Temporal dichotomy in the hydrological and physiological controls of soil respiration in a forested wetland in Southeast US

Background/Question/Methods

The complexity of soil respiration (SR) process determines the difficulties of quantification and mechanism theory development. Recent development of automated chamber technology provides the CO₂ flux datasets with improved frequency, and also brings the uncertainties related with temporal variations. Besides the broadly discussed confounded effects from temperature and water, the diurnal variation also has significant influences on modeling the seasonal variation. The forested-wetland ecosystems have special physical environments controlled by hydrological conditions, which are relatively stable in diurnal scale but have strong seasonal or annual variations. This characteristic offers a good opportunity to explore the potential controlling factors of SR in different time scales and study the effect of temporal variations on modeling. In a forested wetland located at Alligator River National Wildlife Refuge in eastern North Carolina, we carry out the soil respiration research in order to identify the controlling factors, quantify the factor effects and estimate the soil carbon loss.

Results/Conclusions

Soil respiration (SR) was measured in 2010 using a four-chamber automated system. During peak fluxes in June and July the mean (±SD) soil temperature at 5 cm was 22.6±1.3 °C, groundwater depth (GWT) was -31.8±6.2 cm and SR was 9.9±2.8 μmol·C·m^-2·s^-1, excluding days with precipitation>10 mm·d^-1. During non-flooded conditions the temperature-normalized base respiration (R₂₀) increased with the drawdown of groundwater. Similarly, Q₁₀ varied with groundwater depth, but exhibited an optimum at GWT=-5 ~ -10 cm. Inclusion of the hydrological control of R₂₀ and Q₁₀ in the model increased the amount of explained daily variation from 53% to 77%. However, temperature and water table fluctuation (when soil temperature and groundwater depth remained invariant) did not explain a significant proportion of diurnal variation, which accounted for 70% of overall variation at this period. Instead, diurnal variation in SR was perhaps primarily accounted for by gross canopy photosynthesis. Given the magnitude of diurnal variation in SR (5 ~ 14 μmol·C·m^-2·s^-1), proper accounting of it is crucial for accurate temporal integration. At our study site, ignoring the diurnal variability in SR could result in an underestimate on monthly and annual timescales.