Modern pollen–climate relationships in north Xinjiang, northwestern China: Implications for pollen-based reconstruction of Holocene climate. (July 2017)

• Record Type: Journal Article
• Authors: Li, Furong; Zhao, Yan; Gaillard, Marie-José; Li, Huan; Sun, Jinghui; Xu, Qinghai
• Is Part Of: Holocene. Volume 27:Number 7(2017)
• Page Start: 951
• Page End: 966
• Publication Date: 2017-07
• Subjects: Altay Mountains -- China -- Jungar Desert -- modern pollen–climate relationships -- pollen ratios -- variation partitioning; Geology, Stratigraphic -- Holocene -- Periodicals; Paleoclimatology -- Periodicals; 333.7
• DOI: 10.1177/0959683616678464
• Languages: English
• ISSNs: 0959-6836

Abstract

Fossil pollen records are widely used to reconstruct past climate. Such reconstructions require that the relationships between pollen assemblages, vegetation, and climate are well understood. These can be studied in present circumstances given we assume that modern vegetation and climate are analogous to past ones. In this study, we analyze pollen–vegetation–climate relationships in the Jungar desert and Altay Mountains, northwestern China, a region for which careful reconstruction of past climate is needed to answer unsolved questions on past climate in an area located at the boundary between two different climate regimes (westerlies and monsoon). We use a dataset of 66 surface pollen samples from forest, meadow, steppe, and desert vegetation and six related climate variables, T ann, T Jan, T Jul, P ann, P Jan, and P Jul . Principal components analysis, redundancy analysis, Monte Carlo permutation tests, and variation partitioning are applied to quantify these relationships. We also assess pollen ratios as indices of aridity. We find that (1) P ann is the major climatic factor influencing pollen assemblages, followed by P Jul, (2) the two variables are not correlated, and (3) the shared effect of (1) P Jan and P Jul, (2) P Jan and P ann, (3) P Jul and T ann, and (4) T ann, T Jan, and T Jul explains a larger portion of the variation in pollen data than the individual effect of each variable. Therefore, robust pollen–climate transfer functions can be developed for P ann and P Jul, and several climate variables treated in combination. Artemisia /Chenopodiaceae is a strong index of aridity and Artemisia /Gramineae might be a useful index of P ann and P Jul .