High-Resolution geochemical data in coral skeleton and proxy for paleoclimate reconstruction in the Mediterranean Sea

One of the major issues in climate studies, being of tremendous interest in the recent years, is the understanding of the anthropogenic contribution to climate changes. In particular, it is fundamental to understand how anthropogenic forcing and natural variability have interacted in the recent past and quantify the real anthropogenic contribution on natural climate. This implies a better understanding of natural climate variability from multiple locations and back in time. The development of numerical models has significantly contributed to the progress in the understanding of climate processes in the past and in the present, and has allowed previsions for the future. The General Circulation Models (GCMs) simulate climate variability and the interaction between atmosphere, oceans, lithosphere and biosphere and providean important tool for understanding climate system dynamics. Although GCMs can simulate current climate fairly faithfully, they contain many approximations and uncertainties that are difficult to quantify and remove. Thus, in order to make GCMs more reliable and accurate in the simulation of climate variability and in the prediction of future changes, they need to be tested and evaluated experimentally against records of high-resolution climate changes in the past. The only possibility to extend our climate database far beyond the instrumental record is studying well-dated natural archives of climate variability, available from a variety of sources, e.g. tree rings, ice cores, fossil pollens, corals and speleothems. Corals and speleothems are one such natural recorder that can provide the opportunity to evaluate model performance and can quantify precise correlation of the climatic conditions between oceanic and continental regions. One of the great advantages of corals and speleothems as paleoclimatic archives stems from the fact that they offer absolute dating possibilities with the 14C method, the U/Th decay series and the band counting, thereby allowing the records of climate variability to be extended into the late Quaternary. In addition, they systematically incorporate stable isotopes and trace elements into the carbonate material and this incorporation is directly driven by climate processes that can be quantified. Another important criterion for choosing corals and speleothems is the possibility to calibrate their information to instrumental climate data, which allow to obtain accurate calibration equations. Considerable effort has been made during the last two decades to reconstruct important climatic parameters, such as sea surface temperature (SST), salinity, ocean circulation changes, air temperature and precipitation variability etc. at high temporal resolution using both shallow-water corals (Druffel and Griffin, 1993; Linsley et al., 1994; Quinn et al., 1998; McCulloch et al., 1999; Gagan et al., 2000; Hendy et al., 2002; Felis et al, 2004; Zinke et al., 2004) and speleothems (Bar-Matthews et al., 1997; McDermott et al., 1999, 2001; Desmarchelier et al., 2000; Neff et al., 2001; Baldini et al., 2002; Bard et al., 2002; Treble et al., 2003; Fleitmann et al., 2003, 2004). The great potential of shallow water corals have been investigated mainly in the tropical regions where corals are abundant and are characterized by a fast linear extension rate (10 mm/year). In the Mediterranean Sea coral-based paleoclimatic studies are extremely rare and they mostly employ sclerochronological data to infer information regarding SST (Peirano et al., 2004). In order to understand variations of past SST in the Mediterranean Sea three samples of shallow-water corals C. caespitosa (Linneo,1758) from three different locations along the Italian coast have been analyzed for a suite of minor and trace elements by laser ablation ICP-MS. No meaningful signals were obtained for Mg, U, and Ba due to their low concentrations and suspected persistent contamination in the coral. Cyclic behavior was observed for B and Sr which was inferred to reflect the seasonal variation in SST, good correlation was found between these two elements. For the other trace elements the variations within the profiles did not perfectly correlate with each other, however, indicating that a number of environmental or physiological factors may influence the seasonal signals. For two of these trace elements B and Sr we used a time base distortion (TBD) method to reconstruct changes in the growth-accretion rate of proxy environmental records, based on the periodic content of the proxy series. In our case we reconstruct growth accretion for non-tropical coral which stop to growth below the temperature of 14 Celsius degrees. This approach is useful for a better reconstruction of the time series. The two weak points of this method are the constraints: that the record has to have a periodic component and the bandwidth of the TBD is limited. The main benefit is that it can reconstruct variation in growth or accretion rate, including short growth or accretion stops. This method is important for transformation of proxy data in temperature past data and it is the starting point to fix the boundary condition in the model simulation. Six calibration equations were calculated, three for B/Ca and three for Sr/Ca vs SST, for the sites of Miramare, Portofino and Taranto. In the case of Sr/Ca trace element the comparisons of calibration equation for Cladocora caespitosa (Linneo,1758) samples with the experimental parabolic curve by Reynauld et al. (2006), obtained by studying the branching tropical coral Acropora sp. under controlled condition in aquaria, were showed. We concluded that the differences in offset (i.e. intercepts) among the three regression lines and the parabola are probably related to the different chemical composition of the water masses in the sampling sites and that the three different slopes of the three regression lines likely depend on the different range of temperature in which the Cladocora caespitosa (Linneo,1758) samples were living. Based on these findings, we suggest that the Sr/Ca composition for this coral species, analyzed under controlled conditions, might be have similarly to the tropical coral Acropora sp, following a parabolic behavior. Finally in the last section of this study we discussed detailed laser ablation analyzes on specific microstructures of two coral species Cladocora caespitosa (Linneo,1758): zooxanthellate coral and Lophelia pertusa (Linneo,1758): azooxanthellate coral). Very important results showed that the Li/Mg ratios is a very promising temperature tracer, being independent to the coral physiology. Sr/Ca and B/Ca ratios, which have been extensively used to reconstruct the SST variations in tropical seas, seem to be affected by the thermal stress, as observed from the specimens experimentally cultured in the four tanks. Both elements show a clear negative linear regression with temperature between 15 and 21 Celsius degrees and a discrepancy from linearity at 23 Celsius degrees. Conversely, the Li/Mg ratios are negatively correlated to the temperature along the whole temperature range (i.e. between 15 and 23 Celsius degrees) and they seem not influenced by the biological factors.