Redox-driven colloidal mobility and its effects on carbon cycling in temperate paddy soils

Rice paddy soils are known to represent a large proportion of global terrestrial carbon (C) stocks (ca.10 Pg), accumulating organic C in the topsoil due to cultivation under submerged conditions. Apart from the limited mineralization under anoxic soil conditions resulting from frequent field flooding, other mechanisms involving the dynamic interactions between organic C and redox-active minerals particularly Fe (oxy)hydroxides, together with the transport of organic C to deep mineral horizons, can lead to long-term C stabilization. Our previous studies have shown that up to 30-50 g m^-2 of dissolved organic C (DOC, defined as <450 nm) and 25-40 g m^-2 of Fe²⁺ may be mobilized and translocated into the subsoil over a rice cropping season in temperate rice paddies, contributing to an increase in belowground C stocks. However, little is yet know on influence of frequent redox fluctuations on the contribution of colloidal organo-mineral associations to C mobilization and accrual in paddy subsoils.

We hypothesized that (i) redox fluctuations may lead to an overall increase in colloid dispersion (via reductive dissolution of Fe oxides, changes in soil pH, as well as neoformation of colloidal organo-mineral associations), and that (ii) colloidal mobility may represent an important C input to paddy subsoils. In order to evaluate the effects of redox fluctuations on colloid dynamics in situ, water-dispersible fine colloids (WDFC) were isolated from soils collected from different horizons along two profiles opened in adjacent plots under long-term paddy (P) and non-paddy (NP) management in NW Italy. Moreover, WDFC were also isolated from anaerobically-incubated topsoil samples to evaluate the changes in colloid dispersion under reducing conditions as a function of management. Colloidal size-fractionation and their elemental compositions were evaluated by asymmetric flow field-flow fractionation (AF4) coupled with OCD or ICP-MS. 

Our results evidenced that redox cycling favours colloidal stability in the topsoils, with a preferential dispersion of the smallest-sized colloidal C (<30 nm and 30-240 nm fractions), even though larger-sized colloidal C (>240 nm) contributes predominantly to the WDFC. Consequently, under long-term paddy management colloidal dispersion and transport along the soil profile were probably responsible for the lower amounts of colloidal C (and Fe) observed in the Ap topsoil horizons of P with respect to NP, as well as for the significant accumulation of colloidal C in correspondence with the Brd subsoil horizons just beneath the plough pan. These illuvial horizons were also particularly rich in small-sized (30-240 nm) colloidal Fe, Al and Si possibly due to mineral phase changes induced by redox fluctuations.

Our findings therefore indicate that downward mobilization of colloidal C associated with Fe (hydr)oxides (e.g. coprecipitates) or small aluminosilicate minerals, rather than dissolved organic C, may represent an important process driving organic C accrual in paddy subsoils. However, further insights are still required to entangle the contribution of the different mechanisms involved.