2.1. ETHICAL STATEMENTS:
This study was carried out with the agreement of the Tamil Nadu Dr. J. Jayalalithaa Fisheries University ethical committee in Nagapattinam, Tamil Nadu, India.
2.2. Experimental system
The experiment was conducted for 90 days at the Marine Research Farm Facility, Fisheries College and Research Institute, Thoothukudi (India) using 250-L rectangular FRP tank (120 × 55 × 38cm) with adequate oxygen levels in tanks were maintained for fish rearing. Silver pompano seeds (0.35 ± 0.06 g) were procured from the ICAR-Vizhinjam regional center of Central Marine Fisheries Research Institute, India and acclimatized for 4 week with 50 % crude protein commercial diet (Avanti Feeds Ltd., India). After the acclimatization, three hundred seeds were randomly distributed into 15 FRP tanks each consisting of 20 fish and all the fish were bulk weighed to determine the mean weight of a fish (0.93 ± 0.07 g).
2.3. Chemical analysis
Chemical analysis of ingredients, experimental diets, fish and feces were examined in triplicates using standard protocols Association of Official Analytical Chemists (AOAC 2005) as follows: moisture content was determined by drying the sample in the oven at 105 ⁰C for 24 h; Crude protein (N × 6.25) was determined by Kjeldahl method with FOSS 8400 analyzer (Tecator systems, Hoganas, Sweden); Crude lipid was determined by Soxhlet extraction using ether in Soxhlet apparatus (FOSS soxtec2043) and crude fiber was determined by using Fibertec system. Total ash content was measured by combustion of the sample at 600 ⁰C for 6 h in a muffle furnace. The gross energy content of the sample was determined using an adiabatic bomb calorimeter (Lab-X, India). Chromium (Cr) content in the diet and feces sample were determined by an inductive coupled plasma atomic emission spectrometer (Arcos-SOP, Spectro Analytical Instruments GmbH, and Kleve, DE). At the final sampling, fish from each tank were randomly taken for carcass composition analysis.
2.4. Experimental diets
Dried silkworm pupae was obtained from a silk processor (Silver mine silk processors, Erode, India) and ground into powder. The powdered meal was deoiled by the solvent extraction method (using n-hexane and ethanol) and sieved through a 500-micron test sieve. Five isonitrogenous (42%) and isocaloric (4000 kcal/kg) experimental diets ((RD, 25FMODSWP, 50FMODSWP, 75FMODSWP and 100 FMODSWP) were prepared by substituting fish meal with DSWP at 0, 25, 50, 75 and 100 % levels. The feed ingredients were obtained from the local shop were finely powdered and sieved using 200 micron sieve. The proximate composition of major ingredients were analyzed for the preparation of experimental feed formulae (Table 1). In all the experimental diets, 1% chromic oxide was added as an external digestibility marker. Feeds were pelletized using pelletizer, dried in hot air oven at 45⁰C and stored until use.
2.5. Feeding and sampling of fish
The fish were hand-fed to apparent satiation three times daily (09:30, 12:30 and 15:30h) with the respective experimental feeds. Initially 1 mm dia feed was given to the fish and later increased the feed size to 1.2 and 1.4 mm according to the mouth size of the fish during the course of experiment. Water quality parameters such as dissolved oxygen (DO), water temperature, salinity, pH, hardness, alkalinity, ammonia, nitrite and nitrate were measured for each tank by using standard procedures American Public Health Association (APHA 2005) and about 90 % of water was exchanged daily. Fish were weighed (bulk) for every 15 days to assess the mean body weight (g) using a digital electronic balance (Shimadzu Corporation, Philippines). At the end of growth trial, fish in each tank were weighed (bulk). The growth performance was assessed using the following standard formulae for different growth parameters and body indices.
2.6. Digestibility trial design
The digestibility trial was conducted in the circular experimental troughs (100 L) for 4 weeks with one reference and four groups in triplicate as described by Manikandan and Felix (2021). After the completion of growth study, fish with a size (11.28± 2.57 g) were stocked at a density of 10 fish per trough. Experimental troughs were closed using a net to prevent leaping out of fish. Fishes were hand-fed to apparent satiation three times daily (9:30, 12:30 and 15:30h). Water exchange was carried out daily morning (08:00h) and fish were fed with 1.8mm size feed without feed fines. After 30 minutes of first feeding, uneaten feed was siphoned out and discarded. Two hours after the first feeding, the first feces collection was done by using plankton net. The collected feces in the plankton net were dried and stored in the freezer as described by Cruz-suarez et al. (2009). The protocol was repeated after the second and third feeding of fishes and feces were collected 3 times in a day. For analysis, the feces from 4 consecutive weeks were pooled. Thus, a total of five samples (1 sample/ treatment) of feces were obtained and each sample was subjected to triplicate analyses. The Apparent Digestibility Coefficient (ADCs) was assessed by following standard formulae
2.7. Sample collection
At the end of the growth trial, fish from each tank in a group were randomly collected, anesthetized with 0.10 mL clove oil / L water (Hamackova et al., 2006) and used for body indices analysis, digestive enzyme analysis, hematological analysis, serum biochemical analysis and histopathological analysis (in duplicates).
Body indices analysis
For body indices analysis, the liver and intestine were dissected from three anesthetized fish in each group, weighed and measured by using the following standard formula.
Hematological and serum biochemical analysis
After anesthetizing fish using clove oil (0.10 mL/L), blood samples were collected by caudal vein puncture (Prabu et al., 2016). Three blood samples were collected from each group using a 1 mL disposable syringe fitted with a 26-gauge needle rinsed with 2.75% EDTA. The collected samples were immediately transferred from the syringe to a separate heparin-coated vial placed on crushed ice for hematological analysis. Then another three blood samples were taken without an anticoagulant, transferred to plain vials and used for serum biochemical analysis. Hematological parameters were determined by placing the whole blood samples in an automated hematology analyzer (Medonic M-series, Boule, India). Serum cholesterol was estimated by the cholesterol oxidase/peroxidase (CHOD/PAP) method (Meiattini et al., 1978) using a cholesterol kit (Coral Clinical Systems, Goa, India). Serum triglyceride was determined by the glycerol phosphate oxidase/peroxidase (GPO/PAP) method (Fossati and Prencipe, 1982) using a triglyceride kit (Coral clinical systems, Goa, India). Lipoproteins (HDL and LDL) were analyzed using commercial kits (Coral clinical systems, Goa, India) and Serum very low-density lipoprotein (VLDL) level was calculated as per the methods of Friedewald et al. (1972),.VLDL = Triglycerides/5. Serum total protein was determined by the Biuret method (Reinhold, 1953) using a protein estimation kit (Coral clinical systems, Goa, India). Albumin was estimated by the Bromocresol green binding method (Doumas et al., 1971) using the kit (Coral clinical systems, Goa, India), whereas globulin level was determined by deducting the values of serum albumin from serum protein.
Digestive enzyme analysis
The liver and intestine samples were collected from the fish in each replicate tank. Immediately after collection of liver and intestine, homogenized with 0.25 M sucrose buffer by taking 5% (w/v) and centrifuged at 10000 rpm for 30 min. The supernatant was collected in tube (15 mL) and stored at -20°C. The supernatant was used as an enzyme extract. Protease activity was determined by the casein digestion method (Drapeau 1976). The α-amylase activity on carbohydrates was estimated by the dinitro-salicylic acid method (Rick and Stegbauer 1974) and lipase activity was determined by following the method described by Zamani et al. (2009).
Histopathological analysis
One anaesthetized fish from each tank (3 fish/ group) were randomly taken for histopathological analysis. The intestine of fish was dissected and collected in 10 % neutral buffered formalin fixative solution. The method described by Luna (1968) was followed to analyze histopathological observation of intestine. A light microscope was used to observe any abnormalities on the intestine of experimental fishes.
2.8. Statistical analysis
The results are presented as means ± SD. Parameter-wise data sets were statistically analyzed using analysis of variance (ANOVA) followed by Tukey's test. Statistical significance was tested at p<0.05. All the statistical tests were performed using IBM SPSS statistical software (Version 26.0). Second-order polynomial regression analysis was carried out to optimize the level of inclusion of DSWP to replace FM in the diet of silver pompano juveniles. All the statistical tests were performed using IBM SPSS statistical software (Version 26.0).