and ponds, rice yields were not reduced because rice was planted along the edges of the trenches and ponds. When fertilizers or pesticides were applied or when the field was sun-dried, the fish were drawn into the trenches or ponds. The ricefields with fish did not need weeding.

All plots were inspected during the last 2 weeks of October. Each fish species was counted, weighed, and measured, and field management notes were examined (Tables 1-3).

Adult Fish Culture in Ricefields

In 1986, eight breeds were cultivated in the three sites (silver chub and shortnose catfish were not used). The number of plots was decreased from 30 to 24, and the area was decreased from 1.3 ha to 1 ha. Eight plots were used for monoculture with and without feeding (without replicate experimental plots). The area of each plot was the same as in 1985. Two plots were used for equal-quantity mixed culture with feeding; the rest were used for 3-breed mixed culture with feeding (the three breeds were nile tilapia, grass carp, and local carp). Each breed was cultivated separately as the main breed in two plots using the same techniques used in 1985. For each breed, 4500 fish/ha were stocked in the ricefield. In equal-quantity mixed culture, 562 fish of each breed were stocked per hectare (each breed made up 12.5%). In unequal-quantity mixed culture, 2250 fish were stocked per hectare: 50% of the main breed plus 7% chub, 3% variegated carp, and 20% each of two other breeds. Breeds cultivated the previous year were harvested at a length of 10 cm, and stocking was completed by early May. Results are presented in Tables 4-6.

Results

Growth of Different Breeds (1985)

Nile tilapia. Even without feeding, high unit yields were obtained when nile tilapia was monocultivated or used to supplement mixed cultures. Survival rates, however, were not high because the fingerlings were small when stocked.

Grass carp. Survival rates were average; however, when grass carp were used to supplement mixed cultures, survival was high. Unit yield was low in monoculture without feeding. With feeding in both monoculture and mixed cultures, yields were higher than for other breeds.

Silver crucian carp. Survival rates were usually higher than for other breeds, but because the fish took longer to grow and body size was small, unit yields were low.

Local carp. Survival rates were high. Unit yields were high in mixed cultures. When local carp were used as the main breed in mixed cultures, unit yield was higher than for any of the other nine breeds. In monoculture, yields were low.

Pouch red carp. In monoculture or mixed culture, survival rate and unit yield were low.

Xingguo red carp. In monoculture without feeding, both survival rate and yield were fairly high, but in mixed culture, both survival rate and unit yield were rather low.

Glass red carp. Survival rates and unit yields were high in monoculture. In mixed cultures, survival rates were low, but unit yields were high.

Shortnose catfish. Survival rates and unit yields were low because the fingerlings used were small. Outside the experimental sites, some farmers stocked larger fingerling and obtained good harvests.

Variegated carp. In monoculture, survival rates and unit yields were the lowest of any breed. In mixed cultures, unit yields were low, but survival rates were high.

Chub. In monoculture, survival rates and unit yields were low. In mixed cultures (in the proportion 2.5-6.5%), survival rates and unit yields were the highest among the 10 breeds. Chub also grow fast.

Stocking Methods

Survival rates were lower in monoculture (30.6%) than in mixed cultures (36.6%) and lower in monoculture without feeding than in monoculture with feeding. Survival rates in equal-quantity mixed cultures were slightly lower than in unequal-quantity mixed cultures. For example, when used as the main breeds in mixed cultures, survival rates were: silver crucian carp > local carp > grass carp > nile tilapia. In mixed cultures, the survival rate of the main breed was usually lower than when that breed was used to supplement other breeds either in unequal-quantity mixed culture or in equal-quantity mixed culture.

Unit yields were lower in monoculture without feeding (16.5 kg) than in monoculture with feeding (276 kg); lower in monoculture with feed (276 kg) than in mixed cultures with feed (460.5 kg); and lower in unequal-quantity mixed cultures (354 kg) than in equal-quantity mixed cultures (519 kg). In unequal-quantity mixed culture, unit yields were: local carp > silver crucian carp > grass carp > nile tilapia.

Growth of Different Breeds (1986)

Nile tilapia. Fingerlings can be bred in the ricefield. In monoculture without feeding, the survival rate was not high, but in the other culture methods, survival rates were higher than for other breeds. Unit yields, however, were lower those for grass carp and local carp.

Grass carp. Survival was low, especially when grass carp were cultivated as the main breed in mixed culture without feeding. Unit yields were highest with most culture methods.

Local carp. Survival rates were higher than grass carp in most cases, but were lower than grass carp in equal-quantity mixed cultures with feeding. Unit yields were lower than grass carp with most culture methods.

Xingguo red carp. Survival rates were the highest of the four carp breeds, but unit yields were lower than local carp. In monoculture without feeding, unit yield was lower than pouch red carp and glass red carp.

Pouch red carp. Survival rates were the highest. Unit yields were low in mixed culture, but fairly high in monoculture, especially in monoculture without feeding.

Glass red carp. In monoculture without feeding, survival rates and unit yields were high. In monoculture with feeding as well as in mixed culture, survival rates and unit yields were low.

Variegated carp. Survival rates were fairly high, especially when cultivated without feeding. Unit yields, however, were low.

Chub. Survival rates were fairly high, but unit yields were low. In monoculture, survival rates and unit yields were lower than variegated carp. In equal-quantity mixed culture, both survival rate and unit yield were higher than variegated carp.

Stocking Methods

Survival rates in monoculture (61.6%) were slightly higher than in mixed cultures (59%) and lower in equal-quantity mixed cultures than in unequal-quantity mixed cultures. When cultivated as the main breeds in three kinds of unequal-quantity mixed culture, survival rates were nile tilapia > local carp > grass carp.

Unit yields in monoculture (19.6 kg) were much lower than in mixed cultures (32.4 kg); lower in monoculture without feeding than in monoculture with feeding; and higher in unequal-quantity mixed culture than in equal-quantity mixed culture. The unit yields of the three breeds cultivated as the main breeds in unequal-quantity mixed culture were ranked from lowest to highest as follows: grass carp > local carp > nile tilapia.

Discussion

The 2-year experiment on rice-fish culture was conducted under natural conditions (e.g., floods, droughts, and birds) and certain artificial factors (e.g., management level of staff, funds, the quality of fish breeds). It involved 50-60 farmers and was carried out in several sites in three counties, one in the South, one in the North, and another in the centre of Jiangxi Province. A lead group and a technical group were organized to undertake the experiment on the basis of unified leadership, unified planning, and unified standards. Limitations in the experimental methods included differences between sites in soil fertility, biological resources, water temperature, water quality, and water sources. There were also differences in rice-growing skills, management level, stocking, and time of harvest.

Conclusion

The fish breeds best suited to rice-fish culture are grass carp, common carp, and nile tilapia. These breeds can be used as the main breeds for rice-fish culture. The breeds most suitable for use as supplementary breeds in rice-fish culture are crucian carp (mainly silver crucian carp), local red carps, chub, and variegated carp. Shortnose catfish can be used as a commodity fish under certain conditions.

In this method of rice-fish culture, trenches and ponds occupy 6-8% of the total area of each plot. The trenches should be 0.35-m deep and the pond 1-m deep. Rice-fish culture can be carried out while a stable increase in rice yield is maintained. Fish yields of 450-600 kg/ha of adult fish and 150-225 kg/ha of fry can be obtained.

Wan Banghuai is with the Aquatic Products Department, Jiangxi Provincial Bureau of Agriculture Animal Husbandry and Fishery, Nanchang, Jiangxi Province; Zhang Qianlong is with the Animal Husbandry and Fishery Department, Yichun Prefecture, Yichun, Jiangxi Province.

Rice-Fish Culture in Ricefield Ditchponds

Luo Guang-Ang

Rice-fish culture has been traditionally practiced in rural areas. Modern rice cultivation techniques such as intercropping, fertilizers, and chemical sprays often conflict with fish growth. In addition, other problems such as monoculture of fish, delays in putting fish into the pond, short growing periods, and lower output and economic profits have arrested the development of rice-fish farming. Recent reforms in the economic system have sparked new initiatives in agricultural production.

Renewed interest in rice-fish culture has prompted scientists to study different patterns of rice-fish culture (e.g., ditch-and-pit, wide ditches, and ditchponds). Research suggests that rice-fish culture in ditchponds can eliminate the conflicts between fish and rice and take full advantage of the symbiosis between rice and fish.

Rice-fish culture in a ditchpond is a three-dimensional agricultural production system. The artificial ecosystem simulates the biostructure of the natural ecological system and consists of an economic crop (rice) and a commercial animal (fish). The system provides the high output of an intensive fishpond farming system and makes full use of the ecological conditions in the ricefield. Fish feed on the abundant aquatic organisms and coexist with the rice. Both fish and rice develop harmoniously, each promoting the growth of the other, and problems between fish-rearing and field care can be reasonably handled. Both crops can achieve their full productive potential. At the same time, liana can be grown on an awning above the pond, taro and beans can be grown on the bank, duckweed can be cultivated with the fish in the water, and loach can be grown in the mud. This integrated and intensive farming system produces higher yields of rice, fish, and vegetables.

Management Techniques

The pond, which takes up 5% of the field area, is dug to a depth of 1-1.5 m on one side of the ricefield. On a third of the pond, a shed with an awning is built. Ditches are dug 40-60 cm wide and 20-30 cm deep. The ditches are dug according to the size of the field: if the field is less than 0.07 ha, ditches are dug in a straight line; if the field is about 0.07 ha, ditches are dug in the shape of a cross; and if the field is larger than 0.14 ha, parallel ditches are dug in both directions (in the shape of a double cross).

Engineering Construction and Management

Five basic steps are undertaken.

· Select a suitable field. The soil must be fertile, have moderate texture, and hold water well. The water supply must be adequate and drainage and irrigation must be available to ensure stable yields during droughts or excessive rain.

· Construct a bank around the field. The bank is raised to a height of 50-70 cm and a width of 30-40 cm. The bank must be solid enough to withstand heavy rains.

· Install fish screens. A semicircle of fish screens about 0.8-m high and 1-m wide are installed in the water inlets and outlets. The gaps between fish-screen bars should be 0.3-0.4 cm. A screen with a gap of 1.5-2.5 cm is placed at the juncture of the ditch and pond to prevent large fish from damaging rice seedlings. This screen is removed after the rice heads.

· Monitor water and fertilizer. Water and fertilizer have a significant effect on the growth of fish and rice. Excess nitrogen makes rice seedlings spindly and results in the closing of crop rows too early and densely, which is fit for neither fish nor rice growth. Therefore, the amount of base manure, potassium, and phosphorus fertilizer should be raised but nitrogen should be controlled to prevent the rice from becoming spindly. Water management is also important. Irrigation and drainage should be controlled separately. The water level should be kept at a depth of 1-1.2 m in the pond; however, the water in the ricefield should be 5-10 cm deep before rice tillering and 10-16 cm deep after rice tillering. When pesticides and fertilizers are needed, the water level in the field should be lowered slowly and the fish should be driven into the ditch and pond to separate the fish and rice for a short period.

· Patrol fields. Care is required to identify and deal with problems such as drought, waterlogging of the fields, diseases, insect pests, escapes of fish, and theft.

Improvements in Culture Techniques

To improve rice-fish culture ditch ponds, several changes are needed in traditional techniques.

· Transform monoculture into polyculture. Traditionally, common carp are the only fish grown in the ricefields. However, now many other species are available, e.g., local carp, wuyuan pouch red carp, wang-an glass red carp, xingguo red carp, feng carp, mirror carp, grass carp, black carp, silver carp, big-head carp, nile tilapia, and crucian carp.

· Change from single-cropping to double cropping of fish. Traditional systems of rice-fish production raised fish once a year However, new techniques now mean that fish can be grown in ricefields throughout the year.

· Institute earlier stocking of fish. In traditional systems, the fish are stocked after the rice heads. New techniques allow the fish to be stocked after the rice is transplanted because of the ponds and ditches in the ricefield.

· Increase stocking density of fish. In the past, less than 1500 fish were stocked per hectare of rice-fish field. Currently, 15 000-30 000 fish per hectare are used. The density depends on the objective of the rice-fish system. For fish fry, the density of summer fry is usually 19 500-30 000 fish/ha; for food fish, 6000-9000 spring fish fingerlings per hectare are recommended. In both cases, grass carp, common carp, and nile tilapia make up about 90% of the total and black carp, silver carp, bighead carp, and crucian carp make up about 10%.

· Introduce feeding of fish. Water plants, plankton, and aquatic animals in the ricefield were the traditional sources of food for the fish. Green grass, duckweed, algae, rice bran, bean residue, distillers’ grains, and manure are now added to supplement the feed for the fish.

Integrated Management

In addition to fish and rice, diverse agricultural products can be obtained from ricefields. For example, soybean, peppers, tomatoes, sorghum, corn, taro, mustard can be grown on the banks; vine crops such as musky pumpkin, wax gourd, balsam pear, Lagenaria vulgaris, and cow gram can be grown on the awning of the shed; and duckweed can be grown under the shed awning as pig feed.

Impact of Rice-Fish Culture

Production experience has demonstrated that fish farming in ditchpond makes good use of the land, water, and biotic and abiotic resources of ricefields. The rate of use and conversion efficiency of material and energy in the rice-fish ecosystem are higher than in rice-only fields. The system is a desirable production model that combines fish with rice.

Economic Benefits

Fish rearing in ditchponds has a low cost of investment but yields diverse products, which in turn increase total agricultural income. This is demonstrated by some examples from Shangyou County, Jiangxi Province.

In 1984, fish were raised in ditchponds on 86.5 ha of ricefields. The output of fresh fish averaged 760.5 kg/ha, and the yield of rice increased by over 2900 kg/ha. The value of the total output increased by CNY 5780/ha. Zhong Linying, a farmer in Qixing, practiced monoculture of nile tilapia in a late ricefield and produced 3200 kg of fish per hectare. Cai Yunging, in Henglin, adopted polyculture of grass carp, silver carp, bighead carp, and common carp and produced an output of CNY 15 860/ha (net income CNY 14 700/ha).

In 1985-87, experiments in various parts of Shangyou County produced positive results. For example, a test of fish-rearing was undertaken by Kong Quingrong in a 0.08-ha double-cropped ricefield (area of pond 0.01 ha). The rice harvest was 976.7 kg (9% more than from the same field planted to rice the previous year). In addition, by using the field-bank ridge and shed awning to grow crops, he produced 5 kg of soybean, 5 kg of balsam pears, 15 kg of tomatoes, 30 kg of white gourds, 40 kg of musky pumpkins, 53 kg of kidney beans, and some other vegetables and livestock feeds. The net income, including CNY 264 from the fish, was 3.6 times as much as in the previous year (an increase of CNY 4590/ha).

In 1986, this production system was extended in Dongshang Township. Results from a 0.9-ha field distributed into 23 pieces of land were 40.2 kg (603 kg/ha) of fish and a rice yield that averaged 11 063 kg/ha (1.8% higher than in the previous year). In addition, average income was CNY 304/ha and total value reached CNY 5680/ha.

In 1987, results were obtained from five households in Dongshan Township. Kang Renhai obtained 15 018 kg of rice, 975 kg of fresh fish per hectare, and CNY 294 from other interplanted crops (e.g., taro, soybean, and pumpkin) for a total output of CNY 20 400/ha. In contrast, an adjacent field that was the same size but had no fish had an output that was 3.15 times lower in value.

Ecological Effects

The rice-fish system is an artificially controlled ecosystem in which rice and fish coexist, depend on, and promote each other. Fish play several roles in the system.

Weed control and preservation of soil fertility. In the ricefield, weeds compete with rice for nutrients, land, water, space, and sunlight. As a result they greatly affect the growth of the rice plants. Experiments have shown that 1 kg of grass-carp fingerlings in a ricefield consume about 40-60 kg of weeds, which would absorb about 1.25 kg of nitrogen as they grow. In the rice-fish system, weeds are eaten by grass-eating fish and their waste becomes a field manure that helps conserve and enrich soil fertility.

Huang Xinlian, a farmer of Shuiyan Township in Shangyou County, adopted the rice-fish system with a ditchpond for 3 years. In 1983, from a 0.06-ha field, she obtained 1083 kg of grain (138 kg more than the previous year) and 38 kg of food fish, 1109 large fingerlings of grass carp, silver carp, and common carp, and 32 Japanese crucian carp. She obtained a total income of CNY 652 and a net income of CNY 451 (CNY 7114/ha). In addition, she saved CNY 12 for chemical fertilizers and pesticides and no tillage or weeding were needed for 3 years. Soil fertility and yields have increased each year.

Control of rice diseases and insect pests. Insects that are harmful to rice are a good food for fish. For example, when they fall into the water, rice borers, rice hoppers, and rice weevils are quickly eaten by the fish. Observations in experiment plots have indicated that densities of pest populations were lower, as was the damage to rice plants, when fish were present in the field. For example, for the third and fourth generations of rice borer the density of the third generation was 1305/ha and the rate of dead hearts in the rice was only 0.4% in the rice-fish field. In the fourth generation, pest density was 1380/ha and the rate of white heads was only 0.9% in the rice-fish field, but in a rice-only field the density was 1650/ha and the rate of white heads was 1.4%.

The vegetable and other crops grown on the field bunds also provide habitat for natural predators. The crops on the bunds and shed awning also enhance fish growth because they shade the water and lower water temperature.

The fish also reduces rice diseases by oxygenating the soil and speeding the decomposition of manure and the release of available nutrients. Large grass carp and common carp remove the basal leaves of the rice plant and diseased leaves, which allows air and sunlight to pass through easily. This promotes rice growth and reduces the incidence of rice diseases.

Early rice sheath-blight disease was investigated in Wangzai County. The incidence of diseased hills was 17.1% and rate of diseased plants was 2.7% in the rice-fish field, compared with 42% diseased hills and a rate of 6.1% for diseased plants in the ricefield without fish.

Social Effects

Making multiple uses of the same field is an important advantage of this system. Large fish fingerlings and commercial fish for market are produced without additional land. For example, five farm households in Shanyou County produced, from an area of 0.2 ha, an average yield of rice of 13 760 kg/ha, 568.5 kg/ha of fish, and an income of CNY 1310/ha from vegetables. Total income reached CNY 9635/ha.

Fish consume weeds and pests, prevent disease, conserve and increase fertilize, reduce or eliminate chemical pesticides, and as a result, also save farm labour. Experiments have demonstrated that 8-10 units of labour power are saved when fish are present. The decrease in the amount of agricultural chemicals also reduces chemical poisoning and environmental pollution. At the same time, natural pest predators such as praying mantis, spiders, and frogs correspondingly increase. In addition, because fish are predators of mosquito larvae and snails, they help prevent malaria, flariasis, and snail fever.

Conclusion

Agricultural production is being changed in Shanyou County. Farmers now know how to use the limited arable land for multiple purposes. There are several advantages to the new production system.

First, the rice-fish system increases economic benefits two or three times compared with a single rice crop. Second, the production system integrates the main crop (rice) with fish and vegetables and improves agricultural production. Third, the principles of a natural ecosystem are borrowed to promote ecological balance and ecological cycles. Fourth, cropping and fish-rearing are linked in a simple way that uses land, water, and biotic and abiotic resources efficiently and features low inputs, quick effects, and excellent benefits.

Because fish-rearing in the ditch pond of a ricefield is a new production system, many items require further research. For example, the proportion of different fishes and the proportion of pond and ditch area must be tested and studied. To further improve the use of ricefield resources, cooperation is required between scientists working in the departments of agriculture, protection, and health.

Rice-fish culture in the ditch ponds of ricefields has great potential. There are 1.3 million ha of ricefields that are suitable for fish-rearing in Jianxi Province. If 25% of these fields were used for rice-fish culture and the unit output of fish was 750 kg/ha, then 250 000 tonnes of fish could be produced. This is 148 000 tonnes more than the total output of fish in 1984. At a market price of CNY 3/kg, the output of fish would increase income by CNY 750 million. If CNY 250 million were produced from interplanted vegetables (CNY 750/ha), the additional amount of revenue would be CNY 1000 million.

Luo Guang-Ang is with the Bureau of Agriculture, Animal Husbandry and Fisheries, Shangyou, Shangyou County, Jiangxi Province.

Rice-Fish Culture in China: Zero-Tillage Ricefields

Techniques for Rice-Catfish Culture in Zero-Tillage Ricefields

Chen Huarong

Yunnan is situated on a low-latitude plateau that has a variety of physical features and many variations in climate. Agricultural production differs dramatically between regions and seasons. Ricefields in Yunnan cover 1 million ha, and more than 132 000 ha are suitable for rice-fish cultivation. The rice-growing regions vary from warm and sunny with plenty of rainfall and good soils to areas with low temperature and poor water-holding capacity.

The ricefields are not efficiently used and economic returns could be improved. Rice-fish fields only cover about 13 300 ha, which is only 10% of the area suitable for rice-fish culture. Average output is also rather low (45 kg/ha in 1982, 62 kg/ha in 1985, and 101 kg/ha in 1987) compared with the national average of 140 kg/ha.

Full exploitation of the potential productivity of the ricefields could remarkably increase economic benefits. Rice-fish cultivation is an important way to increase productivity. Experiments were conducted in Kunming, a rice-growing region of Yunnan Province, at an elevation of 1900 m and with an annual average temperature of 14.5° C.

Fish Species

The choice of species greatly affects fish harvests, the value of the output, and the economic benefits gained from ricefields. The characteristics of the rice-growing regions in the Yunnan plateau are low temperatures, low water temperatures, shallowly flooded ricefields. The traditional fish species grow slowly and the growing period is short. Given these conditions, it was important to find fish species that grew quickly (reached market size in 120 days), tolerated low-oxygen conditions, were of high quality, and produced high outputs. Experiments were conducted in 1986-88 to compare different species.

Vigour and Production

Different species were raised for 120 days in the ditches and beds of zero-tillage ricefields. Growth rates and yields varied greatly (Table 1). Clarias leather grew the fastest. Individual weights increased 168-fold and their average weights were four times that of nile tilapia and 10 times that of the carp. The length of the catfish increased 4.2-fold. Yields of C. leather were the highest. When stocked at 25% of the population C. leather yielded 71.4% of the harvest.

Yields of C. leather were twice as high in monoculture than in polyculture (average 3.1 t/ha in 1988; largest individuals weighed 450-600 g). The size of the young fish used for stocking influenced yield and economic return. Overwintered large-size young fish produced higher yields than smaller-size fish hatched the same year. The survival rate of overwintered fish to food fish was 90%, and the harvest increased significantly when larger-size young fish were used. Fish that were 15-20 cm in length constituted 63% of the average yield; whereas, fish less than 10 cm in length constituted 37% of the yield.

Economic Benefits

Economic benefits were closely related to the species (Table 2). C. leather produced the best economic returns. In mixed culture, C. leather constituted 43% of the total input and produced 74% of the total output value. Net income from C. leather was CNY 5640/ha (93% of total net income) and the benefit ratio was 1:4.5. The output value of nile tilapia and carp was 15-21% of the total output, net income was CNY 75-367.5/ha, and the benefit ratio was between 1:1.1 and 1:1.3. In monoculture, the output of C. leather amounted to CNY 24 552/ha, net income was CNY 15 000, and the benefit ratio was between 1:2.6 and 1:2.9

Experiments and demonstrations in ricefields over 3 years showed that C. leather grows quickly, produces high yields, and is of good quality. C. leather is considered to have 10 advantages. They grow exceptionally fast, individual fish are large, they are omnivorous and capable of eating coarse food materials, they tolerate “humble” living conditions, they are resistant to diseases, they have a high survival rate, they tolerate of low oxygen levels, they are suitable for cultivation in dense populations, they produce excellent output, and they produce good economic benefits for farmers.

Rice-Fish Cultivation

The techniques for rice-fish cultivation have been improved. Different fish species have been used and economic benefits have been increased by better integrating the use of the ricefields. Two patterns of the cultivation are currently used.

Traditional Method

This is the main method used in Yunnan. Although the method does produce some economic benefits, yields are limited because of deep ploughing, close planting of the rice, the use of shallow-flooded fields with few ditches, the stocking of small numbers of small fry, and no additional feeding and management. This pattern of production yields on average less than 150 kg/ha (maximum 300-450 kg/ha). It is important to improve the techniques and gradually encourage farmers to change from low-yield, extensive cultivation to the high-yield, intensive methods of cultivation.

Improved Method

This method combines the culture of fish in ditches and beds in the ricefield with zero-tillage of the fields. This method has been successful in Kunming, a rice-growing region with an annual average temperature of 14.5°C (monthly averages from May to September are 18.9°C, 19.4°C, 19.7°C, 18.9°C, and 17.4°C). Corresponding water temperatures in the ditches are 21.6°C, 22.1°C, 23.3°C, 22.2°C, and 20.3°C. Several steps are involved in establishing this type of rice-fish system. First, choose a ricefield with good water supply and a convenient irrigation and drainage system, good water-retaining properties, and little shade. Second, dig ditches and divide the field into beds (zero-tillage) - ditches should be 0.4-0.6 m in depth and 0.4-0.5 m in width, the beds 2-3 m wide, and the ditch area should be 10-15% of the total area of the ricefield. Third, spread manure over the field before the ditches are dug, pulverize and level the top soil, and dig out the ditches to cover the manure before the rice seedlings are planted. Fourth, raise and reinforce the surrounding small dikes with the subsoil dug from the ditches. The ditches should be straight, level, and have a flat bottom. Tillage should not be needed for 5 years.

The main features of this cropping pattern are: zero-tillage fields with deep ditches and wide beds; rice planted in shallow water and rice and fish grown together; symbiosis of rice and fish, which reduces stress on both; intensive management and multiple use of water; and increased income from both rice and fish. Additional ditches can be added without reducing yield, the ricefield can be dried without damaging the fish, and chemicals can be applied to the rice without killing the fish.

Income generation. Rice-fish yields increased each year. Rice yields were 7-12% higher in the zero-tillage rice-fish system than in ploughed fields without fish. Fish yields were more than 3000 kg/ha, which was over 10 times the yield of fish from ricefields cultivated in the traditional way. The value of rice plus fish was CNY 27 525/ha and net income was CNY 16 815/ha (an increase of 8.6- to 19-fold, Table 3).

Ecological benefits. Increases in yields of rice and fish were closely related to the patterns of zero-tillage, bed division in the ricefield, rice and fish being grown together, and the selection of specific fish species. The rice and fish derived mutual benefit from the system, and ecological conditions in the ricefield were improved in several ways (Table 4).

· Improved habitat. A mutually beneficially habitat was provided for the rice and the fish. Circulation of air and penetration of light were improved between beds and rows. Light penetration in fields with bed divisions was 12.6% higher than in fields without divisions. Water temperatures were 0.7°C higher, there were 6-10 more rice grains per panicle, and the rate of empty, shrunken grains was 5-10% lower. These changes in light and water conditions favoured growth of both fish and rice.

· Higher yields of rice. Experiments have shown that in zero-tillage fields, shallow-planted plants constitute 94.6% of the population. In ploughed fields, deep-planted plants make up 77.2% of the population. (Shallow-planting less than 6.7 cm, deep planting more than 10 cm). Shallow-planted plants recover more readily, tiller earlier, and more vigorous, have a larger number of productive tillers, a higher percentage of ear-bearing tillers, a lower percentage of empty, shrunken grains, and have heavier panicles with less diseases (Table 5). Zero-tillage is a new practice that solves the problem of plants being planted too deep and also saves energy and labour.

· Improved soil fertility. Rice-fish culture is an efficient way to accelerate soil enrichment. Ricefields provide fish with a rich source of natural food, and the fish excrete manure, loosen the soil, and eradicate weeds. Soil analysis (Table 6) shows that rice-fish cultivation with successive zero-tillage increased organic matter and active nitrogen, but reduced phosphorus in the soil. It is necessary to apply large amounts of manure and smaller amounts of fertilizers. With successive rice-fish cultivation, it is particularly important to control nitrogen and increase phosphorus to maintain steady rice growth during the entire crop cycle.

Because the fish loosen the soil, the soil has higher permeability, which promotes decomposition of complex soil nutrients, manure, and fertilizer and improves rice growth. Root systems are also better developed (25.5% increase) and more widely and deeply distributed. Large amounts of weeds were eaten by C. leather, which has a good appetite and also devours insects. As a result, the ricefields were nearly free of weeds without cultivation or weeding. In contrast, fields planted to rice, but not stocked with fish, require weeding twice during the growing season and still contained 2000 kg of weeds per hectare at harvest. Oriental army worms and rice planthoppers were eaten by the fish; therefore, no pesticides were needed and labour for spraying was saved. The percentage of diseased rice plants decreased by 7-14% compared with fields without fish.

Conclusion

The most desirable fish species for rice-fish cultivation in the Yunnan plateau is C. leather. It grows quickly, is of good quality, tolerates low oxygen, and can be stocked at high densities. The use of ditches and beds, combined with zero-tillage, produced the best economic returns from rice-fish cultivation in Yunnan. This system produces increased outputs and income from both rice and fish and imparts additional ecological benefits to the ricefield.

Chen Huarong is with the Yunnan Academy of Agricultural Sciences, Junming, Yunnan Province.

Demonstration of High-Yield Fish Farming in Ricefields

Cai Guanghui, Ying Yuguang, Wu Baogan, He Zhangxiong, and Lai Shengyong

Traditional fish farming in ricefields yields about 150 kg of fish per hectare. To improve economic efficiency and stimulate the development of a commodity economy in rural areas, improvements have been made in rice-fish culture since 1984 in several major rice producing provinces (e.g., Sichuan, Hunan, Hubei, Jiangxi, Guangxi, Anhui, Jiangsu, Zhejaing, Guizhow, Guangdong, and Yunnan).

In 1986, the Science and Technology Commission of Guangxi Autonomous Region assigned a rice-fish project through the Spark Program to the Guangxi General Station of Aquatic Technical Extension. Demonstration experiments were conducted in ricefields for 2 years. High yields of rice and fish were obtained over large areas using a ditch and pit method. In 1986, the experimental area amounted to 48 ha and 29 657 kg of fish were harvested. The average yield was 620 kg/ha which was 3.8 times the average yield of other rice-fish systems in the region. The average rice yield was 10 700 kg/ha.

In 1987, the demonstration area was increased to 55.5 ha, the total fish yield was 40 268 kg (average 725 kg/ha), and the average rice yield was 11 610 kg/ha. These results surpassed the target of the specialized contract by 525-600 kg/ha of fish and more than 7500 kg/ha of rice.

Experimental Methods

Experimental Fields

The two methods that were used were the pit and ditch method and the ridge and ditch method.

Pit and ditch method. In 1986, there were 48 ha of demonstration fields and 529 households participated. In 1987, there were 55.5 ha of fields and 658 households. The experimental fields were distributed from north to south in a total of 26 towns in 14 counties.

Water resources in the experimental fields were abundant, drainage and irrigation were convenient, and yields were ensured despite drought or flood. Before the early rice was transplanted, 6-8% of the ricefields were dug into fish pits that were 65-100 cm deep. At the same time, the footpaths between ricefields were dug 20 cm deeper to form fish ditches that connected all of the fish pits.

Ridge and ditch method. In 1987, in Cenqui County of Wuzhow District, experiments were conducted using a ridge and ditch method on 3.4 ha. Seven days before the early rice was transplanted, the ricefields were tilled and base manure was spread. All of the water was then drained from the field. A ditch (50 cm wide and 40 cm deep) was then dug around the field. If the field was large area, an x- or #-shaped ditch was dug. After the ditch was dug, the ridges were made in an East-West direction. The ridges were 24-26 cm wide and 20-22 cm high. The ditches were 36-40 cm wide and 20-22 cm deep.

Stocking Methods

Two methods of stocking were used: single stocking and two stockings. Single stocking is usually done before the end of April. When two stockings are used, the first is between early April and early May when the early rice is transplanted. The second stocking occurs during the second half of July. Larger fish are partly harvested before or after the harvest of the early rice, and fish fry (4-7 cm long) are stocked.

In 1987, fish fry were stocked at the rate of 11985 fry/ha. The fish stocked were 80% common carp, 6.1% grass carp, 11% nile tilapia, 2% silver carp, and 0.6% bighead carp. Various experiments were conducted to investigate the effects of: stocking carp at different densities, different species (common carp, grass carp, and nile tilapia), and different feeds (fresh plants, concentrates, and fresh plants mixed with concentrates). Additional comparisons were made between rice yields with and without fish in one ricefield divided into two parts. These experiments were carried out at all the 14 counties.

Day-to-Day Management

Rice production in the demonstrations field was conducted according to the conventional methods used by farm households. In most demonstration fields no weeding was done. The depth of water in the ricefields was generally between 6 and 10 cm; however, during the flowering period the water depth was increased to 15 cm. Atmospheric temperature varied between 16.5°C and 39°C, water temperature was 20-36.5°C, and pH was 6.4-7. Bran cake, green forage, and fertilizer were applied as required.

Collection of Data

In each of the 14 districts, one observation station was established for every 3.3 ha of ricefields and 37 farm households were used for observation. Aquatic scientists and technicians visited the experimental fields periodically to make observations and record data.

Results

Yield of Fish and Rice

Fish and rice were grown in the experimental fields for between 176 and 320 days. In 1986, the 48 ha of ricefields yielded an average 620 kg of fish per hectare. This represented an increase of 4.8 times the average yield of 106 kg/ha obtained from rice-fish culture in the district. The average rice yield with fish was 10 700 kg/ha, an increase of 4.8% compared with ricefields without fish at the same location.

In 1987, there were 55.5 ha of demonstration fields. The average fish yield was 725 kg/ha, which was an increase of 5.5 times compared with the average yield in the district. The average rice yield with fish was 11 605 kg/ha, an increase of 7.2% compared with fields without fish. There were 3.4 ha of rice-fish in the ridge and ditch method. The average rice yield was 12 307 kg/ha, an increase of 8.9% compared with other ricefields.

Survival Rate and Average Weight

In 1987, 665 711 fish were stocked in the demonstration fields and 439 557 fish were harvested. The average survival rate was 66%. Survival rates for each species were: carp 65% (52-79%); grass carp 67% (52-75%); nile tilapia 65% (50-78%); silver carp 68% (62-76%); and bighead carp 75% (71-77%).

The average weights of the different fish varieties were: carp 69.6 g; grass carp 242.6 g (the largest fish was 400 g); nile tilapia 116.1 g; silver carp 255.3 g; and bighead carp 343 g. The yield proportion of the different species was: carp 60%, grass carp 16%, nile tilapia 14%, silver carp 6%, bighead carp 3%, and others 1.1%.

Stocking Densities

In 1987, at Guangyang County, Guilin District 3, 0.16 ha were used for trials of different stocking densities (4500, 9000, and 15 000 fish/ha). Stocking at a density of 9000 carp fry per hectare produced the highest yield (834 kg/ha). At a density of 4500 carp fry per hectare yield was 495 kg/ha. At 4500 carp fry per hectare, the average weight per fish was highest. The largest fish was 73.1 g. At a density of 9000 fish per hectare, the largest fish was 70.8 g; at 15 000 fish per hectare the largest fish was only 37.9 g.

Varieties

In 1986, trials were conducted with crucian carp, grass carp, and nile tilapia as the main stocked fish. High yields could be obtained for all species. When summer grass carp were stocked at 27 000 fish per hectare, with the objective of rearing large fingerlings, 14 400 fry larger than 10 cm were harvested per hectare (survival rate 53%).

Feeding

In 1986, trials of different fish feeds were undertaken. No matter what kind of feed was used (e.g., farmyard manure, green fodder, bran cakes, or combinations of all three), the yield of fish could be increase. A combination of concentrated feed, green fodder, and farmyard manure produced the highest yields.

Rice-fish Compared with No Fish

In 1987, in a total of 1.7 ha of ricefields owned by 22 households in four districts, experiments were conducted to compare rice-fish farming with rice-only farming. The heading rate, fruiting rate, number of grain per head, 1000-grain weight, and nitrogen, phosphorus, and potassium levels in the soil were higher in the rice-fish systems. The average yield from the rice-fish fields was 5.6% higher (Tables 1 and 2).

Weeds were reduced by 430 kg/ha in rice-fish fields in Yulin District; 3117 kg/ha in Qinzhou District. Weeds in rice-fish fields in Guilin District were 7.7% of the level in the control field.

Economic Efficiency

The direct economic gain from the experimental fields during the 2-year period was increased to CNY 222 840. Of this amount, CNY 181 902 was net income derived from fish farming. The demonstration experiments on 1070 ha of ricefields produced an average fish yield of 293 kg/ha and net income was increased to CNY 11 530/ha. The total benefit (direct plus indirect benefits) was over CNY 1 million. The ratio between input and output was 1:5.51.

Ecological Effects

Fish grown in the ricefield eat mosquitoes and control diseases. Research conducted in 1987 in Quanzhou County by the Parasitic Disease Research Institute of Guanxi Autonomous Region, and the Sanitation and Antiepidemic Station of Quanzhou County. Carp and grass carp raised in ricefields successfully controlled mosquitoes. The relative density index (RDI) was between 0.9 and 42.9. In general, the number of larva and pupa in rice-fish fields was remarkably lower than in the fields without fish.

Conclusion

Several new techniques have been introduced to help improve traditional rice-fish culture:

· Pit, ditch, and ridge and ditch fish-farming systems in ricefields are modifications of traditional methods. The objective of these new methods is to reduce conflicts between fish farming and rice growing and to increase the yields of both rice and fish and raise the economic efficiency of the ricefields.

· Carp, grass carp, nile tilapia, silver carp, and chub are now mixed for rearing. The traditional method of raising only carp was changed to increase fish yields, promote rice production, and improve sanitary conditions in the rural areas.

· The use of over-wintered fingerlings has significantly increased fish output compared with the traditional method of breeding fingerlings the same year.

· The combination of the production of edible fish with the culture of fingerlings has changed the habit of breeding only fingerlings or only edible fish. This has provided edible fish and abundant fish varieties for market.

· The introduction of appropriate feeding has promoted fish growth and increased income.

· One species of fish adapted to local conditions is now used as the major species. Selection of brood stock has also been improved.

The key techniques used to increase fish yields in ricefields are:

· Use reasonable stocking densities and ratios of multiple fish varieties. Grass carp, nile tilapia, silver carp, and variegated carp were stocked at a rate of about 12 000 fish/ha.

· Breed and stock over-wintered fingerlings.

· Dig fish pits and fish ditches to solve the contradictions between fish farming and rice management.

· Apply appropriate fertilizers and feeds to supplement natural feeds found in the ricefields.

Cai Guanghui is with the Guangxi General Station for Extension of Aquatic Technology, Nanning, Guangxi Autonomous Region; Ying Yuguang is with the Livestock and Aquatic Bureau of Yulin Prefecture, Yulin, Guangxi Autonomous Region; Wu Baogan is with the Aquatic Technology Extension Station, Guilin Prefecture, Guilin, Quangxi Autonomous Region; He Zhangxiong is with the Aquatic Technology Extenion Station, Wuzhow Prefecture, Wuzhow, Guangxi Autonomous Region; and Lai Shengyong is with the Agriculture, Animal Husbandry, and Fishery Bureau, Quanzhou Prefecture, Quanzhou Autonomous Region.

Rice-Azolla-Fish in Ricefields

Chen Defu, Ying Hanquing, and Shui Maoxing

The yield per hectare of traditional fish-raising methods is only 75-150 kg in China. In Zhejiang Province, a high-yield rice-azolla-fish system was developed and extended to farmers. The technique has now been adopted in many regions in Zhejiang Province, and rice production and fish yields have both increased.

In 1987, a demonstrative farmer, Shao Shousheng in Yuhang County, tested the high yield rice-azolla-fish system. The Zhejiang Academy of Agricultural Sciences, Yuhang Agricultural Committee, and Yuhang Science and Technology Society conducted the experiments. The techniques used to simultaneously raise rice, azolla, and fish are discussed.

Test Methods

The test field was located in Xingqiao Village, Yuhang County, 2 km from Hangzhou City. The field was 0.3 ha and had been used in 1986 as a test field for high-yield fish culture. Therefore, the farmer, Shao Shousheng, had practical experience. The fish ditch was 3 m in width, 1 m in depth, and occupied 21% of the total ricefield. Fine feed was used as the main fish food, and organic manure and fertilizers were used as supplements. Omnivorous, carp and crucian carp were raised. In 1986, the yield of rice was 9730 kg/ha and the net yield of adult fish was 3426 kg/ha. A comprehensive experiment on rice-azolla-fish was started in 1987 and several changes were made in rearing techniques:

· Instead of using fine feed for the fish, the fine feed was first fed to pigs and the pig manure was fed to the fish.

· Rice-azolla-fish were grown together in 1987, and the azolla were used as the main feed for the herbivorous fishes.

· Herbivorous fishes were chiefly raised in 1987 (silver carps, common carp, and crucian carp).

Test Results

Yield of Rice and Fish

The rice yield was 9786 kg/ha in 1986. The early japonica rice (Biyuzaonuo) grew well in 1987 and produced an additional 980 kg/ha. However, transplanting of the late rice was delayed to nearly the beginning of Autumn because the early rice was late to mature. It was also difficult to keep the grass carp in the fish ditches away from the rice on the sides of the ditches. Because the fish eat some of the rice, rice yield decreased to 1233 kg/ha, but the yield of adult fish increased greatly (Table 1). In 1987, the yield of adult fish from the rice-azolla-fish system was 6990 kg/ha, an increase of 71% over the 4200 kg/ha in the rice-fish system in 1986. If the yield of fish fry is excluded, the net yield of adult fish was 5500 kg in the rice-azolla-fish system, an increase of 60% over the 3444 kg/ha in the rice-fish system.

Feed Consumption and Cost

Fine feed was the main food source for the fish in 1986, and 7300 kg/ha of feed were consumed. Therefore, the cost of fish raising was rather high. Major modifications were made in 1987. The fine feed was first used to feed pigs. The pig manure was then fed to the fish and used to fertilize the azolla. This modification made full use of material inputs and added four pigs to the output with limited input. The cost of feed for fish raising was also reduced. The unit cost of feed in the rice-azolla-fish system was only 47% of the cost in the rice-fish system, and the feed cost of fish per kilogram was 29% of the cost in 1986 (Table 2).

Economic Efficiency

The cost of fish raising was almost equal both years. The output value and net profit in the rice-azolla-fish field increased significantly because the adult fish yield in the rice-azolla-fish field was higher than that in the rice-fish field. The output value was CNY 17 100/ha and net profit was CNY 9278/ha in the rice-fish field in 1986; whereas, in 1987, the output value was CNY 25 404/ha and the net profit was CNY 17 528/ha in the rice-azolla-fish field (an increase of 89%, Table 3).

Analysis and Discussion

There were several reasons for the yield increase and the cost decrease in the rice-azolla-fish fields.

Use of Herbivorous Fishes

Herbivorous fishes (grass carp and bream), especially grass carp, are fond of azolla and grow quickly. When there is a sufficient supply of azolla there is great potential for yield increase. The omnivorous fishes (carp and crucian carp) require high-quality fine feed, but grow slowly. These omnivorous fishes were the most numerous in 1986 (79% of the fish and 72% of the total weight). Herbivorous fishes comprised 14% of the fish and 15% of the total weight. Among the herbivorous fishes, grass carps accounted for 5% of the fish and 4% of the weight. In 1987, azolla was grown in the ricefields and herbivorous fish were the main species raised (48% of the fish and 31% of the weight). Grass carp accounted for 32% of the herbivorous fish and 29% of the weight, and bream 16% of the fish and 1% of the weight. Omnivorous fishes comprised 7% of the fish and 11% of the weight. Oligophagous fish (silver carp) accounted for 45% of the fish and 58% of the weight (Table 4). The azolla were consumed by the herbivorous fish, and the manure from the grass carp increased the amount of plankton, which raised the yield of silver carp.

Proportions and Harvest of Fish

In 1986, too few fish were raised early in the year and too many were raised later. At high densities, all fish were almost the same size, which made batch harvesting impossible. The market size of the fish harvested at the end of a year was also low. In 1987, the density of the fish was reduced in rice-azolla-fish system. In addition, 983 larger fish (3760/ha) were stocked, which accounted for 27% of the total number of fish raised (1123 kg). Of these larger fish, 188 were the grass carp (719/ha), 300 were the common carp (1148/ha), and 495 were silver carp (1757/ha), with the mean weight of 0.42 kg, 0.15 kg, and 0.38 kg, respectively (Table 5). In this way, the fish-holding capacity early in the year increased to 102 kg in 1987 from 45 kg in 1986. Grass carp and common carp fed on azolla in mid-March; whereas, the larger grass carp fed heavily on azolla during April to June. Grass carp and silver carp were caught in batches to supply the market. During April to October, 851 kg of large fish were harvested.

Azolla

The growth of fish and azolla in rice-azolla-fish system were different. Azolla grows quickly in the spring when the fish are small, grow slowly, and eat little. At this time, the azolla supply exceeds demand. In July and August, azolla grows slowly and the fish grow quickly; therefore, the demand for azolla exceeds supply. Three methods were used to mitigate this problem.

Harvest adult fish. At the end of June, adult fish were caught for market to decrease the fish-holding capacity of the field when the azolla were growing slowly. The grass carp grow very quickly during September, which is the second peak of azolla growth.

Supplement supply of azolla. The azolla supply in rice-azolla-fish fields could not satisfy the demand during July to August. Azolla from nearby ricefields, ditches, and ponds were used to supplement the supply. Green-stored azolla and dried azolla were also used.

Adjust feed. Less fine feed was used when azolla were plentiful, and more fine feed or grass was fed when azolla levels were insufficient.

Stopping Fish From Eating Rice

Two methods were used to stop the fish from eating the rice. The fish were kept in the fish pits with dikes and nets after the rice was transplanted. In addition, a grass field and feed platform were established in the fish pit for the omnivorous fish. Tender grass was placed in the grass field. Fine feed was placed on the feed platform when the tender grass was almost completely eaten.

Raising Fish in Flowing Water

Flowing water has a high oxygen content, which favours fish growth. New irrigation water was added at regular intervals. During cloudy, muggy weather, when the fish lacked oxygen, a submersible pump was used to make the water flow and increase the oxygen content. Once a month, 75 kg/ha of quick lime were diluted in water and sprayed over the field to decrease the concentration of acid in the water. This treatment promotes the breakdown of organic materials and helps sterilize the water and prevent fish diseases.

Pig-Azolla-Fish-Rice System

This system reduced the cost of raising fish. Pigs were fed with fine feed. Fish and azolla were fed with pig manure. Azolla were used to feed the pigs. Fish manure also enriched the field. In this way, costs were reduced while net profit was increased.

Conclusion

Experiments in 1987 indicated that the high-yield rice-azolla-fish system was a success. Fish yields and net profit were increased and rice yields were maintained. This system can improve the economic efficiency of the ricefield; nevertheless, there are still some problems that require further study.

· Grass carp quickly grow to a large size. Silver carps are smaller and have a lower commercial value. The appropriate proportion of silver carp to grass carp must be studied. Bream grow slowly and should be raised in small proportions in rice-fish-azolla fields.

· In 1987, dikes and fish nets were used to prevent the fish from eating rice. This was expensive in both capital and labour. Many late rice seedlings were eaten by the fish in 1987, which resulted in a decrease in rice yield. Changes to the design of the fish ditch are proposed to allow the farmers to block the fish with one dike and one net. This would make the operation easier and more convenient.

· Azolla continue to grow in the winter and spring. This potential should be developed to build up resources that can be used as fish feed.

· The rice yield might be improved by transplanting the rice seedlings earlier. In 1987, rice seedlings were transplanted during the second 10 days of May. This caused the ripening stage to be postponed. The rice yield might be increased if seedlings were transplanted during the first 10 days of May.

· The harvesting method for the fish should be changed. Fish should be raised and harvested on a rotational basis. This means that different species should be raised in different proportions. The appropriate size and time of harvest also needs to be established for each species.

Chen Defu, Ying Hanquing, and Shui Maoxing are with the Zhejiang Academy of Agricultural Sciences, Hangzhou, Zhejiong Province.