Pollen-based mapping of Holocene vegetation on the Qinghai-Tibetan Plateau in response to climate change

Abstract

Abstract Improved studies of past vegetation change are required to better understand the variation of alpine ecosystems on the Qinghai-Tibetan Plateau (QTP) in response to future climate change. Spatial and temporal variations of past vegetation can be traced by fossil pollen data mapping. In this paper, we synthesized 57 continuous pollen records on the QTP covering the past 15 kyr to depict large-scale vegetation change and its response to climate variations. In order to minimize potential chronological biases, age-depth models were revised using a state-of-the-art and consistent method for all the records. The spatial and temporal variation of major pollen taxa were examined based on interpolated pollen maps at 1000-year intervals. The arboreal pollen (AP, mainly of Pinus, Betula and Abies/Picea) content expressed significant climate signals over a broad spatial and temporal gradient. During the late glacial period, high proportions of AP widely occurred in regions that are presently unforested owing to the sparse local vegetation coverage. For the Holocene period, AP showed relatively high contributions in records from the southeastern margin of the QTP, with a decreasing gradient in abundance towards the northwest. The transportation of AP to unforested regions corresponds closely to the intensity of monsoon wind, which can be used to track the Holocene evolution of the summer monsoon. The dominant shrub and herbaceous taxa (including Artemisia, Chenopodiaceae, Cyperaceae, Poaceae and Ephedra) generally represent developments of local vegetation responding to climate variations. In addition, the persistent increase in Poaceae pollen during the mid to late Holocene correlates possibly to regional human activities. The inferred spatial and temporal patterns of major pollen types on the QTP provide significant knowledge about long-term vegetation change and its potential response to climate variations.