Multi-proxy Holocene palaeoclimate records from Walton Moss, northern England and Dosenmoor, northern Germany, assessed using three statistical approaches

Abstract

Plant macrofossil analysis is a robust and established method for estimating palaeoclimatic changes from raised peat bogs. There remains uncertainty, however, over the best statistical method to use for converting the palaeoecological data to an index of bog surface wetness (BSW). To test this, a 9 m core (WLM22) from Walton Moss, northern England was analysed for changes in plant macrofossils, testate amoebae and humification. The plant macrofossil data were then converted to a single curve of bog surface wetness (BSW) using three different techniques: weighted averaging by the Dupont Hydroclimatic Index (DHI) and ordination by Detrended Correspondence Analysis (DCA) and by Non-metric Multidimensional Scaling (NMDS). Through comparison with testate amoebae and humification data from the same core, it was observed that a statistically better fit was found with the wetness curve from DHI than that from either ordination. It is suggested that the weighted averaging method of DHI can provide a good indicator of BSW variation in long peat records, where the profile incorporates data from both Sphagnum-dominated and Eriophorum/ericaceous-dominated bog deposits, which is common to NW European bogs. The data also imply that, with sufficient diversity of taxa, there is potential for the further development of plant macrofossil transfer functions. Application of DHI to an additional data set from Dosenmoor, northern Germany and comparison with published records from NW Europe revealed coherent regional climatic variations.