Dietary protein requirements of tire track eel (Mastacembelus favus Hora, 1923) fingerling

Nguyen Huu Yen Nhi, Tran Kim Ngoc, Trinh Thi Lan, Nguyen Thi Thuy Hang, Nguyen Thi Bich Hanh, Dang Van Ty¹, Torbjörn Lundh² and Anders Kiessling²

An Giang University, Vietnam National University Ho Chi Minh City, Viet Nam
nhynhi@agu.edu.vn
¹ Mekong Aquafeed Company Limited, Phu Cuong, Tam Nong, Dong Thap, Viet Nam
² Department of Animal Nutrition and Management, Swedish University of Agricultural Sciences, 75007 Uppsala, Sweden

Abstract

Tire track eel Mastacembus favus is an important candidate species for commercial aquaculture in Vietnam, therefore, a ninety day-feeding experiment was carried out to determine appropriate protein levels for efficiently rearing tire track eel fingerlings to complete the procedure of raising this species by using commercial feed. The fish were fed one of seven diets formulated with different levels of protein (24%, 29%, 34%, 39%, 44%, 49% and 54%) at the same energy and fat levels. Triplicate groups of 50 tire track eel fingerlings (initial weight ranged 2.67 - 2.7 g) were randomly assigned to 21 identical conical 500-L tanks equipped with an aeration system and substrates for shelter. Feed intake (FI), protein intake (PI), daily weight gain (DWG), specific growth rate (SGR), feed conversion ratio (FCR), feed conversion efficiency (FCE), protein efficiency (PE), survival rate (SR), net protein utilisation (NPU) and body indices were determined at the end of the experiment. The highest growth rate of fish fed different levels of protein was obtained in group fish fed 44% protein. DWG of this group (0.093 g/day) was higher than those of the groups (24%, 29%, 34% and 54%, p < 0.05), but there was not different from the remaining groups (39% and 49%, p > 0.05). SGR of fish in group fed the 44% protein (1.59%/day) was similar to those of the groups (39% - 54%) but was higher than the remaining groups (24% - 34%). The optimal protein requirement of tire track eel was determined based on the quadratic regression method of Zeitoun et al (1976) with the equation y = - 0.0014 x² + 0.1265 x – 1.5351 (R ² = 0.8355). From this equation, the value x-max = 45.2 is the protein content for the fish to achieve maximum growth. After ninety-day treatment period, SR of tire track eel were relatively high ranging from 76.7% to 84.7% as well as body indices was no difference between treatment groups (p > 0.05). FCR was lowest in the 44% protein treatment group (1.43). This study revealed that 44% protein content stimulated the best growth and the lowest FCR as well as the value of 45.2% protein in the feed allowed the fish grow optimally.

Keywords: protein requirement, tire track eels, optimum growth, Mastacembelus

Introduction

Tire track eel is distributed in Asian countries (e.g., India, Lao, Pakistan, Indonesia, Sri Lanka, Thailand, China, Malaysia) and in Southeast Asia, it is commonly present in the basins of the Mekong Rivers (Rainboth 1996; Petsut and Kulabtong 2015; Ahmad et al 2018). In Vietnam, it is found in almost all regions of the country where the water flows lightly, e.g., crevices, rock embankments, bridge feet (Khoa and Huong 1993).

Tire track eel is a popular table fish owing to its delicious taste and high nutritional value (Gupta and Banerjee 2016) so it is one of the most economically important species and is now being farm-raised in Vietnam on a small scale to support local demand (Jamaluddin et al 2019). However, there is almost no research on the requirement of essential nutrients in food for tire track eel, but only a few studies on population genetics (Jamsari et al 2014) with genetic structural variants in Southeast Asia (Jamaluddin et al 2019) isolated and identified. The research results on the use of feed also show that during rearing, this species prefer to use live feed (e.g., Moina and worms), especially worms for higher growth rate and survival rate than other feeds (Loan et al 2010; Trung et al 2010; Lenh 2010). However, live food is often scarce, difficult to produce and not active in farming households, so it greatly affects the development process of this fish. To develop sustainable fish farming and meet the requirements of fish farming industry about quality demand and traceability, research on nutritional requirements of tire track eels (Mastacembelus favus Hora, 1923) is essential. Therefore, this study was conducted to determine appropriate protein contents for effectively rearing tire track eel fingerling, contributing to complete the procedure of raising this species using commercial feed.

Materials and methods

Experimental animal and conditions

This experiment was carried out on an experimental farm in An Giang University, Long Xuyen city, An Giang province, Vietnam. Tire track eel fingerlings were obtained from a local hatchery in Thoai Son, An Giang province, Vietnam. They were transported to the experimental unit in 0.5 m ³ plastic containers with proper aeration. All the fingerlings were given a prophylactic dip in a solution of 3% NaCl for 5 minutes and were then acclimatized to the laboratory conditions for two weeks prior to the experiment, during which they were fed commercial feed with 42% crude protein level for snakehead fish (Uni-President Vietnam Co.,LTD, Vietnam). The fish were selected to have a relative uniform in size with initial weight ranged 2.6 g/fish and initial length ranged 10 cm/fish.

Photo 1. Tire track eel

Each of the 21 experimental tanks was aerated via one air stone connected to a low-pressure electrical blower (Resun GF-370, China). Each of the 21 experimental tanks was equipped with substrates as plastic cluster for shelter. Experiment tanks were constructed by Quang Đat Company, Can Tho city, Viet Nam.

They were connected as a clear water recirculation system. It included 21 parallel-connected composite settlement tanks with a volume of about 500L per tank, connected to a sedimentation tank containing sand and stones (1- 2 mm Ø), functioning as a biological filter. The tanks were housed in an open wall and closed roof construction on a concrete foundation. The water source was running municipal tap water dechlorinated by aeriation for 24 hours before use. Water was circulated in the system at a rate of 3L per min into each tank. About 30% of the water was replaced twice daily.

Photo 2. Tanks were used in the experiment

The experiment was arranged in a completely randomized design with 21 tanks (0.5m³/tank) with 7 treatments and three replicates (n = 3). Seven diets were formulated with different levels of protein: 24%, 29%, 34%, 39%, 44%, 49% and 54% with the same energy and fat levels. The energy level from 18-20 kJ/g feed and fat from 8-12% in the feed were selected based on the data obtained from analysing the chemical composition of the feed during the local survey.

Fish in the experiment were reared for 3 months at the density of 100 individual/m³.

Experimental diets

Seven iso-energetic (19.5 MJ /Kg) and lipid (10.6% crude protein) diets were produced with fish meal, soybean meal, wheat flour, fish oil, vegetable oil, earthworm liquid, premix vitamins - minerals, and CMC (Table 1 and 2) at the level of 24%, 29%, 34%, 39%, 44%, 49% and 54%. Diet recipes are given in Table 1. In Table 2 is the chemical composition and level of selected essential amino acid given.

All ingredients were thoroughly mixed and then pelleted by using an electronic meat grinder (Quoc Hung company, Long Xuyen city, Vietnam) with pellet diameter and length in the range of 1 mm. All diets were dried by oven in 60 °C until the feed’s moisture less than 10%. Then diets were weighed and stored in sealed plastic bags in small portions at 5 °C until use.

The process of mixing ingredients and making pellets was described in Diagram 1.

Diagram 1. Process for making feed pellets

Fish management

The tire track eels were fed to satiation (approximately 10% of BW) manually twice daily (between 7:00 - 9:00 and between 16:00 - 18:00) by using a plastic tube to put feed into the nets at the bottom of the tanks. Each feeding was closely monitored in each tank, the feed residue was collected, and the feeding rate was adjusted depending upon the feed consumed on the previous day. The amount of feed used was recorded during the experiment and used to calculate true feed intake. In addition, the feeding and swimming behaviours of the fish were monitored and recorded. The water of rearing tanks was changed about 30 % every day and feed residue were collected before next feedings. During the experiment, water environmental factors (e.g., temperature, pH, O₂ and TAN and NO ₂^-) were daily monitored and controlled. Temperature in the experimental treatments ranged from 26.5 to 29.8^oC, pH from 7.5 to 8.1, dissolved oxygen from 5.8 to 6.5 mg/L, total protein (TAN) 0.003 - 0.01 mg/L and NO₂^- from 0.1 to 0.3 mg/L were all within the appropriate limits for the normal growth and development of tire track eels (Boyd and Pillai 1985).

Measurements and calculations

All fish were weighted and measured for weight and length before and after experiment. Specific growth rate (SGR), daily weight gain (DWG), feed conversion ratio (FCR), feed conversion efficiency (FCE), protein efficiency ratio (PER), total feed intake per fish (FI), protein intake (PI), survical rate (SR) and net protein utilisation (NPU) were calculated using the equations (Da et al 2012; Nguyen et al 2021; Zhang et al 2016):

SGR (%/day) = [(ln final Wt – ln initial Wt)/days] x 100

DWG (g/day) = (Final Wt – Initial Wt)/days

FCR = Total feed intake (g)/Total wet weight gain (g)

FCE = 1/FCR

PER = Total wet weight gain (g)/Protein intake (g)

FI (g) = total feed intake (g)/number of fish

PI (g) = Feed intake (g) x Protein in the diet (%)

SR (%) = (Total number of fish harvest/Total number of fish cultured) x 100

NPU (%) = [(Protein in final fish – Protein in initial fish)/Protein intake] x 100

In addition, at the end of the experiment, five fish from each tank were sampled randomly and sacrificed by lethal anesthesia using ethylene glycol monophenyl ether at 0.5 mg/L. Viscera, liver, stomach, and intestine were dissected out and weighed/measured. Viscera weight index (VSI, %), hepato˗somatic index (HSI, %), gastro-somatic index (GSI, %) and intestinal quotient (Qi) index were then determined following procedures described by Moreira et al (2012), Nguyen et al (2021) and Da et al (2012) using the equations:

VSI (%) = [100× (Viscera weight) (g)/Body weight (g))]

HSI (%) = [100× (Liver weight (g)/Body weight (g))]

GSI (%) = (Stomach weight/Total weight of the individual) x100

Qi = Intestine length/Total length of the individual

Chemical analysis

Ingredients, feed and experimental fish samples before allocating to the experiments and after harvesting were kept frozen (-20^oC) until analysis. Chemical compositions (e.g., crude protein, crude lipid, crude energy, crude fibre, ash, amino acids and moisture) of these samples were analysed in triplicate and measured in dry mass as described in Nguyen et al (2021). Specifically, dry matter was determined drying in an oven at 105 ^oC until constant weight. Ash content was determined by incineration of the samples at 560^oC for 4 hours (until constant weight). Crude protein was calculated as 6.25 x %N analysed by the Kjeldahl method. Crude lipid was determined by Soxhlet extraction without acid hydrolysis. Crude fibre content was analysed using acid-base digestion (AOAC, 2000). Nitrogen-free extract (NFE) was calculated as NFE (%) = 100 – (% protein + % lipid + % fibre + % ash). The gross energy (Kcal/Kg) was calculated by using gross energy values of 5.64 Kcal/g for crude protein, 4.11 Kcal/g for carbohydrate, and 9.44 Kcal/gfor crude fat (NRC 1993). Amino acid content was determined by high performance liquid chromatography according to the method described by Vázquez-Ortiz et al (1995) at the laboratory of the National Institute of Animal Husbandry, Ministry of Agriculture and Rural Development, Vietnam. Briefly, diet and ingredient samples (10 mg) were hydrolyzed with 6 N HCl for 24 h. Sodium thioglycolate was added to the samples to prevent oxidation. The hydrolysed samples were suspended in sodium citrate buffer (pH 2.2), derivatized with o-phthaldialdehyde (OPA), and injected (10 μL) into a Varian 9012 HPLC apparatus equipped with a fluorescent detector with a 340–380 nm excitation filter and 460 nm emission filter. α-aminobutyric acid was used as an internal standard. Separations were carried out on a reversed-phase C18-column and the flow rate of mobile phase was 1.5 mL per min at 25 - 29 °C. The amino acids were all eluted within 20 min.

Statistical analysis

Protein requirements were determined based on the quadratic regression method of Zeitoun et al (1976) with the equation y = ax² + bx + c (where y is growth and x is the protein content of the feed). From this equation, a y max point and its corresponding x max value were determined. The x max value is the protein content for the fish to achieve maximum growth. Means and standard deviations were calculated. The data were analysed using Minitab 16.0 software. One-way ANOVA and DUNCAN test were used to compare means between treatment groups.

Results and discussion

Growth of tire track eels with different protein levels

There was no significant difference in initial sizes of fish (e.g., weight and length) between treatments (p > 0.05, Table 3). This indicates that the initial weight of fish did not affect the growth of fish after the experiment.

There was a significant difference in growth rates of tire track eels fed different protein contents and the growth rate was highest in the 44% protein diet group and the growth response of eels to increasing level of protein in the diet (Figure 1). In fact, daily weight gain (DWG) of the 44% protein content group (i.e., 0.093 g/day) was not significantly different from those of the 39% and 49% diet groups, but there were significant differences in DWG of this group from the remaining groups (24%, 29%, 34% and 54%). Specific growth rate (SGR) of tire track eels in the treatments using different protein contents ranged from 0.76 to 1.59 %/day. The 44% protein group obtained the highest SGR (e.g., 1.59%/day) which was not significantly different from those of the treatment groups (39% - 54%, p > 0.05), but was significantly different (p < 0.05) from those of the remaining diet groups (24%-34%).

Figure 1. Growth response of eels to increasing level of protein in the diet

Protein-based energy sources are the most expensive (Nunes et al 2014), therefore, non-protein energy sources should be provided as much as possible to reduce the cost of feed. This study indicated that of the seven protein levels ranging from 24% to 54%, the 44% protein content stimulated the best growth rates. This result is similar to the results of research on protein requirements in feed for other species. For example, in a study in snakehead (Channa micropeltes) (Hien et al 2005) fingerling with different seed sizes, e.g., small (2.6 g/fish) and large (6.07 g/fish) were fed 5 different protein contents ranging from 14% to 54% (energy 4.2 kcal/g) for 50 days and the protein content of 44% and 54 % brought about the highest growth rates in large and small fingerlings, respectively. In another research on Asian red tail catfish (Hemibagrus wyckioides) (Le et al 2013), a 45% protein content showed the best SGR. In a study on the protein requirement for bronze featherback (Notopterus notopterus) fingerlings, Tran et al (2013) revealed that the appropriate protein content was 40%-45%, corresponding to the lipid content in the feed of 9% and 6%, respectively. In contrast, compared to some catfish species, the dietary protein of this species was higher than those of fingerling of catfish Pangasianodon hypophthalmus (e.g., 40.5 %), Pangasius bocourti (e.g., 35%) (Hung et al 2002)

Protein requirements in feed for optimal growth

The optimal protein requirement for tire track eel fingerling was determined by the quadratic regression method of Zeitoun et al (1976). The equation was y = - 0.0014 x² + 0.1265 x – 1.5351 with the correlation coefficient R² = 0.8355 (where y is growth and x is the protein content of the feed). From this equation, the maximum point (Y max = 1.4) was determined, and the corresponding X max was 45.2. The value of X max = 45.2 is the protein content for the fish to achieve maximum growth (Figure 2). The results of this study are equivalent to the study on snakehead fingerling as the quadratic curve analysis showed that the protein content for maximum growth in small fingerling (e.g., 2.6 g/fish) was 50.8% and large fingerling (6.07 g/fish) was 46.5% (Hien et al 2005). The optimal protein requirement for clown knifefish ( Chitala chitala) fingerling was 40%-45% matching 9%-6% lipid, respectively, resulting in optimum growth and effective protein utilization (Tran et al 2013).

Figure 2. Correlation between the protein contents and specific growth rates of tire track eel fingerling

Survival rates and feed efficiency

After 90-day treatment period, survival rates of tire track eels were relatively high ranging from 76.7 to 84.7% and there was no significant difference in survival rates between treatment groups (p > 0.05, Table 4). Thus, the results of this study showed that diets with different protein content did not affect the survival rates of tire track eels. Similarly, most studies indicated that diets with different protein content have had no effect on fish survival. For example, in a study on pointed-tailed goby (Pseudapocryptes elongatus) fingerling, diets with different protein contents and energy levels did not affect survival rates of the fish as there was no significant difference in survival rates between treatment groups ranging from 85.7-92.9% (Tran et al 2014). The results of this study are similar to those of Tran et al (2013) on clown knifefish (Chitala chitala) fingerling where survival rates fluctuated from 71.1 to 84.4% and there was no significant difference ( p > 0.05) in the survival rates between diets with different protein contents and lipid levels.

Feed conversion ratio (FCR) seemed to decrease gradually as protein contents in diets increased from 24% to 44% and FCR rose slowly when the protein levels in diets were 49% and 54% (Table 4 and Figure 3). FCR was lowest in the treatment of 44% protein (e.g., 1.43) which was not significantly different (p > 0.05) from those of the groups of 39% and 49% protein but was significantly different from the remaining treatment groups of 24%, 29%, 34% and 54% protein. The results of this study are similar to those of Tran et al (2013) on clown knifefish (Chitala chitala) with the lowest FCR at the protein level of 45%.

Figure 3. Feed conversion rate of eels to increasing level of protein in the diet

Similar to FCR, feed efficiency (FCE) was highest in the 44% protein treatment and lowest in the 24% protein treatment. FCE was not significantly different (p > 0.05) between the 44% and 49% but was significantly different (p < 0.05) from those of the remaining treatment groups (Table 4).

Protein efficiency ratio (PER) is the weight gain of aquatic animals per unit weight of protein ingested. PER varies with the amount and types of protein ingested and changes with the protein content of feed (Hien and Tuan 2009). According to (Lovell 1989), the protein content in feed is the most important factor affecting growth and feed cost in aquaculture. Increasing the protein content of the feed often improves fish performance, especially carnivores, but also increases feed costs. With the same source of protein, PER will be high in low protein content feed because aquatic animals will make the most of the protein source in the feed to build their bodies (Hien and Tuan 2009). The results of this experiment are consistent with the above statement as PER of tire track eels fed different protein contents after 90-day treatment period was highest in the 24% treatment group (1.94) and decreased gradually when the protein contents in the feed increased. The lowest PER was in the 54% protein group (1.14). There were significant differences (p < 0.05) in PER between treatment groups but no significant difference (p > 0.05) in PER between treatments with protein contents from 34% to 44%.

Net protein utilisation (NPU) in the total protein intake ranged from 17.49 to 31.51 %. There were significant differences (p < 0.05) in NPU between protein content groups. NPU tends to be similar to PER as it decreased when the protein content of the feed increased.

Chemical composition of tire track eels before and after the experiment

Feed has a great influence on the biochemical composition of aquatic animals. The results of the biochemical composition analysis of tire track eels before and after the experiment are shown in Table 5.

The moisture of the fish after the experiment fluctuated between 73.8% and 77.3% and was lower than that of the original fish (82.4%). Other biochemical components such as protein, lipid and ash in the body of fish after the experiment were higher than those of the fish at the beginning of the experiment (Table 5).

Protein content of fish after the experiment ranged from 13.5 to 15.8% and was highest in the 44% protein content group. However, there was no significant difference (p > 0.05) in the protein content of fish between this group and other groups with protein contents from 39% to 54%. Lipid and ash content in the fish body after the experiment also showed a similar trend as the protein content (Table 5).

Body indices

Table 6. Body indices of tire track eels after 90 days fed experimental diets with different protein levels
Parameters	Protein levels (%)							SEM	p - value
	24	29	34	39	44	49	54
HSI,%	1.63	1.70	1.67	1.43	1.66	1.45	1.85	0.21	0.82
VSI, %	9.95	10.10	9.24	8.78	9.59	9.11	10.60	0.77	0.69
iGas, %	1.05	1.42	1.18	0.99	1.33	1.29	1.80	0.21	0.14
Qi	0.21	0.25	0.22	0.21	0.23	0.24	0.28	0.02	0.08
Values are given as Lsmean, SEM=Standard Error of the Mean.

Body indices of tire track eels fed different protein levels was show in Table 6. There was no significant interaction was found between diet protein (p>0.05)

Conclusion

• The protein content of 44% produced the best growth of tire track eels, the lowest FCR and the highest FCE.
  
  
• Survival rates and body indices were quite high and did not differ between the treatment groups.
  
  
• Based on the correlation of the second-order curve, it showed that the value of 45.2% protein in the feed brought about optimal growth of tire track eels.

Acknowledgments

This research is funded by Vietnam National University HoChiMinh City (VNU-HCM) under grant number B2021-16-01

References

Ahmad A B, N Zaini, N Nayan, M Fahmi-Ahmad, S A Rizal and Y Yusuf 2018 Freshwater fish diversity of Sungai Setiu, Terengganu, Peninsular Malaysia. Malayan Nature Journal 70 (4):499-507.

AOAC 2000 Official methods of analysis of AOAC International (17th ed.), AOAC International, Gaithersburg, MD, USA (2000).

Boyd C E and V Pillai 1985 Water quality management in aquaculture. CMFRI special Publication 22:1-44.

Da C T, T Lundh and J E Lindberg 2012 Evaluation of local feed resources as alternatives to fish meal in terms of growth performance, feed utilisation and biological indices of striped catfish (Pangasianodon hypophthalmus) fingerlings. Aquaculture 364–365:150-156.

Gupta S and S Banerjee 2016 Food, feeding habit and reproductive biology of tire-track spiny eel (Mastacembelus armatus): A review. Journal of Aquaculture Research and Development 7 (5):429.

Hien T T T, N A Tuan, D T Yen and N T N Lan 2005 Protein requirements of giant snakehead (Channa micropeltes, Cuvier, 1831) fingerlings. Science Journal of Can Tho University (in Vietnamese version) :58-65.

Hien T T T and N A Tuan 2009 Textbook of nutrition and aqua feed. Agricultural publisher.

Hung L T, P T Liem, H T Tu and C Mariojouls 2002 Comparing growth and protein requirements for fingerlings of three catfish of the Mekong river (P. bocourti, P. hypophthalmus and P. conchophilus). J. Aqua. Tro., 17(4), 325-335.

Jamaluddin J A F, N So, B M Tam, A Ahmad, C Grudpan, L M Page, M Z Khaironizam and S A Mohd Nor 2019 Genetic variation, demographic history and phylogeography of tire track eel, Mastacembelus favus (Synbranchiformes: Mastacembelidae) in Southeast Asia. Hydrobiologia 838 (1):163-182.

Jamsari A F J, S Nam, B M Tam and M N Siti-Azizah 2014 Isolation and characterisation of microsatellite loci in the tire track eel, Mastacembelus favus and cross-species amplification. Conservation Genetics Resources 6 (2):477-479.

Khoa T T and T T T Huong 1993 Identification of freshwater fish in the Mekong Delta (in Vietnamese).

Lenh L V 2010 Experimental research on artificial spawning stimulation and rearing of tire track eel (Mastacembelus favus). Department of Scientific Management and International Cooperation, Faculty of Agriculture, An Giang University.

Loan P P, N. T. Anh, V H Vinh, L T Tung, L V Linh, T K Ngoc, N V H Tinh and T T Hong 2010 Developing an artificial seed production process for Tire track eels ( Mastacembelus favus) in An Giang Scientific research project of An Giang University. (in Vietnamese) :70 pages.

Lovell T 1989 Nutrition and feeding of fish (Vol. 260). New York: Van Nostrand Reinhold.

Moreira R L, Silveira L P, Teixeira E G, Moreira A G L, Moura P S D and Farias W R L 2012 Growth and gastrointestinal indices in Nile tilapia fed with different diets. Acta Scientiarum. Animal Sciences. 34, 223-229.

Nguyen H Y N, T L Trinh, K Baruah, T Lundh and A Kiessling 2021 Growth and feed utilisation of Nile tilapia (Oreochromis niloticus) fed different protein levels in a clear-water or biofloc-RAS system. Aquaculture 536:736404.

NRC 1993 Nutrient Requirements of Fish. National Academy Press, Washington, DC.

Nunes A J, M V Sá, C L Browdy and M Vazquez-Anon 2014 Practical supplementation of shrimp and fish feeds with crystalline amino acids. Aquaculture 431:20-27.

Petsut N and S Kulabtong 2015 Field survey of freshwater fishes in Upper Wang River, North Thailand. Biodiversity Journal 6 (2):513-516.

Rainboth W J 1996 Fishes of the cambodian mekong : Food & Agriculture Org.

Tran H T T, B H Nguyen, L M Lam and T C L Tran 2013 Protein and lipid requirements for clown knifefish fingerling (Chitala chitala). Can Tho University Journal of Science

Tran Le Cam Tu, Duong Kim Loan, Trang Tuan Nhi, Tran Thi Thanh Hien 2014 Determination protein requirement of goby fingerling (Pseudapocryptes elongatus, Cuvier 1816) at two different energy levels. Can Tho University Journal of Science 2014: 302-309. https://ctujsvn.ctu.edu.vn/index.php/ctujsvn/article/view/1837 (26):196-204. https://ctujsvn.ctu.edu.vn/index.php/ctujsvn/article/view/114

Trung N T, N T Anh and N Q Thanh 2010 Experiments of artificial propagation of tire track eel (Mastacembelus favus). (in Vietnamese).

Vázquez-Ortiz F, G Caire, I Higuera-Ciapara and G Hernández 1995 High performance liquid chromatographic determination of free amino acids in shrimp. Journal of Liquid Chromatography & Related Technologies 18 (10):2059-2068.

Zeitoun I H, D E Ullrey, W T Magee, J L Gill and W G Bergen 1976 Quantifying nutrient requirements of fish. Journal of the Fisheries Board of Canada 33 (1):167-172.

Zhang N, G Z Luo, H X Tan, W C Liu and Z W Hou 2016 Growth, digestive enzyme activity and welfare of tilapia ( Oreochromis niloticus) reared in a biofloc-based system with poly-beta-hydroxybutyric as a carbon source. Aquaculture 464:710-717.