Geological Survey of Denmark and Greenland Bulletin 15, 2008, 89-92 Reconstructing past secular environmental variations is an important issue in palaeoclimate research. However, most key variables for palaeoclimate reconstructions cannot be measured directly, and reconstructions are therefore based on proxy data. Here, we demonstrate the potential of bivalve shells as an archive of environmental parameters. The Geo lo - gical Survey of Denmark and Greenland (GEUS) has devel- oped a fast and reliable method for chemical analyses of shell material by laser ablation - inductively coupled plasma - mass spectrometry (LA-ICP-MS), and here we present some ex - amples of the use of this method. In tropical and subtropical waters, corals can provide cen- tury-long archives of past water chemistry with annual reso- lution. A comparable archive for temperate and Arctic waters would be highly useful in climate research, and therefore it has been examined whether this can be provided by bivalve shells (e.g. Schoene et al. 2005). Long-lived species may pro- vide archives with annual resolution extending over several hundred years, whereas short-lived, fast-growing species can provide archives with a seasonal or in some cases daily reso- lution over a period of a few years. Most bivalves are sessile, and shells are commonly preserved as fossils. There are, how- ever, a number of challenges related to the use of bivalves as proxy archives: (1) many proxies show species specific behav- iour (Seed 1980); (2) only very few proxies are dependent on a single variable (Wefer et al. 1999); and (3) the effects of biology and ontogeny on the uptake of trace elements and stable isotope fractionation in shell carbonate are largely unknown and have to be evaluated empirically. Therefore, any potential proxy must be calibrated individually for each species of interest before it can be used. A large number of chemical analyses are needed to calibrate a proxy. These are commonly obtained by solution ICP-MS, in which sample preparation is time-consuming and labour-intensive. The use of LA-ICP-MS is therefore a considerable advance in bivalve shell proxy research, as it greatly reduces the effort needed for sample preparation. At the same time, the method requires less material for analysis, thus providing better spatial and hence temporal resolution. Proxies based on bivalve shell carbonate can be used in present-day environmental monitoring, and for environmen- tal reconstructions from shells found as fossils. Shells from museum collections and shells found in archaeological mid- dens can give information on historic and prehistoric envi- ronmental conditions (e.g. Carrell et al. 1987), and fossil shells can be used as archives of environmental parameters on geological timescales (e.g. Hendry et al. 2001). Shell mineralisation Bivalve shells consist mainly of calcium carbonate with im - purities in the form of various elements substituting for cal- cium in the crystal structure. Calcium carbonate represents 95–99% by weight of the shell, the remaining 1–5% being organic matrix, which is dominated by proteins (Marin & Luquet 2004). The shell material is deposited sequentially in growth increments that are often visible in polished sections studied by computer-controlled scanning electron micro - scopy (CCSEM). As a consequence of the growth pattern the increments occur in chronological order, and a relative time line for chemical analyses can be established. A section with the different layers of the shell of the common blue mussel (Mytilus edulis) is shown in Fig. 1, where internal growth increments are also illustrated. Figure 2 shows an image of the actual shell structure. The elements needed for shell mineralisation come from the water or from particles that the bivalve ingests. In order to be included in the shell, the elements have to cross two biological membranes, the outer and inner mantle epithe- lium. These membranes actively discriminate against certain elements, but for some elements this discrimination is influ- enced by external stimuli (e.g. Klein et al. 1996). 89 Laser ablation analysis of bivalve shells – archives of environmental information Maiken Hansen Klünder, Dorothee Hippler, Rob Witbaard and Dirk Frei © GEUS, 2008. Geological Survey of Denmark and Greenland Bulletin 15, 89–92. Available at: www.geus.dk/publications/bull Aragonite Calcite Periostracum Hemolymph M an tl e Sh el l Inner extrapallial fluid (EPF) Outer extrapallial fluid Epithelium (inner/outer) Fig. 1. Section through the margin of shell and mantle of a Mytilus edulis. The crystalline shell consists of two separate layers: a prismatic layer of calcite and an aragonitic layer of nacre. The outermost layer is a protective organic layer (periostracum). The shell is secreted in growth increments in the area be - tween the shell and the mantle. The LA-ICP-MS method Many previous studies of bivalve shells have utilised wet chemical analysis. Samples often consist of powder drilled from the shell with dentist drills and other microdrilling tools. Powder samples are routinely analysed by solution ICP-MS. LA-ICP-MS combines an analytical precision com- parable to that of solution ICP-MS with a significantly shorter and easier sample preparation process. The laser tech- nique is not only time-saving – the fewer steps needed in sample preparation also reduce the risk of contamination. Furthermore, the spatial resolution is much higher, as laser ablation in shell samples can be undertaken with a beam diameter of 30–65 µm, as opposed to the 200–300 µm diam- eter of a microdrill. Sample preparation Any sample of bivalve shell material can be analysed by LA- ICP-MS, but cross-sections through entire valves are pre- ferred in order to constrain the growth history. The shell must be cleaned of soft tissues, epibionts or adhering sedi- ment. The shell material is embedded in epoxy resin to pre- vent it from fracturing during handling. The shell is then cut with a diamond-tipped rock saw to produce a cross-section, and polished to show the shell structure (Fig. 3). Shells longer than 5 cm may have to be divided into two or more sections to fit into the sample chamber of commercially available laser ablation systems. After polishing, the section is cleaned with alcohol and treated ultrasonically to remove possible surface contamination. Analytical techniques The LA-ICP-MS equipment at GEUS is a Finnigan El - ement2 high resolution ICP-MS connected to a new wave research UP213 laser ablation system. For shell analyses, the NIST 612 and NIST 614 glasses are used as standard mate- rials. The elemental concentrations for the standard glasses published by Pearce et al. (1997) are used for concentration calculations. There are potential problems in using non- matrix-matched standards, but at ablation times of less than 80 seconds, these problems are not significant in analyses car- ried out on calcium carbonate (Vander Putten et al. 1999). As an internal standard in the samples, calcium (43Ca) is suitable for the measurement of several trace and minor elements in calcite (Longerich et al. 1996), and SEM analyses of M. edulis have shown that the calcium content in the calcite layer is uniform. We use the Glitter software package for final con- centration calculations from the time-resolved raw data. Relative age and growth rate An advantage of calibrating a proxy on bivalve shells taken from laboratory or field culturing experiments is that mea- surements of the shell length can be made during the experi- ment, so that the chemical analyses can be time constrained. When applying the proxy to fossil shells, it is of course im - possible to carry out multiple shell length measurements on the live individual, so other methods must be used. Many species form annual growth increments that can be used to set the relative age of a specimen. Furthermore, all species show micro-increments that are visible in a microscope. These nar- row growth increments are not always regular, but in a number of species the increments show a periodicity related to moon phases or diurnal or tidal shifts. Utilising what is known about the periodicity of increment formation in the species analysed, one can assign relative ages to chemical analyses and calculate approximate growth rates for the analysed shell. Examples Mg/Ca thermometry The use of the ratio between Mg and Ca in shells as a tem- perature proxy was first suggested because it was found that the Mg/Ca ratio in marine carbonates varies according to lat- 90 5 µm 1 cm Fig. 2. SEM image of a Mytilus edulis shell in cross-section, showing arago- nitic nacre (left side of image) and prismatic calcite (right side of image). The image illustrates the differences in structure between these two shell layers, and the direction of the growth increments. Fig. 3. Computer scan of shell sample Mytilus edulis B218 prepared for LA- ICP-MS analysis. itude (see Henderson 2002). Using calcite from M. edulis taken from field culturing experiments in the Wadden Sea, we found shell Mg/Ca ratios to be temperature dependent. The Mg/Ca ratio of M. edulis shells from Svendborg Sund, Denmark was then used to calculate seawater temperatures. The temperature was calculated from the Mg/Ca ratio using the equation T = 2.22 + 18.2 log (Mg/Ca) (unpublished data, M.H. Klünder). The calculated temperatures have been com- pared with water temperatures measured by the National En - vironmental Research Institute (Fig. 4). It is seen that the Mg/Ca thermometer gives a fair estimate of temperature changes during a summer. Lead pollution The concentration of Pb in shell increments of the bivalve Mya arenaria is a function of the Pb concentration in the water (Pitts & Wallace 1994), and hence the former concen- tration of Pb in the water can be calculated from the lead con- centration in M. arenaria shells. Shell samples from a four- year old specimen from Limfjorden, Denmark, collected in 2005, have been analysed. The results indicate that the Pb con- centration in Limfjorden has varied from 20 to 280 pmol/kg water over the sampled time span (Fig. 5). Hence analysing a single water sample may give a misleading picture of the Pb level. The concentration of Pb in Limfjorden is comparable to that found in the relatively uncontaminated Cape Cod Bay, eastern USA; it is up to ten times higher than pre-indus- trial levels in the Boston area, as calculated from the Pb con- tent of sub-fossil shells from shell middens, and ten times lower than in Boston harbour (Pitts & Wallace 1994). The Pb proxy has also been applied to data from an Arctica islandica individual that was transferred from the Baltic Sea to a Dutch harbour (Fig. 5). The proxy has not yet been cal- ibrated for A. islandica, and the results can only be regarded as qualitative. However, it is seen that the Pb concentration in shell material secreted after transplantation to the harbour is significantly higher than in that secreted in the Baltic. These results indicate that shells of A. islandica can be used to monitor Pb contamination of seawater. Shell Mn/Ca and Ba/Ca – a link to primary production? It has been suggested that the content of Mn and/or Ba in bivalve shells can be correlated with primary production (e.g. Stecher et al. 1996). This would suggest that Mn/Ca or Ba/Ca ratios are related to phytoplankton blooms, providing a proxy 91 0 5 10 15 20 25 0 50 100 150 200 Jan/1 Mar/1 May/1 Jul/1 Sep/1 Nov/1 Jan/1 Calculated temperature Measured temperature T em pe ra tu re ( °C ) Analysis number 0 100 200 300 400 020406080100120140160 0 10 20 30 40 50 60 pm o l P b/ kg w at er Analysis no. A545 Analysis no. 025 10 0 20 30 40 50 C hl -a , n o n- ac id ifi ed μm o l B a/ m o l C a 0.5 1 1.5 2 2.5 3 3.5 4 10 0 20 30 40 50 0.05 0.10 0 0.15 Sep/1 Jan/1 May/1 Sep/1 Jan/1 C hl -a , n o n- ac id ifi ed μm o l M n/ m o l C a A B Fig. 4. Temperatures in Svendborg Sund, Denmark in the summer of 2005 calculated from Mg/Ca ratios in Mytilus edulis (red line) and compared to measured water temperatures (blue line). The calculated temperatures pro- vide a fair estimate of the actual temperature. Fig. 5. Pb content in seawater calculated from shell Pb concentration using the equation of Pitts & Wallace (1994). One specimen of Mya arenaria (blue line, Limfjorden) and one specimen of Arctica islandica (red line, Baltic Sea, transplanted to the Netherlands) were analysed. The arrow shows the time of the transplantation. Fig. 6. Ba/Ca and Mn/Ca ratios of four Mytilus edulis specimens compared with the chlorophyll-a concentrations (red line) of the seawater. A: Ba/Ca ratios in M. edulis shells from the Dutch Wadden Sea. B: Mn/Ca measured in the same shells. for the timing and size of such events. To test this theory, M. edulis shell samples from an aquaculture field experiment site in the Wadden Sea were analysed at GEUS and compared to the chlorophyll-a concentration of the ambient water (Fig. 6). The results are not conclusive, but they suggest that the relationships between the Mn/Ca or Ba/Ca ratios and the chlorophyll-a concentration are not simple linear functions. The Ba/Ca ratio in the shells seems to increase with the chlorophyll-a concentration in the water, but continues to remain at an elevated level after the end of the bloom; how- ever, Mn/Ca seems to have a correspondence with the peaks of some less intensive algae blooms, but is quite low during the most pronounced bloom recorded in May. Clearly, fur- ther research is needed to better understand the link between Mn/Ca and Ba/Ca ratios in shell material and phytoplankton blooms. Final remarks The LA-ICP-MS method is a reliable and advantageous tech- nique for the analysis of a wide range of trace elements in car- bonates. The combination of relatively high precision, low detection limits, high spatial resolution, straightforward sam- ple preparation and fast analysis makes the method especially suited for research and application of calcium carbonate based proxies. Further development of biogenic carbonate proxies will have benefits for both palaeo-climate research and investigation into the processes of biomineralisation, as well as for environmental studies. Acknowledgements We would like to thank M. Sejr for assistance and J. v. Iperen for the chlorophyll-a data. This paper is a contribution to the EuroCLIMATE pro- ject 04 ECLIM FP08 CASIOPEIA. The work has been supported finan- cially by the European Science Foundation under the EUROCORES Programme EuroCLIMATE and the Danish Research Agency for Science, Technology and Innovation (FNU). References Carell, B., Forberg, S., Grundelius, E., Henrikson, L., Johnels, A., Lindh, U., Mutvei, H., Olsson, M., Svärtström, K. & Westermark, T. 1987: Can mussel shells reveal environmental history? Ambio 16, 2–10. Henderson, G.M. 2002: New oceanic proxies for paleoclimate. Earth and Planetary Science Letters 203, 1–13. Hendry, J.P., Perkins, W.T. & Bane, T. 2001: Short-term environmental change in a Jurassic lagoon deduced from geochemical trends in arag- onite bivalve shells. Geological Society of America Bulletin 113, 790–798. Klein, R.T., Lohmann, K.C. & Thayer, C.W. 1996: Bivalve skeletons record sea-surface temperature and δ18O via Mg/Ca and 18O/16O ratios. Geology 24, 415–418. Longerich, H.P., Günther, D. & Jackson, S.E. 1996: Elemental fractiona- tion in laser ablation inductively coupled plasma mass spectrometry. Fresenius’ Journal of Analytical Chemistry 355, 538–542. Marin, F. & Luquet, G. 2004: Molluscan shell proteins. Comptes Rendus Palevol 3, 469–492. Pearce, N.J.G., Perkins, W.T., Westgate, J.A., Gorton, M.P., Jackson, S.E., Neal, C.R. & Chenery, S.P. 1997: A compilation of new and published major and trace element data for NIST SRM 610 and NIST SRM 612 glass reference materials. Geostandards Newsletter 21, 115–144. Pitts, L.C. & Wallace, G.T. 1994: Lead deposition in the shell of the bivalve Mya arenaria: an indicator of dissolved lead in seawater. Estuarine, Coastal and Shelf Science 39, 93–104. Schoene, B.R., Fiebig, J., Pfeifer, M., Gless, R., Hickson, J., Johnson, A.L.A., Dreyer, W. & Oschmann, W. 2005: Climate record from a bivalved methuselah (Arctica islandica, mollusca, Iceland). Palaeo - geography, Palaeoecology, Palaeoclimatology 228, 130–148. Seed, R. 1980: Shell growth and form in the bivalvia. In: Rhoads, D.C. & Lutz, R.A. (eds): Skeletal growth of aquatic organisms, biological records of environmental change, 23–67. New York: Plenum Press. Stecher, H.A., Krantz, D.E., Lord, C.J., Luther, G.W. & Bock, K.W. 1996: Profiles of strontium and barium in Mercenaria mercenaria and Spisula solidissima shells. Geochimica et Cosmochimica Acta 60, 3445–3456. Vander Putten, E., Dehairs, F., André, L. & Bayens, W. 1999: Quantitative in situ microanalysis of minor and trace elements in biogenic calcite using infrared laser ablation – inductively coupled plasma mass spec- trometry: a critical evaluation. Analytica Chimica Acta 378, 261–272. Wefer, G., Berger, W.H., Bijma, J. & Fischer, G. 1999: Clues to ocean his- tory: a brief overview of proxies. In: Fischer, G. & Wefer, G. (eds): Use of proxies in paleoceanography: examples from the South Atlantic, 1–68. Berlin: Springer. 92 Authors’ addresses M.H.K. & D.F., Geological Survey of Denmark and Greenland, Øster Voldgade 10, DK-1350 Copenhagen K, Denmark. E-mail: mhk@geus.dk D.H., Faculty of Earth and Life Sciences, Vrije Universiteit Amsterdam, De Boelelaan 1085, 1081 HV Amsterdam, the Netherlands. R.W., Royal Netherlands Institute for Sea Research (NIOZ), P.O. Box 59, 1790 AB Den Burg (Texel), the Netherlands.