The NIFI feedmill was originally designed around two Ace Model CPM pellet mills each of 1.5 ton/hour capacity. The mill was intended for the manufacture of dry, sinking-type pelleted feed for fish. Design of the plan and ancillary equipment was carried out by staff of the Agricultural Engineering Department of Kasertsart University. With the exception of the two pelleting machines and a small steam boiler, all other components and structures of the mill were of local design and fabrication.
Figures 1 and 2 in Appendix II show the equipment layout of the mill.
The mill was designed to operate as follows : ingredients to be mixed are first weighed and then loaded manually with shovels, at ground level, into a heavy steel buck. The bucket is hoisted up by a motor-powered pulley that is installed on an I-beam over a feed hopper, and moves laterally along this beam under the control of the operator on the ground. One manoeuvered above the hopper, the bucket is made to discharge its load by a worker positioned beside the hopper. Material that has been dumped into the hopper passes through a scalper before it enters the vertical mixer. The scalper removes foreign contaminants such as gunny strings and scrap matter. Mixing takes about 20 minutes, after which the feed is transferred by gravity through the discharge sleeve of the mixer directly into the steam-conditioning compartment of the pelleting machine. Steam is provided by a small package-type steam generator. Hot, moist, pellets emerging from either of the two pelleting machines fall on a small vibrating screen to remove fines. The pellets roll off the slightly sloped screen into the lower leg of a bucket-type elevator and are then discharged into a cooler/dryer. Moisture is removed as the hot pellets fall by gravity against streams of air radiating from the perforated inner cylinder of the cooler. The dry pellets collect at the bottom of the cooler/dryer and are removed through four openings.
The design of this mill appeared to be based on flow characteristics of cereal grains and is probably a scaled-down version of an earlier type grain-drying facility. Such design, however, is not suitable for pelleted feed production. Fresh pellets emerging from the pelleting machine are too moist and soft to be screened without disintegration and clogging up of the sieve. Transport of hot, moist pellets by bucket elevator is generally not recommended for the same reason. The fact that the buckets in this case are actually flat slats renders the conveyor even less suitable for that purpose. Fines that are screened out have to be returned manually to the pelleting machine for reprocessing. The cooler/dryer is also ineffective because of uneven distribution of hot pellets as they enter the cooler, and the rather short drop through the air streams. Futhermore, the uncontrolled dissipation of moist, dust-laden air from the cooler into the external environment is not desirable.
At the time of the author's visit, the mill was not used as designed due to reasons explained earlier.
Handling of feed ingredients for mixing and pelleting were as described. Only one pelleting machine was in use. This was apparently due to inadequate steam supply for both mills to be operated together.1 The complete absence of insulation for pipes leading from the steam generator to the steam-conditioning chamber of the pelleting machine further aggravated the situation. Condensation occuring along the uninsulated pipes not only reduced the quantity of the steam delivered to the pelleting machine but also reduced steam quality as well. The present supply of steam may just be sufficient for slow production of small diameter pellets on one pelleting unit. When used for making large-size pellets (4 mm diameter), the author observed that the finished product lacked strength and was very friable. Drying and cooling was done on the floor on which pellets emerging from the pelleting machine were allowed to fall and then spread out. This did not permit the pellets to cool rapidly enough or to lose all excess moisture.
In proposing changes for improving the mill, the author took into consideration two major factors: recent developments in compact mill design, and cost of remodelling. In view of the considerable expense anticipated for carrying out the necessary engineering work, the author also met with potential donors within the Thai feed milling industry who had expressed a keen interest in aquaculture development in the country, and who had, in the past, supported Government research programmes. in the livestock sector. Successful identification of a possible donor resulted in two alternatives for remodelling the mill.
The first alternative, (A), is provided by the author, and is based on the state of the art in feed milling technology, utilizing as much as possible existing machinery components.
The second alternative, (B), was provided by the prospective donor and is based on a simpler design but involves considerably more new machinery.
The major differences between the two designs lie in material transport and pellet cooling. The first employs a pneumatic conveyance (air lift) system that not only transports material and product, but also provides air for cooling and drying of pellets. The second alternative uses a bucket elevator for transport of feed to the pelleting machine, and pellets are cooled and dried in a vertical-type cooler/dryer. This alternative also does away with one of the pelleting machines. Unique features of each of the two designs are detailed further as follows:
Alternative (A)
Basic to this proposal is the installation of a positive pressure pneumatic conveyor system2 of the type found in large, integrated mills, as well as in recent, small, package-type pellet mills. The unique feature of the proposed system is the sharing of the system for transferring mixed feed from the vertical mixer to the meal-holding bin for pelleting operation and conveying the freshly made product from the pelleting machine to the cooler/dryer. This is made possible by locating the Venturi intake valve of the pneumatic system at a position where the mixed feed or product can enter without mechanical assistance.
Apart from the construction and installation of the pneumatic system, complete with air cyclone, the proposal also calls for the fabrication of two meal bins (one serving each pelleting machine) and the relocation of the vertical mixer and cooler/dryer. The cooler/ dryer will also be modified for link-up with the pneumatic system and should include a conical extension at its base for collection and storage of the finished product.
Figures 3 and 4 show the equipment layout of the modified mill (A)3.
The modified mill (A) will operate as follows:
After weighing, pulverized feed ingredients are dumped into the intake hopper of the vertical mixer. The mixed feed is discharged by gravity from the side of the mixer, via a sleeve connection, into the Venturi intake valve of the pneumatic conveyor system. The meal is taken up past the cyclone and distributor valve set to divert the meal to either of the two meal bins. When the mixing operation and meal transfer is completed, the sleeve linking the mixer to the Venturi valve is disconnected and another sleeve connecting the valve to the outlet of the pellet mill will be put in its place. The emerging pellets are taken up the pneumatic pipe past the cyclone where the bulk of excess moisture is removed and where significant cooling of the pellets also takes place. Leaving the cyclone, the pellets are directed into the modified cooler/dryer which completes cooling and moisture removal of the finished product. The dry pellets collect in the conical extension of the cooler/dryer and may be discharged into bags or other containers for immediate use or permanent storage.
The proposed alterations to equipment and flow process will result in high-quality products. Cooling and moisture removal from freshly made pellets will be rapid and complete. This is largely due to the passage of hot pellets through the air cyclone. Reversal of the direction of air streams into the inner cylinder of the cooler/dryer also increases the volume of cool air available since the surface area (and therefore number of perforations) of the wall of the cooler/dryer far exceeds that of the inner cylinder. The reversal of air ventilation also does away with the environmental hazards associated with the original operational scheme. Furthermore, dust and fine feed particles will be sucked up from the inner cylinder of the cooler/dryer and aggreagated with the sticky pellets as the latter emerge from the pelleting machine.
Operation of the modified facility will be considerably simpler. It is possible for just one skilled worker to operate the entire mill - from loading ingredients to discharging of the finished product.
Alternative (B)
This alternative, proposed by an engineer of the prospective donor company, is more conventional. Use is restricted to one of the two pelleting machines. The proposal employs a bucket elevator and gravity as the principal means of material transfer.
New constructions include: (a) a 10-m high bucket elevator for transfer of mixed feed to a holding bin; (b) a holding bin and, (c) a vertical cooler fitted with a cyclone dust collector. As in alternative (A), the vertical mixer will be brought down to ground level. The pellet mill, however, will be installed on an elevated, sturdily structured frame so that freshly manufactured pellets will enter by gravity into the cooler immediately beneath the pelleting machine. No provision is made for immediate storage of the finished product.
5 shows equipment layout of the modified mill (B).
The modified mill (B) will operate as follows
Pre-weighed feed ingredients are loaded into the vertical mixer, as in alternative (A). The mixed feed is then discharged by gravity through the side sleeve of the mixer into the lower leg of the bucket elevator which then discharges it into the meal bin located above the pellet mill. Feed enters the steam conditioning chamber of the pelleting machine by gravity. Freshly made pellets fall immediately into an automatic cooler/dryer. Air is sucked through vants in the cooler by means of a fan/cyclone system. The cooled pellets have to be removed constantly for uninterrupted operation of the pelleting machine.
The pellets produced, although of good quality with respect to hardness, will not be free from fines since there is no provision for screening and recycling of the latter. However, like alternative (A), the entire operation will be essentially dust-free.
The operation, although simple, will require more workers than in the case of alternative (A). One skilled worker must be stationed on the elevated platform at all times during the pelleting operation. One or two workers will be needed to bag the finished product as it leaves the cooler. As in alternative (A), feed mixing should preferably be completed before the pelleting operation begins.
Cost consideration of the two alternatives
The estimated cost of modifying the mill according to alternative (A) is approximately Baht 250 000 (US$ 11 000)4. No cost estimate was provided by the prospective donor for alternative (B), but from the amount of new constructions required (especially the automatic cooler/dryer system) the author estimates the cost to be considerably higher.
4 US$ 1.00 = Baht 23.00 (approximately)
The operation cost of alternative (A) will be lower since less labour is required. This saving will more than offset the anticipated higher electricity consumption required to power the air-lift system. Maintenance cost will also be lower since the air-lift system. Maintenance cost will also be lower since the air-lift system is self cleaning whereas the bucket elevator of alternative (B) requires periodic servicing and maintenance.
Finally, because both pelleting machines will be in place in alternative (A), there will be no production loss due to regular or unscheduled servicing and maintenance of any one machine.
The nutrition laboratory was equipped primarily for proximate analysis of feeds. Other equipment included a Hobrat mixer/mincer used for the preparation of moist, pelleted, experimental diets. The laboratory is staffed by three research personnel and two technicians. At the time of the author's visit a cold storage room adjoining the nutrition laboratory was being converted for use by the nutrition unit. One staff member was conducting a microbiological assay of vitamin B12 in test diets for Ophicephalus, based on ensiled trash fish.
As mentioned in an earlier section of this report, feeding trash fish to catfish and snakehead often resulted in costly disease outbreaks. Although affected fish ultimately succumbed to bacterial infection, the clinical symptoms were often reminiscent of vitamin deficiency syndromes, particularly with regard to vitamin C and thiamine. In view of the frequency and severity of cases brought to the attention of Government fishery authorities, NIFI decided that it would be necessary to set up a laboratory to investigate the possible relationship between traditional feeding methods and vitamin status of fish. To carry out these investigations, the following equipment were purchased with assistance from the project; a full-range spectrophotometer, a fluorometer and a refrigerated centrifuge.
As an initial step, investigations were limited to the study of levels of the following vitamins in blood: vitamin C, thiamine and pyridoxine. While vitamin C would be measured directly in blood plasma, indirect measurements of thiamine and pyridoxine were to be carried out instead. Fish were either from NIFI pond stocks or purchased in the market.
Blood sampling and handling
Fish were placed in a deep pan usually 40 cm dia., and containing about 3 cm water. A few drops of thiamylal sodium was then stirred into the water. Fish usually became unconscious within minutes.
Class hyperdermic syringes were used with 22-guage needles.
Effectiveness of anti-coagulant was found to be dependent on the species involved. Heparin was less effective than sodium-EDTA for rohu carp, whereas both anti-coagulants could be used for catfish, snakehead, tilapia, Pangasius and sand goby. Since Sodium-EDTA was less expensive to use, an aqueous solution containing 10 mg Na-EDTA/ml was prepared and used at the rate of approximately 0.1 ml/10 ml blood drawn. Before each sampling, some anti-coagulant was drawn into the syringe to wet the barrel and the excess anti-coagulant expelled. A small amount, depending on the quantity of blood to be collected, was placed into a clean, dry, centrifuge tube.
The anaesthesized fish was laid on a towel on its side and the needle inserted at the midline just posterior to the anus and then toward the caudal vein from which blood was drawn. Up to 4 ml of blood was obtainable from fish weighing 250 g without apparent permanent injury to the animal. After blood sampling, the fish was returned to clean water to recover. All fish appeared normal at the time of blood sampling.
Blood analysis
A small aliquot (about 0.1 ml) of the blood sample was taken in a special capillary tube for determination of hematogrit level. The rest of the blood was centrifuged to separate red cells from plasma. Plasma was analyzed for vitamin C and glutamate-pyruvate transminase activity5. Red blood cells (RBC) were washed with isotonic saline and lyzed for analysis of transkelolase activity.6
Blood was collected from the following species for analysis (number of specimens in brackets): Clarias batrachus (7): Labeo rohita (6): Oxyeleotris marmoratus (4): Sarathorodon niloticus (10): Ophicephalus striatus (3): and, Pangasius sutchi (2).
Analytical procedures for vitamin and enzyme activity determinations are described in Appendix III and the results summarized in Table 1, Appendix I.
(a) Hematocrit values
Hematocrit values ranged from a low of 19.0% for tilapia to a high of 59.8% for snakehead (see Table 1). As a group, tilapia showed the lowest average hematocrit value (31.0%) and the snakehead, with 51.9%, the highest. The other species ranged in between. Plasma of the sand goby was milky white in colour, an indication of lipaemia7.
5 Plasma glutamate-pyruvate transaminase activity was determined as a measure of pyridoxine status.
6 Erythrocyte transketolase activity was determined as a measure of thiamine status
(b) Plasma vitamin C
Normal plasma vitamin C ranged from a low of 1.37 to 6.70 mg/100 ml. The rohu carp appeared to have consistently higher plasma vitamin C levels.
(c) Plasma glutamate-pyruvate transaminase (GPT) activity
Plasma-GPT ranged between 195 and 408 units/litre. There did not appear to be species differences. Although pyridoxine is an essential co-factor for GPT interpretation of blood GPT levels should be done with caution since they are linked to other health parameters, especially those concerning the liver. Hence liver or biliary infections will raise plasma-GPT levels high above normal and mask incipient deficiencies of the vitamin. Plasma from haemolyzed blood also gives higher than expected values. Direct measurement of the vitamin will be more acceptable for ascertaining pyridoxine status in fish.
(d) Erythrocyte transkelolase activity
The appearance of hexose and disappearance of substrate pentose in the assay system is a measure of transkelolase activity. Pentose metabolism appeared to be most pronounced in RBC of snakehead and least in RBC of tilapia. The disappearance of pentose, however, did not appear to correlate with increased hexose appearance. This could be due to differences in metabolic status among individual fishes. Values obtained by direct measurement of thiamine using the fluorometric method would decide the reliability of using the more rapid transkelose assay as a diagnostic tool for studying thiamine status in fish.
Work was started on the analysis of thiamine in feedstuffs using the thiochrome method, employing the newly installed spectrofluorometer. A sample of rice bran analyzed gave a value of 23.69 ug/g, fairly close to the value of 22.5 reported by the United States National Research Council.8
Thailand is a major agricultural producer in the Southeast Asian region. Large quantities of feed raw materials are produced and consumed each year. Estimated domestic consumption of feeds for 1980 was in excess of 11 million tons, nearly half of which was for hog production. The chicken industry consumed an estimated 4 million tons and the remainder accounted for by duck producers. Yet the country exports large quantities of feedstuffs ranging from high-valued fish meal to inexpensive tapioca chips. Although commercial fish farming in Thailand has been in existence for as long a period as its livestock husbandry development of feeding in aquaculture has not kept up with similar activities in the livestock industry. Feeding methods for Clarias culture are still largely traditional in charter and based on empirical experience involving two or three ingredients, viz., trash fish, rice bran and broken rice. Failure of the fish farming industry to fully exploit feed material resources is mainly due to a general lack of information on the nutritional requirements of the cultured species, and ignorance among fish farmers on whether and how other feed ingredients can be effectively fed to fish. Development of this knowledge is essential for aquaculture to remain viable as supplies of trash fish continue to decline.
8 Nutrient Requirements of Warmwater Fishes. NRC, Washington D.C., 1977
The country produces a large surplus of feed grains such as maize and sorghum. Domestic consumption of these feed materials is almost exclusively by the livestock and feed milling industries. Maize and sorghum have not been used for direct feeding to fish but probably constitute a significant percentage in commercially produced floating-type pelleted pelleted feed for catfish. Broken rice, as well as cargo rice, on the other hand, are used extensively by fish farmers engaged in traditional feeding methods. The choice of feed grain usage, apart from traditional preference, may be due to processing difficulties with maize, which needs to be pulverized before it can be heat processed to make water stable fish feed. Broken rice, on average, costs more than maize or sorghum, despite its year-round availability.
By-products from the grain milling and oil seed extraction industries constitute the second largest feed resource in the country.
Rice bran is the principal by-product of the milling industry and is also available year round. It is the second most important ingredient for feeding to catfish in the country. Two varieties of rice bran are available. They are raw rice bran, and solvent-extracted rice bran. Raw rice bran is preferred by catfish farmers for two important reasons: unlike solvent-extracted rice bran, which is only available in larger cities, it is available from small village mills or feed stores; and, the main mistaken belief that oil extracted rice bran is inferior to raw rice bran due to the loss of fat in the former.
By-products from the oil-seed extraction industry are, next to fish meal, the most important protein resource for the livestock industry. They include: soyabean meal, cottonseed meal, peanut meal and sesame meal. Except for commercial manufacture of compounded feeds, their use for fish feeding by traditional methods has not been reported.
Fish meal represents the most important feed resource in the country. It also represents more than half the total value of all animal feedstuff exports. In 1980 it was estimated that 1.2 million tonnes of marine fish catch were reduced to fish meal. Of the 300 000 tonnes fish meal produced, one-third was exported. Ironically, fish meal consumption by the aquaculture industry is insignificant. Instead, raw trash fish is used. There is no estimate of trash fish usage for aquaculture in the country. From best estimates of annual catfish production of 50 000 tonnes from ponds9, it would appear that about 250 tonnes of trash fish are used yearly for that purpose alone.
Other feedstuffs available are: tapioca meal, molasses, leaf meal, brewery by-products, shrimp waste and meat meal. With the expansion of the poultry processing industry, increasing quantities of chicken offal are also becoming available. Practically all of this is sold unprocessed for use in pig farms and, more recently, for snakehead culture.
Due to heavy demand by a sophisticated feed industry and good road transportation, prices generally reflect the feeding value (protein and energy) and quality of respective ingredients.
A list of commonly available feed raw materials that are suitable for fish diet manufacture appear in Table 2, Appendix I.
The animal feed milling industry is one of the country's fastest growing industries. Of the 60 or so registered feed mills, nine large integrated mills account for more than half the 3 million tonnes of compound feed manufactured annually. Despite its size, the industry caters to only one-quarter of the total feed requirements for farm animal production. The balance is met by on-farm mixing by the farmers themselves. Nevertheless, the feed milling industry plays a prominent role in the operation of the modern farm, and by its own involvement in commercial livestock production, has fostered high-quality standards for factory manufactured feeds.
Al though commercial sinking-type dry pelleted feeds for catfish were introduced at least 10 years ago, their high cost then made few converts to their use. Since then new efforts have been made by the feed milling industry, with the introduction by one mill of floating-type pelleted feed, to shift farmers away from traditional feeding practices. Feed conversions as low as 1.5 have been claimed, but comparatively higher costs associated with commercial feeds and conservatism among fish farmers have resulted in slow acceptance of these new products. However, a new awareness in the feed industry towards problems, as well as commercial opportunities in aquaculture will soon lead to the industry applying its substantial resources and expertise for development of new products that will be more acceptable to fish farmers. This awareness was expressed to NIFI by Thailand's leading feed manufacturer, the Chareon Pokphand Group of Companies, in an offer to bear the cost of remodelling the Institute's feed mill. The mill group also expressed their readiness to sponsor selected fish nutrition research projects that have commercial relevance.
A research programme was proposed by the author for the development of optimal diets for pond culture of the two most important commercial species: Clarias and Ophicephalus.
Objectives
The primary objectives of the experiments will be to:
develop diets that will provide the highest economic returns to farmers under the proposed culture systems and
develop knowledge of substitution of raw feedstuff materials to ensure continued economic operation during periods of price changes or shortages of certain key ingredients.
The secondary objectives will be to develop knowledge concerning:
nutritional requirements of the cultured species; and
optimal feeding rates and feeding frequencies under a varies of rearing conditions.
Criteria for test diets
Test diets will be formulated to:
meet all known (and unknown) nutrient requirements of the species
include only locally available feed ingredients;
possess physical and chemical attributes that will ensure ready acceptance by fish as well as good water stability to minimise wastage and pollution load.
Design of experiments and experimental parameters
In view of increasing cost and shortage of fish meal and other conventional animal protein sources for compound feed manufacture, and the serious water pollution and disease (both bacterial and nutritional) problems associated with the use of trash fish in Clarias and snakehead culture, diets containing less expensive vegetable protein substitutes will be formulated and tested. Furthermore, because of the high cost of protein in fish diets, the cost effectiveness of different protein levels will also be studied in each culture system. Ingredient substitution levels and dietary protein levels in each study will be according to the age class of experimental animals. Due to limited resources and facilities presently available at NIFI, the number of ingredient substitution levels factored with dietary protein levels must necessarily be within those restricted to narrow limits set by empirical knowledge concerning the safe use of the substituent ingredients, and by known requirements of the species.
Response of experimental animals to each dietary treatment will be measured by weight gains over the experimental period. Efficiency of each diet will be determined by computation of feed conversion ratios from which the cost of production and profitability of each operation may be calculated.
Analysis of variance of the measured response to each dietary treatment, coupled with cost data derived by computation, will indicate the optimum diet combinations for maximum returns.
Since fairly large differences between treatments may be expected, three replicates will suffice. Each replicate, however, should consist of at least 200 fry (5-6 cm) or 100 fingerlings (10–12 cm) in wet lab studies.
All fish should be closely graded for size before stocking.
In wet lab studies, plastic-lined circular pools of 1 m diameter may be used for fry and 2 m diameter for fingerlings. Water depth will depend upon the species, e.g., 30 cm for Clarias and 60 cm for snakehead.
Fry should be fed four times daily and fingerlings twice daily. Feeding rates will have to be established, any daily feed consumption must be recorded for each tank or pool.
The period of experimentation for fry should last until the fish have attained an average size of 10 cm, at which time the diet change may be made. Experiments with fingerlings may be derminated when the fish have attained 25 cm in average length.
Handling and weighing of fish should be held at a minimum. Young fish may be weighed once every two weeks at the most. There is no necessity for more frequent weighing of fish beyond the fingerling size.
Development of an optimal feeding programme from the fry stage will, therefore, be based on the selection and matching of two, highly efficient diets for the different growth stages. This combination may not necessarily consist of the optimal fry diet and optimal grower diet determined in separate experiments. By factoring out fry raised on the two best fry diets with the two best fry diets with the two best grower diets, an indication of the most profitable combination of diets for the entire production period, can be obtained.
Experimental data
The following experimental data should be collected for each replicate.
Initial fish weight and length measurements
Weights at intermediate periods
Daily feeds given
Weight gains for entire experimental period; length measurements
Total feed consumption for entire experimental period
Daily records of mortality
Daily observations on experimental conditions
Body composition analysis on 5 percent of fish at end of experiment.
Details of work programme for initial studies with Clarias is given in Figure 6, Appendix II. Test diets are described in Table 3 and 4, Appendix I.
Cannibalism during the early growth stages of Macrobrachium is a major cause of economic loss in commercial prawn culture. Although not normally observed in the wild, this behavioural aberration is one of the most important causes of high mortality and low yield under artificial culture conditions. It has been observed to be most prevalent under intensive culture systems where stocking densities are high. Cannibalism is also observed among young, rapidly growing broiler chickens reared under crowed conditions in warm weather. These conditions tend to produce a rapid rise in the basal metabolic rate of the animals. In warm-blooded animals, stress promotes the secretion of glucocorticoids which in turn increases retention of sodium (and usually of chloride) and hence raises the requirement for the mineral. Due to increases metabolic rate, requirement for labile phosphate is also increased. Successful control of cannibalism in chickens by increasing the dietary salt and phosphorus levels, indicates an increased requirement for these minerals under stress conditions. The effect of physiological stress on mineral requirements in Macrobrachium has not been investigated.
An experiment was proposed to determine the effects, if any, of dietary salt and phosphorus levels on cannibalism and growth among post-larvae and juveniles under laboratory conditions. The factorially designed experiment will involve the testing of NIFI Macrobrachium diets reformulated to contain two levels of salt and phosphorus. The proposed work programme for this study appears in Figure 7, Appendix II.
Possible shortages in supplies of trash fish, as well as fish meal, necessitate the development of suitable substitutes based on plant sources. The most readily available plant protein at present is soyabean meal. There are, however, problems associated with its use in compound fish feed, especially at high levels for young fish. These problems are in part related to poor digestibility of the carbonhydrate component of soyabean. Another important constraint in its usage is soyaben's low content of methionine. Furthermore, selenium deficiency, which results in muscle degeneration and unthrifty growth among conventional livestock fed all-vegetable protein diets, has been attributed to low levels of this essential element in vegetable protein sources. 10
Recent studies in the U.S.A. have shown that full fat soyabean meal, produced under high temperature and pressure, when supplemented with a proper amount of synthetic methionine, may be used to replace fish meal in diets for trout. This could indicate modification of oligosaccharides in soyabean by the heat process, rendering them more digestible. Studies involving poultry and livestock have indicated that selenium deficiency symptoms can be prevented by supplementing the all-vegetable protein diets with sodium selenite.
Under these provisions, comparative feeding trials have been recommended for determining the effectiveness of fat-free soyabean meal and full-fat soyabean meal as substituents for fish meal. Diet formulations for this study appear in Table 5. For the study on requirement of supplemental selenium in all-vegetable protein diets, formulations in Table 6 are suggested.
Encapsulated whole egg diets 11 were tested on larvae of four fish species: rohu, clarias, Pangasius and sand goby.
Rohu
Two whole-egg preparations were tested against a third, based on boiled egg yolk involving a total of about 1 500 rohu fry that had just completed yolk sac absorption and had begun to feed. The fish were divided equally into six glass aquaria per treatment, i.e. approximately 250 larvae per aquarium. Rations of encapsulated diets were measured out daily and re-suspended in water before feeding. The rest of the diet preparations were kept under refrigeration. Daily feeds on a dry basis amounted to 20 percent estimated body weight of the fish and were given four times daily. Aquaria were cleaned daily.
The larvae were kept on the diets for ten days, during which their growth and survival were closely monitored. Fish on the encapsulated whole egg diets grew normally and remained active during the period. There was no mortality among larvae feeding on encapsulated diets.
Larvae fed boiled egg yolk suffered mortality from the fourth day of the trial and by the tenth day there were few survivors.
11 For details on their preparation, see ADCP/REP/30/11.
Clarias
The encapsulated egg diet was tested against live Moina and two artificial feeds: dried earthworms from Japan and a larval diet of New Zealand origin. Fish on Moina performed best in terms of growth and survival, followed by the encapsulated diet. Performance of fish on the two, dry diets were very poor.
Sand goby
Larvae of sand goby were fed the encapsulated diet for two weeks. Although there was no observed mortality up to the seventh day, large losses occurred beyond that time period. It was not known whether the losses were due to mortality or because of the increasing presence of water mites from the fourth day of the test.
Pangasius
The test of the encapsulated diet on Pangasius larvae was unsuccessful. The aquaria were too densely stocked at the beginning, resulting in uncontrolled cannibalism among the young fish.
These preliminary studies indicate that encapsulated whole egg is a good substitute for live natural food for rohu and Clarias.
