ABSTRACT: In order to develop an artificial diet, the dietary utility of enzyme-treated fish meal was investigated for juvenile Pacific bluefin tuna Thunnus orientalis (PBT). Diets containing each 63% of Chilean fish meal (FM), enzyme-treated Chilean fish meal (EC) and enzyme-treated Peruvian fish meal (EP), with 10% bonito oil and raw sand lance Ammodytes personatus (SL) were fed to juvenile tuna six times per day for one week. In a different trial, diets EC and SL were fed to tuna six times per day for 2 weeks. Only diet EC sustained similar growth or caused lower survival and higher feed efficiency, hepato- and enterosomatic indices and final carcass lipid content as compared to those of SL. Diets FM and EP led to lower specific growth rate (SGR) but similar feed efficiency, survival and hepatosomatic index, yet higher enterosomatic index. Moreover, PBT fed diet EC for 2 weeks led to similar growth performance but higher final carcass and hepatic lipid contents, and plasma cholesterol and phospholipid levels than those fed SL. Carcass fatty acid composition of diet EC group had lower 20:5 n-3 and 22:6 n-3 levels than the SL group. These results revealed that EC, as a suitable dietary protein source, could sustain growth of PBT, while dietary bonito oil led to higher carcass lipid but lower accumulation of n-3 highly unsaturated fatty acids.
Pacific bluefin tuna (PBT) Thunnus orientalis, together with Atlantic bluefin tuna Thunnus thynnus, is a migratory pelagic fish and grows faster than other teleost fish.1,2 The Fisheries Laboratory, Kinki University, Oshima, near the Cape of Shionomisaki, has carried out basic and practical research for establishing PBT aquaculture since 1970.3 An artificial full-cycle culture of PBT was initially achieved at the laboratory in 2002.4,5 Natural spawning of PBT broodstock ordinarily occurs at 21.6–29.2°C in nearshore areas of Oshima Island from June to August. The diameter of fertilized eggs is 0.926–1.015 mm and decreases with increasing water temperature. More than 50% normal hatching occurs at 21.2–29.8°C and requires 32 h post fertilization at 24°C.6 Embryonic development of PBT selectively requires triglycerides as a main energy source, and organogenesis is accelerated from the stage of Kupffer's vesicle.7 Larvae immediately post hatch (PH) are 2.83 mm in total body length. The hepatic organ, gallbladder and pancreas appear by 36 h PH, and tryptic and amyloritic activities were allocated just before first feeding. The gastric organ appears at approximately day 10 PH (10 dPH) and actively evoked peptic activity on 17–25 dPH.8 Larval PBT showed similar growth to larval red sea bream Pagrus major and metamorphosed to juveniles of approximately 10.6 mm in body length on 20 dPH.6 After 20 dPH, juvenile tuna growth is steeply accelerated and reaches 140 mm in body length on 50 dPH. Transformation from juvenile to young tuna, and high mortality caused by the trauma of flash collision against tank walls, are initiated at 80–100 mm body length.6
The nutritional requirements of PBT (Takii et al.,9 Takii K et al., unpubl. data, 2007), Atlantic bluefin tuna Thunnus thynnus10 and southern bluefin tuna Thunnus maccoyii have been reported.11–13 However, feeding raw trash fish is still recommended, which is concomitant with some problems such as environmental pollution, unfavorable meal composition and infectious disease outbreaks. Therefore, it is necessary to establish an artificial diet for the development of PBT aquaculture rapidly. In a previous study (Takii K et al., unpubl. data, 2007), juvenile PBT showed lower apparent digestibility coefficients for a moist pellet diet composed of equal parts raw sand lance Ammodytes personatus and Chilean fish meal (FM), than those for raw sand lance. Moreover, juvenile PBT fed artificial FM diets appeared to have lower specific growth rates and heavier digestive organ weights than those fed on sand lance (Takii K et al., unpubl. data, 2007). We hypothesized from these results that the digestive and absorptive functions of PBT juveniles could not fully adapt to dietary fish meal processed by heavy steam-boiling and heat-drying. In this study, therefore, we estimated the availability of some enzyme-treated fish meals, in place of FM, as a main dietary protein source for juvenile PBT, together with bonito oil as a main dietary lipid source, for developing an artificial PBT diet.
MATERIALS AND METHODS
Dietary formula and proximate composition are shown in Table 1. The control diet was chopped sand lance (SL), which is commonly fed to juvenile PBT in Japan. Three test diets were included: (i) 63% of Chilean fish meal (FM); (ii) enzyme-treated Chilean fish meal (EC) with main peptides of 100 and 570 molecular weight (BIO-CP, Nagase Biochemical Selling Co., Osaka, Japan), which was prepared with fresh raw horse mackerel Trachurus japonicus; and (iii) enzyme-treated Peruvian fish meal (EP) with main peptides of ≤ 5000 molecular weight (Peptidemeal, Tottori Canning Co., Tottori, Japan), which was prepared with anchovy Engraulis japonica meal as a main protein source. The enzyme-treated fishmeal was expected to improve protein digestibility and utility by hydrolysis. Wheat gluten, bonito oil, α-potato starch, vitamin mixture, mineral mixture, lecithin, taurine and feeding stimulants were commonly included in test diets. The bonito oil was fortified by the addition of 10% (DHA) oil (70% docosahexaenoic acid oil, Harima Kasei Co., Osaka, Japan). The vitamin and mineral mixtures were after Halver.14 We supplemented a mixture of l-alanine, l-glutamic acid, l-histidine, l-lysine and inosine-5′-monophosphate disodium salt hydrate (Na2) as feeding stimulants to the diets. After mixing well, 30% tap water (v/w) was further added and moist type pellets with 1.2 mm diameter were prepared using a laboratory pelleting machine. The diets were stored in a freezer at −20°C until use.
Table 1. Formulation and chemical proximate composition of test diets
The crude protein content of SL was higher than diets FM, EC and EP. Diet EC had the highest crude lipid content followed by diets SL, FM and EP. Sugar content of test diets was reduced to less than 8.3%. The highest crude ash was detected in diet EP followed by diets FM, SL and EC (Table 1).
Fatty acid composition of diets EC and SL is shown in Table 2. Diet EC had higher DHA and eicosapentaenoic acid (EPA) levels than SL.
Table 2. Test diets and final carcass fatty acid composition (%) of juvenile Pacific bluefin tuna in trial 2
Two feeding trials (trials 1 and 2) were conducted at the same time in the present study. Juvenile PBT were sourced from an artificial seedling production facility at the Fish Nursery Center, Kinki University, Uragami, Wakayama. Naturally fertilized eggs were obtained from the Fisheries Laboratory, Kinki University, Amami, Kagoshima and were transported to the Fish Nursery Center by air and land. This transportation took approximately 12 h.
In trial 1, 40 fish with mean body weight of 1.48 g were introduced into each of eight indoor circular 1.4-m3 fiber-reinforced plastic tanks with two replicates for each diet. The fish were fed the diets six times (07:00, 09:00, 11:00, 13:00, 15:00 and 17:00 hours) daily until apparent satiety for one week. The flow rate of filtered sea water into each tank was adjusted to 3 L/min. Water temperature and dissolved oxygen (DO) were 26.9 ± 0.5°C and 6.96 ± 0.5 mg/L, respectively.
In trial 2, 407 and 404 fish, originating from the same batch of fish as trial 1, were introduced to two indoor 15-m3 eight-angular concrete tanks and fed diets EC or SL, respectively. The feeding protocol was as in trial 1, but rearing period was two weeks. The flow rate of filtered sea water into each tank was adjusted to 30 L/min. Water temperature and DO were 27.1 ± 0.5°C and 7.91 ± 1.1 mg/L, respectively.
Measurements and assays
At the end of trials 1 and 2, body weight and length, specific growth rate (SGR) ([lnWi − lnWf] × 100/days, where Wi and Wf are mean initial and final body weights, respectively), feed efficiency (weight gain × 100/feed intake) and condition factor ([body length]3 × 100/body weight) were measured. Hepato-, gastro- and enterosomatic indices (organ weight × 100/body weight) were also calculated.
At the end of the trial 2, blood samples were collected with a heparinized syringe from the caudal vein of five fish fed the SL diet and five fish fed the EC diet. Hematocrit value (Ht) and hemoglobin concentration (Hb) were determined by the microhematocrit method and a commercial kit (Wako Pure Chemical Co., Osaka, Japan), respectively. After centrifugation of blood at 3000 g for 10 min at 4°C, plasma was sampled for assaying total protein, glucose, triglyceride, total cholesterol, phospholipid and nonesterified fatty acid (NEFA) levels by commercial kits (Wako Pure Chemical).
Proximate compositions of diets, carcass and liver were assayed by the Association of Official Analysis Chemists method.15 Fatty acid composition of diets SL and EC was assayed by gas liquid chromatography.16
Data were analyzed by one-way analysis of variance anova. When differences were found among dietary treatments, Duncan's multiple range test was used to compare the mean differences by the SPSS statistical package (SPSS, Chicago, IL, USA). Differences were considered significant at P < 0.05.
The final mean body weights of juvenile PBT fed diets EC and SL were significantly higher than diets FM and EP. Diets EC and SL also led to high SGR without a significant difference, but higher than diets FM and EP. The highest feed intake was obtained for the PBT fed SL, followed by diets FM, EP and EC. Diet EC provided the highest feed efficiency, followed by diets SL, FM and EP (Table 3). Diets FM, EC and EP induced higher hepato-, gastro- and enterosomatic indices than SL. Diets EC and SL supported higher condition factors than diets FM and EP, but no significant differences were detected among them (Table 4).
Table 3. Growth performance of juvenile Pacific bluefin tuna fed diets with difference protein sources for one (trial 1) and two weeks (trial 2)
The dietary treatments led to similar carcass moisture and crude ash levels in juvenile PBT. Diets EP and EC significantly decreased and increased carcass crude protein level and lipid levels, respectively, compared with other diets (Table 5).
Table 5. Proximate composition (%) of whole carcass and liver of juvenile Pacific bluefin tuna fed diets with different protein sources for one (trial 1) and two weeks (trial 2)
Values are mean of two group of fish (n = 2), with 10 fish per group (five fish per tank). Different letters indicate significantly different (P < 0.05).
Standard error of the mean.
SL, sand lance; FM, Chilean fish meal; EC, enzyme-treated Chilean fish meal; EP, enzyme-treated Peruvian fish meal.
Juvenile PBT fed diet EC maintained 80% final mean body weight and SGR of those fed SL. There were no marked differences in survival and feed intake between dietary treatments (Table 3). Diet EC caused higher feed efficiency and hepato- and enterosomatic indices than SL, comparable to trial 1. The gastrosomatic index and condition factor of the fish showed no significant differences between dietary treatments (Table 4).
Diet EC induced significantly higher carcass crude lipid levels than SL, but not carcass moisture, crude protein or crude ash. Otherwise, diet EC provided higher hepatic crude lipid and lower hepatic moisture and crude protein levels than diet SL (Table 5).
There were no significant differences in final Ht, Hb, plasma total protein, glucose, triglyceride and NEFA levels of juvenile PBT between the dietary treatments. Diet EC provided higher plasma total cholesterol and phospholipid levels than SL (Table 6).
Table 6. Hematocrit (Ht), hemoglobin (Hb) and concentration of serum constituents of juvenile Pacific bluefin tuna after end of trial 2
Values are mean with 5 fish per group (n = 5). Different letters indicate significantly different (P < 0.05).
Standard error of the mean.
NEFA, nonesterified fatty acid.
SL, sand lance; EC, enzyme-treated Chilean fish meal.
Total protein (mg/dL)
Total cholesterol (mg/dL)
The 18:1 n-9, 20:1 and 18:2 n-6 compositions of carcass lipid of juvenile PBT fed diet EC was significantly higher than those fed SL. The 20:5 n-3 and 22:6 n-3 compositions of carcass lipid were lower in the PBT fed diet EC than SL (Table 2). The linear regression lines of carcass fatty acid compositions against dietary fatty acid compositions were obtained in the saturated and monoenoic fatty acid fraction and the polyenoic fatty acid fraction. The slope of the saturated and monoenoic fatty acid fraction fed diet EC was higher than that fed SL. The slope of polyenoic fatty acid fraction including DHA and EPA fed diet EC was lower than that fed SL (Fig. 1).
Survival of juvenile PBT in the present study was remarkably lower than for of other cultured fish. This is attributed to the inevitable trauma caused by flash collision against tank walls, resulting in the imbalance of morphological development.6 The trauma obligates a short rearing period and obscures statistics for different variables to some extent, but the final body weight of juvenile PBT fed the diet attained more than two times the initial body weight without serious problems.
Juvenile PBT showed remarkably lower digestibility and utility of a moist pellet diet containing equal parts of an artificial FM diet and SL than juvenile chum mackerel Scomber japonicus (Takii K et al., unpubl. data, 2007). Moreover, artificial FM diets led to markedly lower weight gain and feed efficiency in juvenile tuna and southern bluefin tuna than raw trash fish.13 The juvenile tuna also responded to dietary FM with increase in digestive organ weight and peptic and tryptic enzyme activities than those fed raw trash fish (Takii K et al., unpubl. data, 2007). Other available useful protein sources than FM are needed for developing a practical artificial diet for juvenile tuna. Trials 1 and 2 of the present study showed that enzyme-treated Chilean fish meal was a suitable dietary protein source for juvenile tuna. Preferable growth performance in fish fed EC corresponding to SL indicated that EC will present important information for developing artificial diets and mass seedling production for PBT.
The lower growth performance of fish fed FM suggested that the juvenile PBT could not utilize FM effectively because of low digestive potency of the FM diet (Takii K et al., unpubl. data, 2007), but could utilize EC, which was prepared by partial hydrolysis of the FM peptide bond. Compared with SL, the changes in protein conformation of FM by heavy heating and dry processing might be a reason for reduced digestibility and utility of FM as a dietary protein source for juvenile PBT. The PBT, a representative off-shore migratory fish, might develop proteinase isoforms to digest only raw fish protein, but not for FM protein, which is made indigestable by heavy heat processing. Suzuki et al.17 indicated by their cDNA cloning and phylogenetic analysis that Japanese flounder Paralichthys olivaceus had isoforms of pancreatic trypsin, chymotrypsin, esterase and carboxypeptidase. It was also shown that tryptic enzyme and basic proteinase sensibility for soybean trypsin inhibitors differed among some fish species.18 Otherwise, juvenile tuna fed diet EC had fewer feed intakes but higher feed efficiency than SL. A large part of chopped SL might easily leave small particles and water-soluble nutrients in the rearing water.
Diets FM, EP and EC induced higher enterosomatic indices for juvenile tuna than SL in trial 1. This enteral organ enlargement might be caused by physiological adaptation for less digestible FM, EP and EC. However, adaptation to the diet may be more efficient in EC compared to FM and EP, which resulted in higher growth performance in fish fed the EC diet. The similar trend of high enterosomatic index and preferable growth performance in juvenile tuna fed the EC diet in trial 2 strongly supports the above suggestion. Yellowtail Seriola quinqueradiata enlarged their enteral organ including pyloric ceca to improve assimilation of less digestible FM diets than raw fish feed, without remarkable differences in weight gain and growth performance.19–21
Diets FM, EC and EP supplemented with bonito oil led to higher carcass lipids for juvenile PBT than SL in trial 1. Diet EC also induced higher carcass and liver lipids together with higher plasma cholesterol and phospholipid levels than SL in trial 2. Moreover, linear relationships were obtained between dietary and carcass fatty acid compositions in juvenile PBT fed diets EC and SL. These results indicate that juvenile tuna utilize bonito oil as an energy source less efficiently than SL oil and their carcass fatty acid composition also reflected dietary fatty acid composition. Harpaz22 suggested that juvenile PBT, a stage with a steep weight gain, were deficient in carnitine, which is a vitamin-like substance and a carrier of long-chain fatty acids into mitochondria for oxidation and energy production. Sand lance contain much carnitine in their cells, but in fishmeal carnitine is thought to be low due to processing. Moyes et al.23 showed that cardiac and skeletal muscles of skipjack tuna Katsuwonus pelamis had between 30 and 80% more mitochondrial protein per gram of tissue than common carp Cyprinus carpio, respectively. Further studies are required to ascertain carnitine requirement and fatty acid metabolism in juvenile PBT.
The calorie/protein ratios of diets EC and SL were 435 and 360 kJ/kg diet per % diet protein, respectively. It is suspected that the imbalance of protein and lipids in diet EC provides slightly lower growth performance, and higher carcass lipid content, plasma cholesterol and plasma phospholipid levels in juvenile PBT than SL. Suitable dietary calorie/protein ratios for juvenile yellowtail and red sea bream were near 313 and 350, respectively.24,25 Juvenile PBT may require high dietary protein for supporting a steeper growth than yellowtail and red sea bream. It may also possible that dietary carnitine deficiency leads low lipid utility. Otherwise, there is the possibility that the different lipid class of bonito oil (mainly non-polar lipid) and SL oil (both non-polar and polar lipid) affects lipid utility and fatty acid metabolism of juvenile tuna.
From the present study, we ascertained that EC had high availability and utility as a dietary protein source for formulating an artificial tuna diet. We need to clarify suitable protein, lipid and sugar contents, as well as carnitine and some vitamin requirements, in the near future.
The authors thank Professor H. Kumai, director of the Fisheries Laboratory, Kinki University, and staff for kind advice and assistance. This research was partially supported by a Grant-in-Aid for Scientific Research (S) 14104007 from JSPS and 21st Century COE Program ‘Center of Aquaculture Science and Technology for Bluefin Tuna and Other Cultivated Fish’ from the Ministry of Education, Science, Sports and Culture of Japan.