The effect of various levels of protein, fat, carbohydrates and energy on growth, survival and body composition of Chanos chanos fingerlings
F. Piedad Pascual
Aquaculture Department, Southeast Asian Fisheries Development Center, Tigbauan, Philippines
Summary
Optimum protein, fat, carbohydrates and energy requirements of milkfish fingerlings were determined using growth, survival and body composition as parameters to assess the effectiveness of the diets. Chanos chanos fingerlings weighing 0.5 to 8.0 g were fed semi-purified dry diets consisting of casein and gelatin (4:1), corn oil and cod liver oil (1:1), dextrin, vitamin and mineral mixes, celufil and carboxymethyl cellulose. Treatments consisted of 27 combinations using three levels of protein (15, 30, 45%), fat (0, 6, 12%) and carbohydrates (10, 20, 30%) with two replicates each.
Each replicate consisted of 20 milkfish which were reared for 8 weeks in a flow-through fibreglass tank (40 1 volume) using filtered seawater. Temperature and salinity ranged from 26 to 31°C and 30 to 32 ppt, respectively. Feeding rate of dry pellets was 10% of total biomass.
The results indicate that fingerlings require a protein level of 30-40% depending on their size, a fat level of 10% and a carbohydrate level of 25%. Weight gain did not improve if energy levels exceeded 3500 kcal/kg diet.
Based on response surface analysis a summary of possible optimum combinations of protein (%)-, fat (%)-, carbohydrate (%)- and energy levels (kcal/kg) in the diets are: for survival rate, 30/12/10/2540; for weight gain, 30/6/20/2680; for protein deposition 40/6 to 10/20 to 30/2960 to 3740; for fat deposition 40/6/10 to 30/2560 to 3360; for ash deposition 40/6/20 to 30/2460 to 3740.
Of the treatments, the following 5 diet combinations gave mortality rates less than 50%: 15/0/30/1800, 30/12/10/2680, 45/0/10/2220, 45/0/30/2800 and 45/12/20/3780.
Introduction
Rice and fish are staple foods in the Philippines as well as in other parts of Southeast Asia, where calorie/protein malnutrition is a serious problem. Milkfish, (Chanos chanos) is well-liked by the masses and, therefore, it could be a good source of protein and calories. In traditional culture, milkfish yield is rather low and limited to the carrying capacity of the culture ponds (Lim et al., 1979). To increase production through intensive culture, a diet to supplement the natural food in the ponds is needed. However, practical diet formulation is hampered by the lack of knowledge about the nutritional requirements of milkfish. The present study was carried out to determine optimum protein, fat, carbohydrate and energy levels for milkfish fingerlings using growth,
228
survival, and body composition as parameters for determining the effectivity of the diets. The response surface analysis was explored graphically using a "freehand" technique to search for optimal diets with respect to dietary levels of protein, fat and carbohydrates.
Methodology
Chanos chanos fingerlings (0.5 to 0.8 g) were fed dry semipurified diets consisting of casein and gelatin in a 4:1 ratio as protein sources, cod liver oil and corn oil (1:1) as fat sources, and dextrin as a carbohydrate source. Celufil made up for the difference in dextrin while vitamins and minerals were added. Carboxymethyl-cellulose (CMC) was used as binder (Table 1). Diets were prepared as described in National Academy of Sciences NAS/NRC (1973).
Table 1. Percentage Composition and P/E ratio's of the experimental diets.
Table 1. Percentage composition and P/E ratio's of the experiment diets.					
Diet no.	P-F-C	Casein: Gelatin	Protein as analyzed in %	Fat as analyzed in %	Protein/Energy ratio (P/E in mg P/kcal)
1, 2, 3	15-0-10/20/30		15.9, 15.5, 15.9	0.2, 0.1, 0.3	150, 107, 83
4, 5, 6	15-6-10/20/30	12.67:	15.1, 16.3, 16.6	5.9, 5.7, 5.5	97, 77, 64
7, 8, 9	15-12-10/20/30	3.17	17.4, 16.6, 16.1	11.4, 11.6, 10.2	72, 60, 52
10, 11, 12	30-0-10/20/30		29.8, 29.8, 30.6	0.4, 0.2, 0.5	188, 150, 125
13, 14, 15	30-6-10/20/30	25.35:	31.4, 36.3, 29.6	5.1, 5.1, 4.5	140, 118, 102
16, 17, 18	30-12-10/20/30	6.40	32.5, 30.9, 31.4	10.7, 10.7, 10.4	112, 97, 86
19, 20, 21	45-0-10/20/30		45.9, 45.5, 45.1	0.6, 0.4, 0.3	205, 173, 150
22, 23, 24	45-6-10/20/30	37.60:	45.4, 43.7, 44.8	4.7, 4.3, 4.4	164, 143, 127
25, 26, 27	45-12-10/20/30	9.60	45.7, 46.3, 44.4	9.5, 9.1, 8.2	137, 122, 110
Vitamin mix (mg/kg/diet)
Vit. A, 10.99 mg; Vit. D2, 0.025 mg; Vit. E, 67.06 mg; Vit. K3, 10.0 mg; Choline Chloride, 581.52 mg; Niacin, 99.92 mg; Riboflavin, 19.98 mg; Pyridoxine HC1, 19.98 mg; Thiamine HC1, 19.98 mg; D-Calcium Pantothenate, 49.96 mg; Biotin, 10.0 mg; Folacin, 15.0 mg; Vit. B12, 12.0 mg; Ascorbic Acid, 100.00 mg; Inositol, 100.0 mg; Celufil, 12,303.585 mg.
Mineral mix (mg/kg/diet)
CaHPO4, 16.37; CaCO3, 14.8; KH2PO4, 10.0; KC1, 10; NaCl, 6.0; MnSO4H2O, 0.35; FeSO4.7H2O, 0.50; MgSO4.7H2O, 6.137; K103, 0.01; CuSO4.5H2O, 0.03; ZnSO4.7H2O, 0.347; CoCl2.6H2O, 0.0031;
Diets were pelleted using a meat grinder and oven-dried (60°C) until moisture content was 10% or below. Diets were stored in a refrigerator at 4°C.
Fingerlings were taken from a pond, acclimated to laboratory
229
conditions and seawater, fed a pelleted dry diet for two weeks, and anesthetized with 2-phenoxy ethanol (0.25 ppt) prior to weighing and stocking. Two replicates of 20 milkfish each were reared for eight weeks in 40 liters of filtered sea water in flow-through rectangular tanks (61 cm x 35 cm x 24 cm) . Temperature and salinity were measured twice daily and ranged from 26 to 31°C and 30 to 32 ppt, respectively. Feeding rate was 10% of total biomass. Fish were fed thrice daily.
Fish were pooled per treatment for proximate analyses at the end of the study. Proximate composition of the diets were obtained by methods previously reported (Pascual et al., 1983). Energy levels and protein to energy (P/E) ratios were calculated using kilocaloric values of 4/g protein, 9/g fat, and 4/g carbohydrates (CHO).
Total weights of fish stocked per tank were obtained at the start and at the end of the study. Daily mortalities were recorded.
A 3x3 factorial incomplete block design was followed. The treatments consisted of 27 combinations with three levels (%) of protein (15, 30, 45), fat (0, 6, 12), and carbohydrates (10, 20, 30) with two replicates for each of the 27 treatments. The milkfish were blocked by size such that fibreglass tanks were arranged in three incomplete blocks consisting of nine tanks per block. Analysis of variance was used to show statistical differences among treatment means. Honest significant difference (HSD) to determine differences between treatment means and response surface analysis were used to estimate some possible optimal diets.
Results
Percentage survival of the milkfish fingerlings showed that none of the diets could be singled out as a diet providing good survival. However, of the 27 treatments, 6 diets gave survival rates of more than 50% throughout the experimental period (Table 2 and 3).
Table 2. Diet combinations that gave survival rates of more than 50%.
% Survival	Protein	Fat	CHO	kcal/kg Diet	P/E ratio
53	15	0	30	1800	83
55	30	0	30	2400	125
60	30	12	10	2680	112
55	45	0	10	2220	205
58	45	0	30	2680	150
60	45	12	20	3680	122
230
Table 3. Analysis of variance of the arcsin transformed percentage survival rates of milkfish fingerlings after a period of 56 days.
Source of variation	df	Sum of squares	Mean. square	F
Replicates	1	325.69016	325.60016	5.81*
Blocks within replicates	4	5,513.05370	1,378.26343	24.59**
A (protein)	2	766.45750	383.22875	6.84**
B (fat)	2	269.48409	134.74204	2.40
C (carbohydrate)	2	1,303.14914	651.57457	11.62**
A x B	4	1,192.21117	298.05279	5.32**
A x C	4	222.48304	55.62076	0.99
B x C	4	632.22878	158.05720	2.82**
AxBxC unconfounded	4	346.97976	86.74494	1.55
partially confounded	4	206.46332	51.61583	0.92
Experimental error	22	1,233.12862	56.05130	
Total	53	12,001.32928		
*	Significant (P <0.05)
**	Highly significant (P <0.01)
Response surface analysis for weight gains showed that a well defined maximum appeared at 30% protein, 6% fat, and 20% CHO. As the level of CHO increased to 30%, a higher protein level seemed to be required to obtain optimum weight gain.
Body composition is shown in Table 4.
231
Table 4. Percentage mean weight gain, survival, crude protein, crude fat and crude ash of milkfish fingerlings fed various diets.
Diet No.	P-F-C	Mean weight gain1 %	Survival %	Crude protein2 %	Crude fat2 %	Crude ash2 %
1	15-0-10	43.01	20	71.79	5.63	14.78
2	15-0-20	8.87	38	70.09	4.81	16.43
3	15-0-30	52.64	53	67.40	12.52	13.76
4	15-6-10	30.34	25	68.18	6.91	18.37
5	15-6-20	14.65	18	69.41	7.50	18.85
6	15-6-30	2.96	23	70.00	5.30	19.76
7	15-12-10	-19.52	5	66.38	—	—
8	15-12-20	-21.69	25	67.46	4.62	23.83
9	15-12-30	12.19	28	69.68	6.41	19.97
10	30-0-10	43.12	28	67.97	7.00	17.26
11	30-0-20	36.42	45	66.44	10.60	14.51
12	30-0-30	56.13	55	66.97	9.50	13.65
13	30-6-10	38.80	45	66.65	11.72	15.33
14	30-6-20	149.92	33	62.40	15.38	13.37
15	30-6-30	28.81	25	61.56	12.64	12.23
16	30-12-10	18.46	60	65.78	12.72	15.44
17	30-12-20	3.27	18	60.78	18.70	—
18	30-12-30	4.04	10	64.03	16.88	—
19	45-0-10	32.36	53	68.25	9.03	14.76
20	45-0-20	51.68	38	67.34	9.69	14.59
21	45-0-30	61.44	60	67.84	9.52	13.48
22	45-6-10	68.06	15	61.62	13.10	14.45
23	45-6-20	63.53	25	60.56	13.97	12.72
24	45-6-30	79.33	48	60.72	13.79	12.99
25	45-12-10	43.19	18	67.43	9.64	15.13
26	45-12-20	24.58	58	65.13	11.24	15.73
27	45-12-30	28.04	28	64.53	14.64	14.64
	Std. error of the mean	+24.15	+ 0	+0.451	+0.628	±0.558
	HSD (.05)			2.58	3.58	3.18
1Analyses of variance indicated insignificant differences .
2	Analyses of variance indicated highly significant differences .
Note: In a column of any two means whose difference is larger than the HSD (Honest Significant Difference) values are declared significantly different.
The response surfaces for percentual protein and fat deposition showed that protein and fat deposition occurred at all levels of CHO. The minimum levels for protein deposition are 30-45% protein, whereas the maximum level for fat deposition amounts to some 6% fat.
On the basis of response surface analysis, a summary of possible optimal combinations of diets is presented in Table 5.
232
Table 5. Some possible optimal combinations of diets.
Response variable	Protein	Fat	CHO	kcal/kg diet	P/E ratio
Survival rate	30	12	10	2680	112
Weight gain	30	6	20	2540	118
Protein deposition	40	6-10	20-30	2960-3740	135-107
Fat deposition	40	6	10-30	2960-3360	156-119
Ash deposition	40	6	20-30	2460-3740	163-107
Discussion
Survival up to eight weeks is possible even without dietary fat. According to Gorriceta (1982) milkfish of marketable size are able to desaturate and elongate fatty acids. One of the best natural foods "lab-lab" contains only 0.92% fat, yet it supports growth and provides for good survival. Milkfish avail over amylases and can digest raw starches (Chiu & Benitez, 1981). Likewise, proteases are present (Benitez & Tiro, 1982), but lipases occur in less amounts in marketable size milkfish.
Although digestive enzymes in milkfish fingerlings has not been defined, data suggest that milkfish can survive and grow with little or no dietary fat. Survival of more than 50% was obtained if fish fed diets 15-0-30, 30-0-30, 45-0-10 and 45-0-30, whereas these fish had mean weight gains of 52.6, 56.1, 32.4 and 61.4% respectively, which were among the higher ones obtained.
Total energy levels did not seem to affect survival. Mortality rates were less than 50% as long as caloric levels ranged between 1800 and 3680. Weight gains did not improve at caloric levels beyond 3500 kcal. Likewise, protein and fat deposition were optimal within this range. A protein energy ratio of 112 to 118 mg could be needed for weight gain and survival.
The low survival and relatively poor growth using some combinations could have been due to several factors, one of which may be the type of fat used. Corn oil contains much more w6 than w3 and very little C22 and C24 w3 fatty acids that are needed by marine fish for maximum growth (Yone, 1975). The same author reports poor growth and some toxic effects when high levels of w6 fatty acids were fed to seabream, whereas the NAS/NRC H440 diet mixture (NAS/NRC, 1973) from which the diets were patterned work well for salmonids and some catfish species (Halver, personal communication).
A ratio of 1:1 cod liver oil and corn oil was used in this study and total lipid was 6 and 12%. Since in previous experiments, where only corn oil was used, very poor growth and survival were obtained, a combination of corn oil and cod liver oil was used in this study. The fatty acid profiles of the diets were not analyzed, but published literature on the fatty acid profiles of corn oil and cod liver oil show that
233
the latter has more w3 while corn oil has more w6 fatty acids. When total fat was 12%, diets with 30 to 45% protein gave poor growth, but average gains of 68.06, 63.5 and 79.3% were obtained at 6% dietary fat and 45% dietary protein.
Alava (1985), who worked with various sources of oils (cod liver oil, corn oil, beef tallow, pork lard, sesame oil, and coconut oil) showed that cod liver oil was the best source while corn oil gave the poorest growth, survival, and liver histology. Furthermore in a study using various levels of cod liver oil, diets containing 7 to 10% gave the highest percentage weight gain and survival, whereas cellular and structural changes occurred in milkfish fingerlings fed diets containing less than 7% or more than 10% cod liver oil (Alava & dela Cruz, 1983.).
Benitez & Gorriceta (1985) suggest that chain elongation and desaturation takes place in the liver where significant quantities of 20:2w6, 20:3w6, 20:5w3, 22:4w6, and 22:5w3 were found despite the absence of these fatty acids in the natural food of the milkfish.
Furthermore, they showed that w6 fatty acids are also needed by the milkfish since they were found in milkfish livers and not in the depot fat. Tilapia zillii has been found to require w6 fatty acids in their diets (Kanazawa et al., 1980). Bautista & Cruz (1983) reported that milkfish fingerlings fed linolenic and/or linoleic acid showed good growth and survival, a combination of 1% linolenic acid with 6% lauric acid giving better growth than linoleic acid.
Diet combinations resulting in survival rates of more than 50% in this study show that milkfish are able to utilize protein and carbohydrates for the production of fatty acids, even in the absence of dietary fat.
Lim & Alava (1983) used 42% crude protein, 9% fat, and 33% carbohydrate for milkfish fry and obtained good results. Lim et al. (1979) found an optimum protein level of 40% while Coloso et al. (1988) reported the level to be around 45%. All these studies used milkfish much smaller than in this study.
Another factor causing low survival and/or poor growth using some diet combinations could be a vitamin deficiency in the diet. Vitamin C used was not the stable form. Furthermore, the diet processing, hot water (80°C) being added to the CMC-vitamin mixture followed by oven drying at 60°C for 6 hours, could have caused destruction of some heat-labile vitamins. In another study with practical diets for milkfish, an increase of the vitamin levels from 4 to 6% in the diet improved weight gain and survival in fish fed 40% protein, 25% CHO, and 10% fat (Pascual, 1983).
In a study by Lim et al. (1979) mean weight gains of milkfish fry initially weighing 40 mg, fed 10% body weight per day, were directly related to the levels of protein up to 40%. Lim's diet contained 6% fat (1:1 ratio cod liver oil to corn oil) and 31% CHO, whereas digestible energy was calculated to be 3340 kcal/kg.
Crude protein contents of 70% and above for body composition were observed in fish fed diets with 15% protein, the crude fat contents being much lower than in fish fed 30 to 45% protein. An increase in fat deposition was observed as protein levels in the diet were increased to 30 and 45%. Zeitler et al. (1984) found the protein content of carp (Cyprinus carpio)
234
to decrease with decreasing amounts of protein in the diet while fat deposits increased.
The P/E ratio of 127 mg protein/kcal is similar to that obtained by 'Winfree & Stickney (1981) with Tilapia aureus. They showed that a diet containing 56% protein, 4600 kcal/kg diet, and 123 mg protein/kcal produced the highest gain in fry (2.5 g) while larger fish grew most rapidly when fed a 34% protein 3200 kcal diet with a P/E ratio of 108. Cho et al. (1985) have indicated that it is not valid to assume that protein and complex carbohydrates have similar metabolizable energy values for fish, thus values presently reported will have to be confirmed and metabolizable energy studies in milkfish are urgently required.
Response surface analysis is a relatively fast method of determining optimal dietary nutrient combinations and could possibly hasten the study of nutrient requirements, be it that results can be difficult to interpret. Studies of diets containing 30-40% protein, 6% fat from a source containing various levels of w3 and w6 fatty acids, 20% CHO and optimal vitamin levels will be needed. The energy values presented are also tentative until metabolizable and digestible energies have been determined. Although results are tentative, a diet containing 30-40% protein, 6-10% fat, 20-30% CHO, a protein/ energy ratio of 118-114, and 2500-3500 kcal/kg is recommended for milkfish fingerlings, and this formulation can be reconsidered as soon as more nutritional data become available.
Acknowledgements
My sincere thanks are due to the International Foundation for Science for partially funding the study; to Dr. Rudy Tan for assisting with the experimental design and statistical analyses, and to the Centralized Analytical Laboratory of AQD, SEAFDEC for proximate analysis.
References
Alava, V.R., 1985. Dietary fat sources in dry diets for Chanos chanos (Forsskal) fingerlings. Abstracts Second International Conference on Warmwater Aquaculture Finfish, February 5 to 8, 1985. Abstract No. 1.
Alava, V.R. & de la Cruz, M.C., 1983. Quantitative dietary fat requirements of Chanos chanos fingerlings in a controlled environment. Abstracts Second International Milkfish Aquaculture Conference, Iloilo City, Philippines, October, 4 to 8, 1983. p. 32.
Bautista, M.N. & de la Cruz, M.C., 1983. Effects of dietary linoleic and linolenic acids on growth, survival, fatty acid composition and liver histology of milkfish fingerlings. Abstracts Second International Milkfish Aquaculture Conference, Iloilo City, Philippines, October 4 to 8, 1983. p. 55.
Benitez, L.V. & Gorriceta, I.R., 1985. Lipid composition of milkfish grown in ponds by traditional aquaculture. In: Cho, C.Y., Cowey, C.B. & Watanabe, T., Finfish Nutrition in Asia: Approaches ro research and development. Ottawa, Ont. IDRC. p. 145.
Benitez, L.V. & Tiro, L.B., 1982. Studies on the digestive protease of the milkfish Chanos chanos. Mar. Biol. 71: 309-
235
315.
Chiu, Y.N. & Benitez, L.V., 1981. Studies on the carbohydrates in the digestive tract of the milkfish Chanos chanos. Marine Biology of 61. p. 247.
Cho, C.Y., Cowey, C.B. & Watanabe, T., 1985. Finfish Nutrition in Asia: Methodological approaches to research and development. Ottawa, Ont. IDRC. p. 11.
Coloso, R.M., Benitez, L.V. & Tiro, L.B., 1988. The effect of dietary protein energy levels on growth and metabolism of milkfish (Chanos chanos Forsskal). Comp. Biochem. Physiol., v. 89A, No. 1, p. 11-17.
Gorriceta, I.R., 1982. Studies on the digestive lipases and lipid composition of milkfish, Chanos chanos Forsskal. MSc-Thesis, Univ. of the Philippines, 56 pp.
Kanazawa, A., Teshima, S. & Sakamoto, M. , 1980. Requirements of Tilapia zillii to essential fatty acids. Bull. Jap. Soc. of Sclent. Fish., 46(11): 1353-1356.
Lim, C. & Alava, V.R., 1983. Artificial diets for Chanos chanos (Forsskal) fry. Abstract Second International Milkfish Aquaculture Conference, Iloilo City, Philippines, October 4 to 8, 1983. p. 57.
Lim, Chorn, Sukhawongs, Somchart & Pascual, F.P., 1979. A preliminary study on the protein requirements of Chanos chanos (Forsskal) fry in a controlled environment. Aquaculture , 17: 195-201.
Mead, J.F. & Kayama, M., 1967. Lipid metabolism in fish. In: Stansky, M.E. (ed.). Fish oils. The AVI Publ. Co., Westport, CT, USA, 289-299.
NAS/NRC, 1973. Nutrient requirement of domestic animals no. 11. Nutrient requirements of trout, salmon and catfish. Subcommittee on Fis Nutrition, Committee on Animal Nutritional Agricultural Board, Nat. Res. Council of Nat. Acad. Sci., p. 50.
Pascual-Piedad, F., 1983. Biological evaluation of some practical diets developed for milkfish fingerlings. Abstracts Second International Milkfish Aquaculture Conference, Iloilo City, Philippines, October 4 to 8, 1983. p. 59.
Sinnhuber, R.D., 1969. The role of fats. In: Newhaus, 0. & Halver, J.E. (editors). Fish in Research. Academic Press, New York, N.Y., USA. 245-259.
Takeuchi, T. & Watanabe, T., 1976. Nutritive value of w3 highly unsaturated fatty acids in pollack liver oil for rainbow trout. Bull. Jap. Soc. Scient. Fish, 42: 907-916.
Takeuchi, T. & Watanabe, T., 1977. Effect of eicosapentaenoic and docosahexaenoic acid in pollack liver oil on growth and fatty acid composition of rainbow trout. Bull. Jap. Soc. Scient. Fish., 43: 947-953.
Winfree, Robert A. & Stickney, Robert R., 1981. Effects of dietary protein and energy on growth, feed conversion efficiency and body composition of Tilapia aureus. J. Nutri., 1ll: 1001-1012.
Yone, Yasuro, 1976. Nutritional studies of the red sea bream. Proceedings of the First International Conference on Aquaculture Nutrition. College of Marine Studies, Univ. of Delaware, New York, Delaware, USA. p. 39.
Zeitler, M.H., Kirchgeasner, M. & Schwarz, F.J., 1984. Effects of different protein and energy supplies on carcass composition of carp (Cyprinus carpio L.). Aquaculture, 36: 37-48.
236