Modeling the decomposition signal and correcting bulk organic data from a peat deposit, a case study at low latitudes (Cameroon)

Organic compounds are widely used for paleoclimatic and paleoenvironmental reconstructions. Bulk organic proxies, however, are more complicated to interpret due to the multiple causes of variation in climatic and environmental conditions and the degree of diagenetic alteration. As labile compounds, rich in easily degradable function and generally richer in heteroatoms such as oxygen and nitrogen, decompose, the remaining organic matter becomes progressively richer in refractory carbon and its carbon content increases. Thus, in a peat deposit composed almost entirely of organic matter, total organic carbon (TOC) is expected to increase with time and depth, which could mask the paleoclimatic signal. We propose a simple model for peat sediments to remove the decomposition signal based on a logarithmic function fitted with a partial dataset where decomposition is the main parameter. The subtraction of the obtained logarithmic function to the raw data (i.e., measured data) leads to "residual" data. We discuss the influence of different parameters (water table depth, vegetation, microbial community) on the "residual" data and their possible link to paleoclimatic and paleoenvironmental variations. This method is tested on bulk elemental and isotopic data obtained from a new peat core from the Adamawa Plateau (North-East Cameroon) covering nearly 10 ka cal BP. Comparison with Rock-Eval (R) parameters highlights similar variations between the Hydrogen Index and residual TOC variations, supporting the interpretation based on residual TOC. Our approach allows to extract paleoenvironmental information from decompositionprone bulk organic proxies and can be generalized to peat deposits where decomposition plays a major role in controlling bulk data.