Volume 3, Chapter 15:- The superspecies proposal.

Dr Canio Vosa in Caryologia 57:4, 395-399 (2004) hypothesizes two superspecies for the subgenus Haworthia.  He cites my published works including Haworthia Update Vol 1 (2002).  It is not clear if he has considered all the species that I recognize, or considered that if there are superspecies there may ipso facto be ‘inferiorspecies” too.  It is also not entirely clear that he as followed my argumentation.  He writes…” the morphological characters, as used for species definition of the taxa in question, do not give a clear indication of true discontinuity over their geographical range which in some case(s) is rather restricted”.   As early as 1972 I pointed out that there were fewer species rather than more, and most of my works have featured some kind of statement on the problem of identifying any discontinuities, other than geographic that suggest a realistic concept of species as systems.

A problem I have also had to face is that there is a very strong user group for names in Haworthia in terms of collectors and growers and it is extremely difficult to balance the different needs and perceptions of these non-professional but knowledgeable people against the needs and perceptions of professional botanist and other scientists that have often proved to be highly flawed.  Also to be taken into account is the ground truth and what I would regard as objective reality.  This entire Vol.2 is dedicated to extending this ground truth to a point that I can claim that there is enough now known to make a truthful statement.  This is the point of Chapter 24 of this Volume where I provide what I think constitutes a rational list of names under which all that I have seen can be catalogued with some degree of confidence.  It is from that point that I believe scientists, like Dr Vosa, should hypothesise and test.

In this last chapter, I myself am going to propound a partial superspecies concept.  On request from The South African Institute for Biodiversity, I have already generated a species list for a proposed flora for the Eastern Cape.  That list excludes other geographic areas and here I limit my discussion to principally those species concerned in the chapters of this book and the southwester Cape.

Where Dr Vosa asserts that…”the subgenus Haworthia constitutes a single polymorphic assemblage of species which can be subdivided, albeit with some difficulty in some cases, into two groups”, he is not making a ground-breaking statement and neither is it accurate.  In all three subgenera it is often quite difficult to allot specimens to groups that would then constitute species.  At least one of the chapters explains that this problem extends to the genera of the Asphodelaceae.  The problem in the subgenus Haworthia is simply more acute and there are more than one polymorphic assemblages.

Dr Vosa makes further assertions about the role of specimens, types and herbaria.  These have to be examined because there are profound misconceptions which fuel the opinions of users in any of the user groups.   The role of the type is not intended to designate a species.  It is solely to substantiate a name and link that name to a tangible reference point that may be a herbarium specimen or an illustration.  Choices in this regard are the functions of taxonomists and nomenclaturists which are governed by a very comprehensive International Code for Botanical Nomenclature (ICBN).  A revisioner would look at all those types in the process of reviewing all the literature and additional specimens and then generate a synonymy to put together any names which he thinks can be attributed to the same species set.  This is all confirmed in the list of specimens, and any published or formally deposited illustrations if necessary, which are then cited to formally convey the total concept of what that revisioner thinks constitutes the “species’.  Unfortunately the necessities of nomenclature do not bring the stability and understanding of names to the point where confusion and misunderstanding is totally excluded.  Examining all the taxonomic literature may even suggest that there is a preoccupation with nomenclature which displaces the purpose and function of classification.  This is only partly relevant to the present discussion in view of the choice by Dr Vosa of H. cymbiformis as the typical species for one of the superspecies.

Why I say this is because in my presentation for a Flora of the Eastern Cape, I have seen the necessity for a superspecies concept and my decision has been that H. cooperi is the superspecies and H. cymbiformis an independent species.  There is a problem created by the priority requirements of the ICBN.  H. cooperi can in the field, seen to be the dominant and most widely spread element whereas H. cymbiformis comprises a smaller part of this complex.  Were they amalgamated in a superspecies, the epithet cymbiformis would have priority and implementing this would distort the perceptions of the relationship.

I fully agree with Dr Vosa that species such as H. bolusii and H. decipiens could be included in one superspecies with H. cooperi  but I do feel that too much usefulness will be lost if this is implemented..  While there is no doubt that very strong cymbiformis look-alikes emerge from ecotypification of H. cooperi, there is a concomitant set of populations at a higher level which have an independent distribution.  The evidence for converse ecotypification is less persuasive.  Certainly to document this kind of evidence by way of herbarium record is totally impractical, as Dr Vosa points out.  The object of this book and of its predecessor was primarily to overcome this obstacle by providing an extensive pictorial record.

Regarding the two superspecies conjectured by Dr Vosa.  Here I have to flatly disagree on the candidature of the two lists.  Quite plainly H. reticulata is bound to the excluded H. herbacea and cannot be included in the superspecies H. cymbiformis.  I started my intensive field career trying to understand the relationship of the former two species and how they related in terms of superspecies thinking to their geographic counterparts.  My observation was the curious one that they were geographically contiguous or complementary.  At several places they grew virtually together and hybridized whereas certain populations could not be positively referred to either of the species.  Wider experience suggests that this kind of situation is deeply imbedded in classification and identification of many plants.  I have not considered the problems this causes as a sound rationalization for lumping.  I would rather suggest a growth process or learning curve that user groups and individuals truly examine their own definitions and understanding of species and classification.  This will require a paradigm shift in how name are used without any tinkering with the system in place.

Dr Vosa presents six colour illustrations which, in the absence of explanation, are presumed to substantiate the claim that they are indistinguishable as discrete species.  It is difficult to refute such a claim against the concomitant one that his years of field experience and careful observation support this claim.  Certainly to the uninitiated this could be true whereas any true aficionado would recognise the untruth of the statement.  H. chloracantha also cannot possibly be included in this same superspecies H. cymbiformis.  It is confounded with H. floribunda and this in turn is central to the recognition of H. retusa as a superspecies.  Furthermore the connection has to be followed through geographic distribution and the historic confounding of the identities of H. chloracantha with H. angustifolia and I turn the relationship of that species with H. cymbiformis.  Recognising superspecies will simply move the problem of continuity to another area or level of classification.

I would again regard Dr Vosa’s superspecies H. retusa as a rational option but it very largely skirts the problems contained in and surrounding the suggestion.   H. bayeri and H. springbokvlakensis do not enter into the picture.  They are “inferior” species in that they do not have problematic continuities.  My own independent view is expressed in a letter that I wrote to Steven Hammer…”Your comment ‘illusion of understanding’ also resonates in a comment in the literature where the author, D.A.Levin, quotes from Raven, Berlin and Breedlove’s, “The origin, expansion and demise of plant species” (Oxford University Press, ca 2000)…“our system of names appears to achieve a reality which it does not possess”.  Levin himself writes…“our taxonomic system communicates little about the organism being considered, although it appears to communicate a great deal”.  This is the conundrum that I get quite lost in.  Species have a reality and a meaning to me quite different to what I sense in people I have talked to over such a long period..

Re pygmaea – I need to find a superspecies solution to the SW Cape …”

I presented an outline of a proposal which I now modify as follows:-

H. retusa – to include mutica, turgida.

H. pygmaea – to include dekenahii, argenteomaculosa, splendens, fusca, esterhuyzenii, acuminate and vincentii (it should be noted that these names refer to elements which have a fairly restricted distribution and are highly sought after collector items.  I do not personally consider them species although I do not doubt the user pressure for a Latin binomial.  For those who have followed my classification it is important to note that I have previously allied them with H. magnifica whereas I am now suggesting that they have their closer origin in H. turgida).

H. mirabilis – to include maraisii, magnifica, emelyae and heidelbergensis.

H. floribunda – to include parksiana (?), chloracantha, floribunda and variegata.

There are other elements viz:-

H. herbacea – to include or cover reticulata, maculata and pubescens

H. rossouwii

The essential problem is that we have three species in the west viz. mutica, turgida and mirabilis, while in the east we have two viz. turgida, and pygmaea.  In the intervening area H. turgida ad H. retusa are continuous with each other.  The retusoid element expresses itself in the west as H. mutica whereas in the east it seems to be transferred to the mirabiloid set suggested above while H. turgida remains independent.

There is a problem in this summation in the very complex relationships that also exist between retusoid, and mirabiloid elements elsewhere.  These extend to floribundoid and rossouwoidH. turgida crosses the distribution area of H. mutica to befuddle the proposal that they are elements of a clearly definable H. retusa superspecies.  Similarly there is a suggestion that H. mutica has some degree of continuity with H. mirabilis sensu stricto.  There is a huge problem to segregate H. floribunda which has close ties to the mirabiloids as well as to H. chloracantha, H. variegata, and thus also H. monticola and perhaps eventually H. zantneriana and H. angustifolia.

I do have preliminary ideas for extending the concept of superspecies to the entire subgenus Haworthia but am reluctant to do so while I am aware of the deficiency in terms of field work.  More fieldwork is required to examine known problems in the area between Uniondale and Oudtshoorn and also in the mountains south of Prince Albert.  These problems suggest that similar situations to those discussed particularly in other chapters, exist in many more areas.  Comprehensive as the field work has been, it still does not provide clearer insight.  Nevertheless the system of names I now suggest in Chapter 17 does provide a framework of communication that will work if users properly recognized the definitions and premises on which it is based and the reservations that accompany it.

One of the most significant and important user groups is conservation bodies and environmental impact assessors.  They require definitive names and lists.  No shuffling and reshuffling of names can produce such a list.  The reality is that species are intrinsically highly variable and this variability simply has not been exposed as it has been in Haworthia.  Hence conservationists have to realize that all pristine and semi-pristine vegetation is valuable and that conservation is not about species, but about habitats with attributes that promote diversity and minority species.

The recognition of supergroups could mean that user groups would have to deal with the problem of using Latin trinomials and higher.  This is where the paradigm shift will be the most acute and yet it would be simply solved in common usage by the substitution of the superspecies name where the genus name is presently cited.  My conclusion is that the recognition of superspecies is an essential process in synthesising and extending our understanding of the genus but that it would have no merit in terms of the communication value of Latin binomials. Far more would be gained from a proper analysis of the many sources of misunderstanding of taxonomy and an educational process instituted where users and practitioners acquire better insights into biological diversity and how it is recorded.

My closing statement is that scientists seeking a better understanding and classification of Haworthia would do well to try and properly follow what I have written.  I am confident that this is the text on which proper and valid hypotheses can be formulated and then tested.  I would further add that my field work and observations have been so extensive and so thorough that they could and should be used to test the validity of sophisticated technological inquiry.


The above manuscript was submitted to Dr Vosa for comment and I deeply appreciate and value his response printed below.  Dr Vosa, as a pre-immanent geneticist, cannot possibly be expected to be aware of the confusion that exists in the minds of non-scientific user groups and I consider that he has, as he set out to do, indeed made a major contribution to demonstrating that the taxonomic emperor is unclothed…

“Dear Bruce,

Thank you very much for your e-mail message and welcome criticism of my Haworthia paper and also for the rather important attachment (Chapter 33). The purpose of my initial effort has been to make clear the problems facing who is trying to classify Haworthia in a plain understandable and possibly correct way.  My paper was also conceived as a kind of provocation aimed chiefly to the “classical” conservative taxonomists.  I have never taken into account the opinion of the many “user groups” interested as they seem in “splitting” and in maintaining as many species and varieties as possible for sheer lust of collecting or,, worst, “economic” reasons.  I agree very much with all you say about my very faulty and, of course, incomplete “groupings” but I considered them in the light of my research experience relative as it is but in all good faith.

I consider your Chapter 33 as very well written and argued and as an important and basic contribute to that knowledge we all earnestly seek.  The Chapter itself will be for me as a sort of main guideline.  Above all, I think very appropriate and wise your statement “that there is a preoccupation with nomenclature which displaces the purpose and function of classification”.

I came to choose H. cymbiformis as the typical species for one of my two “superspecies” in the light of my field experience in the Eastern Cape.  However, taking into account the evidence, I think that your decision to consider H. cooperias a superspecies as perfectly valid.

Bruce, I certainly understand fully you disagreement on the ‘candidatures’ as suggested in my two lists which were compiled and meditated using my, alas, fragmentary notes and material in hand.  In any case, the reasoned purpose of my paper was to put forward the concept of superspecies as contributing to progress and as a discussion point for constructive criticism and useful argumentations and suggestions as yours.

Thanking you again very much,

With regards, Canio”