Multi-proxy approach involving ultrahigh resolution mass spectrometry and self-organising maps to investigate the origin and quality of sedimentary organic matter across a subtropical reservoir. (January 2021)

• Record Type: Journal Article
• Authors: Gontijo, Erik Sartori Jeunon; Herzsprung, Peter; Lechtenfeld, Oliver J.; de Castro Bueno, Carolina; A.C. Barth, Johannes; Rosa, André H.; Friese, Kurt
• Is Part Of: Organic geochemistry. Volume 151(2021)
• Publication Date: 2021-01
• Subjects: Kohonen neural network -- 13C NMR -- FT-ICR-MS -- Stable isotopes -- Humic substances -- Source discrimination; Organic geochemistry -- Periodicals; Biogeochemistry -- Periodicals; Géochimie organique -- Périodiques; 553.205
• DOI: 10.1016/j.orggeochem.2020.104165
• Languages: English
• ISSNs: 0146-6380

Abstract

Graphical abstract: Highlights: Changes in sedimentary humic substances revealed by set of complementary techniques. Differences between samples from upstream and dam areas linked to in-lake processes. Evidence of influence of land use on catchment in humic and fulvic acids quality. Fulvic acids were better than humic acids for finding inter-site differences. Self-organising maps successfully used for integrated analysis of all data. Abstract: Humic substances (HS) in sediments play an important role in carbon and nutrient biogeochemical cycles and fate of contaminants in the environment. However, information regarding HS quality and transformations that may affect their behaviour in reservoirs is still limited. The aim of this investigation was to track sources and changes in sedimentary HS across a subtropical reservoir, connecting them to in-lake processes and land-use influences. Surface sediments were collected at seven sampling sites in Itupararanga Reservoir (Brazil). Humic (HA) and fulvic (FA) acids (components of HS) were extracted from the sediment samples followed by in-depth characterisation via UV/VIS, fluorescence spectroscopy, elemental (C, N) and isotopic analysis (δ 13 C, δ 15 N), nuclear magnetic resonance ( 13 C NMR) and Fourier-transform ion cyclotron resonance mass spectrometry (FT-ICR-MS). All data were analysed by self-organising maps. The results showed that samples from the upstream part of the reservoir were older and more decomposed. They likely originated from C3 land-plants (δ 13 C varied from −26.3‰ to −22.4‰), having more aromatic, oxygen-poor (O/C < 0.5) and unsaturated compounds (H/C < 1.1). In contrast, near-dam samples were younger and had larger contributions of autochthonous material. This was confirmed by oxygen-rich (O/C > 0.5) and partly more unsaturated compounds for FA as well as oxygen-poor and saturated compounds with H/C > 1.1 for HA. Self-organising maps pointed out these differences between upstream and dam areas and indicated that agriculture lands were related to microbially-derived HS. Changes in HS composition revealed that internal reservoir processes may have influenced HS quality across the reservoir.