11. Lipids in Marine and Freshwater Organisms
277
important saturated fatty acid and, with 14:0 (myristic) and 18:0 (stearic) acid,
adds up to 20-40% of the total fatty acids. The monounsaturated fatty acids
similarly tend to be dominated by 18: 1 eo9 (oleic). Prior to 1980, most fatty acid
analyses were conducted on packed GLC columns and the oleic acid peak always
included 18:1eo7 (cis-vaccenic). Capillary GLC conveniently reveals many such
isomer details. As can be seen from Figure 11.4, some 16: 1eo7 is usually elongated
to 18: 1eo7 but is nevertheless often important by itself (Tables 11.1 and 11.2). It is
also rational to expect some 20:1eo9, elongated from 18:1eo9, and yet two
shoulders, for 20: 1 eo 11 (by de saturation of 20:0) or 20: 1 eo7 (by elongation from
18: 1 eo7), are also commonly observed. The 22: 1 family pose a unique problem in
that the dominant isomer is 22: 1eo11. In the North Atlantic and Pacific, this is
derived from the corresponding fatty alcohol in the wax esters of copepods as
already discussed. This alcohol is a structural anomaly in respect to biosynthesis
of the unusual bond position but is mostly oxidized to the corresponding fatty acid
on digestion by fish (Sargent et aI., 1979) and, together with 20:1, heavily influences the total monoethylenic fatty acids of species such as herring or capelin
feeding on these zooplankton (Ratnayake et aI., 1979a,b). In fact, in the production of fish oil and meal by the Atlantic fish meal production industry a ratio of
I,22: 1 > I,20: 1 in the oil, totaling 30-40% of fatty acids, is almost a sure sign of
herring as the source of a fish oil. Because 22: 1 and 20: 1 are so abundant, 18: 1 and
even 16: 1 may be present at reduced levels in these oils. The fish feeding directly
on phytoplankton that do not have the 20: 1 and 22: 1 of zooplankton, notably the
American menhaden (Brevoortia tyrannis) (Fig. 11.4), instead break down the
plant polysaccharides and synthesize additional 14:0 and 16:0. The common
assembly route to fish triacylglycerols as first proposed by Brockerhoff (Litchfield, 1972) is shown in Figure 11.5. The final insertion of fatty acids in the sn-3
position (sometimes marked as the a'-position) depends on whatever fatty acid is
circulating and available. Thus, if the menhaden finds 14:0 or 16:0 readily available, or the herring 20: 1 or 22: 1, or in some cases a polyunsaturated fatty acid is
available, then these will be inserted. The DHA is preferred by the phospholipid
intermediates for the 2-position in this process. This explains why 22:6eo3 may be
primarily in the 2-position and 20:5eo3, originally from phytoplankton, more
scattered between the 2-position and the 3-position (Ando et ai., 1992; Litchfield,
1972). Certain fish oils show an ability to "winterize" (Ackman, 1988). An oil
clear at 30°C may tum cloudy at 20°C, and on cooling to 15°C a solid layer
crystallizes out. The two fractions of the winterized oil do not differ radically in
most fatty acids, but the polyunsaturated fatty acids are higher in the liquid layer
(Ackman, 1980), suggesting that combinations of more than one of 20:5eo3 plus
22:6eo3 or plus 18:4eo3 may be present in a proportion of the molecules. Other
molecules may have an excessive proportion of 14:0 plus 16:0, promoting this
crystallization (Ackman, 1988). The sharp separation described is hindered if too
much 20: 1 and 22: 1 are present, and the whole mass of the oil may simply solidify.
Consideration of these matters and of recent assemblages of fish oil triacylglycerol compositions (Moffat, 1995; McGill and Moffat, 1992), and their stereochemistry (Ando et aI., 1992), suggests that there is one basic fatty acid composi-
277
important saturated fatty acid and, with 14:0 (myristic) and 18:0 (stearic) acid,
adds up to 20-40% of the total fatty acids. The monounsaturated fatty acids
similarly tend to be dominated by 18: 1 eo9 (oleic). Prior to 1980, most fatty acid
analyses were conducted on packed GLC columns and the oleic acid peak always
included 18:1eo7 (cis-vaccenic). Capillary GLC conveniently reveals many such
isomer details. As can be seen from Figure 11.4, some 16: 1eo7 is usually elongated
to 18: 1eo7 but is nevertheless often important by itself (Tables 11.1 and 11.2). It is
also rational to expect some 20:1eo9, elongated from 18:1eo9, and yet two
shoulders, for 20: 1 eo 11 (by de saturation of 20:0) or 20: 1 eo7 (by elongation from
18: 1 eo7), are also commonly observed. The 22: 1 family pose a unique problem in
that the dominant isomer is 22: 1eo11. In the North Atlantic and Pacific, this is
derived from the corresponding fatty alcohol in the wax esters of copepods as
already discussed. This alcohol is a structural anomaly in respect to biosynthesis
of the unusual bond position but is mostly oxidized to the corresponding fatty acid
on digestion by fish (Sargent et aI., 1979) and, together with 20:1, heavily influences the total monoethylenic fatty acids of species such as herring or capelin
feeding on these zooplankton (Ratnayake et aI., 1979a,b). In fact, in the production of fish oil and meal by the Atlantic fish meal production industry a ratio of
I,22: 1 > I,20: 1 in the oil, totaling 30-40% of fatty acids, is almost a sure sign of
herring as the source of a fish oil. Because 22: 1 and 20: 1 are so abundant, 18: 1 and
even 16: 1 may be present at reduced levels in these oils. The fish feeding directly
on phytoplankton that do not have the 20: 1 and 22: 1 of zooplankton, notably the
American menhaden (Brevoortia tyrannis) (Fig. 11.4), instead break down the
plant polysaccharides and synthesize additional 14:0 and 16:0. The common
assembly route to fish triacylglycerols as first proposed by Brockerhoff (Litchfield, 1972) is shown in Figure 11.5. The final insertion of fatty acids in the sn-3
position (sometimes marked as the a'-position) depends on whatever fatty acid is
circulating and available. Thus, if the menhaden finds 14:0 or 16:0 readily available, or the herring 20: 1 or 22: 1, or in some cases a polyunsaturated fatty acid is
available, then these will be inserted. The DHA is preferred by the phospholipid
intermediates for the 2-position in this process. This explains why 22:6eo3 may be
primarily in the 2-position and 20:5eo3, originally from phytoplankton, more
scattered between the 2-position and the 3-position (Ando et ai., 1992; Litchfield,
1972). Certain fish oils show an ability to "winterize" (Ackman, 1988). An oil
clear at 30°C may tum cloudy at 20°C, and on cooling to 15°C a solid layer
crystallizes out. The two fractions of the winterized oil do not differ radically in
most fatty acids, but the polyunsaturated fatty acids are higher in the liquid layer
(Ackman, 1980), suggesting that combinations of more than one of 20:5eo3 plus
22:6eo3 or plus 18:4eo3 may be present in a proportion of the molecules. Other
molecules may have an excessive proportion of 14:0 plus 16:0, promoting this
crystallization (Ackman, 1988). The sharp separation described is hindered if too
much 20: 1 and 22: 1 are present, and the whole mass of the oil may simply solidify.
Consideration of these matters and of recent assemblages of fish oil triacylglycerol compositions (Moffat, 1995; McGill and Moffat, 1992), and their stereochemistry (Ando et aI., 1992), suggests that there is one basic fatty acid composi-