Terrestrial Organic Matter in Marine Sediments
567
upwelling intensity are also evident. The step-like
increase in TOC throughout stage 12 suggests
three periods of enhanced marine organic fluxes.
Organic petrologic data closely follow these three
periods showing highest proportions of marine OM
(65-70%) during periods of maximum productivity
(Fig. 12). ol3C org' in contrast, does not parallel this
trend but remains at similar levels throughout glacial and interglacial conditions. With regard to eolian
supply of terrigenous OM to Site 663 it is interesting to note that the transition from glacial to interglacial conditions is characterized by a continuous
decrease in HI and marine OM (based on organic
petrology) whereas TOC persists at high levels
(Fig. 12). This pattern suggests that paleoupwelling already started to decrease during glacial stage 12 reaching minimum intensities at the
beginning of interglacial stage 11. Hence, an alternative mechanism is required to explain high TOC
values at the end of stage 12. We favor a final pulse
of eolian material causing a strong terrigenous overprint atthe end of isotopic stage 12. Records from
organic petrology, Rock-Eval pyrolysis and
phytholith accumulation support this assumption
(Fig. 12).
Analysis of a Late Pleistocene core located
below the central Equatorial Divergence showed
that pollen and spores are rare and their
acculumulation rates are extremely low in these
sediments (Marret 1994). There is a steep gradient of sporomorph fluxes away from the continent
reflecting the decreasing transport energy of the
trade winds with increasing transport distance.
Finer terrestrial detritus and smaller particles may
therefore have a wider range. Instead of a position below the central Equatorial Divergence, we
chose site 776 located nearer to the continent and
still under the winter dust plume for further investigation (Fig. Sa; Jahns et al. subm). Accumulation
and percentages of sporomorphs in core 776 reveal a drop at the transition from glacial to interglacial stages (Fig. 13). Site 776 and other records
from the eastern tropical Atlantic register an early
increase of the rain forest vegetation at about 435
ka BP (Dupont and Agwu 1992; Dupont 1993;
Dupont et al. 1998). The accumulation rates oftotal pollen in core 776 decrease because of the much
lower pollen production by rain forest trees as compared to grasses. As a result, the proportion of rain
forest pollen increases but those of grass pollen remain more or less constant (Jahns et al. 1998). That
the proportion of grass pollen does not decrease is
attributed to additional eolian pollen transport.
Palynological records thus indicate a persistently
high eolian supply throughout glacial stage 12 which
abruptly decreases towards the beginning of isotopic stage 11.
From the previous discussion on dust source areas reaching the eastern and central Equatorial
Atlantic it has been expected that OM in glacial
deposits contains some admixture of C4-debris.
Phytoliths which typically reveal increased accumulation rates during glacial stages of the late
Quaternary (Pokras and Mix 1985; DeMenocal et
al. 1993; Fig. 12) and the grass pollen record from
core 776 (Jahns et al. 1998; Fig. 13) bear witness
of glacial supply of grass-type vegetation. The
phytholith record of Site 663 is fairly well paralleled
by the ratio of inertinite/vitrinite (Fig. 12), meaning
that peaks in phytoliths are reflected by increased
proportions of oxidized OM. Charcoal sensu strictu
was rarely identified in samples of Sites 663 and
959 (Wagner 1999) although it was anticipated with
regard to observations made by Verardo and
Ruddiman (1996). However, it cannot be excluded
that inertinite and some partly oxidized fraction of
vitrinite in a broader sense belong to this type of
OM. Total contents of inertinite and inertinite plus
vitrinite reveal maximum ranges between 4-33%
and 13-67%, respectively. In comparison, estimates
of "charcoal" presented by Verardo and Ruddiman
(1996) of 50 to 80% during Stage 2 to 6 appear
much too high which is probably related to the analytical technique. Nevertheless, the general statement ofVerardo and Ruddiman (1996), that airborn terrigenous OM essentially contributes to the
bulk organic carbon preserved in open Equatorial
Atlantic deposits, is supported by our results.
In analogy to modern conditions grass pollen
records of stage 12 support the hypothesis that C4plant material contributes to the bulk OM. Since
there is little variability in ol3C O ,g recorded in
sediments of stages 12 and II (Fig. 12) it is assumed that wind-driven C4-debris might have
masked the marine isotopic signal. Estimated maximum fractions of C4-debris range from zero to
567
upwelling intensity are also evident. The step-like
increase in TOC throughout stage 12 suggests
three periods of enhanced marine organic fluxes.
Organic petrologic data closely follow these three
periods showing highest proportions of marine OM
(65-70%) during periods of maximum productivity
(Fig. 12). ol3C org' in contrast, does not parallel this
trend but remains at similar levels throughout glacial and interglacial conditions. With regard to eolian
supply of terrigenous OM to Site 663 it is interesting to note that the transition from glacial to interglacial conditions is characterized by a continuous
decrease in HI and marine OM (based on organic
petrology) whereas TOC persists at high levels
(Fig. 12). This pattern suggests that paleoupwelling already started to decrease during glacial stage 12 reaching minimum intensities at the
beginning of interglacial stage 11. Hence, an alternative mechanism is required to explain high TOC
values at the end of stage 12. We favor a final pulse
of eolian material causing a strong terrigenous overprint atthe end of isotopic stage 12. Records from
organic petrology, Rock-Eval pyrolysis and
phytholith accumulation support this assumption
(Fig. 12).
Analysis of a Late Pleistocene core located
below the central Equatorial Divergence showed
that pollen and spores are rare and their
acculumulation rates are extremely low in these
sediments (Marret 1994). There is a steep gradient of sporomorph fluxes away from the continent
reflecting the decreasing transport energy of the
trade winds with increasing transport distance.
Finer terrestrial detritus and smaller particles may
therefore have a wider range. Instead of a position below the central Equatorial Divergence, we
chose site 776 located nearer to the continent and
still under the winter dust plume for further investigation (Fig. Sa; Jahns et al. subm). Accumulation
and percentages of sporomorphs in core 776 reveal a drop at the transition from glacial to interglacial stages (Fig. 13). Site 776 and other records
from the eastern tropical Atlantic register an early
increase of the rain forest vegetation at about 435
ka BP (Dupont and Agwu 1992; Dupont 1993;
Dupont et al. 1998). The accumulation rates oftotal pollen in core 776 decrease because of the much
lower pollen production by rain forest trees as compared to grasses. As a result, the proportion of rain
forest pollen increases but those of grass pollen remain more or less constant (Jahns et al. 1998). That
the proportion of grass pollen does not decrease is
attributed to additional eolian pollen transport.
Palynological records thus indicate a persistently
high eolian supply throughout glacial stage 12 which
abruptly decreases towards the beginning of isotopic stage 11.
From the previous discussion on dust source areas reaching the eastern and central Equatorial
Atlantic it has been expected that OM in glacial
deposits contains some admixture of C4-debris.
Phytoliths which typically reveal increased accumulation rates during glacial stages of the late
Quaternary (Pokras and Mix 1985; DeMenocal et
al. 1993; Fig. 12) and the grass pollen record from
core 776 (Jahns et al. 1998; Fig. 13) bear witness
of glacial supply of grass-type vegetation. The
phytholith record of Site 663 is fairly well paralleled
by the ratio of inertinite/vitrinite (Fig. 12), meaning
that peaks in phytoliths are reflected by increased
proportions of oxidized OM. Charcoal sensu strictu
was rarely identified in samples of Sites 663 and
959 (Wagner 1999) although it was anticipated with
regard to observations made by Verardo and
Ruddiman (1996). However, it cannot be excluded
that inertinite and some partly oxidized fraction of
vitrinite in a broader sense belong to this type of
OM. Total contents of inertinite and inertinite plus
vitrinite reveal maximum ranges between 4-33%
and 13-67%, respectively. In comparison, estimates
of "charcoal" presented by Verardo and Ruddiman
(1996) of 50 to 80% during Stage 2 to 6 appear
much too high which is probably related to the analytical technique. Nevertheless, the general statement ofVerardo and Ruddiman (1996), that airborn terrigenous OM essentially contributes to the
bulk organic carbon preserved in open Equatorial
Atlantic deposits, is supported by our results.
In analogy to modern conditions grass pollen
records of stage 12 support the hypothesis that C4plant material contributes to the bulk OM. Since
there is little variability in ol3C O ,g recorded in
sediments of stages 12 and II (Fig. 12) it is assumed that wind-driven C4-debris might have
masked the marine isotopic signal. Estimated maximum fractions of C4-debris range from zero to