Metabolism and growth performance of pacu ( Piaractus mesopotamicus ) juveniles submitted to different feeding frequencies

The aim of this study was to evaluate the effects of different feeding frequencies on productive performance, hepatocytes integrity, frequency of muscle fibers distribution, enzymatic activity, and proximate composition of pacu juveniles (Piaractus mesopotamicus). One hundred and sixty fish were distributed in 20 tanks arranged in a recirculation system. The proposed managements of feed offer were the partition of the daily amounts into one, two, three or four times during 60 days. Feed consumption were higher in fish fed four times a day, whilst the hepatosomatic index was higher in fish fed once, but similar to fish that received food two or four times a day. As for plasmatic glucose levels, fish fed twice a day displayed an increase in this parameter. The livers of fish fed two or four times a day presented larger hepatocytes and cellular nuclei. Regarding the proximate composition of carcasses, fish that received one daily feeding presented higher ether extract. It was concluded that feeding frequency is a factor that promotes changes in P. mesopotamicus nutritional metabolism, thus the offer of feed four times a day could be recommended for this species in this life stage.


Introduction
Providing high-quality diets that meet the nutritional demands of a species is fundamental for the productive success of fish, since it allows a better use of the available nutrients by the animal (NRC, 2011), and also the adoption of adequate feeding frequencies is essential, seen that it relates directly to the animal's metabolism, healthiness and to the deposition of nutrients in its carcass (Bittencourt et al., 2013;Santos et al., 2015). Factors such as species, age, water quality and temperature interfere in the number of daily feeding events (Hayashi et al., 2004).
The determination of feed supply frequency and quantities to be offered are important in relation to feeding management, in order to avoid excessive amounts of feed, which can act in the reduction of both waste generation and production costs, besides maintaining the water quality. On the other hand, feeding frequencies below the optimal rate are also undesirable (Cho et al., 2003). Besides, it composes an important factor within the feeding management, since it stimulates the fish to search for the feed in pre-determined moments, thus contributing for the reduction of feed conversion rates and enabling the possibility of observing the health status of the animals (Carneiro and Mikos, 2005), as well as modifying the fish's proximate composition (Burkert et al., 2008), which may exert an important influence on the quality of breeding water (Cho, 2003).
Providing information about productivity and physiology of fish during food restriction can generate relevant information regarding the applicability of food managements in distinct production systems. In this sense, the dietary frequency in fish rearing might influence the levels of protein, cholesterol and triglycerides in plasmatic metabolism (Dieterich et al., 2013), as well as batch uniformity (Pouey et al., 2012), growth (Santos et al., 2014), and body composition with alterations in content moisture, protein, lipid and ash (Daudpota et al., 2016). Productive performance parameters (e.g. weight gain, specific growth rate, and apparent feed conversion) may also be affected, as reported by Tsuzuki et al. (2014) and Frasca-Scorvo et al. (2017). Thus, studies on feeding dynamics could highlight the implications of the energy resources used by reared fish and provide information about the biochemical pathways involved in feeding, and such a comprehension of these biochemical processes may be useful to satisfy the energetic needs of aquatic organisms.
The pacu (Piaractus mesopotamicus), a migrating fish native from South America has opportunistic omnivorous food habit, being well adapted to production systems (Vaz et al., 2000). Its distribution includes the Paraná, Paraguay and Uruguay rivers basins (Calcagnotto and DeSalle, 2009). According to Urbinati and Gonçalves (2005), this species is considered one of the fish species with higher market value in Brazil, and significant volumes in native fish farming, as it is considered an easyadaptive species to rearing conditions (IBGE, 2016;Mourad et al., 2018). This species displays high fillet yield, and its meat is appreciated for its excellent flavor.
Based on the aforementioned, the aim of this study was to verify the effects of different feeding frequencies on productive performance, plasma glucose, hepatocytes integrity, frequency of muscle fibers distribution, enzymatic activity and proximate composition in P. mesopotamicus.

Methodology
The experimental procedures adopted in this study were previously approved by the Ethics Committee on Animal Use of the State University of Western Paraná the protocol number 04/18.

Experimental design
A total of 160 pacu juveniles with mean weight of 139.16 ± 8.86 g and mean length of 19.75 ± 0.67 cm were used. In order to accommodate the animals, 20 tanks of cylindrical shape with conical bottoms were used (useful volume of 0.5 m 3 ), coupled with constant aeration and adapted to a recirculation system.
The proposed feeding management was the parceling of feed supply (Ruohonen, Kettunen and King, 2001), split into one, two, three and four daily offers, according to the scheme described below: (1) Fish fed once a day; (2) Fish fed twice a day; (3) Fish fed three times a day; (4) Fish fed four times a day. A commercial diet was used in this experiment contained 35% crude protein, 5.5% ether extract, 5.0% ash, 13% fibers, 3% calcium and 1% phosphate.

Plasmatic metabolic variable
In order to obtain blood plasma, three animals from each experimental unit was submitted to blood sampling by puncture of the caudal vein using disposable syringes containing EDTA (ethylenediaminetetraacetic acid) at 10%. For glucose determination, the methodology of Dubois et al. (1956) was performed with the aid of a spectrophotometer at 450 nm. A standard glucose solution of 1 mg mL -1 was used for the estimation.

Growth performance analysis
At the end of the experimental period of 60 days, all fish underwent a fasting period of 24 hours, followed by anesthesia with a benzocaine solution (100 mg L -1 ) that was applied inside the aquariums where the fish were immersed, as recommended by Bittencourt et al. (2013). Subsequently, biometrics were performed to assess the animals' weight and length, and the following calculations were made: weight gain -WG (g) = (final weight -initial weight); specific growth rate -SGR (% day -1 ) = (100x [(ln final weight -ln initial weight)/time]; daily feed intake -DFI (g) = (total feed intake/time); Feed conversion rate-FCR = (supplied feed/weight gain); protein efficiency rate -PER (%) = (100 x (weight gain/consumed crude protein); survival -SU (%) = (number of fish at the end of the experiment/number of fish at the beginning of the experiment) x 100, and whole eviscerated fish-WEF (g) = (weight of whole fish without viscera).
After euthanasia, samples were collected for the evaluation of somatic indexes and histological analysis. The evaluated somatic indexes were viscerosomatic fat index (VFI (%) = 100 x [weight of visceral fat/body weight]) and hepatosomatic index (HI (%) = 100 x [weight of hepatic tissue /body weight]).
The proximate composition of the fish carcasses was determined according to the methodology of AOAC (2005), in which all animal were eviscerated and frozen (-20 °C) before analysis. These samples were dried for 24 hours (55 °C in an aircirculating oven) for dry matter analysis and burned (550 °C) for 6 hours to access ash content. Crude protein content was analyzed by the Kjeldahl method and the ether extract by the Soxhlet method.

Enzymes assay
Tissues sub-samples of the fish's liver were homogenized with a Teflon pestle motor-driven tissue cell disruptor (IKA_ T10 basic Ultra Turrax, Staufen, Germany) into 0.02 M Tris / 0.01 M phosphate buffer pH 7.0 plus glycerol 1/1 (v/v), at 4º C under ice bath and kept a ratio 1 : 10 (tissue: buffer). Homogenates were centrifuged at 13400 x g for 3 min (Centrifuge Eppendorf 5415R), and the supernatants (crude homogenates) were preserved for analyses.
The homogenates from tissue disruption were centrifuged at 600 x g for 3 min, and the supernatants were centrifuged at 6000 x g for 8 min at 3 ºC. Resultant supernatants were used as crude enzyme source. The enzyme assay was done in the following reaction medium Alanine aminotransferase (ALT) and Aspartate aminotransferase (AST) activities were determined by an end-point type assay and read at 430 nm. The reaction medium of both enzymes was: 222 mM alanine (ALT) or 44.4 mM aspartate (AST), 11.6 mM -ketoglutarate, 0.22 mM arsenate, 0.27 mM pyridoxal phosphate and 0.1M phosphate buffer pH 7.0 (Reitman and Frankel, 1957).

Histological analysis
The fragments of dorsal white skeletal muscle and liver were immersed in a buffered formaldehyde solution for 24 hours and then washed in alcohol 70%. After fixating the samples, fragments were dehydrated, diaphanized and included in paraffin. Then, cuts of 2-6 µm were obtained and stained with hematoxylin-eosin (HE) (skeletal muscle) or period acid of Schiff/hematoxylin (PAS-H) (liver samples). Microscopic analysis and material documentation were performed in a microscope Olympus BX41 (Tokyo, Japan).
In muscle samples, 200 muscular fibers were measured by animal, which were grouped in three diameter classes (<20, 20-50, and >50 µm), in order to evaluate the contribution of hyperplasia and hypertrophy to muscle growth (Almeida et al., 2008) in each treatment. In liver samples, we measured hepatocytes with pyknotic nuclei (%), hepatocyte area, area of the pyknotic nuclei and area of normal nuclei.

Statistical analysis
The used experimental design was entirely randomized with four treatments and five replicates, with eight animals being placed in each experimental unit. All data were submitted to a homogeneity and normality test (Levene) and to a variance analysis (ANOVA) at a 1% significance level and mean (standard deviation) were compared by the Tukey's test by software Statistic 7.1®.

Growth performance and proximate composition of carcasses
Feeding frequency influenced both daily feed consumption and HI of fish (p < 0.01). The animals that received four Research, Society and Development, v. 10, n. 4, e28710413979, 2021 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v10i3.13979 5 daily feedings consumed a larger amount of feed (p < 0.01) in comparation to one, two and three feeds per day. Fish that received feed on a daily basis presented higher HI (p < 0.01) in relation to those who consumed the diets in three distinct moments throughout the day (Table 1). The parameters regarding the proximate composition of carcasses presented differences for ether extract of fish fed once a day (p < 0.01) ( Table 2). Fish that received two feedings a day presented higher glucose indexes in comparison to other treatments (p < 0.01).
No significant differences were found regarding the enzymatic activity of the protein metabolism, ALT and AST in the animals' liver (Table 3).  (2021).

Liver histopathology
Feeding frequencies promoted an increase in the size of both normal and pyknotic hepatocytes, as well as in the nuclei, and the percentage of normal and pyknotic cells (p < 0.01), when fish were fed twice or four times a day, as shown in Table 4. The percentage of normal cells was higher when fish were fed one or three times a day.  Concerning the evaluation of the animals' livers, regardless of the used feeding frequency, the tissue did not present any apparent lesion (Figure 1). The histological cuts demonstrated a cord distribution of hepatocytes, and the sinusoids were coated with endothelial cells (Figure 1A, C). Hepatocytes presented clear cytoplasm with central nuclei and round shape ( Figure 1B) and in some regions, it was dislocated to the cell's periphery ( Figure 1D).

Muscle histomorphometric
The distribution frequencies of skeletal muscle fibers of pacu did not present significant differences among feeding frequencies (p > 0.01), as shown in Table 5.

Growth performance
Increasing feeding frequencies led to higher feed ingestion, although it did not interfere with the growth rate of fish.
The lack of effects on growth might be related to feeding habits and metabolism, factors that may interfere the utilization of nutrients and consequently the nutritional status of fish (Sampaio et al., 2007). Souza et al. (2014) showed that distinct feeding frequencies did not affect productive performance and physiological variables of tambaqui juveniles (Colossoma macropomum), while Sousa et al. (2012) demonstrated similar results for Nile tilapia juveniles. However, the authors suggest that when higher fractionations of offered feeds were performed, tilapia juveniles displayed a better use of nutrients -which was reflected in its productive performance, thus highlighting that metabolic processes are related to intrinsic characteristics of a species' development.
According to Marques et al. (2008), lower feeding frequencies reduces the amounts of consumed feed, due to the limited storage capacity of the fish's digestive system, thus influencing its physiological processes of compensatory growth (Hornick et al., 2000), and in the maintenance of the body right after re-feeding, so that the animals reach its adequate growth (Cho et al., 2006). In several species as the goldfish (Carassius auratus), tilapia (Oreochromis niloticus), common snook (Centropomus parallelus) and pacu (Piaractus mesopotamicus), increased feeding frequencies leads to improved productive performance, especially due to the increase in feed consumption and use efficiency (Dieterich et al., 2013;Tsuzuki et al., 2014;Santos et al., 2014;Zhao et al., 2016).
Feed intake increased along with feeding frequency, being higher in fish fed four times a day, thus demonstrating that fish submitted to higher feed frequencies grew equally in comparison to other groups. It is noteworthy that lower feed frequencies lead to increased appetite, aggression and voracity in fishes (Qin et al., 1997), as the catabolic process that unleash the activation of hormones such as cortisol is related to stress in animals (Nelson and Cox, 2011). An optimal frequency may improve the intestinal digestive function of fish, leading to higher growth and utilization of dietetic components (Wu et al., 2018). High feed rates promoted an increase in gastric motility, hampering the digestive and absorption functions of nutrients (Volkoff et al., 2017).
Alterations in HI are expected to occur due to changes in the offer of feed to fish. Santos et al. (2015) reported that the dynamic of endogenous utilization of energy reflects in the use of lipids and products formed through cellular stock, in both absorption and use in metabolic demand. The possible alterations related to increased feeding frequencies suggest that fish use food as a priority in order to meet the energetic demands of body maintenance and reposition of catabolized tissues, which may still be directed to growth (Pouey et al., 2012;Ribeiro et al., 2012;Zhao et al., 2016). In this sense, the somatic index suggests the use of energetic reserves. Fish fed only once a day presented similar growth in comparison to other treatments, although the liver apparently had to act more in the maintenance and supply of tissues, as shown by the HI.
The deposition of body fat in fish was directly influenced by feeding strategies. With the provision of feed once a day, metabolic processes promoted a greater fat deposition, which in addition to being available for the maintenance and functional development may benefit meat composition by improving the quality of deposited polyunsaturated fatty acids, according to the offered diet (Li et al., 2015, Zhao et al., 2016. Throughout fasting periods, the reserves of protein, glycogen and lipids tend to be reduced (Silveira, Logato and Pontes, 2009;Calvo et al., 2017).

Glycemic activity and hepatic aminotransferases
The maintenance of glucose levels was observed when fish were fed once, three or four times a day, which reinforces increased glycaemia. Throughout fasting periods, the maintenance of glycaemia is directly related to the capacity of mobilization of hepatic glycogen (glycogenolysis) -at least during the early fasting stages, and also depends of the later activation of hepatic gluconeogenesis (Souza et al., 2014) and of the reduction in the utilization rate of glucose (Moon and Foster, 1995). Although the hepatic levels of glycogen were not assessed in this study, the use of glycogen reserves for the maintenance of glycaemia probably did not occur, seen that the maintenance of the hepatosomatic index was observed.
The plasma glucose of reared fish may suffer alterations according to the adopted dietary management, due to the dynamics of the animal's energetic reserves (Souza et al., 2014). Fish have a limited capacity to metabolize glucose, i.e. remove this molecule from the plasmatic compartment, thus resulting in possible hyperglycemia that may remain for hours (Ribeiro et al., 2012) due to the increase in hepatic glycogenesis and lipogenesis (Seixas-Filho, 2004).
Several studies demonstrated that teleost fish are capable of tolerating periods of unfavorable feeding conditions by activating behavioral changes and biochemical, physiological and structural adaptive responses (Ashouri et al., 2019;Mozanzadeh et al., 2015). Fish have the ability of reducing their energy costs in fasting periods and reusing proteins (Salem et al., 2007). Plasmatic glucose levels is a routine index to determine nutritional status (Pottinger et al., 2003) and nutritional stress (Furné et al., 2012;Pérez-Jiménez et al., 2012). It is noteworthy that the maintenance of this metabolite during fasting has been reported for other fish species (Ashouri et al., 2019;Bermejo-Poza et al., 2019;Mozanzadeh et al., 2015;Yarmohammadi et al., 2012), which reveals the capacity and plasticity of some fish species in the face of short-term fasting. In general, serum glucose are maintained by glycogenolysis (the break of glycogen in the liver by gluconeogenesis), generating glucose from lactate, amino acids and glycerol (Favero et al., 2019;Ashouri et al., 2019).
In this study, no differences were found in the liver enzyme content of the animals (AST and ALT), which did not indicate damage in hepatocytes, seen that these enzymes are produced in these cells and its increased concentration is associated with acute hepatic lesions. The enzyme AST is found mainly in mitochondria, while ALT is a hepatic cytosolic enzyme (Nelson and Cox, 2011). As transamination enzymes did not display significant differences, it could be inferred that no protein catabolism occurred. Studies of severe restrictions reported increased activity of these enzymes, suggesting higher gluconeogenesis from substrates such as amino acids and lactate during fasting periods (Favero et al., 2019;Ashouri et al., 2019), with prominent nutritional stress states (Furné et al., 2012;Pérez-Jiménez et al., 2012). However, when fish were fed once again, serum levels returned to normal. In this study, fasting was not severe in order to mobilize other molecules to supply the energy demand, with circulating glucose coming from feed.
Similar levels of AST and ALT activity indicate metabolic stability in the liver tissue, regardless of the longer period of catabolism in fish that were fed less meals a day (Thrall et al., 2015;Menga et al., 2018). Based on the findings of this study, splitting feeding portions from up to four feedings per day does not alter the protein metabolism.

Liver and muscle morphology
Due to the central role in fish metabolism, the liver is responsible for metabolic dynamics. Initial hepatic alterations during the digestion process mainly affect the structural and functional characteristics of hepatocytes (Carter and Houlihan, 2001;Fujimoto et al., 2016).
The observed morphological characteristics in the liver of pacu are in accordance with the performed descriptions for other species of Neotropical fish (Fujimoto, Cruz and Moraes, 2008). No alterations were observed regarding organ overload, which may occur due to excessive protein levels in the diet and adaptations to nutrients and additives (Honorato et al., 2013).
When fish were fed twice or four times a day, an increase in the number of hepatocytes with pyknotic nuclei was observed. Variations in the size of hepatocytes and its nucleus in response to changes in feed supply was evidenced by Ostaszewska et al. (2005) in P. mesopotamicus submitted to diets containing different sources of protein. According to Segner and Braunbeck (1988), the size of the hepatocyte's nucleus is directly related to the metabolic activity of the liver cells. In addition, protein synthesis rates may be shifted, leading to increases or reductions of its activity, possibly due to variations in Research, Society andDevelopment, v. 10, n. 4, e28710413979, 2021 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v10i3.13979 diet composition and quantity of provided nutrients in feed (Carter and Houlihan, 2001;Fujimoto et al., 2016).
The growth of skeletal muscle fibers is influenced by several external factors, such as nutrition (Koumans and Akster, 1995). The fact that there were no differences between the distributions in growth classes indicates that the feed contained available nutrients (e.g. amino acids), thus the fish maintained their anabolic rates and were able to synthetize proteins and form muscle cells (Brown and Cameron, 1991), a fact that can also be reinforced by the protein efficiency rate, which did not present significant differences among feeding frequencies.
The frequency of muscle fiber distribution indicates that the fish had hypertrophic growth, as a great share of the fibers were 20-50 or above 50 µm, which was also verified by Almeida et al. (2010) for the same species. A small frequency of fibers lower than 20 µm was noted in all sampled animals, showing once again that the animals continued to recruit new fibers for muscle growth, which is very important for this species, which has a great economic importance. Almeida et al. (2008) suggested that there is a strong association between MyoD and myogenin (factors that regulates muscle growth) and the growth stages of pacu and related them to hyperplasic and hypertrophic growth.

Conclusion
Different feeding frequencies did not compromise the growth of pacu (P. mesopotamicus). Thus, feed supply into four feeding events is recommended for the development of this species at this life stage.