Abstract: Soil respiration is an important component of global carbon cycle. The rate of soil respiration varies with time. Accurate measurement of soil respiration rate is necessary to evaluate soil carbon emission. To evaluate soil respiration accurately, we conducted an incubation experiment with soil collected from a Zoigê alpine wetland. Soils were randomly separated into several parts after homogenization, and then were subjected to two or four drying-rewetting cycles. During each drying-rewetting cycle, CO₂ was collected for eight or four times. CO₂ emissions between each sampling time and the whole drying-rewetting cycle were compared. The results showed that the optimum sampling times were on the 4th day of the highfrequent drying-rewetting cycle (8 days per cycle). Under the lowfrequent cycle (16 days per cycle), the optimum sampling time was different between the high and low intensities, which was on the 6th day and the 4th day, respectively. An alternative approach was to collect the samples on the 4th day of the first cycle, and then sampled every 8 days. These findings suggest that the optimum sampling time of soil respiration depends on soil dryingrewetting cycle. Sampling during the middle period of drying-rewetting cycle accurately represents soil respiration during each cycle. Optimizing the sampling time could contribute to accurate evaluation and prediction of soil CO₂ emission in alpine wetland ecosystem under climate change.

Key words: winter wheat, cultivation pattern, radiation use, harvest index, yield.