coding flower and pollen dimorphism

Elmar Robbrecht E.Robbrecht at BR.FGOV.BE
Fri Oct 31 14:21:45 CST 1997


For our research on Rubiaceae-Cinchonoideae and Hedyotideae, we are looking
for a solution to code characters in relation to heterodistyly and pollen
dimorphism. None of the classical theoretical textbooks seems to give
information on this. A particular problem is the common occurrence of iso-
and heteromorphic taxa within the study group. One can easily design a set
of characters and states for the brevi- and longistylous morphs, but how to
combine these with the same (single) set for isostylous taxa?

For a cladistic matrix, one could (i) code dimorphic characters as a
polymorphism 0,1, or (ii) split the character for iso- and heterostylous
taxa and use ? (non applicable).

In a numerical phenetic analysis, it seems that there is no fundamental
obstacle against splitting the taxa, i.e. using the morphs as OTU's.

Zoologists probably do not encounter similar problems (groups with males,
females AND hermaphrodites non existing?).

Elmar Robbrecht
Suzy Huysmans
Steven Dessein
An Scheltens

Prof. Dr. Elmar Robbrecht
acting head of Department S.P.
National Botanic Garden of Belgium
Domein van Bouchout, B-1860 meise (Belgium)
Tel. **32 2 269 39 05; fax ** 32 2 270 15 67
email robbrecht at

NEW: information on N.B.G. on the internet (

NEW: Direct consultation of the database Genera of the Rubiaceae:

More information about the Taxacom mailing list