Diatom-based pH reconstruction of lake acidification using canonical correspondence analysis

Methods of quantitative pH reconstruction based on diatoms in sediment cores were considered. Current methods using multiple linear regression were limited by the inappropriate assumption that biological variables were monotonically related to environmental variables such as pH. The use of canonical correspondence analysis and weighted averaging was more appropriate. This method was applied to a data set and to surface samples. There was good agreement between measured pH and reconstructed values. The technique was applied to sediment core data from south-west Scotland. There was good agreement with the results of other pH reconstruction methods.