Persistence of old soil carbon under changing climate: The role of mineral-organic matter interactions

Globally, soils store between 1500 and 2800 Pg of organic carbon (OC). The physical and chemical stability of these terrestrial soil carbon stores under plausible climate change scenarios is unclear. Soil organic carbon (SOC), especially in volcanic soils, is stabilized through mineral matrix interactions. How susceptible are these mineral-organic matter interactions to environmental change? Here we present a study of SOC age along a climate gradient of andisols from Kohala volcano on the Island of Hawai'i. We measure carbon isotope composition (C-14/C-12, C-13/C-12) in bulk samples and extracted biomarkers for 4-8 horizons of 15 soil profiles to understand variability in SOC age and persistence across incremental differences in mean annual precipitation. Bulk OC in the subsoil has radiocarbon fraction modern (Fm) values as low as 0.28 to 0.16 (similar to 10,160 to similar to 14,630 conventional radiocarbon years). Coexisting plant-derived long chain fatty acids (LCFAs) are older, over 22,500 yrs. (Fm = 0.060), implying that these are among the most stable compounds in the soil, while corresponding shorter-chain (C16) fatty acids are much younger, consistent with an origin from active microbial communities assimilating young OC percolating from surface horizons. There is significant Fe loss at higher mean annual precipitation (MAP) (>2200 mm yr(-1)) sites associated with episodic soil saturation and microbial Fe reduction. %OC is higher at these sites, consistent with the expectation that saturated conditions promote SOC storage. However, in these higher MAP sites iron depletion is associated with much younger bulk SOC and LCFAs C-14 ages (similar to 2900 C-14 years) than at equivalent sample depths in sites that retain most Fe (similar to 14,200 C-14 years). The remaining mineral matrix consists primarily of Si, Al, and Ti as SRO minerals. The data imply that modest increases in precipitation resulting from environmental change at locations near a potential saturation or redox threshold could result in destabilization of Fe-SOC complexes, rendering previously stabilized carbon available for rapid degradation, potentially irreversibly decreasing the size of the old SOC reservoir. The destabilization of an old, persistent Fe-SOC reservoir can decrease SOC storage and ultimately increase the amount of CO2 released to the atmosphere.