Long- and short-term precipitation effects on soil CO2 efflux: Patterns revealed only with high frequency measurements

Background/Question/Methods

Atmospheric CO₂ has risen 40% since the industrial revolution, and this increase is likely to influence the terrestrial C cycle and the environmental feedbacks that regulate it in complex ways. Soil efflux (E_soil), the main pathway that C moves from the biosphere to the atmosphere, is a critical flux in the terrestrial C cycle. As E_soil is variable spatially and temporally, documenting patterns at appropriate scales is paramount in our ability to make inferences regarding how changing environmental variables will affect C processes. Soil moisture, one of the most influential abiotic regulators of E_soil, can directly affect it by lowering gas diffusion through the soil by reducing its air-filled porosity, and can indirectly affect it by altering patterns in aboveground C assimilation and belowground C allocation. Disentangling these complex patterns and controls will require better approaches to rigorously measure E_soil patterns and examine its controls.

Our objectives were to test the impact of an altered precipitation regime on total belowground carbon allocation (TBCA), root biomass, and E_soil. We used two methods to quantify E_soil and TBCA: daytime biweekly manual measurements and automated continuous measurements. Our hypothesis is that the former method fails to capture short- and long-term effects of soil moisture on E_soil and TBCA, while the latter method allows the effects of soil moisture to be discerned. We carried out a field experiment in control and irrigated plots in a longleaf pine wiregrass savanna, a system wherein productivity is more constrained by water than temperature or nitrogen.

Results/Conclusions

Irrigation treatments increased soil moisture by 41.2% over non-irrigated plots. Root biomass was significantly higher in irrigated plots, with the increase largely concentrated in 0–24 cm depth. Root biomass in non-irrigated plots was more evenly distributed with depth. The method used, and thus measurement frequency, to estimate TBCA was important in the ability to detect significant differences between treatments. Estimations using only the manual biweekly measurements showed no difference between treatments; however, irrigated TBCA was significantly greater than that of the control when estimated with the continuous measurements. This is likely due to 1) the delayed increase in E_soil following irrigation pulses of soil moisture, variation that the biweekly manual measurements could not capture, and 2) the timing of biweekly manual measurements was restricted to mid-day, which because peak E_soil occurred ~4 hours earlier than peak solar radiation, this method likely systematically biased TBCA estimates lower than the continuous automated method.