142
C. Lee
trinsically resistant to biodegradation and thus may also be selectively preserved. It is also
possible that refractory organic or inorganic matrices may protect intrinsically labile organic substances (Knicker et al. 1996). A key difference among these protection mechanisms is that humification and selective preservation should result in pronounced changes
in organic structure, whereas protection does not require major compositional changes.
Hedges et al. (2001) used solid-state l3C NMR to investigate the composition of the
bulk organic carbon used in the analyses shown in Figs. 5.7 and 5.10. They found that
the major biochemical composition for bulk organic carbon was similar to that obtained by chromatographic analysis. This result fits more with the hypothesis of physical protection rather than selective preservation as described above. Several mechanisms could account for physical protection of organic matter by minerals. For example,
as mentioned earlier, some organic matter is incorporated into silicate and carbonate
tests during biological deposition of these minerals (Lowenstam and Weiner 1989). This
material is entombed within the mineral matrix and will not be exposed to decomposing enzymes until it is released by dissolution of the mineral phase. Organic matter may also be protected on the surface of mineral particles if adsorption of the organic matter into. Mayer (1994) postulated that adsorption of organic matter into
micropores of inorganic sedimentary material removes it physically from the action
of large hydrolyzing enzymes. Inorganic matter makes up most of the total mass of
sinking particles, so it would not be surprising if physical protection of the organic
fraction by mineral components occurred.
The idea that minerals preserve associated organic matter and the fact that they
are responsible for adding the ballast to particles that allows them to sink has suggested new concepts about organic matter fluxes in the ocean (Armstrong et al. 2002).
Typical mixes of organic carbon compounds have densities 1.1 times that of sea water,
while the density of typical inorganic mineral ballast is about 2.5 times that of sea water.
Since sinking velocities are proportional to the excess in density over that of the fluid
through which particles sink, a particle that is half organic matter and half inorganic
ballast will sink many times faster than a particle of comparable size composed totally of organic matter. Wind-blown dust particles, opaline silica (produced by diatoms and radiolarians), and calcium carbonate (produced by coccolithophorids and
foraminifera) are the major types of mineral ballast in the world ocean (Honjo 1996).
Dust, for example, does not dissolve appreciably with depth, so that organic matter
adsorbed to dust will be protected during its transit to the sea floor. In contrast, organic matter external or internal to opal and carbonate tests will be subject to decomposition as the biomineral dissolves. The mineral phase with which organic material
is associated may therefore affect organic carbon flux in two ways: through mineralspecific differences in the amount of organic carbon that can be protected per unit
ballast mineral, and through differential dissolution of the ballast minerals themselves.
These thoughts have been incorporated into a new model of organic matter decomposition in the ocean (Fig. 5.n; Armstrong et al. 2002). Considering mineral ballast and
protection, a quantitative description of POC remineralization must account both for
POC that is "protected" by its association with ballast and for POC that is "unprotected"
from degradation. Both types of POC are assumed to be associated with the same sinking aggregates (flocs and/or faecal pellets); the same ballast would then provide the
excess density needed for both types of carbon to sink. Only further research on organic-inorganic associations will allow progress in making quantitative and predic-
C. Lee
trinsically resistant to biodegradation and thus may also be selectively preserved. It is also
possible that refractory organic or inorganic matrices may protect intrinsically labile organic substances (Knicker et al. 1996). A key difference among these protection mechanisms is that humification and selective preservation should result in pronounced changes
in organic structure, whereas protection does not require major compositional changes.
Hedges et al. (2001) used solid-state l3C NMR to investigate the composition of the
bulk organic carbon used in the analyses shown in Figs. 5.7 and 5.10. They found that
the major biochemical composition for bulk organic carbon was similar to that obtained by chromatographic analysis. This result fits more with the hypothesis of physical protection rather than selective preservation as described above. Several mechanisms could account for physical protection of organic matter by minerals. For example,
as mentioned earlier, some organic matter is incorporated into silicate and carbonate
tests during biological deposition of these minerals (Lowenstam and Weiner 1989). This
material is entombed within the mineral matrix and will not be exposed to decomposing enzymes until it is released by dissolution of the mineral phase. Organic matter may also be protected on the surface of mineral particles if adsorption of the organic matter into. Mayer (1994) postulated that adsorption of organic matter into
micropores of inorganic sedimentary material removes it physically from the action
of large hydrolyzing enzymes. Inorganic matter makes up most of the total mass of
sinking particles, so it would not be surprising if physical protection of the organic
fraction by mineral components occurred.
The idea that minerals preserve associated organic matter and the fact that they
are responsible for adding the ballast to particles that allows them to sink has suggested new concepts about organic matter fluxes in the ocean (Armstrong et al. 2002).
Typical mixes of organic carbon compounds have densities 1.1 times that of sea water,
while the density of typical inorganic mineral ballast is about 2.5 times that of sea water.
Since sinking velocities are proportional to the excess in density over that of the fluid
through which particles sink, a particle that is half organic matter and half inorganic
ballast will sink many times faster than a particle of comparable size composed totally of organic matter. Wind-blown dust particles, opaline silica (produced by diatoms and radiolarians), and calcium carbonate (produced by coccolithophorids and
foraminifera) are the major types of mineral ballast in the world ocean (Honjo 1996).
Dust, for example, does not dissolve appreciably with depth, so that organic matter
adsorbed to dust will be protected during its transit to the sea floor. In contrast, organic matter external or internal to opal and carbonate tests will be subject to decomposition as the biomineral dissolves. The mineral phase with which organic material
is associated may therefore affect organic carbon flux in two ways: through mineralspecific differences in the amount of organic carbon that can be protected per unit
ballast mineral, and through differential dissolution of the ballast minerals themselves.
These thoughts have been incorporated into a new model of organic matter decomposition in the ocean (Fig. 5.n; Armstrong et al. 2002). Considering mineral ballast and
protection, a quantitative description of POC remineralization must account both for
POC that is "protected" by its association with ballast and for POC that is "unprotected"
from degradation. Both types of POC are assumed to be associated with the same sinking aggregates (flocs and/or faecal pellets); the same ballast would then provide the
excess density needed for both types of carbon to sink. Only further research on organic-inorganic associations will allow progress in making quantitative and predic-