Constraining riverine δ13C-DIC using Late Cretaceous and Early Paleogene freshwater bivalve mollusks (Unionoidea) form Montana

Abstract:

Interpretation of carbon isotope variation in freshwater unionoid mollusk shells (δ¹³C_SHELL) is not straightforward because of the variable contributions of metabolic (i.e., food) and dissolved inorganic carbon (DIC). Bivalve shells typically contain between 0 and 50% metabolic carbon (C_M), which has a δ¹³C value close to the animal’s food source. In marine systems, the food source (usually phytoplankton) has a δ¹³C value typically around -20 ‰ and a d13CDIC value around ~0 ‰. In freshwater systems, these numbers can vary considerably, with food sources ranging from -35 to -10 ‰. Typically, δ¹³C-DIC values range between -25 to 0‰ and are dependent on numerous factors; carbonate weathering and equilibrium with the atmosphere typically leading to high values and respiration of organic matter and oxidation of methane leading to lower values. Therefore, δ¹³C-DIC values reflect numerous processes occurring in the watershed.

Nevertheless, here we suggest δ¹³C_SHELL values can constrain the lower bounds of riverine δ¹³C-DIC values, despite the influence of C_M. The metabolic end-member δ¹³C value is typically lower than the DIC end member and consequently will lead to higher calculated δ¹³C-DIC when using δ¹³C_SHELL values. Therefore, if the C_M fraction is set to 0 %, δ¹³C_SHELL values will provide the lowest possible riverine δ¹³C-DIC values (after accounting for fractionation). Applying this method to modern shells from waters with known δ¹³C-DIC values (ranging from -3.2 to -12.8 ‰) results in calculated δ¹³C-DIC values from -6.0 to -12.4 ‰, which is close to measured DIC data from the waters in which the mussel grew. This can then in turn be applied to well-preserved fossil shells.

Freshwater unionoid shells from the uppermost Cretaceous Hell Creek Formation and the lower Paleogene Fort Union Formation are exceptionally well preserved. Applying this method to these shells results in δ¹³C-DIC values ranging from -6 to -11‰, which is consistent with values from shallow, well-mixed streams. This is markedly different than a model of wide, slow-moving, meandering rivers flowing over the coastal plain to the Cretaceous Western Interior Seaway. Our results suggest bivalve δ¹³C_SHELL values may be useful indicators of ancient riverine environmental conditions.