This is the old United Nations University website. Visit the new site at

Back to the table of contents page

Discussion lll

The assumption that cassava or its residues could be used in bioconversion processes to make animal feed is not well founded. Surprisingly little waste is available from cassava production and, in any event, cassava's main use in South-East Asia is for direct human consumption.

Microbiar treatments of cassava are known and practiced at the village level, but as a means of enriching the cassava with protein for human use. If significant surpluses become available it will be necessary to decide what to do with them. Cassava could be used to produce food, feed, or fuel according to what is required. If calories alone are needed for food or feed, there is no need for enrichment processes. The need will arise if protein production is the objective.

Cassava leaves were mentioned as possible feed ingredients. There is a danger in using them without some form of treatment to decompose the cyanogenetic glucosides they contain. One method is to allow them to ferment naturally, another is to inoculate them with a fungal preparation. Both methods are said to reduce the HCN content of the leaves to 5 per cent of that originally present, leaving less than 50 mg per kg in the final product. This is considered a safe level for use in feeds.

It did not escape notice that more than 2 million tons of cassava are being exported each year from South-East Asia to Europe to be used in animal feeds. The suggestion was made that this could be used in South-East Asia for protein enrichment and thus reduce soybean imports.

The use of molasses as a substrate for single-cell protein production was thought to be unsuitable for village conditions because of the equipment needed to harvest the SCP. On the other hand, it was felt that enriching the molasses with protein by growing an organism on it and using the total biomass could be feasible. It is, of course, possible to use molasses for the production of power alcohol, but this is not always encouraged by the authorities because alcohol can be put to purposes other than fuel.

The Philippine work on banana rejects excited considerable interest, not least in the manner of getting the technology into the villages. This is done by collaboration between the research team, farmers' cooperatives, and an organization called BLISS (Bagong Lipunan Integrated Sites and Services; bagong lipunan means "new society")

The problems of handling agro-industrial wastes were discussed. It was felt that these could, in practice, be solved only by research and development work in the industries themselves.

A general comment, applicable to all fermentation processes, concerned the upsetting of the ecological balance of micro-organisms in the vicinity of operations where there could be a build-up of substrate in and on the equipment being used. In order to detect and control a concentration of spores that could be toxigenic, a regular screening of the microbial population is advisable. This could be done by the team that had developed the process or by other suitably qualified persons.

Utilization of trash fish and fish wastes in Indonesia

Production of fish silage
The nutritional value of fish silage
Proposal for further research

I. Putu Kompiang
Ciawi, Bogor, Indonesia

The total amounts of fish wastes and trash fish available in Indonesia are not recorded in official fisheries statistics. However, the estimated minimum amount in 1976 was 450,000 tons. Unfortunately, only relatively small quantities are produced at each landing port.

Trash fish are found only in areas where there is shrimp trawling, such as in the Arafuru Sea, the Java Sea, and the Indian Ocean. Fish wastes are produced as byproducts of processing plants. ln 1978, 63,000 tons of frozen shrimp were exported, but the percentage of wastes that came from shrimp culture or from fishing in the open sea was not recorded. It is also not known how much of the total waste is supplied by small-scale traditional fishermen, who generally utilize the by-catch themselves, and how much comes from the larger trawlers. The ratio of shrimp to trash fish is generally very low, being about 1:5 during the season and as low as 1:20 in the off-season. It has been estimated that a minimum of 200,000 tons of trash fish per Year is returned to the sea as a result of shrimping in the Arafuru Sea, and 27,000 tons per year in Central Java, of which 600 tons is by fishing boats from Tegal. Most of this trash fish is dumped into the sea because the fishermen prefer to keep space available for possible large catches of shrimp.

In addition to trash fish, large gluts occur at times from the sardine industry in the Bali Strait. Mechanization of the boats and recent introduction of purse seining has resulted in greatly increased catches. During the heavy fishing season, which usually lasts for at least 30 days, the daily surplus is about 150 tons, giving a total excess of 5,000 tons.

Fish wastes are mainly produced at scattered processing plants. The estimated waste from the canning factory at Muncar, East Java, is 1,600 tons per year, and from Bali, 450 tons per year. Wastes consisting of gills and guts from the fish-freezing plant in Menado, North Sulawesi, amount to about 1,000 tons per year.

Wastes from shrimp-processing plants amount to 8,000-10,000 tons per year, and from frog-processing plants, 7,000 tons per year. As a result of mishandling there are also wastes at all landing ports. Lack of ice for preservation, and transport difficulties are the main causes of this loss. It has been estimated that 15 per cent of the catch, or 200,000 tons per year, is wasted in this way.



The Ministry of Agriculture has strongly emphasized that fish should be used directly for human consumption. Fish-meal production for livestock feeding should therefore be restricted to the use of fish wastes. Unfortunately, this is often not possible due to daily and/or seasonal variations in the size of the catch, transport difficulties, and/or inadequate processing facilities. Thus, complete utilization of fish resources is rarely possible in the tropics.

As a consequence of these seasonal and day-to-day fluctuations and the availability of only small quantities of waste fish and fish wastes at any one location, the production of fish meal is usually commercially unattractive. In spite of this, there are some small, cottage industry fish meal plants. Their operations are not efficient and there is no drying equipment. The fish is boiled or steamed and then pressed. The presscake is sun-dried and the press liquid (stickwater) is discarded. Up to 40 per cent of the protein is lost (table 1); if the raw material is not fresh, as is often the case, the loss is even greater. Stickwater disposal also creates environmental problems. The cheapest way of drying the presscake is in the open sun. Drying is often difficult, particularly in the wet season, so local fish meal is often of poor quality. The moisture content of these meals is generally above 10 per cent, with some samples as high as 17 per cent (table 2). The protein concentration is variable, ranging from 31 to 51 per cent. Good fish meal should have less than 12 per cent moisture to prevent the growth of moulds and should contain not less than 50 per cent protein.

TABLE 1. Percentage of Presscake from Boiled, Fresh, and By-catch Fish



Wet weight


Dry weight








Press-liquid percentage composition equals 100 minus presscake percentage.

TABLE 2. Chemical Composition of Indonesian Fish Meals

Sample No.
















































In some areas, trash fish wastes and frog wastes are fed directly to ducks. In North Sulawesi, fish wastes are used mainly as fertilizer. A fermented fish sauce, locally known as bakasang, is produced from fish guts. Some fish and shrimp wastes are used in fish paste.


Production of fish silage

The poultry industry in Indonesia is growing rapidly (table 3). Fish meal and soybean meal are the main sources of protein for poultry feed, but the supply is inadequate and some feed must be imported. A cheap and effective method of preserving fish wastes and trash fish would benefit Indonesia.

Following an FAO feasibility study in 1976, a group was set up to investigate the use of trash fish as fish silage for animal feed. Fish silage has many advantages: (i) the process is virtually independent of scale; (ii) the technology is simple; (iii) the capital required is small, even for large-scale production; (iv) effluent and odour problems are reduced; (v) production is independent of weather; (vi) silage can be produced aboard ships, so trash fish do not require chilled storage, and (vii) the ensiling process is rapid in tropical climates.

TABLE 3. Population and Production of Indonesian Poultry




Average Annual

Increase (%)

Population (millions)


Village chickens




Commercial chickens








Production (thousand tons)


Poultry meat




Eggs from


village chickens




commercial chickens








Some disadvantages should also be considered: (i) silage is a bulky product that causes storage and transportation difficulties; (ii) many tropical fish have a high oil content (e.g., Bali Strait sardines contain up to 25 per cent oil), which complicates the use of silage and may give an oily taint to the flesh of animals consuming it.

Fish silage can be prepared by adding minerals or organic acids (chemical silage), or by microbial fermentation supported by the addition of carbohydrate (biological silage). I have used both methods successfully and have evaluated their nutritional value. Formic acid and propionic acids were used for the production of chemical silage. Three per cent (w/v) of a 50:50 mixture of formic acid and propionic acid recommended by Gildberg and Raa (2) was needed, probably because of the high buffering capacity of our fish.

The addition of propionic acid prevents the growth of mould. This preservative action was maintained when the silage was mixed with a carbohydrate carrier. Moist mixtures of equal amounts of silage and cassava or corn remained free from moulds for at least three months at room temperature (30C). Without the addition of propionic acid, the moist mixture usually became mouldy, bacterial growth caused a pH increase, and putrefaction followed within a few weeks (3).

Chemical silage was found to be very stable, and storage for 21 days caused no significant change in the relative concentrations of the amino acids, although there was 1.3 per cent amino nitrogen loss through ammonia production.

TABLE 4. The pH Value of Biological Fish Silage in Indonesia

Fish/Molasses Fermentation Period (Days)
Ratio 0 3 7 14 21
20:1 6.9 5.0 5.5 - -
10:1 6.8 4.6 4.5 4.5 4.8
6.7:1 6.7 4.5 4.3 4.4 4.4
5:1 6.6 4.4 4.3 4.3 4.3

Figures are averages of six observations.

TABLE 5. Chemical Composition of Chemical and Biological Fish Silages after 21 Days of Storage

  Moisture Protein Soluble Proteina NH3-Nb pH
Chemical stagec 62.5 18.0 81.3 1.4 3.6
Biological silage          
10:1 molasses (w/w) 64.0 17.2 25.6 16.0 4.76
6.7:1 molasses (w/w) 63.2 17.3 24.9 11.6 4.40
5:1 molasses (w/w) 63.0 14.7 25.7 14.8 4.33

a. Percentage of total protein.
b. Percentage of total nitrogen as ammonia.
c. 100 kg fish: 1 5 litres formic acid: 1.5 litres propionic acid.

Biological silage was prepared by natural fermentation after adding molasses to the minced fish. The minimum ratio of fish to molasses for a stable silage is 10:1 (table 4); however, a ratio of 10:1.5 is recommended. After six months in storage, the silage was organoleptically stable and had a fresh, acid smell. As with the chemical silage, ammonia is produced during storage; after 21 days 10 per cent of the total nitrogen is ammonia, similar to the level found by Rydin (personal communication). The compositions of chemical and biological silage are shown in table 5.


The nutritional value of fish silage

The nutritional value of chemical silage when fed to pigs (4) and to common carp (5) was the same as that of the fish-meal control. Similar results have been reported by other workers. However, when chickens in Indonesia were fed at a high level (23 per cent dry silage) compared with usual fish meal diets, the nutritional value was inferior to that of fish meal (table 6). The factors that inhibit the growth of chickens fed chemical silage may be associated with the lipid fraction (3).

TABLE 6. Performance of Chicks Fed Fish Silage (One-Week-Old Chicks Fed for Three Weeks)

Treatment Weight Gain (g/3 weeks) g Feed/g Gain
Fish meal (23 %) 554a 1.75b
Chemical silage (23 %) 364c 1.96a
Biological silage (23 %) 462b 1.90a

a, b, and c indicate significant differences (p < 0.05).

The nutritional value of biological silage, even though inferior to fish meal, was significantly better than that of chemical silage. This is difficult to explain, but it has been reported that some antibiotics and B group vitamins are produced during microbial fermentation (6-8). When biological silage was used at normal levels (8 per cent dry silage in the ration), the body weight gain and feed efficiency were similar to results from fish-meal feeding (9).


Proposal for further research

The Indo-Pacific Fishery Council has suggested that the core research on fish silage in the region should be concentrated in Indonesia. The proposed research includes: (i) production of bulk quantities of fish silage, (ii) incorporation of silage into livestock feed, and (iii) modification of trawlers for silage production.

We are currently studying the production of chemical silage and the use of other locally available acids. Since there are trawlers dumping large quantities of waste fish, the technical and economic feasibility of producing silage on board fishing vessels should be evaluated.

Other carbohydrate sources, such as rice bran or cassava waste, should also be investigated for the production of biological silage. Cassava waste is being studied at the Department of Microbiology, Institute of Technology, Bandung. At present, it appears that biological silage will be more difficult to prepare than chemical silage on board ships. However, as the nutritional value of biological silage is greater than that of chemical silage, research on biological silage should be continued using on-shore fish wastes.

Studies to improve the nutritional quality of all silage should be continued.



1. J. Sumner, Proceedings of the Indo-Pacific Fishery Council Working Party on Fish Technology and Marketing. Colombo, Sri Lanka, 1976 (FT/76/4).

2. A. Gildberg and J. Raa, in J. Sci. Food Agric., 28: 647-653 11977).

3. I.P. Kompiang, R. Arifudin, and J. Raa, in Proceedings of the International Conference on Fish Science and Technology, Aberdeen, Scotland, UK, 1979.

4. L. Batubara and M. Ranguti, in Proceedings of the Indo-Pacific Fishery Council Working Party Meeting on Fish Technology and Marketing and Workshop on Fish Silage, Jakarta, Indonesia, 17-21 Sept. 1979.

5. D. Hidayat and D. Roestami, in Proceedings of the Indo-Pacific Fishery Council Working Party Meeting on Fish Technology and Marketing and Workshop on Fish Silage, Jakarta, Indonesia 17-21 Sept. 1979.

6. H.C. de Klerk and l.N. Coetzee, Nature, 192: 340 (1961).

7. P. Reeves, Molecular Biology, Biochemistry and Biophysics, vol. 2: The Bacteriocins (Springer Verlag, Berlin and New York, 1972).

8. S. Lindgren and G. Cleustrand, in Swedish J. Agric. Res., 8: 61-66 (1978).

9. I.P. Kompiang, Yushadi, and D. Cressell, in Proceedings of the Indo-Pacific Fishery Council Working Party Meeting on Fish Technology and Marketing and Workshop on Fish Silage, Jakarta, Indonesia 17-21 Sept.1979.