The stable carbon isotope composition (δ13C) of plant components such as plant wax biomarkers is an important tool for reconstructing past vegetation. Plant wax δ13C is mainly controlled by photosynthetic pathways, allowing for the differentiation of C4 tropical grasses and C3 forests. Proxy interpretations are however complicated by additional factors such as aridity, vegetation density, elevation, and the considerable δ13C variability found among C3 plant species. Moreover, studies on plant wax δ13C in tropical soils and plants have focused on Africa, while structurally different South American savannas, shrublands and forests remain understudied. Here, we analyze the δ13C composition of long-chain n-alkanes and fatty acids from tropical South American soils and leaf litter to assess the isotopic variability in each vegetation type and to investigate the influence of climatic features on δ13C. Rainforests and open vegetation types show distinct values, with rainforests having a narrow range of low δ13C values (n-C29 n-alkane: -34.5 +0.9/-0.6 ‰ ; Suess-effect corrected) allowing for the detection of even minor incursions of savanna into rainforests (13C-enriched). While Cerrado savannas and semi-arid Caatinga shrublands grow under distinctly different climates, they can yield indistinct δ13C values for most compounds. Cerrado soils and litter show pronounced isotopic spreads between the n-C33 and n-C29 alkanes, while Caatinga shrublands show consistent values across the two homologues, thereby enabling the differentiation of these vegetation types. The same multi-homologue isotope analysis can be extended to differentiate African shrublands from savannas.

We present a regional study of the Pampean loess in South America based on a detailed analysis of three sections across the core of the Pampas. High-resolution luminescence dating resulted in a new chronology that covers a period from Marine Isotope Stage 3 to the early Holocene. Reliable estimations of mass accumulation rates (MARs) for loess were used to infer the temporal dust flux variation during the last glacial/interglacial transition in southern South America (SSA). Minimum MARs in each section were identified for the Last Glacial Maximum (LGM), contrasting with high dust fluxes observed in more distal Southern Atlantic Ocean (SAO) and East Antarctica. We hypothesize that the power of the Pampean loess as a sink of dust was reduced during the LGM, allowing long-range transport of SSA dust to SAO and East Antarctica. This hypothesis is consistent with proxy data and models suggesting drier conditions in the Pampas during the LGM, which would have shut down loess accumulation. It is also consistent with isotopic evidence that points to northern Patagonia and southern central-western Argentina as main contributors of dust to East Antarctica during glacials, given that the prevailing regional wind system implies that dust emitted from these regions would have necessarily passed through the Pampas in its way to the SAO and East Antarctica. Forthcoming Nd, Sr, and Pb isotope results for the Pampean loess will allow further testing of this hypothesis.