IMPLICATIONS OF VARIABILITY IN FLUVIAL δ13CDIC FOR ANCIENT CARBON CYCLE RECONSTRUCTIONS

The Neoproterozoic carbon isotope record is characterized by large variations in δ¹³C values measured from carbonate rocks (δ¹³C_carb), ranging from +10‰ to -15‰. The origin of these excursions cannot be explained by the traditional steady state carbon cycle model, which is often used to interpret such records. Foundational assumptions in this carbon cycle model include (1) δ¹³C_carb values accurately reflect marine dissolved inorganic carbon δ¹³C (δ¹³C_DIC), (2) the major carbon input to the ocean is delivered by rivers, and (3) the δ¹³C value from riverine input (δ¹³Criv) is equal to the average mantle δ¹³C value (δ¹³Cmantle = -5‰). The validity of ancient carbon cycling and paleoclimate interpretations depend on the accuracy of these assumptions. The third assumption - that δ¹³Criv is equal to δ¹³Cmantle - has not been tested in the modern. This study aims to quantify modern δ¹³C_DIC variability in rivers and describe the dominant controls on the observed variability. Data compilation from 140 papers yielded 3,990 δ¹³C_DIC values from 445 rivers around the globe. These values of δ¹³C_DIC have a flux-weighted mean of -9.3‰ and an enormous range from -27.5‰ to +7.0‰. We consider what controls this range, including river catchment bedrock lithology (e.g., the relative abundances of carbonate, siliciclastic and igneous lithologies) inferred from geological maps compiled in the Macrostrat database. Better understanding the large-scale controls on fluvial δ¹³C_DIC variability may be especially important for interpreting δ¹³C_carb values from Neoproterozoic shallow water carbonates, as these often form in marginal marine and mixed carbonate-siliciclastic systems, where water column DIC values may be more sensitive to local river input.