236
Mackensen and Bickert
glacial ocean distribution patterns constructed by
using a simple nutrient analogy of either Cd oro\3C
are not compatible, especially in the Southern
Ocean (Boyle and Rosenthal 1996). Glacial maximum benthic O\3C values suggest that Southern
Ocean deepwater had higher nutrients than even
the eastern tropical Pacific (Oppo et al. 1990),
whereas Cd data suggest a situation more similar
to today' s distribution pattern with highest nutrient
concentrations in the Pacific (Boyle 1992; Boyle
1994).
One possible explaination to solve contradictory
proxy results in the LGM Southern Ocean
became apparent when for the first time live
F. wuellerstorfi 013C and the contemporaneous
bottom water 013C LC02 from Southern Ocean sites
were sampled and measured from the same multiple-corer subcore (Macken sen et aJ. 1993). It
turned out that similar to the artifact discovered
earlier in infaunal living Uvigerina peregrina
(Zahn et aJ. 1986), F. wuellerstorfi, although
preferrentially living epifaunal, may calcify its tests
in 13C-depleted water. It was suggested that in
areas with strongly seasonal production where
particulate organic matter is rapidly deposited, a
phytodetritus layer (BiJlett et al. 1983; Gooday 1988;
Thiel et al. 1989) is developed that causes a 013C
gradient within this layer above the actual sediment/
water interface. Consequently, the steep o\3C gradient in the uppermost sediment would extend into
the phytodetritus layer. This layer affects epibenthic
foraminifera in a twofold way: (I) the food supply
triggers chamber building and reproduction (cf.
Gooday and Turley 1990; Corliss and Silva 1993;
Ohga and Kitazato 1997) and (2) calcification will
take place in times of very strong 013C gradients
at the sediment/water interface. This process may
cause an effect for F. wuellerstorfi in areas of seasonally high particulate organic matter fluxes,
which we prefered to call "phytodetritus effect" but
now commonly is tenned "Mackensen effect" (e.g.
Boyle et al. 1995; McCorkle et aJ. 1997; Raymo et
al. 1997). This explanation was corroborated by a
strong Ol3C o'g enrichment at the sediment/water interface occurring at sites with high organic carbon
fluxes in the vicinity of the Subantarctic Front and
the Polar Front (Wefer and Fischer 1991;
Mackensen et al. 1993). Parts of these isotopic
changes are attributed to high benthic respiration
rates in the fluffy layer producing CO 2 that is depleted in 13C relative to the supplied phytodetritus
(Deniro and Epstein 1978). High amounts of 13 C
depleted CO 2 released in the fluffy layer may thus
alter the isotopic composition of the 013C rc02
therein and hence may influence the Ol3C of
F. wuellerstorfi.
The doubts the Mackensen et al. study casted
on the reliability of paleoceanographic reconstruction based on benthic Ol3C values were at least
partly contradicted by a study carried out on Pacific sites (McCorkle and Keigwin 1994). In the
western equatorial Pacific on the Ontong Java Plateau and on the Emporer Seamount in the northwest Pacific F. wuellerstorfi faithfully records
bottom water 013C LC02 values. In a later study by
the same first author, however, these foraminiferal
O\3C values in 1994 just accepted to represent reliable bottom water 013C LC02 values, were suspected to be significantly biased by possible calcite
dissolution effects (McCorkle et al. 1995). A summary of the findings concerning epibenthic 013C
values thus far may be formulated following Boyle
and Rosenthal (1996): "F. wuellerstorfi is a reliable indicator of bottom waterol3CLC02 under conditions of low and moderate productivity", but, we
may add, it shows significant negative shifts under
regions of strongly seasonal productivity and
phytoplankton blooms with subsequent rapid sedimentation of phytodetritus and the development of
a fluffy layer on the sea floor (Macken sen et al.
1994).
To further corroborate this point we here
present an extended data set including new data
on live F. wuellerstorfi and other cibicidoids from
the southernmost Weddell Sea (Table I).
This data set is, compared to the Mackensen
et al. (1993) data, significantly extended by analyses oflive specimens and additional data from the
Antarctic continental margin in the southern
Weddell and Lazarev Seas. On the first glance it
clearly supports our former findings such that a
significant number of cibicidoid 013C values are
depleted by more than 0.2 %0 relative to ambient
bottom water o!3C LC02 values (Fig. 4).
Although, a more thorough analysis of the new
data reveals that few cibicidoid 013C values from
Mackensen and Bickert
glacial ocean distribution patterns constructed by
using a simple nutrient analogy of either Cd oro\3C
are not compatible, especially in the Southern
Ocean (Boyle and Rosenthal 1996). Glacial maximum benthic O\3C values suggest that Southern
Ocean deepwater had higher nutrients than even
the eastern tropical Pacific (Oppo et al. 1990),
whereas Cd data suggest a situation more similar
to today' s distribution pattern with highest nutrient
concentrations in the Pacific (Boyle 1992; Boyle
1994).
One possible explaination to solve contradictory
proxy results in the LGM Southern Ocean
became apparent when for the first time live
F. wuellerstorfi 013C and the contemporaneous
bottom water 013C LC02 from Southern Ocean sites
were sampled and measured from the same multiple-corer subcore (Macken sen et aJ. 1993). It
turned out that similar to the artifact discovered
earlier in infaunal living Uvigerina peregrina
(Zahn et aJ. 1986), F. wuellerstorfi, although
preferrentially living epifaunal, may calcify its tests
in 13C-depleted water. It was suggested that in
areas with strongly seasonal production where
particulate organic matter is rapidly deposited, a
phytodetritus layer (BiJlett et al. 1983; Gooday 1988;
Thiel et al. 1989) is developed that causes a 013C
gradient within this layer above the actual sediment/
water interface. Consequently, the steep o\3C gradient in the uppermost sediment would extend into
the phytodetritus layer. This layer affects epibenthic
foraminifera in a twofold way: (I) the food supply
triggers chamber building and reproduction (cf.
Gooday and Turley 1990; Corliss and Silva 1993;
Ohga and Kitazato 1997) and (2) calcification will
take place in times of very strong 013C gradients
at the sediment/water interface. This process may
cause an effect for F. wuellerstorfi in areas of seasonally high particulate organic matter fluxes,
which we prefered to call "phytodetritus effect" but
now commonly is tenned "Mackensen effect" (e.g.
Boyle et al. 1995; McCorkle et aJ. 1997; Raymo et
al. 1997). This explanation was corroborated by a
strong Ol3C o'g enrichment at the sediment/water interface occurring at sites with high organic carbon
fluxes in the vicinity of the Subantarctic Front and
the Polar Front (Wefer and Fischer 1991;
Mackensen et al. 1993). Parts of these isotopic
changes are attributed to high benthic respiration
rates in the fluffy layer producing CO 2 that is depleted in 13C relative to the supplied phytodetritus
(Deniro and Epstein 1978). High amounts of 13 C
depleted CO 2 released in the fluffy layer may thus
alter the isotopic composition of the 013C rc02
therein and hence may influence the Ol3C of
F. wuellerstorfi.
The doubts the Mackensen et al. study casted
on the reliability of paleoceanographic reconstruction based on benthic Ol3C values were at least
partly contradicted by a study carried out on Pacific sites (McCorkle and Keigwin 1994). In the
western equatorial Pacific on the Ontong Java Plateau and on the Emporer Seamount in the northwest Pacific F. wuellerstorfi faithfully records
bottom water 013C LC02 values. In a later study by
the same first author, however, these foraminiferal
O\3C values in 1994 just accepted to represent reliable bottom water 013C LC02 values, were suspected to be significantly biased by possible calcite
dissolution effects (McCorkle et al. 1995). A summary of the findings concerning epibenthic 013C
values thus far may be formulated following Boyle
and Rosenthal (1996): "F. wuellerstorfi is a reliable indicator of bottom waterol3CLC02 under conditions of low and moderate productivity", but, we
may add, it shows significant negative shifts under
regions of strongly seasonal productivity and
phytoplankton blooms with subsequent rapid sedimentation of phytodetritus and the development of
a fluffy layer on the sea floor (Macken sen et al.
1994).
To further corroborate this point we here
present an extended data set including new data
on live F. wuellerstorfi and other cibicidoids from
the southernmost Weddell Sea (Table I).
This data set is, compared to the Mackensen
et al. (1993) data, significantly extended by analyses oflive specimens and additional data from the
Antarctic continental margin in the southern
Weddell and Lazarev Seas. On the first glance it
clearly supports our former findings such that a
significant number of cibicidoid 013C values are
depleted by more than 0.2 %0 relative to ambient
bottom water o!3C LC02 values (Fig. 4).
Although, a more thorough analysis of the new
data reveals that few cibicidoid 013C values from