palaeoceanographical proxy relationships in foraminifera Vital effects and beyond: a modelling perspective on developing Geological Society, London, Special Publications

This paper mainly reviews our recent work on the biology and geochemistry of foraminifera with respect to their use as palaeoceanographic proxies. Our approach to proxy validation and development is described, primarily from a modeler’s point of view. The approach is based on complementary steps in understanding the inorganic chemistry, inorganic isotope fractionation, and biological controls that determine palaeo-tracer signals in organisms used in climate reconstructions. Integration of laboratory experiments, field and culture studies, theoretical considerations and numerical modelling holds the key to the method’s success. We describe effects of life-processes in foraminifera on stable carbon, oxygen, and boron isotopes as well as Mg incorporation into foraminiferal calcite shells. Stable boron isotopes will be used to illustrate our approach. We show that a mechanism-based understanding is often required before primary climate signals can be extracted from the geologic record because the signals can be heavily overprinted by secondary, non-climate related phenomena. Moreover, for some of the proxies, fundamental knowledge on the thermodynamic, inorganic basis is still lacking. One example is stable boron isotopes, a palaeo-pH proxy, for which the boron isotope fractionation between the dissolved boron compounds in seawater was not precisely known until recently. Attempts to overcome such hurdles are described and implications of our work for palaeoceanographic reconstructions are discussed. Development and validation of palaeoceanographic proxy relationships in foraminifera have evolved rapidly over the past few years. During the early years of palaeoceanography, offsets from isotopic and elemental geochemical equilibrium that were attributed to life processes were often referred to as biological ‘vital effects’. In the case of stable carbon and oxygen isotopes, the black box was opened, resulting in a precise characterization of biological effects on geochemical signals recorded in the calcite shells of foraminifera. Interspecific variations have long been recognized in the stable carbon and oxygen isotope system (for review, see Wefer & Berger 1991; Spero 1998). However, the breakthrough in understanding interas well as intraspecific isotope variability came with culture experiments of live foraminifera under controlled laboratory conditions (Bijma et al. 1998; BouvierSoumagnac & Duplessy 1985; Hemleben et al. 1985; Spero & DeNiro 1987; Spero & Williams 1988; Spero & Lea 1993, 1996) as pioneered by Be et al. (1977) and Hemleben et al. (1977). The profound consequences of controlled culture experiments for palaeoceanographic interpretations were widely recognized in 1997, when Spero and coworkers demonstrated that the seawater carbonate chemistry significantly affects dC and dO in planktonic foraminifera (Spero et al. 1997; Bijma et al. 1999). This phenomenon has been referred to as the ‘carbonate ion effect’. While palaeoceanographers had long been aware that temperature and seawater dO affect foraminiferal dO (Emiliani 1955; Shackleton 1967), another important player, the ocean’s CO2 chemistry, had to be added to the From: AUSTIN, W. E. N. & JAMES, R. H. (eds) Biogeochemical Controls on Palaeoceanographic Environmental Proxies. Geological Society, London, Special Publications, 303, 45–58. DOI: 10.1144/SP303.4 0305-8719/08/$15.00 # The Geological Society of London 2008. list. As a result, dO-based temperature estimates are likely too low for geologic periods in the more distant past of high atmospheric CO2 concentrations and low oceanic pH (Zeebe 2001; Royer et al. 2004; Bice et al. 2006). With respect to quantitative modelling of life processes in foraminifera, a first attempt to open the ‘vital effect’ black box by means of a mathematical approach was provided by Spero et al. (1991). Their work outlined an abstracting concept, transforming geometry and fluxes in the living organism (Fig. 1) into mathematical equations which allowed the calculation of stable carbon isotope fractionation in a model foraminifer. But it was not until after the discovery of the carbonate ion effect that more sophisticated tools such as numerical models of the chemical microenvironment (Fig. 2) were developed to understand life processes, stable isotope fractionation, the carbonate ion effect and, prospectively, trace metal incorporation into foraminiferal calcite (Wolf-Gladrow & Riebesell 1997; Wolf-Gladrow et al. 1999; Zeebe et al. 1999; Zeebe 1999). Elderfield et al. (1996) proposed a Rayleigh distillation model for trace element incorporation into foraminiferal CaCO3, which is consistent with data in benthic but not in planktonic foraminifera. As suggested by Zeebe & Sanyal (2002), the process of metal incorporation is, particularly in the case of Mg2þ, most likely intimately intertwined with the energetics of the precipitation mechanism itself (cf. also Erez 2003). Development of a comprehensive theory of element incorporation in foraminifera by means of mathematical and numerical modelling is currently an active area of palaeoceanographic research. Parallel to the refinement of well-established proxy relationships such as dC and dO in foraminifera, other important geochemical proxies have been revitalized or newly developed over the past years, including metal/Ca ratios of Mg, Sr, U, Li and stable calcium and boron isotopes. Stable boron isotope ratios in foraminifera provide a tool for reconstructing the pH of ancient seawater (e.g. Spivack et al. 1993; Hemming & Hanson 1992; Sanyal et al. 1995; Pearson & Palmer 2000; Honisch & Hemming 2005). The biogeochemical and physicochemical aspects of this ‘palaeo-acidimetry’ proxy have been intensively examined over the past years by culture studies with live planktonic species, inorganic precipitation experiments, and theoretical means including ab initio molecular orbital theory (Sanyal et al. 1996, 2000, 2001; Honisch et al. 2003; Honisch & Hemming 2004; Zeebe et al. 2001, 2003; Zeebe 2005a). In this paper, we highlight some of our recent work on the development and validation of palaeoceanographic proxy relationships in foraminifera, primarily from a modeller’s point of view. ‘Vital effects and beyond’ briefly describes the philosophy of our approach which is spelled out in terms of stable isotope fractionation in ‘Stable isotope fractionation’. In ‘Foraminifera dramatically alter their chemical and isotopic micro-environment’ we show that foraminifera strongly perturb their chemical and isotopic microenvironment, which has immediate consequences for the palaeoceanographic interpretation of stable isotopes in foraminifera from the fossil record. Application of our method to stable boron isotopes and a downcore reconstruction of Late Pleistocene glacialinterglacial cycles in surface ocean pH is presented in ‘Planktonic forminifera are reliable recorders of the ocean’s palaeo-pH’. Finally, ‘Foraminifera appear to control their shell-Mg/Ca ratio by a luxurious method’ describes our findings that foraminifera seem to control their Mg/Ca ratio by a rather expensive method in terms of energy requirements. The final section also points to several gaps in our understanding of biomineralization in foraminifera. Vital effects and beyond: the approach Our approach to proxy validation and development is based on complementary steps in exploring the inorganic chemistry, inorganic isotope fractionation and biological controls on proxy relationships in organisms relevant to climate reconstructions. In many cases, the integration of laboratory experiments, field and culture studies, theoretical considerations and numerical modelling has turned out to be a successful method for this task. The foremost goal of this research is to improve climate reconstructions. Climate signals extracted from the geological record can be heavily overprinted by secondary, non-climate related phenomena because in the case of foraminifera, climate fluctuations are recorded by living organisms rather than by chemical compounds of inorganic origin. The long-term prospect of this work is to achieve refined palaeoceanographic interpretations of proxy relationships and to apply those relationships to the actual down-core record. The practical field application of our work to deep-sea sediment records has been documented by several authors of the present paper (e.g. Sanyal & Bijma 1999; Zeebe 2001; Spero & Lea 2002; Honisch & Hemming 2005). Stable isotope fractionation A great deal of our recent efforts has been focused on developing a comprehensive theory of stable isotope fractionation in foraminifera, focusing on the elements of carbon, oxygen and boron. While the inorganic CO2 chemistry and isotope R. E. ZEEBE ET AL. 46