To better understand the biotic and abiotic factors that control soil CO2 efflux, we compared seasonal and diurnal variations in simultaneously measured forest-floor CO2effluxes and soil CO2 concentration profiles in a 54-year-old Douglas fir forest on the east coast of Vancouver Island. We used small solid-state infrared CO2 sensors for long-term continuous real-time measurement of CO2 concentrations at different depths, and measured half-hourly soil CO2 effluxes with an automated non-steady-state chamber. We describe a simple steady-state method to measure CO2 diffusivity in undisturbed soil cores. The method accounts for the CO2 production in the soil and uses an analytical solution to the diffusion equation. The diffusivity was related to air-filled porosity by a power law function, which was independent of soil depth. CO2 concentration at all depths increased with increase in soil temperature, likely due to a rise in CO2 production, and with increase in soil water content due to decreased diffusivity or increased CO2production or both. It also increased with soil depth reaching almost 10 mmol mol−1 at the 50-cm depth. Annually, soil CO2 efflux was best described by an exponential function of soil temperature at the 5-cm depth, with the reference efflux at 10 °C (F10) of 2.6 μmol m−2 s−1 and the Q10 of 3.7. No evidence of displacement of CO2-rich soil air with rain was observed.
Effluxes calculated from soil CO2 concentration gradients near the surface closely agreed with the measured effluxes. Calculations indicated that more than 75% of the soil CO2 efflux originated in the top 20 cm soil. Calculated CO2 production varied with soil temperature, soil water content and season, and when scaled to 10 °C also showed some diurnal variation. Soil CO2 efflux and concentrations as well as soil temperature at the 5-cm depth varied in phase. Changes in CO2 storage in the 0–50 cm soil layer were an order of magnitude smaller than measured effluxes. Soil CO2 efflux was proportional to CO2 concentration at the 50-cm depth with the slope determined by soil water content, which was consistent with a simple steady-state analytical model of diffusive transport of CO2 in the soil. The latter proved successful in calculating effluxes during 2004.