• Plants & Soils
• Rhododendrons
• Chemical Taxonomy
• Plant Publications

• Links
• Contact Us
• Search

• Prof. David Rankin
• School of Chemistry
• University of Edinburgh

Chemical Taxonomy

Being able to name plants is essential for communication about them, and it so has important consequences, not least for conservation. However, it is not always straightforward to identify a particular specimen using its appearance alone. Of course, the ideal would be to obtain a DNA profile, but this is not sufficiently routine (or cheap) to be used routinely, and in any case the DNA must be related to other physical characteristics which can define the boundaries of species.

Rhododendrons are particularly difficult to classify, as many species, sub-species and varieties seem to merge with one another in at least some part of their range, and there are many wild hybrids. Cultivated plants and herbarium specimens are not always truly representative of wild populations, and plants named as one taxon may in fact be another, while hybrids are often named as species. We are using analyses of the waxes on leaves of Rhododendron, particularly sub-section Taliensia, to help to resolve some of these problems of identity. The components of the waxes are separated by gas chromatography, and where necessary identified by gas chromatography-mass spectrometry.

• The waxes include a major non-polar component consisting mainly of straight-chain alkanes, C_nH_2n+2, where n can be odd numbers in the range 23 to 35. Each specimen has a distribution of alkanes, usually centred at either n = 27 or at n = 31. This provides a simple, unequivocal distinction between two groups of taxa within the sub-section Taliensia. Sometimes the results are surprising. Rhododendron phaeochrysum var. phaeochrysum is a C₃₁ taxon, whereas R. p. var. levistratum is a C₂₇ taxon. Such a fundamental difference raises the question whether these taxa should be separated at a higher level than merely varieties.
• By fitting the intensities of the component alkanes to a smooth curve, the maximum of the distribution can be located to within about 0.2 (units of n). This quite fine resolution allows distinctions to be drawn between more taxa, and so a partial key to the sub-section, based entirely on the alkanes in the leaf waxes, has been drawn up.
• Hybrids between species with n = 27 and n = 31 may have different wax distributions, with either a double maximum, or a wide distribution centred near n = 29. Thus some specimens, previously thought to be of a pure species, have been shown to be hybrids. In other cases, an unknown parent of a plant which was believed to be a hybrid can be identified. For example, a plant grown from seed from R. taliensia (n = 27) has wax with a distribution centred at n = 31. Of the possible neighbours of the source plant, only R. lacteum has this maximum, and the identification of the hybrid as R. taliensia X lacteum can thus be confirmed by both the wax chemistry and morphology.

Relevant publications