Growth, biochemical response and liver health of juvenile barramundi (Lates calcarifer) fed fermented and non-fermented tuna hydrolysate as fishmeal protein replacement ingredients

Ethics statement

The growth trial of the fish was performed in accordance with the Australian code of conduct for the use and care of animals for scientific purposes. The procedures and protocols used in this experiment for treating fish were approved by the Animal Ethics Committee of Curtin University, Australia (Approval Number: AEC_2015_41).

Fish and rearing conditions

Juvenile barramundi, obtained from the Australian Centre for Applied Aquaculture Research, Fremantle, WA, Australia were used for the experiment. Prior to the commencement of feeding, the fish were acclimated to the laboratory conditions for one week; during that period they were fed on a commercial barramundi diet containing 470 g protein kg⁻¹ diet and 20.0 MJ kg⁻¹ dietary gross energy three times a day. The barramundi juveniles, with a mean initial weight of 6.75 ± 0.16 g fish⁻¹, were stocked randomly into fifteen tanks (capacity: 250 L each), at a stocking density of 20 fish tank⁻¹. The water in all the tanks used for this experiment was recirculated independently from an external bio-filter with a water refreshment capacity of 10 L min⁻¹ and was constantly aerated to ensure sufficient levels of dissolved oxygen.

The water temperature, salinity, pH, and the level of dissolved oxygen in the water were monitored daily, and the levels of nitrite and ammonia were monitored twice a week. The water quality parameters maintained during the experimental period were as follows: water temperature = 27.8–29.7 °C, salinity = 31–38 ppt, pH = 7.20–8.10, dissolved oxygen concentration = 5.65–7.45 mg L⁻¹, nitrite <0.25 mg L⁻¹, and total ammonia nitrogen <0.5 mg L⁻¹. The room for the experiment was maintained under a 12-h light–12-h dark cycle, using automatic indoor-light switches (Clipsal, Australia). Prior to feeding, the daily feed ration for the fish was divided into three equal portions, and then the fish were hand-fed to satiation with the respective experimental diets, three times a day, at 0,800 h, 1,300 h, and 1,800 h. After 0.5 h of each time, the uneaten feed was removed from the tank by siphoning, transferred to aluminum cups, and dried to constant weight, in order to evaluate the feed conversion ratio.

Preparation of fermented TH

Tuna hydrolysate (TH) supplied by SAMPI, Port Lincoln, Australia, was fermented by following the technique proposed by Himawan et al. (2016). Baker’s yeast Saccharomyces cerevisiae (Instant dried yeast, Lowan^®, Glendenning, Australia), and Lactobacillus casei in the form of a skim-milk product (Yakult^®, Tokyo, Japan), were used at 10% and 5% (cell density for Lactobacillus casei = 3 × 10⁶ CFU g⁻¹ meal) of the total weight of the meal mixture, respectively, and distilled water was used at approximately 70% of the total weight of the meal ingredients; all these ingredients were homogenized together in a food mixer. The mixture was transferred into an Erlenmeyer flask, which was then covered with aluminum foil and incubated at 30 °C for 4 d. Following this, the fermented product was dried in an oven at 60 °C for 24 h, and then used as a feed ingredient.

Experimental diets

All the feed ingredients used in this study, except TH, were obtained from Specialty Feeds Pty. Ltd, Great Eastern Highway, Western Australia; TH was provided by SAMPI, Port Lincoln, Australia. The diets were formulated to fulfill the nutritional requirements of the juvenile barramundi set by the NRC (2011). Two series of experimental diets, isonitrogenous and isocaloric, were formulated for the juvenile barramundi, containing approximate 47% crude protein (CP) and 20 MJ kg⁻¹ gross energy (GE), respectively. The diets were labeled as TH₅₀ and TH₇₅ (non-fermented TH), and FTH₅₀ and FTH₇₅ (fermented TH). The diets were formulated to replace FM at 50% and 75% inclusion levels of TH/FTH. A basal diet with FM as the sole protein source was used as the control diet. Another set of diets was formulated with the addition of 5 g chromic oxide (Cr₂O₃) per kg of diet as an inert marker for assessing the digestibility in the fish (Cr₂O₃; Thermo Fisher Scientific, Scoresby, VIC, Australia). All the test diets were processed with the addition of water to about 35% mash dry weight of the mixed ingredients in order to form a dough. The dough was passed through a mincer in order to create pellets of desired diameter (3 mm). The moist pellets were oven dried at 60 °C for 48 h and then, cooled at room temperature, sealed in plastic bags, and stored at −15 °C until further use. The formulations and the proximate compositions of the experimental diets are presented in Table 1.

Digestibility assessment

In order to estimate the apparent digestibility coefficients (ADCs) for dry matter, protein, and lipid, fecal matter was collected using the stripping technique (Austreng, 1978), 15 h post feeding, prior to terminating the feeding trial. All the fecal matters collected from each tank within each period were pooled and frozen at −20 °C immediately. Prior to commencing the analysis, the fecal samples were oven-dried to a constant weight at 105 °C. The chromium oxide content in the diet formulations and the fecal samples was analyzed by the method described by Cho, Slinger & Bayley (1982). The ADCs for each nutritional component in the test diets were calculated based on the formula given below: ${\displaystyle \text{ADC}=100-100\ast {\left[\text{marker in feces}\left(\text{\%}\right)∕\text{marker in diet}\left(\text{\%}\right)\right]}^{\ast }}$ ${\displaystyle \left[\text{nutrient in feces}\left(\text{\%}\right)∕\text{nutrient in diet}\left(\text{\%}\right)\right]}$

Sampling and chemical analysis

At the termination of the trial, the fish fasted for 24 h. The fish were then anesthetized with 5 mg L⁻¹ of AQUI-S (Australia), followed by bulk weighing in order to assess the final body weight (FBW), specific growth rate (SGR), feed conversion ratio (FCR), and survival rate for the fish. Blood samples were collected from the caudal veins of three fish from each tank, using a 1-mL plastic syringe and a 22G × 1∕2″ straight needle. The extracted blood sample was transferred to heparinized tubes for the analysis of biochemical indices of blood, such as hemoglobin (Hb) content, hematocrit concentration, leucocrit concentration, and the glutathione peroxidase (GPx) enzyme activity. Hematocrit (Hct,%) and leucocrit (%) concentrations were determined by using the standard McLeay and Gordon’s method (McLeay & Gordon, 1977), following a centrifugation at 2,000 rpm for 5 min. The hemoglobin (Hb,%) content was determined by using an Hb kit (Randox Laboratories, Antrim, United Kingdom). The GPx activity (GPx units g⁻¹ Hb) in the red blood cells was assessed by using the Ransel RS–505 assay kit (Randox, Antrim, United Kingdom).

The moisture, crude protein, crude lipid, ash, and gross energy contents in the experimental diets were assessed using the methods and procedures given by the Association of Official Analytical Chemists (AOAC, 2006). Moisture in the samples was determined by oven-drying to a constant weight at 105 °C; ash content was determined by combustion at 550 °C for 24 h in an electric furnace (Carbolite, Sheffield, UK); crude protein content (N × 6.25) was determined by following the Kjeldahl method, using a Kjeltec Auto 1030 analyzer (Foss Tecator, Höganäs, Sweden); crude lipid was analyzed by following the Soxhlet technique, using a Soxtec System HT6 (Tecator, Höganäs, Sweden); and the gross energy content was determined by using a bomb calorimeter (Heitersheim, Germany). The composition of amino acids, except tryptophan, in the tested diets, was analyzed by using high-performance liquid chromatography (HPLC), following an acid hydrolysis.

Histopathology

In order to analyze the histopathological conditions, one liver segment from an individual fish, i.e., six liver segments from each treatment were sampled. The liver tissue samples were excised and preserved in 10% buffered formalin until they were processed using the standard histological procedures. The blocks of the designated samples were dehydrated in 100% ethanol, and then, embedded in paraffin wax. The sections of approximately 5 µm in size were cut and stained with Hematoxylin-Eosin (H&E) stain, for histological examination under a light microscope (BX40F4, Olympus, Tokyo, Japan). All the samples were prepared using the standard histological techniques (Luna, 1968).

Calculations

The fish fasted for 24 h prior to weighing and sampling, and the following parameters were measured after the growth trial of 56 d: ${\displaystyle \text{Weight gain, WG}=\left[\text{(mean final body weight}\text{—mean initial body weight)}\right]}$ ${\displaystyle ∕\left[\text{mean initial body weight}\right].}$ ${\displaystyle \text{Specific growth rate, SGR}\left(\text{\%}∕\text{day}\right)=\left[\text{(ln mean final body weight}\text{—in mean}\text{initial body weight)/number of days}\right]\times 100}$ ${\displaystyle \text{Feed intake, FI}=\text{dry feed consumed/number of fish.}}$ ${\displaystyle \text{Feed conversion ratio, FCR}=\text{dry feed fed/wet weight gain.}}$ ${\displaystyle \text{Condition factor, CF}\left(\mathrm{g}{\left({\mathrm{cm}}^3\right)}^{-1}\right)=\left[\left(\text{body weight, g}\right)∕{\left(\text{length, cm}\right)}^3\right]\times 100.}$ ${\displaystyle \text{Hepatosomatic index, HSI (\%)}=\text{[(liver weight, g)/(body weight, g)]}\times 100.}$ ${\displaystyle \text{Viscerosomatic index, VSI (\%)}=\text{[(visceral weight, g)/(body weight, g)]}\times 100.}$ ${\displaystyle \text{Survival rate, SR}=\text{(final number of fish/initial number of fish)}\times 100}$ ${\displaystyle \text{Skewness}=\left[1∕n{\Sigma }_{i=1}^n{\left(x_i-\overline{x}\right)}^3\right]∕{\left[1∕n{\Sigma }_{i=1}^n{\left(x_i-\overline{x}\right)}^2\right]}^{3∕2}}$ where Σ is the summation for all the observations (x_i) within a sample, and $\overline{x}$ is the sample mean.

Statistical analysis

The statistical analyses were performed using SPSS version 24 for Windows, IBM, Curtin University, Australia. The experimental data for growth performance, digestibility, hematological parameters, and body composition were subjected to one-way ANOVA, followed by Duncan’s multiple-range test. The data on length and weight distribution represented through skewness were executed together with the normal distribution test. Data values were expressed as a mean ± standard error in the triplicate tanks, and the threshold of statistical significance was set at p < 0.05.

Growth performance, feed utilization and somatic indices

The growth performance, feed utilization, and somatic indices for the juvenile barramundi fed on diets containing different inclusion levels of fermented and non-fermented TH for 56 d are presented in Table 2. At the end of the growth trial, it was observed that the growth performance parameters, including FBW, WG, and SGR, decreased as the levels of inclusion of TH and FTH in the diets increased. The control treatment demonstrated best growth performance, and the lowest growth performance was observed for the 75% replacement level in both fermented and non-fermented treatments. Similarly, highest FI was observed in the control group of fish, which was fed on the FM-based diet. The FCR was significantly higher in the fish fed on TH₇₅ diet compared to the fish which were fed on all the other diets, except for the fermented-TH group of the same replacement level—FTH₇₅. No significant difference was observed in the FCR among the groups of fish fed on the control, TH₅₀, and FTH₅₀ diets. The body indices, such as VSI, HSI, and CF, and the survival rate of the fish were not affected by the inclusion of TH or FTH in the diets. The distributions of length and weight of the juvenile barramundi fed on different diets are presented in Figs. 1 and 2, and the respective values for skewness are presented in Table 2. The variations in the length and weight of the fish that were fed on different diets were not significant compared to control. Negative skewness, in terms of length, was obtained for the control, TH₅₀, and FTH₅₀ groups; whereas, negative skewness, in terms of weight, was observed for the control and TH₅₀ groups.

Digestibility

The digestibility coefficients for dry matter (DM), protein, and lipid in the juvenile barramundi fed on the diets that contained TH and FTH at various inclusion levels are enlisted in Table 3. As seen in Table 3, the higher inclusion levels of TH and FTH in the diets resulted in a significant decrease (p < 0.05) in the apparent digestibility coefficients (ADCs) for DM, protein, and lipid, with the lowest and the highest values of the ADCs for DM, protein, and lipid obtained for the TH₇₅ and the control group, respectively.

Whole fish body composition

The whole-body proximate compositions (moisture, protein, lipid, and ash) and the energy content for the fish belonging to the different treatments are presented in Table 4. As seen in Table 4, the whole-body proximate composition and the gross energy of the juvenile barramundi were neither influenced by the TH types (fermented or non-fermented) nor by the levels of replacement (p > 0.05).

Biochemical status

The blood biochemical indices and the glutathione peroxidase (GPx) enzyme activity for the juvenile barramundi are presented in Table 5. The dietary inclusion of TH and FTH exhibited no significant effect on the blood hemoglobin, hematocrit, and leucocrit levels in the fish. However, the incremental inclusion of TH and FTH exhibited significant effects on the glutathione peroxidase (GPx) enzyme activity in the juvenile barramundi. The fish fed on the FM-replacement diets TH₅₀, FTH₅₀, and TH₇₅ exhibited significantly reduced GPx activity compared to control; whereas, the 75% FM-replacement diet FTH₇₅ exhibited no significant difference in the GPx activity compared to the control. Furthermore, the GPx activity was significantly increased in the fish fed on fermented diet compared to the fish fed on the non-fermented diet, when the replacement level was 75%; however, at 50% replacement level, the difference in the GPx activity was not significant between the fermented and the non-fermented diets (p > 0.05). The antioxidant glutathione peroxidase (GPx) activity values obtained for the juvenile barramundi belonging to different treatments are presented in Fig. 3.

Liver histopathology

The control fish exhibited a normal liver condition, which is characterized by hexagonal hepatocytes with a round, central nucleus, and a rare occurrence of cytoplasmic vacuolization or granules (Fig. 4A). The excessive use of TH in the diets resulted in several alterations in the liver tissue of the fish. The alterations, including the occurrence of cytoplasmic vacuolization with lipid accumulation (steatosis), were observed in the fish fed on the TH₇₅ diet (Fig. 4B). In the fish fed on the FTH₇₅ diet, severe and irreversible damages were observed in the liver, which were characterized by disorganized hepatic cordons, cellular deformation, and nuclear hypertrophy. In further damaged livers, in addition to the other alterations, the absence of nucleolus and the presence of necrotic foci were observed (Fig. 4C).