2 \0
P.F. Landrum and S. W. Fisher
organic solvent combinations. Most frequently, the solvent chosen for the lipid
extraction is determined by that which most efficiently extracts the contaminant of
interest from the various matrices. This often leads to incomplete extraction,
sometimes in excess of 50%, of the total lipids and is generally a failure to extract
the polar lipids (Ewald, 1996). Further, the solvents specifically used for lipid
extraction sometime fail to completely extract the contaminants, particularly in
leaner organisms with a difference of nearly 40% in some cases (Ewald, 1996).
There has been little effort to intercalibrate the relative extractability of the
various solvent systems with a more conventional system, chloroform:methanol,
to extract lipids (Bligh and Dyer, 1959; Folch et aI., 1957). These extraction
differences may lead to substantial differences in comparing the lipid-normalized
bioconcentration across studies and among species of varying lipid composition.
We suggest that the substantial variability in lipid extractability among data sets
could be eliminated by standardizing to the Bligh-Dyer extraction scheme, which
uses a chloroform:methanol extracting solvent and is known to extract both polar
and nonpolar lipids (Randall et a!., 1991). It would not be necessary for all studies
to use a single extraction technique, but there should be an intercalibration between the extraction solvent of choice and the Bligh-Dyer scheme. This would
permit calculation to common units. This scheme can even be adapted for the
measurement of lipid content for small sample masses (Gardner et aI., 1985).
9.3.2, Lipid Composition and Bioaccumulation
The relative lipid compositions among species also contributes to the observed
variation in bioaccumulation potential. The generally held view is that neutral
storage lipids are the most important class for the bioaccumulation of nonpolar
contaminants. Thus, species with higher neutral lipid fractions would presumably
accumulate higher contaminant concentrations on a lipid-normalized basis. However, the evidence for the role of lipid composition is limited. Several studies have
focused on the relative extractability of the contaminant of interest and the corresponding lipid to suggest the role of lipid composition on bioaccumulation
(Ewald, 1996; de Boer, 1988; Schneider, 1981). Another approach examining the
role of lipid composition to compare across species and contaminant characteristics used different portions of the lipids as normalizing factors. When this is done,
prediction of bioaccumulation was better related to total lipid content than to any
one lipid fraction (Kucklick et aI., 1996; Stange and Swackhamer, 1994).
The relative distribution of the moderately lipophilic herbicide triallate was
investigated by using autoradiography (Arts et aI., 1995) and mass spectrometric
techniques (Arts et al., 1996) and showed that the contaminant resides primarily in
the storage lipids and in the highly lipophilic nervous system tissues. Although the
triallate was found in storage lipids as expected, the bioaccumulation did not
always correlate with the triglycerol content of the amphipods (Arts et aI., 1995).
This suggests a role for other lipids or other storage sites in some species that were
not detected by the above techniques.
P.F. Landrum and S. W. Fisher
organic solvent combinations. Most frequently, the solvent chosen for the lipid
extraction is determined by that which most efficiently extracts the contaminant of
interest from the various matrices. This often leads to incomplete extraction,
sometimes in excess of 50%, of the total lipids and is generally a failure to extract
the polar lipids (Ewald, 1996). Further, the solvents specifically used for lipid
extraction sometime fail to completely extract the contaminants, particularly in
leaner organisms with a difference of nearly 40% in some cases (Ewald, 1996).
There has been little effort to intercalibrate the relative extractability of the
various solvent systems with a more conventional system, chloroform:methanol,
to extract lipids (Bligh and Dyer, 1959; Folch et aI., 1957). These extraction
differences may lead to substantial differences in comparing the lipid-normalized
bioconcentration across studies and among species of varying lipid composition.
We suggest that the substantial variability in lipid extractability among data sets
could be eliminated by standardizing to the Bligh-Dyer extraction scheme, which
uses a chloroform:methanol extracting solvent and is known to extract both polar
and nonpolar lipids (Randall et a!., 1991). It would not be necessary for all studies
to use a single extraction technique, but there should be an intercalibration between the extraction solvent of choice and the Bligh-Dyer scheme. This would
permit calculation to common units. This scheme can even be adapted for the
measurement of lipid content for small sample masses (Gardner et aI., 1985).
9.3.2, Lipid Composition and Bioaccumulation
The relative lipid compositions among species also contributes to the observed
variation in bioaccumulation potential. The generally held view is that neutral
storage lipids are the most important class for the bioaccumulation of nonpolar
contaminants. Thus, species with higher neutral lipid fractions would presumably
accumulate higher contaminant concentrations on a lipid-normalized basis. However, the evidence for the role of lipid composition is limited. Several studies have
focused on the relative extractability of the contaminant of interest and the corresponding lipid to suggest the role of lipid composition on bioaccumulation
(Ewald, 1996; de Boer, 1988; Schneider, 1981). Another approach examining the
role of lipid composition to compare across species and contaminant characteristics used different portions of the lipids as normalizing factors. When this is done,
prediction of bioaccumulation was better related to total lipid content than to any
one lipid fraction (Kucklick et aI., 1996; Stange and Swackhamer, 1994).
The relative distribution of the moderately lipophilic herbicide triallate was
investigated by using autoradiography (Arts et aI., 1995) and mass spectrometric
techniques (Arts et al., 1996) and showed that the contaminant resides primarily in
the storage lipids and in the highly lipophilic nervous system tissues. Although the
triallate was found in storage lipids as expected, the bioaccumulation did not
always correlate with the triglycerol content of the amphipods (Arts et aI., 1995).
This suggests a role for other lipids or other storage sites in some species that were
not detected by the above techniques.