Contributions to the carbon isotopic composition of pedogenic carbonate in Aridisols

Pedogenic carbonate is a common feature of arid and semi-arid soils. The stable carbon isotopic composition of soil carbonate is widely used for paleoenvironment reconstruction. The use of soil carbonate as a paleo-proxy depends on the applicability of the diffusion-production model of the isotopic composition of soil CO₂ under conditions that may be highly variable, and unconstrained in the fossil record. In order to assess the applicability and appropriate model parameters of the diffusion-production model, we studied pedogenic clast coatings from Aridisols on alluvial fans across west-central Utah. To maximize potential variation in lithogenic vs. biogenic carbon, we selected sites from eight alluvial fans with different drainage basin lithologies, which included quartz sandstone, rhyolite, marine carbonate, and sulfide-rich black shale and marine carbonate. We targeted sites with abundant juniper vegetation to minimize variation in the isotopic composition of biogenic carbon and obtain δ¹³C values of C₃-dominated vegetation. To constrain variation in the δ¹³C values of soil organic carbon (SOC) with elevation, we collected soil organic matter from vegetation zones along an elevation transect along the Stansbury Range. We observed decreasing δ¹³C values with increasing elevation with each vegetation zone having a characteristic δ¹³C value. In order to minimize variation in the atmospheric carbon contribution to soil carbonate, all clast coatings were collected from ~60cm depth. The δ¹³C values of SOC from the 8 sites had a narrow range from -25.4 to -23.8‰, suggesting we successfully minimized variation in the isotopic composition of biogenic carbon by targeting the juniper vegetation zone. The δ¹³C values of soil carbonate from all sites ranged from -6.6 to -3.3‰. We expected to see lithogenic carbon contribution in soils with sulfide-rich black shale, however the limited variation in δ¹³C_PC suggests that biogenic carbon is the dominant contribution to the carbon isotopic composition of soil carbonate. However, the observed ~20‰ difference between δ¹³C_SOC and δ¹³C_PC indicates that assuming a cumulative ~15‰ fractionation to estimate δ¹³C of paleo-biomass from δ¹³C_PC may be problematic. Soil respiration rates must be lower than is generally assumed to account for this difference under the assumptions of the diffusion-production model of soil CO₂-carbonate. Respiration rates between 1x10^-11 to 2x10^-11 mol cm^-2 s^-1 would account for the observed difference in the carbon isotopic composition of soil carbonate and SOC. Measured respiration rates peak (~1.4x10^-10 mol cm^-2 s^-1) early in the morning and decrease during the day to ~2x10^-11 mol cm^-2 s^-1. Our results suggest caution if using pedogenic clast coatings from Aridisols to reconstruct paleo-vegetation. Lithogenic carbon does not contribute to the carbon isotopic composition of soil carbonate but without accounting for low respiration rates the δ¹³C values of soil carbonate would suggest a greater C₄ contribution when using assumptions of respiration rates on the order of 1.5x10^-10 mol cm^-2 s^-1 or greater.