A novel proxy for terrestrial organic matter in sediments based on branched and isoprenoid tetraether lipids

Abstract

We propose a novel tracer for terrestrial organic carbon in sediments based on the analysis of tetraether lipids using high-performance liquid chromatography/mass spectrometry (HPLC/MS). Analysis of terrestrial soil and peats shows that branched tetraether lipids are predominant in terrestrial environments in contrast to crenarchaeol, the characteristic membrane lipid of non-thermophilic crenarchaeota, which is especially abundant in the marine and lacustrine environment. Based on these findings, an index was developed, the so-called Branched and Isoprenoid Tetraether (BIT) index, based on the relative abundance of terrestrially derived tetraether lipids versus crenarchaeol. This BIT index was applied to surface sediments from the Angola Basin, where it was shown to trace the outflow of the Congo River. Furthermore, analyses of particulate organic matter from the North Sea showed relatively higher BIT indices in water column particulate organic matter near large river inputs. A survey of globally distributed marine and lacustrine surface sediments shows that the BIT index in these environments correlates with the relative fluvial input of terrestrial organic material making this index generally applicable. The new proxy allows the rapid assessment of the fluvial input of terrestrial organic material in immature sediments up to 100 Ma old.

Introduction

Large amounts of terrestrial organic carbon are annually transported from the continents to the oceans mainly by fluvial transport or, in lower amounts, by aeolian dust. They represent a large source for organic carbon in the marine environment (ca. 4×10¹⁴ g C/year; [1]) and are thus an important part of the global carbon cycle. Information on the modes and distances of transport of terrestrial carbon allows the reconstruction of, for example, the proximity to continents and wind strengths. Hence, detailed knowledge on (past) variations in transport of terrestrial carbon to marine environments is of importance for reconstructing (past) carbon cycles.

Typically, the relative amounts of terrestrial organic matter in marine sediments are determined by analyzing the ¹³C contents and C/N ratio of bulk organic matter ([2] and references cited therein). Present-day terrestrial organic matter tends to have a more ¹³C-depleted content and a higher C/N ratio than marine phytoplanktonic organic matter. However, variations in ¹³C contents of phytoplanktonic organic matter can be large (e.g., [3]) and terrestrial organic matter containing an important fraction of C₄-plants can have substantially enriched ¹³C values (e.g., [3]). Furthermore, early diagenesis can significantly alter C/N ratios by, for example, selective loss of amino acids. Hence, this approach can easily lead to erroneous interpretations of (changes in) the relative amounts of terrestrial organic matter in marine sediments.

An alternative approach is the analysis of specific molecular tracers for terrestrial organic matter and use them as a proxy for the terrestrial contribution to total sedimentary organic matter [2]. Long-chain odd-carbon-numbered n-alkanes are the most widely found terrestrial compounds in marine sediments and are derived from the surface waxes of terrestrial higher plant leaves [4]. These compounds can be washed from the leaves by rain or eroded with soils and transported by rivers to the coastal marine environment. In addition, dust can ablate the waxes from leaves, thereby transporting them through the atmosphere to locations far from the coast (e.g., [5], [6]). Long-chain n-alkanes are therefore found in both coastal and open ocean sites [2]. The stable carbon isotopic composition of these n-alkanes allows to deconvolute sources of these n-alkanes and the reconstruction of vegetation belts on continents through time (e.g., [7], [8]). Long-chain even-carbon-numbered fatty acids and n-alcohols are also derived from higher plants and are mainly fluvially transported (e.g., [9]). Specific triterpenoids such as oleanene and taraxerol allow tracing specific plant inputs from gymnosperms and mangroves, respectively (e.g., [10]). Terrestrial organic components not directly amenable to gas chromatographic analysis are lignin and cutin, biopolymers which occur abundantly in vascular plant tissues [11] and are mainly fluvially transported (e.g., [12], [13]). Through cuprous oxide degradation, reaction products can be obtained from these biopolymers, which are typical for different plants and tissues (see references in [2]). By determining their stable carbon isotopic composition, inferences can be made on the relative inputs of C₃ and C₄ plants in marine sediments (e.g., [12], [13]). Finally, reconstruction of terrestrial input is also often based on the relative abundance of C₂₉ sterols compared to C₂₇ and C₂₈ sterols [14] since C₂₉ sterols are dominant in terrestrial plants. However, C₂₉ sterols have also shown to be ubiquitously occurring in marine algae (e.g., [15]) and thus may not represent a pure terrestrial signal in marine sediments.

A large range of molecular proxies for terrestrial organic matter is thus available. However, quantification of the relative inputs of terrestrial carbon is difficult due to large variations in concentrations of compounds in the different plant materials. Also, terrestrially derived biomarkers have different degradation rates both compared to each other and to marine-derived compounds. For instance, we recently showed that preservation factors of long-chain n-alkanes, typical markers for terrestrial input, are substantially higher than those for long-chain alkenones, typical markers for prymnesiophyte algae [16], [17]. Hence, relative changes in the amounts of terrestrial n-alkanes compared to marine compounds may not only be due to changes in terrestrial contribution, but also to changing oxygen exposure times.

An alternative approach to reconstruct terrestrial input into the marine environment is not to use a tracer derived from higher plants but from organisms thriving in soils and peats. For instance, Prahl et al. [18] used diploptene in Washington coastal sediments as a tracer for soil organic carbon. Although this worked quite well in these sediments, general application of this tracer is made difficult by the fact that a number of marine bacteria also make diploptene or precursor compounds from which diploptene can be formed (e.g., [19], [20]).

The development of a high-performance liquid chromatography/mass spectrometry (HPLC/MS) technique for the analysis of glycerol dialkyl glycerol tetraethers (GDGTs) [21] enabled us to recognize a group of non-isoprenoidal GDGTs (structures I, II, and III in Fig. 1) that was recently identified with 2D NMR techniques after isolation from a Dutch peat [22]. Its biological origin is as yet unclear, but a survey of recent sediments indicates that it is derived from organisms living in the terrestrial environment [23]. In addition, we identified a structurally related isoprenoid GDGT of marine planktonic archaea, “crenarchaeol” (structure IV, Fig. 1; [24]). This compound occurs abundantly and ubiquitously in marine and lake sediments (e.g., [23], [25], [26]), the marine water column (e.g., [27], [28]) and the only available uni-archaeal “culture” of the marine pelagic crenarchaeota, Cenarchaeum symbiosum[24]. In marine sediments, this biomarker is, together with GDGT V (a less specific GDGT but also predominantly derived from planktonic archaea), probably the single most abundant component [23].

Here, we show that the amount of branched GDGTs compared to crenarchaeol in marine and lacustrine sediments, quantified in the so-called Branched and Isoprenoid Tetraether (BIT) index, is correlated with the relative amount of fluvial terrestrial input. This gives a new approach to reconstruct the fate of soil organic carbon, and thereby fluvial terrestrial input, in marine and lacustrine environments based on the analysis of GDGTs.