Climate and vegetation codetermine the increased carbon burial rates in Tibetan Plateau lakes during the Holocene

The Tibetan Plateau (TP), also known as the “Asian Water Tower,” includes thousands of lakes containing sizable lacustrine sediments that have been proven to be important carbon pools of inland ecosystems. To develop an in-depth understanding of the status of TP lake carbon storage and its response to global changes and human activities, we accurately quantified the change trend of long-term organic carbon burial in lakes over the TP, compared levels of differentiation within the TP, and explored related driving factors of the carbon accumulation rates (CARs) by compiling 24 lake sediment cores. The results show that the lake CAR over the TP presented a rising trend during the Holocene, with an average of 6.50 g C m−2 yr−1 and levels ranging from 1.75 to 18.41 g C m−2 yr−1. The eastern TP (ETP), with warmer and wetter climate conditions and more vegetation coverage, had almost twice the CAR as that of the central and western TP (CWTP). We demonstrate that climate and vegetation could be the main drivers of the lake CAR on the TP despite environmental heterogeneity contributing to subregional discrepancies. During the early Holocene, superior summer temperatures and precipitation favored vegetation development, providing abundant material for lake carbon deposition. In the middle Holocene, together with vigorous vegetation, decreasing summer temperatures led to a decline in mineralization, which enhanced carbon storage efficiency and resulted in an increase in the CAR across the ETP. In contrast, for the CWTP, herbaceous-dominated vegetation continued to expand during the early-mid Holocene, which constrained detritus supply, thus maintaining a relatively stable CAR. However, intensive human activities could be responsible for the CAR peak in the late Holocene on the TP, especially since 2000 cal yr BP. In addition to the anthropogenic impacts, variations in CAR among lakes resulting from local factors lead to large uncertainties in future C predictions for the whole region of TP.