Drivers of Soil Respiration Partitioning in a Seasonally Dry Mid-Latitude Forest

When considering belowground ecosystem responses to climate change, the soil is often treated as a monolith, i.e., considering root and soil microbial respiration together as the total soil respiration flux. However, plant roots and soil microorganisms represent fundamentally different components of belowground ecosystems with different potential sensitivities to drought, temperature extremes, and diurnal, seasonal, and inter-annual changes in plant productivity. We measured the climate sensitivity of belowground processes across multiple temporal scales (i.e., diurnal, seasonal, inter-annual), using long-term measurements from the Missouri Ozark AmeriFlux (MOFLUX) site. This heavily instrumented Quercus-Carya (oak-hickory) forest is located in the Ozark Border Region of Central Missouri, which experiences seasonal soil water deficits because of high precipitation variability, evapotranspiration, and vapor pressure deficit. The MOFLUX site has measured hourly soil respiration using 8 or 16 automated chambers since 2004, with a subset of chambers over plots where roots were excluded since 2017. Using a combination of statistical, machine learning, and wavelet coherence analysis, we found that heterotrophic (i.e., microbial) respiration is most responsive to soil temperature at daily and seasonal timescales, while autotrophic (i.e., root) respiration is most responsive to aboveground productivity (using LAI as a proxy) and the time of the year. Microbial and root respiration have similar responses to temperature but not to LAI or time of year, and as a result, the time of year and LAI are the most influential factors on the partitioning between microbial and root respiration, suggesting that the partitioning can be inferred from these variables. Soil moisture exerts the strongest influence on soil respiration on synoptic weekly-to-monthly timescales. We used the Multi-Assumption Architecture and Testbed (MAAT) to explore the ability of different model structures to capture climate extremes, with a focus on dry-wet transitions. Different representations of soil water dynamics in models result in diverging predictions, but long-term measurements of belowground plant-water-energy dynamics can both constrain historical predictions and inform future projections.