Carbon dioxide efflux and pCO 2 in soils of threeQuercus ilex montane forests

Soil CO2 efflux and pCO2 in the soil atmosphere were measured during one year at three montane sites of Mediterranean sclerophyllous forests in NE Spain. Two sites were located in the upper and lower slopes of a small catchment in the Prades mountains (mean precipitation 550 mm year-'), and a third site was located on a lower slope in the Montseny mountains (mean precipitation 900 mm year-'). The three sites were similar in bedrock and vegetation, but differed in soil characteristics and water availability. Seasonal variation of CO2 efflux and soil pCO2 were affected by soil temperature and, to a lesser extent, by soil moisture. Annual mean soil CO2 efflux (considered as soil respiration) was similar at Montseny and at the comparably located site at Prades (83 ± 18 S.E. vs. 75 4 9 mg CO2 m - 2 hour- , respectively), and was highest at the Prades upper slope site (122 ± 22 mg CO2 m -2 hour-l). Despite those relatively similar CO 2 effluxes, mean soil pCO2 was much higher at both Prades sites than at Montseny. Soil pCO2 always increased with depth at Prades while maxima pCO2 at Montseny were often at 20-30 cm depth. A model based on gas diffusion theory was able to explain why soil pCO2 was much higher at Prades than at Montseny, and to reproduce the shape of the vertical profile of pCO2 at the Prades soils. Nevertheless, the model failed to simulate the soil pCO2 maximum found at 20-30 cm depth at the Montseny site. Model simulations using a time-variable CO2 production rate suggested that pCO2 maxima at intermediate depth could be the result of a transient situation instead of an equilibrium one.