AQUAFEEDS IN ASIA - A REGIONAL OVERVIEW
Michael B. New¹ and Imre Csavas²

¹ ASEAN - EEC Aquaculture Development Programme (AADCP) P.O. Box 1006, Kasetstart Post Office, Bangkok 10903, Thailand

² FAO Regional Office for Asia and the Pacific Phra Atit Road, Bangkok 10200, Thailand

NEW, M.B. and I. CSAVAS. 1993. Aquafeeds in Asia - a regional overview, p.1-23. In M.B. New,A.G.J. Tacon and I. Csavas (eds.) Farm-made Aquafeeds. Proceedings of the FAO/AADCP Regional Expert Consultation on Farm-Made Aquafeeds, 14-18 December 1992, Bangkok, Thailand. FAO-RAPA/AADCP, Bangkok, Thailand,434p.

INTRODUCTION

A major purpose of the meeting at which this paper is being presented is to exchange views and experiences on economical feeding strategies and feed production. The emphasis is on farm-made feeds but the country review papers will summarize national feed ingredient availability, industrial and on-farm aquafeed manufacturing, on-farm feed formulation and feeding strategies and will provide details of information on the major institutions involved in aquafeed research and development, national feed regulations, current aquafeed problems and constraints and trends foreseen. We hope that this consultation will provide unique information on these topics, which will be published by FAO and AADCP.

We do not, in our introductory paper, intend to pre-empt the information which we expect to gain, either from the country papers or from the documents which we have invited other speakers to prepare. Rather, we aim to present a broad overview of the current scale of aquafeed production in Asia and some thoughts on future trends.

SCALE OF AQUACULTURE

Total global aquaculture production in 1990 was 15.3million tons, with a value of US$ 26.5 billion. 84 % of the global production and 78% of the value was attributable to Asia(Figure1). The average annual growth rate in aquaculture production (1980-90) is running at 7.3% in Asia and the Pacific, compared to an increase of 5.8 % for poultry meat production and 3.9% for wheat (Figure 2). Population rose by 1.8% in that decade. Because of the large quantities of molluscs and seaweeds produced in the region(Figure 3)only about 54% of the total of 12.9 million 't produced by Asian aquaculture in 1990 was finfish and a surprisingly low 4.5% crustaceans. However, whereas Asian production of finfish grew from 1.8 million t in 1975 to 7.0 million t in 1990, crustacean production leapt from 26,000 t to 576,000 t in the same period, nearly six times the rate of increase of finfish (Figures 4 and 5).

Figure 1. Aquaculture production in 1990 by volume and by value (FAO 1992)

Figure 2. Growth of population and food production in Asia and the Pacific between 1980 and 1990 (FAO/RAPA 1991)

Figure 3. Aquaculture production by major commodity groups in 1990 (FAO 1992)

Figure 4. Growth of cultured finfish production in Asia and the Pacific (Csavas 1988; FAO 1992)

Figure 5. Growth of cultured crustacean production in Asia and the Pacific (Csavas 1998;FAO 1992)

Finfish aquaculture production

While diadromous fish, mainly salmonids, form 38% of finfish aquaculture production in the rest of the world(Figure 6), in Asia the proportion is only 9%. Conversely the proportion of both freshwater and marine fish is greater in Asia than in the rest of the world. Freshwater fish dominate the finfish aquaculture production of Asia (87 %) and the growth rate of this sector is exponential. Carnivorous fish represent 7 % of Asian finfish production, compared to 50 % in the rest of the world (Figure 7). The proportion of carnivorous finfish grown in “developed” countries of the region is high (90 %) compared to their “developing” neighbours(3 %); however most of the total finish production is reared in the “developing” countries(95 %).

While most finfish feeds are produced for carnivorous fish, aquafeeds are also manufactured for some non-carnivorous species. The most important species or groups in the latter category are common carp, the tilapias and milkfish. In discussing aquaculture production in terms of aquafeed demand this paper therefore considers carnivorous finfish, shrimp, prawns, and other crustacea, together with common carp, tilapias and milkfish. Feeds are made for other types of fish but they form a very small proportion of the total aquafeed production and have not been taken into account.

Figure 6. Production of cultured finfish by major species groups in 1990 (FAO 1992)

Figure 7. Cultured carnivorous and non-carnivorous finfish production in 1990 (FAO 1992)

In 1990, of the 510,000t of carnivorous fish produced in Asia(Figure 8), 11 % consisted of salmonids. The other important species groups were catfishes (15 %),yellowtails (32 %),eels (19 %) and seabreams (14 %). Produced of the three non-carnivorous fish species groups (Figure 9) totalled 1.6 million in 1990 of which 53 % was common carp, mostly produced in China.

A glance at the geographical disparities in Asian carnivorous fish production(Figure 10) immediately highlights those countries where commercial aquafeed production is current or has potential - two of the three “developed” countries of the region (Australia and Japan, together with those “developing” countries where the total production of carnivorous fish is more than 2,500t per year-Taiwan (Province of China), India, China, Thailand, Indonesia, Republic of Korea, Hong Kong, and the Philippines, in descending order of magnitude. At the moment (Figure 8), more carnivorous fish are produced by aquaculture in the rest of the world (nearly 708,000 t) than in Asia (510,000 t).

Figure 8. Cultured carnivorous fish by major species groups in 1990 (FAO 1992)

Figure 9. Cultured non-carnivorous fish by major species groups in 1990 (FAO 1992)

Figure 10. Major producers of cultured carnivorous fish in Asia and the Pacific in 1990 (FAO 1992)

Crustacean aquaculture production

Globally, about 716,000 t of crustacea were produced by aquaculture in 1990(Figure 11). 80% of this total was produced in Asia(nearly 576,000 t), almost all of it in the developing countries of the region. China (27 % of the global total) and ASEAN (35%) mainly Indonesia (14%), the Philippines (8%) and Thailand (14%) were the major producing areas. Vietnam, Taiwan, Bangladesh, India and the Democratic Peoples Republic of Korea were also significant producers.

Figure 11. Production of cultured crustaceans by major species groups in 1990 (FAO 1992)

CURRENT AQUAFEED MARKET IN ASIA

Finish feeds

In Asia, fish feed needs centre round those for yellowtails, eels, seabreams, catfishes, salmonids and the three non-carnivorous fish species (Figure 12). Much of the feed requirements of yellowtails, seabreams, and “other” carnivorous fish are still met by moist feeds consisting of a large proportion of trash fish or low-value marine fish.

Figure 12 Aquaculture production in Asia and the Pacific in 1990 and estimates for 2000 (commercially fed species only)

We have estimated (Table 1(), using a series of assumptions (Table 2), that the Asian market for carnivorous fish feeds in 1990 was about 472,000t. By far the largest type of carnivorous finfish feed is for eels (37%), almost entirely in Taiwan and Japan. Salmonid feeds constitute 16%. Feed for catfish (13%) and for “other carnivores” (15 %) are the largest species groups, followed by yellowtails (10 %).

The remainder is feeds for seabreams (9 %). By country, Japan (54 %) and Taiwan (23 %) were the main users of feeds for carnivorous finfish in 1990. Total consumption for all ASEAN countries was estimated at less than 8%, most of it in Thailand. About 7 % was used in the Indian sub-continent.

Feeds are also made for the non-carnivorous fish listed in Table 1. It is estimated that over 554,000 t of commercial feeds were made for common carp (31 %), tilapia (29 %) and milkfish (40 %) in 1990. Of these most were for the market in Taiwan (49 %). ASEAN (25 %), mainly Indonesia and the Philippines, was also a very significant market for feeds for non-carnivorous finfish feeds.

* and Ayu

* includes Ayu (non-carnivorous but 100% commercially fed)

Crustacean feeds

Using a similar series of assumptions (Table 2) we have estimated that the Asian crustacean feed market in 1990 was about 533,000 t (Figure 12), mostly for marine shrimp (90%).

Indonesia and Thailand, each with a 26% share, were by far the largest markets for crustacean feeds in 1990 (Table 1). The ASEAN total (over 360,000t) constituted 67% of the Asian total. China (13%) and Taiwan (11 %) were the other two main centres, with the Indian sub-continent (3%) and other countries (4%) in a fledgling state. While Japan was the major user of commercial feeds for carnivorous finfish in 1990 it was a very minor market for crustacean feeds (less than 1 %).

It should be noted that our estimate of the marine shrimp feed market in 1990 (481,000 t) is about 30 % less than that of Akiyama (1992a), which was 691,000 t. His estimate of 250,000 t of shrimp feed production in China is, in our view, far too high. This is also so, to a lesser extent, for Thailand and Taiwan. On the other hand, our estimates for Indonesia and the Philippines are higher than his. Since China, Thailand and Taiwan account for 585,000 t of Akiyama's Asian total of 691,000 t, this represents the major cause of our disparity. Naturally, we believe our estimates are more realistic !

Total Asian aquafeed market in 1990

The current Asian aquafeed market for all species is estimated to total 1.6 million t. ASEAN was the major market for commercial aquafeeds in 1990 (34 %) - Indonesia (14 %), Thailand (11 %) and the Philippines (9 %) being the focal points. The commercial aquafeed market in Taiwan, at 28 % of the Asian total, was by far the largest of any country, being more than Indonesia and Thailand combined. Japan was also a huge market, at 19 % of the Asian total. Aquafeeds in the Indian sub-continent (8 %) and China (7 %) approached the level for the Philippines (9 %). Aquafeeds for carnivorous finfish (30%), non-carnivorous finfish (36 %) and crustaceans (34 %) constituted similar proportions of the total 1.6 million t market in 1990.

It should be noted that our estimate of 1.6 million t of aquafeeds in Asia in 1990 is about 1 million t less than that of Akiyama (1992b), due principally to a gross over-estimate by the latter for China (PRC). Megisson (1990) gave a global estimate of 4 million.t for aquafeed production in 1988. In our opinion this also is much greater than reality; we believe that it was less than 3 million t. Gill (1992a) stated that aquaculture accounts for 3 % of total world feed production, stated to be 610 million t in 1992. This would equate to 18.3 million t of aquafeeds, which is clearly wrong; this quantity would be sufficient to feed 100 % of all species of cultured finfish and crustaceans ! We believe that, amongst the several figures available, ours of 1.6 million t in Asia in 1990 is the most reasonable estimate and certainly the only one for which the assumptions have been published.

FUTURE PROSPECTS FOR AQUAFEEDS

What of the future: what will be the scale and nature of the Asian aquafeeds industry in the year 2000 ? Using a series of assumptions on the expansion of aquaculture production, improvements in food conversion efficiency, and the proportion of stock fed with commercial feeds, which vary by species group and location (Table 2), we have derived forecasts of the Asian aquafeed market in the year 2000 (Table 3). Aquaculture production in 1990 and forecasts for 2000 have been compared in Figures 12 and 13, while the potential aquafeed markets in 1990 and 2000 are illustrated in Figures 14 and 15.

* and Ayu

Figure 13. Aquaculture production in Asia and the Pacific in 1990 and estimates for 2000 (commercially fed species only)

Figure 14. Aquafeed markets in Asia and the Pacific in 1990 and forecasts for 2000 by species

Figure 15. Aquafeed markets for Asia and the Pacific in 1990 and forecasts for 2000 by country

Finfish feeds in the year 2000

Growth rates for farmed carnivorous fish in Asia are extremely difficult to predict because, unlike crustacea, there are so many species which are or will be produced, in varying locations, each with entirely different markets: some local, some global.

Asian production of carnivorous fish is expected to grow by over 40 % above 1990 levels to 736,000 t by 2000, while production of the three major non-carnivorous species which are partially fed by commercial feeds is forecast to increase to 2.26 million t in 2000, 41 % over 1990. The Asian aquafeed market in 2000 is expected to include 817,000 t of feeds for carnivorous fish and 886,000 t for non-carnivorous finfish. This represents increases of 73 % and 60 % respectively over 1990.

In the year 2000 the principal markets for carnivorous finfish feeds in Asia will still be Japan (51 %), and Taiwan (18 %). However, while expansion in these countries will be significant, the market for carnivorous finfish feed by 2000 will grow faster still, and is expected to more than double over 1990, in China, the Indian sub-continent and ASEAN. Eel feed as a proportion of the total feeds for carnivorous finfish is expected to decrease from 37 % in 1990 to 25 % in 2000, the demand being expected to increase by only 28,000 t in the decade. Demand for salmonid feeds will be almost static, increases in aquaculture production being compensated by improvements in food conversion ratio (FCR). Conversely, the market for commercial feeds in Asia is expected to treble by 2000 for yellowtails, to about 145,000 t, to increase by a factor of 2.5 to 152,000 t for catfishes and to double for seabreams and other carnivorous fishes.

The market for feeds for the three non-carnivorous finfish groups in Asia will be around 886,000 t by 2000. This represents a 60 % increase over 1990. The proportions of the increased market for aquafeeds for these three species groups, common carp (31 %), the tilapias (30 %) and milkfish (39 %) will remain almost identical to 1990. However, the share destined for Taiwan (42 %) will decrease. ASEAN's share is expected to increase from 25 % in 1990 to nearly 29 % in 2000. The market for non-carnivorous finfish feedsin China in the year 2000, though still a small proportion of the Asian total (10 %) is expected to be about seven times greater by weight than in 1990.

Crustacean feeds in the year 2000

This Asian market is forecast to expand by 387,000 t (73 %) over 1990. The market for fresh water prawn feeds will nearly treble to 112,000 t. 444,000 t of crustacean feeds are expected to be used in ASEAN countries in 2000, a significantly smaller share (48 %) of the total than in 1990. The Chinese market, which is expected to be over 250,000 t by 2000, will constitute a 28 % share of Asian crustacean feed production. Taiwan will retain a sizeable share (15 %) but less than Thailand (19 %) or Indonesia (18 %). The market in the Indian sub-continent will still be small (5 %).

The estimates of the crustacean feed market in the year 2000 have been based on an expansion of aquaculture production from the official statistics of 576,000 t in 1990 (Table 1) to 817,000 t in 2000 (Table 3).

Total Asian aquafeed market in the year 2000

The total market for commercial aquafeeds in Asia is expected to grow by 1.0 million t (68 %) to about 2.6 million t by the year 2000 (Table 3). Of this total, about 35 % will consist of feeds for crustacea and 34 % for non-carnivorous finfish (Figures 14 and 15). Taiwan is expected to remain the largest single aquafeed market, at over 650,000 t; however, at 25 %, this is a decreasing proportion of the Asian total. ASEAN (over 786,000 t) is expected to constitute 30 % of the total, also a significantly lower proportion than in 1990, while Japan's share remains static. China is expected to consume 15 % of Asian commercial aquafeeds by 2000 (7 % in 1990), increasing its feed production by a factor of 3.8 to nearly 400,000 t.

FARM-MADE AQUAFEEDS

As we have shown above, it is possible to estimate commercial aquafeed production in 1990 and to attempt to predict how this market may change by the year 2000 in response to anticipated increases in aquaculture production, improvements in feed conversion efficiency and increasing reliance on compounded feed. However, what we have not shown is the amount of farm-made feed made, either now or in the future.

We do not propose to attempt to do so now because there is insufficient data on which to make such estimates and predictions. We hope that the expert consultation, to which our paper is an introduction, will provide sufficient information on the utilization of farm-made aquafeeds in Asia and the Pacific to make such estimates possible.

The estimates we have made for commercial aquafeed production are based upon the current production of carnivorous finfish, crustacea and three species groups of non-carnivorous finfish, and upon our forecasts for increased aquaculture production of these groups by the year 2000. Using similar assumptions on FCR to those used in deriving our estimates for commercial aquafeed production it can be shown that, in 1990, most aquaculture production, even of these types of fish and crustaceans, was not produced through the use of commercial aquafeeds (Figure 16).

Figure 16. Volume of selected cultured species produced through the use of commercial aquafeeds and other types of feeds in Asia and the Pacific

By 2000 this may have changed significantly for carnivorous finfish and crustaceans, so that more than 50 % will be cultured through the use of commercial aquafeeds. However, even then, most common carp and tilapias will still be produced through other nutrient inputs. The same will apply to milkfish outside Taiwan.

Our calculation (Table 4) highlights our belief that, both now and in the year 2000, around 2.0-2.3 million t of carnivorous finfish, crustacea and common carp, tilapias and milkfish will be produced without the use of commercial feeds. A significant proportion is reared through the use of dry and moist farm-made feeds, ranging from simple mixtures, sometimes made into a doughballs or other forms by hand, to relatively complex pelleted feeds. The scope for greater use of farm-made feeds is immense. The need for improvements in formulation, and in production and storage technology for the small-scale farmer is similarly great and provides the challenge which we hope this meeting will emphasize.

* carnivorous finfish, selected non-carnivorous finfish (common carp, tilapias and milkfish) and crustaceans
** assuming average FCR of 2:1 for 1990 and 1.8:1 for 2000
*** natural productivity; fertilization; integrated aquaculture; farm-made feeds

LIST OF REFERENCES

Akiyama, D.M. 1992a. “Guesstimated” world production of shrimp feed. World Shrimp Farming 17.11.37.

Akiyama, D.M. 1992b. Futureconsiderations for the aquaculture feed industry, p. 5-9. In D.M. Akiyama and R.K.H. Tan (eds.) Proceedings of the Aquaculture Feed Processing and Nutrition Workshop. Thailand and Indonesia, 19-25 September 1991. American Soybean Association, Singapore.

Anon. 1990a. Feed boom forecast. Fish Farming International 17(3):1.

Anon. 1990b. CP Feed mill regains lost popularity with local stock investors. Bangkok Post, Bangkok. April 30, 1990:18.

Anon. 1990c. World feed panorama: Census 1990. Feed International, January 1990: 7-9.

Csavas, I. 1988. Shrimp farming developments in Asia, p. 63-92. In INFOFISH (eds.) Shrimp '88 Conference Proceedings. 26-28 January 1988, Bangkok, Thailand. INFOFISH, Kuala Lumpur, Malaysia.

Csavas, I. 1990a. Shrimp aquaculture development in Asia, p. 207-222. In M.B. New, H. de Saram and T. Singh (eds.) Technical and Economic Aspects of Shrimp Farming: Proceedings of the Aquatech '90 Conference, 11-14 June 1990, Kuala Lumpur, Malaysia. INFOFISH, Kuala Lumpur, Malaysia.

Csavas, I. 1990b. New developments in Asian aquaculture, p. 11-57. In Aquaculture International congress Proceedings, september 4-7 1990, Vancouver B.C., Canada.

FAO. 1992. Aquaculture Production (1985-1990). FAO Fisheries Circular No. 815, Revision 4, FIDI/C 815 Rev. 4, Statistical Tables. 206 p.

FAO/RAPA. 1991. Selected indicators of food and agriculture development in Asia-Pacific region, 1980-90. RAPA Publication 1991/18. 211 p.

Gill, C. 1992a. Focus on feed: Global perspective. Feed International, January 1992:4-6.

Gill, C. 1992b. More feed, fewer feedmillers. Feed International, January 1992:10-16.

Megisson, P.A. 1990. Future world of aquaculture. Paper presented at the Rovithai International Shrimp Seminar, July 1990, Bangkok, Thailand.

New, M.B.1989. Formulated aquaculture feeds in Asia : some thoughts on comparative economics, industrial potential, problems, and research needs in relation to the smallscale farmer, p.19-30. In Report of the Workshop on shrimp and finfish development, Johore Bahru, Malaysia, 25-29 October 1988, Report No. ASEAN/SF/89/GEN/11. ASEAN/UNDP/FAO Regional Small-Scale Coastal Fisheries Development Programme, Manila, Philippines.

New, M.B. 1990. Compound feedstuffs for shrimp culture, p. 79-118. In M.B. New, H. de Saram and T. Singh (eds.) Technical and Economic Aspects of Shrimp Farming: Proceedings of the Aquatech '90Conference, 11-14 June 1990, Kuala Lumpur, Malaysia. INFOFISH, Kuala Lumpur, Malaysia.

New, M.B. 1991. Turn of the millennium aquaculture: navigating troubled waters or riding the crest of the wave? World Aquaculture 22(3): 28-49.

New, M.B. and U.N. Wijkstrom. 1990. Feed for thought: some observations on aquaculture feed production in Asia. World Aquaculture 21(1): 17-19 and 22-23.

Rosenberry, R. (ed.). 1990. World Shrimp Farming 1989. Aquaculture Digest 28 p.

Rosenberry, R. (ed.). 1991. World Shrimp Farming 1990. Aquaculture Digest 40 p.

Rosenberry, R. (ed.). 1992. World Shrimp Farming 1991. Aquaculture Digest 44 p.

shuau. S.Y. and I. C. Liao. 1989. Structure and outlook for the fish feed industry in Taiwan, p. 142- 159. In Proceedings of the 29th Annual Conference of the International Association of Fish Meal Manufacturers, 20-24 November 1989, Hong Kong.

Wijkstrom, U.N. and M.B. New. 1989. Fish for feed: a help or a hindrance to aquaculture in 2000 ? Infofish International 6/89:48-52.

Yamamoto, T. 1992. Development of aquaculture in Japan within her fisheries. Resource Paper SEM-18-92 presented at the APO Seminar on Aquaculture, 25 August to 4 September 1992, Tokyo, Japan.

SUPPLEMENTARY FEEDING IN SEMI-INTENSIVE AQUACULTURE SYSTEMS
Sena S. De Silva

Faculty of Aquatic Sciences, Deakin University, P.O. Box 423, Warrnambool, Victoria 3280, Australia

DE SILVA, S.S. 1993. Supplementary feeding in semi-intensive aquaculture systems, p.24-60. In M.B. New, A.G.J. Tacon and I. Csavas (eds.) Farm-made aquafeeds. Proceedings of the FAO/AADCP Regional Expert Consultation on Farm-Made Aquafeeds, 14-18 December 1992, Bangkok, Thailand. FAO-RAPA/AADCP, Bangkok, Thailand, 434p.

INTRODUCTION

A ‘supplement’ is defined as a “thing added to supply deficiencies”. In an aquacultural nutritional context it will be tantamount to the supply of feed to meet one or more nutrient deficiencies of the system for the well being of the stock. It is a truism that the supply of deficiencies in relation to nutrition of the stock in semi-intensive aquaculture practices is one of the least addressed aspects of the industry. Therefore, it continues to remain a rather grey area in our Knowledge. The aquaculture industry in Asia, in particular, is embarking on an increasing phase of intensification, driven by a multitude of economic forces amongst which are an increasing need for even the small-scale farmer to transform from subsistence aquaculture to an income generating enterprise. In the light of increasing competition for primary resources, such as water and land, such economic needs become even more demanding. This increasing intensification of the industry at large however, has to take place within the frame work of certain potential global constraints on the industry such as the “fishmeal trap” (Wijkstrom and New 1989). In such a scenario issues related to supplementary feeding are likely to come to the forefront.

Supplementary feeding in semi-intensive aquaculture has to be considered in the perspective of the past-and future trends of the industry. Moreover, the treatment will be meaningful only if supplementary feeding is considered in the context of other strategies that are prevalent in Asian aquaculture which directly or indirectly reduce nutritional deficiencies. In this paper an endeavour is made to address these issues and in particular in the context of the envisaged growth of the industry.

GENERAL TRENDS IN ASIAN AQUACULTURE

The development of semi-intensive aquaculture (to be defined later) in Asia and the Pacific, and in tropical Asia were recently reviewed by Csavas (1989) and Bailey and Skladany (1991) respectively. A majority of the recent reviews however, have tended to highlight the growth of shrimp and carnivorous finfish culture, particularly from the point-of-view of the aquafeed industry and its potential problems (Wijkstrom and New 1989; New 1991). In this presentation an attempt will be made to focus on other aspects of the aquaculture industry in Asia which are of relevance to supplemental feeding.

The backbone of Asian aquaculture is finfish. Finfish contribute over 50% to the total aquaculture production in Asia (Table 1; Figure 1), and has been steadily increasing over the last five years. In addition, finfish production in Asia accounts for more than 80% of the world's total production of cultured finfish (Figure 2), and it ranks second in relative terms only to the cultured seaweed production of the world. In the latter case the cultured aquatic plant production of Asia contributes nearly 98% to the world's total.

The mainstay of Asian aquaculture historically, as well as at present and in the foreseeable future, is non-carnivorous finfish (Table 2; Figure 3). The noncarnivorous finfish production in Asia has been increasing steadily, and presently (1989) contributes 90% to the total, cultured non-carnivorous finfish production of the world (Figure 4). On the other hand, the relative contribution of cultured carnivorous finfish production in Asia has been on the decline. The reason for this decline was addressed by Csavas (1990).

Csavas (1990) also focused on some of the key features of the Asian aquaculture industry, some of which are directly relevant to the present topic. Those concerning finfish and shrimp culture, the two practices in which supplementary feed and feeding are likely to have a direct bearing, are as follows:

• very little hard data exist regarding the level of intensity of Asian aquaculture. Nevertheless, the typical freshwater culture systems are beyond doubt extensive or semi-intensive polycultural systems with some fertilization and supplementary feeding. This is reflected in the major species cultured and the average value of the product. For example in 1988 it was US$ 1.18/kg in developing countries of Asia, as opposed to US$ 3.21/kg elsewhere;

• by and large freshwater aquaculture is a small-scale activity in Asia. A “small-scale activity” is not easily definable either. Perhaps in the present context small-scale can be considered as non-corporate culture activities which are single family owned and managed;

• the development of non-carnivorous finfish production still shows an exponential growth, and has continued to exceed that of carnivorous species;

* % Asia¹ = Asian contribution of the particular commodity to the world production of thatcommodity;
% Asia² = contribution of the commodity to total Asian aquacultureproduction

Source: FAO (1992)

• in the developing countries of the region traditional extensive or semi-intensive pond culture dominates finfish and shrimp aquaculture. The general tendency over the past decade was to intensify these to shift from an uncontrolled “polyculture” to monoculture of shrimp by replacing trapping of wild juveniles with stocking;

• in coastal areas the dominant function of aquaculture is income generation, the production of cash-crops sold on distant markets;

• in many cases even semi-intensive shrimp farms have shifted to use factory made feeds, and this trend is likely to increase in the ensuing years.

* % Asia¹- Asian contribution to the particular group
% Asia²- Asian contribution of the commodity to total world finfish production

Source:FAO (1992)

Figure 1. Share of major commodity groups in the aquaculture production of Asia (based on FAO 1992)

Figure 2. Contribution of Asia to the overall aquaculture production of the world (based on FAO 1992)

Figure 3. Trends in carnivorous and non-carnivorous finfish production in Asia (A) and the world (W) (based on FAO 1992)

Figure 4. Contribution of Asia (A) to the world (W) production of carnivorous and non-carnivorous finfish and to the world's total finfish production

SEMI-INTENSIVE AQUACULTURE: A SEARCH FOR A DEFINITION

Perhaps semi-intensive aquaculture will continue to remain a term and a practice that will never be defined precisely. On the other hand, some might argue that such a definition is not strictly needed, and that perhaps broadly accepted guidelines are sufficient.

One of the best available guidelines depicting semi-aquaculture practice was presented schematically (Figure 5) by Tacon (1988). Accordingly, semi-intensive aquaculture would constitute a significant quantum in the continuum between extensive and intensive aquaculture. Within this broad spectrum the semi-intensive practices could differ from each other in the degree of management, provision of feed, the stocking density and yields. For example Csavas (1990) suggested that in respect of Asian shrimp culture such practices could result in yields between 1,500-5,000 kg/ha/yr. He related the scale of intensity with improving external inputs and increasing experience in production techniques and farm management. Yields ranging from 5,890-14,000 kg/ha/yr have been reported recently from the semi-intensive composite culture practices of Indian major carps from Andhra Pradesh, India (Veerinaet al. in preparation). The upper range of this production level far exceeds that from most intensive culture practices in temperate waters.

On the other hand, Schroeder (1978) reported yields of up to 30 kg fish/ ha/day (common carp and tilapia) through intensive manuring, and Zhuet al. (1990) reported an increased net yield of 10.2 kg/ha/day in Chinese integrated farm ponds as a result of manure application. The purpose of focusing on these data is not to emphasise the importance of fertilization in semi-intensive aquaculture but to draw attention to the fact that perhaps yields are one of the weakest criteria in determining the degree of intensity of a culture practice. Secondly, and more importantly, it raises the question whether fertilization needs to be considered as a feed supplement.

The use of manure in fish farming is an old Asian practice, and has been adequately reviewed by Wohlfarth and Schroeder (1979) and Wohlfarth and Hulata (1987). The influence of manure on aquaculture is schematically depicted in Figure 6. Manure is thought to act in supplying a deficiency; in this case primarily nutrients in the culture system, supplementation of which enhances the growth of natural food organisms. Hence, I am inclined to suggest that manure be considered as an indirect feed supplementation; perhaps a supplementation that is least competitive with the other forms of direct or indirect nutrient supplementation that are used in semi-intensive aquaculture. On the other hand, Jana and Chakraborty (1990) suggested that a better approach to carp culture would be to introduce live plankton rather than direct manuring. Here again, the question of wide-scale adoption of such a practice by the small-scale farmer arises.

Figure 5. Schematic representation of the range of aquaculture practices in relation to the inputs (modified after Tacon 1988)

Figure 6. Schematic depiction of the influence of application of manuare in comparison to other inputs on yield in semi-intensive aquaculture

SUPPLEMENTARY FEEDS

Interactions with natural food supplies

Supplementing the naturally available food in a culture system is the most simplistic functional interpretation of supplementary feeds. In reality, the role of supplementary feeds in the nutrition of the cultured organism is much more intricate and is closely interlinked to the dynamics of the pond/system. The intricacy is analogous to the interpretation of interactions between species in a simple aquatic community explained in terms of a food chain as opposed to a food web.

The natural food organisms in a pond are a rich source of protein, often the mix of organisms containing 50-60% of protein (dry matter basis). According to Albrecht and Breitsprecher (1969) the mean protein, carbohydrate and lipid content of fish food organisms in ponds were 52.1, 27.3 and 7.7% respectively, and the calorific value ranged from 1.6-5.7 kcal/g dry matter, with a mean of 3.9 kcal/g. Proximate composition of some of the common individual food organisms is given in Table 3. Thus it is evident that the nutritional value of natural food organisms is adequate to meet the nutritional requirements of most of the semi-intensively cultured organisms in Asia, which according to Prus (1970) range from 28-35% of protein and 96 mg/kcal protein to energy ratio. It has also been shown, for example, that the food of tilapia (Oreochromis mossambicus), from quasi-natural impoundments, even when it contains a high proportion of detritus, is well digested (De Silva et al. 1984).

In semi-intensive culture there are complex interactions between the natural food organisms and supplementary feeding practice(s). The possible qualitative interactions in a well prepared pond which is to be subjected to supplementary feeding is schematically depicted, in Figure 7. The above simplification does not take into account the possible changes in the feeding habits of the cultured organisms. The trophic niches of fish species are known to change with natural food productivity, the standing crop of fish and the plasticity of the feeding behaviour. Such changes are documented for a number of species (see Werner and Hall 1976; De Silva et al. 1984; Odum 1970). Furthermore, in polyculture systems fish species are more capable of changing their diet readily (Weatherly 1963). Also Spataru et al. (1980), in their studies on common carp, have shown that supplementary feeding can affect feeding habits and food selection, the fish tending to select a narrower range of food items when on supplementary feeding regimes.

There is some experimental evidence to show that, as the difference between standing crop (SC) of the cultured organism(s) and the critical standing crop (CSC) of natural food organisms increases, the deficit in natural protein supply also increases. Under such conditions supplementation with “protein rich” diets becomes necessary to maintain growth and production.

* since the values are averages of figures collected from different sources they do notnecessarily add up to 100%

Source: modified from Hepher (1988)

Figure 7. Schematic representation of the qualitative changes that occur in a semi-intensive culture pond. Also indicated is the potential utility of “supplementary feed” strategies.

The changes in protein requirements in supplemental feed(s) in relation to standing crop are depicted in Figure 8. The increase in supplementary protein requirements in relation to increasing SC is gradual. For example Hepher et al. (1971) and Hepher (1975) have shown that in common carp under monoculture, a pelleted feed of sorghum (9% protein), or diets of 22.5% and 27.5% protein did not influence the growth until a SC of 800kg/ha was reached. However, when the SC increased further the 9% protein diet was not effective; the 22.5% diet became ineffective when the SC reached 1,400kg/ha. Hanley (1991) on the other hand, reported that in Nile tilapia (Oreochromis niloticus) the dietary lipid level did not influence the growth rate or the food conversion ratio (FCR) and recommended that diets for tilapia have to be compounded on the basis of protein rather than energy considerations.

Unlike in an intensive culture system the utilization efficiency of a supplementary feed in semi-intensive culture is dependent on the natural food supply, in addition to other factors such as standing crop and body weight. The latter effects are common to all systems. This phenomenon is best illustrated by the work of Hepher (1978) on semi-intensive culture of common carp (Figure 9). It is apparent from this figure that at a SC of about 500kg/ha, sorghum as a supplementary diet was inadequate and consequently the FCR increased sharply, as opposed to those maintained on a pelleted feed of 25% protein. As pointed out by Hepher (1978) this sharp increase in the FCR is critical from an economic point of view.

Hepher (1988) suggested that in such circumstances one of three options is open to the farmer:

• improve the diet by adding the missing nutrients. This results in an increase in the cost, which may not necessarily be paid off by corresponding increase in yield;

• reduce the SC. This results in a decrease in the yield; and

• increase productivity by manuring.

The above observations indicate that supplementary feeding in semi-intensive aquaculture has many facets; to be economical the farmer has to be alert about the type and quantity of feed to be used. It would be apparent therefore, that feed supplementation in semi-intensive aquaculture, if properly carried out to gain the maximum return, is not an ad hoc process. In effect, because of the interactions between the natural food supply and supplementation it is more complex than providing the stock with a nutritionally wholesome diet under intensive culture conditions.

Figure 8. Schematic depiction of changes in the natural food organisms and fish yields, in relation to standing crop of the cultured organism and the ensuing protein needs of the supplemental feed (s)

Figure 9. Supplementary feed conversion ratio at different standing crops of common carp in ponds (after Hepher 1978)

Supplementary feeds

Supplementary feeds used in semi-intensive aquaculture in Asia range from kitchen wastes on the one extreme to almost nutritionally complete compounded feeds, such as those used typically in shrimp culture operations. Supplementary feeds can be in the form of single ingredients, simple mixtures of powdered ingredients or ingredients compounded into a dough or pellet form.

In Asia, where aquaculture is mostly a rural occupation, the selection and utilization of supplementary feeds is linked to other agricultural activities in that particular region. This is so because the bulk of the ingredients used as supplementary feeds consists of agricultural by-products, or by-products of the animal husbandry industry. Fish are not capable of efficiently utilizing most of these ingredients, in particular the grasses Table 4 gives a list of commonly used feed types and the FCR values reported in respect of each. The relatively inefficient utilization of grasses and other cellulose-rich materials is to be expected because fish do not possess the enzyme cellulase.

Information available in the literature on the types and extent of usage of different supplementary feeds on farms is scanty. Over the last decade an almost explosive growth of a semi-intensive, polyculture practice of Indian major carps has been witnessed in the south eastern state of Andhra Pradesh in India. Yields ranging from 5,890-14,000 kg/ha/yr have been reported (Veerina et al. in preparation). In a survey covering 189 farms in four districts of the State,it was found that 9 major ingredients and 7 feed types are used by the farmers. On average 27,000 kg/ha/yr of feed was utilized as supplementary feed (Tables 5 and 6; Figure 10). This data provides an indication of the diversity of supplemental feeds that are utilized in semi-intensive finfish culture, even within a single region, and also variation in farmer preference for particular ingredients and or ingredient mixes. The scientific rationale behind the adoption of some of the supplemental feed practices is not immediately obvious, however; for example the use of common salt in the feed mix by some Andhra Pradesh farmers. Our knowledge on use of supplementary feeds at farmer level per se is very scanty. Until the recent survey by Veerina et al. (in preparation) there had been no documentation that common salt was used as an additive to supplementary feed(s). The above is just one example; perhaps there are many more which the scientific community is unaware of and the scientific basis of usage unexplained. This further exemplifies the need for a deviation from the traditional nutritional research based on dose-response approach, as advocated by De Silva and Davy (1992).

* from Ling (1967)
** food quotients for herbivorous fish species in China (Yang et al. 1985)

Source: Tacon (1988)

Source: Veerina et al. (in preparation)

* CSM = cottonseed meal; DOB = deoiled (rice) bran; DOC = deoiled (groundnut) cake;GNC = groundnut cake; MIL = millets; NIL = none; RB = rice bran;SBM = Soybean meal; SFM = Sunflower meal; SOR = sorghum.

Source: Veerina et al. (in preparation)

Over the last decade or so there had been a shift in emphasis in respect of supplementary feeds; the term supplementary feeds and supplementary ingredients being almost completely discarded from the fish nutrition jargon. No longer is there an emphasis on supplementary ingredients per se but on the utilization of such ingredients in compounded feeds in relation to inclusion levels as a substitute for fishmeal. Such ingredients are now commonly referred to as non-conventional feedstuffs, a term which was originally used in respect of the animal husbandry industry. Comprehensive accounts on such feed resources are given by Devendra (1985), Tacon and Jackson (1985), Pantastico (1988) and Wee (1991). The proximate composition of feed ingredients, their suitability for incorporation into practical compounded feeds and other relevant information are dealt with in these reviews. Table 7 summarises the results of some of the experimental work based on the utilization of feed ingredients in fish feeds for tropical species. It is apparent that most oilseed meals and grains, when incorporated into compounded feeds at levels up to about 35% of dietary protein, resulted in achieving more than 80% of the growth reported for control diets where fishmeal was used as the primary source of protein.

The above mentioned findings are unfortunately generally confined to laboratory experiments or small-scale pilot studies. Due to a multitude of factors, these hardly get translated into practical feeds (De Silva and Davy, 1992). Nevertheless, these findings have given some indirect indications of the effectiveness of use of agricultural by-products in compound feeds and/or in simple mix forms, as are commonly utilized by the rural farmer.

Figure 10. Extent of use of different supplementary feed combinations in semi-intensive carp culture practices in four districts in Andhra Pradesh, India (Veerina et al., in preparation)

Source: modified from Wee (1991)

DISPENSATION OF SUPPLEMENTARY FEEDS

In a great majority of semi-intensive culture practices in Asia, particularly in small-scale, single-owner rural operations, the most common method of dispensation of supplementary feeds is in powder form. There is very little scientific information on the efficacy of this form of feeding. Swingle (1958) reported that yields of channel catfish fed on a dry mixture of ingredients and in pelleted form were 1,247 kg/ha (FCR 3.3:1) and 2,648 kg/ha (FCR 1.6:1) respectively. To the author's knowledge, comparable studies from Asian aquaculture are lacking and this focuses on the problem of relative dearth of on-farm based nutritional research in Asian aquaculture.

The reasons for the decreased ineffectiveness of this form of feeding are manifold; significant wastage of food, individual fish being unable to ingest sufficient quantities of each of the constituent ingredients and thereby not obtaining a nutritionally favourable diet as well as the digestibility of the individual ingredients being different when ingested on their own. However, farmers continue to use this practice, sometimes in modified form. Farmers continue to provide supplemental feed, in one form or the other, only because they experience an obvious advantage through the practice (Table 8). This applies not only to the practice per se but, as pointed out earlier, to the use of additives in feeds of which the scientific rationale is not that obvious. It is conceivable that the direct influence of the feed per se on growth and yield are marginal. Nonetheless, feed waste and the undigested material enhances natural food production in the system which results in increased yields over and above what would have been achieved without supplementation.

In semi-intensive culture practice in Andhra Pradesh, India, one aspect of the culture practice remains uniform; the method of feeding in the grow-out ponds. Feed, a mixture of ingredients in powdery form (Figure 10) is presented in perforated polythene bags suspended in the pond at a number of points. Input-output analysis of the farms have shown that the feed ingredients, such as groundnut cake, cottonseed meal, de-oiled rice bran, sunflower seed and mineral mix, apparently provided as direct supplemental feed have a positive but moderate influence on yields (Veerina et al. in preparation). A few random samples of fish from these farms in 1989 were found to have only animal and plant material in the stomachs. This information, though grossly inadequate, reinforces the need for more scientific investigation to determine the extent to which supplementary feeds in most aquaculture practices contribute directly to the nutrition of cultured organisms. It is expected that the findings from such studies will lead to the next step of improving such feeds used by farmers, using the simplest of technologies available to them.

In a general sense, the feeding practices in semi-intensive aquaculture in a rural farming scenario are as diverse as the feeds used. In most semi-intensive aquaculture practices, feeding is rather ad hoc and feed in a powder form is broadcast manually, once or twice a day. Use of feeding trays or comparable devices, such as use of perforated bags with feed, is becoming increasingly popular. The use of the latter practices, though less labour intensive and/or time saving, is bound to degrade feed quality, through continued leaching of nutrients. There is also the possibility of establishment of prominent hierarchical effects which could result in the prevention of access to the food source to a significant proportion of the stock. Perhaps the greatest challenge lies in the development of efficient ways of dispersing feed in shrimp culture; to make feeding more effective whilst preserving the quality of the feed for a longer period with these rather sluggish feeders.

* integrated system (No. of ducks stocked = 15)
** integrated system (No. of ducks stocked = 30)

The quantity of feed required per unit body weight (or % body weight) for maintenance and growth decreases as the fish increase in size. However, the relative feed required (g per fish) increases due to increase in weight. Under semi-intensive culture conditions therefore, supplementary feed requirements have to be balanced, taking into account increasing body weight and natural food availability. Based on pond studies, Marek (quoted by Hepher 1988) developed a feeding table for common carp culture in Israel (Table 9). Such carefully planned feeding strategies are rarely used in the rest of Asia, apart from feeding guidelines that have been developed for different species by various researchers. Some of these guidelines are summarised by Tacon (1988).

There is evidence to show that, under experimental conditions or in pilot scale trials, most cultured fish species tend to perform better when the frequency of feeding is increased, there being an optimum feeding frequency (Chiu 1989; Tuang and Shiau 1991). However, there is no data available on the effect of feeding frequency for on-farm semi-intensive culture systems in Asia. Perhaps, this is another area of investigation which should be given priority, based on which possible means of obtaining higher yields can be recommended to small-scale farmers.

More recently the extension of the hypothesis developed by De Silva (1985), that the daily provision of high protein diets is wasteful, has been tested for semi-intensive polyculture of two Indian major carps (catla and rohu) and common carp. Nandeesha et al (in preparation) found that 15.3% of protein input in the feeds and 10-20% of overall feed costs can be saved by adopting one of three mixed feeding schedules, based on rice bran and rice bran: mustard oilcake mix, supplemental feeds commonly used in Indian carp culture. Obviously, there is more to be done and understood in adopting mixed feeding schedules in semi-intensive culture. There are also practical implications, such as the storage of two or more types of feeds and how to dispense the correct feed according to strict schedules, which tantamounts to increasing farmers' awareness of the practice. Adoption of mixed feeding schedules, with feeds already in use, as a feed cost saving strategy, is perhaps relatively more easily extendible than new types of feeds.

Generally, there is a reluctance in the scientific community to readily accept new concepts; more so if such concepts originate from the third world (Gaillard 1991). It may be some time, therefore, before mixed feeding schedules become common practice in semi-intensive culture in Asia.

* based on four pelleted diets of increasing protein: 12% (I), 18% (II), 25% (III), and 30% (IV).

Source: Marek (1975), cited in Hepher (1989)

COMPOUNDED FEEDS AS SUPPLEMENTATION

The use of compounded diets as supplementary feeds is becoming increasingly popular in Asian aquaculture (Csavas 1990). For example, the Asian shrimp industry is almost totally dependent on such feeds. In the light of increasing competition for primary resources such as land and water, Asian aquaculture, particularly semi-intensive finfish culture which essentially caters for lower and middle income groups, can only develop (or may only survive) in the ensuing years if such practices become more profitable and productive. One of the main keys to profitability even in semi-intensive aquaculture will be a reduction in feed cost, which is generally estimated to exceed 50% of operating costs.

The basic nutrient requirements of the major species cultured in the world are fairly well known (Wilson 1991); at least to the extent that permits reasonably good diets to be formulated. New (1987) and Tacon (1988) gave some of the compounded feeds used in the culture of tropical species, semi-intensively. However, most of the feed formulations cited were experimental (test diets) and the overall proximate composition of the feeds were not easily obtainable. I will make no attempt to repeat this information, but I do consider conceptually the validity of usage of test diets in semi-intensive aquaculture. Some of the common features recognisable in these diets are:

• most of them contain a significant proportion of fishmeal, as a protein source; and

• the formulations have been made in order to meet the complete nutrient requirements of the species in question.

The need to incorporate micro-nutrients such as vitamins in supplementary feeds is also questionable. Dickson (1987) demonstrated that vitamins incorporated into supplemental diets for tilapia culture had no influence on growth and reproduction. Comparable observations have been reported on survival and growth in juvenile Penaeus monodon in modified extensive culture systems (Trino et al. 1992) and for P. vannamei (Castille and Lawrence 1989). In a similar vein Castell et al. (1988) commented that it was unlikely that severe essential fatty acid (EFA) deficiencies would be encountered in aquaculture practice; perhaps the probability is further reduced in semi-intensive culture in tropical Asia.

In semi-intensive aquaculture, where external feed input is expected to supplement natural food production, the use of a nutritionally balanced feed seems to be questionable; a waste of resources, as well as an economically unsound practice. In addition, there is increasing evidence to show that there are significant differences between the dietary protein requirements for optimal growth as opposed to the economically optimal protein level; the concept seems to be applicable at least to two groups of fish which are commonly cultured in Asia (De Silva et al. 1989; De Silva and Gunasekera 1991). It follows, therefore that formulation of compounded diets for semi-intensive aquaculture has to focus on the fact that the diets need not necessarily be nutritionally complete and, to be economical, that the protein level in the diets could be reduced even further.

Recently, Green (1992) demonstrated that substitution of organic manure (chicken litter - 1,000 kg total solids for a 0.1 ha earthen pond) for a 23% protein pelleted feed in monosex culture of Oreochromis niloticus resulted in better net returns. The total production cost for feed only, manure followed by feed, and manure plus simultaneous feeding, resulted in yields of 5,305, 4,794, and 4,351 kg/ha and costs of US$ 5,336/ha, US$ 4,645/ha and US$ 3,471/ha respectively. This brings me back to the point that I raised earlier; should not manure be considered as a feed input, and accordingly some of the ensuing research priorities be directed to such studies?

Perhaps, for most countries and/or regions the whole emphasis on formulations and manufacture has to be looked at afresh. Rural aquaculture in Asia is such that, in order to make any impact, new strategies may have to be considered. A schematic representation of a possible strategy is given in Figure 11.

A LOOK INTO THE FUTURE

Globally, the potential resource limitation of fishmeal on further expansion of the aquaculture industry has been focused upon by Wijkstrom and New (1989). According to New (1991), Asian aquaculture in 1988 utilized about 53.5% of global aquafeed production which was almost equally distributed between the production of carnivorous and non-carnivorous finfish and shrimp (Table 10). The aquafeed industry has grown very rapidly over the last five years. For example, the feed industry in the ASEAN alone had only 5 feed mills producing 27,000 t of feed in 1985, but by 1988 the number of feed mills had increased to 53, producing 264,000 t of feed (Boonyaratapalin and Akiyama 1989). New (1991) estimated that the Asia-Pacific shrimp industry will require nearly 1.1 million t of feed by the year 2000 (to produce 1.084 million t of shrimp, of which 52% will be produced through intensive culture). Csavas (1989), based on a predicted increase of carnivorous fish production to 800,000 t and shrimp production to 1.2-1.3 million t, estimated that the actual fishmeal requirements of the industry by year 2000 would be 375,000 t, approximately a doubling of the present usage.

Figure 11. Schematic representation of the steps involved in the formulation of economic compounded diets for semi-intensive aquaculture

*
(a) - feed utilization as a percent of Asian total
(b) - feed utilization as a percent of global total

Source: New (1991)

Most predictions on fish meal requirements for the growing aquaculture industry are based on the known FCR values of cultured species. The high feed cost for aquatic organisms is attributed to the relatively high protein requirements of fish. However, this is being increasingly disputed now (Bowen 1987; Lovell 1989). Apart from this, there is increasing controversy regarding the “real requirements” of even some of the most studied species such as the rainbow trout (Murai 1992). It is also interesting that recent work on the optimum dietary protein requirements of the bighead carp Aristichthys nobilis, (Santiago and Reyes 1991) has shown that it is closer to the economically optimal dietary protein level as predicted by De Silva and Gunasekera (1991). In effect therefore, there is increasing information being gathered which indirectly shows that we have been wasting dietary protein, wittingly or unwittingly. There is a trend towards more stringent legislation being introduced in developed countries with regard to fish feed specifications which primarily originated in an attempt to curtail environmental degradation from aquaculture activities. Whatever the reasons for the introduction of such legislation, these new feed specifications entail a reduction (Table 11) of the protein level in the feeds (amongst other components) without compromising growth and production.

*
 % dry matter;
 + maximum

Source: after Kiaerskou (1991)