Disentangling the Effects of Temperature, Moisture, and Substrate Availability on Soil CO₂ Efflux

Soil respiration ( R s ), the largest carbon emission flux in ecosystems, is usually modeled as an empirically parameterized function of temperature, and sometimes water availability. The likely contribution by other factors, such as carbohydrate substrate supply from photosynthesis, has been recognized, but modeling capacity to use this information is limited. Wavelet transformations of the residuals of a seasonal Q 10 temperature response model indicated structure at different temporal scales. We hypothesize that this indicates the lack of explicit representation of relevant processes in the models. Using cross‐spectral analysis, we found that time series of photosynthetically active radiation explained most of the diurnal variation, temperature, explained variability at multiple time scales (diurnal‐synoptic), whereas volumetric soil water content correlated with variability in R s at scales 15‐30 days. The results suggest that the time domains of influence for different driving variables of R s are discrete, and largely nonoverlapping, and represent functional relationships between soil biological activity and its constraints. Analysis of phase angles showed that R s lagged photosynthetically active radiation by 1.5‐3.0 hr. As this time lag was the same in both young and mature trees, with more than fivefold difference in transport distances, we hypothesize that this finding adds to the body of literature that support the pressure‐concentration‐wave model of carbohydrate availability in plants.