M. Hofmann et al., Influence of a [CO2(aq)] dependent biological C-isotope fractionation on glacial C-13/C-12 ratios in the ocean, GLOBAL BIOG, 13(4), 1999, pp. 873-883
Planktonic foraminiferal shells buried in deep ocean sediments record lower
delta(13)C values of surface water dissolved inorganic carbon during glaci
al times than during Holocene, In the Southern Ocean and at high northern l
atitudes, a drop of between 0.3 parts per thousand and 0.9 parts per thousa
nd is observed, whereas the mean change in low and middle latitudes is only
+/-0.15 parts per thousand. However, a stronger biological carbon pump suf
ficient to explain the 80 ppmv lower atmospheric pCO(2) values during glaci
al times would raise the surface ocean delta(13)C values of dissolved inorg
anic carbon by about 1.0 parts per thousand. Here the results of a three-di
mensional ocean circulation model study are presented which demonstrate tha
t the increase of delta(13)C values in the sea surface due to a strengtheni
ng of the biological carbon pump is counteracted by processes which drive t
he delta(13)C values in the opposite direction, This was found by performin
g simulations employing the three-dimensional Hamburg Model of the Oceanic
Carbon Cycle (HAMOCC) combined with a [CO2(aq)] dependent parameterization
of the biological carbon isotope fractionation. The difference in the biolo
gical carbon isotope fractionation between Glacial and Holocene is responsi
ble for a lowering of delta(13)C values in surface water dissolved inorgani
c carbon by about 0.3 parts per thousand. The additional effects of the gla
cially elevated CO32- concentration (0.25-0.50 parts per thousand) combined
with the 0.35 parts per thousand lowering of delta(13)C values for the who
le ocean due to a transfer of terrestrial organic carbon from the biosphere
to the ocean-atmosphere reservoir also contribute to a further delta(13)C
drop of 0.6-0.85 parts per thousand. Hence a small glacial decrease of the
planktonic foraminifera delta(13)C of the order of 0.25 parts per thousand
instead of an increase is predicted.