Addendum 9. Observations on cytology as a character source in Gasteria.


The classification of Aloe, Haworthia, Gasteria, Astroloba and smaller related genera is in the public domain in the sense that there is an immense public interest in the collection and cultivation of these plants.  Unlike more general horticultural elements, succulents plants are of interest to collectors as natural “species”  and hence the taxonomy and nomenclature are central to the activity and communication which takes place in the collector citizenry.  Unfortunately the degree of interest and activity of competent (as opposed to trained) scientists (ie. botanists) has not reached the same degree and intensity.  The need for information and classification has by default passed to non-scientists/botanists.  A paper in respect of this topic is published in Asklepios (Bayer, Aug./Sept. 1999).  This particular article is written to further examine the nature of the data obtained by scientists and presented as such to the amateur community.  It is written particularly in the light of the conflict which has clouded the literature concerning Haworthia since at least 1947, and which continues unabated.

There is clearly a myth in the amateur community that scientific technology will solve all problems, and that excavating even in an amateur way at the base of the intellectual pyramid will bring understanding.  This paper will show that this will not happen, and that scientists themselves may be as largely deluded and uninformed as the amateurs they may scorn (see Smith & Mossmer, 1995, Crouch et al, 1999).  Instead of knowledge and understanding filtering down to the lower levels, it is the intellectual pyramid which will be found to be often, and largely, built on sand.

This article takes into account only a single scientific paper, Vosa and Bennet in Caryologia 43:235-247 (1990) entitled “Chromosome Studies in the Southern African Flora. 58-94. Chromosome Evolution in the Genus Gasteria Duval.”  This is a further part in the same series in which karyotypes of various ‘species’ of Haworthia were illustrated and described.  It is by no means the only publication in a reputable refereed scientific journal which can be used to exemplify the kind of science available to the amateur who must presume it to be professionally acceptable.


Vosa and Bennet present the cytological analysis of 37 specimens of Gasteria, of which 12 were unidentified.  I have taken that same data and attempted to interpret it myself.  I have done this in terms of identifications by E.J. Van Jaarsveld himself who wrote the revision of the genus.  Firstly I have tried to establish if there is any pattern at all (as Vosa and Bennet have tacitly assumed does exist), and secondly to see if the new Van Jaarsveld’s classification (identification) helps to interpret the observations, or the converse.


The results are disappointing in the extreme.  For three specimens there is no data at all other than the comment “The karyotype of this species (the word ‘taxon’ is used twice) is well within the range of the variation found in the genus”.  Most specimens are registered as “this species”.  The terminology used does not appear to be consistent and there is no inadequate explanation for it.  Thus we have ‘allocyclic segments’, ‘allocyclic constrictions’, ‘allocyclic gap’, ‘allocyclic regions’ and plain ‘segments’.

Specimen 59 (G. brachyphylla) is said to be similar to 58 (G. maculata) “except that there is a small allocyclic constriction in the short arms of chromosome L1”.  In 58 the comment is that there are “allocyclic segments” in the short arms of L2 and L3, and in the long arm of S3 and so the comparison has to assume this is also the case for 59.

Specimen 66 (G. batesiana) is said to be similar to 64 (G. carinata var. glabra “with only minor differences”.  The comparison is meaningless without some suggestion of what these difference might possibly be.

Specimen 72 (G. nitida) is said to resemble that of 68 (G. bicolor) with small allocyclic constrictions in short arms of L1 and L2.  However, 72 also has an “almost metacentric” S3 which 68 does not have.

Specimen 77 (G. bicarinata) is said “This species is characterised by distally located segments” in L2 short arm, and S3 long arm.  But specimens 77 and 88 are also this species, and the statement is obviously untrue.  Specimen 84 (G. pillansii) is similarly said to be characterised by “allocyclic constrictions” in L2 short arm and long arm S3.  This is not true for 61, 70, and 89 which are also G. pilllansii.

In specimen 79 (G. baylissiana) no allocyclic “regions’ are reported, and this is the only specimen for which such a statement is made.

In looking for pattern in the data myself, the only significant observation I can make is that 76 (G. rawlinsonii – spelled rowlinsonii) is the same as 86 (G. carinata).  This is significant because G. rawlinsonii is so notably a different order of Gasteria that the late Frank Stayner maintained that it could not be in the same genus.  Specimen 75 (G. pulchra) is the same as 91 (G. croucheri) except that the feature in the L1 short arm is called an “allocyclic ‘gap'”, and in 91 all the features are called “allocylic regions”.  This would make 75 the only specimen with a “gap” and 91 the only specimen with “regions”.  The other specimen of G. croucheri which was examined (94) is reported as “characteristically, all chromosomes in this taxon appear to have allocyclic segments”.

Photomicrographs of four chromosome complements are included in the paper.  Six arrows are used to draw attention to particular features.  As is often the case in cytological publications, one has to assume that the authors can see somethings there in fuzzy photographs which are not visible to the untrained eye.  Alternatively, visible in the original preparation but no longer visible in the published photograph.


There is no statement by Vosa and Bennet on how their specimens were identified nor by whom.  They conclude that.. “The slight variability (observed) is probably not more that that found between the individuals of a normal natural population of a single species”.  Thus as far as their cytological observations can take them, there may be only one Gasteria species.  The three references to characterisation of taxa by the observed karyotype is thus also fallacious.

I conclude that the paper has virtually no credibility at all.  The data is so presented that it is not possible to make any real comparisons or draw any conclusions.  The nomenclature is inconsistent and the data is also not presented in a consistent manner.   The specimens are treated as taxa and there is repeated reference to the ‘specimens’ as ‘species’.  The specimens are not fully representative of the genus as currently classified by van Jaarsveld.  Thus it is not possible to use the data to meaningfully examine that revsion.  I find no pattern to suggest that it might, even where the species G. bicolor and G. carinata, represented in the analysis by 9 and 7 specimens respectively, are concerned.  In the paper, the authors write “Gasteria has been somehat neglected by taxonomists, and indeed it is a difficult genus to classify owing to the difficulty of obtaining satisfactory herbarium specimens.  Any meaningful classification .. must rely on living collections.”  This is patently not true.  Gasteria has not been any more neglected than any other genus.  It is regarded (as is the Liliaceae generally) as a difficult genus to classify because of the absence of synoptic and distinguishing characters by which “species” (as indeterminate as this concept is in taxonomic botany)  can be distinguished and circumscribed.  It is thus also “difficult” because of that absence of a species definition and concept by which any classification of the species could have been derived as a general truth for botanical classification.  Finally there is the significant fact that there is an amateur community demanding a classification.  The conclusion one has to consider is that the scientific community has not been able to present a workable classification where there is critical and perhaps sceptical audience.  Prof. E.A. Schelpe did attempt to do so, and could not.  This conclusion is as real as the scenario of a good classification for which the audience does not have the critical or sceptical faculties or skills to recognise.  What would botanical classification look like under a general spotlight of informed and critical community interest?

As we move rapidly to a point where cladistics is, and DNA sequencing is becoming, a sacred altar of classification, the amateur community had better be warned.  Schumacher (1980) cites Burr (1972) to say “dethronement of chemistry… with all its DNA-mythology of molecules becoming information systems, is certainly a big step in the right direction.”  The statement is even more true for cytology where at one point in time it seemed that, as in the Vosa/Bennet paper, chromosome morphology under oil-immersion microscopy would reveal what could not be seen in the whole plants themselves by naked eye.  The less that is said about cladistics which imposes linearity and dichotomous order on what may be an amorphous and reticulate (even chaotic) system, the better.

It is evident in a spate of books written on the subject, or even peripherally, that there is a need for a new paradigm in science.  The mystique of technology is as frightening as religious dogma, blind faith and disorganised belief.