A novel procedure for quantitative regional paleoclimatic reconstruction using surface pollen assemblages

Abstract

Since the inception of pollen analysis, climatic reconstruction using fossil pollen records has been one of its major applications. However, modern pollen-climate relationships have been mainly established using surface soil pollen assemblages (including moss cushions and trap samples), and pollen assemblages from the surface sediments of lakes (and peatlands) have been used less frequently. Previous studies have indicated that surface soil pollen assemblages are derived from a single (local) plant community, with an additional contribution from the regional vegetation, whereas pollen assemblages from the surface sediments of lakes are derived from multiple plant communities within the watershed. This may lead to substantial uncertainties in the application of surface soil pollen assemblages to establishing pollen-climate relationships for use in paleoclimate reconstructions using fossil pollen records from lake sediment cores. Here, we present a novel procedure for quantitative regional paleoclimatic reconstruction using a regional integration method which attempts to resolve mismatches between surface soil pollen assemblages and surface lake sediment pollen assemblages. Principal Co-ordinates Analysis (PCoA) of pollen assemblages from surface lake sediments and surface soils, and regional integrated pollen assemblages indicates that the composition of regional integrated pollen assemblages and surface lake sediment pollen assemblages is very similar. However, the integrated optimal scales of surface soil pollen assemblages are various between different lakes. Based on the regional integrated pollen dataset, reconstructed mean annual temperature and annual precipitation for different lakes in North China are close to modern observations, suggesting that the dataset has great potential for quantitative regional paleoclimatic reconstruction.