Stable Isotopes of Pteropod Shells as Recorders of Sub-Surface Water Conditions
193
value of past surface or subsurface water temperatures. In addition, certain large species such as C.
pyramidata may also reveal seasonal temperature
changes as shown by Kalbereret al. (1993). These
authors used oxygen isotope measurements of several sub-samples of single shells of C. pyramidata
which were attributed to certain seasons according to their life spans (Van der SpoeI1973); by this
means, seasonal sub-surface water temperatures
could be reconstructed from one single pteropod
shell.
The 813Ccomposition of pteropods seems to be
controlled by a combination of inherent (e.g. shell
growth) and environmental factors (e.g. temperature, 8 13 C of dissolved inorganic carbon). Grossman
and Ku (1986), for instance, reported an inverse
relationship between temperature and the 8 13 C of
aragonitic benthic foraminifera and benthic molluscs. This was also found by Ganssen (1983) and
Kalberer et al. (1993). As concluded by Grossman
and Ku (1986), this relationship may be used as a
basis for a paleothermometer but can provide only
a sensitivity of ± l.6°C. However, temperature
changes alone cannot explain the large carbon isotope variations found in skeletal aragonite
(Grossman and Ku 1986; Mook and Vogel 1968).
Obviously, the 8 13 C of pteropods, or other molluscs,
is strongly dependent on the 8 13 C of the dissolved
inorganic carbon in ambient sea water (e.g. Mook
and Vogel 1968; Grossman and Ku 1986; Jasper and
Deuser 1993) which in turn is controlled by photosynthesis prevalent in surface water and ocean circulation.
In this study, we compared a four-year oxygen
and carbon isotope record of the mesopelagic
pteropod L. injlata (juveniles) with a record produced by a planktonic foraminifera (G. ruber, white
variety) sampled by sediment traps at a mesopelagic
site off Cape Blanc (Mauretania). We discuss these
records in comparison to measured seasonal sea
surface temperatures (SST), organic carbon export
fluxes and variations in ocean circulation.
Material and Methods
Four sediment trap and current meter moorings
were located off Cape Blanc at about 20 o N, 200W
(CBI-4) (Fischer et al. 1996) and sampled from
22.03.88to 19.11.91 Water depths ranged from 3646
to 4108 m, sampling depths were around 730 m
(shallow traps CB3-4) and between 2195 and 3562
m (deep traps). For sample preparation and particle flux data see Fischer et al. (1996). Pteropod
preparation and isotope analysis are described in
detail by Kalberer et al. (1993). All pteropods larger
than 150lim were removed under the microscope
and washed briefly with fresh water and dried at
60°C. High numbers of pteropods (mainly L.
injlata) were found only during 1989. From this
year, counts of three size fractions of L. injlata
(150-500 lim, 500-1000 lim, > 1 mm) are available
(Kalberer et al. 1993). From the fraction> Imm
(adults), pteropods were manually removed from
the total amount of material, from the smaller fraction a wet split of 1164 was used for pteropod analysis. Planktonic foraminifera (G. ruber (w)) were
picked out by hand from a wet 1164 or V. split and
dried at 60°C. For G. ruber the fraction> 150 lim
was counted, stable isotope analysis was performed
on specimens ofthe size fraction ranging from 200
to 300 lim (Table 1).
Carbon and oxygen isotope measurements ofL.
injlata were performed on shells of about 250 lim
in diameter (juveniles) and weighing merely several
lig. Mostly, about 20-30 individuals were collected
and dissolved in 100% orthophosphoric acid at 75°C
and measured with a MAT 251 mass spectrometer
equipped with an automatic preparation line (KIEL
device). From the adult species (> lmm), one single individual was taken for isotope analysis. Results are reported in the standard 8-notation. Calibration of the laboratory standard to PDB was
achieved using NBS 18, 19 and 20 standards. Analytical precision was better than 0.07%0 fOr(; 18 0
and 0.05%0 for 8!3C based on internal carbonate
standards.
Water temperatures were calculated from the
oxygen isotope records of G. ruber and L. injlata
applying the paleotemperature equation established
by Epstein et al. (1953). 8 18 0 of sea water was
computed on the basis of its relationship to salinity.
We used the equation by Ganssen and Sarnthein
(1983) with a very shallow slope (8 18 0 SW = 0.11 S
- 3.15) which fits between the regression lines established for the eastern and western equatorial
Atlantic (Fairbanks et al. 1992). The following
193
value of past surface or subsurface water temperatures. In addition, certain large species such as C.
pyramidata may also reveal seasonal temperature
changes as shown by Kalbereret al. (1993). These
authors used oxygen isotope measurements of several sub-samples of single shells of C. pyramidata
which were attributed to certain seasons according to their life spans (Van der SpoeI1973); by this
means, seasonal sub-surface water temperatures
could be reconstructed from one single pteropod
shell.
The 813Ccomposition of pteropods seems to be
controlled by a combination of inherent (e.g. shell
growth) and environmental factors (e.g. temperature, 8 13 C of dissolved inorganic carbon). Grossman
and Ku (1986), for instance, reported an inverse
relationship between temperature and the 8 13 C of
aragonitic benthic foraminifera and benthic molluscs. This was also found by Ganssen (1983) and
Kalberer et al. (1993). As concluded by Grossman
and Ku (1986), this relationship may be used as a
basis for a paleothermometer but can provide only
a sensitivity of ± l.6°C. However, temperature
changes alone cannot explain the large carbon isotope variations found in skeletal aragonite
(Grossman and Ku 1986; Mook and Vogel 1968).
Obviously, the 8 13 C of pteropods, or other molluscs,
is strongly dependent on the 8 13 C of the dissolved
inorganic carbon in ambient sea water (e.g. Mook
and Vogel 1968; Grossman and Ku 1986; Jasper and
Deuser 1993) which in turn is controlled by photosynthesis prevalent in surface water and ocean circulation.
In this study, we compared a four-year oxygen
and carbon isotope record of the mesopelagic
pteropod L. injlata (juveniles) with a record produced by a planktonic foraminifera (G. ruber, white
variety) sampled by sediment traps at a mesopelagic
site off Cape Blanc (Mauretania). We discuss these
records in comparison to measured seasonal sea
surface temperatures (SST), organic carbon export
fluxes and variations in ocean circulation.
Material and Methods
Four sediment trap and current meter moorings
were located off Cape Blanc at about 20 o N, 200W
(CBI-4) (Fischer et al. 1996) and sampled from
22.03.88to 19.11.91 Water depths ranged from 3646
to 4108 m, sampling depths were around 730 m
(shallow traps CB3-4) and between 2195 and 3562
m (deep traps). For sample preparation and particle flux data see Fischer et al. (1996). Pteropod
preparation and isotope analysis are described in
detail by Kalberer et al. (1993). All pteropods larger
than 150lim were removed under the microscope
and washed briefly with fresh water and dried at
60°C. High numbers of pteropods (mainly L.
injlata) were found only during 1989. From this
year, counts of three size fractions of L. injlata
(150-500 lim, 500-1000 lim, > 1 mm) are available
(Kalberer et al. 1993). From the fraction> Imm
(adults), pteropods were manually removed from
the total amount of material, from the smaller fraction a wet split of 1164 was used for pteropod analysis. Planktonic foraminifera (G. ruber (w)) were
picked out by hand from a wet 1164 or V. split and
dried at 60°C. For G. ruber the fraction> 150 lim
was counted, stable isotope analysis was performed
on specimens ofthe size fraction ranging from 200
to 300 lim (Table 1).
Carbon and oxygen isotope measurements ofL.
injlata were performed on shells of about 250 lim
in diameter (juveniles) and weighing merely several
lig. Mostly, about 20-30 individuals were collected
and dissolved in 100% orthophosphoric acid at 75°C
and measured with a MAT 251 mass spectrometer
equipped with an automatic preparation line (KIEL
device). From the adult species (> lmm), one single individual was taken for isotope analysis. Results are reported in the standard 8-notation. Calibration of the laboratory standard to PDB was
achieved using NBS 18, 19 and 20 standards. Analytical precision was better than 0.07%0 fOr(; 18 0
and 0.05%0 for 8!3C based on internal carbonate
standards.
Water temperatures were calculated from the
oxygen isotope records of G. ruber and L. injlata
applying the paleotemperature equation established
by Epstein et al. (1953). 8 18 0 of sea water was
computed on the basis of its relationship to salinity.
We used the equation by Ganssen and Sarnthein
(1983) with a very shallow slope (8 18 0 SW = 0.11 S
- 3.15) which fits between the regression lines established for the eastern and western equatorial
Atlantic (Fairbanks et al. 1992). The following