Diurnal and seasonal variability of soil CO2 efflux in a cropland ecosystem on the Tibetan Plateau

Seasonal variation of soil CO2 efflux is highly dependent on climate and plant phenology. On the Tibetan Plateau with an average altitude about 4000 m a.s.l., variations of CO2 efflux may be strongly associated with the unique ecophysiology and climate over there. Our objectives were to quantify diurnal and seasonal variations of soil CO2 efflux, and to investigate the effects of daily and seasonal variations of soil temperature and phenology on soil CO2 efflux. CO2 efflux was measured in Lhasa River valley on the Tibetan Plateau using a static closed chamber technique and gas chromatography for 2 full years from September 1999 to August 2001. Soil CO2 effluxes showed similar diurnal change patterns and fluctuated from minimum at 5:00 h to maximum at 11:00–14:00 h in different phenological stages. Soil CO2 efflux exhibited pronounced variation corresponding to crop phenology and soil temperature, with a minimum value of 0.4 g CO2 m−2 d−1 in January and a maximum value of 15.0 g CO2 m−2 d−1 in mid June. The observed mean soil CO2 effluxes were 6.6 g CO2 m−2 d−1 in the growing season and 2.8 g CO2 m−2 d−1 in the non-growing season, with an average of 6.0 g CO2 m−2 d−1 in the 2 full measured years. While soil CO2 efflux was strongly dependent on soil temperature with highest correlation found with 5 cm depth temperature, maximum values of CO2 efflux coincided with maximum values of leaf area index (LAI) and live root biomass (LRB) in mid June but not with maximum soil temperature in July. CO2 efflux was positively correlated with LAI and LRB in the growing season from March to August. Soil moisture had relatively little effect on soil CO2 efflux due to frequent irrigation. Simulation of soil CO2 efflux with soil temperature in the whole 2 years of observation led to overestimates in non-growing season and underestimates in the growing season. Taking account of the influence of crop phenology, a temperature dependent exponential model was separately used to fit CO2 efflux in growing season and non-growing season. This alteration increased 7.5% of explained variance of seasonal variability and provided more accurate prediction of soil CO2 efflux. Q10 values ranged 2.0 in growing season and 2.5 in non-growing season. Total annual loss of C from soil respiration was estimated to be 579 ± 13 g C m−2 per year at the site. The results suggest that soil temperature is the determinant factor controlling temporal variation of soil CO2 efflux and crop phenology modifies the temperature dependence of soil CO2 efflux in different growing periods. Our results also indicate that root respiration in the growing season can be estimated approximately by using the discrepancy between CO2 efflux relations in the growing season and non-growing season of cropland ecosystem.