Alpine wetlands play a sensitive function in global carbon cycle during the ongoing climate warming, yet the temporal patterns of carbon dynamics from in situ ground-based long-term observations remain unclear. Here, we analyzed the continuous net ecosystem CO 2 exchange (NEE) measured with the eddy covariance technique over an alpine peatland on the northeastern Qinghai-Tibetan Plateau from 2007 to 2016. The wetland acted as a net CO 2 source with a positive NEE (120.4 ± 34.8 gC m −2 year −1 , Mean ± SD), with the mean annual gross primary productivity (GPP) of 500.3 ± 59.4 gC m −2 year −1 and annual ecosystem respiration (RES) of 620.7 ± 74.2 gC m −2 year −1. At the seasonal scale, the classification and regression trees (CART) analysis showed that aggregated growing season degree days (GDDs) were the predominant determinant on variations in monthly NEE and monthly GPP. Variations in monthly RES were determined by soil temperature (Ts). Furthermore, nongrowing season Ts had a significant positive correlation with the following year annual GPP (p < 0.05). Nongrowing season RES only accounted for about 25% of annual RES but had significant correlation with annual RES and annual NEE (p < 0.05). The further partial correlation analysis showed that nongrowing season air temperature (Ta, p = 0.05), rather than precipitation (PPT, p = 0.25) was a predominant determinant on variations in annual NEE. Our results highlighted the importance in carbon dynamics of climate fluctuations and CO 2 emission from the nongrowing season in alpine wetlands. We speculated that the vast peadlands would positively feedback to climate change on the Tibetan plateau where the nongrowing season warming was significant.

Biomass temporal stability plays a key role in maintaining sustainable ecosystem functions and services of grasslands, and climate change has exerted a profound impact on plant biomass. However, it remains unclear how the community biomass stability in alpine meadows responds to changes in some climate factors (e.g., tem-

The amplitude of the diurnal temperature (ADT) has been decreasing under climate change, with substantial anticipated effects on alpine grassland carbon budgets. Here, we quantified the temporal response of the growing seasonal CO2 fluxes to ADT over alpine shrubland on Qinghai‐Tibetan Plateau (QTP) from 2003 to 2016. At a daily scale, net ecosystem exchange (NEE) and gross primary production (GPP) quadratically responded to ADT with optimum values of 15.4 and 13.4°C, respectively. Ecosystem respiration (RES) negatively linearly correlated with ADT. Partial correlation, and classification and regression trees (CART) analysis, both showed that the maximal (MaxTa) or minimal air temperature (MinTa), rather than ADT, played much more important role in daily variations of CO2 fluxes. At a monthly scale, GPP and NEE were both positively and negatively controlled by MaxTa while RES was negatively determined by MinTa, respectively. Monthly ADT exerted a negligible influence on monthly CO2 fluxes. At an annual scale, only MaxTa played a significant role in variations of GPP and RES. NEE did not significantly respond to ADT, MaxTa or MinTa. The little direct correlations between NEE and ADT at daily, monthly or annual scales contradicts a previous hypothesis that a larger ADT would enhance carbon sequestration capacity over alpine ecosystems. Given the positive impact of MaxTa on GPP and MinTa on RES, our study would suggest that a decreasing ADT could indirectly stimulate more carbon loss and weaken the carbon sequestration capacity of alpine shrublands under the scenario of further increases in MinTa over QTP.