Hunt Institute Archives Text Discovery Platform

Page transcription

Critique of "A Numerical Taxonomic Study of
The Mexican Species of Solanum, section Tuberarium"

1. Has the material been published previously in the same or similar form?
Studies on Solanum, using some sort of statistical technique have been
published before, (Heiser, et. al., 1965) and, there have been many "objective"
comparisons on other groups made between the statistical methods and some
"classical" classification, and most are invalid, for the following reasons.
(1) The statistical methods are based upon some assumptions which may or may not be
valid, and these assumptions are quite different from the assumptions expressed
or inherent in the classical methods. (2) The authors do not state the bases for
accepting one or the other statistical package, and leave the reader to guess why
they chose the ones they did. (3) There is no effort to explain (if they indeed can)
whether the mathematical bases underlying the statistical assumptions are satisf-
actory for the purposes of taxonomy.

2. Has the research been carried far enough to warrant publication?
Clearly, it has not. There are no clear-cut decisions as to the arrangement,
or re-arrangement of the taxa included in the section, and this is the only basis
for publication of some work in BRITTONIA. We are left at the end with no decision
which method of numerical taxonomy is valid, if any, nor why there are differences
in the output. If the information of the species taken from cultivated specimens
differs from the information on the wild-grown plants, we clearly deserve to have
these differences reconciled. Why, one asks, were not both sets of data run
together?

The authors do not seem aware that, having used nearly 50% of the information
on leaf characters, they have heavily weighted the classifications toward the
vegetative information about the plants. This, in itself, could help to explain why
their various results do not agree with the earlier classifications where I am
sure the authors considered reproductive or chromosomal differences uppermost.
Furthermore, in nearly every case of the measurement characters, they have duplicated
the character in such a way that they have introduced false biological weighting.
It is not conceivable that there are that many independent genetic systems at work
in determining leaf dimensions, and if our classifications, numerical or otherwise,
do not make some effort to reflect genetic variation in the plants under study,
we have no reason for calling ourselves taxonomists. However, if the authors had
clearly stated at the outset that their objectives were to build a classification
predominantly on information from the leaves, then they may proceed, and all will
understand them.

3. Is BRITTONIA the most suitable journal for the publication of this
material?
No, nor any other, considering the present state of the work.