My experience with Haworthia dates back to my childhood and on to nearly 70 years of observation. However, my interest was only able to properly manifest when I began work at the Karoo Botanic Garden in 1969 and it has since been through many phases. I wrote a formal taxonomic revision of the genus in 1999 and have spent a good bit of the last nine years adding to and verifying what I wrote. Haworthia has always been regarded as a problem child of botany to be avoided by professional taxonomists for various reasons including an apparent phobia of the many amateur collectors peering over the shoulder while at work. This has puzzled me because it seemed to me that if the need for good classification and identification was so strong there was an obligation on botany to provide the service. So my involvement has been largely by default. I was trained in an agricultural and entomological tradition with a totally different and unsophisticated approach to things like taxonomy, systematics and nomenclature. In the infant science that agriculture then was in South Africa, I can barely claim that my MSc is much more than an indication that I tried to learn something beyond normal schooling. While trained as an agricultural entomologist, my leaning was to plants and I eventually came to the Karoo garden to do what I liked best viz. exploring plants. Unfortunately the route is via identification and names and so I have walked a long road through the minefield that this is. Was this only in respect of Haworthia? No! This is a persistent misconception. Haworthia is only different because it has attracted such close and sustained amateur interest by so many for so long. I experienced failing classification in many other genera. To be fair I think the real reason is the lack of importance attached to the whole function of plant classification. It even seems as if many modern botanists pursue the study of plant relationship under the guise of systematics that is not committed to providing formal names and identifications.

The problem in the genus Haworthia begins with weak botany from the moment Linnaeus decided that there were four small plants that belonged together as one species in the genus Aloe. This alone generated a nomenclatural problem the dust of which is still to settle. A worse and obscured fact is that because of their small flowers these plants were and are still treated as one genus – one set. This is not true. The allied genera of the Alooideae of the Asphodelaceae are not comfortably classified and the three sets of plants that constitute the genus Haworthia are not only florally distinctive, they are also phytogeographically and behaviorally different. Thus while there was or is this huge flaw at this classification level, it is actually bizarre that sense can ever be thought to have reigned at the species level. The proliferation and confusion of names coupled with the associated argumentation about names has even led me to doubt my sanity, or even lose it as my critics will happily claim. My impression now as a non-botanist is that taxonomic botany somehow lost sight of the requirement that a biological classification should mirror and reflect the evolutionary processes or phylogeny that underlies our view of species and their origins. Worse, is the actual absence of a clear and unsullied definition of what a species is and hence what a Latin binomial as a supposedly formal scientific term of reference actually means. Cladistic methodology in plant taxonomy is in my opinion nothing more than a belated and even misguided attempt to arrive at or return to some semblance of scientific truth. Molecular biology is now being desperately seen as a holy grail that will summarily correct reluctantly conceded failures of the past.

There is no doubt that the mess in the classification of Haworthia is the product of non-botanists, but it is absolutely incorrect to exculpate the contributory role of taxonomic botany and professional botanists who are or were taxonomists. My interest has by no means been confined to Haworthia and the need for names has largely been driven by a perpetual curiosity about many different plants in the field environment. I am, by the route my life has taken me, unusually and almost uniquely experienced in the relation between botany, botanists and interest of the layman in plants, as it is reflected in the names we use for them and how we respectively do so. Consequently I think that I am now qualified to express some opinions on what Haworthia can teach us quite apart from what the hostilities that wage across the whole taxonomic front suggest.

I pointed out rather timorously that in Oxalis, morphological characters were not as indicative of species as revisions and identification keys generally seem to imply. The dramatic changes in our classification systems that the results of molecular biology now seem to threaten, is a manifestation of this simple fact rather than any great new understanding of what really goes on in the plant kingdom. My fear is that molecular biology is likely to fudge most of its promise because the basic tenets of classification, the mistakes of the past and the reality of the field are still not properly appreciated. This is what I think Haworthia has to tell us. I made use of Oxalis to test my observations that we can look at Haworthia as composed of elements that reflect the very nature of the habitats and environments where they are found. I did this because Haworthia plants simply do not have any array of characters by which species can be arrived at whereas Oxalis has a plethora of such apparent (taxonomically bewitching, tempting, and misleading) means. While the few facts that there are in respect of differentiating Haworthia into three groups (subgenera as presently constituted) has been lost and confounded in the broader classification of the Alooideae there is not much else to help verbally to define the species. It is not to say that modern technology has not been resorted to, to help clarify the issue. Both cladistic and molecular studies have been done but the results flawed by weak hypothesis and perhaps just weak science too.

How then do I think sense can be made of Haworthia?  It is firstly by defining the species as dynamic fractal systems that illustrate the responses of plants to skeletal habitats and geological substrate in the wider context of landscape and topography across the total distribution area. Furthermore, there is an obvious correlation to be made with vegetation and other biological systems both plant and animal. I cannot pretend to know a great deal about any of these things, but also suggest that nothing more is needed other than awareness that this is what biological diversity and species are essentially about anyway.

It has taken me nearly 40 years to arrive at the big picture of Haworthia. It is as a non-taxonomist that I say categorically that the three main elements in the present genus need to be seen and registered as “different”. Then we can focus on the subgenus Haworthia where the greater difficulty lies and which I believe offers the greatest insight into what plant species are and what the difficulties are in recognizing and defining them. Also into what Latin binomials should or might mean.

For this I am going to take just two species systems and ignore some of the peripheral facts that might confuse the issue rather than illustrate the profound nature of species interactions. I take H. retusa (Figures 1 to 4) and H. mirabilis (Figures 5 to 10). In order to do this I have to bypass even my own conservative classification by considering H. turgida and its many variants to be the riverine and cliff dwelling version of H. retusa. H. mirabilis is more complex and I have to include three other species viz H. maraisii, H. magnifica and H. heidelbergensis for which spatial factors and major geological substrate are the prime differentiators. Both my two selected species are highly variable both within and between populations. It is in fact the intra-population variability that I underestimated. Growing plants from field collected seed can result in a hundred or more individual plants that are all different in appearance.

The two species share the same distribution range, but with an extension of H. mirabilis into the Worcester/Robertson Karoo, while H. retusa extends further east from Albertinia to Great Brak. H. mirabilis flowers in late summer and H. retusa flowers in spring. They never grow in direct association! The populations of each are highly localized and when in proximity are usually separated by hundreds of meters, and difficult to ascribe to habitat conditions given the ranges of habitat that respective populations may occupy. I have already noted and warned of the reality that there is evidence of interaction between these two clearly defined systems. In addition, somehow or other, the species H. pygmaea emerges as discrete in the eastern Mossel bay area, while H. mutica sits in an island between Swellendam and Riviersonderend in the west.

What is most striking is the variability within each system and this is where the crunch comes and what my pictures should illustrate. They are but a small sample of many hundreds and I have adopted an informal method of captioning the pictures as the formalities of nomenclature would need to be followed to depart as I have from the existing status. Not only are my pictures supposed to illustrate the problem of variability within species, figures 2 and 3 show variability within a population and figures 3 and 10 show two very similar plants from two quite different species systems. Note that the plant in figure 3 is in flower and that the similarity to figure 10 is rather exaggerated by photograph scale as in H. mirabilis ‘toonensis’ the plants are smaller.

Taken individually from the morphological extremes between and even often within populations, it will stretch the credibility of the most ardent taxonomic lumper to claim that any two of these are representatives of the same species. It is an unpalatable truth that things that look different are the same just as the converse is that things that look the same are different. When this reality is properly conceded by the professional taxonomist, perhaps we will succeed in demystifying and deglamorising the Latin binomial. Botanists who are not taxonomists will arrive at a better understanding of their plant world and consequently we who look to them for understanding will be better served. We need to be more realistic about plant description and identification and also correspondingly more conscious of the real nature of diversity. Conservation is not to be seen as maintaining a list of names for which there is a use. It is to be seen as maintaining the options for change and the conservation of a far greater diversity than any list of names can hold. It is botanists (quoted in Levin, 2000) who wrote “our system of names appears to achieve a reality which it does not in fact actually possess”. I think this implies that even the Latin binomials we so modestly, reluctantly or even portentously use do not convey what they should; but rather the product of our individual limited experience, knowledge and perception. My feeling is that species do have a reality and for Latin names to have any sense, they need to be more circumspectly applied. There is little merit in having a highly regulatory nomenclatural code and little value in what the names actually refer to. ♦