Continuous measurements of soil CO2 concentration and soil respiration from a grazed degraded peatland in the Sacramento-San Joaquin Delta

Soil respiration is an important component of the terrestrial carbon cycle, and is the primary process by which carbon is lost from drained soils in the Sacramento-San Joaquin Delta, California. Over the past century, the elevation of the soil in the Delta has subsided from 10-30 meters above sea level to roughly 10 meters below sea level due primarily to peat soil oxidation following drainage of the region for agriculture. Understanding the processes and factors controlling soil respiration, and reducing the uncertainties associated with its measurement is essential to improving our understanding and estimates of soil subsidence and carbon budgets of drained agricultural ecosystems in the Delta. In this study, we measured CO2 concentrations at various depths in the soil in a drained, grazed pasture on Sherman Island using solid-state CO2 sensors, and assessed the ability of the soil CO2 gradient method to estimate continuous soil respiration. The soil CO2 gradient method estimates soil respiration using both measurements of vertical gradients of soil CO2 concentration and CO2 diffusivity (Ds) in soil. Several models have been developed to estimate Ds, however, notable differences in Ds and hence soil respiration may be observed depending on which model is used. In this study, we directly measured diffusivity in undisturbed soil cores, and obtained a power law relationship between the relative gas diffusion coefficient (Ds/D0, where D0 is the diffusion coefficient in air) and air-filled porosity using the diffusion chamber method. Periodic chamber-based measurements of soil respiration were made to validate the continuous gradient method measurements. We also measured net ecosystem exchange (NEE) using the eddy covariance technique, and ecosystem photosynthesis and ecosystem respiration were obtained by partitioning gap-filled NEE. Results from this study highlight the importance of selecting and testing the appropriate model for estimating Ds in soil as large differences (up to an order of magnitude) in soil respiration were observed depending on which model was used. In this study, we developed a model specifically for soil at our site based on diffusion rates through undisturbed soil cores. Rates of soil respiration from the grazed pasture were on average 2.1 μmol m-2 s-1 throughout the study, with higher average respiration rates during winter (rainy season) than summer (dry season) (2.4 and 1.8 μmol m-2 s-1, respectively). Large pulses (up to 9 μmol m-2 s-1) in soil respiration were observed immediately following rain events. We found that the diurnal pattern of soil respiration during the winter was driven by soil temperature, while soil respiration in the summer was decoupled with soil temperature. During the summer soil respiration was correlated with photosynthesis, but with a lag of several hours. This was likely due to the slow translocation of photosynthates to the roots of a perennial plant (pepperweed) that dominates the canopy from April to October, and the associated microbial activity. This was not observed during the winter since the pepperweed senesces, and the canopy is dominated by a C3 grass with a much shallower root system. These results corroborate the findings of other studies that photosynthesis drives soil respiration in addition to soil temperature and moisture.