Shifting mineral and redox controls on carbon cycling in seasonally flooded mineral soils

Abstract

Abstract. Although wetland soils represent a relatively small portion of the terrestrial landscape, they account for an estimated 20 %–30 % of the global soil carbon (C) reservoir. C stored in wetland soils that experience seasonal flooding is likely the most vulnerable to increased severity and duration of droughts in response to climate change. Redox conditions, plant root dynamics, and the abundance of protective mineral phases are well-established controls on soil C persistence, but their relative influence in seasonally flooded mineral soils is largely unknown. To address this knowledge gap, we assessed the relative importance of environmental (temperature, soil moisture, and redox potential) and biogeochemical (mineral composition and root biomass) factors in controlling CO2 efflux, C quantity, and organic matter composition along replicated upland–lowland transitions in seasonally flooded mineral soils. Specifically, we contrasted mineral soils under temperature deciduous forests in lowland positions that undergo seasonal flooding with adjacent upland soils that do not, considering both surface (A) and subsurface (B and C) horizons. We found the lowland soils had lower total annual CO2 efflux than the upland soils, with monthly CO2 efflux in lowlands most strongly correlated with redox potential (Eh). Lower CO2 efflux as compared to the uplands corresponded to greater C content and abundance of lignin-rich, higher-molecular-weight, chemically reduced organic compounds in the lowland surface soils (A horizons). In contrast, subsurface soils in the lowland position (Cg horizons) showed lower C content than the upland positions (C horizons), coinciding with lower abundance of root biomass and oxalate-extractable Fe (Feo, a proxy for protective Fe phases). Our linear mixed-effects model showed that Feo served as the strongest measured predictor of C content in upland soils, yet Feo had no predictive power in lowland soils. Instead, our model showed that Eh and oxalate-extractable Al (Alo, a proxy of protective Al phases) became significantly stronger predictors in the lowland soils. Combined, our results suggest that low redox potentials are the primary cause for C accumulation in seasonally flooded surface soils, likely due to selective preservation of organic compounds under anaerobic conditions. In seasonally flooded subsurface soils, however, C accumulation is limited due to lower C inputs through root biomass and the removal of reactive Fe phases under reducing conditions. Our findings demonstrate that C accrual in seasonally flooded mineral soil is primarily due to low redox potential in the surface soil and that the lack of protective metal phases leaves these C stocks highly vulnerable to climate change.