Benthic foraminiferal stable carbon isotopes are widely used. Essentially, the d¹³C of oceanic bottom waters depends on the equilibrium between the downward flux of isotopically light phytodetrital organic matter, and the ventilation of the ocean bottom by deep water circulation (Berger and Vincent, 1986). As such, the d¹³C has primarily been used as a proxy for deep water ventilation (e.g. Duplessy et al, 1984; Michel et al, 1995; Bickert and Wefer, 1996; 1999; Vidal et al, 1999), whereas the Dd¹³C_{planktonic-benthic} (difference between planktonic and benthic d¹³C) has been tentatively applied as a marker of export production (Shackleton et al., 1983). It has further been suggested (Zahn et al., 1986; McCorkle et al., 1988; Wefer et al., 1999; Mackensen et al., 2000) that the simultaneous analyses of two or more benthic foraminiferal taxa, living at different depths in the surficial sediment, could inform about the degradation of organic matter in the sediment, and thus, about the flux of labile organic matter to the ocean floor. One of the two main objectives of our research project is to follow the latter suggestion, and to develop a reliable proxy of paleo-export production on the basis of multi-species d¹³C analyses.

Benthic foraminiferal taxa live at slightly different depths (microhabitats) in the upper sediment layers (Corliss, 1980; Jorissen et al., 1995; 1998; Jorissen, 1999, Fontanier et al., 2001). This makes that some of them should accurately record bottom water characteristics, whereas others are more typical for pore water characteristics. This difference complicates the interpretation of the stable isotope records, but, if several taxa are analyzed simultaneously, a much more complete picture of the former benthic environment may be obtained.

. Today, live benthic foraminiferal assemblages are usually divided into four microhabitat categories (Corliss and Chen, 1988; Jorissen, 1999):

1)       Epifaunal taxa, supposedly preferring a life position on top of the sediment, in our study represented by Fontbotia wuellerstorfi and the Cibicidoides group. These taxa are supposed to adequately register the ancient bottom water characteristics.

2)       Shallow infaunal taxa, living within the topmost cm of the sediment, here represented by the Uvigerina peregrina group. This group is supposed to represent the former d¹³C in the upper part of the pore waters.

3)       Intermediate infaunal taxa, represented by some representatives of the Melonis barleeanus group, live at the lower end of the oxygen-containing sediment.

4)       Deep infaunal taxa, represented by the Globobulimina affinis group, live around the zero oxygen level, and may even be found rather abundantly in the upper part of the totally anoxic sediments.

By performing d¹³C analyses for each of these four microhabitat groups, it is possible to reconstruct past pore water d¹³C gradients. Such paleo-d¹³C pore water profiles will allow us to quantify the extent of organic matter degradation in the ocean bottom waters, as well as within the sediment (Gehlen et al., 1999). The amount of organic matter remineralisation depends directly on the downward flux of labile organic matter. Such a detailed insight into former pore water d¹³C profiles will also allow us to deconvolve the sea bottom d¹³C record, and to better separate the impact from paleoproductivity changes from the impact of changes in bottom water ventilation. This deconvolution may be reinforced by independent estimates of bottom water oxygenation and/or export production. Estimates of bottom water oxygenation can be obtained from the study of the benthic foraminiferal assemblage structure (see for example Kaiho, 1994; Loubere, 1996). Recent publications showing that benthic foraminiferal faunas vary according to the downward organic (Altenbach et al., 1999; De Rijk et al., 2000; Morigi et al., in press) suggest that they could also provide independent estimates for export production.

The modern thermohaline circulation mode is driven by saline gradients leading to a major deep water formation in the northern part of the North Atlantic Ocean. Several paleoclimatic reconstructions indicate that this mode was probably not the only one active and several authors suggest that rapid climatic transitions are linked to abrupt shifts between different states of the thermohaline circulation (Vidal et al, 1999; Stocker, 2000; Keigwin and Boyle, 1999). They propose that these “ seesaw ”shifts between different circulation modes are due to freshwater input in the North Atlantic either by melting events of icebergs or by atmospheric hydrological perturbations (Ganopolski and Rahmstorf, 2001). Alternative explanations give more weight to salty changes in the southern ocean (Keeling and Stephens, 2001). It is important to discriminate between these different hypotheses to understand the real mechanisms of the climatic system and it is crucial to reconstruct independently temperature and salinity changes in the deep water masses.

Stable oxygen isotopes records based on benthic foraminifera integrate temperature, salinity and global ice volume changes (Shackleton, 1974). Nevertheless reconstructions of deep water temperature variations and then density variations using only stable isotopes have been tempted (Labeyrie et al, 1987; 1992; Lynch-Stiegliz et al, 1999; Duplessy et al, 2002).

In recent years, elemental ratios (equally measured on benthic foraminiferal shells), such as Mg/Ca and Sr/Ca have been used more or less successful to reconstruct former ocean floor temperature; a prerequisite to better interpret benthic foraminiferal d¹⁸O records. After correction, using these independent bottom water temperature estimates, the benthic foraminiferal stable oxygen isotopes may next be used to calculate former sea bottom salinities.

We propose to analyze in parallel d¹⁸O and Mg/Ca of calibrated benthic species to reach a better knowledge of the temperature and salinity characteristics.

Factors complicating the use of proxy methods based on benthic foraminiferal carbonate

All of the aforementioned proxies are based on measurements of the calcitic material composing the tests of a relatively small number of benthic foraminiferal taxa. Fontbotia wuellerstorfi is one of the favorite taxa, because of its tendency to life in epifaunal microhabitats, and it may be thought that it forms its test in equilibrium with the characteristics of the bottom water. Other taxa which are frequently used are the closely related, equally superficially living Cibicidoides group, and the Uvigerina peregrina group, which lives in the topmost sediment layers. As outlined before, the slightly deeper living Melonis barleeanus and Globobulimina affinis groups, could provide valuable information about the fate of labile organic matter within the sediment. In each of these four taxa, complex groups of morphotypes are used, often with a confuse taxonomy. Even within these four species groups, the various morphotypes may occupy different microhabitats, and in consequence, do not represent the same ocean bottom micro-environment. Furthermore, it should be realized that the choice of the species used for the geochemical analysis is in many cases restricted, because fossil deep oceanic sediments do only contain a limited number of taxa, especially in the amounts needed for some of the more sophisticated geochemical analyses. In such cases, other taxa should be selected to represent the various micro-environments.

Although a rather solid working routine has been developed for most of the aforementioned proxies, which allows a successful first order interpretation, many problems arise when a more detailed interpretation of these proxies is attempted. The most evident problem is that of the deconvolution of the various environmental parameters which each influence the same proxy record. This has been outlined for the benthic foraminiferal d¹³C, which is influenced both by the downward organic flux and by the deep water ventilation, and can also be seen in the recent trials to use the d¹⁸O to reconstruct former bottom water salinity concentrations.

The most important factors which at present hamper a more precise application of the aforementioned proxy methods is the relative lack of knowledge about a number of complicating parameters, which each may influence the isotopic and elemental composition of the foraminiferal carbonate, and which may seriously obscure the target environmental parameter which the proxy tries to reconstruct.

Some of the most evident questions common to each objective, which are addressed in this proposal, are:

1)       Do the investigated taxa of benthic foraminifera secrete their test in equilibrium with the chemical composition of the surrounding bottom and pore waters, and, if so, where exactly do the various taxa of benthic foraminifera form their test?

In order to answer this double question, a joint geochemical/ecological approach is needed. Information about the living depth of the foraminiferal taxa during the various stages of their life cycle has to be combined with measurements of foraminiferal and pore water d¹⁸O, d¹³C and elemental ratios for several depth intervals within the sediment. Furthermore, it can be imagined that some species change their microhabitat during their life history. For instance, species may reproduce at the sediment surface, whereas adult specimens may live deeper in the sediment (Fontanier et al., submitted). In such cases, the isotopic composition will be a composite of the conditions dominating the different depth levels. For this reason, most geochemical laboratories perform there analyses on foraminifera from a rather narrow size class (when present!). Modern techniques, however, allow the study of these ontogenetic changes, either by comparing measures for different size classes (if possible with single foram measurements), or by measurements on single chambers (such as is possible with UV laser ablation ICPMS).

2)       What is the impact of seasonal, interannual or more episodic variability of the organic flux, and/or of the target parameter itself, on the proxy record?

Recent evidence suggests that a major part of the fossil benthic foraminiferal faunas is formed during relatively short periods of time, when major input of fresh organic matter takes place (e.g. Gooday, 1988; 1993; Gooday et al., 1992). Such rather extraordinary conditions, following surface water bloom periods, may be seasonal, annual, or even much more episodic. The impact may be twofold:

§         In the case of the d¹³C, important phytodetritus deposits may be colonized by surface-dwelling benthic foraminifera, which consequently will isolated from their bottom water micro-environment, and will show extremely low d¹³C-values, a phenomenon which is known as the “Mackensen”-effect (Mackensen et al., 1993). The taxon Fontbotia wuellerstorfi, which shows a rather systematical d¹³C offset with respect to bottom water values, is supposed to reproduce preferentially during such periods of intensive organic matter input.

§         The physico-chemical conditions (temperature, salinity) reigning during the short time intervals, when the foraminiferal carbonate is formed, may be very different from the "average" oceanographic, long term conditions we try to reconstruct with our proxy methods.

A multi-strategy approach is necessary to better define the impact of this episodicity on the foraminiferal-based geochemical proxy records:

a)         field studies over prolonged periods of time, aiming at describing the seasonal, interannual and interdecadal variability of the faunas, and of the geochemical composition of their tests.

b)         detailed studies of recently formed fossil faunas should inform what part of the fossil faunas is formed during the high productivity episodes.

c)         geochemical analyses of individual foraminiferal tests can inform us about the variability in fossil assemblages, and the variability of the target parameter in the period of time represented by the sample.

3)       How large is the intra-and interspecific variability in the species and species-groups which are subject to the geochemical analyses?

Although the species Fontbotia wuellerstorfi is rather well defined (an exception!), in the case of Cibicidoides spp. or Uvigerina peregrina very large species complexes are concerned, with a very variable morphology, and a rather opaque taxonomy. Very probably, different morphotypes will occupy different microhabitats, and will form their tests in different equilibrium conditions. It is essential that micropaleontologists and geochemists together define a practical, taxonomically correct nomenclature, and study the ecological characteristics of the different morphotypes, as well as their isotopic composition. A comparative, ecological/geochemical approach is needed, involving morphotypes of a large number of areas.

4)       What is the impact of small scale variability (patchiness) on the ocean floor?

The combination of bottom currents and microtopography of the ocean floor causes a concentration of fresh phytodetritus in depressions; relatively elevated areas will have a much lower organic input. Such differences may largely influence the d¹³C of the organisms inhabiting these different microenvironments. A combined geochemical micropaleontological study of patchiness of benthic foraminiferal faunas in deep oceanic environments is needed to answer this question.

5)       In what way are the benthic foraminiferal faunas and their geochemical signals changed by diagenetic processes?

Diagenetic processes are responsible for important losses during the transition from living to fossil faunas. A detailed comparison of recent, living faunas with subrecent fossil faunas can inform us about the extent of this phenomenon. Furthermore, it is known that calcite dissolution taking place on the sea floor results in the preferential removal of Mg-enriched calcite of foraminifer shells, leading to a decrease of the Mg/Ca ratio in the remaining tests (Savin and Douglas, 1973; Rosenthal and Boyle, 1993; Rosenthal et al., 2000; Lea et al., 2000). This dissolution effect strongly biases the Mg/Ca paleothermometry results towards lower estimates of water mass temperatures. Among the dissolution proxies that can be used for this purpose, we recently tested the “calcite cristallinity”. From their study along a depth transect on Ontong Java Plateau, Bonneau et al. (1978) have shown that the cristallinity of planktonic foraminifer tests improves as dissolution takes place, which results from the fact that the poorly cristallized calcite is removed first during dissolution processes. We recently showed that calcite cristallinity can be used indeed to correct Mg/Ca for dissolution effects (Bassinot et al., 2001). Our next goal within this program will be to apply the “calcite cristallinity” index to correct Mg/Ca in benthic foraminifers.

This research project aims to obtain a better evaluation of the impact of each of these complicating factors on all proxies based on benthic foraminiferal carbonate. 

In order to do so, we propose a coupled ecological/geochemical study of the most commonly used groups of benthic foraminifera. We intend to combine ecological field studies with detailed geochemical measurements. Already existing, intensive collaborations with research teams in the Universities of Utrecht (Netherlands) and Tübingen (Germany), will allow to study the formation of the geochemical signals in microcosms, under controlled environmental conditions.