Stable carbon isotopic compositions of archaeal lipids constrain terrestrial, planktonic, and benthic sources in marine sediments

Abstract Archaea occupy an important niche in the global carbon cycle and their lipids are widely used as indicators of environmental conditions in both paleoenvironmental and modern biogeochemical studies. The principal sources of archaeal lipids in marine sediments are benthic archaea, fossil remnants of planktonic archaea, and allochthonous sources such as soils. However, the relative contributions of these sources to the sedimentary lipid pool have not been comprehensively constrained, complicating a mechanistic understanding of archaeal lipid proxies. In order to provide insights into the relative contributions of these sources and identify signals derived from sedimentary activity, we performed a systematic survey of stable carbon isotopic compositions (δ13C) of both core and intact archaeal lipids via analyses of their phytanyl (Phy) and biphytanyl (BP) moieties in diverse marine sediments. The sample set consisted of 44 sediment horizons from the Mediterranean and adjacent basins and represented diverse sources of organic matter and depositional conditions. Complementary geochemical data enabled the comparison of lipid distributions and carbon isotopic signatures with prevailing redox conditions. The δ13C of tricyclic BP (BPcren) from the core and intact forms of crenarchaeol ranged from -19.1 to -28.6‰ and –18.1 to –27.4‰, respectively. δ13C values of core and intact BPcren did not differ, suggesting that intact crenarchaeol is either a fossil relic from planktonic archaea or a product of lipid recycling by benthic archaea, as opposed to being synthesized de novo by sedimentary archaea. δ13C values of BP0 derived from core and intact forms of glycerol and butanetriol dibiphytanyl glycerol tetraethers (GDGTs and BGDTs, respectively), but predominantly from caldarchaeol (GDGT-0), ranged from –19.4 to –32.0‰ and –20.9 to –37.0‰, respectively. In contrast to BPcren, intact-lipid derived BP0 was often 13C-depleted relative to its core counterpart, consistent with in situ production by sedimentary archaea. This relative depletion was most pronounced in sulfate reduction zones, likely due to heterotrophic activity. Core and intact archaeol exhibited the largest ranges in δ13C values, from –21.6 to –42.1‰ and –22.7 to –58.9‰, respectively. This strong 13C–depletion relative to both total organic carbon and dissolved inorganic carbon is consistent with mixtures of functional sources of sedimentary chemolithoautotrophic, methanotrophic, methanogenic and heterotrophic archaea. Based on the substantial 13C-depletion of BPcren and BP0 in samples in the vicinity of the Rhone River delta relative to a distal marine reference site, we infer that the terrestrial soil contribution of archaeal lipids to these sediments is as high as 43%. Hence, terrestrial input of archaeal lipids, including their intact forms, can be substantial and suggests caution when using existing molecular proxies aimed at constraining riverine input. In summary, our comparative isotopic analysis of sedimentary core versus intact archaeal lipids improves the apportionment of their diverse sources and confidence in distinguishing in situ lipid production by sedimentary archaea.