Abstract:CO2 efflux (ECO2) from stems and branches is important in the regulation of biomass productivity and maintenance of carbon balance in terrestrial ecosystems. ECO2 is a temperature-sensitive physiological process. An increase in temperature could promote the activities of enzymes, decrease the gas solubility, and improve the diffusion coefficient of CO2. Until recently, great uncertainties relevant to the variations in ECO2 and their temperature-sensitivity (expressed in terms of Q10) remain poorly assessed because of the lack of continuous sampling, both in time series and space sequence. Understanding the effect of temperature on ECO2 at different time scales is important to accurately upscale the chamber-based CO2 measurements to the whole-stem. To investigate the temporal dynamics and vertical variations in woody-tissue ECO2 and their temperature sensitivity, the hourly ECO2 at four different heights (10, 130, 240, and 370 cm) of the four typical tree species (Sophora japonica, Salix matsudana, Larix principis-rupprechtii, and Platycladus orientalis) was observed using an Li-8100 automatic instrument (LI-Cor, Inc, NE, USA) on two consecutive days per month throughout 2014 in an artificial mixed forest plantation near Beijing. Simultaneously, the meteorological conditions (air temperature, relative humidity, and woody-tissue temperature) and stem radial increment were monitored using the HOBO Pro V2 probes (Onset Computer Corporation, Bourne, MA, USA) and dendrometer (CR10X, Campbell Scientific, Logan, UT, USA), respectively. The results indicated that:(1) the ECO2 of the four tree species showed similar seasonal variations, characterized by a significantly higher ECO2 in the growing months than in the dormant months. Together, the woody-tissue temperature (TW) and stem radial diameter increment (Di) of the trees explained the seasonal variation of ECO2; (2) Our study showed that the temperature-sensitivity coefficient, Q10, was not a constant, as assumed in some models, but was instead highly dependent on the measurement temperature. A lower Q10 was observed in the summer months, but a higher value was observed in the dormant months. A strong vertical variation in Q10 was observed during the growing season, showing a gradient increment with height along the trunk, and the vertical variation disappeared in the dormant months; (3) Based on the strength of correlations between stem (TW) and air temperature (TA), the daytime was divided into four periods. In growing months, the night-time temperature-sensitivity coefficient Q10 was significantly higher than that of day-time values, especially for the two deciduous species (Sophora japonica and Salix matsudana). This result implies that ECO2 is more sensitive to temperature in the night than in the daytime, and the temperature responses of ECO2 estimated by only daytime measurement can lead to the underestimated stem respiration, especially considering that the temperature increase is faster during the night-time. In contrast, the diurnal asymmetry of Q10 disappeared in the dormant months, i.e., no significant difference was observed between each of the four periods. The quantitative analysis of the temporal dynamics and vertical variations of ECO2, as well as what factors are driving these changes, would help us to improve our knowledge of the ECO2 processes and understand how climatic changes affect forest carbon budgets.