Saintpaulia taxonomy

Timothy S. Ross rosst at CGS.EDU
Thu Jan 4 21:54:13 CST 1996

        In the face of the subtle suites of characters that currently
define many of the Saintpaulia species, I can understand the apprehension
felt at how to treat the taxa as natural units.  A certain amount of this
apprehension may come from the diversity of "species concepts" that are
floating around out there.  Before I got into field botany, I was exposed
to the simplistic species concepts proffered in high school level textbooks
that are great for memorizing and regurgitating on the final test, but that
are often highly irrelevant to organismal units in the real world.  One of
the cornerstones of these simplistic definitions was that "if two organisms
can hybridize and produce fertile offspring then they are one species,"
pure and simple.
        My own species concepts have evolved considerably since I moved to
California and began doing field botany (I say "concepts" because I do not
have only one that is broadly applicable).  Since there are different
evolutionary and isolating mechanisms at work in different taxa, a
simplistic definition of "what constitutes a species" has yet to be
reached.  A room full of taxonomists can get into a bloody brawl on this
topic.  Each taxon needs to be assessed in a context that is relevant to
its NATURAL distribution patterns and NATURAL isolating mechanisms.  If, in
Saintpaulia, the geographically isolated taxa can be defined on the basis
of suites of morphological characters and there is no natural hybridization
encountered "in the wild," then that should be taken into account.  Homo
sapiens (ssp. intoxicans, bibiens, belligerens, or whatever), the ultimate
transporter and tinkerer, has managed to bring together all kinds of
organisms out of their natural settings.  The existence of ligers in a zoo
(= lion x tiger) would not negate the validity of treating the parental
taxa as "good" species since the likelihood of them doing hanky-panky in
the wild is virtually nil.
        Originally, African violet cultivars were derived from hybridizing
what were considered differing clones of S. ionantha (the first species
described), but with the popularity of the plants and the need to try and
surpass a million named selections by the year 2000, people began using
some of the other species that were being discovered and brought into
cultivation (S. goetzeana, S. magungensis, S. orbicularis, S. shumensis, S.
pendula, S. rupicola, etc.); hence, we now have our miniature saintpaulias,
our trailing saintpaulias, and so forth.  Despite their ability to
hybridize in cultivation, one does not typically encounter hybrid swarms of
African violets in nature, and, in producing a meritorious taxonomic
revision of the genus, an understanding of the NATURAL distribution
patterns and concomitant variation will be critical.
        Many (perhaps most?) of the Saintpaulia species that have been
discovered in the last 40 years or so were formally described by B.L.Burtt,
who has spent a lifetime studying palaeotropical Gesneriaceae.  Burtt is an
outstanding taxonomist who probably had good cause for treating his taxa at
specific rank.  This is not to say that the genus does not need revision.
However, I think that extensive study of the taxa in situ would be
essential prior to proposing taxonomic innovations.
        With that as my semi-solicited caveat, I would offer a wish of
"Good luck" on the study.  [My three cents worth.]


>   I am doing so taxonomic work with African violets (Saintpaulia) and
>need some advice on what others have done with plant groups that show
>similar taxonomic problems.
>African violets belong to a genus with a relative small geographical
>range (parts of Tanzania and Kenya).  Although 20 species have been
>described, it is now becoming apparent that most of these "species"
>were described from small populations of individuals.  When brought
>into cultivation, almost every species will cross easily with another
>and give fertile offspring.  There does not appear to be any genetic
>barriers between the species.
>Question - in other plant groups, what guidelines have been used to
>define a "species" in situations like this?  Should a species be one
>of the local populations that can be identified by morphological
>characteristics, or should the ease with which the plants hybridize
>be taken into account?
>The pollinators appear to be thrips (very small flying insects) so
>crossing between geographical populations in the wild is doubtful.
>I'd appreciate any advice and throughts on how to define the species
>under these circumstances.
>Jeff Smith
>06jlsmith at

Sr. Curatorial Asst.
RSA-POM Herbarium
Rancho Santa Ana Botanic Garden
1500 North College Avenue
Claremont, CA 91711, U.S.A.
(909) 625-8767 ext. 233
FAX (909) 626-7670
rosst at

"At the end of a fortnight, I fired myself for willful incompetence."
              -- Donald Culross Peattie (The Road of a Naturalist, 1941)

More information about the Taxacom mailing list