|Fish size/age||Feed form||Protein (min)||Lipid (min)||Fibre (max)||Moisture (max)|
|1 to 4 cm||powder||40||10||8||12|
|4 cm to 1 month||expanded||32||4||8||12|
|1 to 3 months||expanded||30||4||8||12|
|3 months to market||expanded||25||4||8||12|
Fresh feed materials
This refers to feed materials that can be used without the need to process them mechanically. These include poultry viscera, fish gills and viscera, kitchen refuse and bread. These feed materials can be directly dumped into ponds upon arrival at the farms. The nutritional values of two fresh feed materials are presented in Table 3. Poultry viscera have been regarded as one of the most effective feeds for grower catfish. On a dry matter basis, this ingredient contains 52.9% protein and 42.4% lipid (New 1987). Besides being nutritious, poultry viscera are highly palatable to catfish and naturally float on the water surface. Catfish fed on poultry viscera are fatty with shiny yellowish-skin, which is in high demand. The demand for poultry viscera by catfish farmers is quite competitive. Larger farms with capital resources gain access to consistent supplies of poultry viscera from chicken slaughterhouses year-round by winning contracts at auction. Small farms, which need smaller quantities of poultry viscera either collect them from poultry stalls in the market or buy from larger farms. The price of poultry viscera ranges from US$ 0.02-0.20/kg, depending upon the bargain struck. The feed conversion ratio for poultry viscera is approximately 4:1 which means that the feed cost could be less than US$ 0.80/kg of catfish produced.
|Feed materials||Protein (%)||Lipid (%)||NFE (%)||Ash (%)||Moisture (%)|
New (1992) defined farm-made feeds as small-scale feed manufacture encompassing everything from simple hand-formed doughballs to small feed production units. Farm-made feed is an alternative to commercial feeds. The majority of farm-made feeds are prepared by catfish farmers who began their operations prior to the availability of commercial feeds in 1980. These farmers have good experience in processing trash fish-rice bran diets. There is no standard composition of current farm-made feeds for catfish. They may be composed of a single c r many sources of raw materials processed by cooking or grinding to some extent. Table 4 shows the composition and nutritional value of some farm-made feeds once used or still in use by farmers. The diversity of the composition of farm-made feeds is dictated by the availability of the individual raw materials. The formulation in farm-made feeds may vary, depending on the amount of each raw material delivered to the farm each day.
From the farmers' point of view, the quality and proportion of feed is not so important as the amount required to feed the ponds. The success of certain farms in using farm-made feeds is usually determined by the availability and cost of raw materials.
|Farm-made feeds||Protein (%)||Lipid (%)||NFE (%)||Ash %||Moisture %|
|Trash fish - rice bran (8:2)||15.6||5.6||10.9||5.4||62.5|
|Trash fish - broken rice (6:4)||13.8||2.6||30.6||3.3||49.7|
|Soybean cake - chicken bone - noodle waste||12.2||0.1||n.a.||n.a.||48.7|
Serm farm in Minburi, on the outskirts of Bangkok, is an example of a successful clarias catfish farm which depends mainly on farm - made feeds. The farm is over 30 years old and has 10 ponds for catfish ranging from 0.16-1.6 ha. Moist feed is used in the farm. The ingredients are moist tofu waste, chicken bones, and instant noodle wastes or green mustard waste (Table 5). Chicken bones are bought from local chicken slaughterhouses. Instant noodle wastes and green mustard waste are purchased on a daily basis from the processing plants in the region. The supplies are brought to the farm and dumped on a wooden feed preparation platform (6m × 4m), which has been built over the feed processing machinery. The ingredients are roughly mixed by shovel on the platform and swept down into a first mincer through a hole in the platform. The minced material from the first mincer is then dropped into a second mincer, installed just below the first one, to be minced again. The mincer die is 2.5 cm thick with a 1.9 cm die-hole size. The two mincers are powered through two sprockets connected by a single drive chain to a 40 HP motor. When only one mincer is needed, as the fish grow bigger, the screw, knife and die plate of the lower mincer is dismantled. A polypropylene tube is tied to the mincer at the feed discharge point, thus making an extension which prevents the minced feed falling into the second mincer; it is also used to guide the feed into a container.
|Feed ingredient||Cost (US$/t)||Inclusion rate (%)||Notes|
|Moist tofu waste||40||variable||Mixture of ingredients was minced in 40 HP mincer to form moist feed|
|Chicken bone||200||variable||with 50% moisture content. Protein and lipid content was 12.2% and|
|Instant noodle waste**||120||variable||1.4% respectively. Average feeding rate was 1.25 t/ha/day. FCR was 3.6:1|
* harvested as 200-300 g fish, 4-6 months after stocking. Total yield was 100 t in a 1.6 ha pond(62.5 t/ha). The fish were sold at a farm-gate value of US$
** sometimes substitued by green mustard waste
The nutritional content of the sampled feed from Serm Farm is presented in Table 5. The proportion of the ingredients in the feeds changes daily, depending on the ingredient supplies. The feeds are usually not processed until soybean cake, chicken bone and instant noodle wastes are all available. Lack of any of these ingredients will result in the postponement of feed processing. This results in deterioration in quality of the ingredient(s) which arrived early. The amount of feed processed is dictated by the amount of ingredients available, not the daily requirement of the fish. For this reason, more feed than is needed is sometimes processed. The excess amount of feed is piled up on a plastic sheet on the pond bank for use on the following day. The feed is distributed into the ponds by shovel. Feeding rate depends on visual observations, grow-out ponds being fed once per day at 15.00-17.00 hours. The fish are harvested by size grading about 4 months after the stocking of 2-3 cm fry at the rate of 31 fry/m². One of Mr. Serm's 1.6 ha hybrid clarias catfish ponds had a total yield of 100 t at the latest harvest (62.5 t/ha). According to his data, the calculated FCR for his farm-made feed was 3.6:1. He estimated unit cost of his feed to be US$ 0.12-0.24/kg.
Kantho Farm in Chachoengsao once depended on its own farm-made compound feed for Clarias batrachus. During 1985-1989 the 9.6 ha farm, with 10 ponds of catfish, was run by Mr. Um Kantho. With assistance from his son, Dr. Uthai Kantho, who is a professor in Animal Nutrition in Kasetsart University, Mr. Um was able to produce floating catfish pellets from simple farm machinery. The composition of the feed and its nutritional value are given in Table 6.
|Fish meal (60% protein)||15|
|Freshwater fish oil||2|
* plus vitamin mix, as directed
The ingredients were locally bought. Corn and cassava chips need to be finely ground in a cyclone hammer mill driven by a 20 HP motor. The hammer mill has a grinding capacity of 200 kg/hr. through a 1 mm screen. The other ingredients were claimed to be fine enough, needing no further grinding. Each ingredient was weighed separately and mixed in a horizontal mixer with a capacity of 100 kg/batch. After 15 minutes of mixing, 30% of water and the required amount of oil were added before another 10-15 minutes of mixing. The wet mixture was then extruded through a mincer (No. 52) powered by a 5 HP motor to form noodles. The noodles were immediately cut by a knife attached to the shaft of the mincer outside the machine. The length of the pellets could be selected by adjusting the distance between the die and the knife blade. The pellets were sun dried on a 400 m² concrete surface for one sunny day and bagged for further use.
Pellets made in this way were able to float for 5 minutes, which was sufficient for catfish to feed on them. Mr. Um showed that three factors are necessary for the production of floating pellets:
low density ingredients such as cassava chips, Leuceana leaf meal and rice bran must be used in the formula;
ingredients must be finely ground to produce rigid compaction of the pellets, slowing down the penetration of water;
a thick die plate is important to allow mixtures to stay in the mincer longer before extrusion, resulting in partial gelatinization of some starchy materials. According to Mr. Um, the thickness of his die plate was 0.6 cm.
Drying the pellets was the process that bothered Mr. Um most, because it was not possible on rainy or cloudy days. To minimize the problem, he suggested that the concrete surface should not be too thick. A thin concrete surface heats faster and accelerates the drying of the pellets. A 0.4 cm thick concrete surface was best.
Ponds were stocked with 2 week-old Clarias batrachus at the rate of 63/m². The fish were fed 30% protein commercial feeds for the first week as they were too small to be fed on his own feed. Thereafter the fish were given farmmade feeds once a day, in the evening, for 6 months until they averaged 200 g. The feed conversion rate was 1.3:1. Survival rate was only 6% and the production was 0.75 kg/m². The poor survival rate was due to low pH in the pond water, resulting from acid soil.
This farm-made feed cost US$ 0.30/kg in 1987. At this price, Mr. Um said that even if the farm-gate value of clarias catfish dropped to US$ 0.80/kg, he would have still made a profit if higher survival rate had been acheved. By 1991, Mr. Um had developed his system into a large feed mill, manufacturing floating catfish feed and marine shrimp feed for commercial purposes. The business is now run by his family, with a great deal of success, while Mr. Um is still running the catfish farm; naturally he uses his own commercial catfish feed.
The quality of feeds for snakehead is more important than for catfish. Unlike clarias catfish, snakehead are fed mainly on fresh trash fish. Minced trash fish is the sole feed for snakehead fry and is fed ad libitum, the feeding rate being adjusted after observing fish behaviour during feeding. The fish are fed three times a day from fry to the fingerling stage. The inclusion of trash fish in snakehead feed is reduced to 80% by the addition of 20% of rice bran when the fish reach the fingerling size. In some farms a mixture of trash fish and cooked broken rice is used. Trash fish is normally delivered to the farms but some farmers who want to select better quality trash fish collect it directly from ports.
Trash fish and rice bran or cooked broken rice are mixed and minced through a meat mincer to form strands, which are placed on 2 x 0.5 m wooden feeding platforms suspended along both sides of wooden piers. The number of platforms vary with the size of the ponds. Usually a 0.16 ha pond needs 4-6 platforms for thorough feeding of snakehead. The feed conversion rates of such trash fish based feeds range from 3.1:1 to 8.1:1.
Chockchai Farm, one of the most successful snakehead farms in the Suphanburi province of Thailand has used farm-made feeds for more than 20 years. The feeds consist of trash fish, chicken offal (including heads, bones and legs) and rice bran. Trash fish are collected daily from the landing port. Chicken offal is supplied by a slaughterhouse with which the farm has a contract. Rice bran is locally available. The proportion of trash fish:chicken offal:rice bran is approximately 7:2:1. These ingredients are mixed on a concrete surface by shovel and then shovelled into an elevator leading to the mincers. The farm has two mincers, one powered by a 20 HP electric motor and the other driven by a fuel engine providing a back-up in times of power failure. Each mincer can produce up to 2 t/hr of feed. The mincers are operated for 6 hours daily since the farm also supplies feed to contract farmers in the vicinity.
Mr. Chockchai, the owner of the farm, provided information on one of his successful snakehead grow-out ponds. The 0.32 ha pond was stocked with 4,500 kg of 70 g snakehead (1.4 kg/m²). The fish were fed twice a day. The average daily feeding rate for the entire growing period in this pond, which was 180 days, was approximately 748 kg. One critical factor mentioned by the farmer is that the trash fish used for feeding snakehead must be “daily fresh” otherwise the fish will not accept the mixed feed. The feed conversion rate was approximately 6:1 and the yield of this pond was nearly 27 t (84.2 t/ha).
Production and feed data for the catfish and snakehead grow-out operations described above is summarized in Table 7.
|Ingredients||Cost (US$/t)||Inclusion (%)||Note|
|Trash fish||220||70||FCR was 6:1|
|Chicken offal (heads, bones and legs)||120||20|
* two rai (0.32 ha) pond was stocked with 4.500 kg of 70 g snakehead (14.1 t/ha), and rearedto marketable size of 700 g in 6 months. The total yield was 26.950 kg (84.2 t/ha/crop)
ECONOMICS OF FARM-MADE FEEDS
The ultimate goal of aquaculture enterprises is to make profits. Profitability is determined by the difference between total revenue and total cost. Therefore, cost minimization is one way to achieve more profit. In many aquaculture operations today, feed costs account for over half of the total variable operating costs. Since commercial fish feeds are quite expensive, making good quality farm-made feeds might be more economical and result in more profit. The economic viability of farm-made feeds is place and time specific ( New 1987). Farm-made feeds may be suitable for certain farms but not for others. Tacon (1990) mentioned the factors to be considered before choosing to make farm-made feeds as the following: market value of the farmed species, financial resources, farming traditions, time available for the farming activities, labour availability, availability of services, feed ingredient availability and cost, feeding habit, feeding behaviour and nutrient requirements, water quality requirements of fish, and food and feeding cost per unit of production per unit time.
The nutritional and physical qualities of farm-made aquafeeds may not match those of commercial feeds but the former can utilize a wide range of locally available ingredients. However, the needs of the species to be farmed require careful consideration. Clarias catfish, by its nature, consumes almost any feed material offered. This allows great opportunities for making feeds on-farm if the farmer has access to the raw materials. Serm Farm is a good example of the economic feasibility of farm-made catfish feeds. Its success is notable (Tables 8 and 9) and is due to the low cost of its moist farm-made feed which has a reasonably good nutritional value and gives a high fish yield. The feed cost in 1992 was US$ 749/t. At this feed price, Serm Farm would still make a gross profit from the operation (all other factors remaining the same as in Table 8) even if the farm-gate value of the fish produced fell to US$ 0.79/kg.
Making compound feeds generally requires more capital than making moist feeds, as a hammer mill and a mixer are required, in addition to a mincer. If conventional raw materials such as fish meal, soybean meal, rice bran, etc. are used, the unit cost of these ingredients and their inclusion rate in formulae will dictate the feed cost. Making compound feeds for clarias catfish is economically feasible in some areas, where swine and poultry rearing enterprises are clustered, because ingredients can be selected and formulation modified according to time specific availability and price. In Thailand, these areas include Nakorn Pathom, Chachoengsao, and their adjacent provinces. The Um Farm once produced compound feed for clarias catfish at US$ 302/t. As shown in Table 9, Um Farm was not economically viable in 1992. However if it was not for its acid soil, which resulted in very poor survival rates, the farm would have made a profit by using farm-made feed.
Snakehead farms still rely on trash fish in farm-made feeds, although the cost of trash fish is rising. This is because the farm-gate value of snakehead is reasonably high (US$ 1.92-2.56/kg), well above feeding costs. Cost of snakehead feed per se is, therefore, not so important as for clarias catfish, which have a low farm-gate value. Taking the Chockchai Farm as an example (Table 9) feed cost for snakehead is US$ 1.51/kg of fish produced. The farm thus spends a huge amount of money on feed alone, However the farm would still break even (under the conditions shown in Table 8) if the farm-gate value of snakehead did not fall below about US$ 1.53/kg.
It is interesting to note that the Mungkorn clarias farm, which relied solely on commercial feed, failed to make a gross profit (Table 9).
|Cultured species||Clarias hybrid||C. batrachus||Clarias hybrid||Snakehead|
|Stocking density (pc/m²)||31||63||42||1.4(kg/m³)|
|Stocking size (cm)||2-3||2-3||10||66.7(g)|
|Growing season (days)||180||180||90||180|
|Average feed input|
|Feed conversion rate||3.6:1||1.3:1||1.4:1||6:1|
* the farm totally depended on commercial catfish feed
*** low production was due to acid soil
|Farm Species||Serm Clarias hybrid||Um C.batrachus||Mungkorn Clarias hybrid||Chockchai Snakehead|
|Gross profit (loss)|
|Farm-gate value at|
* cost of farm-made feeds includes 30% mark-up on ingredient cost as suggested byNew (1987). Feed costs are mid-1992 except for the Um Farm
** based on production (yield - stocking weight)
*** farm-gate values based on prices in July 1992, except the Um Farm (1987)
The majority of farm-made feeds for clarias catfish are moist, comprising a variety of unconventional raw materials. Their composition and nutritional value is therefore highly variable from place to place. Although making moist farm-made feeds is practical and cost-effective for certain farms, it may not be so for others since the raw materials used are unconventional. As long as snakehead rearing depends on trash fish it will be difficult to increase the overall production, since farming would not be possible where trash fish was not available. Production data for snakehead over the past decade support this prediction.
There is no doubt that farm-made feeds are economically viable for farms with access to inexpensive feed materials. However, the quality and quantity of moist farm-made feed cannot be easily controlled, This results in unpredictable and unmanageable fish production. Moist feeds disintegrate easily and therefore pollute the aquatic environment. The quality and quantity of farm-made feeds can be improved by processing conventional feed ingredients into a compound form. However, most of the compound feeds made are of the sinking-type, which are not attractive to farmers. Commercial feeds may be another alternative for catfish and snakehead farmers. However, the cost and palatability of each type of feed, the production management system in use, and the farm-gate value of fish need to be considered prior to making any decision.
LIST OF REFERENCES
Bardach J.E., J. H. Ryther, and W.O. McLarney. 1975. Aquaculture, the farming and husbandry of freshwater and marine organisms. Wiley-Interscience, New York, USA. 868 p.
Boonyaratpalin, M. 1980. Protein requirement of pla chon (Ophicephalus striatus), p. 37-38. In National Inland Fisheries Institute Annual Report (in Thai). Department of Fisheries, Bangkok, Thailand.
Boonyaratpalin, M. 198 la. Lipid requirements of snakehead fingerlings. Progress report of the regional project RAS/76/003, Network of Aquaculture Centres in Asia, Bangkok, Thailand. 30 p.
Boonyaratalin, M. 1981 b. Vitamin requirements in snakehead diets. Progress report of the regional project RAS/76/003, Network of Aquaculture Centres in Asia, Bangkok, Thailand. 18 p.
Boonyaratpalin, M. 1988. Catfish feed. National Inland Fisheries Institute. Extension Paper No.528 (in Thai), Department of Fisheries, Bangkok, Thailand. 17 p.
Butthep, C., P. Sitasit, and M. Boonyaratpalin. 1983. Water-soluble vitamins essential for the growth of Clarias, p.118-129.In C.Y.Cho, C.B.Cowey and T.Watanabe (eds.) Finfish nutrition in Asia, IDRC, Ottawa, Canada.
Chotiyarnwong, A., and W. Chuapoehuk. 1981. Protein requirements of Clarias batrachus. (in Thai). Faculty of Fisheries, Kasetsart University, Bangkok, Thailand. 5 p.
Garling, D.L. and R.P.Wilson. 1976. Optimal dietary protein to energy ratio for channel catfish fingerlings, lctalurus punctatus. J. Nutr. 106: 1368-1375.
Jantrarotai, W., P. Sitasit and S.Raichapakdee. 1992. Optimum dietary level of broken rice for growth and performance of hybrid walking catfish. National Inland Fisheries Institute, Technical Paper (in press). Department of Fisheries, Bangkok, Thailand.
Lovell, R.T. and E.E. Prather. 1973. Response of intensively fed catfish to diets containing various protein to energy rations. Proc. Ann. Conf. Southeast Assoc. Game and Fish Comm. 17: 455-459.
Luquet, P. and Y. Moreau. 1990. Energy-protein management by some warm water finfishes, p. 751-756. In Advances in tropical aquaculture. Actes Colloq. No. 9. IFREMER, Paris, France.
New, M.B. 1987. Feed and feeding of fish and shrimp. ADCP/REP/87/26. FAO,Rome, Italy. 275 p.
New, M.B. 1992. Farm-made feeds. Lecture presented at the AADCP Component 4 Training Course, 19-30 October 1992. AADCP Working Paper WP/20. ASEAN-EEC Aquaculture Development and Coordination Programme, Bangkok, Thailand. 23 p.
Sitasit, P. 1968. Study on the chemical condition of water in pla duk pond at Samutprakam Province, p. 107-139. In Annual Report of the Pond and Experimental Culture Section (in Thai). Department of Fisheries. Bangkok, Thailand.
Sitasit, P. 1970. The experiment of culture of Clarias batrachus in cages, p. 160-176. In Annual Report of the Pond and Experimental Culture Section ( in Thai). Department of Fisheries, Bangkok. Thailand.
Sitasit, P., N. Unpresert and W. Jantrarotai. 1984. Vitamin requirement for growth and survival rate of Clarias macrocephalus fry. National Inland Fisheries Institute Paper No. 36. Department of Fisheries, Bangkok, Thailand. 31 p.
Srisuwantach, V., R. Soungchomphan and P.Sitasit. 1981. Comparison of the effects of trash fish and pelleted diets in Clarias grow-out operations. Report on Programme for the Development of Pond Management Techniques and Disease Control (DOF-UNDP/FAO THA/75/012), Thailand. 21 p.
Tacon, A.J. 1990. Standard methods for the nutrition and feeding of farmed fish and shrimp. Argent Laboratories Press, Redmond, Washington, USA. 208 p.
Tacon, A.J. and M. Beveridge. 1981. Analysis of NIFI Clarias Diet No. 12. Report on Pro- gramme for the Development of Pond Mangement Techniques and Disease Control (DOFUNDP/FAO THA/75/012), Thailand, 2 p.
Taechajanta, K. and P. Sitasit. 1981. Assesment of a vitamin and mineral premix in an artificial feed for pla duk oui (Clarias macrocephalus) fry. Report on Programme for the Development of Pond Management Techniques and Disease Control (DOF-UNDP/ FAO THA/75/012), Thailand, 6 p.
Tanomkiate, K. 1984. Effect of feed with various protein to energy ratios on growth and survival of Clarias batrachus. Masters Thesis, Kasetsart University, Bangkok, Thailand. 60 p.
Thongutai, K. 1969. Study on the growth rate of pla duk in different stocking rate, p. 114-128. In Annual Report of the Pond and Experimental Culture Section (in Thai). Department of Fisheries, Bangkok, Thailand.
1 Feed Quality Control and Development Division National Inland Fisheries Institute Kasetsart University Campus, Bangkok 10900, Thailand
2 ASEAN-EEC Aquaculture Development and Coordination Programme (AADCP), P.O. Box 1006, Kasetsart Post Office Bangkok 10903, Thailand
BOONYARATPALIN, M. and M.B. NEW. 1993. On-farm feed preparation and feeding strategies for marine shrimp and freshwater prawns, p. 120-134. 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, 434 p.
Our paper describes some examples of farm-made aquafeeds for marine shrimp and freshwater prawns currently in use in ‘Thailand and shares some farmers’ experiences on the topic. It is hoped that this information may be of use to other countries in the region.
In Thailand, aquafeeds were made for shrimp and prawns long before commercial feeds became available. Previously chicken feeds were used for freshwater prawn culture but these were unsatisfactory, not only nutritionally but also because of their poor water stability. Furthermore, neither the ingredients nor the compounded feed were of an appropriate particle size. Farm-made feed production for Macrobrachium started around 1980 during a period of rapid expansion initiated through the Chacheongsao Fisheries Station and assisted by an UNDP/FAO project. To this date, most freshwater prawn farms in Thailand use farm-made feeds.
By contrast, farm-made feed manufacture for marine shrimp in Thailand is very rare; complicated formulae, the necessity for careful sourcing and quality control of ingredients, and the need for sophisticated feed technology generally being stated as the reasons for this. However, it should be borne in mind that most Thai marine shrimp culture is intensive or super-intensive, productivity usually being five to ten times higher per unit area than that for freshwater prawn culture.
These higher risk systems consequently require dependable, complete diets. Making such feeds requires suitable equipment, high quality ingredients, efficient formulation and an experienced plant engineer. Thus high investment and qualified staff are essential.
Farm-made crustacean aquafeeds in Thailand include larval feeds, Artemia fattening feeds, and grow-out feeds for Macrobrachium and, to a lesser extent, Penaeus spp. Our paper describes the method of manufacture of some specific examples of these types of feeds but does not cover their use on farms in detail.
Larval feeds are made for both freshwater prawns and marine shrimp and are confined to government hatcheries and small “backyard” hatcheries. Many of the more than 1,000 “backyard hatcheries” in Thailand are no longer smallscale. As they increase in size there is a tendency for them to convert to the use of commercial larval diets, because of the labour involved in making large quantities of larval feed on-farm. However, many hatcheries continue to make their own larval feed because the commercially available feeds are expensive, varying between US$ 24-80/kg, and their quality and texture is variable. In addition, they are still not complete feeds; in other words, live feeds remain an essential feature of hatcheries. For example, one commercial micro-encapsulated feed requires 15-30 Artemia nauplii/larva/day, in addition to large quantities of “artificial” feed (Tables 1 and 2).
Farm-made larval feeds in Thailand originate from the work of the Chacheongsao Fisheries Station, with UNDP/FAO assistance, in the late 70's and early 80's. Basically they consist of an egg custard, made from a mixture of whole eggs and the dried milk marketed for human babies (Figure 1). Sometimes a vitamin mix (three commercial shrimp vitamin premixes are marketed in Thailand) is added, though the other ingredients, such as mussel meat and fish flesh which used to be added a decade ago, have fallen out of fashion. Like the “artificial” feeds, farm-made larval feeds for crustacea are used in conjunction with Artemia nauplii, which are fed at 10-20 nauplii/ml/day. Feeding rates for egg custard are about 2-2.5 times as great as those recommended for the commercial encapsulated larval diets (Table 1) because they are moist. Since farm-made feeds require time to make, the first feed of the day consists of Artemia. Conversely the first feed of the day, when commercial larval feeds are used, is of dry feed (Table 3).
|Artificail feed||Natural live feed**|
|Quantity (g/t of water/day)||Frequency (#/day)|
|Stage||microcapsule size||flake*||Diatoms; Chatoceros, Skeletonema (cells/ml)||Artemia (nauplii per ml)|
|No. 1||No. 2||No. 3|
* after P10 stage, give flake feed at 15% of body weight/day
** constant or average level
|Stage||Size of microcapsules||Feeding quantity||Daily feeding frequency|
|Zoea*||No. 1||40-50 g/1 million zoea/day||4-5|
|Mysis**||No. 2 No. 3||0.12 mg/l mysis/day or 120 g/l million mysis/day||4-5|
|Postlarva||Flake||15% of body weight per day and increase at 12% per day||4-5|
* should also feed with diatoms
** should also feed with 15-30 Artemia/larva/day
|Type of feed||Feeding times|
Farm-made crustacean larval feeds are manufactured in Thailand as follows. Sixty eggs are cracked into a bucket (both egg-whites and egg-yolks are used), and powdered full-fat baby milk added at a rate of 2 teaspoonfuls per egg. The mixture is then whisked until it thickens and rises (Figure 2). It is then poured into trays and steamed for 30 minutes. Usually the trays are layered above the boiling water. The resultant steamed egg custard is then forced through a series of sieves using high pressure water and retained in a plankton net; sieve sizes vary according to the size of larvae to be fed (Table 4). The “egg custard” feed is then added to the larval tank (Figure 3).
Figure 1. Basic ingredients for farm-made crustacean larval feed
Figure 2. Farm-made egg custard feed for larval crustacea after cooking and before sieving
|Stage||Particle size (microns)||Mesh size||Artemia nauplii (per ml)|
|Mysis 3 - P4||177||80||20|
|P5 - P13||350||42|
|P13 - P20||500||35||10|
Figure 3. Farm-made crustacean larval feed at feeding time
ARTEMIA CULTURING FEED
Another type of farm-made larval feed used in Thailand is for culturing Anemia nauplii to an adult size. These Artemia are used as a supplement to the normal feeding schedule for Penaeus monodon, from post-larval stage 12 (P12) onwards, and also for 1 cm (one month-old) sea bass and grouper larvae. The major ingredient used is rice bran; 4 kg is added to 10 litres of sea water with 2 kg of sea salt and blended for 10 minutes. The blended mixture is then poured into buckets so as to allow the larger particles to settle. After being filtered through a nylon net it is squeezed by hand, washed and squeezed again to obtain as much filtrate as possible. The solid material is discarded, while the filtrate is filtered again through a finer (60 micron) nylon mesh. The concentrated Artemia feed is then ready for use or can be stored under refrigeration (0.5°C) for later use. Before use, 20 litres of the concentrated feed is diluted to 500 litres with water. An enrichment mixture is then prepared, which consists of 1 egg yolk, and 50 ml of “EPA oil” (eicosapentaenoic acid; EPA oil is available in the local markets), diluted to 1 litre with water. One litre of the enrichment mixture is then added to each 500 litres of Artemia fattening feed. The complete feed is added to Artemia rearing tanks by hand, or automatically through pipes equipped with a submersible pump and a timer; the feed is applied every 2-3 hours.
GROW-OUT FEEDS FOR MARINE SHRIMP
As noted earlier, few Thai marine shrimp farms use farm-made aquafeeds now because they are characteristically intensive or super-intensive, requiring reliable, nutritionally complete, and water stable feeds. However, two farms which have made their own feeds are described here.
The first farm, which is situated in Rayong, previously used mincing machinery but does not do so now because the quantity of feed was thought to be too large (however, very large through-put mincers are available in China and Japan). The 100 kg/hr mincer used was not the limiting factor; the main problems encountered were with drying, and with the use and regular supply of trash fish. A typical formula used when the extrusion process was employed, which used trash fish as a binder and protein source and had a high moisture content (35-40 %), is given in Table 5.
The first version of their new feed plant was relatively simple and included a hammer mill, a pelleter and a drier. Basically, ingredients were mixed, elevated to a triple conditioner heated by heat lamps rather than steam, pelleted and dried in a rotary drier also heated by lamps. The plant, capable of producing 300 kg/hr, cost approximately US$ 15,000. A typical formulation for this first pelleting plant is presented in Table 6. No trash fish was used in the formulation but cooked rice was employed, which resulted in a moisture content of 23-25 %.
|Good quality trash fish||20.0||23.8|
|Shrimp shell meal||10.0||11.9|
|Squid liver meal||4.0||4.8|
|Vitamin and mineral mix||0.5||0.6|
* extrusion through mincer
|Ingredient||Inclusion rate (%)|
|Shrimp shell meal||10.0|
* approximately 20 kg after boiling
** 30 whole eggs, without shell
Later, a new 500 kg/hr production plant was installed which was modified from a chicken feed plant, adding an electrically heated rotary drier. Dry ingredients were ground, transferred to a horizontal mixer and then by screw elevator to a vertical mixer. The mixture then passed through a conditioner/ feeder to a pelleter. Following pelleting the feed passed through two driers, one electrically heated and the second a vertical cooler/drier, before packaging. A typical example of the formulation used in the second pelleting plant is given in Table 7. However, there were further problems. The lack of trash fish or boiled rice in the formulation, caused by the necessity to reduce moisture to 14-16%, and the inadequate conditioner (which was the weak point in the whole system), meant that a mixture of three binders had to be used to obtain the necessary water stability; the binders used were guar gum (1%), aqua-gel (0.5%) and wheat gluten (7%). However, poor conditioning resulted in problems with the pelleter and this plant is now shut down.
|Ingredient||Inclusion rate (kg)|
|Shrimp head meal||10.0|
* 30 whole eggs, without shell
** 3 litres of water were also added to this formula during manufacturing
The experience of the Rayong plant neatly demonstrates the problems that can occur when a small production unit is scaled up to make complete feeds for intensive culture. Design faults quickly caused plant breakdown and the company could not afford to hire a good nutritionist or a good plant engineer. Keeping the plant simple would have been more satisfactory. In retrospect, it would have been perfectly feasible to continue to extrude the feed through a mincer rather than a pelleter. Often it is thought that this technique, which enables the use of moist ingredients such as trash fish, would not be feasible because of the large area required for sun-drying. However, as the next example shows, mincing can be followed by steam drying to solve this problem.
Our second example of farm-made feed manufacture for marine shrimp is from a small cooperative of poor farmers in Surat Thani. These farmers, unlike those which were supplied by the Rayong plant, do not rear their shrimp very intensively. Their feed unit was started in 1986 and employs ten workers producing 1-5 mt of shrimp feed per day. Three types of shrimp feeds are produced with 40-42% protein for use in the first month (from P15 onwards), 38% for the second and third months and 30% for use after that. No formulae were provided but the ingredients used include fish meal, soybean meal, broken rice, rice bran, shrimp head and shell meal, squid meal, yeast, fish liver oil, squid liver oil, a premix, dicalcium phosphate, and a binder.
The small size of the mixer used by these farmers necessitates ten separate mixing cycles of 100 kg to produce each one-ton batch. Thus the workers work in two shifts, one in the morning and one in the afternoon. The feed manufacturing equipment employed consists of an ingredient grinder, a mixer (100 kg capacity), a meat grinder (used for extrusion), a small boiler (used for boiling the broken rice included in the formulation to increase its digestibility), and a drier. Wood is used to fuel the main boiler, which generates steam for the steam jacketed three-phase drier. Ingredients are weighed on a simple scale (shrimp meal and soybean meal being first re-ground) and then added to the mixer. Major ingredients are inserted first, except broken rice, which is preboiled. Minor ingredients, such as squid meal, yeast, vitamin premix and binder are premixed by hand before being added to the mixer. Liquid ingredients (squid or fish liver oil) are then mixed in, and finally the boiled broken rice is added. After the last ingredient is added, mixing is continued for 15 minutes. The material is then discharged from the mixer before transfer to the grinder where the mixture is extruded in spaghetti or noodle-like form. The moist pellets are then elevated to the first two stages of the steam-jacketed drier and the semi-dried material then elevated again to the third-stage drier. Finally the dried material is discharged and the pellets left on the floor overnight to cool before being bagged for transport to the members of the cooperative. Though effective, this production
unit could be improved by adding a cutter after the extruder and a cooler after the drier.
GROW-OUT FEEDS FOR FRESHWATER PRAWNS
Farm-made aquafeeds for Macrohrachium are common in Thailand. Compared to marine shrimp feeds, freshwater prawn feed formulations are simple because the culture system is comparatively semi-intensive. Where marine shrimp culture is still semi-intensive, as for example in some other Asian countries, similar farm-made aquafeeds are feasible.
Freshwater prawn farms which make their own feeds are common in the province of Suphanburi, north of Bangkok. Many farms also have their own inland hatcheries, thus presenting simple examples of vertical integration. At the first farm we are describing, ingredients are weighed and mixed together on the concrete floor. A “wok” is used to boil the broken rice. Two examples of the feed formulations used on this farm are given in Table 8. Each ingredient is carefully added to the apex of a heap of ingredients on the floor to ensure adequate initial mixing. The mixture is then shovelled into bowls for transfer to the mincer for extrusion (and further mixing) through a coarse die plate. The material is collected in deep baskets where it remains while a finer die plate is fitted. The result of the second extrusion is collected in shallow baskets (Figure 4), which are easier to carry to the sun-drying area. The moist pellets are then spread thinly on a concrete floor and turned regularly with a simple rake.
Other Macrohrachium farms in Suphanburi use a similar process. Our second example has been manufacturing feed for the past six years. The formulation is shown in Table 9. Feed sufficient for two or three days is made at one time, depending on the availability of trash fish. The mixed ingredients are extruded twice (coarse, followed by fine die plate) through a mincer (powered by a converted car engine) and then sun-dried. The feed is extruded in the morning (07.00-08.00) and, if sunshine is bright, drying is completed within one day on a concrete surface (Figure 5). If necessary, feed is collected and put out to dry again on a second day. The water stability of the pellet is reported to be up to two days. Feeding rate depends on visual observations; grow-out ponds being fed once per day (at 15.00-17.00) and nursery ponds twice per day. Feed is broadcast by hand from the pond bank and no fertilizers are applied. The average farm gate value of the crop in June 1992 was reported to be US$ 4,600/t.
|Ingredient||Formula #1||Formula #2|
|Vitamin and mineral mix||0.5||0.15||0.5||0.14|
|Broken rice (boiled)||30.0*||8.88||30.0*||8.51|
|Chicken layers feed||50.0||14.81||60.0||17.01|
|Shrimp shell meal||-||-||15.0||4.25|
* weight before boiling
Figure 4. Extrusion of farm-made freshwater prawn feed
|Ingredient||Cost (US$/t)||Inclusion rate|
* lipid not less than 4: protein not less than 25; fibre not more than 8%
** composition not known
*** contains vitamins A. D, C. E and an unspecified antibiotic (oxytetracycline ?). Availablelocally and made specifically for Macrobrachium. Manufacturer suggests inclusion inthe diet at 0.5-1.0%
**** it was said that no water was added; trash fish provides sufficient moisture
Figure 5. Sun-drying a 450 kg batch of freshwater prawn feed
Examples of a farm-made crustacean larval feed, an Artemia fattening feed, and a number of grow-out feeds for marine shrimp and freshwater prawns have been provided and the manufacturing process described*. Experience in Thailand indicates that, as the hatchery grows large or the (marine shrimp) farming system becomes intensive, operators move towards the purchase of commercial feeds, mainly because the large-scale production of farm-made crustacean feeds is time-consuming and labour intensive.
In the case of small hatcheries, simple egg custards are still popular, both for marine shrimp and freshwater prawns, and this is likely to remain so until commercial larval feeds no longer require supplementation with Artemia nauplii or other live feeds. These egg custard feeds are readily applicable outside Thailand. One important fact is that farm-made feeds for crustacean larvae in Thailand include both the yolks and the whites of the chicken eggs incorporated. Many fish larval diets exclude the egg white, which is a mistake in our view. Egg white is a valuable, high-protein ingredient, which will not foul the water in larval tanks if the diet is prepared (homogenized) correctly. In fact the egg white helps to encapsulate the egg-yolk to make a water stable larval feed.
Farm-made feeds for marine shrimp culture have not really proved feasible when the culture system is intensive or super-intensive. Feeds for this type of culture, which must supply all the nutritional requirements of the shrimp, must be of consistently high quality to produce optimum growth rates, food conversion efficiency, and survival. Intensive shrimp culture is a high-risk business and poor or inconsistent quality feeds are anathema. Good quality shrimp feeds require high quality fish meal. Feed plants only making shrimp feeds cannot utilize poor quality batches of fish meal in feeds for other farmed animals. The production of reliable feeds for intensive systems of shrimp culture requires the installation of good production equipment in a well-designed plant, a reliable supply of high quality ingredients, careful least-cost formulation to ensure the economic provision of all known nutritional requirements, a qualified plant engineer and well trained production staff.
However, the use of farm-made feeds in less intensive crustacean farms remains a viable option. Even large farms or cooperatives of a number of smaller farms could overcome the space requirements for sun-drying extruded pellets by installing steam-fired (or electric) driers. For the small farm sun-drying on a clean concrete surface is feasible, even in the rainy season. The common and long-term use of farm-made feeds by Macrobrachium farmers in Thailand demonstrates that they are an acceptable, practical and profitable option. This applies not only to freshwater prawn culture but also to marine shrimp farms where supplemental, rather than complete, feeds are appropriate to the culture system used. The systems which we have described can, with minor adaptation, also be used for producing farm-made fish feeds.
* Editors' note: when this paper was presented, over 70 slides were shown, which illustrated the manufacturing procedures in detail.
Grain and Food Processing Section
Natural Resources Institute
Central Ave., Chatham Maritime, Kent ME4 4TB, U.K.
WOOD, J. 1993. Selecting equipment for producing farm-made aquafeeds, p. 135-147.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, 434 p.
When addressing the subject of preparing feeds for fish it is important to recognise that we are in fact seeking to prepare a foodstuff with specific nutritional and physical properties to meet the differing feeding habits of a range of aquatic consumers. The purpose of the feed manufacturing process is therefore to prepare a food which, as far as possible, meets the gastronomic habits of the target consumer. The need to prepare feeds for slow and fast feeders, water surface, mid water or bottom feeding species has been well recognised.
If a farmer is to be successful in feed preparation it is important that he examines the options available to him. In practice, most farmers have come to believe that pelleted feed is the most desirable since this is the feed form which has resulted in high yields on what are perceived to be financially successful large commercial fish farms. This is an understandable conclusion to make but its validity on a nutritional and practical basis still requires solid confirmation for semi-intensive culture. Nevertheless, in this paper we will consider the factors affecting the selection of equipment, principally for making feeds of high water stability in pellet or noodle form, since these types of feed potentially enable the maximum number of consumers to be fed at any one time. This paper will outline the principles behind equipment selection, and in particular comment on the properties of feed raw materials which will influence such decisions.
WHAT DO FARMERS WANT?
In seeking to meet the requirements of the fish we must also examine the requirements of the farmer. What are his objectives and are they achievable? The two may not necessarily be compatible in all cases. Annex 1 summarises the primary objectives of the farmer concerning feed preparation, but the degree to which they are achievable will vary from farm to farm, region to region and country to country. There is no single solution that will be practical and cost effective for all farm-made aquafeeds. We are assuming that the farmer already has a feeding strategy but wishes to improve upon it.
FACTORS INFLUENCING THE SELECTION
AND INSTALLATION OF FEED PROCESSING EQUIPMENT
The supplementary feeding of cultured fish has become an established practice in many parts of the world and numerous designs and sizes of machines have been used for farm and commercial feed manufacture. Some farmers have made wise choices while others have regretted expenditure on equipment which has failed to perform adequately with the raw materials available, or was of an inappropriate capacity.
If a farmer is starting a new venture there are important factors to be evaluated before any selection of specific equipment should be considered. The response to their evaluation will then form the framework for determining the machinery requirements to meet the feed production objectives within the financial resources available (see Annex 2).
This evaluation of interacting factors must be conducted with an understanding of the physical and functional properties of the raw materials available. These aspects will be discussed in more detail later in this paper.
PROCESSING EQUIPMENT OPTIONS
There are many and various options for processing equipment for aquafeed processing, as can be seen from the practices of aquaculturists in any locality. Processing equipment options include electric or petrol/diesel engine driven machinery, and also the use of hands and feet. The most common processing operations can be summarised as:
raw material size reduction;
raw material blending;
The range of equipment options within each processing operation are summarised in Table 1. It is not appropriate to describe these machinery options in detail since almost all are well known and specific details can be obtained from manufacturers.
|Process operation||Equipment||Raw material/product|
|Size reduction||Mortar and pestle||dry or moist grinding or blending|
|Mincer||wet materialse.g. trash fish/offals|
|Hammer mill||coarse-fine dry materials|
|Plate mill||coarse-fine dry materials|
|Blending||Physical||Hand||for small quantities variable efficiency|
|Mechanical mixers||Bowl||moist doughs|
|Horizontal||dry powders or moist crumbs|
|Cooker extruder||semi-moist/dry pellets or noodles|
The choice of equipment will be limited firstly by financial resources, secondly by the desired water stability of the feed (and thus raw materials availability), and thirdly by the required scale of feed manufacture. However, on the assumption that finance is not limiting and all forms of machinery are potentially available, then the factor which will govern machinery choice is the spectrum of raw materials on offer for feed formulation and manufacture.
FUNCTIONAL PROPERTIES OF FEED RAW MATERIALS
Commercial fish feed manufacture is predominantly associated with the processing of dry ingredients and the manufacture of a dry product. This is not necessarily the case for farm feed manufacture. Commercial processors require dry products for long term storage and transport. Farm feeds can utilise local raw materials which may be:
high in moisture and of short shelf life;
of insufficient quantity to become commercially viable;
grown on the farm or be a byproduct of local agriculture;
available during certain seasons only;
having functional properties which have not been damaged through industrial pre-treatment.
The question which must then be asked is: “In what way can the nutritional and physical characteristics of the raw materials available meet the desired nutritional requirements of the fish to be farmed, and be in a form which will stimulate feed intake and be water stable throughout the feeding period?”
To enable this question to be addressed we should consider the general functional properties of feed nutrients and the effect which moist heat is likely to have on them. The functional properties of raw materials, or more specifically of feed nutrients, are those which affect the ability of feed materials to:
cross bond with each other;
alter their viscosity characteristics;
change from granular to plastic consistency;
release bound moisture;
induce water stability.
The effect of moist heat on the functional properties of feed raw materials is important since this is the kind of treatment which many are subjected to during processing before they, or their byproducts, are actually considered as raw materials for fish feed (or other animal feed).
The important functional properties of the major nutrients are summarised in Table 2. These properties are, of necessity, expressed in general terms since there are differences in specific properties depending upon the source of the nutrient. From Table 2 it is evident that the nutrients with the most important functional properties are the proteins and the starches. In most circumstances, both nutrients will also be present in high proportion in the mixed diet and therefore have the potential to significantly alter its properties. Let us examine the functional properties of these two nutrients in a little more detail.
|Nutrient||Normal state||Effect of moist heat||Change in functional property|
|Proteins||colloidal fibrous globular viscous||denaturing||soluble/hydrable to insoluble|
|Starches||inert granules||gelatinisation||insoluble granule to soluble gel|
|Fats||liquid or solid||some cross bonding with starch amylose||difficult to extract with organic solvents|
|Sugars||soluble solid or liquid||reaction with lysine caramelization||minimal|
|Minerals||soluble or insoluble solid||minimal||minimal|
|Vitamins||soluble or insoluble solids||some heat labile||minimal|
In their native state starches, whether from cereal or root crops are found in the form of starch granules which are essentially inert when placed in cold water. They resist water absorption, and there is minimal adhesion between granules. In this form starches are also of low digestibility to aquaspecies, and have minimal properties for binding other feed components. It is only when starches are heated in water, causing the granules to rupture and gelatinise forming viscous pastes and gels, that starch has desirable properties for binding feeds.
However, since the starch has also become more soluble during the gelatinisation process (a property which is retained after rapid starch drying), feeds bound with gelatinised starch alone will disintegrate as the feed hydrates and continues to absorb water, although the degree of binding is considerably better than when feeds contain no gelatinised starch.
In their native or natural form, proteins associated with plants tend to be “globular” in shape whereas animal proteins are more characteristically “fibrous” in form. Apart from those proteins which are totally insoluble, such as those comprising wool, silk and hair, proteins are soluble in water, salt solutions or mild alkaline or acid solutions. When subjected to heating, proteins tend to coagulate or denature. This is a process which is well recognised during the cooking of egg albumin, when the raw soluble protein is converted to an irreversibly insoluble protein gel.
To illustrate the potential interactions of raw materials let us examine what processing options could be available to a farmer for fish feed production. For example, let us consider the use of three raw materials potentially available in many developing countries i.e. trash fish, sun dried cassava and full fat soya beans. These raw materials may be described in the following way:
trash fish: hydrated, fibrous protein, with some fat associated, but no starch;
cassava: non-hydrated starch granules with little protein or fat;
full fat soya beans: non-hydrated globular protein with high association with fat and low in starch.
We will assume that each raw material will be used at the same dry matter level in each feed, and that the blend of raw materials will meet the desired nutritional requirements of the culture system being operated. The oil present in the soya beans is a required nutrient for the diet.
Let us assume a process whereby all three raw materials are treated individually as industrial agro-products for processing and storage. They will then be blended to form a feed of the desired nutrient specification.
|RAW MATERIAL||PROCESS OPTION||EFFECT OF PROCESS ON FEED BINDING|
|Trash fish (souble protein)||Hot air dried to fish meal (coagulated protein) (denatured protein)||Loss of ability to cross link with other proteins|
|Cassava (inert granules)||No further processing (inert granules)||Inert granules with no gelling chractersisitcs|
|Soya beans (hydratable protein)||Oil expelling; toasting of soya meal (coagulated protein)||Loss of ability to cross link with other proteins|
Result: Three separate raw materials with no inherent ability to bond with each other. These materials will require a binder and careful steam conditioning to partially gelatinise the cassava before any degree of binding between particles can be obtained.
|RAW MATERIAL||PROCESS OPTION||EFFECT OF PROCESS ON FEED BINDING|
|Trash fish (soluble protein)||Sun dried to produce fishmeal (coagulated protein) (partially denatured)||Partial loss of ability to cross link with other proteins|
|Cassava (inert granules)||Cook in water (gelatinised starch)||Viscous base for aiding feed binding|
|Soya beans (hydratable protein)||Oil expelling; toasting of soya meal (coagulated orotein)||Loss of ability to cross link with other protines|
Result: Improved stability of the feed due to partial hydration of some fish proteins and viscosity of gelatinised starch. Heat denatured soya proteins are inert and add little to feed stability.(Note: Even slow sun drying results inpartial denaturation of fish proteins such that rehydration does not result in the reformation of colloidal protein material).
|RAW MATERIAL||PROCESS OPTION||EFFECT OF PROCESS ON FEED BINDING|
|Blend the materials together and co-extrude while heating||Formation of intermeshing matrix of proteins and starch which are respectively denatured and gelatinished simultaneously|
Result: This process will give the most effective bonding between the feed components, and thus the maximum feed stability. The process is in effect that which occurs during the process of cooker extrusion, and is the reason why this process has become so popular amongst commercial fish feed manufacturers. Furthermore, when using full fat soya as a raw material, the process enables the destruction of the trypsin inhibitors which would otherwise significantly depress protein digestion and assimilation by the fish.
A further advantage of this form of processing is that it can incorporate wet fish as an ingredient into the feed mixture without the need for pre-drying, or the need to extract the oil from soya beans prior to their use.
ARE THESE METHODS APPROPRIATE FOR
FARM-SCALE AQUAFEED MANUFACTURE?
The foregoing options illustrate the desirability for the controlled processing of protein- and starch-containing raw materials by methods which will maximise the effects of their functional properties. Feed materials must be examined in terms of their physical or functional properties and not solely as sources of nutrients.
On many fish farms, feeds are being prepared which take into account the factors discussed above. The benefits of cooking starchy bases such as cassava or rice have been recognised, both for developing the viscous gelatinisation of starch and the resulting improvement in digestibility.
However, the stability of formulated feeds which include cooked starches may also be due to a degree of starch retrogradation during cooling and sun drying. During retrogradation, starch gels tend to lose their consistency as starch molecules and form into crystalline structures (for example, starch retrogradation is often associated with the staling of bread). Once a starch has retrograded it is often more difficult to solubilise, and is less readily hydrolysed by enzymes. From a fish feed perspective therefore, starch retrogradation in association with a protein matrix may enhance feed resistance to disintegration in water. The negative effect will be reduced starch digestibility.
Where farmers use trash fish as their base, they are introducing a valuable source of both nutrients and functional characteristics to the feed. For example: a Thai farmer produced his own feed for freshwater prawns from the following ingredients: wet trash fish (44.4%), fish meal (13.3%), rice bran (6.8%), soya meal (8.9%), poultry feed (13.3%), and broken rice (13.3%). The rice was cooked to a thick paste and then mixed with the ground dry ingredients and the minced trash fish. The moist dough was then formed into strands through a mincer die plate, and the product sun dried on a concrete pad. The dried feed, which had a water stability of more than 24 hours in static water before disintegration took place, was bound with gelatinised and partially retrograded starch and by the matrix of partially denatured fish proteins produced during sun drying. Of particular interest is the addition of poultry feed. The farmer commented that it had been added to the feed mix to overcome the tendency of the fibrous rice bran to break the structure of the pellet during drying. Practically and nutritionally, the farmer may have had the same success by removing the rice bran and poultry feed from the diet and replacing them by soya meal and broken rice.
The process used in the preparation of this feed was quite labour intensive but, since the farmer was able to obtain most of his dry materials in a pre-ground form, his machinery requirements were reduced to a mincer and a simple heated bowl for cooking the rice. The mincer also served as the device for shaping the feed dough into long strands for sun drying prior to crumbling to feeding size pellets, and for storage.
This situation is, however, not typical of fishfarms which are situated away from coastal regions, where supplies of trash fish are non-existent. The only ingredients available may be the byproducts of other agro-industries which have heat treated the materials and denatured the proteins during processing, e.g. in oilseed cakes. Under these circumstances farmers may be able to obtain sufficient raw materials to meet the nutrient requirement of the culture system being operated, but have few materials which will help in developing any degree of water stability. The most sophisticated machinery will be of little use to the farmer unless he is able to obtain additional raw materials to induce binding, while maintaining the nutritional integrity of the desired feed.
As aquaculture expands throughout the world there is a requirement that farmers use their water resources wisely. This applies not only for the protection of their own stock, but for the protection of those who will abstract pond discharge waters for other agricultural uses, and also for the protection of the consumers of cultured species.
One of the first environmental requirements of an aquaculture farm is to maintain the quality of the aquatic environment by attempting to ensure that feed added to the pond reaches the target animal. This is the purpose of preparing water stable feeds.
A further requirement is to ensure the microbiological safety of aquatic feeding systems. Since many animal proteins are associated with potential pathogens such as salmonella, it is desirable that animal (including fish) proteins should be pasteurised before they are ingested by cultured fish. This is particularly important should the trash fish have been grown in septic ponds or other water courses potentially contaminated with faecal material. Similarly animal viscera, such as poultry offal, should also be treated before it is consumed by fish or other aquatic species. At the moment, few farm feed processors have the facilities to pasteurise protein material of this kind while including it as valuable nutrients and functional protein within feeds.
There is therefore a need to develop low cost farm scale feed processing equipment which will enable raw materials to be processed to produce feeds with acceptable microbiological and water stability characteristics. Of particular benefit would be a heated die plate for attachment to a mincer or similar feed forming machine which would act as a feed pasteurisation unit while also developing a degree of starch gelatinisation and protein denaturation. However, such a device could not replace the functions of a cooker extruder.
In presenting the above information an attempt has been made to demonstrate that the selection of machinery for the manufacture of farm made aquafeeds is not a simple exercise of selecting items from a catalogue in accordance with the quantity of feed desired. It is important that machinery is selected in relation to the properties of the raw materials available for formulating and processing. We are aware of what is desirable, but the task is to select the most appropriate equipment for use with non-ideal raw materials. This challenge will be much greater for some farmers than others.
|Objectives||Potential for achievement|
|Prepare feeds of ideal nutritional and physical quality which are cheaper than existing feeds||Maybe|
|Use raw materials which are locally and constantly available,||Doubtful|
|Produce feeds that are attractive and non-polluting||Possibly|
|Utilize processing equipment which has low capital and running costs and is available locally||Perhaps|
|Take minimal time and effort for processing||Uncertain|
|Have technology which is known only to him, is specific to his needs, and gives him an advantage over competitors||unlikely|
|Have a technology which is operable by family members||Hopefully|
Annex 2. Preliminary factors for quantifying desired aquafeed production capability
Identify and evaluate the following:
target animals and their feeding behaviour;
raw material availability and continuity of supply;
feasibility of obtaining the desired feed conversion and crop yield;
output of feed required in relation to fish growth phase;
desired frequency of manufacture in relation to fish growth phase;
characteristics of the proposed site: access, power supply and its reliability. building design, storage facilities for raw materials and finished feeds, security:
desired lifespan of equipment;
possible future developments/expansion on site;
locally made or imported equipment;
process equipment options, flexibility, power requirements;
compatibility of options for formulations and equipment;
access to finance.
INDECO/Winrock International Agency for Agricultural Research and Development Jalan Salak 22, Bogor 16151, Indonesia
CHONG, K.C. 1993. Economics of on-farm aquafeed preparation and use, p. 148-160. In M.B. New, A.G.J. Tacon and 1. 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.
Fish (used generically here to refer to finfish and shellfish), when fed a balanced diet, clearly respond well to the feed provided. The artificial (compound) aquaculture feed industry came into existence to supply the feed needed by fish farmers. These feeds are invariably based on 25-65% fish meal. Apart from micronutrient premixes fish meal is usually the most expensive of all commonly used feed ingredients.
Because they are commercially manufactured, feed quality can be controlled in terms of uniformity, nutrient composition and use convenience. Because of these advantages, the manufacturing process is costly and becoming increasingly so as the supply of fish meal declines and demand for fish meal increases, together with fish meal-based feeds. Feeds are often expensive because they tend to be over-formulated. This tendency arises due to the lack of information on the nutrient requirement of the fish under culture conditions.
At the outset it should be clearly stated that, contrary to the popular and widespread myth that fish meal is not (or should not be) used in farm-made aquaculture feed, fish meal can be included as one of the ingredients in on-farm feed formulation. However, the economic gains or returns from including or excluding it should be closely examined. The term fish meal is taken here to include both fish meal itself and trash fish.
I term aquaculture feeds as both single ingredient and compound feeds. The former consists mainly of natural feeds and is thus of variable quality while the latter is usually artificially manufactured and thus of more uniform and high
quality. A distinction is also made between factory-manufactured aquaculture feed and on-farm prepared aquaculture feed. Since it first became commercially available, factory-manufactured aquaculture feed has been accepted and used as a necessary input in fish farming, in spite of its apparent high cost. Such aquaculture feed is generally used in the production of high-value (and normally low-volume) species. Because of this, its use is economical and profitable.
On the other hand, it is claimed that it is relatively cheaper to prepare your own feed than to buy commercially available feed, even when the capital cost of an on-farm feedmill is taken into account and even if the feedmill is underutilised! Such is the economic impetus and rationale for on-farm feed preparation. There is a 250-ha shrimp farm in Indonesia with its own shrimp feedmill which shows this to be the case. The implication is that the profit margin of commercial aquaculture feed is still large.
The apparently high cost of high-protein fish meal-based commercial aquafeed is still generally recognised as the main constraint to greater fish output from culture. Because of growing scarcity of fish meal supply due to overfishing and (traditional) competitive demand, for livestock feed and other uses, alternative aquaculture feeds and/or sources of feed ingredients are being actively explored. One of the options is on-farm prepared feed, using locally available feed ingredients, preferably with little or no fish as a protein and energy source.
The economic importance of feed on final farm profit is too critical to be ignored or taken lightly. At present, commercial aquafeeds cost an average of US$ 0.50/kg (from US$ 0.30-1.00/kg). Feed cost accounts for an average of 30-60% of production cost, sometimes even higher. As such, profit is highly sensitive to feed cost. Every improvement in feed conversion efficiency means a bigger profit margin for the fish producer.
Supplier induced demand
The purchase and use of commercial aquafeeds remains very much a supplier induced demand. By this is meant that the feed suppliers induce fish farmers to buy the feed. Fish farmers do so because they need to feed their fish. They (actually) do not have much choice as there seem few viable alternatives. This is especially so if they want higher yields using intensive systems. One alternative, which is to rely on natural feed, does not give as good a result as feeding commercial feed.
Another alternative, which is to fertilize the fish ponds with inorganic and organic manures to stimulate natural productivity in the water column, also does not give as good a result as commercial aquafeed. Given such production circumstances, fish farmers having access to capital rely on the feed supplied by manufacturers. Feed manufacturers are able to induce fish farmers to buy their feed through persuasive sales advice and recommendations because they are not as informed on fish nutrition.
Because of this supplier-induced demand, there is a real tendency for fish farmers to over-feed, either knowingly or unknowingly. This arises because farmers tend to follow manufacturers' feeding schedules, causing unnecessary waste. In addition, such feeds tend to be over-formulated and become uneconomical. More and more fish farmers and manufacturers are recognising the economic significance of feed.
It can be observed that the overall economic status of feed manufacturers is generally better than that of fish farmers. This is because of the supplierinduced demand for feed phenomenon. Fish farmers, being price-takers, have little or no influence on feed price determination. As a result, manufacturers tend to become prosperous more readily. In addition, farmers rarely keep records of feed use; if they do, the data is seldom rigorously analysed to improve farm performance.
Ratafia and Purinton (1989) estimated that there were over 118 feed companies worldwide producing aquafeeds. This is a low estimate as many small feed companies, especially in developing countries, have not been accounted for. The number of aquafeeds companies is projected to grow as the aquaculture industry itself expands. As aquafeed demand is at present largely induced by suppliers, producers need to closely scrutinize their feed costs as a cost centre in their farm management economic analysis. Economic farm performance indicators can be developed to guide them in the use of feed.
Purpose of study
Where on-farm aquafeeds are concerned, formulation, preparation and use are still largely governed by technical and technological considerations rather than economic criteria. There remains a general lack of information on the nutritional requirements of fish under culture conditions. By incorporating economic performance indicators, greater profits can be obtained.
The economics of on-farm aquafeeds is examined in this paper as a viable alternative to commercial feed and as one which is environmentally sustainable and inexpensive.
ECONOMIC BASIS OF ON-FARM FEED PREPARATION AND USE
Supplementary feed-based aquaculture
According to Akiyama (1991), most of the 15 million t global aquaculture output is produced without the benefit of artificial feeds. Most is produced by using fertilizers, both inorganic and organic. Reliance on organic manures to grow fish is quite prevalent in Asia, especially in China, and is usually identified with integrated crop-livestock-fish culture. Since Asia accounted for 85% of the world's aquaculture production in 1988 (Csavas, 1991), it can be inferred that approximately 80% of global aquaculture production is produced without using commercial aquafeeds. Thus, great potential exists to increase the use of feed to produce more fish. Feed can significantly increase yield. As pointed out above, fish meal is used in feed formulation for both commercial and farm-made aquafeeds. Commercial aquafeeds are heavily dependent on fish meal as a key feed ingredients.
However, on-farm aquafeed holds greater promise in a world increasingly concerned with environmental quality and natural resources system productivity, stability, sustainability and equitability. Even though farm-made feed can be environmentally more damaging than commercial feed, because of its higher biological oxygen demand (BOD), the former should be actively promoted as a viable alternative because of its use of byproducts and waste recycling basis. Farm-made aquafeed uses locally available raw materials as ingredients which are found mainly on or in the vicinity of the farm.
Since such wet ingredients are usually incorporated without further processing and the resulting ground feed is directly fed to the fish, its moisture content is very high. The use of such feeds, combined with improper feeding practices, not only results in unnecessary waste but also leads to more rapid (bottom) pond soil deterioration due to leftover uneaten feed. Constituting organic matter of relatively high protein content, waste feeds become rapidly fouled, giving rise to poor water quality, in turn creating stressful growing conditions. Loss of appetite, poor or slow/stunted growth and disease outbreaks causing high mortality can occur, all translating into economic loss.
Although effluents from intensive fish farms have been alarmingly viewed as an environmental problem, they may not necessarily be so. Ways and means can be found to strip the nutrient-rich pond effluents for recycling, including for fish farming (Chong, 1992a).
At present, many types of aquaculture feeds with many brand names are available in the market. For now, the market appears to be a captive one, as fish producers are more or less dependent on them. This is especially true in countries where feed technology is not well developed yet.
Economic value of feed in aquaculture
This paper is not concerned with pond fertilization but with the use of feed, especially the prospects and contribution of farm-made aquafeeds to grow fish. It will try to answer the million dollar question: is farm-made feed use economical and can such feed replace commercial feed in aquaculture?
The use of feed in salmon, trout, catfish, shrimp and eel culture, among others has been shown to be profitable. However, these commercial feeds rely heavily on fish meal as a major ingredient. As a result of the acceptance of feed as a key input in aquaculture, a sizeable aquafeed industry has developed and prospered. From less than 1 million t output in the early 1970's the industry is reputed to have grown by more than fourfold to 4 million t* in 1988 (Megisson, 1990). At an average price of US$ 500/t, this would be worth about US$2 billion. Because of the profitable use of feed for high value species the aquafeed industry, and fish farmers persuaded by feed salesmen, have recently turned their attention to the use of feed for the production of other species of lower market value.
Aquafeed is now available for carp, tilapia, milkfish and other similar species but these feeds still rely on fish meal for protein and energy although at lower levels. Fish meal is becoming scarce and expensive. Can non-fish meal ingredients be found as possible replacements? Is fish meal indispensible? Is the use of non-fish meal-based feed formulated and prepared on the farm economical and profitable?
On-farm feed economics
The preparation and use of farm-made aquafeeds can benefit from close observations of the feeding behaviour of the fish under natural conditions. The types of natural fish foods available in the wild and fish preference, feeding frequency and quantity consumed, are excellent nutrient/dietary indicators and an economic basis for the systematic evaluation, preparation and use of on-farm aquafeeds using locally available non-fish meal raw materials.
In the wild, fish do not completely depend on other fish for nutrition; other animal and plant organisms also form part of the diet. The aquatic food web is complex and interdependent. A balanced nutritional regime is more appropriate than one based on a single food item. Also information available from both private and publicly supported research and development should be used in
* Editors' note: however, we believe this to have been an over-estimate (see New and Csavas 1993). formulating on-farm aquafeed. Such information can be applied to either grow the necessary feed ingredients on the fish farm itself or to gather them from the wild. or to purchase them from suppliers.
Once the available ingredients and their nutrient profile are known, including their pro- and anti-nutritive factors, the economics of on-farm feed production must be calculated and its use compared to commercial aquafeed. Such economic comparison must be made against the background of the farm gate value of the fish produced, which may differ according to the quality of the final product and be affected by the type of feed used. Different feeds may affect the flavour and texture of the fish flesh produced.
The unit cost of on-farm feed production, unit price of commercial feed and the cost of fish production using each type as well as the unit profitability of each system of fish production must be compared before one type of feed is selected. Growth and production performance of the fish cultured using the two different feeds have to be taken into account.
The environmental cost reflected in the social cost of using each feed must also be closely examined before one is selected. There are potential tradeoffs between the two types of feeds in terms of the use of raw materials and their affect on the environment, for example, the fisheries on which fish meal reduction is dependent. Shrimp feed cost is projected to increase due to the growing competing demand for the largely stagnating supply offish meal caused by the overfishing of the fisheries resources. Thus greater effort should be exerted to find or develop substitute ingredients. Mussels and soybean meal, among others, have been evaluated highly promising fish meal substitutes for aquafeed in Thailand. In Bangladesh, fish meal is unavailable and many other human foodstuffs such as slaughterhouse waste, including fresh coagulated blood, have successfully been used as fish feed, especially for fry and fingerlings.
Feed ingredient interchangeability
Feed ingredient interchangeability refers to the ability of feed ingredients to become partial or complete substitutes for each other. Of the 40 or so essential nutrients required by fish (including crustaceans), precise knowledge on quantitative (and possibly qualitative) requirements is available for only a few tropical species (World Bank 1991).
Methodologies are now available not only to assess the nutrient requirements and nutritional status of fish but also to assess the nutrient profile and bio-availability of feed ingredients such as digestibility, biological value of proteins, serum or tissue nutrient or enzyme concentrations and activities (World Bank 1991) and potential pro- and anti-nutritive factors. Finding inexpensive substitutes for fish meal and other expensive or increasingly scarce ingredients for fish
feed remains an important task for fish nutritionists and aquaculturists. Fortunately, data compiled on raw materials suitable for livestock feed manufacture are also useful for evaluating their use in fish feed. This is especially so for onfarm feed preparation. Much of the scientific literature on livestock feed ingredients is in the public domain.
There is in fact voluminous information on ingredients for the formulation and preparation of on-farm fish feed. The main problem lies in how to get it to the farmers. There is also no lack of information on overcoming antinutritive factors of candidate ingredients through processing, either by physical, chemical or mechanical means. Numerous studies have been done on the replacement of fish meal by plant proteins in aquaculture feed. Also, as mentioned above, slaughterhouse waste is an excellent source of animal protein.
Training fish farmers on how to utilise such information is also of equal, if not more importance. This aspect is the least developed area of aquaculture extension in many developing countries. It has been overlooked until recently partly due to the availability of commercial aquafeed. It thus calls for greater government attention to upgrade fish farmers' skill in being self-sufficient in the different aspects of fish farming.
On-farm aquafeed suffers from certain limitations not true of commercial feed. For example, it seldom, if ever, benefits from the normally wellsupported feed and nutrition research and development programmes of the commercial feed manufacturers.
On-farm feed producers usually rely on whatever published information is available, which is usually scanty being highly proprietary in nature. This is in direct contrast to the information available on livestock nutrition and animal husbandry. Public domain information is produced by publicly-supported research and development in government fish feed laboratories or universities. Although not proprietary, the usefulness and degree of detail of such information is frequently limited. It is also well-known that feed companies going into joint-venture in other countries do not reveal their entire formulation. They retain certain ingredient “mixes” which have to be imported from their country of origin. It has frequently been alleged that, even when they seem to share their formulation, it is usually their old formula.
Farm-made feed is variable in quality and generally found to be more costly per unit of nutrient based on growth and final fish production, that is profit performance. It seldom benefits from economies of scale which can reduce production cost. Also, for those feed ingredients which have to be bought, the quantity discounts available for bulk purchases cannot be obtained. Such disadvantages, however, are counter-balanced by the relatively low cost of grinding and milling (inclusive of capital cost).
The pros and cons of farm-made aquafeed vis-a-vis commercial feed discussed so far do not necessarily imply that the former is unattractive. The problems farm-feed faces can be overcome by making more information available to fish farmers.
It is reported that the feed processing cost of commercial feed is very high in relation to other cost components like feed ingredient cost. For farm-made feeds, ingredient quality and cost is the most important consideration. Very little information is available on the proximate chemical analysis of potential less-conventional ingredients, which are not traded in the market place but are available in the vicinity of the farm.
Many commercial aquafeeds are over-formulated and are thus expensive, beyond the capital or financial capability of small fish farmers. Many farmers do not apply any production inputs, either those which increase or maintain yield. In many developing countries, most fish farms are owned by small farmers and are grossly underutilised. By encouraging them to use more inputs such as fertilizers and feeds (e.g. farm-made aquafeed), much greater output can be obtained thus increasing the supply of fish.
According to Akiyama (1991), currently available commercial feeds are believed to be using excessive amounts of protein for energy. He also stated that total phosphorus levels are excessive and not highly available to aquatic animals.
Having considered all the above factors, the remaining consideration is the nutritional and economic performance of the two kinds of feeds: farm-made versus commercial.
ON-FARM AQUAFEED FORMULATION
Ingredients and formulation
To encourage fish farmers, especially small fish farmers, to produce their own aquafeed, the cost must be low and its quality comparable to commercial feed. Thus, the promotion of farm-made feeds concerns sourcing locally available ingredients to provide cheap protein, carbohydrate and fat without excessive sacrifice of the required amino acid requirements of the fish. The protein content of such feeds should be utilised for growth and not to provide energy, which should preferably come from the fat and carbohydrate component. The aquafeed formulated should fulfil the nutritional requirements of the fish being grown.
Non-fish meal feed ingredients
Major candidate feed ingredients consist of macrophytes either of aquatic or terrestrial origin, agricultural, industrial and household/restaurant wastes and by products as well as animal organisms either of aquatic or terrestrial origin.
Kitchen ponds and live feed
During the author's survey of milkfish farmers in Taiwan, Province of China, the Philippines and Indonesia in the late 1970's and early 1980's fish farmers purchased “lumut” or filamentous algae and aquatic plants from other fish farmers who specialised in their cultivation. This is especially widespread in the Philippines where a “jeepney” load of “lumut” can often be seen. “Lumut” is also purchased to feed to siganid fish in the Philippines. The little snails, aquatic organisms, insects and plankton trapped in the “lumut” also provide valuable nutrients to the fish.
The concept of the kitchen pond where live food is produced to feed the fish constitutes a farm-made aquafeed option which merits further research and development. In Thailand, Artemia and other microscopic aquatic organisms are cultured in special ponds and the enriched pond water is then pumped into grow-out fish ponds as feed. This is actually an extension of pond fertilization during pond preparation. Further, there are also commercial formulations of such organisms available for the inoculation of pond water to boost planktonic food.
Plant origin feed ingredients
Feed ingredients of plant origin which have been traditionally used in farm-made aquafeed include the following:
grains and bran (e.g. broken rice and rice bran);
oil cakes and meals (e.g. copra cake and soybean meal);
tapioca (Cassava manihot) leaves;
Ipil ipil (Leucaena leucocephala) leaves;
brewery wastes and byproducts.
Only certain species of fish, with specialized anatomical and physiological mouth parts and digestive apparatus for food ingestion, mastication, digestion and absorption can make full use of such feed ingredients.
In his review of the use of non-conventional plant feedstuffs in fish feed, Wee (1991) found that it is not possible to utilise them at high levels without comprising growth and production. In other words, such feedstuffs, which are low in protein and fat, cannot completely replace such conventional feed ingredients as fish meal. However, there are methods to overcome such deficiencies or to enhance their nutritive values, by proper processing to increase the bioavailability of nutrients and reduce anti-nutritional factors and by the inclusion of appropriate additives. Soybean meal is a good example of a byproduct (of soybean oil extraction) which has been successfully processed into a feed ingredient which is increasingly becoming a partial substitute for fish meal in aquafeeds.
According to Akiyama (1991), aquatic animals are more sensitive to feed quality than terrestrial animals. This implies that closer attention must be given to aquafeed formulation, especially when fish farmers are preparing them. At the same time, the ingredients must also be cost-effective.
As noted above, aquafeeds are costly or expensive inputs for fish culture. The demand for aquafeed at present is still largely induced by feed suppliers. As such, fish farmers have to guard against overfeeding caused by too closely following the feeding recommendations of feed suppliers. Fish farmers are well advised to closely scrutinize their feed costs to improve their profits.
In this paper, a distinction has been made between commercial aquafeed and farm-made aquafeed. Commercial aquafeed has long been accepted as a necessary production factor in fish farming, in spite of its high cost. Such aquafeed is generally used in the production of high-value species for which it is economical and profitable.
The high cost of high-protein fish meal-based commercial aquafeed is still generally recognised as the main constraint to increased fish output from culture. Because of growing scarcity of fish meal due to overfishing and (traditional) competitive demand for fish meal for livestock feed and other uses, alternative aquafeeds and/or sources of non-fish meal ingredients are being actively explored. One of these is farm-made feed using locally available ingredients.
The economic importance of feed on final farm profit is too critical to be taken lightly. At present, commercial aquafeeds cost an average of US$ 0.50/kg and feeding accounts for an average of 30-60% of production costs. As such, profit is highly sensitive to feed cost. Every improvement in feed conversion efficiency means a bigger profit margin.
The purchase and use of commercial aquafeed remains very much a supplier-induced demand. As such feed cost, as a cost centre in the overall fish production costs, remains outside the control of fish farmers. Feed prices are determined by the feed suppliers; fish farmers, especially small-scale operators, have little or no influence on the final determination of feed prices. Furthermore, feed suppliers encourage fish farmers to use their feeds by providing samples for feeding trials, extending credit and easy terms of payment and other such product promotional schemes.
Because of this supplier-induced demand, there is a real tendency for fish farmers to over-feed, either knowingly or unknowingly. This arises because farmers tend to follow feed suppliers feeding schedules and their guidelines on quantity and frequency. In addition, few farmers keep records as a means to evaluate their farm performance and even if they do so, seldom critically analyse the data.
The economics of farm-made aquafeed has been discussed and found to be not only a viable alternative to commercial aquafeed but one which is environmentally sustainable and inexpensive.
LIST OF REFERENCES
Akiyama, D.M. 1991. Future considerations 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 held in Thailand and Indonesia, 19-25 September 1991. American Soybean Association, Singapore.
Akiyama, D.M. and R.K.H. Tan (eds.) 1991. Proceedings of the Aquaculture Feed Processing and Nutrition Workshop held in Thailand and Indonesia, 19-25 September 1991. American Soybean Association, Singapore. 241 p. Anon. 1988. Fish feeds. Infofish International 3/88:18-20.
Bostock, T. 1991. Better feeds for small-scale shrimp farmers. Bay of Bengal News 42:22-26.
Chong, K.C. 1992a. The DGF fisheries planning model, FAO/UNDP Project on Assistance in Agricultural Planning (INS/88/001), Food and Agriculture Organization of the United Nations and Directorate-General of Fisheries (DGF) and Bureau of Planning, Ministry of Agriculture, Jakarta, Indonesia.
Chong, K.C. 1992b. Improving profitability of shrimp aquaculture. Paper presented at the Asian Productivity Organization (APO) Seminar on Aquaculture, 25 August -4 September, 1992, Tokyo, Japan.
Crampton, V., H. Kreiberg and J. Powell. 1990. Feed cost control in salmonids, p.99-108. In Proceedings of the International Congress on Aquaculture, 4-7 September 1990, Van-couver, British Columbia, Canada.
Cruz, E.M. 1992. Preliminary assessment of the production and economics of culturing common carp (Cyprinus carpio) in running water system in North Sumatra, Indonesia. J. Aqua. Trop. 7:7-14.
Csavas, I.1991. Physical and nutritional properties of aquafeeds. Paper presented at VICTAM-Asia'91 Symposium, Feed Production Tomorrow, 24-26 January 1991, Bangkok, Thailand.
De Silva, S.S. 1989. Reducing feed costs in semi-intensive aquaculture systems in the tropics. Naga, The ICLARM Quarterly 12(4):6-7.
De Silva, S.S. (ed.). 1991. Fish nutrition research in Asia. Proceedings of the Fourth Asian Fish Nutrition Workshop, Asian Fisheries Society, Philippines and International Development Research Centre, Canada. 205 p.
FAO. 1979. Training in fish feed technology. A report of the FAO/UNDP Training Courses in Fish Feed Technology held in the University of Washington, Seattle, U.S.A., 9 October-15 December 1978. ADCP/REP/79/8. FAO, Rome, Italy. 16 p. Megission, P.A. 1990. Future world of aquaculture. Paper presented at Rovithai International Shrimp Seminar, July 1990, Bangkok, 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. Ratafia, M. and T. Purinton. 1989. Emerging aquaculture markets. Aquaculture Magazine 15(4):32-46. Tacon, A.G.J. 1992. The food and feeding of farmed fish and shrimp: an annotated selection of FAO Field Documents, 1973-1991, FAO Fisheries Circular No. 849, FIRI/C849. FAO, Rome, Italy. 92 p.
Tacon, A.G.J. and C.B. Cowey. 1985. Protein and amino acid requirements, p. 155-183. In P. Tytler and P. Calow, (eds.) Fish Energetics - New Perspectives. Croom Helm Press Ltd., London.
Wee, K.L. 1991. Use of non-conventional feedstuff of plant origin as fish feeds - is it practical and economically feasible? p. 13-32. In S.S. De Silva (ed.) Fish Nutrition Research in Asia. Proceedings of the Fourth Asian Fish Nutrition Workshop, Asian Fish. Soc. Spec. Publ. 5. Asian Fisheries Society, Manila, Philippines.
World Bank. 1991. Tropical aquaculture development: research needs, World Bank Tech.Pap. No. 151 (Fisheries Series), The World Bank/United Nations Development Programme/ Commission of the European Communities/Food and Agriculture Organization, Washington DC., USA. 52 p.