Predicting the Effect of Decadal Soil Warming on Temporal CO2 Losses from a Temperate Forests Soil

Abstract:

As a consequence of the predicted warming of the atmosphere over the coming decades, soils are expected to warm at similar rates. This warming might have profound consequences for the rate at which biogeochemical processes take place in soils, such as an enhanced rate of soil organic carbon mineralization to CO₂. Making reliable predictions of the response of soil organic carbon to soil warming is therefore necessary to assess the extent to which this process might constitute a positive feedback to global climatic change over the coming decades.

Therefore, we applied a state-of-the-art soil biogeochemical model (ReSOM, Tang and Riley (2015)) to two long-term soil warming experiments (+ 5 °C) at Harvard Forest (Massachusetts, USA), which have been running for 17 and 29 years. Our aims are to (1) assess the extent to which ReSOM is able to correctly simulate the effect of long-term soil warming on soil organic carbon stocks and temporal patterns of CO₂ fluxes, (2) assess the importance of including thermal adaptation of soil microbes (using macromolecular rate theory) and temporal patterns in soil moisture content in ReSOM to correctly simulate intra- and inter-annual soil organic carbon dynamics, and (3) check the reliability of model predictions to make forecasts on decadal timescales when calibrated using data from a decade of soil warming.

Our results show that including thermal adaptation of soil microbes is necessary to reliably predict the effect of decadal soil warming on intra- and inter-annual CO₂ fluxes. In addition, the limiting effect of soil moisture on microbial decomposition of soil organic carbon needs to be simulated in order to correctly simulate the effect of soil warming at different forest sites in the same region. Last, we assessed the effect of including or excluding the effect of (1) soil moisture and (2) thermal adaptation of soil microbes on simulated changes in soil organic carbon stocks over the next century. We found diverging predictions of soil organic carbon losses under these different scenarios, showing the importance of correctly including and parameterizing these processes.