Feeding of Rainbow Trout with Non-Fish Meal Diets

Transcription

1 Fisheries Science 63(2), (1997) Feeding of Rainbow Trout with Non-Fish Meal Diets Takeshi Watanabe, Visuthi Verakunpiriya, Kanako Watanabe, Viswanath Kiron, and Shuichi Satoh Department of Aquatic Biosciences, Tokyo University of Fisheries, Konan, Minato, Tokyo 108, Japan (Received August 5, 1996) As a part of the on-going efforts to increase the use of alternative proteins in practical fish feeds, this study was conducted to investigate whether or not fingerling rainbow trout can grow normally on the same non-fish meal diets that resulted in the green-liver for yellowtail. Two feeding experiments were carried out with fingerling rainbow trout for 20 weeks. The first ex periment was conducted to examine the effect of supplemental crystalline amino acids, with or without coating, in the non-fish meal diets on the growth performance of rainbow trout. In the second, the effect of supplemental levels of crystalline amino acids in the non-fish meal diets was investigated. Feeding of non-fish meal diets for 20 weeks produced excellent growth and feed performances for rainbow trout without any abnormalities. Supplementation of essential amino acids (EAA) in the non fish meal diets to balance amino acid profiles comparable to the control fish meal diets resulted in no fur ther enhancement of growth over the control fish meal diet group. The experimental fish could grow well on the non-fish meal diets without EAA supplement. This has shown that the content of EAA in the non-fish meal diets used in this study was sufficient to satisfy the EAA requirement of rainbow trout. Key words: non-fish meal diets, growth performances, rainbow trout, FAA The shortage of quality fish meal in Japan due to a quick decline of sardine catch during these several years has made it necessary to increase the use of alternative pro teins in practical fish feeds. Therefore, efforts are now being directed to find alternative protein sources of good quality which are ideally less expensive and readily availa ble as a substitute for the expensive fish meal component in fish diets. Our previous experiments1-2) on rainbow trout have al ready shown that a combination of 30% defatted soybean meal (SBM), 15% corn gluten meal (CGM), and 15% meat meal (MM) as a replacement for 90% of fish meal in the diet produced growth comparable to the fish meal diet at about 44% dietary protein level. This combination of alter natives could also replace 50-60% of fish meal in diets for carp,3) red sea bream,*1 and yellowtail.4-6) Subsequent tri als to eliminate fish meal entirely from the diets for yellow tail,*2 revealed that fish grew normally for the first 6 to 8 weeks until the so-called 'green-liver' appeared in the ex perimental fish fed the non-fish meal diets, but not in those on the fish meal diet. These results suggested that yellow tail could grow on non-fish meal diets, irrespective of the appearance of green-liver, but it remained unclear if the ab normal liver was induced by the deletion of fish meal. In order to ascertain if any link existed between non-fish meal diets and the appearance of green-liver, rainbow trout in this experiment were fed the same non-fish meal diets which caused the green-liver in yellowtail. Experimental Materials and Methods Diets Two feeding experiments ( T and U) were carried out with fingerling rainbow trout Oncorhynchus mykiss. In Ex periment I, the supplemental effects of coated and uncoat ed crystalline amino acids in non-fish meal diets on growth performances of rainbow trout were examined. Experi ment U examined the effects of different supplemental lev els of crystalline amino acids in the non-fish meal diets. Experiment I: The composition of the experimental diets and their proximate values analyzed according to methods described previously1-3) are shown in Table 1. Since the experimental diets were originally prepared for yellowtail, the crude protein (CP) and crude lipid (CL) con tents of the test diets were higher than the requirement for rainbow trout. A control diet (diet 6) with 67% fish meal (CP: 68%) as a main protein source was formulated; this being close to a practical soft-dry pellet (SDP) for yellow tail.7-8) In the test diets, the fish meal component included in the control diet was entirely eliminated and replaced by a combination of alternative protein meals such as soy pro tein concentrate (SPC; CP: 70%; a product of Aarhus Olie called Danpro-A), SBM (CP: 46%), CGM (CP: 65%), and MM (CP: 80%) at different levels. The SPC was used as a main protein source, ranging from 40% in diet 1 to 20% in diets 4 and 5, CGM being correspondingly includ ed at a level of 3% in diet I and 23% in diets 4 and 5. All the test diets were also incorporated with 2% krill meal in *1 T. Watanabe et al.: Abst. Metg. Japan. Soc. Fisheries Sci., April, 1995, p. 36 (in Japanese). *2 M. Maita et al.: Abst. Metg. Japan. Soc. Fisheries Sci., April, 1995, p. 36 (in Japanese).

2 * Non-Fish Meal Diets for Advanced Aquaculture 259 Table 1. Composition of the non-fish meal diets used for feeding fingerling rainbow trout 1 Monobasic. *2 Mg-L-Ascorbyl-2-Phosphate. 3 Lysine 1.5, methionine 0.5, threonine 0.5 and tryptophan 0.2. *4 Lysine 1.5, methionine 0.5, thereonine 0.5, tryptophan 0.2 and others 2.3. *5 Not determined. order to enhance palatability of diets.9) The mineral and vitamin mixtures were prepared to satisfy the requirement of yellowtail.7) A mixture of 1.5% lysine, 0.5% methio nine, 0.5% threonine, and 0.2% tryptophan was sup plemented in diets 1 to 4 to match the amino acid profile of these diets with that of the control fish meal diet.1-2) The same amino acid mixture, but coated with 1.5% lysine, 0.5% methionine, 0.5% threonine, 0.2% tryptophan, and 2.3% others, was added to diet 5 which had the same dieta ry composition as diet 4, to compare the availability of crystalline amino acids in terms of absorption speed from the intestine. The test diets formulated to have almost same CP and CL levels of about 45% and 20%, respec tively, were prepared as dry pellets by a small size of twin screw extruder (Suehiro Iron Works Co., Ltd.), while the control fish meal diet was prepared as SDP by a large size twin screw extruder as described previously.7) The pre pared diets were kept in a cold storage room (5 Ž) during the feeding period. The essential amino acid (EAA) composition of the test diets analyzed by Japan Food Research Laboratories is shown in Table 2. Since the SPC has been demonstrated to have certain disadvantages, including the deficiency of some EAAs; the test diets were all prepared to be compara ble to the control fish meal diet (diet 6), and contained adequate amounts of EAA to satisfy the requirement of rainbow trout.10) The fatty acid composition of the ex perimental diets analyzed by the same methods described elsewhere11) is shown in Table 3. The percentage of n-3 highly unsaturated fatty acids (n-3 HUFA), the essential fatty acids (EFA) for fish, did not differ between the test diets, but was quite higher in the control diet since it con tained both supplemental fish oil (pollack liver oil) and that derived from the fish meal. The content of n-3 HUFA in all the diets satisfied the EFA requirements of rainbow trout.11) Experiment U: The contents of SPC, CGM, and MM were kept constant in all diets at levels of 30%, 10%, and 13%, respectively. A part of SBM in the diets was replaced by the amino acid mixture used in the Experiment I at lev els of 0, 2.7, 5.4, and 8.1%, respectively in diets 7 to 10, in order to examine the effect of different supplemental levels of amino acids on feed performances. Diet 2 used in the Experiment I was arranged as the control diet in order to compare it with diet 7, the former with 2.7% amino acid mixture and the latter without it. In Experiment 1, it was found that the absence of fish meal in the diets resulted in marginal levels of minerals like P (Table 1), the content being around 0.8%, slightly higher than P requirement of rainbow trout and yellowtail (around %). However, in order to ensure the safe availability, the level of P in the test diets must be adjusted such that 85% of the mineral is absorbed.12) This prompted the mineral fortifica tion of the diets, particularly with iron (Fe) and phospho rus (P), in Experiment U. The level of the vitamin mixture was also increased from 1.5% in Experiment I to 2.0% in Experiment U. The test diets were formulated to have almost the same proximate composition as those in the Ex-

3 260 Watanabe et al. Table 2. Amino acid composition of the non-fish meal diets fed to rainbow trout and a comparison of the amino acid requirements of carp and rainbow trout *1 Control diet. *2 Ogino & Takeda.12) periment I in terms of CP, CL, and n-3 HUFA levels, but the protein content was found to be slightly higher in Ex periment U. The diet was prepared using a small size twin screw Feeding extruder. Methods Fingerling rainbow trout with an average body weight of about 10.9 g were randomly sampled from stock which had been fed with a commercial rainbow trout diet for 4 weeks. They were divided into ten 60 litre tanks, 45 1-water volume (6 tanks for Experiment I and 4 for Experiment U), 25 fish in each, with a continuous water supply at a rate of /min at 12.1 }3.9 Ž (ranging between 18.0? 6.8 Ž). The experimental fish were fed with one of the var ieties of dietary preparations for a total of 20 weeks. Feed ing frequency was arranged to suit the fish size; three times daily from the initiation to the 12th week and twice a day from the 13th week till the termination, each time to satia tion. Analytical Methods At the beginning and at the end of the feeding experi ments, fish were starved for 24 hours and anesthetized with ethyl aminobenzoate (100 ppm) before being weighed individually (initial and final body weight). Prior to start of the experiment, twenty fish were sacrificed and stored at -20 Ž until analysis. The pooled fish sample was homogenized and analyzed for whole body and muscle proximate composition following the method described previously. 1-2 At the end of the experi ment after individually weighing, ten fish from each tank of each dietary treatment were collected for whole body, muscle, and liver composition analysis. The values ob tained were employed for calculating the protein efficiency ratio (PER), net protein utilization (NPU), and retention of protein and energy according to the methods described elsewhere.1-2) Blood samples were collected by cardiac puncture with heparinized syringes from 5 fish of each treatment which were starved 24 hr prior to the termination of the feeding experiment. The samples were kept at 4 Ž and then hematocrit values (Ht) were determined within 2 hr of blood sampling. The hemochemical properties of the in dividual blood samples from each treatment were deter mined using plasma obtained by centrifuging the blood at 3000 rpm for 10 min. More details on methodology are provided elsewhere.4) Feed performance was evaluated from bi-weekly group

4 Non-Fish Meal Diets for Advanced Aquaculture 261 Table 3. Fatty acid composition of the experimental diets used for fingerling rainbow trout *1 Trace amount (<0.005). *2 Not detected. weighings on the basis of weight gain, growth rate, and feed efficiency ratio. Daily observations were made on mor phological changes, pathological symptoms (clinical signs), unusual behavior, and mortality. Analysis of Variance (ANOVA) was performed to exa mine the effect of the test diets on the growth of the fish. Duncan's New Multiple Range Test was then employed to compare the individual treatment mean difference at the 0.05 significance level. Feed Performances Results and Discussion Experiment I: The mean weight gain, growth rate, feed efficiency, and average feeding rate are summarized in Ta ble 4. Fish fed the experimental non-fish meal diets showed generally good growth throughout the feeding period of 20 weeks, as represented by the growth curves in Fig. 1; but comparatively, the fish on diets 1 to 3 containing relatively high amounts of SPC were slightly lower in weight gain. The growth of fish fed diets 4 or 5 containing 20% SPC and 23% CGM was highest among treatments, being equal to that of the control group fed the fish meal SDP. Since the replacement of SPC by CGM improved growth, it may suggest that there could be a limit for the inclusion level of SPC in the diet. Further there was no particular difference when the fish were fed diets supplemented either coated or uncoated EAA mixtures. Thus, the non-fish meal diets which could not sustain normal growth in yellowtail for more than 8 weeks, provided excellent growth for rainbow trout which grew from 10.9 g at the initiation up to around 180 g after the 120 days of feeding period. Feed efficiency was also good in all the treatments (Fig. 2), ranging from 1.23 for the control diet to 1.44 for diet 5. A slightly lower value obtained in the control diet was probably due to the higher feed intake of this group. The high feed efficiency of the non-fish meal diets was partly due to the relatively high energy contents, compared to normal rainbow trout diets, and also due to the effective utilization of the alternative protein sources used in the for mulations. Generally, adding high percentages of soy products in fish diets can cause unpalatability and unacceptability lead-

5 262 Watanabe et al. Table 4. Growth and feed performances of rainbow trout fed non-fish meal diets for 20 weeks *1 Mean }SD *2 g/25 fish per 120 feeding days (rearing periods: 20 w eeks)., n=25 and the values within column with the same superscript are not significantly different (P>0.05). Fig. 2. Daily feed intake of fingerling rainbow trout fed the non-fish meal diets for 20 weeks. Values are mean of bi-weekly data. Fig. 1. Growth of rainbow trout fed non-fish meal diets for 20 weeks (Experiment I). ing to diminished growth as noted in salmonids1,13) and yel lowtail.4) Contrary to this, the present non-fish meal diets were palatable and acceptable to rainbow trout, irrespec tive of the high amount of soy components contained therein. Daily feeding rate ranged from during the initial 2 weeks to during the final 2 weeks (Fig. 3), being highest for the control group throughout the feeding period. Experiment U: The results of growth and feed utiliza tion are presented in Table 4 and Figs Growth rate was highest in fish fed diet 7 or diet 8, followed by the con trol diet (diet 2 in the Experiment I). The composition of diet 7 was almost the same as that of diet 2 except for the supplementation of 2.7% EAA mixture in the latter and some changes in the combination of protein sources. Diets 2 and 8 contained 2.7% EAA, and were dissimilar only on account of the differing protein source ratios. The changes of protein combinations in diet 8 did not cause any retarda tion of growth (Fig. 4) as demonstrated previously.1-3) This indicated that there was no effect of EAA mixture sup plementation in the non-fish meal diets on growth perfor mances in rainbow trout. Conversely, EAA contents of the experimental non-fish meal diets were sufficient to satis fy the EAA requirement of rainbow trout, 10) Reduction of

6 Non-Fish Meal Diets for Advanced Aquaculture 263 growth rate in fish fed diets 9 and 10 with the EAA supple ment twice and thrice respective to that of diet 8, might have been caused by an imbalance resulting from overload ing of crystalline amino acids. Feed efficiency and daily feeding rate were similar to those in the Experiment I. Palatability and acceptability were also good even in fish fed diets with 5.4% or 8.1% EAA mixtures, judging from bi-weekly feed consumption. Further, the fortification of minerals and vitamins in the diets also did not show any supplemental effect on growth performance. Neither in Experiment I nor in Experiment U was the green-liver abnormality noted in rainbow trout, even though the fish were fed the very same non-fish meal diets that had caused green-liver in yellowtail.*2 During the course of the daily observations no clinically suggestive re marks were made and fish of all the treatments grew nor mally. Fig. 3. Feed efficiency in fingerling rainbow trout fed the non-fish meal diets for 20 weeks. Values are mean of bi-weekly data. Protein and Energy Utilization Protein and energy retention in whole body and muscle of fish in the Experiment I and II are shown in Table 5, together with PER and NPU. Experiment I: PER values ranged from , being highest in diet 5 and lowest in the control diet. The low PER of the control group might be due to a higher protein intake which affects protein utilization efficiency. The PER of diet 5 was higher than that of diet 4, probably enhanced by the coating of EAA, as observed for feed efficiency. The same tendency prevailed for protein retention and NPU, the indicators of protein utilization efficiency. Pro tein retention varied between 40.9 and 50.3% in whole body, and 50.8 and 60.2% in muscle, being highest in fish fed diet 5 and lowest in those on the control fish meal diet in both the portions. The NPU values ranged from 50.7 to 62.0, the trend identical to that for protein retention and PER. However, these values were quite high for all the diets regardless of the high protein contents, compared to practical diets hitherto used for rainbow trout,14-15) indicat ing that protein utilization efficiency was effectively en hanced by elevation of dietary energy levels. Similar results were also obtained in rainbow trout fed high-pro tein high-energy diets produced by a twin screw ex truder.*3 Energy retention showed no marked variations among the dietary treatments, and was higher in whole body ( %) than muscle ( %). Experiment U: The values for PER, protein and energy retention, and NPU are summarized in Table 5. These values showed overall same tendency, higher in fish fed diets 7 and 8, and lower in fish with lower growth rate. The protein utilization efficiency in terms of PER, protein retention, and NPU seemed to be improved slightly by sup plementation of 2.7% EAA mixture, when these values were compared between fish on diets 7 and 8. The reten tion of energy was also lower in fish fed diets 9 and 10 with higher amounts of the EAA supplement. Fig. 4. Growth of rainbow trout fed the non-fish meal diets for 20 weeks (Experiment U). Proximate Composition of Whole Body, Muscle and Liver Proximate composition of the experimental fish is presented in Table 6. *3 T. Watanabe et al.: Abst. Metg. Japan. Soc. Fisheries Sci., April, 1995, p. 37 (in Japanese).

7 264 Watanabe et al. Table 5. Retention of protein and energy in whole body and muscle of fingerling rainbow trout fed non-fish meal diets for 20 weeks *1 Protein efficiency ratio. *2 Net protein utilization. Table 6. Proximate composition of the rainbow trout fed non-fish meal diets for 20 weeks *1 Not determined. Hepatosomatic *2 index (Meant }SD, n=5). Experiment I: The whole body of the initial fish tended to have higher moisture (75.3%) and ash content (1.9%) but lower lipid (6.6%) and protein (16.1%) content than those of the final fish (moisture: %; ash: %; lipid: %; and protein: %). The low body fat content in the initial fish may have resulted from feeding a commercial rainbow trout diet which had only 3% lipid. There was only a slight increase in the body protein of the final fish, whereas their lipid content was more than double the initial value, reciprocating the reduced moisture content. The proximate values of pro tein and lipid also increased in the muscle of final samples, though it was not particularly related to the dietary treat ments. In the liver, the protein content ranged from 17.2 to 19.9% and lipid varied between 4.6 and 6.5%. The lipid contents remained low in all the groups regardless of high protein high-energy intake. Therefore, hepatosomatic in dex was also low, ranging from (Table 6). Experiment U: As in Experiment I, there were no definite trends for the proximate values of whole body and muscle, linked to the treatments. There were no marked differences in the proximate composition of whole body, muscle and liver between fish on diets 7 and 8 or the con trol diet (diet 2 in Experiment I), though liver protein con tent was slightly high in fish fed diet 2. The diets were quite similar to each other except for the 2.7% EAA supplement ed to diets 2 and 8. Thus, we find that the supplementation of 2.7% EAA did not influence the proximate values, as observed for growth data. Blood Biochemical Parameters The blood properties of the experimental rainbow trout fed the non-fish meal diets for 20 weeks, in the two experi ments are summarized in Table 7. Experiment I: The fish in all the treatments fed the non, fish meal diets seemed to be in good health. The

8 Non-Fish Meal Diets for Advanced Aquaculture 265 Table 7. Results of the hemochemical assessment made on rainbow trout fed the non-fish meal diets for 20 weeks*1 *1Mean }SD, n=5. hematocrit values were unaffected by dietary treatments, and did not indicate extreme nutritional deprivation or overt disease.16) The values in these groups for the enzyme activities, GOT= IU/1, GPT=16-201U/1, and ALP= IU/1, were comparable to the control fish meal group (diet 6). The values of GOT and phospholipid indicated a relatively better status for the control group. Among the non-fish meal groups (diets 1-5), PL value in diet 4 was relatively high, and inferring from this hemochemical data, in conjunction with the growth and feed utilization data; diets 4 and 5 could be considered to be as effective as the control fish meal diet. On the other hand, an explanation supporting the benefit of EAA coat ing (diet 5) could not be provided based on the values of protein metabolism. Further, there were no representative changes in the values of other blood components, except for the high values of lipid metabolites in the control fish meal group, probably due to dietary sources. Experiment U: There was no marked difference in en zyme activities and blood metabolites among the treat ments, and they were as good as the control group (diet 2). The addition of supplemental levels of EAA mixture did not show any negative results when compared with the con trol group, as noted for growth or feed performances, in dicating that the overloading of EAA mixture in the non fish meal diets does not adversely affect the fish and they maintained good health throughout the twenty-week feed ing period. The extruded diets without fish meal which could not sustain normal growth of yellowtail and caused the green liver, seemed not only to be harmless to rainbow trout, but also efficient in growth promotion and performance. It has to be noted that the green-liver condition in yellowtail was caused by a parasite Myxobolus spirosulcatus found in the bile ducts of the liver. The expression of this abnormal con dition in yellowtail was observed only in those fish fed non-fish meal diets, and hence a dietary link cannot be ruled out. Though this correlation has to be verified, the present result on rainbow trout proves that the diets are not solely responsible for the abnormality. Supplementation of EAA in the non-fish meal diets to match the amino acid profiles with that of the control fish meal diet, did not prove to be beneficial. Since the fish could grow well on the non-fish meal diets without EAA supplement, it is probable that the EAA already contained in the non-fish meal diets used in the current study satisfied the requirement of rainbow trout. Fish meal based diets usually contain sufficient amounts of EAA at a level above the requirement of the fish; therefore, the need to balance the EAA profile rarely arises. When new sources are being tested, the EAA content of the dietary proteins and diges tibility of the protein or each EAA should be considered for assessing the adequacy of the EAA profile of the diet in toto. If overlooked, this might cause an overloading of EAA mixture, as in the present experiment, wherein the diets with levels of 5.4 and 8.1 % reduced growth and feed efficiency together with protein utilization efficiency. Differ ence in performance of the non-fish meal diets between yel lowtail and rainbow trout may partly be due to difference in growth speed, feed intake, and EAA requirement; although it was reported recently that EAA requirement did not differ significantly between the two species.1-5,10) The results of this current study have shown the possibility of developing practical fish feeds with a low amount of fish meal or even without fish meal as a protein source. Fish meal is the widely sought after fish feed ingredient because it is a rich source of essential amino acids, essen tial fatty acids, energy and minerals. Moreover, it is very palatable and highly digestible to most freshwater and ma rine fishes. Recently fish meal is being sparingly used in commercial fish feeds because of the shortage of quality fish meal due to the decline in the landings of meal grade fish. Among the plant sources being considered as possible replacers for fish meal, the less expensive and more abun dant, soybean meal and their products are of vital interest. Eventhough, soybean meal has better amino acid profiles

9 266 Watanabe et al. than other plant protein foodstuffs, it has some nutritional limitations when being considered as replacement in fish feeds.1-6) Soybean meal is low in digestibility of nitrogen and energy, deficient in minerals, contains indigestible components such as oligosaccharides, and some anti-nutri tional factors (trypsin inhibitors, antigenic proteins, lec tins, saponins, etc.), and generally results in poor growth when being considered as a partial or total substitute for fish meal.1-6,12,17) However, recent advances in feed manufacturing processes have enabled better utilization of soybean in fish feeds. 1-2,17,20) SPC is a potentially good fish feed ingredient because during the processing procedure, most of the soluble carbohydrates are removed, oligosac charide concentration is reduced and protein content is elevated. Rumsey et al.21) reported that after manufacture, trypsin inhibitor activity for SPC is also lower than the crit ical point (5 mg/g) at which fish proteolytic digestive en zymes are adversely affected.22) Furthermore, the same researchers demonstrateeed that the combination of SPC and low temperature fish meal rotein ratio of 50:50 fed to rainbow trout gave good growth that was compara ble to low temperature fish meal group. However, the use of SPC as a sole protein source in rainbow trout diet did not produce the overall benefits of fish meal.23) Neverthe less, the results obtained from this study showed that the combination of SPC and other non-fish meal protein sources produced growth on a par with control fish meal fed fish. Though SPC was included as a main protein source, this component is also expensive, and further ex periments are needed to entirely replace fish meal with more practical, less expensive protein ingredients in fish feeds. References 1) J. Pongmaneerat and T. Watanabe: Utilization of soybean meal as protein source in diets for rainbow trout. Nippon Suisan Gak kaishi, 58, (1992). 2) T. Watanabe and J. Pongmaneerat: Potential soybean meal as a protein source in extruded pellets for rainbow trout. Nippon Suisan Gakkaishi, 59, (1993). 3) J. Pongmaneerat, T. Watanabe, T. Takeuchi, and S. Satoh: Use of different protein meals as partial or total substitution for fish meal in carp diets. Nippon Suisan Gakkaishi, 59, (1993). 4) V. Viyakarn, T. Watanabe, H. Aoki, H. Tsuda, H. Sakamoto, N. Okamoto, N. Iso, S. Satoh, and T. Takeuchi: Use of soybean meal as a substitute for fish meal in a newly developed soft-dry pellet for yellowtail. Nippon Suisan Gakkaishi, 58, (1992). 5) T. Watanabe, V. Viyakarn, H. Aoki, H. Tsuda, H. Sakamoto, M. Maita, S. Satoh, and T. Takeuchi: Utilization of alternative protein sources as substitute for fish meal in a newly developed soft-dry pellet for yellowtail. Suisanzoshoku, 42, (1994). 6) T. Watanabe, H. Aoki, V. Viyakarn, M. Maita, Y. Yamagata, S. Satoh, and T. Takeuchi: Combined use of alternative protein sources as a partial replacement for fish meal in a newly developed soft-dry pellet for yellowtail. Suisanzoshoku, 43, (1995). 7) T. Watanabe, H. Sakamoto, M. Abiru, and J. Yamashita: Develop ment of a new type of dry pellet for yellowtail. Nippon Suisan Gak kaishi, 57, (1991). 8) H. Sakamoto, T. Watanabe, and T. Takeuchi: Optimal levels of protein and lipid in a newly developed soft-dry pellet for yellowtail, Seriola quinqueradiata. Suisanzoshoku, 43, (1995). 9) C. Shimizu, I. Allahpichay, T. Tokoro, and Y. Shirakawa: Feeding stimulation in seabream, Pagrus major, fed diets supplemented with Antarctic krill meals. Aquaculture, 86, (1990). 10) C. Ogino: Requirement of carp and rainbow trout for essential ami no acids. Nippon Suisan Gakkaishi, 46, (1980). 11) T. Takeuchi, T. Watanabe, and C. Ogino: Dietary levels of methyl laurate and essential fatty acid requirement of rainbow trout. Nip pon Suisan Gakkaishi, 43, (1977). 12) C. Ogino and H. Takeda: Requirement of rainbow trout for dietary calcium and phosphorus. Nippon Suisan Gakkaishi, 44, (1978). 13) L. G. Fowler: Substitution of soybean and cottonseed products for fish meal in diets fed to chinook and coho salmon. Prog. Fish Cult., 42, (1980). 14) C. Ogino, J. Y. Chiou, and T. Takeuchi: Protein nutrition in fish IV. Effects of dietary energy sources on the utilization of proteins by rainbow trout and carp. Nippon Suisan Gakkaishi, 42, (1976). 15) T. Takeuchi, M. Yokoyama, T. Watanabe, and C. Ogino: Opti mum ratio of dietary energy to protein for rainbow trout. Nippon Suisan Gakkaishi, 44, (1978). 16) NRC (National Research Council): Nutrient requirements of col dwater fishes, National Academy Press, Washington, D.C., ) K. Dabrowski, P. Poczyczynski, G. Kock, and B. Berger: Effect of partially or totally replacing fish meal protein by soybean meal pro tein on growth, food utilization and proteolytic enzyme activities in rainbow trout (Salmo gairdneri). New in vivo test for exocrine pan creatic secretion. Aquaculture, 77, (1980). 18) A. G. J. Tacon, J. V. Haaster, P. B. Featherstone, K. Kerr, and A. J. Jackson: Studies on the utilization of full-fat soybean and sol vent extracted soybean meal in a complete diet for rainbow trout. Nippon Suisan Gakkaishi, 49, (1983). 19) J. Pongmaneerat and T. Watanabe: Effect of extrusion processing on the utilization of soybean meal diets for rainbow trout. Nippon Suisan Gakkaishi, 59, (1993). 20) C. D. Webster, J. H. Tidwell, L. S. Goodgame-Tiu, and D. H. Yancey: Evaluation of distillers grains with solubles as an alterna tive plant protein in aquaculture diets, in "Nutrition and Utiliza tion Technology in Aquaculture" (ed. by C. Lim and D. J. Sessa), AOCS Press, Champaign, Illinois, 1995, pp ) G. L. Rumsey, J. G. Endres, P. R. Bowser, K. A. Earnest-Koons, D. P. Anderson, and A. K. Siwicki: Soy protein in diets of rainbow trout: Effects on growth, protein absorption, gastrointestinal histol ogy, and non-specific serologic and immune response, in "Nutri tion and Utilization Technology in Aquaculture" (ed. by C. Lim and D. J. Sessa), AOCS Press, Champaign, Illinois, 1995, pp ) R. P. Wilson and W. E. Poe: Effects of feeding soybean meal with varying trypsin inhibitor activities on growth of fingerling channel catfish. Aquaculture, 46, (1985).