1. INTRODUCTION
2. FATTY ACID COMPOSITION OF FISH
3. BODY LIPID COMPOSITION AND DIETARY LIPID REQUIREMENTS
4. POLYUNSATURATED FATTY ACID REQUIREMENTS OF FISH
5. ESSENTIAL FATTY ACID REQUIREMENTS OF FISH
6. FATTY ACID METABOLISM IN FISH
7. NEGATIVE ASPECTS OF LIPIDS IN FISH NUTRITION
8. REFERENCES
J. E. Halver
University of Washington
Seattle, Washington
Lipids are the generic names assigned to a group of fat soluble compounds found in the tissues of plants and animals,: and are broadly classified as: a) fats, b) phospholipids, c) sphingomyelins, d) waxes, and e) sterols.
Fats are the fatty acid esters of glycerol and are the primary energy depots of animals. These are used for long-term energy requirements during periods of extensive exercise or during periods of inadequate food and energy intake. Fish have the unique capability of metabolizing these compounds readily and, as a result, can exist for long periods of time under conditions of food deprivation. A typical example is the many weeks of migration by salmon in their return upstream to spawn; stored lipid deposits are burned for fuel to enable body processes to continue during the strenuous journey.
Phospholipids are the esters of fatty acids and phosphatidic acid. These are the main constituent lipids of cellular membranes allowing the membrane surfaces to be hydrophobic or hydrophylic depending on the orientation of the lipid compounds into the intra or extracellular spaces.
Sphingomyelins are the fatty acid esters of sphingosine and are present in brain and nerve tissue compounds.
Waxes are fatty acid esters of long-chain alcohols. These compounds can be metabolized for energy and to impart physical and chemical characteristics through the stored lipids of some plant and several animal compounds.
Sterols are polycyclic, long-chain alcohols and function as components of several hormone systems, especially in sexual maturation and sex-related physiological functions.
Fatty acids can exist as straight chain or branch chain components; many of the fish fats contain numerous unsaturated double bonds in the fatty acid structures. A short bond designation for. fatty acids will be used throughout where the w number identifies the position of the first double bond counting from the methyl end. Linolenic acid would be written 18:3w 3. The first number identifies the number of carbons; the second number, the number of double bonds; and the last number, the position of the double bonds.
Many reviews of fish nutrition have been published which contain information on lipid requirements. Most work on lipid requirements of fish has been with salmonids. Rainbow trout have an essential fatty acid (EFA) requirement for the linolenic of w 31 series rather than for linolenic or w 6 as required by most mammals. The main emphasis on lipid requirements has been on EFA and on the energy value of lipids.
2.1 Environmental Influences
2.2 Effects of Diet
2.3 Seasonal Variation
2.1.1 Salinity
The difference between fatty acid compositions of marine and freshwater fish has been noted by several authors. Some examples of fatty acid patterns are given in Table 1. Although these fish lipids are higher in w 3 fatty acids, it is clear that freshwater fish have higher levels of w 6 fatty acids than marine species. The average w 6/w 3 ratios are 0.37 and 0.16 for freshwater and marine fish, respectively. Fish in general contain more w 3 than w 6 polyunsaturated fatty acids (PUFA) and should have a higher dietary requirement for w 3 PUFA; thus the dietary EFA requirement of marine fish for w 3 PUFA may be higher than that of freshwater fish.
The same type of difference in the w 6/w 3 ratio between freshwater and seawater is seen when some species of fish migrate from oceans to streams or vice versa. The PUFA ratio of sweet smelt (Plecoglosus altivelis) changes drastically in only one month as they migrate from the sea to a freshwater river. A similar but reverse change occurs in the masu salmon (Oncorhynchus masu) as they migrate from freshwater to seawater. Even within the same species of fish, the salinity of the water seems to cause a dramatic change in the fatty acid pattern.
The difference between marine and freshwater fish may be due simply to differences in the fatty acid content in the diet or it may be related to a specific requirement in fish related to physiological adaptations to the environments. The phospholipids are generally considered to be structural or functional lipid, being incorporated to a large extent in the membrane structure of cell and subcellular particles. The triglycerides are more often storage lipids and reflect the fatty acid composition of the diet to a greater extent than do the phospholipids. In Table 2, the fatty acid compositions of the triglyceride and phospholipid fractions of fish lipids are presented. It can be seen that the effect of changing environment on the fatty acid composition of the phospholipid is as great in the case of salmon, and considerably greater in the case of the sweet smelt, than it is on the triglyceride composition. Rainbow trout on diets containing either corn oil, which is high in w 6 but low in w 6 PUFA, showed a higher mortality and growth reduction in seawater than in freshwater over the twelve-week feeding period.
2.1.2 Temperature
There are several other factors besides the salinity of the water which affect the fatty acid composition and especially the PUFA of fish. In Tables 1 and 2 it can be seen that the salmonids, even in freshwater, tend to have a higher total PUFA of the 20 and 22 carbon chain length, and a lower w 6/w 3 ratio than the other fish. The salmonids are mostly cold-water fish. The fatty acids from a number of marine animals from temperate and arctic waters show some significant differences in the general pattern; unfortunately analysis included fatty acids longer than 20:1. There are a number of other experiments demonstrating the effect of environmental temperature on fatty acid composition of aquatic animals. The general trend toward higher content of long chain PUFA at lower temperatures is quite clear. The w 6/w 3 ratio decreases with a decrease in temperature (Table 3). If the trends in fatty acid composition can be taken as clues to the EFA requirements of fish, the w 3 requirement would be greater for fish raised at lower temperatures. Fish raised in warmer waters, such as common carp, channel catfish, and tilapia may do better with a mixture of w 6 and w 3 fatty acids.
Some of the fatty acid compositions listed in Table 3 may be seriously affected by the dietary lipids. The mosquito fish and guppies were fed trout pellets which had an w 6/w 3 ratio of 2.75. The catfish were fed diets supplemented with either beef tallow or menhaden oil, with w 6/w 3 ratios of 18.13 and 0.15, respectively. These fish were able to alter the dietary w 6/w 3 ratio in favour of w 3 fatty acid incorporation into the flesh lipids even at the highest temperature. Commercially available trout pellets are often low in w 3 PUFA and high in w 6 fatty acids. It is important not to ignore the effect of dietary lipid composition on fatty acid composition of fish fed artificial diets. It is clear from the data in Table 3 that the w 6/w 3 ratio of the fish lipids is greatly affected by the w 6/w 3 ratio of the dietary lipids. When the dietary ratio is very high in w 6 fatty acids supplied by animal lard or vegetable oils, there is a tendency for fish to alter the ratio of PUFA incorporated in favour of w 3 fatty acids. When the dietary oil is a fish oil high in to3 fatty acids, there is little change in the w 6/w 3 ratio of lipids incorporated into the fish. This is further suggestive evidence of an EFA requirement of fish for w 3 PUFA.
Table 3 Effect of Environmental Temperature on Fatty Acid Composition of Fish Lipids 1/
Seasonal variations in the fatty acid composition of fish species have often been reported. Seasonal changes have been observed in total lipid and iodine values of herring oils. The iodine value or degree of unsaturation of the oil was minimal in April and maximal in June. The great increase in unsaturation corresponded to the onset of feeding in spring. The absence of a gas liquid chromatograph (GLC) at the time precluded identification of changes in individual fatty acids.
Flesh and viscera lipid content of the sardine Sardinops melanosticta vary from 3.9 to 10.77 percent and from 10.9 to 38.3 percent, respectively. The fatty acids of principal interest with respect to EFA metabolism are 20:4w 6, 20:5w 3, and 22:6w 3. There was considerable variation in all of these fatty acids in both neutral and polar lipid from both tissues. In the flesh, the 20:4 w 6 was consistently higher in the neutral lipid than in the polar lipid. The total 20:5w 3 plus 22:6w 3 was consistently higher in polar lipid than in the neutral lipid. Thus, in spite of the major fluctuations in fatty acids caused by changes in diet and temperature throughout the seasons, there was a consistent preferential incorporation of PUFA of the w 3 series into the polar or phospholipid fraction of the lipids.
One of the best clues to the EFA requirements of a species can be gained from the fatty acid composition of the lipids incorporated into the offspring or egg. The act of reproduction or spawning also has a significant effect on the seasonal fluctuation of lipids in fish. Fatty acid composition of fish egg lipids is probably distinctive for each species and contains increased levels of 16:0, 20:4 w 6, 20:5 w 3 and 22:6w 3 compared to the liver lipids of the same female fish (Ackman, 1967).
Elevated levels of 16:0, 20:5w 3, and 22:6 w 3 and reduced 18:1 in the ovary occurred compared to mesenteric fat of Pacific sardine fed a natural copepod diet. The blood fatty acids of the sardine fed the natural diet were similar to those of the ovary. When the sardines were fed trout food, both the blood and mesenteric fat responded to the diet with elevated 18:2w 6 and reduced 20:5w 3 arid 22:6w 3. The effect of the diet on ovary fatty acid content was considerably less, as relatively high levels of 20:5w 3 and 22:6w 3 were retained.
The ovary lipids of the sweet smelt show an increase in 16:0, and a reduction in the PUFA, especially in the phospholipids, compared to the lipids from the flesh of fish caught at the same time of year. The w 6/w 3 ratio of the ovary was lower than that of the flesh lipids, 0.21 and 0.17 for the ovary compared to 0.31 and 0.20 for the triglycerids and phospholipids of the flesh, respectively.
The hatchability of eggs from common carp fed several different formulated feeds is greatly reduced when the 22:6w 3 of the egg lipids is less than 10 percent. Further, the muscle, plasma, and erythrocyte fatty acid compositions are more affected by dietary lipid than those of the eggs.
The EFA requirements of a number of species of fish have been investigated in nutritional studies. The fish themselves have given ample evidence for EFA preference by the types of fatty acids they incorporate into their lipids. Fish, in general, tend to utilize w 3 over w 6. This is especially observed when the dietary lipids are high in w 6, as the fish tend to alter the w 6/w 3 ratio toward the w 3 fatty acids in the tissue lipids. The lipids of the egg must satisfy the EFA requirement of the embryo until it is able to feed. The fatty acid composition data suggest that the w 3 requirement is greater in seawater than in freshwater and higher in cold water than in warm water.
Detailed information is still lacking on the dietary lipid requirements of many species of fish, but there is an abundance of information on the fatty acid composition of fish oils. Information on the lipid composition of fish can be used to make some guesses about dietary lipid requirements. Linolenic acid (18:3w 3) resulted in some sparing action and growth promotion in rats, and fatty acids of the w 6 EFA prevented all of the EFA-deficiency symptoms. Research with homeothermic land-dwelling animals showed that the w 6 series of fatty acids are the "essential fatty acids", while the w 3 series are considered to be non-essential or only have a partial sparing action on EFA-deficiency. The w 6 series of fatty acids have been shown to be essential to enough species that it began to become accepted that these are the essential fatty acids for all animals.
It was assumed by many that fish also required w 6 fatty acids. Many researchers began by supplementing fish diets with vegetable oils, such as corn, peanut, or sunflower oil, which were rich in linoleic acid. The main sympton observed during the development of EFA deficiency in chinook salmon fed fat-free diets was a marked depigmentation that can be prevented by addition of 1 percent trilinolein, but not by 0.1 percent linolenic acid.
Although the w 6 fatty acids are considered to be essential, one of the general characteristics of fish oils is the low levels of w 6 series fatty acids and the higher levels of w 3 type fatty acids. There is evidence that polyunsaturated fatty acids (PUPA) of the w 3 series, which are present in relatively large concentrations in fish oil, play the role of essential fatty acid for fish.
When a test diet containing 13 percent corn oil and 2 percent cod-liver oil was fed to rainbow trout, subsequent deletion of the cod-liver oil from the diet depressed growth and produced some kidney degeneration which might be attributed to a lack of sufficient w 3 PUFA present in significant quantities in cod-liver oil (McLaren et al., 1947). Dietary fish oil is superior to corn oil in promoting growth of rainbow trout (Salmo gairdneri) and the yellow-tail (Seriola guingueradiata). Dietary linolenic acid or ethyl linolenate (18:3 w 3) gives a positive growth response for rainbow trout which may be attributed to a dietary requirement for w 3 fatty acids.
5.1 Rainbow Trout
5.2 Channel Catfish
5.3 The Common Carp
5.4 The Eel
5.5 The Plaice
5.6 The Turbot
5.7 The Red Sea Bream
5.8 Other Species
One of the most widely accepted theories explaining the presence of such high levels of 20:5w 3 and 22:6w 3 fatty acids in fish oils is related to the effect of unsaturation on the melting point of a lipid. The greater degree of unsaturation of fatty acids in the fish phospholipids allows for flexibility of cell membrane at lower temperatures. The w 3 structure allows a greater degree of unsaturation than the w 6 or w 9. This theory is consistent with the fact that cold water fish have a greater nutritional requirement for w 3 fatty acids, while the EFA requirement of some warm water fish can be satisfied by a mixture of w 6 plus w 3.
Rainbow trout, a cold water fish, requires w 3 fatty acids as EFA in the diet. The EFA requirement can be met by 1 percent 18:3w 3 in the diet. Inclusion of 18:2w 6 in the diet may result in some improvement in growth and feed conversion compared to EFA deficient diets; however, the w 6 fatty acids will not prevent some EFA deficiency symptoms such as the "shock syndrome". Although it is clear that rainbow trout require w 3 fatty acids, it remains to be shown conclusively whether some dietary level of w 6 fatty acid is essential.
In all the above studies with rainbow trout, dietary 18:2w 6 or 18:3w 3 were readily converted to C-20 and C-22 PUFA of the same series, and 18:3w 3 or 22:6w 3 had similar EFA value for rainbow trout. Either 20:5w 3 or 22:6 w 3 is superior to 18:3w 3 in an EFA value for rainbow trout, and the former two fatty acids in combination are superior to either alone. This is consistent with data for mammals, where 20:4w 6 has higher EFA value than 18:2w 6. The superior nutritional value of C-20 and C-22 carbon w 3-PUFA is further supported by the excellent growth promoting effects of dietary fish oils such as pollock liver oil and salmon oil for rainbow trout.
One of the most important warm water fish in North America is the channel catfish (Ictalurus punctatus). The quantitative EFA requirement of the catfish has not yet been determined. However, the evidence is convincing that the w 3 requirement is not as great as that of rainbow trout. Analysis of fatty acids of lipids from catfish purchased at five processing plants showed very low levels of 20:4w 6, 20:5w 3, and 22:6w 3; 0.8 - 5.5, 0.2 - 1.3, and 0.6 - 6.1 percent of the total fatty acids, respectively. It was shown that corn oil added to a semipurified casein based diet initially resulted in a positive growth response and protein sparing, but later growth inhibition was observed. The apparent repressive effects of corn oil may be due to its 18:2 w 6 content since 20:5 w 3 and 22:6w 3 present in menhaden oil had no apparent detrimental effects. The growth suppressing effects of 18:2w 6 were also noted when 3 percent corn oil was added to 3 percent beef tallow and 3 percent menhaden oil. The growth suppression caused by unsaturated fatty acids does not appear to be limited to w 6 fatty acids. Linseed oil (high in 18:3w 3) in the diet of catfish resulted in growth suppression similar to that caused by corn oil compared with dietary beef tallow, olive oil and menhaden oil.
The picture for another warm water fish, the common carp (Cyprinus carpio) is much clearer than that for the channel catfish. This fish has an EFA requirement for both w 3 and w 6 fatty acids. The best weight gains and feed conversions are obtained in fish receiving a diet containing both 1 percent 18:2 w 6 and 1 percent 18:3w 3. With the carp, 20:5w 3 and 22:6w 3 at 0.5 percent of the diet are superior to 1 percent of 18:3u3. Carp fed a fat-free, or EFA deficient, diet incorporated high levels of 20:3w 9 in their lipids, especially in the phospholipids.
The eel (Anguilla japonica), another warm water fish, has a requirement for both w 3 and w 6 fatty acids. Corn oil (high in w 6) and cod liver oil (high in w 3)in a 2:1 mixture are most favourable for the growth of eels. The eel requires w 6 and w 3 in the same proportion as the carp, but at a lower level in the diet; namely, 0.5 percent of each, rather than 1.0 percent of each PUFA.
The plaice becomes depleted of both w 3 and w 6 PUFA when fed a fat-free diet. The addition of 12:0 and 14:0 to the diet result in the synthesis of saturated and monoenoic fatty acids of chain lengths up to C18; however, increased levels of 20:3w 9 noted in trout and mammals have not been reported in plaice. Plaice fed dietary 18:2w 6 and 18:3w 3 will not produce significant amounts of 20:4w 6, 20:5w 3, or 22:6w 3.
The growth of turbot (Scophthalmus matimus) is much better with w 3 PUFA than with w 6 or saturated fat (hydrogenated coconut oil) in the diet. The turbot also appears to be unable to convert dietary 18:2 w 6 to 20:4w 6 when fed corn oil, or to convert endogenous 18:1w 9 to 20:3w 9 when fed the EFA deficient diet. Although it appears to have an EFA requirement for w 3 fatty acids such as are present in cod liver oil, this requirement is not satisfied by 18:3w 3. The chain elongation and desaturation of 18:lw 9, 18:2 w 6, or 18:3w 3 has been found to be very limited (3-15 percent) in turbot compared to the rainbow trout where 70 percent of the 18:3w 3 was converted to 22:6w 3. The required level of long-chain w 3 fatty acids for turbot is at least 0.8 percent of the diet.
The red sea bream (Chrysophyrys major) grows better when the dietary lipid is of marine origin (pollock residual oil) rather than a vegetable oil (such as corn oil). The EFA requirement of the red sea bream is not satisfied by either linoleic acid of corn oil or supplemented linolenate. A mixture of 20:5w 3 and 22:6w 3 supplemented to the corn oil diet has been shown to be effective in improving growth and condition of these fish. Thus, even in warm water, marine fish seem to require not just w 3 fatty acids but 0)3 fatty acids of 20 to 22 carbon-chain length. A direct correlation between feed efficiency and the 18:1 level in the lipids of the red sea bream has been postulated.
Among warm water marine fish, mullet and fundulus possess the ability to chain, elongate, and desaturate 18:2w 6 or 18:3w 3 PUFAs. The process is, however, inhibited in fundulus by high levels (about 5 percent) of these PUFAs of 18:2w 6 or 18:3 w 3 in the diet.
It appears that high levels of 18-carbon w 6 or w 3 fatty acids inhibit the synthesis and metabolism of 18:lw 9. It is interesting to note that the channel catfish, which also exhibits negative growth response to dietary 18:2w 6 or 18:3w 3, incorporates very high levels of 18:1 into its body lipids. The inclusion of either 18:2w 6 or 18:3w 3 in the diet reduces the levels of 18:1 fatty acids in body lipids. A similar reduction has also been observed in red sea bream liver phospholipid when either of the PUFAs is added to the diet.
The competitive inhibition of chain elongation and desaturation of members from one series of fatty acids for members of another series is well established, with w 3 > w 6 > w 9 being the usual order of potency for inhibition.
The pathways of fatty acid metabolism have been reviewed by Mead and Kayama (1967). Fish are able to synthesize, de novo from acetate, the even-chain, saturated fatty acids, as shown in Figure 1. Radio tracer studies have shown that fish can convert 16:0 to the w 7 monoene and 18:0 to the w 9 monoene. The w 5, w 11 and w 3 monoenes are proposed based on the identification of these isomers in the monoenes of herring oil.
Fish are unable to synthesize any fatty acids of the w 6 and u3 series unless a precursor with this w structure is present in the diet. Fish are able to desaturate and elongate fatty acids of the w 9, w 6, or w 3 series as outlined in Figure 1. There is competitive inhibition of the elongation desaturation of fatty acids of one series by members of the other series. The w 3 fatty acids are the most potent inhibitors, the w 9 are the least. The ability to elongate and desaturate fatty acids is not the same in all species of fish, as was noted earlier. The turbot was able to desaturate and elongate only 3-15 percent of 18:1w 9, 18:2w 6, or 18:3 w 3, when given the C14 labelled fatty acid; in the rainbow trout, 70 percent of the label from 18:3w 3 (C14) was found in 22:6 w 3.
The essential fatty acids are not unique in their ability to supply energy. The b -oxidation of fatty acids in fish is basically the same as in mammals. The EFA and saturated and monoenoic fatty acids are all equally utilized by fish for energy production.
Fig. 1 Flow diagram for fatty acid synthesis mechanisms in fish - Saturated and monoenoic fatty acids (Adapted from Castell, 1979)
Fig. 1 Flow diagram for fatty acid synthesis mechanisms in fish - Polyunsaturated fatty acids (Adapted from Castell, 1979)
Increased swelling rates of liver mitochondria occur in rainbow trout fed diets deficient in w 3 fatty acids. It is possible that EFA plays an important role in the permeability as well as the plasticity of membranes. The role of w 3 fatty acids in membrane permeability may be one of the factors accounting for differences in content of this family of fatty acids between freshwater and marine fish.
Fish mitochondria with high levels of the w 3 PUFA and very low levels of w 6 fatty acids are very similar to mammalian mitochondria with respect to cytochrome content, b -oxidation of fatty acids, operation of the tricarboxylic acid cycle, electron transport, and oxidative phosphorylation. The w 3 PUFA may play the same role in fish that the w 6 fatty acids play in rats. The EFA play another role in the mitochondria. In addition to their importance in membrane structure, the EFA are important in some enzyme systems.
Unsaturated fatty acids play an important role in the transportation of other lipids. It has been repeatedly shown that feeding PUFA will lower the cholesterol levels in animals with above-normal blood lipid and cholesterol levels. Fish oils are more effective in lowering cholesterol levels than are most dietary lipids. The major portion of the fatty acids absorbed across the intestinal mucosa are transported as protein-lipid complexes stabilized by phospholipids. The low body temperature in fish probably results in a greater importance for unsaturation in transport of lipids than in homeothermic animals.
The requirement by fish for PUFA of the w 3 series creates problems with respect to feed storage. These types of fatty acids are very labile on oxidation. The products of lipid oxidation may react with other nutrients such as proteins, vitamins, etc., and reducing the available dietary levels or the oxidation products may be toxic. The effect of oxidized lipids on dietary proteins, enzymes and amino acids have been demonstrated.
The use of oxidized menhaden oil in the diets of swine and rats caused decreased appetite, reduced growth, yellowish-brown pigmentation of depot fat, and decreased haemoglobin and haematocrit levels. The negative effects of the oxidized fish oils were reversed by the addition of alpha-tocopherol acetate or ethoxyoquin to the diet.
Much of the use of vegetable oils in fish diets in the 1950s and 1960s might, in part, have been based on their greater stability in prepared diets. It has been demonstrated that rancid herring and hake meals in fish feeds caused dark colouration, anaemia, lethargy, brown-yellow pigmented liver, abnormal kidneys, and small gill clubbing in chinook salmon. The symptons can be alleviated by addition of alpha-tocopherol to the diets containing rancid fish meals. The addition of vitamin E would prevent the toxic or negative effects of adding 5 percent highly oxidized salmon oil to the diet of rainbow trout. This same sparing effect of alpha-tocopherol can also apply to rancid carp feed.
The positive nutritional value of w 3 fatty acids in fish lipids for fish feeds can become a negative factor if adequate care is not taken in the preparation and storage of feeds. Only fresh oils with low peroxide values should be used in feeds. Fish feed ingredients such as fish meals should be protected against oxidation. The level of vitamin E added to the diet should be increased as the PUFA level is increased. The finished feed, if possible, should be stored in air tight containers at reduced temperatures with minimum exposure to UV radiation and other factors accelerating the rate of lipid oxidation. The problems of rancidity or antioxidation of lipids in fish feeds should not be ignored.
Ackman, R.G., 1967 Characteristics of the fatty acid composition and biochemistry of some freshwater fish oils and lipids in comparison with marine oils and lipids. Comp.Biochem.Physiol., 22:907-22
Castell, J.D., 1979 Review of lipid requirements of finfish. In Finfish nutrition and fishfeed technology, edited by J.E. Halver and K. Trews. Proceedings of a World Symposium sponsored and supported by EIFAC/FAO/ICES/IUNS, Hamburg, 20-23 June, 1978. Schr.Bundesforschungsanst.Fisch.,Hamb., (14/15) vol.1: 59-84
Cowey, C.B. and J.R. Sargent, 1972 Fish nutrition. Adv.Mar.Biol., 10:383-492,
Cowey, C.B. and J.R. Sargent, 1977 Lipid nutrition in fish. Comp.Biochem.Physiol. (B Comp.Biochem.) 57:269-73
McLaren, B.A. 1947 et al., The nutrition of rainbow trout. 1. Studies of vitamin requirements. Arch-Biochem. 19:169-78
Meed, J.F. and M. Kayama, 1967 Lipid metabolism in fish. In Fish oils, edited by M.E. Stansby. Westport, Conn., Avi Publ. Co., pp. 289-99
National Research Council, 1973 Subcommittee on Fish Nutrition, Nutrient requirements of trout, salmon and catfish. Washington, D.C., National Academy of Sciences, (Nutrient requirements of domestic animals), 11:57 p.
National Research Council, 1977 Subcommittee on Warmwater Fishes, Nutrient requirements of warmwater fishes. Washington, D.C., National Academy of Sciences, (Nutrient requirements of domestic animals), 78 p.
Sinnhuber, R.O., 1969 The role of fats. In Fish in research, edited by O.W. Newhaus and J.E. Halver, New York, Academic Press, pp. 245-61
