A proxy for all seasons? A synthesis of clumped isotope data from Holocene soil carbonates

Abstract

Soil carbonates are important paleoclimate archives, but interpretations of their isotopic compositions (ẟ18O, ẟ13C, and ∆47) are hampered by uncertainty in the annual timing of their accumulation. Several previous studies have inferred the annual timing of soil carbonate formation by comparing modern air/soil temperatures with temperatures estimated from the clumped isotopic composition (T∆47) of Holocene soil carbonates. Here, we compile the existing ∆47 data to first consider if recent changes in ∆47 standardization methods alters the apparent seasonal biases. Then we explore the importance of various environmental parameters on soil carbonate T∆47 and discuss implications for associated paleoclimate reconstructions. Most soil carbonates record T∆47 values higher than mean annual air temperature, but the residual varies from -4 to +24 °C. ∆47 values for most observed soil profiles do not vary with depth. We find that the grain size of the soil matrix, timing of precipitation, and presence of vegetation each explain a portion of the observed variance in seasonal bias and can be used to aid in interpretations of T∆47 values from paleosols. In some soil carbonates, a warm-season bias in T∆47 is accompanied with calculated ẟ18O values of soil water that appear to be biased toward the ẟ18O of summer rainfall. Most estimated values of ẟ18O of soil water are within 2 ‰ of the ẟ18O values of mean annual rainfall. Where possible, paleoclimate reconstructions should consider precipitation timing, soil texture, and vegetation cover to aid in identifying the seasonal bias of soil carbonate stable isotopic compositions.