Poultry by-product meal as an alternative to fish meal in the juvenile gilthead seabream (Sparus aurata) diet

Aquaculture 511 (2019) 734220

Contents lists available at ScienceDirect

Aquaculture journal homepage: www.elsevier.com/locate/aquaculture

Poultry by-product meal as an alternative to fish meal in the juvenile gilthead seabream (Sparus aurata) diet

T

⁎

Maroua Sabbagha, ,1, Roberta Schiavonea, Giulio Brizzib, Benedetto Sicuroc, Loredana Zillia, Sebastiano Vilellaa a

University of Salento, Di.S.Te.B.A Laboratory of Comparative Physiology, Ecotekne, Lecce, Italy Chlamys Srl, Marine Ecology and Aquaculture, Trani, Italy c Department of Veterinary Sciences, University of Turin, Italy b

A R T I C LE I N FO

A B S T R A C T

Keywords: Fish nutrition Poultry by-product meal Gilthead seabream Growth performances Fish welfare Fillet quality

The aim of the present study was to evaluate the eventual possibility of fish meal (FM) substitution in the diet of farmed gilthead seabream (Sparus aurata) by poultry by-product meal (PBM) and to verify PBM suitability in aquafeeds. Inclusion of PBM in the seabream diet was therefore investigated by a multidisciplinary analysis to evaluate its possible effects on fish growth performance, fish welfare and fillet quality. A control diet and two experimental diets, Feed A and Feed B containing PBM with 50% and 100% of FM substitution respectively, were formulated. All diets were isoproteic (45%), isolipidic (20%) and isoenergetic. The growth trial lasted 110 days, including 20 days of fish acclimatization. Juvenile gilthead seabream with an initial average weight of 73.57 ± 10.47 g were allotted at random in nine tanks (three replicates per diet), fed once a day by hand (feeding rate 1%). As a result, average weight gain increased in all fish groups without any statistically significant difference (P > .05). Measured zootechnical parameters were similar among fish groups, the condition factor as an indicator of fish condition was around 2 and the survival rate was 100%. Investigations through haematological parameters, digestive enzymes and liver histology analyses demonstrated that no statistical difference existed among dietary treatments and this clear evidence attests that PBM inclusion in seabream diets did not have a negative effect on fish welfare. Protein patterns obtained from fish fed the control diet and PBM diets showed a similar expression of structural proteins such as actin, tropomyosin, myosin light chain 1 (MLC1), myosin light chain 2 (MLC2) and fast skeletal myosin light chain (MLC3). Results for fish fillet compositions were comparable in all fish groups with some exceptions for fatty acid composition. The gross energy content of seabream muscle was also not affected by PBM. The present study demonstrated that the total substitution of FM with PBM in the commercial diet of gilthead seabream (S. aurata) is achievable without compromising fish growth performance, fish welfare and fillet quality and suggests that PBM could be considered as a good sustainable raw material for fish food.

1. Introduction The growth of aquaculture and the global consumption of farmed fish have been steadily increasing in the last decade (FAO, 2018). Feed represents one of the major running costs in aquaculture, and understanding the effect of nutrition strategy on the quality and marketability of farmed fish is of crucial importance for farming profitability (Baghaei Jezeh et al., 2014). Fish meal (FM) is currently one the major source of high value proteins in fish feed (Fasakin et al., 2003; Tacon and Metian,

2008); however, its limited availability as a result of various factors as overfishing, stagnation of the global capture fisheries and climate change particularly the El Niño phenomenon (Hardy and Tacon, 2002; Schreiber et al., 2011), has notably increased the cost of this commodity (Auchterlonie, 2017). Consequently, the substitution of FM presented a real challenge for fish feed suppliers and has stimulated nutritionists and researchers worldwide (Hernández et al., 2007; Martínez-Llorens et al., 2009; Pavan Kumar et al., 2014) to look for sustainable, effective and cheaper protein sources (Hardy, 2006, 2010; Shepherd et al.,

⁎

Corresponding author. E-mail addresses: [email protected] (M. Sabbagh), [email protected] (R. Schiavone), [email protected] (B. Sicuro), [email protected] (L. Zilli), [email protected] (S. Vilella). 1 Present affiliation: Chlamys Srl, Marine Ecology and Aquaculture, Trani, Italy https://doi.org/10.1016/j.aquaculture.2019.734220 Received 30 November 2018; Received in revised form 10 June 2019; Accepted 11 June 2019 Available online 12 June 2019 0044-8486/ © 2019 Elsevier B.V. All rights reserved.

Aquaculture 511 (2019) 734220

M. Sabbagh, et al.

(Benedito-Palos et al., 2008; Grigorakis, 2007; Roncarati et al., 2010; Yildiz et al., 2008). Furthermore, the essential amino acid (EAA) deficiency is regarded as one of the most important issues on FM substitution (Kaushik and Seiliez, 2010), and unbalanced dietary EAA levels have been reported as one of the principal causes for growth depression in fish fed animal by-products-based diets (García-Gallego et al., 1998; Millamena, 2002; Xavier et al., 2014). In the present study a multidisciplinary analyses such as digestive enzymes activities, haematological parameters, liver histology, amino acid and fatty acid profiles of fish fillet, lipid composition, lipid peroxidation and proteomic analysis were carried out, at the aim of evaluating the possible effects of FM substitution by PBM in the diet of farmed gilthead seabream.

2005). This may help in shortening the pathway of raw materials supplying, increasing the ecological sustainability of fish farming and supporting fishing effort reduction on wild stocks to produce FM. FM substitution by less expensive plant proteins has progressively increased in the aquafeed industry, although the use of these resources in fish diets is hampered by the presence of a variety of endogenous anti-nutritional factors (Francis et al., 2001; Gatlin et al., 2007; Kaushik, 1990), deficiency in some indispensable amino acids (Hardy, 2006) and reduction in protein digestibility (Santigosa et al., 2008). Poultry by-product meal (PBM) is considered to be a suitable alternative protein of FM (Gaylord and Rawles, 2005; Rawles et al., 2006; Thompson et al., 2007) in artificial diets for carnivorous and omnivorous aquaculture species because of its high production volume, nutritional composition and price compared to FM (Hardy, 2000; Yu, 2004). The EC Regulation No 56/2013 has authorized the use of PBM, as processed animal protein (PAP), in the diets of non-ruminant farmed animal including fish. PBM is a palatable and high-quality feed ingredient in fish nutrition, because of its high protein content and essential amino acids profile similar to FM, except for lysine and methionine (Bureau et al., 1999, 2000; González-Rodríguez et al., 2016; NRC, 2011; Zhou et al., 2004). As demand in aquafeeds is increasing, several authors confirmed that PBM has a considerable potential as a feed ingredient in fish production systems (Bureau et al., 2000; Fasakin et al., 2005; Millamena, 2002). Some studies have shown that PBM could replace 75% or even 100% of FM without significant decrease in fish growth. In fact, PBM could replace 75% of FM in the diet of juvenile gilthead seabream (S. aurata) without amino acid supplementation (Nengas et al., 1999) and up to 100% for red seabream (Pagrus major) (Takagi et al., 2000). The recent study of Karapanagiotidis et al. (2019) demonstrated that 50% of FM can be replaced by PBM without compromising growth performance, feed utilization, proximate composition and did not have negative effects on digestive protease activities (trypsin and chymotrypsin), and on haematological and biochemical parameters. Cuneate drum, Nibea miichthioides, was fed successfully with 50% of FM replaced by PBM (Wang et al., 2006). Kureshy et al. (2000) were able to raise juvenile red drum, Sciaenops ocellatus on a diet with 66.7% of PBM. The inclusion of PBM in the diets of freshwater fish was also proven successful. Muzinic et al. (2006) replaced 100% of FM protein with turkey meal without any negative effects on the growth of sunshine bass (Morone chrysops x M. saxatilis). Gibel carp (Carassius auratus gibelio) grew well by the substitution of 50% of FM by PBM protein (Yang et al., 2006). However, González-Rodríguez et al., 2016 demonstrated that PBM could replace only 25% of FM protein in the diet of tench (Tinca tinca) without impairing growth performance. Gilthead seabream is among the most important marine finfish species in the Mediterranean (Gómez-Requeni et al., 2004) and its production is still expanding rapidly. In the Mediterranean, S. aurata production in Yones and Metwalli, 2015 was about 160,000 t (according to FAO, 2017). Only a few studies have tested the substitution of FM by PBM in the diet of S. aurata (Alexis, 1997; Karapanagiotidis et al., 2019; Lupatsch et al., 1997; Nengas et al., 1999) and the possible effects of significant percentage of PBM in the diet of gilthead seabream are not yet fully understood. The effects of dietary changes on fish growth performance, welfare and fillet quality are investigated through the assessment of digestive enzymes activities that reflect the suitability of protein sources in the diets (Silva et al., 2010). The assessment of haematological parameters represents an effective and sensitive index to monitor fish physiological changes (Satheeshkumar et al., 2011) and health status (De Pedro et al., 2005; Metochis et al., 2017), particularly interesting when possible physiological alterations induced by fish diet composition are investigated. The characterization of fish tissue proteome investigates how tissues protein pattern of the organism is affected by nutritional changes (Rodrigues et al., 2012; Addis, 2013), and muscle lipid composition may be affected quantitatively and qualitatively by fish diet

2. Materials and methods 2.1. Experimental diets Two different diets, Feed A and Feed B, were formulated with increasing levels of PBM inclusion (50% and 100% of FM substitution respectively). The substitution of FM by PBM was made on the basis of their inclusion levels in the diets. These diets were tested against a control diet (containing 36% of FM) without PBM. Fish feeds were prepared on request by a commercial company – the Veronesi Group (Verona, Italy) – as extruded pellets (3 mm in diameter). All diets were isoproteic (45% crude protein) and isolipidic (20% crude fat) and isoenergetic (Table 1). Formulation and proximate composition analysis of the three diets were carried out and certified by the feed manufacturer. The experimental diets were formulated at a balanced amino acid profile that complies with the nutritional requirements of gilthead seabream (S. aurata) (Table 2). Amino acids profile of the three diets was carried out and certified by the feed manufacturer. 2.2. Experimental design and growth trial A growth trial of 110 days was carried out in an industrial aquaculture facility belonging to the commercial company Panittica Pugliese Srl (Torre Canne, Brindisi, Italy). A total of 1848 juvenile seabreams (S. aurata) were used. An initial random sampling of 30 fish was carried out to determine the initial mean body weight (IBW) and the initial mean standard length (SL). The remaining fish (1818) were allotted at Table 1 : Formulation and proximate composition of the experimental diets. Diets

Control diet

Feed A

Feed B

18.6 13.0 8.0 7.0

18.6 13.0 – 10.8 4.5 4.5 7.5 0.1 0.250 0.750 45 20 1.6 8.5 2.15 1.35 0.18 23.1

Formulation(%) Soy meal Wheat Pea protein Wheat gluten Rapeseed oil Soy oil Fish oil D-lysine DL-methionine Vitamin and Mineral premix

9.0 7.5 – 0.160 0.735

18.6 13.0 – 3.0 4.5 4.5 7.5 – 0.200 0.700

Proximate composition (%) Crude Protein Crude Fat Crude Fiber Ash Calcium Phosphorus Sodium Gross energy (MJ/Kg)

45 20 1.8 7.5 1.22 1.04 0.45 23.3

45 20 1.6 9.6 2.2 1.48 0.4 22.9

Formulation data are expressed as % of ingredients on wet matter. Proximate composition data are expressed as % on dry matter. 2

Aquaculture 511 (2019) 734220

M. Sabbagh, et al.

Fish muscle dissection took place as follows: a piece of muscle was taken from the fish, the skin was removed, and the central core of the muscle used for fillet quality analysis (proteomic analysis, amino acid profile, fatty acid profile, lipid peroxidation, etc.). Muscle samples were extracted from the loin muscle, under the dorsal fin and above the lateral line and then conserved at −80 °C for further analysis.

Table 2 : Amino acids profile of the three experimental diets (% of diet). Amino Acids Profile (%)

Control diet

Feed A

Feed B

Methionine Methionine + Cysteine Lysine Threonine Valine Arginine Tryptophan Isoleucine Leucine Phenylalanine Glycine

1.08 1.65 2.87 1.72 2.11 2.87 0.5 1.86 3.29 2.03 2.81

1.08 1.58 2.72 1.71 2.07 2.85 0.48 1.8 3.18 1.87 3.46

1.12 1.68 2.78 1.65 2.08 2.87 0.47 1.82 3.26 1.99 3.49

2.6. Blood plasma analysis Before the dissection procedure, blood (1 ml) was collected from all sampled fish from the caudal vasculature in heparinized syringes and plasma was separated by centrifugation at 2500g for 20 min and stored at −20 °C for future use. Plasma protein concentration was measured by the Bradford assay (Bradford, 1976) and plasma osmolality was measured using an automatic osmometer (5520 VAPRO, Delcon). Cortisol was determined by a commercially available immunoassay kit (EIA cortisol kit, Cayman Chemical, Ann Arbor, MI). Commercial kits (Sigma-Aldrich) were used to measure alanine aminotransferase (ALT) and aspartate aminotransferase (AST) activities.

random to nine tanks, each containing 202 fish. Each tank had a volume of 1000 L. The initial average fish density was 14.86 kg/m3. Water renewal was by running water at rate of 1.5 renewals/h. The water temperature was stable at 19.4 ± 0.16 °C. Oxygen was dissolved by pump into the water input and was > 100% of saturation. These parameters were measured every two days using an oxymeter. Before starting the growth trial, fish were acclimatized to the experimental tanks for 20 days and fed only the control diet. The experimental design was monofactorial, balanced with three replicates per diet (3 × 3) and the experimental factor was the diet. Fish were weighed every 15 days in order to check fish biomass gain and to adjust the 1% of biomass feeding rate (once a day). Feeding rate was chosen at a percentage close the apparent satiation at the given temperature. The three diets were given manually each day for seven days a week.

2.7. Digestive enzyme assays Nine fish per treatment (three fish per tank) were used for enzyme analysis where 0.3 g of each organ (liver and intestine) were homogenized in 3 ml of 1.1% NaCl, 0.5 mM phenylmethanesulphonyl fluoride (PMSF; 100 μl) buffer using a mechanical polytron homogenizer (Kinematica, GmbH). All analyses were carried out immediately after homogenate preparation at 4 °C. Enzyme activities were measured by the spectrophotometric method. All enzyme analyses were conducted in three replicates. In order to establish the specific activities of the enzymes, protein concentrations were determined in the enzyme extracts by the Bradford assay (Bradford, 1976), with bovine albumin as standard. Alkaline phosphatase (AP) activity was measured both in the gut and the liver following the Bessey et al. (1946) method. Lipase determination was adapted from the Albro et al. (1985) method. Leucine aminopeptidase was measured following the Storelli et al. (1986) protocol.

2.3. Sampling procedure After two weeks of adaptation, an initial sampling was carried out and 30 fish were sacrificed in iced water for 40 min, as is the normal practice in Italian fish farming. Total weight and standard length were measured to calculate the morphometric parameter (initial mean body weight, initial mean length). Each month, 24 fish were sampled randomly and sacrificed from each experimental group. Before each handling, the fish were anesthetized with phenoxy-2-ethanol, sacrificed in iced water and then transported within one hour of sample collection in a cool bag with ice to the Laboratory of Comparative Physiology (D.i.S.Te.B.A.) at the University of Salento.

2.8. Histological examination of the liver For liver histology analysis, five fish per tank were collected at each sampling. Samples were routinely dehydrated in ethanol and embedded in paraffin according to standard histological practices. Transverse sections were cut to a thickness of 5 μm and stained with haematoxylin and eosin for examination with light microscopy. The morphology of the liver was evaluated on a scale of 1 to 3 using the criteria of McFadzen et al. (1997). Grade 1 corresponds to a healthy liver with lightly granular, small and distinct nuclei and hepatocyte cytoplasm with a structure varying in texture and scattered granules with eosinpositive patches. Grade 2 corresponds to an intermediate condition. The nuclei have abundant dark granules, and the nucleoli are enlarged or indistinct. The cytoplasm is homogeneous and is vacuolised only to a very limited degree. Grade 3 indicates a degraded liver with small, dark, pyknotic nuclei. The cytoplasm is hyaline with a lack of texture and with large sinusoidal spaces. To assess liver condition, 28 sections for each diet and in each sampling were observed and analysed.

2.4. Growth performance and somatic indexes On arrival at the laboratory, the fish were individually weighed using a technical balance (Gibertini CENT 2000CAL) and the standard length was measured with a metric ruler. For each sampling, the following mean individual growth performance indeces were calculated per treatment: average weight gain (AWG), condition factor (K), specific growth rate (SGR), feed conversion ratio (FCR), viscerosomatic index (VSI), hepatosomatic index (HSI) and survival rate (SR). Of the fish, 24 per treatment (8 fish per tank) were used to determine productive parameters and 15 were used for somatic index determination. 2.5. Dissection procedure The peritoneal cavity of the fish was opened along a ventral midline incision. The entire viscera was extracted and weighed to determine the VSI. The liver was then removed and weighed to calculate the HSI. Individual livers and intestines were isolated, washed with distilled water and conserved at −80 °C for further analysis (digestive enzymes activities). A part of each liver was conserved immediately after dissection in phosphate-buffered formalin (10%) to be used in histological analysis.

2.9. Proteomic analysis At each sampling date, three fish were sampled at random in order to carry out the analysis. All analyses were performed in triplicate runs (90 gels). Proteomic analysis was carried out following the Schiavone et al. (2008) protocol. Analysis was performed by protein extraction, isoelectric focusing (IEF) using IPGphor (Amersham Biosciences) and SDS3

Aquaculture 511 (2019) 734220

M. Sabbagh, et al.

treatment were used for all lipid analyses. Cholesterol quantification was performed using the commercial Sigma-Aldrich cholesterol quantification kit. Phospholipid (PL) analysis was performed using the commercial Sigma-Aldrich phospholipid assay kit, in order to assess the effect of diets in fish muscle phospholipid content. The gross energy content of fish muscle and diets was measured by combustion in a Parr bomb calorimeter 1241 (Parr Instrument Company, Moline, IL, United States), using benzoic acid as the standard.

PAGE separation. Imagemaster 2D Elite software 3.1 (Amersham Biosciences) was used for the acquisition and analysis of 2D gels. However, we focused our study on the partial proteome with optimal resolution since for spots corresponding to proteins at higher molecular weight (> 60 kDa) the resulting resolutions were scarce. Spot differences were initially examined by visual inspection and subsequently by using image analysis software. Spots of interest were selected according to the criteria used in the Schiavone et al. (2008) study. The amount of protein in a spot was taken as the area of the spot multiplied by its density and referred to as the volume. Following removal of background, the spot volumes were normalized to the total protein detected for each gel by dividing the individual spot volume by the sum of all spot volumes and multiplying by 100. The normalized spot volume is referred to as abundance. The normalized volume of each spot considered in this study was evaluated as mean ± SD of spot normalized volumes measured in all gels. Considering that these data did not show normal distribution, a non-parametric statistic was used in the statistical elaboration (i.e. Mann-Whitney test).

2.13. Lipid peroxidation: TBARS method A commercial lipid peroxidation (MDA) Sigma-Aldrich assay kit was used in order to evaluate the effects of the experimental diets on lipid peroxidation in fish fillets. Lipid peroxidation was determined by the reaction of malondialdehyde (MDA) with thiobarbituric acid (TBA) to form a colorimetric (532 nm) product, proportional to the MDA. Thiobarbituric acid-reactive substances (TBARS) were expressed as nanomoles MDA per μl sample volume.

2.10. Amino acid analysis The amino acid profile in fish muscle (edible part) was ascertained according to the Bosch et al. (2006) protocol. Acid hydrolysis was used for all amino acids except cysteine and methionine for which performic acid oxidation and then acid hydrolysis were used. The amino acid contents were determined using an HPLC system (Agilent 1100 series) consisting of two pumps, a manual sampler and a fluorescence detector. Aminobutyric acid was added as an internal standard before hydrolysation. The amino acids were derivatised with AQC (6-aminoquinolylN-hydroxysuccinimidyl carbamate). Chromatographic separation was carried out using a C18 column (Gemini® 5 μm 150 × 4.6 mm). The column was thermostated at 37 °C and the flow rate was 1.0 ml/min. The injection volume was 10 μl and injection was carried out manually. Mobile phase A consisted of AccQ.Tag eluent A (100 ml AccQ.Tag A concentrate +1 l Milli-Q water). Mobile phase B consisted of eluent B acetonitrile-water solution (60:40, v/v). The gradient conditions shown were selected in a previous study (Reverter et al., 1997). At the beginning of the gradient, the column was equilibrated in 100% A for 10 min. Detection was carried out by fluorescence (λ excitation 250 nm and λ emission 395 nm).

2.14. Statistical analysis All data were analysed by one-way analysis of variance (ANOVA) using SPSS software (IBM SPSS Statistics 23). After the ANOVA test, the differences among means were determined by Tukey's multiple comparison test, using the significance level of P < .05. Results are presented as means ± SD. 3. Results 3.1. Growth performance and somatic index Productive parameters and somatic indexes were measured (Table 3). A special care was given to assure that all feed supplied was consumed, in fact all diets were well accepted by fish. The survival was 100% in all fish groups. The final average body weight of fish fed-FM Table 3 : Productive parameters and somatic indeces of fish fed-FM and fish fed-PBM at 50% and 100% of fish meal replacement.

2.11. Fatty acid analysis Fillet lipid extraction was performed using gas chromatography (6890 Series GC System, Agilent Technologies) with a capillary column DB - 5MS (60 m × 0.32 mm) and a manual sampler. The GC unit is connected directly to a mass spectrometer detector (MSD) (5973 Network Mass Selective Detector). Pure helium (82 kPa flow) was used for the analysis. The helium had an air flow of 50 kPa and a hydrogen flow of 60 kPa under the following temperature conditions: the initial temperature was 150 °C for 5 min: it was then increased by 5 °C/min up to 170 °C, which was maintained for 10 min. Finally, the temperature was increased at a rate of 5 °C/min up to 220 °C and maintained for 20 min. However, fillet fatty acid analysis was performed by lipid extraction, transesterification and fatty acid separation using GC–MS. Qualitative analysis was determined by mass spectrometry and retention time corresponding to the standard mixture, while quantitative analysis was determined as the percentage of identified fatty acid of the total fatty acids in each sample.

a

SR IBW FBW d SL e SGR f AWG g K h HSI i VSI j FCR b c

CONTROL

FEED A

FEED B

100 73.57 ± 10.47 150.46 ± 22.07 17.65 ± 0.85 0.79 ± 0.05 76.88 ± 6.15 2.75 ± 0.07 2.83 ± 0.28 8.29 ± 1.21 1.23 ± 0.13

100 73.57 ± 10.47 138.71 ± 20.87 17.25 ± 0.82 0.71 ± 0.04 70.24 ± 3.58 2.73 ± 0.07 2.23 ± 0.37 8.16 ± 0.47 1.45 ± 0.18

100 73.57 ± 10.47 145.19 ± 22.20 17.52 ± 0.83 0.76 ± 0.06 71.62 ± 1.91 2.72 ± 0.04 2.79 ± 0.35 7.40 ± 0.79 1.32 ± 0.15

All values are given as the mean ± standard deviation (SD). a SR: Survival Rate (%). b SL: Standard Length (cm). c IBM: Initial Body Weight. d FBW: Final Body Weight. e Specific growth rate (SGR) = 100* (ln final weight – ln initial weight)/time (days), (n = 24); f Average weight gain (AWG) = final weight (g) – initial weight (g), (n = 24); g Condition factor (K) = [weight of the fish (g) / standard length of fish3 (cm)] *100, (n = 24); h Hepatosomatic index (HSI) = (liver weight/body weight) *100, (n = 15); i Viscerosomatic index (VSI) = (visceral weight/body weight) *100 (n = 15); j Feed Conversion Ratio (FCR) = feed intake (g) * wet weight gain−1 (g);

2.12. Total lipid, cholesterol, phospholipid and gross energy content in muscle Fillet total lipid content was determined using the commercial Cell Biolabs lipid quantification kit. Three fish per tank and nine per 4

Aquaculture 511 (2019) 734220

M. Sabbagh, et al.

Table 4 : Protein concentration (n = 9), osmolality (n = 9), cortisol level (n = 5), ALT (n = 9) and AST (n = 9) in gilthead seabream plasma at the end of growth trial.

Protein concentration (mg/ml) Osmolality (mmol/kg) Cortisol (pg/ml) AST (mU/ml) ALT (mU/ml)

Control

Feed A

Feed B

10.99 ± 0.99

10.59 ± 1.71

10.78 ± 0.93

615 ± 18.67 918.86 ± 329 24.66 ± 1.65 9.56 ± 1.23

601.4 ± 8.99 582.44 ± 243 24.20 ± 1.42 10.01 ± 0.97

618.4 ± 42.00 1137.97 ± 330 23.96 ± 1.09 9.45 ± 1.72

All values are given as the mean ± standard deviation (SD).

resulted higher than fish fed-PBM but no statistical difference was found. Furthermore, no statistical difference was found on final fish length among all fish groups. Condition factor and somatic indeces (HSI and VSI) were unaffected by PBM diet and were similar to those found in fish fed-FM at the end of feeding trial. 3.2. Fish welfare 3.2.1. Blood biochemical assays Cortisol, protein levels, osmolality, ALT and AST activities in fish plasma (Table 4) were not significantly (P > .05) affected by the partial (50%) and the total (100%) replacement of FM by PBM. 3.2.2. Digestive enzyme activities No statistically significant difference was found among experimental fish groups with regard to liver alkaline phosphatase. The intestinal enzymatic activities (alkaline phosphatase, lipase and leucineamino peptidase) at the end of the trial showed similar values in all experimental groups (Table 5). 3.2.3. Liver histological analysis At the beginning of the trial, fish showed a healthy liver characterized by normal irregular shaped hepatocytes with very prominent circular nuclei and conspicuous centrally located circular nuclei. At the end of the trial, most of the livers examined from all groups showed healthy hepatocytes (Fig. 1). However, fish fed control diet showed 93% healthy livers (only 7% at the intermediate grade). This percentage was 91% in fish fed A and to 90% in fish fed B. No liver was classified as grade 3 during the growth trial. 3.3. Fillet quality 3.3.1. Fillet protein profile: proteomic analysis Image analysis using ImageMaster™ 2D Platinum in fish muscle revealed 700–800 spots. No statistical difference was found on total spot numbers between fish groups at the end of the trial (743 ± 22, 751 ± 14 and 756 ± 18 in the control, Feed A and Feed B groups respectively). Only 12 spots were selected for investigation during the experimental trial (Fig. 2). The spots named 2, 4, 5, 6, 10, 11 and 12 showed changes in the normalized volume (Fig. 2). The molecular weight (MW) and isoelectric point (pI) of the spots selected for analysis are summarized in Table 6. The results (Fig. 3) demonstrated that the normalized volume of

Fig. 1. : Liver microscopic structure at the end of growth trial. (C) Control diet, (A) Feed A, (B) Feed B. Normal hepatocytes with circular conspicuous and centrally located nuclei (arrows) (H and E staining).

Table 5 : Liver and intestine enzyme specific activities of gilthead seabream fed control and experiment diets (n = 9) at the end of growth trial. Enzyme activities (mU/mg)

Control

Feed A

Feed B

Liver alkaline phosphatase Intestine alkaline phosphatase Lipase Leucine-amino peptidase

89.24 ± 18.41 71.95 ± 20.00 1.76 ± 0.68 21.16 ± 3.21

89.66 ± 18.06 110 ± 22.67 2.52 ± 0.97 27.65 ± 6.00

88.02 ± 19.72 100 ± 19.00 2.35 ± 0.57 18.54 ± 4.92

All values are given as the mean ± standard deviation (SD). 5

Aquaculture 511 (2019) 734220

M. Sabbagh, et al.

Table 7 : Amino acid profile in fish fillet at the end of growth trial (data are expressed as g 100 g−1 wet weight, mean ± SD, n = 3).

Arginine Histidine Isoleucine Leucine Lysine Methionine Phenylalanine Threonine Valine ∑ EAA Alanine Aspartate Cystine Glutamine Glysine Proline Serine Tyrosine ∑ NEAA EAA/NEAA

Fig. 2. : 2-D Page of gilthead seabream muscle proteins selected for investigation. Table 6 : Molecular weight (MW) and isoelectric point (pI) of selected spots. Spot number

MW kDa

pI

1 2 3 4 5 6 7 8 9 10 11 12

43 38 33 32 26 26 21 19 19 18.5 18 17.5

5.1 5.4 4.6 5.3 5.4 5.2 4.6 3.5 4.7 4.8 6.1 6.6

Control

Feed A

Feed B

3.31 ± 0.45 0.81 ± 0.11 0.61 ± 0.07 1.39 ± 0.21 0.90 ± 0.07 0.78 ± 0.14 0.71 ± 0.06 1.75 ± 0.10a 1.15 ± 0.21 11.40 ± 1.32 1.67 ± 0.21 2.81 ± 0.40 0.18 ± 0.04 3.69 ± 0.53 0.33 ± 0.04 1.00 ± 0.08 1.14 ± 0.16 0.98 ± 0.06a 12.01 ± 2.34 0.95 ± 0.07

3.38 ± 0.17 0.79 ± 0.08 0.66 ± 0.05 1.35 ± 0.11 0.84 ± 0.13 0.70 ± 0.10 0.70 ± 0.04 1.46 ± 0.15b 0.97 ± 0.09 10.85 ± 1.10 1.49 ± 0.13 2.31 ± 0.12 0.14 ± 0.06 3.20 ± 0.32 0.30 ± 0.04 0.87 ± 0.18 1.00 ± 0.08 0.72 ± 0.15b 10.04 ± 1.07 1.08 ± 1.01

3.45 ± 0.21 0.77 ± 0.10 0.61 ± 0.03 1.41 ± 0.16 0.83 ± 0.09 0.68 ± 0.17 0.68 ± 0.08 1.46 ± 0.12b 0.93 ± 0.10 10.80 ± 1.52 1.45 ± 0.11 2.18 ± 0.76 0.16 ± 0.03 3.10 ± 0.82 0.30 ± 0.06 0.88 ± 0.18 0.95 ± 0.09 0.71 ± 0.08b 9.74 ± 1.96 1.11 ± 0.09

Means in the same row followed by different letters are significantly different P < .05.

decrease in normalized volume after 30 days of trial (Fig. 3) but, at the end of the trial, no significant difference (among fish groups) was observed between them and with regard to the control (Fig. 3). In addition, our results demonstrated that the normalized volume of spot 10 increased in the muscle of fish fed B after 30 days but at the end of the trial it showed a normalized volume similar to the control. The normalized volume of spots 1, 3, 7, 8 and 9 did not show any change during all the trials. Based on previous studies (Addis et al., 2010; Schiavone et al., 2008), these spots were identified as follows: spot no. 1 − actin, alpha skeletal muscle; spot no. 3 − tropomyosin 1, alpha chain; spot no. 7 − myosin light chain 1 (MLC 1); spot no. 8 − myosin light chain 2 (MLC 2); and spot no. 9 − fast skeletal myosin light chain 3 (MLC 3). 3.3.2. Muscle amino acid profile At the end of the trial, the results showed that the amino acid profile in the muscle of seabream was similar in all dietary groups, with the exception of tyrosine (NEAA) and threonine (EAA) (Table 7). 3.3.3. Muscle total lipid, total phospholipid, total cholesterol and gross energy content Lipid concentration increased at the end of the growth trial in all the experimental groups and no statistical difference was detected in fillet fat content (Table 8). Analysis of the amount of phospholipids did not reveal any differences between the experimental groups. At the end of the growth trial no statistical differences were found on total cholesterol content among fish groups (Table 8). Moreover, free cholesterol and cholesterol esters measured in the muscle of all fish

Fig. 3. : Spots with increased normalized volume during growth trial. The upper part of the figure presented the difference of normalized volume of spot No. 2; spot No.4; spot No. 5 and spot No. 6. The dotted line square presented the normalized volume of spots No. 11 and No. 12 at 30 days of trial in the muscle of fish fed control and PBM-diets and their normalized volume at the end of the growth trial. (C: Control group, A: Fish fed A; B: Fish fed B).

Table 8 : Total lipid (n = 3); total phospholipids (n = 5); total cholesterol (n = 3); and lipid peroxidation (n = 5) in fish muscle.

Total lipid (mg/dl) Total phospholipids (μM) Total cholesterol (pg/μl) Lipid peroxidation MDA(nmole/ μl) Gross energy (Kcal/100 g)

spots 2, 4, 5 and 6 reached a maximum increase at 90 days both in fish fed A and B. Visual inspection of gels also suggests that the spot expression increased with the increase of PBM content in diets. A different pattern of volume variation was observed for spots 11 and 12. With regard to the control group, these spots in fish fed A and B showed a

Control

Feed A

Feed B

0.38 ± 0.07 437 ± 25 3.85 ± 1.2 0.28 ± 0.04

0.48 ± 0.08 425 ± 36 4.18 ± 1.2 0.30 ± 0.03

0.53 ± 0.15 428 ± 18 3.14 ± 1.0 0.32 ± 0.01

145 ± 8

148 ± 8

149 ± 10

All values are given as the mean ± SD. 6

Aquaculture 511 (2019) 734220

M. Sabbagh, et al.

ingredient (proximate composition, amino acids and fatty acid content) and on its effect on fish growth (Ayadi et al., 2012; Barreto-Curiel et al., 2016; Gümüş and Aydin, 2013). There is a lack of deep information available on the effect of PBM on the physiological parameters (such as enzyme activities), or on fillet quality such as protein patterns and lipid peroxidation in gilthead seabream (S. aurata). The present study showed that the total replacement of FM by PBM did not result in any negative effect on fish growth parameters and somatic indexes. These findings confirm those of Alexis (1997), which demonstrated that the PBM inclusion of 75% and 100% in the diet of S. aurata fingerlings gave productive results comparable to that of fish fed by FM. The present results are also in agreement with a previous study, carried out on seabream by Nengas et al. (1999), and to those obtained in other marine fish species fed PBM such as red seabream (Takagi et al., 2000), sunshine bass (Rawles et al., 2011) and grouper (Shapawi et al., 2007). The recent study of Karapanagiotidis et al. (2019) reported that both 50% and 100% replacement of FM by PBM resulted in reduced growth performance in juveniles of gilthead seabream, due to a lower feed intake as well as low levels of dietary methionine and lysine. In fact, it has been demonstrated by several studies that PBM is characterized by some amino acids deficiency, especially in methionine and lysine that may affect fish growth (González-Rodríguez et al., 2016; Yigit et al., 2006; Zapata et al., 2016). Our experimental diets were formulated to have a balanced amino acids profile that meets the nutritional needs of S. aurata, which explain the positive results in agreement with other studies (Hernández et al., 2010, 2014; Zhou et al., 2011). As in other feed ingredients obtained by rendered animal products, PBM composition is variable and depends on the processing methods and quality of source by-products (Johnston, 1999; Shapawi et al., 2007). In fact, other studies on PBM- fed S. aurata suggested that the total replacement of FM could be realized when using a high-quality poultry meal (Alexis, 1997; Nengas et al., 1999; Karapanagiotidis et al., 2019). The effectiveness of PBM as a substitute of FM was further supported by the observation that all the productive parameters measured in this trial were similar within all the experimental groups. Several studies have demonstrated that FM replacement with PBM did not affect either SGR or AWG in various species such as: juvenile black seabass (up to 60% PBM; Sullivan, 2008), tilapia (100% PBM, Yones and Metwalli, 2015) and cobia (up to 60% PBM, Zhou et al., 2011). On the contrary, these parameters were affected by high levels PBM inclusion in the diet of golden pompano (up to 100% PBM, Ma et al., 2014), humpback grouper (100% PBM, Shapawi et al., 2007), juvenile spotted rose snapper (75% PBM, Hernández et al., 2014) and gilthead seabream (50% and 100% PBM, Karapanagiotidis et al., 2019). In the present study, the obtained condition factor (K) values > 1.4, indicates a good to excellent condition (Pésic et al., 2015). HSI and VSI were also unaffected by PBM-diets. The above results were similar to those found in other fish species (Ma et al., 2014; Rawles et al., 2011; Shapawi et al., 2007).

Table 9 : Fatty acid composition of fish muscle fed control and two experimental diets (Feed A and Feed B) at the end of the trial (mean ± SD; n = 3).

C14:0 C16:1 C16:0 C18:2n-6 C18:1n-9 C18:0 C20:5n3 (EPA) C20:1n9 C22:6n3 (DHA) SFA MUFA PUFA n-3 n-6 n-9 EPA/DHA n-3/n-6

Control

Feed A

Feed B

2.56 ± 0.40 4.27 ± 0.67 16.78 ± 0.48 18.93 ± 0.18a 33.26 ± 1.23a 4.27 ± 0.27 9.22 ± 1.86 3.39 ± 0.80 8.01 ± 1.47a 23.61 ± 1.05 40.92 ± 0.73a 36.16 ± 0.90a 17.23 ± 0.08a 21.67 ± 1.08 36.65 ± 0.28a 1.15 ± 0.29 0.80 ± 0.35

2.73 ± 0.09 5.06 ± 0.33 16.24 ± 0.59 19.95 ± 0.30b 38.36 ± 0.54b 4.24 ± 0.16 6.32 ± 1.23 2.37 ± 0.35 4.73 ± 1.11b 23.21 ± 0.92 45.79 ± 0.70b 31.00 ± 0.57b 11.05 ± 0.11b 19.95 ± 0.30 40.73 ± 0.41b 1.34 ± 0.18 0.55 ± 0.09

2.32 ± 0.19 4.17 ± 0.49 16.74 ± 0.75 20.44 ± 0.39b 36.96 ± 1.89b 4.46 ± 0.40 6.36 ± 1.07 2.82 ± 0.12 5.72 ± 1.23ab 23.52 ± 0.57 43.96 ± 0.30b 32.52 ± 0.62b 12.08 ± 0.22b 20.44 ± 0.39 39.79 ± 0.16b 1.11 ± 0.22 0.59 ± 0.07

Data are expressed as % of individual fatty acid of total fatty acids in samples. All values are given as the mean ± standard deviation (SD). Means in the same row followed by different letters are significantly different P < .05.

groups showed no statistical difference (data not reported). Lipid peroxidation was investigated in order to evaluate the possible effects of FM substitution with PBM on fillet quality and resulted similar in all fish groups (Table 8) at the end of the trials. The gross energy content of fish muscle was similar in all fish groups and no statistical difference (P > .05) was detected. 3.3.4. Fillet fatty acid profile The most abundant fatty acid in gilthead seabream fillet was oleic acid (C18:1n-9). It was higher (P < .05) in the muscle of fish fed PBMdiets than that found in the muscle of the control fish group (Table 9). As for oleic acid, the linoleic acid (C18:2n-6) content in the fish muscle of the experimental groups was higher than that found in the control fish muscle (Table 9). At the end, the percentage of palmitic acid C16:0 was about 16% of total fatty acids in the muscle of all fish groups (P > .05). The content of the remaining fatty acids identified (C14:0; C16:1; C18:0; C20:1n-9; C20:5n-3; and C22:6n-3) was < 10% of total fatty acids in fish fillet. Eicosapentaenoic acid (EPA, C20:5n-3) and docosahexaenoic acid (DHA, C22:6n-3) levels decreased in the muscle of fish fed the Feed A and Feed B in comparison with the control group at 90 days of trial, but these were significantly (P < .05) lower only in the case of DHA (Table 9). The total amount of monounsaturated fatty acids (MUFA) was higher (P < .05) in the muscle of seabream fed with PBM compared with those fed the control diet, while the opposite occurs for the quantity of polyunsaturated fatty acids (PUFA), (Table 9). The n-3/n-6 ratio at the end of the growth trial was lower in fish fedPBM than in fish fed-FM but no statistical difference was found. With regard to the EPA/DHA ratio, no significant difference (P > .05) was detected among treatments.

4.2. Fish welfare Researches evaluating the effect of PBM on gilthead seabream welfare are limited. The present results clearly suggest that a PBM mealbased diet did not cause, per se, any alteration in welfare conditions in this fish species. Our study demonstrated that partial (50%) and total (100%) FM replacement by PBM did not affect negatively haematological parameters indicating a good fish health. Plasma cortisol concentrations as well as plasma protein levels were similar in all fish groups and a comparable results were found in tench fed up 100% PBM (Panicz et al., 2017), tilapia fed up 100% PBM (Yones and Metwalli, 2015) and cobia juveniles fed 60% PBM (Zhou et al., 2011). Alanine aminotransferase (ALT) and aspartate aminotransferase (AST) are produced in the liver and in the spleen. Their plasma levels are low in healthy specimens, but increase when fish are stressed or

4. Discussion 4.1. Growth performance Poultry by-product meal (PBM) is considered to be a suitable alternative to fish meal (FM), particularly for its specific nutritional characteristics comparable with those of FM. PBM has been investigated in the diet of several aquaculture fish species. Some researchers focused their attention on the nutritional value of this 7

Aquaculture 511 (2019) 734220

M. Sabbagh, et al.

(small but significant) change of the normalized volume of only 7 over 800 total spots, thereby suggesting a minimal effect of PBM on protein muscle composition.

when tissue necrosis occurs (Jung et al., 2003; Wells et al., 1986), indicating liver damage. In the present study ALT and AST were not affected by the high levels PBM inclusion in the diet of S. aurata. Similar results were also found in tilapia (Yones and Metwalli, 2015), Atlantic salmon (Liland et al., 2015) and tench (Panicz et al., 2017) fed with PBM. Different haematological and blood biochemical parameters were investigated by Karapanagiotidis et al. (2019) and reported, accordingly to our study, that fish health was unaffected by PBM inclusion in juveniles gilthead seabream. To date, little information exists regarding the effect of PBM on digestive enzymes in gilthead seabream or in other fish species. Research on the digestibility of PBM suggested that it is well digested by several fish species (Yang et al., 2006). In the present study, changes in enzyme activities (hepatic and intestinal AP, intestinal lipase and leucine aminopeptidase) showed same pattern for all fish groups throughout the experimental period, thus confirming a PBM high digestibility by gilthead seabream, as demonstrated by other studies (Alexis, 1997; Davies et al., 2009; Karapanagiotidis et al., 2019; Nengas et al., 1999). Knowledge on PBM effects on fish morphology and their possible implications for fish health is limited. Histology has been extensively used to describe tissue changes and possible sub-pathological alterations in the liver induced by substitution of fish meal in artificial feeding of farmed species (Caballero et al., 2003; Ostaszewska et al., 2005). In the present study, the replacement of FM by PBM did not produce steatosis (a massive synthesis and deposition of triglycerides in the vacuoles) and did not cause any liver alteration. However, these results demonstrated that the amounts of dietary lipid in PBM-diets did not exceed the capacity of hepatic cells to oxidise fatty acids. To the best of our knowledge, no information exists about liver histological analysis in gilthead seabream fed with PBM. In the study of Aydín et al. (2015), PBM inclusion in the diet of Nile tilapia (Oreochromis niloticus) up to 100% did not reveal any effect on the histological examination of the liver tissue and livers showed, as in our results, normal-shaped hepatocytes. Similar results were found using different animal by products. Jasim et al. (2016), demonstrated that 75% of fish biosilage inclusion in the diet of common carp did not result in any adverse alterations or abnormalities in fish liver structure while Belghit et al. (2019), declared that insect meal could replace totally FM in the diet of Atlantic salmon (Salmo salar) without compromising effects on liver lipid accumulation. Contrarily, Hu et al. (2013) demonstrated a negative effect of FM replacement by animal protein blend (a mixture of 40% poultry by-product meal, 35% meat and bone meal, 20% spraydried blood meal and 5% hydrolyzed feather meal) in japanese seabass (Lateolabrax japonicus) livers that were characterized by enlarged hepatocytes and apparent hepatic steatosis with intense vacuoles in the hepatocytes resemble lipids.

4.3.2. Fillet amino acid content The high protein level (450–650 g/kg) and the amino acid profile of PBM (except methionine and lysine) are similar to FM (GonzálezRodríguez et al., 2016; NRC, 2011; Zhou et al., 2011). In this study, PBM inclusion did not negatively affect the fillet composition of gilthead seabream in terms of amino acid profile. Similar results were found in the fillets of juvenile spotted rose snapper (Lutjanus guttatus) fed with PBM diets, up to a 75% of FM substitution (Hernández et al., 2014). Furthermore, Karapanagiotidis et al. (2019), demonstrated that whole-body and muscle protein contents in juvenile gilthead seabream were unaffected by PBM-diet. 4.3.3. Fillet lipid and energy content Lipid composition in the edible part of fish fillet influences the taste and appearance of cooked flesh for the consumer (Grigorakis, 2007). In the present study, the inclusion of PBM in the diet of gilthead seabream did not alter fillet lipids and energy content even at high level of FM replacement. These results may be explained by the similar dietary lipid levels in the three experimental diets, as well as by their similar energy content. Karapanagiotidis et al. (2019), and Nengas et al. (1999), reported that high level of FM replacement by PBM, reduced muscle lipid and energy content in S. aurata, due to the lowest dietary lipid levels compared to FM-diet. On the other hand, it has been demonstrated that high levels of FM substitution by PBM resulted in higher lipid content in Nile tilapia (El-Sayed, 1998), rainbow trout (Baboli et al., 2013) and tench (González-Rodríguez et al., 2016). 4.3.4. Fillet fatty acid profile In the present study, considerable differences were observed in the muscle fatty acid profiles of fish fed PBM diets when compared with fish fed FM-diet. ΣMUFA increased significantly with higher dietary PBM. This is mainly due to the abundance of C18:1n-9 in fish muscle, which is a consequence of the high oleic acid content of PBM (Sullivan, 2008). However, high PBM levels in gilthead seabream diets resulted in low ΣPUFA levels. This was because of the decrease in eicosapentaenoic acid (EPA, C20:5n-3) and docosahexaenoic acid (DHA, C22:6n-3) in the muscle of fish fed PBM diets. PBM doesn't contain EPA and DHA (Bureau et al., 2000), and this fact explains the lower values of EPA and DHA found in the fillet of fish fed the PBM-based diet. No data were found on the fillet fatty acid profile of gilthead seabream fed with PBM; similar results were found in the muscle of carp fry (Gümüş and Aydin, 2013) and in juvenile black seabass (Sullivan, 2008) fed with PBM. Moreover, our results show that Σn-3 fatty acid decreased in the muscle of fish fed Feed A and Feed B. Consequently, the Σn-3/Σn-6 ratio decreased and this may be related to the low content of Σn-3 fatty acid in PBM (Bureau et al., 2000). These findings demonstrated that PBM inclusion in the diet of S. aurata notably reduced PUFA levels in fish fillet, which may consequently affect the nutritional value in terms of fatty acids available for human consumption.

4.3. Fillet quality 4.3.1. Fillet protein pattern In order to obtain systematic characterization of the gilthead seabream muscle proteome and investigate any modification caused by PBM inclusion in diets, a 2-DE study was carried out to monitor fillet quality. There is no available information in the literature on the protein pattern in seabream fillet fed with PBM. Concerning fillet quality and shelf-life, preliminary data on postmortem changes in seabream muscle were obtained by Schiavone et al. (2008). Muscle protein maps of gilthead seabream fed either with the control diet, the Feed A and the Feed B, displayed the same protein spots and resembled the farmed seabream maps investigated by Addis et al. (2010). The protein patterns obtained in all fish groups showed a similar expression of structural proteins such as actin (the predominant protein), tropomyosin, MLC1, MLC2 and MLC3, and the results were comparable to those reported previously (Addis et al., 2010; Schiavone et al., 2008). The inclusion of PBM in the seabream diet determined a

4.3.5. Fillet cholesterol content Fish are universally considered to have a low level of cholesterol (Harlioğlu et al., 2016). Cholesterol, as well as lipid content, depends on many factors including fish diets (Izquierdo et al., 2003). The present study suggests the absence of any negative effect related to PBM based-diets on fillet S. aurata cholesterol content. Moreover, this result may be attributed to a well balanced diet formulation, supplied at an appropriate feeding ration. 4.3.6. Fillet lipid oxidation Lipid peroxidation is of great commercial importance (because of its 8

Aquaculture 511 (2019) 734220

M. Sabbagh, et al.

negative impact on flavour, colour and nutritional characteristics) (Monahan, 2000), given the concerns of consumers regarding farmed products. The TBARS method has been utilized in some studies on fish flesh quality (Menoyo et al., 2002; Sicuro et al., 2010). In the present study PBM inclusion did not induce lipids peroxidation. In literature it was demonstrated that the greatest peroxidation levels are closely related to high PUFA concentrations in tissues (López-Bote et al., 2001; Menoyo et al., 2002). In the present study, fillet PUFA contents were lower in fish fed-PBM than that fed-FM, which may explain the absence of fillet lipid peroxidation and, consequently, be considered as a valuable aspect for fillet quality.

hydroxysccinimidyl carbamate (AQC) reagent to the RP-HPLC determination of amino acids in infant foods. J. Chromatogr. B Anal. Technol. Biomed. Life Sci. 831, 176–183. Bradford, M.M., 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248–254. Bureau, D.P., Harris, A.M., Cho, C.Y., 1999. Apparent digestibility of rendered animal protein ingredients for rainbow trout (Oncorhynchus mykiss). Aquaculture 180, 345–358. Bureau, D.P., Harris, A.M., Bevan, D.J., Simmons, L.A., Azevedo, P.A., Cho, C.Y., 2000. Feather meals and meat and bone meals from different origins as protein sources in rainbow trout (Oncorhynchus mykiss) diets. Aquaculture 181, 281–291. Caballero, M.J., Izquierdo, M.S., Kjørsvik, E., Montero, D., Socorro, J., Fernández, A.J., Rosenlund, G., 2003. Morphological aspects of intestinal cells from gilthead sea bream (Sparus aurata) fed diets containing different lipid sources. Aquaculture 225, 325–340. Davies, S.J., Gouveia, A., Laporte, J., Woodgate, S.L., Nates, S., 2009. Nutrient digestibility profile of premium (category III grade) animal protein by-products for temperate marine fish species (European sea bass, gilthead sea bream and turbot). Aquac. Res. 40, 1759–1769. De Pedro, N., Guijarro, A.E., Lopez-Patino, M.A., Marinez-Alvarez, R., Delgado, M., 2005. Daily and seasonal variation in haematological and blood biochemical parameters in tench Tinca tinca. Aquac. Res. 36, 85–96. El-Sayed, A.F.M., 1998. Total replacement of fish meal with animal protein sources in Nile tilapia, Oreochromis niloticus (L), feeds. Aquac. Res. 29, 275–280. FAO, 2017. Fisheries and aquaculture. FishStatPlus – Universal software for fishery and aquaculture statistical time series. In: FAO Fisheries and Aquaculture Department. Statistics and Information Branch (FIAS). FAO, 2018. The state of world fisheries and aquaculture 2018 – Meeting the Sustainable Development Goals. (Rome). Fasakin, E.A., Balogun, A.M., Ajayi, O.O., 2003. Evaluation of full-fat and defatted maggot meals in the feeding of Claiid catfish, Clarias gariepinus fingerlings. Aquac. Res. 34, 733–738. Fasakin, E.A., Serwata, R.D., Davies, S.J., 2005. Comparative utilization of rendered animal derived products with or without composite mixture of soybean meal in hybrid tilapia (Oreochromis niloticus×mossambicus) diets. Aquaculture 249, 329–338. Francis, G., Makkar, H.P.S., Becker, K., 2001. Anti-nutritional factors present in plantderived alternate fish feed ingredients and their effects in fish. Aquaculture 199, 197–227. García-Gallego, M., Akharbach, H., de la Higuera, M., 1998. Use of protein sources alternative to fish meal in diets with amino acids supplementation for the European eel (Anguilla anguilla). Anim. Sci. 66, 285–292. Gatlin, D.M., Barrows, F.T., Brown, P., Dabrowski, K., Gaylord, T.G., Hardy, R.W., Herman, E., Hu, G.S., Krogdahl, A., Nelson, R., Overturf, K., Rust, M., Sealey, W., Skonberg, D., Souza, E.J., Stone, D., Wilson, R., Wurtele, E., 2007. Expanding the utilization of sustainable plant products in aquafeeds: a review. Aquac. Res. 38, 551–579. Gaylord, T.G., Rawles, S.D., 2005. The modification of poultry by-product meal for use in hybrid striped bass (Morone chrysops×M. saxatilis) diets. J. World Aquacult. Soc. 36, 363–374. Gómez-Requeni, P., Mingarro, M., Calduch-Giner, J.A., Médale, F., Martin, S.A.M., Houlihan, D.F., Kaushik, S.J., Pérez-Sánchez, J., 2004. Protein growth performance, amino acid utilisation and somatotropic axis responsiveness to fish meal replacement by plant protein sources in gilthead seabream (Sparus aurata). Aquaculture 232, 493–510. González-Rodríguez, A., Celada, J.D., Carral, J.M., Sáez-Royuela, M., García, V., Fuertes, J.B., 2016. Evaluation of poultry by-product meal as partial replacement of fish meal in practical diets for juvenile tench (Tinca tinca L.). Aquac. Res. 47, 1612–1621. Grigorakis, K., 2007. Compositional and organoleptic quality of farmed and wild gilthead sea bream (Sparus aurata) and sea bass (Dicentrarchus labrax) and factors affecting it: a review. Aquaculture 272, 55–75. Gümüş, E., Aydin, B., 2013. Effect of poultry by-product meal on growth performance and fatty acid composition of carp (Cyprinus carpio) fry. Turk. J. Fish. Aquat. Sci. 13, 827–834. Hardy, R.W., 2000. Diet preparation. In: Halver, J.E. (Ed.), Fish Nutrition, 2nd ed. 151. Academic Press, New York, NY, pp. 475–548. Hardy, R.W., 2006. Worldwide fish meal production outlook and the use of alternative protein meals for aquaculture. Avances en Nutrición Acuícola VIII. 410–419 ISBM 970-694-333-5. Hardy, R.W., 2010. Utilization of plant proteins in fish diets: effects of global demand and supplies of fishmeal. In: Journal Compilation. Blackwell Publishing Ltd.. Hardy, R.W., Tacon, A.G.J., 2002. Fish meal historical uses, production trends and future outlook for sustainable supplies. In: Stickney, R.R. (Ed.), Sustainable Aquaculture New York. CABI. Harlioğlu, A.G., Yilmaz, Ö., Oray, I.K., Aydin, S., 2016. A comparison of fatty acid, cholesterol and vitamin composition in seabass and seabream from three cage farm areas: Antalya and Muğla (Turkey) and Ïskele (Northern Cyprus). J. Appl. Ichthyol. 32 (3), 577–582. Hernández, M.D., Martínez, F.J., Jover, M., García, B.G., 2007. Effects of partial replacement of fish meal by soybean meal in sharpsnout seabream (Diplodus puntazzo) diet. Aquaculture 263, 159–167. Hernández, C., Olvera-Novoa, M.A., Hardy, R.W., Hermosillo, A., Reyes, C., González, B., 2010. Complete replacement of fishmeal by porcine and poultry by product meals in practical diets for fingerling Nile tilapia Oreochromis niloticus: digestibility and growth performance. Aquac. Nutr. 16 (1), 44–53. Hernández, C., Osuna-Osuna, L., Hernández, A.B., Sánchez-Gutiérrez, Y., GonzálezRodríguez, B., Domínguez-Jiménez, P., 2014. Replacement of fish meal by poultry byproduct meal, food grade, in diets for juvenile spotted rose snapper (Lutjanus guttatus). Lat. Am. J. Aquat. Res. 42 (1), 111–120. Hu, L., Yun, B., Xue, M., Wang, J., Wu, X., Zheng, Y., Han, F., 2013. Effects of fish meal

5. Conclusions In conclusion, this multidisciplinary study demonstrated the possibility of partial (50%) and total (100%) FM replacement by PBM in the diet of gilthead seabream, without any adverse effects on growth performance and fish welfare. Amino acids profile and protein pattern in fillet were not affected by PBM inclusion. Nevertheless, the fillet nutritional value, in term of fatty acids available for human consumption, was impaired by the low PUFA content. Acknowledgements The authors acknowledge Prof. Giuseppe Mele and Dr. Maria Rosaria Lazzoi of the Laboratory of Organic and Inorganic Chemistry, Department of Innovation Engineering of the University of Salento for their help and collaboration in fatty acid analysis. The authors also gratefully thank the Veronesi Group and their technical team especially Dr. Pierluigi Perantoni and Dr. Paolo Chignola for providing fish feed and Panittica Pugliese for their collaboration, in particular Dr. Massimo Caggiano. This research has received no specific grant from any funding agencies in the public, commercial or non-profit sectors. References Addis, M.F., 2013. Proteomics in Foods. In: Proteomics Foods Princ Appl. Springer US, pp. 181–203. Addis, M.F., Cappuccinelli, R., Tedde, V., Pagnozzi, D., Porcu, M.C., Bonaglini, E., Roggio, T., Uzzau, S., 2010. Proteomic analysis of muscle tissue from gilthead sea bream (Sparus aurata, L.) farmed in offshore floating cages. Aquaculture 309, 245–252. Albro, P.W., Hall, R.D., Corbett, J.T., Schroeder, J., 1985. Activation of nonspecific lipase (EC 3.1.1.) by bile salts. Biochim. Biophys. Acta 835, 477–490. Alexis, M.N., 1997. Fish meal and fish oil replacers in Mediterranean marine fish diets. In: Tacon, A.G.J., Basurco, B. (Eds.), Feeding tomorrow's fish. Zaragoza, CIHEAM, pp. 183–204. Auchterlonie, N., 2017. Fish meal, fish oil and “the need for feed”. International Aquafeed 20 (10), 18–21. Ayadi, F.Y., Rosentrater, K.A., Muthukumarappan, K., 2012. Alternative protein sources for Aquaculture feeds. Aquaculture Feeds science and Nutrition. 4 (1), 1–26. Aydín, B., Gümüş, E., Balcí, B.A., 2015. Effect of dietary fish meal replacement by poultry by-product meal on muscle fatty acid composition and liver histology of fry of Nile tilapia, Oreochromis niloticus (Actinopterygii: Perciformes: Cichlidae). Acta Ichthyol. Piscat. 45 (4), 343–351. Baboli, M.J., Dawodi, M., Gorjipor, A., 2013. Effect of replacement fish meal by poultry meal on growth, survival and body composition of rainbow trout. Int. Res. J. Appl. Basic Sci. 5 (3), 296–300. Baghaei Jezeh, S., Baboli, M.J., Askary Sary, A., 2014. Influence of supplemental barley and formulated dry feed feeding on the growth, survival and fillet chemical composition of farmed common carp (Cyprinus carpio L.). Int. J. Biosci. 5, 137–143. Barreto-Curiel, F., Parés-Sierra, G., Correa-Reyes, G., Durazo-Beltán, E., Viana, M.T., 2016. Total and partial fishmeal substitution by poultry by-product meal (petfood grade) and enrichment with acid fish silage in aquafeeds for juveniles of rainbow trout Oncorhynchus mykiss. Lat. Am. J. Aquat. Res. 44 (2), 327–335. Belghit, I., Liland, N.S., Gjesdal, P., Biancarosa, I., Menchetti, E., Li, Y., Waagbø, R., Krogdahl, A., Lock, Erik-Jan, 2019. Black soldier fly larvae meal can replace fish meal in diets of sea-water phase Atlantic salmon (Salmo salar). Aquaculture. 503, 609–619. Benedito-Palos, L., Navarro, J.C., Sitjà-Bobadilla, A., Bell, J.G., Kaushik, S., PérezSánchez, J., 2008. High levels of vegetable oils in plant protein-rich diets fed to gilthead sea bream (Sparus aurata L.): growth performance, muscle fatty acid profiles and histological alterations of target tissues. Br. J. Nutr. 100, 992–1003. Bessey, O.A., Lowry, O.H., Brock, M.J., 1946. A method for the rapid determination of alkaline phosphatase with five cubic millimeters of serum. J. Biol. Chem. 164, 321–329. Bosch, L., Alegría, A., Farré, R., 2006. Application of the 6-aminoquinolyl-N-

9

Aquaculture 511 (2019) 734220

M. Sabbagh, et al.

Ostrand, S., Webster, C.D., 2011. Effects of replacing fish meal with poultry byproduct meal and soybean meal and reduced protein level on the performance and immune status of pond-grown sunshine bass (Morone chrysops x M. saxatilis). Aquac. Nutr. 17, 708–721. Reverter, M., Lundh, T., Lindberg, J.E., 1997. Determination of free amino acids in pig plasma by precolumn derivatization with 6-N-aminoquinolyl-N- hydroxysuccinimidyl carbamate and high-performance liquid chromatography. J. Chromatogr. B Biomed. Sci. Appl. 696, 1–8. Rodrigues, P.M., Silva, T.S., Dias, J., Jessen, F., 2012. Proteomics in aquaculture: applications and trends. J. Proteome 75 (14), 4325–4345. Roncarati, A., Sirri, F., di Domenico, A., Brambilla, G., Iamiceli, A.L., Melotti, P., Meluzzi, A., 2010. Survey of qualitative traits of European sea bass cultivated in different rearing systems. Eur. J. Lipid Sci. Technol. 112, 770–779. Santigosa, E., Sánchez, J., Médale, F., Kaushik, S., Pérez-Sánchez, J., Gallardo, M.A., 2008. Modifications of digestive enzymes in trout (Oncorhynchus mykiss) and sea bream (Sparus aurata) in response to dietary fish meal replacement by plant protein sources. Aquaculture 282, 68–74. Satheeshkumar, P. Ananthan, Senthil, G., Kumar, D., Jagadeesan, L., 2011. Haematology and biochemical parameters of different feeding behaviour of teleost fishes from Vellar estuary, India. Comp. Clin. Pathol. 21 (6), 1–5. Schiavone, R., Zilli, L., Storelli, C., Vilella, S., 2008. Identification by proteome analysis of muscle proteins in seabream (Sparus aurata). Eur. Food Res. Technol. 227, 1403–1410. Schreiber, M.A., Niquen, M., Bouchon, M., 2011. Coping strategies to deal with environmental variability and extreme climatic events in the Peruvian anchovy fishery. Sustainability 3 (6), 823–846. Shapawi, R., Ng, W.K., Mustafa, S., 2007. Replacement of fish meal with poultry byproduct meal in diets formulated for the humpback grouper, Cromileptes altivelis. Aquaculture 273, 118–126. Shepherd, C.J., Pike, I.H., Barlow, S.M., 2005. Sustainable feed resources of marine origin. European Aquaculture Society Special Publication 35, 59–66. Sicuro, B., Daprá, F., Gai, F., Palmegiano, G.B., Schiavone, R., Zilli, L., Vilella, S., 2010. Olive oil by-product as a natural antioxidant in gilthead seabream (Sparus aurata) nutrition. Aquac. Int. 18, 511–522. Silva, F.C.P., Nicoli, J.R., Zambonino-Infante, J.L., Le Gall, M.M., Kaushik, S., Gatesoupe, F.J., 2010. Influence of partial substitution of dietary fish meal on the activity of digestive enzymes in the intestinal brush border membrane of gilthead seabream, Sparus aurata and goldfish, Carassius auratus. Aquaculture 306, 233–237. Storelli, C., Vilella, S., Cassano, G., 1986. Na-dependent D-glucose and L-alanine transport in eel intestinal brush border membrane vesicles. Am. J. Phys. 251, 463–469. Sullivan, K.B., 2008. Replacement of Fish Meal by Alternative Protein Sources in Diets for Juvenile Black Seabass. University of North Carolina Wilmington (Master thesis). Tacon, A.G.J., Metian, M., 2008. Global overview on the use of fish meal and fish oil in industrially compounded aquafeeds: trends and future prospects. Aquaculture 285, 146–158. Takagi, S.T., Hosokawa, H., Shimeno, S., Ukawa, M., 2000. Utilization of poultry byproduct meal in a diet for red sea bream Pagrus major. Nippon Suisan Gakkaishi 66 (3), 428–438. Thompson, K.R., Metts, L.S., Muzinic, L.A., Dasgupta, S., Webster, C.D., Brady, Y.J., 2007. Use of Turkey meal as a replacement for menhaden fish meal in practical diets for sunshine bass grown in cages. Aquaculture 69, 351–359. Wang, Y., Guo, J., Bureau, D.P., Cui, Z., 2006. Replacement of fish meal by rendered animal protein ingredients in feeds for cuneate drum (Nibea miichthioides). Aquaculture 252, 476–483. Wells, R.M., McIntyre, R.H., Morgan, A.K., Davie, P.S., 1986. Physiological stress responses in big gamefish after capture: observations on plasma chemistry and blood factors. Comparative Biochemistry and Physiology A: Physiology 84, 565–571. Xavier, T.O., Michelato, M., Vidal, L.V.O., Furuya, V.R.B., Furuya, W.M., 2014. Apparent protein and energy digestibility and amino acid availability of commercial meat and bone meal for Nile tilapia, Oreochromis niloticus. J. World Aquacult. Soc. 45, 439–446. Yang, Y., Xie, S.Q., Cui, Y.B., Zhu, X.M., Lei, W., Yang, Y.X., 2006. Partial and total replacement of fishmeal with poultry by-product meal in diets for gibel carp, Carassius auratus gibelio Bloch. Aquac. Res. 37, 40–48. Yigit, M., Erdem, M., Koshio, S., Ergun, S., Turker, A., Karaali, B., 2006. Substituting fish meal with poultry by-product meal in diets for black Sea turbot Psetta maeotica. Aquac. Nutr. 12, 340–347. Yildiz, M., Şener, E., Timur, M., 2008. Effects of differences in diet and seasonal changes on the fatty acid composition in fillets from farmed and wild sea bream (Sparus aurata L.) and sea bass (Dicentrarchus labrax L.). Int. J. Food Sci. Technol. 43, 853–858. Yones, A.M.M., Metwalli, A.A., 2015. Effects of fish meal substitution with poultry byproduct meal on growth performance, nutrients utilization and blood contents of juvenile Nile tilapia (Oreochromis niloticus). J. Aquacul. Res. & Develop. 7, 389. Yu, Y., 2004. Replacement of fish meal with poultry by-product meal and meat and bone meal in shrimp, tilapia and trout diets. In: Cruz Suárez, L.E., Ricque Marie, D., Nieto López, M.G., Villarreal Cavazos, D., Scholz, U., Gonzalez Félix, M., Pérez Velázquez, M. (Eds.), Avances en Nutrición Acuícola VII. Memorias del VII Simposium Internacional de Nutrición Acuícola. 16–19 November 2004. Hermosillo, Sonora, Mexico. Zapata, D.B., Lazo, J.P., Herzka, S.Z., Viana, M.T., 2016. The effect of substituting fishmeal with poultry by-product meal in diets for Totoaba macdonaldi juveniles. Aquac. Res. 47, 1778–1789. Zhou, Q.C., Tan, B.P., Mai, K.S., Liu, Y.J., 2004. Apparent digestibility of selected feed ingredients for juvenile cobia Rachycentron canadum. Aquaculture 241, 441–451. Zhou, Q.C., Zhao, J., Li, P., Wang, H.L., Wang, L.G., 2011. Evaluation of poultry byproduct meal in commercial diets for juvenile cobia (Rachycentron canadum). Aquaculture 322-323, 122–127.

quality and fish meal substitution by animal protein blend on growth performance, flesh quality and liver histology of Japanese seabass (Lateolabrax japonicus). Aquaculture 372-375, 52–61. Izquierdo, M.S., Obach, A., Arantzamendi, L., Montero, D., Robaina, L., Rosenlund, G., 2003. Dietary lipid sources for seabream and seabass: growth performance, tissue composition and flesh quality. Aquac. Nutr. 9, 397–407. Jasim, B.M., Al-Noor, S.S., Najim, S.M., 2016. Effects of using fish biosilage as fish meal replacer on feeding, growth and gut histology in common carp Cyprinus carpio L. fingerlings. Basrah J. of Vet. Res. 14 (1). Johnston, I.A., 1999. Muscle development and growth: potential implications for flesh quality in fish. Aquaculture 177, 99–115. Jung, S.H., Sim, D.S., Park, M.S., Jo, Q., Kim, Y., 2003. Effects of formalin on haematological and blood chemistry in olive flounder, Paralichthys olivaceus (Temminck et Schlegel). Aquac. Res. 34, 1269–1275. Karapanagiotidis, I.T., Psofakis, P., Mente, E., Malandrakis, E., Goomazou, E., 2019. Effect of fishmeal replacement by poultry by-product meal on growth performance, proximate composition, digestive enzyme activity, haematological parameters and gene expression of gilthead seabream (Sparus aurata). Aquac. Nutr. 25, 3–14. Kaushik, S.J., 1990. Use of alternative protein sources for the intensive rearing of carnivorous fish. In: Floss, R., Tort, L., Torres, P. (Eds.), Mediterranean Aquaculture. Ellis Horwood, pp. 125–138. Kaushik, S.J., Seiliez, I., 2010. Protein and amino acid nutrition and metabolism in fish: current knowledge and future needs. Aquac. Res. 41, 322–332. Kureshy, N., Davis, D.A., Arnold, C.R., 2000. Partial replacement of fishmeal with meatand-bone meal, flash-dried poultry by-product meal, and enzyme-digested poultry by-product meal in practical diets for juvenile red drum. N. Am. J. Aquac. 62, 266–272. Liland, N.S., Hatlen, B., Takle, H., Venegas, C., Espe, M., Torstensen, B.E., Waagbø, R., 2015. Including processed poultry and porcine by-products in diets high in plant ingredients reduced liver TAG in Atlantic salmon, Salmo salar L. Aquac. Nutr. 21, 655–669. López-Bote, C.J., Diez, A., Corraze, G., Arzel, J., Alvarez, M., Dias, J., Kaushik, S.J., Bautista, J.M., 2001. Dietary protein source affects the susceptibility to lipid peroxidation of rainbow trout (Oncorhynchus mykiss) and sea bass (Dicentrarchus labrax) muscle. Anim. Sci. 73, 443–449. Lupatsch, I., Kissil, G.W., Sklan, D., Pfeffer, E., 1997. Apparent digestibility coefficients of feed ingredients and their predictability in compound diets for gilthead seabream, Sparus aurata L. Aquac. Nutr. 3, 81–89. Ma, X., Wang, F., Han, H., Wang, Y., Lin, Y., 2014. Replacement of dietary fish meal with poultry by-product meal and soybean meal for golden pompano, Trachinotus ovatus, reared in net pens. J. World Aquacult. Soc. 45 (6), 662–671. Martínez-Llorens, S., Tomas-Vidal, A., Jauralde-Garcia, I., Pla-Torres, M., Jover-Cerdá, M., 2009. Optimum dietary soybean meal level for maximizing growth and nutrient utilization of ongrowing gilthead sea bream (Sparus aurata). Aquac. Nutr. 15, 320–328. McFadzen, I.R.B., Coombs, S.H., Halliday, N.C., 1997. Histological indices of the nutritional condition of sardine, Sardina pilchardus (Walbaum) larvae off the north coast of Spain. J. Exp. Mar. Biol. Ecol. 212, 239–258. Menoyo, D., Lopez-Bote, C.J., Bautista, J.M., Obach, A., 2002. Herring vs anchovy oils in salmon feeding. Aquat. Living Resour. 15, 217–223. Metochis, C.P., Crampton, V.O., Ruohonen, K., El-Mowafi, A., Bell, G.J., Adams, A., Thompson, K.D., 2017. Effects of marine protein-, marine oil-and marine-free diets on the growth performance and innate immune responses of Atlantic salmon (Salmo salar, L.) post-smolts. Aquac. Res. 48, 2495–2515. Millamena, O.M., 2002. Replacement of fish meal by animal by-product meals in a practical diet for grow-out culture of grouper, Epinephelus coioides. Aquaculture 204, 75–84. Monahan, F.J., Faustman, C., 2000. Oxidation of lipids in muscle foods: fundamental and applied concerns. In: Decker, E.A., Lopez-Bote, C.J. (Eds.), Antioxidants in muscle foods: nutritional strategies to improve quality, pp. 3–24. Muzinic, L.A., Thompson, K.R., Metts, L.S., Dasgupta, S., Webster, C.D., 2006. Use of Turkey meal as partial and total replacement of fish meal in practical diets for sunshine bass (Morone chrysops × Morone saxatilis) grown in tanks. Aquaculture 12, 71–81. Nengas, I., Alexis, M., Davies, S.J., 1999. High inclusion levels of poultry meals and related by-products in diets for gilthead seabream Sparus aurata L. Aquaculture 179, 13–23. NRC, 2011. Nutrient requirements of fish and shrimp, pp. 57–101. National Research Council. National Academies Press, Washington, DC. Ostaszewska, T., Dabrowski, K., Czumínska, K., Olech, W., Olejniczak, M., 2005. Rearing of pike-perch larvae using formulated diets − first success with starter feeds. Aquac. Res. 36, 1167–1176. Panicz, R., Żochowska-Kujawska, J., Sadowski, J., Sobczak, M., 2017. Effect of feeding various levels of poultry by-product meal on the blood parameters, fillet composition and structure of female tenches (Tinca tinca). Aquac. Res. 48, 5373–5384. Pavan Kumar, B., Raghu Ramudu, K., Chamundeswari Devi, B., 2014. Mini review on incorporation of cotton seed meal, an alternative to fish meal in aquaculture feeds. Int. J. Biol. Res. 2 (2), 99–105. Pésic, A., Durovic, M., Joksimovic, A., Markovic, O., Ikica, Z., 2015. Condition of Fish, Seabream (Sparus aurata) and Seabass (Dicentrarchus labrax) from Fish Farm in Bokakotorska Bay (Southeast Adriatic). Zbornik Predavanja. Int. Conference on Water and Fish. Rawles, S.D., Riche, M., Gaylord, T.G., Webb, J., Freeman, D.W., Davis, M., 2006. Evaluation of poultry by-product meal in commercial diets for hybrid striped bass, Morone chrysops×M. saxatilis, in recirculated tank production. Aquaculture 259, 377–389. Rawles, S.D., Thompson, K.R., Brady, Y.J., Metts, L.S., Aksoy, M.Y., Gannam, A.L.,

10