LAUSR

Tibetan alpine permafrost soils functioned as a significant net carbon sink during 2000 to 2010 (1), causing a negative climate–carbon feedback under a changing climate; however, the current and future fates of this sink across the Tibetan Plateau are unknown. In PNAS, Wei et al. (2) present interesting results with combinations of 32 eddy covariance in situ measurements and 16 warming manipulative experiments as well as 18 model simulations across Tibetan alpine ecosystems. They claim that plant uptake of CO2 outpaces CO2 losses from permafrost and plant respiration on the Tibetan Plateau and that plants played a dominant role in cooling this plateau under a warming climate. In principle, we agree with this notion; however, we also propose that multiple anthropogenic activities, e.g., peatland draining and grassland grazing, should be considered, particularly on the eastern Tibetan Plateau (3), to robustly estimate net ecosystem production (NEP) across Tibetan alpine ecosystems. In PNAS, Wei et al. (2) mainly consider the potential influences of ecosystem type, altitude, and permafrost status (continuous or noncontinuous permafrost) but neglect the potential influences of multiple human disturbances on the net ecosystem carbon balance on the Tibetan Plateau. Using the NEP datasets from Wei et al. (2), we found significantly (P < 0.001) higher annual NEP for natural alpine peatlands (mean ± 95% confident interval: 1,784 ± 454 kg C ha 1 y ) than for drained peatlands (–1,180 ± 208 kg C ha 1 y ) (Fig. 1A), while there were no significant differences in annual NEP between natural and grazed meadows and shrublands (Fig. 1B). In addition, livestock intake removed considerable amounts of net primary production from drained and grazed peatlands and grasslands, while less net primary production was returned through livestock excreta (stocking rate ranges from 1.48 to 3.49 sheep unit ha 1 y ) (4); therefore, one can deduce that multiple human activities will further reduce the capacity of alpine peatlands and grasslands to serve as net carbon sinks. Given that previous reports estimated that 27% of the Zioge natural peatlands (the world’s largest plateau peatlands) have been drained and used for grazing (5) and that grazing was a dominant land-use type for Tibetan alpine grasslands (most widely distributed alpine ecosystems) (6), Wei et al. (2), to a certain extent, overestimate net carbon uptake by Tibetan alpine ecosystems. Ignoring these important carbon export components may collectively explain why Wei et al. (2) found an approximately four times higher carbon fixation than that in previous model predictions (7–10) or soil resampling measurements on the Tibetan Plateau (1). From this perspective, multiple anthropogenic activities should be considered in evaluating the net ecosystem carbon balance of Tibetan alpine ecosystems in future studies. Incorporating this information Fig. 1. NEP of alpine ecosystems under different disturbances. ***P < 0.001; n.s., not significant. The datasets are directly extracted from Wei et al. (2).