Female Tilapia Strategising Energy Mobilisation Differently For Growth Or Reproduction Depend On Living Environments

This study was conducted to investigate the energy mobilization and ionoregulation pattern of tilapia living recirculating aquaculture system (RAS) and cage culture environments. Three different groups of tilapia were compared as tilapia cultured in RAS (Group I - RAS), tilapia cultured in open water cage (Group II - Cage) and tilapia transferred from cage to RAS (Group III - Compensation) as physiology compromising model. Results revealed that Group II tilapia mobilized glycogen as primarily energy for daily exercise activity and promoted growth, whilst tilapia from Group I and III mobilised lipid to support gonadogenesis and protein reserved for somatogensis. The gills and kidney NKA activities remained relative stable to maintain balance homeostasis with a electrolytes level. As a remark, this study revealed that tilapia re-strategized their energy mobilization pattern in accessing glycogen as easy energy to support exercise metabolism and mobilized lipid and protein for growth and gonadal development.


Introduction
Tilapia from the genus Oreochromis is the second most farmed species after cyprinid that accounted for more than 6.93 million tonnes in 2020 and expected to continue rising with the increasing global population (FAO, 2020;Zeng et al., 2021). In 2021, the market of tilapia had continuously survive even under Covid-19 pandemic and expected to continue gains steady to reach 7.3 million (FAO, 2021a).
Popularity of tilapia farming crossing more than 140 countries with global production market values recorded at USD12 billion in 2018 and expected to reach USD25 billion by 2028 (FAO, 2018).
Tilapia as a hardy sh which can be farmed from extensive to super intensive production systems from conventional pond to cage culture systems or either partial or complete recirculating tank system (Kabir et al., 2019). However, high density farming leads to stress in competition for growth which may induce stress related immune inhibition and increase disease susceptibility (Munangandu et al., 2016). However, production lost that worth about USD10 billion that accounted 40% in tilapia farming is predicted by year 2028, if production line continuous expose to environment stresses, diseases and quality stock (Owens, 2012;Joshi et al., 2021). Some reports highlighted that open water cage culture of tilapia exhibited slow growth and mortality reaching 60-70% occurred yearly especially during the dry season (Siti-Zahrah et al., 2008). The causes are unclear and require convincing evidence to elucidate the issue thoroughly. In Malaysia, Como River -Kenyir Lake located at Terengganu state of Malaysia is categorised for open cage farming. However, many farmers experience production lost and slow growth, which resulted in abandonment of their farm operation (Siti-Zahrah et al., 2008). By putting disease infection apart, the losses of the production could be related to physiological stress affected by living environment. One of the possibility may be due to extra energy expenses to cope with the living condition and maintain the homeostasis of active ion uptake. Active swimming is known require extra energy expenses to support exercising metabolism (Liew et al., 2012;Liew et al., 2013). In addition, maintaining active ionoregulation is crucial to retain and remain ionic stability in the body for all biology processes. Performing active ionoregulation is an energetically expensive process that require about 1-20% of the total ATP demand (Evan and Claiborne 2008; Moyson et al 2015). Thereby, prioritizing energy metabolism is believe reserve for exercise and ionoregulation that be promote somatogenesis in tilapia cultured in cage system (Waldrop et al., 2018;Inoue et al., 2019).
However, due to Covid-19 pandemic with restriction in operation under Movement Control Order (MCO) in many countries (Waiho et al., 2020), domestic food supply and chain become important to ensure nation food security during pandemic situation. As tilapia is one of the most popular choice in Malaysia markets. Thus, micro tilapia farming operation have restarted by using abandoned backyard earth ponds to support local community needs and contribute to household income during this pandemic period. Nevertheless, a lot of these micro tilapia farmers complaining that the growth rate of tilapia cultured in pond was not comparable to cage culture tilapia. With this question raised up by local farmers, we speculated that tilapia that cultured in pond system tend to spend energy for secondary maturation and reproduction. As tilapia is a mouthbrooder species that reached sexually maturation at size of about 100 g body weight (Specker and Kishida, 2000). The male tilapia started establish territory site digging, cleaning and guarding his territory to attract female tilapia. Once spawning process take place, female tilapia immediately collect fertilized eggs and incubate in her mouth for two weeks (FAO, 2021b). Thereby, our rst hypothesis assume that tilapia tend to priority their energy for reproduction especially female tilapia which slow growth is expected. Whereas, tilapia that cultured in cage system spent their energy for exercise and growth, because of cage condition not suitable for reproduction. Secondly, maintaining homeostasis balancing is important in order to order to perform all basal metabolism needs. As compared tilapia that cultured in open cage with recirculating aquaculture system (RAS), tilapia in cage culture would have to maintain high level of ionoregulatory activity such as sodium pump (NKA activity) to maintain essential ion. While, high NKA activity also expected in RAS cultured tilapia which believed in associate with ammonia excretion. Increase ammonia excretion e ciency concurrently induces an increase NKA activity also been reported previously on other species such as common carp (Liew et al., 2012), rainbow trout , climbing perch (Chew et al., 2014).
With this hypothesis as background, the objective of this study aimed to investigate the energy mobilisation pattern and ionoregulation of tilapia cultured in different living conditions. The status of energy expenses and ionoregulation of freshwater teleost in the open water cage culture still remains to be investigated. Therefore, with special interest on the metabolic adjustments that include energy mobilization and ionoregulation pattern were pursued on the tilapia farmed at open water cage culture at Como River, Kenyir Lake (Group II -Cage) in compared with tilapia cultured in recirculating aquaculture system (Group I -RAS). In terms of energy mobilisation, we hypothesized that lipid was the priority energy being mobilized followed by glycogen and protein with greater ionoregulatory activities under this circumstances. In addition, we also examined their physiological remodelling strategy of tilapia by transferring the tilapia from the open water cage of Kenyir Lake back to indoor recirculating aquaculture system (RAS) and cultured for 4 weeks (Group III -Compensation) to compare the energy and ionoregulation patterns.

Source of specimens and management
Hybrid red tilapia Oreochromis sp. was used in this study, the experimental tilapia were divided into three different groups. Group I -RAS referred to the hybrid red tilapia that were cultured in enclosed recirculating aquaculture system (RAS) for two month at the hatchery facility in AKUATROP, UMT. Group II -Cage referred to the hybrid red tilapia that were cultured in the crystal clear open water oating cages (5°02'22.1"N 102°50'41.1"E) at Como River, Kenyir Lake, Terengganu. Fish from Group II were harvested after two months period of cultivation. Group III -Compensation referred to the hybrid red tilapia that were cultured at the crystal clear open water oating cages, which were then transferred back to the hatchery AKUATROP hatchery and cultured in RAS facility for two months. Fish were fed 3% of body weight (BW) with commercial tilapia pellet (TP-2 Star-Feed®) (28% protein and 3% crude fat) twice daily at 8:30h and 16:30h. During cultivation period for all groups, water pH was maintained at 6.5-7.5, temperature at 26.5-28.5°C and dissolved oxygen at 5.2-6.8 mg/L. A volume of 50 ml water samples were collected from all groups for ionic measurement.
Experiment series I -physiological compensatory strategy comparison of hybrid red tilapia living in Group I -RAS and Group II -Cage In order to compare the physiological responses of hybrid red tilapia living in different conditions, both groups of tilapia and water samples were analysed. For Group I (RAS) -20 hybrid red tilapia that have been cultured for 2 months were sampled directly from RAS at a body weight (BW) of 167.69 ± 4.35 g and body length (BL) of 21.12 ± 0.20 cm. For Group I, 20 tilapia were introduced randomly into a rounded berglass tanks at a capacity of 2000 L equipped with an external ltration system at a capacity of 1 tonne in volume. Filter system consisted of 5 compartments with particle settlement as the rst unit followed by sponges and bio-rings units to maintain good water quality level. Throughout culturing period, only settlement and sponges units were cleaned weekly to remove excessive wastes. Water in the system were refreshed weekly at 40% and monitored by using YSI multiple meter (YSI-556 MPS). shes per tank and cultured for 4 weeks (compensation recovery strategy). Weekly, 10 shes were sampled randomly from the system to examine their physiological strategy to compare their physiological parameters. During cultivation, similar feeding schedule and feed were applied for compensation recovery strategy group. Water in the system were refreshed weekly at 40% and monitored by using YSI multiple meter (YSI-556 MPS).

Sampling procedures
At every sampling time point intervals, a total of 10 female shes were collected randomly for biometric characteristics measurement and tissue collection for biochemistry analysis. During sampling process, all selected shes were anesthetized with clove oil at 10 mg/L (Thalid et al., 2021). The clove oil was rst mixed with ethanol to make a stock solution at a ratio of 1:10 (clove oil:ethanol) before use in order to assist emulsi cation. After sh showed passive operculum movement and loss of equilibrium, they were immediately removed and blotted for biometric measurement followed by blood sampling. Blood were drawn via caudal peduncle using 1 ml heparinized syringe and carefully expelled into heparinized 1.5 ml bullet tube. Samples were immediately centrifuged at 5,000 g under 4°C for 30 sec. Thereafter, plasma samples were transferred into another 1.5 ml bullet tube and immediately frozen in liquid nitrogen (N 2 ). In order to collect other tissues, sh were euthanized with a sharp blow to the head. Thereafter, gills, liver, kidney, muscle tissues and gonad (if available) were excised quickly. Wet liver mass was measured and all other tissues were wrapped in aluminium foil individually. All samples were frozen in liquid nitrogen immediately and stored in -80°C freezer until analysis. Both liver and muscle tissues were used for bioenergy analysis, while gills and kidney samples were used for enzymatic electrolytes ATPase transporters analysis. Biometric measurement were used to calculate condition factor, whereas liver mass and gonad volume were used to calculate hepatosomatic index (HSI) and gonadosomatic index (GSI).
Condition factor was calculated as K-factor = ¼ 100 x (W/L 3 ), where, W is the body weight of sh (g) and L 3 is the total length (cm). Whereas, HSI = 100 x (Lw/BW), where Lw is weight of the liver mass (g) and BW is body weight (g). GSI was calculated as GSI = 100 x (Gw/BW), where Gw is the weight of ovary (g) and BW is the body weight (g). In this study, GSI was measured only for female tilapia, as female is the parent performing mouthbrooding incubation with no food intake during this period.

Tissue metabolites
For bioenergy analysis, both liver and muscle tissues weighed about 2g was homogenized using handheld homogenizer under ice chilled condition . Homogenization was performed at 5x folds dilution factor with ultrapure water (Milli-Q grade). Thereafter, total bioenergy of the liver and muscle tissue were analysed for lipid, protein and glycogen contents. Lipid was extracted by methanolchloroform and measured with a tripalmitin standard reference (Bligh and Dyer, 1959). Protein measurement was performed by following Bradford method (Bradford, 1976) and using bovine serum albumin as standard reference. While, glycogen content was measured using Anthron method with glycogen standard as reference (Roe and Dailey, 1966).

Plasma osmolarity and electrolytes
Plasma osmolality levels were measured using Osmometer (Advanced Instrument Inc. -Model 3320) with unit expressed as mOsm/l. Plasma electrolytes such as Na + , K + , Cl − , Ca 2+ and Mg 2+ were measured using Ion Chromatography Analyzer (Metrohm 81 Compact IC Plus -Model 883) with unit expressed as mmol/L. Gills and kidney enzymatic Na + /K + ATPase activity Gills and kidney Na + /K + ATPase (NKA) activity was measured according to the method described by McCormick (1993) and Liew et al. (2015). A total of 8 samples from each gills and kidney samples were randomly selected for electrolytes enzymatic ATPase activity analysis. Selected samples were homogenized with the mixture of ice cooled neutralized SEI/SEID buffer solution (SEI − 150 mM sucrose; 10 mM EDTA; 50 mM imidazole solution / SEI with 0.1% sodium deoxycholate solution) with buffer solution pH 7.5 at ratio of 4:1. Thereafter, samples were centrifuged for 1 min at 5,000 g under 4°C to obtain enzyme supernatant. During enzymatic measurement, duplication of 10 µl supernatant samples were pipetted and carefully transferred into 96-wells microplate in two series. A freshly made 200 µl mixture cocktail assay solution A (400 U lactate dehydrogenase; 500 U pyruvate kinase; 2.8 mM phosphoenolpyruvate; 0.7 mM ATP; 0.22 mM NADH; 50 mM imidazole) were added into the rst series supernatant and 200 µl mixture cocktail assay solution B (mixture assay A with 0.4 mM ouabain) were added into the second series supernatant on the microplate. The NKA activity were measured kinetically by using spectrophotometer (MultiskanTM FC microplate photometer, ThermoFisher Scienti cTM) read at 340 nm for 10 min with 15 s intervals. Adenosine diphosphate (ADP) was used as standard reference. NKA activity were calculated by subtracting the oxidation rate of NADH in the presence of ouabain from the oxidation rate to the NAD in the absence of ouabain. The crude homogenate protein was determined by using bovine serum albumin (US Biochemical, Cleveland, OH, USA) as standard reference and read at 430 nm according to Bradford (1976). The NKA activity unit was expressed as µmol ATP/h/mg protein .

Statistical analysis
The results of growth indication, plasma osmolality, electrolytes, bioenergy and NKA ATPase activities were presented as mean ± standard error mean (SEM) (n = 10). Prior to signi cance analysis, all data were checked for normality distribution by using Shapiro-Wilk test and homogeneity of variance by using Levence test. In case of failure to ful l normality and homogeneity requirement, data were either log or arcsine square root transformed prior further analysis. Data collected from experiment series I and II (Group I-RAS, Group II-Cage and Group III-Compensation) on weekly progress were compared by using one-way analysis of variance (ANOVA). Tukey HSD post-hoc test was performed to identify signi cant differences among experimental series groups set at 95% con dent limit at p < 0.05.

Biometric indication
In term of growth performance in term of BW and BL, no obvious differences was noticed (Table 2). Highest BW was recorded in Group II at 212.11 ± 3.79g, which was signi cantly highest compared to Group I at 167.69 ± 2.35g and Group III at week-1 with BW recorded only at 162.31 ± 2.49g. Whereas, BL have no signi cant different at all groups. As shown in Table 2, the average hepatosomatic index (HSI) of tilapia from Group I was recorded at 1.45 ± 0.04, which was signi cantly higher compared to tilapia from Group II at 1.02 ± 0.05 (P < 0.01). For Group III, the HSI increased signi cantly from week-1 until week-3 of culture period. The HSI noted at week-1 was 1.23 ± 0.18, had shown a signi cant increasing pattern with 1.51 ± 0.10 at week-2 and reached the highest at week-3 with HSI of 1.95 ± 0.21. However, a slightly decreasing trend was noticed at week-4 with HSI at 1.57 ± 0.11. Whereas, GSI for Group I was recorded at 4.48 ± 0.15 and Group II at 3.15 ± 0.51, respectively. When in compared with Group III, GSI for week-1 was 3.36 ± 0.24, week-2 at 3.45 ± 0.24 and week-3 at 4.09 ± 0.59, which were insigni cant compared to Group II. However at week-4, GSI was recorded at 4.61 ± 0.56 that similar at Group I, but signi cantly higher compared to Group II and those from Group III at week-1, -2 and − 3, respectively (Table 2). Condition factor (K-factor) varied considerably from 1.54 ± 0.01 and 2.07 ± 0.04 (Table 2). Lowest K-factor was recorded in Group I at 1.54 ± 0.01 and the highest K-factor was recorded in Group II at 2.07 ± 0.04 (Table 2). K-factor for Group III tilapia showed a decreasing trend from 2.02 ± 0.02 at week-1 and decreased gradually to 1.99 ± 0.02 at week-2. This value was further decreased to 1.95 ± 0.02 at week − 3 and reached to the lowest point at week-4 with K-factor of 1.74 ± 0.01 (P < 0.05; Table 2).

Tissue bioenergy
Overall muscle glycogen in all groups of tilapia were considerably low. As for tilapia from Group I and Group II, muscle glycogen levels were low but not signi cantly different compared to Group III (P > 0.05; Fig. 1.A). Muscle glycogen for Group I was recorded at 1.41 ± 0.20 mg/g, which was signi cantly higher compared to Group II at 0.90 ± 0.14 mg/g. Nevertheless in Group III, tilapia that adapted to RAS for the week-1 of recovery had the lowest muscle glycogen level at 0.85 ± 0.12 mg/g. However, muscle glycogen levels showed a signi cant increment at week-3 with 1.56 ± 0.20 mg/g and week-4 with 1.61 ± 0.33mg/g as compared to week-1 at 0.86 ± 0.12 mg/g (P < 0.05; Fig. 1.A). While for liver glycogen, the lowest value was recorded in Group II at 40.44 ± 2.89 mg/g and the highest liver glycogen level was recorded in Group III at week-2 at 62.17 ± 3.59 mg/g (P < 0.05; Fig. 1.B).
Both muscle and liver protein from Group I were recorded at 18.45 ± 0.39 mg/g and 30.31 ± 2.14 mg/g ( Fig. 1) respectively, which were signi cantly lower compared to Group II with muscle protein recorded at 23.71 ± 0.87 mg/g and liver protein at 53.80 ± 5.39 mg/g (P < 0.05; Fig. 1.C). Liver protein level found in Group II was recorded at 53.80 ± 5.38 mg/g considerably higher as compared to Group I and Group III from week-2 onwards (P < 0.05; Fig. 1.D).
Muscle lipid levels remained relatively stable among all groups of tilapia from Group I, Group II and Group III (P > 0.05; Fig. 1.E). Differently, lowest liver lipid was recorded in Group I and Group III at week-4 with 40.59 ± 3.44 mg/g and 39.54 ± 2.44 mg/g, respectively (P < 0.05; Fig. 2.B). Surprisingly, liver lipid level from Group II was not signi cantly different compared to Group III at week-1,-2 and − 3 (P > 0.05; Fig. 1.F).

Plasma osmolarity and electrolytes
Plasma osmolarity and electrolytes concentration was presented in Fig. 4. Plasma osmolarity for all groups were remained relatively stable with 295.54 ± 1.44 mOsm/L recorded for Group I and 291.52 ± 1.65 mOsm/L (P > 0.05; Fig. 4.A). Meanwhile, plasma osmolarity level in Group III were recorded at a range of 300.05 ± 1.78 mOsm/L to 306.28 ± 2.81 mOsm/L, which were relatively stable within 4 weeks (P > 0.05; Fig. 4.A). The lowest plasma sodium (Na + ) was found from Group II with only 135.39 ± 1.60 mmol/L compared to all other groups (P < 0.05; Fig. 4.B). Plasma Na + level for Group I and Group III for all weeks were maintained relatively stable at range of 146.95 ± 4.23 mmol/L to 156.24 ± 7.69 mmol/L (P > 0.05; Fig. 4.B). Potassium (K + ) is the second important cation for biological processes in organism. In this study found that plasma K + levels in all groups were not signi cantly differences ranging from 3.71 ± 0.78 to 4.60 ± 0.89 mmol/L (P > 0.05; Fig. 4.C). Differently for plasma chloride (Cl − ) levels in all groups, Group I exhibited lowest plasma Cl level for Group I was recorded at 128.67 ± 2.78 mmol/L and the lowest plasma Cl − was noted in Group II at 122.58 ± 1.82 mmol/L compared to Group III at week-1 and week-2, which were 143.85 ± 1.48 mmol/L and 145.05 ± 4.56 mmol/L (P < 0.05; Fig. 4.D). Similarly trend was also noticed for plasma calcium (Ca 2+ ) with the lowest plasma Ca 2+ observed at Group II at only 6.57 ± 0.29 mmol/L compared to other groups (P < 0.05; Fig. 4.E). Plasma Ca 2+ for Group I and III were remained insigni cantly different ranging from 7.58 ± 0.26 mmol/L to 8.43 ± 0.59mmol/L (P > 0.05; Fig. 4.E). Whereas for plasma magnesium (Mg 2+ ), lowest value was noticed in Group I with only 0.69 ± 0.04 mmol/L (P < 0.05; Fig. 4.F). Overall, highest plasma Mg 2+ was noticed in Group II at 1.05 ± 0.09 mmol/L, but was not signi cantly different compared to Group III ranging from 0.82 ± 0.12 mmol/L to 0.97 ± 0.07 mmol/L (P > 0.05; Fig. 4.F).
Gills and kidney enzymatic Na + /K + ATPase activity Sodium pump Na + /K + ATPase activity (NKA) for the gills and kidney of tilapia cultured under different environment were presented in Fig. 5. In general, both gills and kidney NKA activities were presented in similar pattern (Fig. 5.A and 5.B). As for gills NKA no signi cant differences noticed for all groups. Although, Group I expressed lowest NKA activities with only 3.61 ± 0.49 µmol Pi/mg protein/h compared to all other groups (P > 0.05; Fig. 5.A). High gill NKA activities was recorded in tilapia from Group II with NKA activities at 4.80 ± 0.67 µmol Pi/mg protein/h. Whereas NKA activities for Group III were recorded ranging from 3.85 ± 0.55 µmol Pi/mg protein/h to 4.54 ± 0.35 µmol Pi/mg protein/h, respectively (P > 0.05; Fig. 5.A).
Similar trend was also noticed in kidney NKA activity with no signi cant difference observed for all groups (P > 0.05; Fig. 5.B). Nevertheless, highest kidney NKA activities noticed in Group II at 2.62 ± 0.22 µmol Pi/mg protein/h, followed by Group III on week-1 and week-2 at 2.25 ± 0.31 µmol Pi/mg protein/h and 2.21 ± 0.36 µmol Pi/mg protein/h (P > 0.05; Fig. 5.B). Again similar as gills, lowest kidney NKA activities were recorded from Group III on week-4 at 1.83 ± 0.22 µmol Pi/mg protein/h, which was comparable with Group I at 1.99 ± 0.61 µmol Pi/mg protein/h, respectively.

Growth and energy mobilisation pattern of tilapia living in different environments
The K-factor of a sh re ects physical and biological quality of a species or individual of shes (Binner et al. 2008). Through its variations, information on the physiological state of the sh in relation to its welfare can be recognized . K-factor is also an indicator of the general sh condition because condition factor re ecting interactions between biotic and abiotic factors in the physiological condition of shes. Moreover, body condition provides an alternative to the expensive in vitro proximate analyses of tissues (Sutton et al. 2000). Therefore, information of K-factor can be vital to culture system management because they provide the producer with information of the speci c condition under which organisms are developing (Araneda et al. 2008). The values of K-factor recorded in the present study were 1.54 ± 0.01 for Group I, 2.07 ± 0.04 for Group II and range from 1.74 ± 0.01 to 2.02 ± 0.02 for Group III respectively. Condition factor of greater than one showed the well-being of shes (Datta et al. 2013). Although all shes were in a good condition, low K-factor in Group I and Group III at week-4 could be related to reproduction pheromone when more energy were channelled to reproduction. Whereas, sh in Group II were cultured in cage culture system. This may head to unsuitability for reproduction. Thereby energy intake was speculated to be used for growth. Similarly for tilapia in Group III at week-1 and − 3. The K-factor decreases at the start of the spawning period due to high metabolic rates as noticed with increased of GSI. There is normally a gradual increase in the condition factor during the reproductive period and normalization occurs immediately afterwards Gupta et al. 2011). Energy that is surplus to the essential standard metabolic requirements (i.e., maintenance, locomotion, predation avoidance, and feeding activity) is allocated to somatic growth, energy storage, or reproduction after the sh reaches sexual maturation. The priority with which this surplus energy is allocated to each of the above biological functions differs among sh species (Nunes et. al. 2011). Fish is under the indeterminate growth as must consider the survival costs and the available energy for reproduction and must make an allocation decision between current and future reproduction as an adaptation to the uctuating environmental conditions.
Hepatosomatic Index (HSI) is de ned as the ratio of liver mass to body weight, where HSI is commonly use as reference to de ne status of feeding and nutrition intake (Facey et al. 1999) of an organism with energy storage for growth and reproduction (Nunes et al., 2011;Sadekarpawar & Parikh, 2013). As data obtained in this study, lower HSI was observed in Group II which could be attributed to the important role of the liver in nutrient metabolism, which was in associated with high activity and metabolic rates. As tilapia from Group II were actively swimming in circle that led to high energy expenditure for exercise while maintained e cient aerobic metabolism. Thereby, accelerates energy mobilization thus reduced HSI value. Differently for tilapia in Group I and Group III, HSI values relatively higher compared to tilapia from Group II. This is because tilapia from Group I and III were cultured in the RAS system with low exercise capacity, thus reserves energy for secondary maturation and reproduction. Evidently, this was noticed in Group III tilapia with signi cant HSI increment till week-3 and started reduces on week-4, which was believed in associated with energy mobilisation for reproduction. HSI is used as a good indicator of total energy reserves on Atlantic cod Gadus morhua, which also correlated positively with sh K-condition (Lambert and Dutil, 1997).
The HSI re ected total energy reserved in the liver of tilapia, this was obvious based on bioenergy mobilisation pattern of tilapia in Group I where both liver protein and lipid were highly mobilized to support reproduction process, while reserving glycogen for routine metabolism needs. This was supported with spawning process and mouth incubation occurred during the study period. Similar energy mobilisation pattern was noticed in Group III, where tilapia from cage culture system transferred back to RAS showed signi cant energy reserved for the rst two weeks acclimation. Thereafter, week-3 onwards liver protein and lipid were mobilized signi cantly which was associated with reproduction process. At week-4, liver protein and lipid reached to the similar level as tilapia from Group I with mouthbrooding behaviour noticed.
Together HSI indication and energy mobilisation pattern showing a contribution to support GSI development as an indication of readiness for spawning process especially tilapia from Group III with an increasing GSI noticed. This also in agreement with our hypothesis that tilapia cultured in RAS condition (Group I and Group III) more suitable for spawning as bottom base of the tank provided spawning ground as territory site and allow female tilapia to collect fertilized eggs after spawning. As compared to tilapia that culture in cage system (Group II) without bottom base which is di cult to perform fertilization and eggs collection processes, therefore lower GSI was obtained in Group II.
Glycogen is one of the important energy sources to maintain basal metabolism (Javed & Usmani 2015) and serves as readily energy supply to fuel metabolic needs under unfavourable environment challenges (Mehjbeen & Nazura 2015). This corresponded to the tilapia from Group II having the lowest liver glycogen, which was believed to be related to their exercise capacity in cage. On the other hand, muscle glycogen reached to the lowest level on week-2 and restored to higher level on week-3 onwards in Group III. Possibility rst two weeks tilapia from Group III could be related with acclimation process from cage to RAS conditions. Meanwhile, liver glycogen reached to peak level was recorded at week-2 and returned to level similar as Group I and Group III at week-1, -3 and − 4, although the liver glycogen levels were not signi cantly different. This shows that tilapia reserved liver glycogen as readily energy to support exercise activity when required and this indicated that all nutrients intake were su cient and able to deposit in the liver.
In sh, protein is known to be more e ciently catabolized into energy sources to support aerobic exercise performance as compared to lipids and glycogen (Ballantyne, 2001;Liew et al., 2012;Rahmah et al., 2020). As Group II tilapia cultured in oating cage was not suitable for spawning, therefore protein was reserved for somatogenesis as higher muscle and liver protein were recorded. Whereas, protein level for tilapia from Group I and III were displayed relatively similar level. Protein was mobilized to support reproduction process, where mating and spawning process were noticed (FAO 2021b). This was in agreement with previous studies reported that at mature age, sh mobilized protein to prioritise gonadosomatic development (Van Dijk et al. 2005;Encina & Granado-Lorencio 1997;Santos et al. 2010).
Protein is known as a central role in production that allows the sh to reallocate energy used for growth to other metabolic needs at different life stage based on priority (Ferrari et al., 2011;Moyson et al. 2015) as well as to improve adaptability performance in different environment changes (Cara et al. 2007).
On the other hand, lipid mobilisation were not noti ed in muscle lipid, but liver lipid mobilisation were distinguished in tilapia from Group I and Group II week-4. Signi cantly liver lipid mobilisation is believed not only to support basal metabolism and somatogenesis, but also to support secondary maturation for reproduction (Kolditz et al. 2008). Energy requirements for gonad maturation appeared to come from liver reserves and it is noted that 1g of lipid contains 2 times higher energy than 1g of carbohydrates or 1g of protein (Jobling et al. 1998;Robb et al. 2002;Wood et al. 2003).

Ionoregulation of tilapia living in different environments
Higher plasma osmolality was recorded in Group II tilapia indicated that tilapia increased their osmolality to facilitate active ion uptake as well as enhance metabolites such as glucose and/or glycogen for routine and active metabolic activities. Facilitating active ion uptake was in parallel with high NKA activity found in both gill and kidney. According to Morgan et al. (1997), maintaining or increasing plasma osmolality is important to conserve stable ionic concentration in body uid with support of active osmoregulation. High NKA activities in Group II tilapia was also believed to relate with swimming activity as highlighted in gold sh and common carp when forced to swim actively, signi cantly accelerated the gill NKA actively . Contradictory, lower plasma Na + was noticed in Group II tilapia, while plasma Na + in tilapia from Group I and III were relatively stable. Although higher NKA activities in gills and kidney were noticed in Group II, this does not retain Na + level e ciently. Loss of Na + might occur in Group II tilapia could be in associated with the living condition in the lake water which is relatively clean and low ionic levels compared to tilapia that lives in the RAS water. In Group III, gills and kidney NKA activities were noticed relatively stable as for tilapia in Group I.
On the other hand, plasma K + levels were relatively stable, except Group I tilapia that had slightly higher K + level. High plasma K + could be released from the tissue into the blood stream in cooperate NKA activity not only to facilitate Na + uptake but also helped in ammonia excretion . Another possible explanation could be correlated with defensive behaviour of Group I tilapia due to active swimming and defending territory for mating and spawning. Active exercise resulted in tissue K + leaked into body uid was reported in common carp (Knudsen and Jensen, 1997). Na + , K + and Cl − are important ions which provide the sustainability of the osmotic pressure of body uids as well as acid-base balance (Keleştemur 2012;Karnaky 1998). As one of the essential ions, plasma Cl-level was maintained consistently in Group III, but lightly decreased in Group I and II tilapia. Stability of plasma Cl-is probably correlated with an increase of Na + uptake via the Na + /Cl − exchanger via dietary intake .
Feeding is known to provide excessive base which consequently led to the uptake of Cl − via branchial Cl − /HCO 3 exchanger as reported in rainbow trout (Bucking and Wood, 2006;. As important ions, the association of unidirectional in ux and e ux of Na + and Cl − (Perry and Fryer, 1997) to be adjusted to a net ux via Cl − uptake at gills Cl − /HCO 3 − exchanger was noted during alkalosis metabolism (Perry and Goss, 1994).
As all shes in all groups were fed twice daily, dietary Ca 2+ uptake seemed to su ciently support basal metabolic needs. Higher plasma Ca 2+ levels in Group II and Group III seemed to be a strategy to retain Ca 2+ in the body active uptake through Ca 2+ transporter and Ca 2+ channel. Ca 2+ is known to play important role in bone and scale formation. Therefore, it is highly essential for tilapia from Group II and III to maintain su cient level of Ca 2+ to support basal metabolic needs. Differently from Group I that lived in enclosed system at all time with consistent water Ca 2+ level as well as received su cient Ca 2+ from dietary supply, which allowed tilapia from this group to maintain Ca 2+ for basal metabolism. Mg 2+ is the second most abundant cation that exists in intracellular uid that acts as a functional co-factor for enzymes as well as play an important role in neurochemical impulse transmission and muscle excitability (Keleştemur 2012). Changes in plasma Mg 2+ level is always associated with stress or environmental changes ). This phenomenon was noticed in this study where plasma Mg 2+ levels were inconsistent where higher Mg 2+ level found in Group II tilapia, lower in Group I and uctuate in Group III. It was believed that the different levels of ions was in uenced by the living conditions which could be associated with territory competition for mating and water ionic status especially in Group II condition.

Conclusion
The present study revealed that tilapia prioritize energy mobilization differently for growth and reproduction according to living conditions. As tilapia living in open cage mobilized glycogen to fuel swimming activity and prioritized energy for growth performance. Whilst, tilapia living under enclosed condition mobilized lipid and protein to prioritise reproduction purpose. This remark supported our hypothesis highlighted that energy mobilization was con rmed for tilapia living in enclosed condition.
Whereas, our second hypothesis on the gills and kidney NKA activities was rejected. The gills and kidney NKA activities in all groups of tilapia remained steady to support balance homeostasis for basal metabolism, without in uenced by living conditions. The ionic concentrations of sodium (Na + ), potassium (K + ), chloride (Cl -), calcium (Ca 2+ ) and magnesium (Mg 2+ ) in water sources from Group I -(RAS) hatchery recirculating aquaculture system tilapia culture system, Group II -(Cage) water sample from Como River cage culture water and Group III -(Compensation) RAS system from compensation. Different superscript letters in the same row indicate signi cant differences of ionic concentration amongst different types of water (P < 0.05).

Declarations
Results are presented as mean ± standard error mean (mean ± SEM). The biometric data of hybrid red tilapia Oreochromis sp. from Group I (RAS), Group II (Cage) and Group III (RAS-Compensation) which include nal body weight, nal body length, HIS, GSI and K-Factor. Different lowercase letters indicate signi cant differences among groups (RAS, Cage and RAScompensation). Results are presented as mean ± standard error mean (mean±SEM, P<0.05, n = 10). Figure 1 Total energy of (A) muscle glycogen (B) liver glycogen (C) muscle protein (D) liver protein (E) muscle lipid and (F) liver lipid levels of hybrid red tilapia Oreochromis sp. rom Group I -RAS (white bar), Group II -Cage (black bar) and Group III -Compensation (grey bars) for week-1, week-2, week-3 and week-4. All values are means ± standard error of the mean (SEM) (n=10). Superscript small letters indicates signi cant differences amongst cultured in different groups (P<0.05).