Pollen-vegetation/land use relationships in southeastern China: Complexity and applicability for paleoenvironmental reconstruction

Abstract

Fossil pollen, as a key indicator of vegetation landscapes, can provide abundant information of long-term land use dynamics. However, pollen-vegetation relationships are complex, and the relevant studies are rare in the large sedimentary basins associated with strong human disturbances. This study aims to investigate the pollen-vegetation/land use relationships at a large regional scale, and to supplement reference for more accurate paleoenvironmental reconstruction in the future. A total of 124 modern pollen samples were collected from 66 lakes in the Yangtze-Huaihe River Basin (YHRB), southeastern China. We conducted a contrastive analysis of both pollen and land use data in different spatial extents and categories. Our results reveal that the pollen assemblages of the YHRB are dominated by both arboreal and herbaceous taxa, such as Pinus, Quercus, Pterocarya, Juglans, Poaceae, Artemisia, Cruciferae, Chenopodiaceae, Compositae, and Humulus. The spatial differences of pollen assemblages in the YHRB are obvious, resulted from the confounded effects of natural environment and human disturbance. Hydrologic process plays a key role in pollen dispersal and deposition in the YHRB. Waterborne pollen is an important component of the lacustrine pollen assemblages, and fern spores and total pollen concentration can be used to indicate the degree of hydrologic process. The main source of lacustrine pollen here generally derives from sub-basins, and the source area is about 80 km. Anthropogenic impact is another significant factor determining the pollen-vegetation relationships in the YHRB. Cereal and anthropochore pollen as well as pollen diversity indexes are ideal indicators of agricultural activities. We suggest that the 4 types of estimated pollen/land-use ratios in this study, covering forest, farmland, rural land, and wetland could be used as a reference for reconstructing past land cover. Our study also demonstrates the existing problems and the potential improvements for the further studies on pollen-based paleoenvironmental reconstruction.