Effect of Fasting and Refeeding on Juvenile Leopard Mandarin Fish Siniperca scherzeri

Simple Summary The development of live prey feed domestication techniques for the carnivorous leopard mandarin fish (Siniperca scherzeri) requires establishing nutritionally complete diets and research into optimal feeding regimes. Accordingly, the purpose of this experiment was to investigate the effects of varied feeding regimes for S. schezeri on compensatory growth. Individuals were divided into a daily feeding control group (C), a 5-day fasting group (F5), a 10-day fasting group (F10), and a 14-day fasting group (F14). The results demonstrated that after four weeks of treatment, all experimental groups experienced full compensatory growth and reached the control group weight. Therefore, this information will be valuable for effective management when fasting is necessary in S. schezeri aquaculture. Abstract To verify the effect of fasting on juvenile leopard mandarin fish (Siniperca scherzeri mean weight, 14.7 g), compensatory growth, body composition, and blood content of juveniles were investigated for six weeks following two-week feeding treatments: fed continuously (control), and fasted for 5 (F5), 10 (F10) and 14 days (F14). Full compensatory growth was evident after four weeks of food resupply in all fasting groups. Specific growth rate, feeding rate, and feed efficiency in all fasting groups were significantly higher than those of the control after the first 2 weeks of food resupply. At the end of fasting, the lipid content, ratio of lipid to lean body mass, hepatosomatic and viscerosomatic indices in all fasting groups, or total cholesterol content in F14 significantly decreased compared to the control. These results indicated that juvenile leopard mandarin fish subjected to 5–14 days of food deprivation could achieve full compensatory growth after feeding resumption for 4 weeks and that the morphological and biochemical indices, as well as body and blood composition, remained comparable to the control group after the completion of the study under our experimental conditions.


Introduction
Fish farmers employ fasting or limited feeding strategies in short-and long-term durations to reduce mortality or stress during water quality degradation, disease outbreaks, sorting, transferring, or harvesting or to simply improve feeding efficiency and reduce costs [1][2][3][4][5][6]. In contrast to fish fed continuously, those subjected to restricted feeding or fasting can experience greater growth once satiation feeding is resumed [3,5,7,8]. Such tendencies to return to original weights or growth trajectories are referred to as compensatory or catch-up growth [2,[6][7][8][9][10][11].

Experimental Design and Management
Following acclimatization, 300 juveniles (initial mean body weight: 14.7 g) were randomly assigned to one of four different feeding groups. Each feeding group consisted of 5 tanks (i.e., replicates), and 15 juveniles were stocked per tank. Control group (C) was fed to apparent satiation twice daily (09:00 and 17:00) during the experimental periods, whereas the three remaining groups fasted for: 5 days (F5) from days 10-14 after initiation of the experiment; 10 days (F10) from days 5-14; 14 days (F14) from days 1-14 before satiation refeeding in week 3. The three fasting groups were fed to satiation same as control. Daily feeding was performed by dropping pellet (the same as during the acclimation period) into each tank every few minutes until the fish ceased to eat due to satiation. Care was taken to ensure that all pellets provided were consumed by the fish. The container of the pellets was weighed before the first feeding and after the last feeding daily. Uneaten pellets were collected from each tank by siphon and were dried in the oven to a constant weight. The actual daily consumed feed amount was determined as the difference in weight between the feed supplied into the tank and the uneaten feed removed by siphon. The experiment lasted 8 weeks.
Prior to the experiment, all the fish in each group were fasted for 24 h for gut evacuation, followed by anesthesia with a 2-phenoxyethanol solution (150 mg L −1 ; Sigma, St. Louis, MO, USA) to minimize the stress before body weight and body length measurement. Before measurement, all excess water on the fish skins was removed by blotting with a paper towel and then measured to the nearest 0.01 g and 0.1 cm.
To evaluate the effect of fasting and refeeding, all fish in each group were bulk-weighed at the end of weeks 2, 4, and 6 and at the end of the experiment (i.e., week 8), individual body weight and body length were measured using the same method above. Fish were not fed on the day of measurement.

Analysis of Blood and Body Content
The blood composition, hepatosomatic index (HSI), viscerasomatic index (VSI), protein, lipid, ash, and moisture content of the four feeding groups, as well as the lipid:lean body mass ratio (lipid/LBM) were measured at the ends of weeks 2 and 8. At the end of week 2, 2 tanks in each feeding group were randomly selected, and 7 fish per tank were randomly subjected to blood composition, HSI, and VSI analyses, while the remaining 8 fish per tank were subjected to whole-body analysis. At the end of week 8, the fish in the remaining three tanks of each feeding group were sampled for analysis of blood and body compositions using the same method as at end of week 2. Fish were anesthetized in 150 mg/L of 2-phenoxyethanol solution for 1 min after the measurement of body weight or body length. Then, they were stored at −40 • C for the analysis of body composition.
Before sacrificing the fish, blood samples of the fish were collected from the tail artery using a heparinized syringe following anesthesia with 2-phenoxyethanol solution (150 mg L −1 ). The blood collected was centrifuged at 8870× g for five minutes to extract serum, after which hemoglobin (Hb), glucose (GLU), total cholesterol (TCHO), glutamic oxaloacetic transaminase (GOT), glutamic pyruvic transaminase (GPT), and high-density lipoprotein cholesterol (HDLC) in the extracted serum were measured with a DRI-CHEM NX500i (Fujifilm Co., Tokyo, Japan). Immediately after blood sampling, the sacrificed fish were also dissected, and the weights of the liver and viscera were measured to determine HIS and VSI.
Prior to whole-body analysis, fish samples were freeze-dried and homogenized. Proximate analysis was conducted according to the AOAC methods [33]. The crude protein content was determined with the Kjeldahl method using the Auto Kjeldahl System (Buchi, Flawil, Switzerland). After 24 h of desiccation in a dry oven at 105 • C, the moisture content was measured. Crude fat was measured using the ether-extraction method, and crude ash content was determined following 4 h of combustion at 600 • C.

Growth Parameters and Statistical Analysis
The following growth parameters were calculated: specific growth rate (SGR, % d −1 ) = 100 × (lnW f − lnW i )/t, feeding rate (FR, % body weight d −1 = 100 × C/[(W f + W i )/2]/t, and feed efficiency (FE, %) = 100 = (W f − W i )/C where W f and W i are final and initial weights (g), t is the feeding duration (day) and C is total feed consumption (g) during t days.
All statistical analyses were performed in SPSS v.20 statistical software (SPSS Michigan Avenue, Chicago, IL, USA). Each replicate was considered an experimental unit, and therefore, the mean value obtained from a replicate within each feeding group was used as a data unit [3,5,8,34]. At the beginning and week 2, all feeding groups had 5 replicates (n = 5), while 3 replicates (n = 3) were used after week 2. One-way analysis of variance (ANOVA) was separately used to compare initial body weight, blood, and body contents among the four feeding groups, and when significant differences were identified, Tukey's multiple range test (p < 0.05) was used to verify the significance between the means. Levene's test was used to verify the homogeneity of variance prior to ANOVA, and percent data were arcsine-transformed. Repeated measures analysis of covariance (ANCOVA) with one factor was used to determine the effects of four feeding groups on body weight from week 2 to week 8 and on SGR, FR and FE from week 3 to week 8. In the ANCOVA of body weight, initial body weight was used as a covariate to remove any biases resulting from the allometric relationship between body weight and growth [3,5,8,34]. In the ANCOVA of SGR, FR, and FE, body weights at the beginning of weeks 3, 5, and 7 were used as covariates to assess growth potential, as it returns to the original growth trajectory [3,5,8,34].
During the first two weeks of feed resupply (weeks 3-4), FRs in all fasting groups were significantly higher than in the control group (F = 33.865, df = 3, p < 0.001; Figure 3). During the second two weeks of feed resupply (weeks 5-6), F14 showed significantly higher FR than the control group, while F5 and F10 did not differ from the control group (F = 7.707, df = 3, p = 0.01). The FR during the final two weeks of resupply (weeks 7 to 8) was not significantly different between the experimental groups (F = 1.034, df = 3, p = 0.428). The overall FR during the six weeks following the fasting period (weeks 3-8) showed the highest FR in F14, followed by F10, F5, and the control group (F = 20.103, df = 3, p < 0.001).

Discussion
In modern aquaculture, producers are seeking to reduce production costs in a manner without adversely affecting productivity or the environment [4]. As feed represents the largest share of aquaculture production costs [35], commercial aquaculture farms aim to develop appropriate feed supply management and saving methods [36]. Compensation growth can be used as a feed supply strategy, whereby satiation feeding after feed deprivation can lead to faster growth and greater feed efficiency, translating into improved economic efficiency by reducing feed and labor costs [37][38][39]. Depending on the strength of fish body weight recovery following fasting, the degree of compensatory growth is classified as over, complete (i.e., full), partial, and non-compensatory [5]. Notably, the degree of compensatory growth is sensitive to both the degree and duration of fasting prior to satiation feeding [5,34,40]. Here, all juvenile S. scherzeri subjected to five-(F5), 10-(F10), and 14-day (F14) fasting showed full compensatory growth that caught up with the weight of the control group four weeks after feed resupply. Similarly, full compensatory growth after ≤2 weeks of fasting was reported to be observed in rock bream (Oplegnathus fasciatus) [5], Chinese longsnout catfish (Leiocassis longirostris) [39], gibel carp (Carassius auratus gibelio) [41], and gilthead seabream (Sparus auratus) [42]. In addition, full compensatory growth was reported in black rockfish (Sebastes schlegeli) after five days of fasting [8], barramundi (Lates calcarifer) [43] and hybrid tilapia (Oreochromis mossambicus × O. niloticus) [44] after one week of fasting, and red seabream (Pagrus major) after 1-3 weeks of fasting [3]. The relative ratio of the body weight of fish subjected to fasting as compared to those fish in the control group has been previously discussed as a critical indicator of full compensatory growth [3,8,41,43]. Tian and Qin [43] determined that after fasting, full compensatory growth occurred for ≥60% of fish body weight in the control group, and the same phenomenon was reported to be observed in gibel carp [41], hybrid tilapia [44], black rockfish [8], Chinese longsnout catfish [40], gilthead seabream [42], rainbow trout (Oncorhynchus mykiss) [45], and rock bream [5]. During the present experiment, F5, F10, and F14 all showed full compensatory growth after fasting for 5 to 14 days for individuals with body weights of ≥~60% (61.7-81.1%) than the control group.
The SGR and FR values after fasting increased during the early feed resupply stages (weeks 3-4) with an increasing duration of food deprivation, whereas similar values to the control group were observed during the latter food resupply (weeks 7-8), comparable to results previously reported for rainbow trout [45] and rock beam [5]. Generally, these phenomena occur during the initial resupply of feed immediately following the period of deprivation due to a hyperphagic response [5,8,44], inducing compensatory growth that decreases with time [5,42] as well as a decrease in the digestibility of excess feed intake [26,32,46,47]. Therefore, during the initial resupply of food in the present study, the FE of the fasting group was very high; however, with time, the FE decreased or was not significantly different from the control group. Similarly, improved FE during feed resupply was documented for European minnow (Phoxinus phoxinus) [48], gibel carp [49], Chinese longsnout catfish [40], Atlantic halibut (Hippoglossus hippoglossus) [50], blackhead seabream (Acanthopagrus schlegelii) [14], and rock bream [5]. Throughout the present study, hyperphagia and improved FE tended to decrease with time, both of which were attributed to the two factors acting as major determinants of compensatory growth of S. scherzeri, as this tendency was seen in all fasting groups, and similar results have been observed in other previous studies as well [3,5,34]. It has been reported, however, that compensatory growth occurs primarily due to the increase in FR and not from improvements in FE during feed resupply after fasting [8,41,45,[51][52][53], suggesting that the major factors driving compensatory growth differ.
In the experiment here, the CF, CVBW, and CVBL of juvenile S. scherzeri were not affected by compensatory growth. Similarly, Silva et al. [17] reported no differences in CF, final weight, and total length between the control group and the replicates of fasting groups. Fasting groups in this study experienced a rise in moisture content after feed deprivation, and similar to previous studies, lipid content and HSI decreased [42,43,54,55]. Using a lipostatic model, one can predict the timing of compensatory growth based on the principle that such growth occurs when the lipid/LBM decreases and ends when the lipid/LBM returns to control group levels [56]. In most animals, the lipid stores indicate the easily activated energy reserves, and the lean body mass (LBM) includes the structural tissues, so the lipid/LBM may be considered an indicator of nutritional state [56]. The lipostatic model hypothesized that restricted feeding induces changes in adiposity through compensatory feeding as a result of negative feedback due to changes in energy balance [56]. Therefore, fat loss after restricted feeding reduces the negative feedback on feed intake and elevated feed intake until normal fat levels and lipid/LBM are maintained. In other words, the lipostatic model represents that the size of body lipid reserves affects the feed intake (i.e., hyperphagic response) of fish via the negative feedback signal to the central nervous system, thereby regulating the feed intake behavior, body weight, and body composition during compensatory growth. Therefore, if the body lipid reserves of fish are repleted very slowly, the compensatory growth through the hyperphagic response can be sustained because the imbalance of the lipid/LBM is maintained for a long time [56]. It is very important to extend the period of imbalance of the lipid/LBM by controlling the repletion of body lipid storage during refeeding after restricted feeding or fasting, and further study is needed in the future. This model has been applied to Atlantic salmon (Salmo salar) [57] and Nile tilapia (Oreochromis niloticus) [58]. The present experiment found that the lipid/LBM of the F5, F10, and F14 groups were significantly lower than that of the control group following fasting (end of week 2), and no differences in SGR and lipid/LBM were observed between the fasting and control groups at the end of the experiment (end of week 8), thus indicating the efficacy of the lipostatic model. Despite this alignment here, numerous previous studies' results were inconsistent with the lipostatic model [3,5,8,[40][41][42]45,59], suggesting that such results may vary by species. Based on the results of the experiment here, the HSI after fasting significantly decreased compared to the control group as the duration of fasting increased. Silva et al. [17] reared pacamã (Lophiosilurus alexandri) for a total of 45 days with a 6:1 and 5:2 feed to fasting regime, and their results showed that the HSI was the lowest in the 5:2 group, which is similar the results found in the present study. As an indicator of fish nutritional status [60], the HSI is a ratio of liver to body weight and can be used to quantify the energy stored in the liver as glycogen [61]. Metabolic and regulatory processes associated with nutrients and lipogenesis originate in the liver of fish [62]. As demonstrated in the present study, the lower HSI values with longer fasting periods indicated the need for greater hepatic energy storage, and Morshedi et al. [15] reported that fasting decreased the amount of hepatic energy available, directly affecting liver weight. At the end of the experiment, the higher HSI in the fasting groups compared to the control was due to an excess accumulation of energy during the recovery process associated with feed resupply. Fish can reduce their energy requirements by reducing the mass of gastrointestinal tissues during a fasting period [63], which is consistent with the findings of the VSI reduction observed after fasting in the present study. According to Jobling [64], fasting increases the digestive tract capacity of fish, allowing them to consume more due to the hyperphagia that may occur throughout the recovery period. This would partially explain the increase in VSI of the fasting group seen at the end of the present experiment and correlate with the hyperphagic behavior observed here as well.
A hematological and biochemical assessment of fish has been used to determine their physiological status [15], and it was revealed that these factors could be affected by biotic and abiotic factors, such as age, sex, water temperature, seasonal patterns, and food intake [65]. At the end of the present study, it was found that fasting and resupplying had no notable effects on the hematological characteristics of juvenile S. scherzeri, and similar results were found with the Siberian sturgeon (Acipenser baerii) [15], Nile tilapia [66], olive flounder (Paralichthys olivaceus) [67], European eel (Anguilla anguilla) [68], red porgy (Pagrus pagrus) [69], and beluga (Huso huso) [70]. During the present study, it was observed that fasting influenced GLU and TCHO levels in the plasma of juvenile S. scherzeri; however, no such increase in plasma GLU was observed in sea bass (Dicentraachus labrax) [71] or pirapitinga (Piaractus brachypomus) [72]. Cholesterol is a structural lipid utilized or synthesized in response to feed deprivation; however, it is also maintained during feed deprivation without increasing, decreasing, or changing depending on the species and fasting duration [72]. For the present study, the cholesterol concentrations decreased after fasting (week 2), similar to the results observed for rainbow trout [73], brown trout (Salmo trutta) [74] and sturgeon (Acipenser naccarii) [75]. Additionally, a decrease in feed intake of S. scherzeri due to either a decrease in feeding frequency or feeding ratio was also associated with a decrease in plasma TCHO concentrations [26,31,32]. In contrast, no change in TCHO concentrations was documented for pirapitinga [72], tamaqui (Colossoma macropomum) [6], and pacamã [17], whereas an increase was observed with pacu (Piaractus mesopotamicus) [76] and climbing perch (Anabas testudineus) [77], similarly suggesting variable results according to species and fasting regime. Therefore, further complementary research is necessary for the future.

Conclusions
The present experiment determined that all juvenile S. scherzeri (initial mean weight, 14.7 g) subjected to 5-14 days of food deprivation experienced full compensatory growth after four weeks of food resupply under our experimental conditions. Responding well to the feeding strategy of fasting and resupplying, S. scherzeri maintained similar body compositions, blood, morphological, and biochemical characteristics compared to the continuously fed control group within 6 weeks after resupply commenced. This information may be helpful to fish farmers who encounter adverse conditions during rearing or require practical feeding strategies to increase feed efficiency while reducing costs. Data Availability Statement: The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.