Oecologia

April 1987, Volume 72, Issue 1, pp 137-147

A critical analysis of techniques for measuring tannins in ecological studies

S. Mole, P. G. Waterman

A series of seventeen plant extracts rich in phenolic materials, including condensed and hydrolysable tannins, have been subjected to a series of chemical analyses in an attempt to gather ecologically significant information about their structure. Procedures investigated were (i) the Folin-Denis and Hagerman and Butler methods for quantifying total phenolics, (ii) the vanillin and proanthocyanidin methods for quantifying condensed tannins, (iii) the iodate and nitrous acid methods for hydrolysable tannins. It was found that the techniques for total phenolics correlated well, the Hagerman and Butler method giving higher estimates where solutions were particularly phenol rich. By contrast there was considerable discrepancy between the methods examined for condensed tannins. This is probably due primarily to the very different chemical reactions that form the basis of these procedures and also to the fact that the extract dependent products of the proanthocyanidin method vary in their A 1 1values. During the study of condensed tannins methods for estimating their contribution to total phenolics and for measuring their average polymer length were examined. In both cases different procedures produced very variable results. Available methods for hydrolysable tannins were found not to be generally applicable across all extracts thought to contain this type of tannin on the basis of chromatographic analysis. An attempt to produce a quantitative spectrophotometric assay for hydrolysable tannins based on changes in reactivity to ferric chloride due to hydrolysis is described. This proved to be of limited sensitivity but may have some merit for estimating levels in hydrolysable tannins in phenol-rich plant extracts that also contain condensed tannins. It is concluded that whilst the overall level of phenolics in extracts can be estimated with some confidence the information imparted by more specific assays is very dependent on the procedures employed, particularly when dealing with extracts from taxonomically highly diverse sources.