The Geography of Phytochemical Races

186 4 Intercontinental Disjunctions
complete shutout in the case of hypolaetin, 0.48–0.0, and tricin, 0.17–0.0. Although
taxa in both areas have the capacity to make fl avonols that lack O-methylation,
that is, kaempferol, quercetin, myricetin, and gossypetin about equally, taxa from
South Africa have a far greater tendency to make O-methylated derivatives of
those same fl avonols. This is particularly pronounced in the formation of larycitrin
and syringetin, which are not known from genera in Australia. The formation of
isorhamnetin (quercetin 3′-methyl ether) [322] and tricin (5,7,4′-trihydroxy-3′,5′dimethoxyfl
avone) [323] happens only in Australian taxa, however, in contrast to
the compounds just mentioned. A study of the O-methyltransferases in these taxa
might uncover some interesting differences in substrate specifi city as in the case of
Chrysosplenium (De Luca and Ibrahim, 1985a, b). Proanthocyanidins are clearly
a major component of South African taxa (0.88) as opposed to Australian (0.04).
In contrast, sulfated derivatives are much more common in Australian taxa (0.27)
than in African ones (0.02). The chemical information clearly supports the newly
restructured taxonomy of the family.
Williams et al. (1998) presented a short discussion of the distribution of the
O-methylated fl avonoids in Restionaceae, with reference to the occurrence of
isorhamnetin in members of the related families, Anarthriaceae, Hopkinsiaceae,
Lyginiaceae, and Ecdeiocoleaceae. They suggested that the absence of isorhamnetin
from African members of Restionaceae, therefore, represents an advanced feature,
and that the presence of hypolaetin in a large fraction of Australian taxa is an
advanced character. The diffi culty in determining polarity in fl avonoid characters
is well known. Suggestions as to the degree to which a given fl avonoid, or
fl avonoid profi le, refl ects a primitive (pleisiomorphic) state or an advanced (apomorphic)
state remain only speculative until some reliable phylogeny of the taxa
involved has been achieved. In the present case, a phylogeny of the families, as
well as the genera within Restionaceae, is absolutely essential. When an understanding
of the phylogenetic relationships among the taxa has been achieved, it
would then become possible to infer the order of evolutionary events underlying
the fl avonoid profi les. Until then, the patterns of fl avonoid occurrence in the Restionaceae
can serve only phenetically. Efforts to use DNA phylogenies to assess
the relationships among groups of secondary metabolites, primarily iridoids, were
reviewed recently by Grayer et al. (1999). Discussions involving fl avonoid profi les
and sequence-based phylogenies can be found in Soltis et al. (1993) concerning
Saxifragaceae s. str., where it was evident that several structural features have been
gained and/or lost a number of times, and a discussion from our group concerning
fl avonoid and sequence data as they relate to the evolutionary position of Itea
and Pterostemmon, two genera whose status within Saxifragaceae is of interest
(J. Y. Yang et al., 1998).
The fl avonoid situation in Restionaceae provides an opportunity to comment on
the importance of establishing phylogenetic relationships within a given taxon. In
addition to affording a better understanding of evolutionary relationships within the
group itself, a reliable phylogeny allows one to investigate the changes that occur
with regard to some particular feature or suite of features, the general topic of character
evolution. This applies no less to microchemicals than it does to morphological