Introduction
Protein is the most important fish dietary nutrient, as it provides essential and non-essential amino acids for growth and energy production.1 However, a given diet with excess protein and low energy leads to ammonia production through amino acid catabolization, a metabolic process that can be harmful to fish.2 Moreover, protein is more expensive than other macro-nutrients in formulated feeds. Therefore, excess dietary protein leads to higher feed costs that can reduce the profits from fish farming. It is well known that the dietary protein requirement of fish depends on dietary energy.3 A diet containing an optimal combination of protein and non-protein energy-supplying nutrients is required for both the growth of cultured fish and the economic efficiency of fish farming. An effect referred to as the protein-sparing activity of carbohydrates and lipids, is commonly recognized.4–11 Lipids can more effectively conserve protein for development than carbohydrates due to their higher energy content, and lipids are readily metabolized by fish.12 In addition to being rich in energy, the inclusion of lipid sources, particularly fish oils, in diets also provides essential fatty acids for the normal performance and health of fish. Excess dietary lipids may cause body fat deposition and deterioration of muscle quality in cultured fish.13,14 Hence, to promote optimal growth without adverse effects such as excessive nutrient loss or low meat quality, the fish diet should comprise an optimal ratio of protein and lipids. So far, a balanced protein‒to‒lipid ratio in diets in order to increase growth, improve flesh quality, and reduce environmental waste has been studied for many cultured fish species.15–19
Mastacembelus eels (striped spiny eel M. pancalus, tire-track spiny eel or zig-zag eel M. armatus, tire track eel or flower spiny eel M. favus) are considered highly valued fish in Asia where they are naturally distributed (Bangladesh, India, Pakistan, Nepal, Thailand, Lao PDR, Cambodia, Vietnam and Peninsular Malaysia). They have a good taste, delicious flesh quality (special flavor, characteristic texture, and high protein, oil, and vitamin C content), high market value, lucrative size, and important production potentials for food and aquarium fish.20–24 The importance of fish farming is underscored by the decline of wild fish stocks, driven by habitat degradation and overexploitation.21 In Vietnam, M. favus (known as chach lau or chach bong) is abundant in the Mekong River Delta (MRD) and was previously misidentified as Mastacembelus armatus, a species found only in countries such as India and Bangladesh.25 M. favus is a carnivorous fish.26 Artificial seed production of M. favus24 and the separate dietary protein and lipid requirements of this species have been studied in the Mekong River Delta of Vietnam.27–29 M. favus is being farmed in the MRD and fed low-value fish and/or homemade feed.30,31 Nguyen et al.32 also studied replacement levels of black soldier fly (Hermetia illucens) larvae meal for fishmeal in the diets of tire track eel fingerlings. To our knowledge, there are no previous studies assessing dietary protein‒to‒lipid ratios on the growth performance of M. favus at any development stages. Thus, to promote the farming of this species in the country, it is necessary to produce formulated feeds with an optimal protein-to-lipid ratio for cost-effective production.
Materials and methods
Study location and duration
The experiment was conducted in the Experimental farm of An Giang University, Long Xuyen City, An Giang Province in the Mekong River Delta of Vietnam from February to May 2021.
Experimental fish and holding conditions
Fingerlings of M. favus with a body mass of 2-3 g each were bought from a local hatchery and transported to the wet lab of An Giang University (a distance of 5 km). The fish were acclimatized in three fiberglass tanks (2,000 L each) and fed commercial pellets for two weeks. The tanks were continuously aerated, and about 30% of the tank water was exchanged every two days to maintain optimal environmental conditions (dissolved oxygen over 5 mg/L, a pH of 7.5-8.0, and a temperature of 27-29⁰C).
Healthy fish with a body weight of 3-5 g were randomly distributed into 27 fiberglass experimental tanks (500 L each) at a stocking density of 50 fish per tank. The average sizes of fish by treatments were from 2.87 ± 0.92 to 3.37 ± 0.28 g/fish. Water flowed through the experimental tanks at a rate of 0.5-0.7 L/min. The tanks were continuously aerated during the experimental period. Four PVC pipes 9.0 cm in diameter and 30 cm in length were placed at the bottom of the tanks to create shelters for the fish. The water quality parameters were measured at 06:00 and 15:00 hours daily using portable DO and pH meters (Hanna Company, Woonsocket, RI, USA).
Experimental design and diets
Due to a lack of available information on the dietary protein and lipid requirements of M. favus. The protein and lipid levels in this study were referred from the research of Hung et al.3 on Asian red-tailed catfish (H. wyckioides), an omnivore, with an adjustment of higher protein contents. The experiment was conducted using a 3 × 3 factorial design with three dietary protein levels (40, 45, and 50%) and three lipid levels (6, 9, and 12%). Each treatment was performed in triplicate simultaneously. The diets were formulated using fish meal, soybean meal, fish oil, cassava flour, bone meal, and premix (Tables 1 and 2). All ingredients were thoroughly mixed and then pelleted by using an electronic meat grinder with a diameter of 1 mm. Diets were dried in an oven at 60°C until the moisture content was less than 10%, and the meal was broken into pellets with a length of 1-1.2 mm. All diets were stored at 5°C in sealed plastic bags until use. The experiment was carried out in accordance with the national guidelines on the protection of animals and experimental animal welfare in Vietnam.33
Feeding management
The fish were manually fed the test diets twice a day from 08:00 to 09:00 and from 15:00 to 16:00 using a feeding rate of about 10% body weight. This species is a bottom feeder; therefore, plastic tubing was used to position the pellets on the bottom of the tanks (close to the shelters). The feed intake of the fish was closely monitored. Uneaten feed was collected, dried, and weighed to estimate actual feed intake. The feeding rate was adjusted depending on the feed amount consumed on the previous day. In addition, the feed intake of the fish was monitored.
Sample collection and analysis
Thirty fish from each tank were individually measured for length and weight at the start (day 1) and end (day 56) of the experiment. A ruler and a digital balance with a minimum precision of 1 mm and 0.1 g, respectively, were used for the measurements.
Three fish at stocking and three fish from each experimental tank (nine fish per treatment) at the harvest were collected for carcass composition analysis. Feed ingredients, experimental diets, and fish samples were preserved at –20⁰C until analysis. The moisture, ash, protein, fat, and carbohydrate (NFE) contents of feed and fish were determined using standard methods.34 Lysine and methionine were calculated based on their values in the ingredients.13 Gross energy (GE) was calculated using values of 5.65 Kcal/g for crude protein, 4.25 Kcal/g for carbohydrates, and 9.45 Kcal/g for crude fat.12
Data analysis
The initial mean weight (Wi) and final mean weight (Wf) of individual fish were determined before and after the experiment. The survival rate (SR), weight gain (WG), daily weight gain (DWG), specific growth rate (SGR), feed intake (FI), feed conversion ratio (FCR), and protein efficiency ratio (PER) were computed as follows:
Survival rate (SR, %) = (number of fish at harvest)/(number of fish at stock) × 100
Length gain (LG, cm) = Lf – Li
Weight gain (WG, g) = Wf – Wi
Daily weight gain (DWG, g/day) = (Wf – Wi)/t
Specific growth rate (SGR, %/day) = [(ln Wf – ln Wi) × 100]/t
Feed conversion ratio (FCR) = dry diet fed/wet weight gain)
Protein efficient ratio (PER) = (Wf – Wi)/protein intake
Daily feed intake (DFI, g/fish/day) = Consumed feed in tank (g)/(number of fish × t)
Where Wi is the initial weight of the fish (g); Wf is the final weight of the fish (g); Lf is the final length of fish; Li is the initial length of fish, and t is the experimental time (day).
Statistical analysis
Treatment means were compared by a two-way ANOVA followed by Duncan’s multiple range test using Minitab software version 16.0. The significance level was set at P < 0.05. Data are presented as mean ± standard deviation. For the protein and lipid requirements, after considering the interaction of the two factors, the difference in each diet was examined to select the best treatment.
Results
Growth and survival of fish
The growth performance of the fish fed nine test diets is presented in Tables 3 and 4. After 56 days, the average final body length (FBL) and length gain (LG) ranged from 12.9 to 14.4 cm and 2.32 to 4.27 cm, respectively (Table 3). There was no significant difference in the FBL or LG of the fish among the treatments. For the nutrient group analysis, the FBL (13.9 cm) and LG (3.49 cm) of the 45% protein diet group were the highest, and those (13.2 cm and 2.72 cm, respectively) of the 40% protein diet group were the lowest. Similarly, the averages of FBL (13.9 cm) and LG (3.54 cm) of the 9% lipid diet group were the highest, and those (13.1 cm and 2.59 cm, respectively) of the 6% lipid diet group were the lowest. The LG of the fish was affected by the dietary lipid content (P < 0.05) but not by the dietary protein content (P >0.05). There was no interaction effect between dietary protein and lipid nutrients on the growth parameters of the fish (P > 0.05). The diet of 45% protein and 9% lipid obtained the best growth of the fish (Table 3).
There were significant interaction effects between dietary protein and lipid nutrients on final body weight (FBW) and weight gain (WG) (P < 0.05), but not on daily weight gain (DWG) or specific growth rate (SGR) (P > 0.05) (Table 4). The FBW ranged from 5.48 to 8.44 g/fish, WG from 2.36 to 5.38 g, DWG from 0.16 to 0.43 g/day, and SGR from 3.70 to 6.80 %/day (Table 4). There was a significant difference (P < 0.05) between the means of FBW, WG, and DWG, but not among those of SGR of the fish (P > 0.05). Based on nutrient group analysis, the average FBW (7.05 g), WG (4.08 g), DWG (0.31 g/day), and SGR (5.75 %/day) of the 9% lipid diets were the highest and those (5.97 g, 2.8 g, 0.19 g/day and 4.36 %/day, respectively) of the 6% lipid group were the lowest. Similarly, at the end of the experiment, the average BW (6.86 g), WG (3.74 g), and DWG (0.3 g/day) of the 45% protein diets and the SGR (5.58 %/day) of 50% protein diets were the highest, and those (6.19 g, 3.18 g, 0.19 g/day, and 4.5 %/day, respectively) of the 40% protein diets were the lowest. The WG and DWG were affected by the dietary protein but not by the dietary lipid (P > 0.05). The diet of 45% protein and 9% lipid showed the best weight gain (Table 4).
The survival rates (SR) ranged from 71.3 to 82.7%. There was no significant difference among the average SRs of fish fed different test diets (P > 0.05). However, the SR of fish fed 45% protein diets was the highest (78.5%) compared to fish in the 40% and 50% protein diet groups (75.1%). The increase in lipid content of the diets from 6% to 9% and 12% resulted in a decline in the SRs of the fish (77.6, 76.4, and 74.7%, respectively). There was no interaction effect between dietary protein and lipid nutrients on the SR of the fish (P >0.05) (Table 3).
Feed utilization efficiency
The average daily feed intake (DFI) ranged from 0.21 to 0.30 g/fish/day. The feed conversion ratio (FCR) and protein efficiency ratio (PER) were 1.91 to 4.07 and 0.81 to 1.70, These parameters were not separately affected by dietary protein or lipid levels except for DFI by dietary protein (P < 0.05). There was a significant interaction of the two dietary nutrients on the FCR but not on the DFI or PER. The DFI value of the diet of 45% protein and 6% lipid resulted in the lowest efficiency, and the diet of 50% protein and 9% lipid had the highest efficiency. The FCR of the diet of 40% protein and 9% lipid was the highest, and that of the diet of 45% protein and 9% lipid was the lowest (Table 5).
For the nutrient group analysis, the average DFI of the 50% protein diets was the highest (0.28 g/fish/day) compared to others (0.24 g/fish/day), and the average DFI of the 9% lipid diets was the highest (0.26 g/fish/day) followed by the 12% and 6% lipid diets (0.25 and 0.24 g/fish/day, respectively). The average FCR of the 40% protein diets was the highest (3.24), followed by the 50% and 45% protein levels (2.99 and 2.75, respectively). The increase in lipid level of the diets from 6% to 12% resulted in the decline of the FCR (3.49, 2.80, and 2.70, respectively). The average PER of the 50% protein diets was the highest (1.49), followed by the 40% and 45% protein diets (1.26 and 1.24, respectively). The increase in the lipid level in the diets also resulted in a decline in the average PER (1.57, 1.22, and 1.20, respectively). There was no interaction effect between dietary protein and lipid on the DFI and PER (P > 0.05), but this effect was documented in the FCR (P < 0.05) (Table 5).
Chemical composition of fish
The proximate composition of the whole body of the fish fed the nine test diets is presented in Table 6. There were significant differences between means in the moisture and lipid contents (P < 0.05) but not between those of the organic matter and protein contents. The proximate composition was separately affected by lipid levels, except for the organic matter content, and was unaffected by the protein levels, except for the protein content (P > 0.05). There were significant interactive effects of the two dietary nutrients on the moisture and lipid contents, but not on organic matter or protein (P > 0.05).
For the nutrient group analysis, the average protein content of the 45% protein diets was the highest (25.4%) (P < 0.05). The increase in the lipid level in the diets resulted in a decline in the average protein content and an increase in the average lipid content. There were no significant interaction effects between dietary protein and lipid nutrients on the organic matter or protein contents, but there was a significant interaction in the moisture and lipid contents (P < 0.05) (Table 6).
Discussion
The water quality parameters of the experimental tanks by treatments were 5.23 ± 0.24 to 5.62 ± 0.28 mg/L for dissolved oxygen, 27.2 ± 0.79 to 28.5 ± 1.97 oC for temperature, and 8.36 ± 1.12 to 8.47 ± 0.14 for pH. These parameters were suitable for the normal development of the fish.35–37
Nguyen et al.28 found that a dietary content of 44% protein stimulated the best growth and the lowest FCR of M. favus fingerlings; they proposed that the value of 45.2% protein in the feed resulted in optimal growth of 2.7 g fish. Nguyen et al.31 also found that a dietary content of 7.5% lipid resulted in the maximum growth of M. favus fingerlings. Therefore, in the experiment involving the combination of three protein levels and three dietary lipid levels, it was reasonable to detect the protein-sparing effect of lipids in M. favus fingerlings. In this study, there was no interaction between dietary protein and lipid nutrients on the growth performance of the fish (Tables 3 and 4). This result was similar to the findings of Hung et al.3 on red-tail catfish (Hemibagrus wyckioides), but different from that of Hien et al.38 on clown knife fish (Chitala chitala) fed diets with a combination of different protein and lipid levels. The protein–sparing effect of lipids in fish has been demonstrated by many authors such as Orire & Sadiku8 in Nile tilapia (Oreochromis niloticus), Fan et al.10 in common carp (Cyprinus carpio), Hung et al.3 in H. wyckioides, and Welengane et al.39 in hybrid fish tambatiga (♀Colossoma macropomum × ♂Piaractus brachypomus). However, Mohammadi et al.40 found that growth indices were reduced by lipid contents increasing from 5 to 13% in combination with dietary protein levels of 15, 22, 29, and 36%, and concluded that the protein–sparing effect of lipids did not occur in O. niloticus. Hien et al.38 found that the growth of the fish fed a dietary lipid level of 9% was the highest compared to the lipid levels of 6 and 12% at different dietary protein levels of 25, 30, 35 and 40%, and also concluded that the protein-sparing effect of lipids was not present in C. chitala. In this study, at low and high dietary protein levels (40 and 50%), the increase in lipid levels from 6 to 12% had no significant impact on length and weight gain. In contrast, at the dietary protein level of 45%, increasing lipid levels from 6 to 9% significantly improved the growth performance of fish, and a higher lipid level (12%) had no significant effect on fish growth (Tables 3 and 4). This implied that the protein-sparing effect of lipids occurred in M. favus fingerlings and the effect was obvious with dietary protein levels close to the optimum level (45%), as was found by Nguyen et al.28 Orire & Sadiku8 found that the growth of O. niloticus fed Palm oil based diets containing three levels of protein (P) and three levels of lipid (L) ratios (15P:25L, 10P:30L and 5P: 35L%) declined when the lipid level was lowered to 5% and crude protein level increased to 35%. To detect the protein-sparing effect of lipids in M. favus, more studies need to be carried out with wider ranges of dietary protein and lipid content, and different fish sizes.
The correlations between changes in DFI and the dietary inclusion ratios of protein and lipids in fish fed diets with varying levels of these nutrients have shown inconsistent patterns. There were no correlations between DFI values and dietary protein or lipid levels in H. wyckioides.3 The DFI of snakehead fish (Channa striata) was not different from the combined protein and lipid diets.41 Total feed intake and DFI were reduced with the increase of dietary lipid levels at all protein inclusion ratios in O. niloticus38 and tambatinga (♀Colossoma macropomum × ♂Piaractus brachypomus).37 Welengane et al.39 also found that increased protein levels lower the DFI of ♀C. macropomum × ♂P. brachypomus juveniles. In the present work, the average of DFI increased with increased dietary protein levels and was not correlated with lipid levels. This result was similar to the findings of Hung et al.,3 but different from the findings on ♀C. macropomum × ♂P. brachypomus39 and O. niloticus.40 The correlation between DFI and dietary protein and lipid nutrients depends on the feeding habits of fish. There were no correlations in studies of carnivorous fish, e.g., H. wyckioides,3 C. striata,41 and M. favus, while such correlations were observed in omnivorous fish, e.g. O. niloticus40 and ♀C. macropomum × ♂P. brachypomus.39
In species showing the protein-sparing effect of lipids, such as Nile perch (Lates niloticus), increasing dietary lipid levels improved the FCR of the fish fed with iso-protein level diets.42 In O. niloticus, the increment of lipid to the highest level of 15% resulted in improved growth performance for all of the lipid sources of groundnut, palm, and fish oil; the palm oil-based diets had the highest level of growth and feed utilization, thereby yielding the lowest FCR.41 With two dietary levels of protein (39 and 44%) and three dietary levels of lipid (6, 9, and 12%), Hung et al.3 found that increasing dietary protein significantly improved FCR, but increasing dietary lipid significantly reduced the FCR of H. wyckioides. However, in species showing no protein–sparing effect of lipids such as C. chitala, in an experiment with four dietary levels of protein (25, 30, 35, and 40%) and three dietary levels of lipid (6, 9, and 12%), the FCR was reduced and reached the lowest values at the levels of 35% protein and 9% lipid.38 Moreover, the FCR of C. striata fed a combined protein and lipid diet showed no differences at all levels of the two nutrients.41 In this work, with three dietary levels of protein (40, 45, and 50%) and lipid (6, 9, and 12%), the FCR values of M. favus were reduced with the increase in lipid levels. A similar result was found in H. wyckioides,3 and FCR was the lowest at 45% protein in C. chitala.38
In the present study, the PER of M. favus fingerlings increased with the dietary protein level and declined with an increase in dietary lipid level. The increase in lipid and decrease in protein levels increased the PER of red-tail catfish.3 The decrease in PER with the increase in dietary lipid levels was found in O. niloticus40 and C. chitala.38 Therefore, the difference in the correlation of PER, as well as of FCR, and dietary protein and lipid levels between this study and others could be due to different experimental conditions, such as species, size, age, nutrient quality, environmental parameters, or other unknown factors.
This work has demonstrated that the body protein contents of M. favus were higher than those of H. wyckioides,3 the hybrid of ♀C. macropomum × ♂P. brachypomus,39 the Nile tilapia O. niloticus40 and clown knife fish,38 but lower than those of L. niloticus.42 The body lipid contents of the fish were lower than those of others.3,36–38,42 An increase in the dietary protein did not alter the carcass of M. favus in terms of moisture, organic matter, or lipid content. This effect was similar to that of H. wyckioides.3 The tendency of decrease in body protein contents with increasing dietary lipid levels of the present study was different from C. chitala,38 but the tendency of increase in body lipid contents was in agreement with patterns observed in L. niloticus,42 the hybrid of ♀C. macropomum × ♂P. brachypomus,39 and O. niloticus.40 The impact of dietary protein and lipid levels on the proximate composition of fish bodies tends to vary across species. However, a common trend observed in fish is that higher dietary lipid levels often lead to increased body lipid deposition.
The survival rate of M. favus eels in the present work was not affected by the experimental diets. Similar results were demonstrated for H. wyckioides,3 C. carpio,10 and C. chitala.38
Conclusions
Based on the growth performances and FCR, the dietary requirements of 3-5 g M. favus fingerlings are 45% protein and 9% lipid, corresponding to an energy level of 3,819 Kcal. kg-1 and a P/E of 118 mg. Kcal-1.
Acknowledgments
This work was supported by the “Research on nutritional needs and formula for producing industrial feed for Tire Track Eel (Mastacembelus favus)” sponsored by the Department of Science and Technology of An Giang province, Vietnam (NO.373.2022.07).
Authors’ Contribution - CRediT
Conceptualization: Phan P. Loan (Lead). Methodology: Phan P. Loan (Equal), Nguyen Thanh Phuong (Equal). Investigation: Phan P. Loan (Lead). Resources: Phan P. Loan (Equal), Nguyen Thanh Phuong (Equal). Writing – original draft: Phan P. Loan (Lead). Supervision: Nguyen Thanh Phuong (Lead). Writing – review & editing: Nguyen Thanh Phuong (Lead).
Competing of Interest – COPE
No competing interests were disclosed.
Ethical Conduct Approval – IACUC
Tire Track Eel (Mastacembelus favus) is not considered endangered or protected, so experiments with this species in Vietnam do not require special permission. The animal experiment received approval from the Scientific and Training Council of the Faculty of Agriculture and Natural Resources of An Giang University (Approval No. 29A-08092023)
Informed Consent Statement
All authors and institutions have confirmed this manuscript for publication.
Data Availability Statement
The data that has been used is confidential.