Reconstruction of historical suspended particulate matter contributions of Rhône River tributaries to the Mediterranean Sea

Abstract

Total metal concentrations are usually used as geochemical tracers to identify sources of suspended particulate matter (SPM) and sediment deposits, but their non-conservative behaviour can lead to large uncertainties. This non-conservatism may be a significant issue when total metal concentrations of contemporary SPM are used to determine historical sediment sources, as diagenesis processes and/or past anthropogenic inputs modify metal concentrations in sediment core profiles. For the Rhône River, which is the main supplier of SPM to the Mediterranean Sea, there is no information on historical SPM source contributions from its main tributaries during major flooding events. The objective of this study was to highlight the relevance of using tracers in the non-reactive fraction to estimate the historical relative SPM contributions of the main tributaries of the Rhône River. These historical SPM contributions (1981–2017) were reconstructed based on a sediment core sampled at the outlet of the river basin and on sampling of contemporary SPM (2011–2019, Rhône Sediment Observatory monitoring network) from each main tributary. Fine stratigraphic resolution enabled us, using this original fingerprinting method, to reconstruct historical tributary contributions to sediment deposits at the scale of past flooding events over the last 37 years. When using tracers in the total fraction of SPM/sediments, the Isère and Ardèche rivers were the main contributors of SPM (32 % and 22 % on average, respectively) during the 1981–2017 period, whereas when using tracers in the non-reactive fraction, the main SPM source was the Durance River (52 %). The results obtained by this fingerprinting approach using the non-reactive fraction are fully consistent with literature data estimated using other methods (e.g. mean interannual SPM contribution of 53 % for the Durance River), and with documented historical flood events. This study showed that using tracers in the non-reactive fraction is more reliable than using tracers in the total fraction, especially for the deepest layers of the sediment core that are influenced by historical anthropogenic inputs.