The influence of seawater carbonate ion concentration [CO2-3] on the stable carbon isotope composition of the planktic foraminifera species Globorotalia inflata

We sampled the upper water column for living planktic foraminifera along the SW-African continental margin. The species Globorotalia inflata strongly dominates the foraminiferal assemblages with an overall relative abundance of 70-90%. The shell delta(18)O and delta(13)C values of G. inflata were measured and compared to the predicted oxygen isotope equilibrium values (delta(18)O(eq)) and to the carbon isotope composition of the total dissolved inorganic carbon (delta(13)C(DIC)) of seawater. The delta(18)O of G. inflata reflects the general gradient observed in the predicted delta(18)O(eq) profile, while the delta(13)C of G. inflata shows almost no variation with depth and the reflection of the delta(13)C(DIC) in the foraminiferal shell seems to be covered by other effects. We found that offsets between delta(18)O(shell) and predicted delta(18)O(eq) in the surface mixed layer do not correlate to changes in seawater [CO(3)(2-)]. To calculate an isotopic mass balance of depth integrated growth, we used the oxygen isotope composition of G. inflata to estimate the fraction of the total shell mass that is grown within each plankton tow depth interval of the upper 500 m of the water column. This approach allows us to calculate the Delta delta(13)C C(interval) (added-DIC); i.e. the isotopic composition of calcite that was grown within a given depth interval. Our results consistently show that the Delta delta(13)C(IA-DIC) correlates negatively with in situ measured [CO(3)(2-)] of the ambient water. Using this approach, we found Delta delta(13)C(IA-DIC)/[CO(3)(2-)] slopes for G. inflata in the large size fraction (250-355 mu m) of - 0.013%o to 0.015%o (mu mol kg(-1))(-1) and of -0.013%o to 0.017%o (mu mol kg(-1))(-1) for the smaller specimens (150-250 mu m). These slopes are in the range of those found for other non-symbiotic species, such as Globigerina bulloides, from laboratory culture experiments. Since the Delta delta(13)C(IA-DIC)/[CO(3)(2-)] slopes from our field data are nearly identical to the slopes established from laboratory culture experiments we assume that the influence of other effects, such as temperature, are negligibly small. If we correct the delta(13)C values of G. inflata for a carbonate ion effect, the delta(13)C(shell) and delta(13)C(DIC) are correlated with an average offset of 2.11. (c) 2005 Elsevier B.V All rights reserved.