Potential in paleoclimate reconstruction of modern pollen assemblages from natural and human-induced vegetation along the Heilongjiang River basin, NE China

Abstract

The relationships among modern pollen, vegetation, climate and human activities can help improving the reliability of reconstruction of past vegetation, regional climate and human activities based on fossil pollen records. We used a dataset of 114 surface soil pollen samples from natural vegetation (wetlands, forests and grasslands) and human-induced vegetation (farmlands and residences) along the Heilongjiang River basin in northeast China to explore the relationships among modern pollen, vegetation, climate and human activities. The results indicated that surface pollen assemblages differentiated modern vegetation well in natural and human-induced vegetation types. The wetlands were mainly composed of Cyperaceae, along with Artemisia, weeds Poaceae (<35 μm) and Sanguisorba. The forests were predominated by Pinus and Betula. Artemisia, weeds Poaceae (<35 μm) and Chenopodiaceae were the most important pollen taxa in grasslands. The farmlands were characterized by Artemisia, Aster, Chenopodiaceae, cereal Poaceae (>35 μm) and Taraxacum. The pollen assemblages of residences were composed of weeds Poaceae (<35 μm), Chenopodiaceae and Salix. Ordination analyses based on main pollen taxa and climatic variables were used to determine the relationships between pollen and climate, suggesting the surface pollen assemblages were primarily influenced by the mean annual temperature (Tann) in northeast China. The statistical performance of transfer function between pollen and Tann were well indicating the modern pollen assemblages could be reliably used in paleoclimate reconstruction in our study area. Furthermore, human-induced vegetation had high frequencies of human-companion pollen taxa, such as Chenopodiaceae, Aster, Taraxacum and cereal Poaceae (>35 μm). Pollen concentrations of human-induced vegetation were lower than natural vegetation types, which could be used as an indicator of human influence intensity.