Jökull - 01.12.1990, Síða 95

where é,sOc is the él80 value of CO2 released from
CaC03 by reaction with H3PO4 at 25°C, and élsOw is
the (5180 value of CO2 in equilibrium with the water
at 25°C.
Prior to interpretation of é,sO as indicator of pa-
leotemperatures the following questions must be con-
sidered:
1- Did the CaC03 deposit in oxygen isotopic equi-
librium with the ocean?
2. What was the oxygen isotopic composition of the
ocean at the time of deposition?
3. Has the original oxygen isotopic record been pre-
served?
Earlier studies have shown that many organisms
and among them molluscs, deposit CaC03 in isotopic
equilibrium with the ocean water. Knowledge of the
isotopic composition of the ocean water remains how-
ever a critical problem and must always be assumed.
Conceming the third question it has been observed
that exchange rates between CaC03 and water are ex-
tremely low at low temperatures and it is generally
accepted that the sample has retained its original iso-
topic record if its crystal stmcture is aragonite. Re-
crystallization of aragonite to calcite can however be
accompanied by isotopic exchange with the pore wa-
ter.
Several fossil shells from marine sediments in Ice-
land have been analysed for é,sO and él3C in order
to study the sea-temperature history around Iceland
at the end of the last glaciation and to compare that
to the temperature information revealed in the Green-
land Ice. All of the shells were radiocarbon dated at the
AMS facilities at the Institute of Physics, University
of Aarhus (Sveinbjömsdóttireí al., 1989) (Table I).
Earlier studies have shown that fluctuations in ma-
rine oxygen isotope records result in part from changes
in global ice volume, which affects the élsO of the
oceans, and in part from changes in ocean temper-
ature (Labeyrie et al., 1987). As shown in Table I
the é13C-values of our samples lie all well above —2
promille, which is the conventional division value be-
tween brackish and sea water molluscs (Weber, 1967).
Consequently no fresh water contamination is obvious
in the present data and hence in the paleotemperature
calculations from equation 1 the é IS0 of the ocean is
assumed to be 0 promille. It is however clear that
future work must include studies aimed at evaluating
the salinity of the ancient sea-water which accordingly
leads to better evaluation of its é180 value. For that
purpose careful studies of the sediments, foraminifera
and molluscs species sensitive to salinity may be the
key factors.
Fig. 6 compares the calculated paleotemperature
data to the é180-curve from the Dye-3 core in Green-
land. Most of the shells represent late Alleröd and
Preboreal age. The calculated temperatures are gener-
ally in good agreement with the ice core data as far as
the end of Alleröd, late Younger Dryas and Preboreal
parts are considered. Some time difference may exist
between the first indications of the cool Younger Dryas
period within the Greenland ice (which responds to
climatic variations instantaneously) and in geological
formations in Iceland. It is however emphasized that
the beginning of Younger Dryas in the Greenland ice is
not accurately determined. As seen in Fig. 6 one shell
from Seltjarnames (AAR 124) falls within Younger
Dryas, however the error bars for that age determi-
nation are large (±330) and it is suggested that the
measured age of the shell is too high and it should ac-
cordingly fall on the Preboreal side of Younger Dryas.
The youngest shells shown on Fig. 6 are relatively
cold and might represent the cold spell suggested by
the Gerzensee profile at about 9500 BP (Fig. 3).
It can be questioned whether the accuracy of the
radiocarbon datings is sufficient for such a detailed
interpretation of the present data. In order to confirm
the temperature curve on Fig. 6 more measurements of
well defined mollusc-profiles are required. It is how-
ever clear from the present data that shells from late
glacial time around Iceland have lived under variable
temperature conditions and reflect therefore the cli-
matic oscillations in Iceland at that time. Accordingly
the results indicate strongly that lsO measurements of
molluscs may become an important tool to reveal the
sea-temperature history around Iceland.
CONCLUSIONS
The oldest sea-sediments in Iceland dated so far are
about 12.500 C-14 years and are situated in the Borg-
JÖKULL, No. 40, 1990 91