Annual variation in δ 13C values of maize and wheat: Effect on estimates of decadal scale soil carbon turnover

Abstract

On sites where C4-plants have replaced C3-plants, changes in soil δ 13C allow the turnover of C3- and C4-derived C to be separated. Studies of decadal scale turnover of soil C following conversion to C4-plants generally lack δ 13C values for previous C4-residue inputs and assume that estimates of C4-derived soil C to be based on a fixed δ 13C value. Further assumptions are that changes in the initial (time-zero) soil δ 13C values are insignificant following conversion to C4-plants. We tested these assumptions by measuring: 1) the δ 13C of annual samples of silage maize biomass (C4-plant) and winter wheat grains (C3-plant) grown during 1988 to 2006, and 2) the δ 13C of soil kept under bare fallow during 1956 to 1983. The δ 13C of plants was related to climate variables, and the impact of maize δ 13C was based on estimates of maize-derived soil C using different approaches to establish the δ 13C in maize inputs. The δ 13C of both maize and wheat decreased with time, but the rate of change and annual variations were considerably larger for wheat than for maize. Maize as well as wheat δ 13C was best related to year (probably reflecting a decrease in atmospheric δ 13C) and the water balance during the active growth period. Using the smallest (−12.44‰) and the largest (−11.26‰) δ 13C measured during 1988 and 2006, estimates of maize-derived C in soil after 18 years ranged from 13.2% to 14.2% of the soil total C. Despite a loss of 31% of the soil C pool under bare fallow, the increase in soil δ 13C was significant only at P < 0.10. We conclude that annual variations in maize δ 13C values and changes in the δ 13C of the soil C fraction derived from the pre-conversion C3-vegetation have only little impact on estimates of maize-derived soil C that cover a few decades. For estimates covering several decades to centuries, the subtle but consistent changes in plant and soil δ 13C need to be accounted for. The variability in δ 13C in wheat grains suggest that the use of any fixed δ 13C value for C3-residues in estimates of C turnover in soils on which C4-plants have been replaced by C3-plants can be associated with considerable uncertainty.