Abstract:

Climate change can significantly affect the vegetation worldwide. Thus, paleovegetation and paleoclimate reconstruction should consider the quantitative relationship between modern vegetation and climate. The specific objectives of this study were (i) to assess the influence of environmental variables on pollen assemblages in the Kanas region, (ii) to reconstruct the evolution of vegetation over the past 3000 years using pollen records and (iii) to quantify historical climate change (including mean annual temperature and total annual precipitation) using a weighted averaging partial least squares regression method (WAPLS) applied to fossil pollen data from the Kanas wetland in Xinjiang, China.

A total of 65 surface and 50 fossil samples were collected from the Kanas wetland and analysed for 14C, pollen and grain size. By combining these data with those obtained from 214 samples of surface pollen assemblages in north Xinjiang, the late Holocene climate was reconstructed using a WAPLS model.

The vegetation in Kanas was dominated by forest for the past 3000 years, undergoing an arbour-vegetation transition from predominantly pine to spruce over that period. The WAPLS model showed that the paleoclimate progressed from cold-wet to warm-dry and subsequently back to cold-wet. Prior to 1350 calibrated years before the present (cal. yr BP), the climate of Kanas was cold and wet, and conditions became increasingly warm and dry until 870 cal. yr BP. The temperature reconstruction model indicated that a ‘Little Ice Age’ occurred ~380 cal. yr BP. These data will help us improve the understanding of abrupt climate change and provide important information regarding the prediction of climate.

Key words: climate change, Kanas wetland, north Xinjiang, pollen, grain size