Bivalve shells offer a great potential as environmental proxies, since they have a wide geographical range and are well represented in the fossil record since the Cretaceous. Nevertheless, they are much less studied than corals and foraminifera and are largely limited to isotopic studies. This is probably due to the fact that the literature has been contradictory regarding the faithfulness of elemental proxies in bivalves. The general aim of this dissertation is to increase our knowledge of proxies in bivalve carbonate. More specifically, d18O, d13C, Sr/Ca, Mg/Ca, U/Ca, Ba/Ca, and Pb/Ca were investigated in both aragonite and calcite bivalve shells and their potential as environmental proxies were evaluated.The most well studied proxy of sea surface temperature (SST) in bivalve carbonate is d18O, and it is well known that in addition to SST, the d18O of the water dictates the d18O value of the shell.This study cIearly demonstrates that unknown d18O of the water can cause severe errors when calculating SST from estuarine bivalve shells; with the example presented here providing calculated SSTs 1.7 to 6.4 °C warmer than measured. Therefore, a salinity independent or salinity proxy would greatly benefit SST reconstructions. In estuaries, shell d13C has long been regarded as a potential salinity indicator. However, more recent works have demonstrated that the incorporation of light carbon from metabolic CO2 interferes with the environmental signal. This study confirms that the amount of metabolic CO2 increases in internal fluids with age, resulting in the strong ontogenic decrease in d13C values of bivalve shells. However, this is not always the case, with Saxidomus giganteus shells showing no discernable decrease over ~10 years growth. On the other hand, this study also demonstrates that the percent metabolic CO2 (%M) incorporated into bivalve shells can be large -up to 35 % in some individuals of Mercenaria mercenaria. An attempt was made to remove this metabolic influence using the relationsbip between %M and shell biometrics; however the inter- and intra-site variability was too large. This was also the case for the relatively short-lived bivalve Mytilus edulis, where the %M varied between 0 and 10%. Within the studied estuary (Schelde) the shells were close to equilibrium, but at the seaward site, where wave action is stronger, the shells contained -10 %M and the absolute d13C values were indistinguishable from specimens within the estuary, despite a salinity difference of 4. Therefore, interpreting d13C values in bivalve carbonate should be done with caution. In addition to d13C, Ba/Ca ratios were investigated as a salinity proxy as well. In the calcite shells of M. edulis a strong linear relationship between shell 'background' Ba/Ca and water Ba/Ca was found in both the laboratory and field. Although each estuary wiII have different relationships between salinity and water Ba/Ca, shell Ba/Ca can be used as an indicator ofsalinity within one estuary. Similar patterns of relatively stable background levels interrupted with sharp episodic peaks were also found in the aragonite shells of S. giganteus, and appear nearly ubiquitous to all bivalves. However, there was an ontogenic decrease in S. giganteus background Ba/Ca ratios, ilIustrating that these proxies can be species specific. Previous hypotheses regarding the cause of the peaks include ingestion of Ba rich phytoplankton or barite. This study ilIustrates that there is no direct relationship between Chl a and Ba/Ca peaks in S. giganteus shells, but they still may be related to blooms of specific species of phytoplankton.The ratios of Sr/Ca and Mg/Ca were investigated as salinity independent SST proxies. Ratios of Sr/Ca were found to be highly correlated to growth rate in S. giganteus, but not in M. mercenaria, contradictory to an earlier study on M. mercenaria. Although growth rates and temperature are often correlated, there was only a poor correlation between Sr/Ca and SST in S. giganteus (maximum R² = 0.27). Similarly, Mg/Ca and U/Ca ratios in S. giganteus were not correlated to SST, with U/Ca exhibiting a strong ontogenic trend.Finally, the use of bivalve shells as recorders of pollution was also assessed. There was both large inter- and intra-specimen variability in Pb/Ca ratios of M. mercenaria shells, but when enough shells were averaged, the typical anthropogenic Pb profile from 1949 to 2003 was evident.Overall, this study demonstrates the difficulties inherent to utilizing bivalve shells as recorders of their environment. It is clear that factors determining proxy incorporation are strongly species specific and that a mechanistic understanding is needed before we can progress further in this line of research. However, this study also ilIustrates that there is indeed environmental information that can be extracted from bivalve shells. Furthermore, the physiological influence on many of the studied proxies may prove to be useful as proxies of bivalve physiology, which in turn could provide information about bivalve paleo-ecology. |
