Indirect Correspondence Analysis and Botanical Macroremains : a case Study
Otto Brinkkemper · Leiden Repository (Leiden University) · 1993
Correspondence Analysis is one of the multivariate Statistical analyses used to explore large data sets.Application to a data set of 106 samples and 246 taxa of waterlogged botanical macroremainsfrom bon Age and Roman Period scttlements on Voorne-Putten (the Netherlands) yieldcd interesting results.The main distinction between the sites corresponded to their location either in the eastern or in the western part of the area studied.This eonfirms the results ohtained hy Cluster Analysis, an alternative multivariate technique.The relation hetween the samples and the taxa is far more clearly expressed in Correspondence Analysis.Bolh salinitx and moisture appear to bc important factors.There is a clear correlation between the occurrence ofwildplants indicating salinity and the erop plant barlcy.This agrees well with experiments revealing the salt tolerante of this erop.Differcnccs in erop plants are also closely related to the daling of the sites.There is a reniarkable rediiction in the numhcr of crops cultivated during the Roman Period.which can he seen as an indication ofarable specialisation.