10
c.c. Parrish
of the extract is used, and precautions have to be taken against including in the
determination any nonlipid material that may have become entrained. To avoid
overestimating with gravimetry, care has to be taken to wash the extract properly,
to evaporate all solvent from the extract, and to minimize any contact with
oxygen, especially for highly unsaturated samples. Such samples may gain as
much as 3.5% in mass due to adsorption of oxygen (Kaitaranta and Ke, 1981).
Other precautions for gravimetric analyses are detailed in Wood (1991). Usually
in ecological studies, only small samples are available, so that microprocedures
such as those described by Gardner et al. (1985) are very useful.
Gravimetry is the most direct method for determining total lipids, provided
there is sufficient material present. If a spectrophotometer is available, colorimetric methods can be more convenient and more sensitive (Parsons et aI., 1989;
Barnes and Blackstock, 1973; Marsh and Weinstein, 1966). These methods tend to
rely on reactions with certain compounds or functional groups (e.g., carboncarbon double bonds: C=C) within the sample matrix and are thus prone to
biasing. In addition, the rate of reaction with those functional groups can vary
according to the type of compound in which they are found (Ahlgren and Merino,
1991). Intercalibration with gravimetry for different sample types has been recommended (Barnes and Blackstock, 1973). As with gravimetry, colorimetric
methods can also suffer from the possibility of inclusion of nonlipid compounds
in the estimation. The only certain way of avoiding this is to perform a chromatographic separation. In the process of separating the lipids from nonlipid contaminants, additional information on the nature of the lipids in the sample is also
usually obtained.
1.2.4. Determination of Lipid Classes
The heterogeneous nature of lipids means that much information can be gained by
quantifying individual classes (Delbeke et aI., 1995; Parrish, 1988). Although
there are some very convenient colorimetric kits available for plasma samples
(bioMerieux, France, or Boehringer Mannheim, Germany), these rarely are
directly applicable to tissue or seston samples (Barnes and Blackstock, 1973). For
these samples, the lipids usually have to be chromatographically subdivided into
classes. For this purpose, thin-layer chromatography (TLC) or HPLC can be used.
In TLC, a lipid extract is spotted at the bottom of a glass plate or quartz rod that
is covered in silica gel, and the lipids are developed with solvents of differing
polarities. After separation on TLC plates, the lipid classes can be visualized and
then scraped off for further analysis (Christie, 1989; Shaikh, 1986). However,
quantification of the separated material directly on the silica gel is much more
convenient and less prone to problems with contamination or recovery. Lipid class
quantification can be performed directly on silica gel-coated rods (Parrish, 1987)
or plates (Olsen and Henderson, 1989; Conte and Bishop, 1988); however, we
prefer to use rod TLC (Fig. 1.4) because in the Chromarod-Iatroscan system there
is a partial scanning facility that permits extensive analysis of a single sample on a
single rod (Fig. 1.5). By separating out all the lipid classes in this way, one obtains
c.c. Parrish
of the extract is used, and precautions have to be taken against including in the
determination any nonlipid material that may have become entrained. To avoid
overestimating with gravimetry, care has to be taken to wash the extract properly,
to evaporate all solvent from the extract, and to minimize any contact with
oxygen, especially for highly unsaturated samples. Such samples may gain as
much as 3.5% in mass due to adsorption of oxygen (Kaitaranta and Ke, 1981).
Other precautions for gravimetric analyses are detailed in Wood (1991). Usually
in ecological studies, only small samples are available, so that microprocedures
such as those described by Gardner et al. (1985) are very useful.
Gravimetry is the most direct method for determining total lipids, provided
there is sufficient material present. If a spectrophotometer is available, colorimetric methods can be more convenient and more sensitive (Parsons et aI., 1989;
Barnes and Blackstock, 1973; Marsh and Weinstein, 1966). These methods tend to
rely on reactions with certain compounds or functional groups (e.g., carboncarbon double bonds: C=C) within the sample matrix and are thus prone to
biasing. In addition, the rate of reaction with those functional groups can vary
according to the type of compound in which they are found (Ahlgren and Merino,
1991). Intercalibration with gravimetry for different sample types has been recommended (Barnes and Blackstock, 1973). As with gravimetry, colorimetric
methods can also suffer from the possibility of inclusion of nonlipid compounds
in the estimation. The only certain way of avoiding this is to perform a chromatographic separation. In the process of separating the lipids from nonlipid contaminants, additional information on the nature of the lipids in the sample is also
usually obtained.
1.2.4. Determination of Lipid Classes
The heterogeneous nature of lipids means that much information can be gained by
quantifying individual classes (Delbeke et aI., 1995; Parrish, 1988). Although
there are some very convenient colorimetric kits available for plasma samples
(bioMerieux, France, or Boehringer Mannheim, Germany), these rarely are
directly applicable to tissue or seston samples (Barnes and Blackstock, 1973). For
these samples, the lipids usually have to be chromatographically subdivided into
classes. For this purpose, thin-layer chromatography (TLC) or HPLC can be used.
In TLC, a lipid extract is spotted at the bottom of a glass plate or quartz rod that
is covered in silica gel, and the lipids are developed with solvents of differing
polarities. After separation on TLC plates, the lipid classes can be visualized and
then scraped off for further analysis (Christie, 1989; Shaikh, 1986). However,
quantification of the separated material directly on the silica gel is much more
convenient and less prone to problems with contamination or recovery. Lipid class
quantification can be performed directly on silica gel-coated rods (Parrish, 1987)
or plates (Olsen and Henderson, 1989; Conte and Bishop, 1988); however, we
prefer to use rod TLC (Fig. 1.4) because in the Chromarod-Iatroscan system there
is a partial scanning facility that permits extensive analysis of a single sample on a
single rod (Fig. 1.5). By separating out all the lipid classes in this way, one obtains