Predicted consequences of increased rainfall variability on soil carbon stocks in a semiarid environment

Research on the impacts of climate change on soil organic carbon (SOC) stocks has focused on the effects of changes in average climate, but the potential effects of increased climate variability, including more frequent extreme events, remain under-examined. In this study, set in a semiarid agricultural landscape in southeastern Australia, we used the Rothamsted carbon (RothC) model to isolate the effects of interannual rainfall variability on SOC stocks over a 50 yr period. We modelled SOC trends in response to 3 scenarios that had the same 50 yr average climate but different interannual rainfall distributions: non-changing average climate, historic variability (H), and increased variability due to more frequent extreme rainfall years (XH). Relative to the non-changing average climate, RothC simulations predicted net decreases in mean SOC stocks to 50 yr of 11% under the H scenario and 13% under the XH scenario. These decreases were the result of predicted SOC decreases (and increased CO₂ emissions) in extreme wet years (ca. 0.26 Mg ha⁻¹ yr⁻¹) that were not counterbalanced by SOC increases in extreme dry years (ca. 0.11 Mg ha⁻¹ yr⁻¹). No significant difference in mean SOC stocks at 50 yr between the H and XH scenarios was likely due to an increase in both extreme wet and counterbalancing extreme dry years in the latter. Strong negative correlations were found between annual changes in SOC stocks and rainfall. Our modelled predictions indicate the potential for extreme rainfall years to influence SOC gains and losses in semiarid environments and highlight the importance of maintaining plant inputs in these environments, particularly during extreme wet years.