More light on the Haworthia pygmaea/retusa/turgida species mystery.

While editing some old pictures I was just stunned that I have so understated the case that retusa, mirabilis, pygmaea, mutica, turgida and emelyae are really one species. The photographic evidence of the plants alone is unbelievable. What I do think has gone under the radar is evidence that I got from the flowers where the flowers of mirabilis, mutica, floribunda are shown to be inseparable.

Warwick Bayer: Do the flowers as a characteristic often fly below the radar as one may struggle to see flowers from different populations together? And one may so often be in the field and not see flowers at all.

Bruce Bayer: Yes, they do mostly because they are seasonal and often these places are not so accessible that it is practical to revisit. The other problem WAS of recording anything the so-similar flowers might be able to say. Dried they are useless. What I ended up doing was digital images of floret profile, side view and face. Very illuminating :)

2019.5.3 – Between the Gouritz Bridge and Herbertsdale. I do not anymore think it is a mystery in itself. The mystery to me is that a rational explanation is neither wanted nor sought.

Here is a map to give a general idea of the locations.
1. Greyton/Genadenal area
2. Robertson Karoo
3. Bredasdorp, Napier
4. Breede River with Infanta, Malgas, Potberg at sea end
5. Tradouw with Buffeljags between this and 4
6. Slang Rivier to SE Heidelberg
7. Duiwenhoks River from Hedelberg to sea
8. Goukou River Riversdale to sea, Kruis and Wegwysers rivers to NE Riversdale
9. Gouritz River with Albertinia to east and Wydersriver to north Albertinia
10. Mossel Bay, Great Brak, Herbertsdale area
11. Montagu area N. Langeberg Range
12. N Garcia Pass in Little Karoo
12. Aasvoelkrans in Little Karoo

2019.5.6 – I thought I would post this to show how confusing classification, plant names and classification can become. I found a folder with these pictures – turgida, pygmaea, retusa, mutica, mirabilis, floribunda, and emelyae. Ridiculous as it seems, they can be linked by a geographically arranged sets of variants that morph them into a single set. In the process my mind just boggled at collectors who just cannot seem to grasp the problem or a solution. Around Mossel Bay we have two species H. turgida and H. pygmaea but in the west around Greyton to Napier, we have three H. retusa, H, mutica and H. mirabilis. In between there are a host of combinations while a fourth species intervenes viz H. floribunda. Difficult enough, but across the Langeberg there is a fifth, H. emelyae. All one? H. floribunda has its own complications. No single picture is remotely adequate to illustrate any one of them. It is this kind of complexity that pervades the plant world and the botanical names we need to communicate about them.

Steven Molteno: I’ve recently got very interested in the relations between floribunda and mirabilis, because of this gradient being so often syntopic with T. marginata. The N.Bredasdorp-Rooivlei Haworthia interest me in particular. Some really remind me of floribunda. More and more I feel the need to draw out some multidimensional lattice in order to try model or understand these relations. A bit like the map you once drew of them Bruce.

Bruce Bayer: Yes Steven – I think this is the real challenge and I find it difficult to believe that there are not botanists who are working at this – confounded by the majority who are stuck in a dead-end. I think I arrived at a classification solution by visualising such a lattice as a reality. What should be done now is to take the end classification and generate layers according to it. This would be a great test of the classification.

 Here are more giant retusas from the same locality – what is true is that populations have characteristics as much as the “species” do. The haworthia species mystery continues …

 2019.5.8 – I only have two pictures of cultivated variants for a place north-east of Albertinia near the Gouritz River = “turgida”. They add to the next three that are the pictures sent to me from a locality a few miles further east right on the Gouritz. These pictures support this fact. If you interpret them in the light of populations to the east you might say the third and fifth pictures are H. pygmaea and the 4th “turgida”. But if you do so from the west you arrive at H. mirabilis and H.”turgida”. In fact, they are not quite either. Why? They come from a common gene pool that is one species! Proliferating all these personal and acquaintance names does nothing but bury the issue under a pile of garbage.

1. 2019.5.7 – I am going to post a heap of habitat pictures of H. retusa/turgida to help enthusiasts understand these plants and their relationships. As with H. cymbiformis and H. cooperi, there is a direct correlation of clump-forming cliff dwellers (turgia) and more solitary ground hugging plains dwellers (retusa). It is quite important to recognise the geographic relationships as well but that is rather difficult to do without maps. The immediate geological formation e.g. less erodable sandstones vs more erodable shales, is also significant. This is H. retusa at the midwestern limits of its distribution on the Breede River near Swellendam.

  What I find extraordinary is the way in which amateur succulent journals resist any effort to rationalise names in accordance with some sort of scientific intent. The Latin binomial seems to be all significant and it has made life in Haworthia absurd. All it needs is a very simple adaptation and the adoption of a convention where the species name is taken as generic and is so understood in the community?

Steven Molteno : This post made me look at my maps again. I see the west-east gradient between mutica-nigra-retusa … I think I also see the habitat-gradation from retusa to turgida in the Eastern Overberg & Riversdale area. On the map it looks like turgida then stretches out in all directions (often along the river-valleys). I get this initial impression because in its westwards extension (i.e. back towards Swellendam), I do not yet see any gradient between turgida and mutica. Would it be correct then to assume that this is because they’re at different “ends” of this entity’s spectrum? (I’m not factoring in the other arms of this entity yet – i.e. mirabilis, pygmaea etc., or my brain will explode. I find it easiest to just examine portions of the Haworthia lattice, one by one, and maybe that’s a human tendency that leads us to the illusion of discrete binary units in nature.)

2. 2019.5.8 – Heidelberg is an extraordinary place for Haworthia and H. retusa has many guises particularly towards its turgida variant. In this population the plants are large and there is a mirabiloid connection as well.

Steven Molteno: I’ve wondered about the mirabilis at Die Plotte west of Heidelberg. Many I thought had features I previously associated with retusa-turgida

Bruce Bayer: Steven – this was the near origin of “heidelbergensis” as very small mirabilis. But looking at only one population does not help much. There are several river systems in the southern Cape and these are associated with the drivers of species variation. At Heidelberg we have the Klipriver and the Duiwenhoks going down to the sea. So, from the sandstone Langeberg to the limestones at Vermaaklikheid we have all the ingredients for difference in the Haworthias. The entire mirabilis/retusa lineages seem to have their origins down those valleys.

 3. 2019.5.9 – H retusa/turgida at Draaihoek NNW Albertinia – a classic as H. dekenahii also originated here but it is of more solitary and flatter to the ground (it fits better into the pygmaea element than into retusa). The 5th picture is significant because it is very mirabiloid and could easily be passed off among southern populations of H. mirabilis.

 4. These are unfortunately the only pictures I have for “H. dekenahii” (=H. pygmaea) Draaihoek. The locality has been severely impacted by bush encroachment associated with fire management along a power servitude. I wonder if there are still plants there?

  Delightful, as I find myself confounded by the very trap I am explaining away. Draaihoek is a principal element in the illusion that H. pygmaea and H. turgida are two different things. When the picture is complete it is very obvious that the two things there are the plains and the cliff forms of the same thing (H. retusa). It later becomes clear that we have a thing south of a mountain range and one north of a mountain range (H. emelyae) where space is the only difference?

5. 2019.5.10 – These posts are not in any sequence and are just taken fairly randomly. So just let us go back to Heidleberg and the Duiwenhoks river that runs to the sea. River systems was an early insight into how this puzzle arises. Tierhoek is a dry area to the SE Heidleberg. It seems to have been unknown or there would probably be a name for the plants!

6. 2019.5.11 – In one of my very early articles I wrote about the continuity of H. turgida from about 16 miles north of Heidelberg down the west bank of the Duiwenhoks river to the town itself. This proved to me that the small spiny green forms in the Sandstones were continuous with the bigger, smoother, yellow, pink and red-coloured plants southward and again to the east all up the Goukou river at Riversdale. The foundation was already laid for the understanding that turgida led on to retusa.

This is an odd population from SE Heidelberg along the Duiwenhoks that in a way replicates the Sandstone forms of turgida (= caespitosa?).

2019.5.12 – As an aside. While looking at these greener things, it occurred to me to make this connection. It brings multiple things to my mind that are fairly new and additive to my story. Is there a suggestion of continuity? What do you think?

Lawrence Loucka: Oh look, 3 different species all in one spot!

Bruce Bayer: Too true! Disastrous! It makes me wonder if the story is worth telling. This is actually “H. paradoxa” in limestone at the mouth of the Duiwenhoks. Taking all the variants, I have no doubt that I could find images from the whole periphery of the distribution of emelyae, mirabilis, retusa, mutica, pygmaea to indisputably say this must be ONE. Strange how unavoidable it is that I use names to explain that only one is actually necessary!

2019.5.13 – The first two pictures are from Kransriviermond where this “turgida” hybridises (sic!) with “mirabilis” to generate “mutica var nigra” – a lineage that can be traced back up the Duiwenhoks, through Heidelberg, along the Klipriver, Van Reenens Crest, Tradouw Pass, Buffeljagts Rivier to “mutica” in the west. The second two are “turgida” again, from a bit south at Dassieklip. That lineage crosses over to the Goukou River to the east and merges in the “retusa” “turgida” “mirabilis” melange of that area.

Jakub Jilemicky: I think plants from Brakkekuil, a bit further South have also both, turgida and retusa nigra.

Bruce Bayer: Yes Jakub, I posted a whole set of those Brakkekuil plants – what is also interesting about them is that it is neither a steep cliff-like habitat nor a flat surface habitat.

https://haworthia-updates.haworthia.org/2012/01/30/volume-5-chapter-4-haworthia-retusa-part-2/

7. 2019.5.13 – Neither “retusa nor turgida” are known near this to the SW of Heidelberg on the Slang River.

7. 2019.5.19 – Perhaps I should have spelled out my hypothesis in the very beginning of this haworthia species mystery. This is that Haworthia was separable into morphological and numerically defined entities that would fit into a clearly defined set of species. A proviso being that the word species was defined as based on the ill-considered and untrue precept of non-interbreeding entities. The nature of variability coupled with this weak species definition result in the initial hypothesis to be totally false. Haworthia does not present such clearly defined entities and in fact there is a very variable set of elements that can only be recognised in respect of their general appearance and their geographic distribution and relationships. These are themselves mildly confounded. In fact, the issue was grossly confounded by the now established fact that Haworthia was, in reality, three very different genera.

When I dismantled G.G. Smith’s putative two-volume manuscript purporting to be a monograph, these three genera were split into 20 sections with specimens of many of the currently recognised species, represented in more than one. I never pretended to be the knight in shiny armour that was a going to slay that windmill of confusion and I put a great deal of effort into looking for someone who could and would. That never materialised.

From my personal perspective (isolated as I am from the popular literature of the last 5 years), all I can say is that there is a publication authored by myself and Dr. John Manning that reasonably presents the truth of Haworthia. It does not entirely do so as this series of posts will show.

Science as a real matter of inquiry, together with objectivity, rationality, logic and reason, are not strongly represented in the human endeavours that have been and are being made to properly understand the phenomenon of the Haworthia species that can be observed. There is no doubt that DNA sequencing offers the very best solution to the issue. My experience suggests that there is still some way to go before that goal is reached and it will not be done until the complexity of issues like this one are seen to be real and understood as they are.

This is the reference … Bayer, M. B. and J. C. Manning. 2012. A rationalization of names in Haworthia: a list of species with new combinations and new synonyms. Haworthia Update, Essays on Haworthia, vol. 7. Preston: Alsterworthia International.

Haworthia Nonemclator 10-7-2012

8. 2019.05.14 – The Slang River first runs south as does the Duiwenhoks, but instead of going on to the sea, turns westward and enters the Breede River further west. So, we find retusa again near Malgas – further south than my first post. Note I have abandoned using the name “turgida”! This is symptomatic of the whole issue viz. that application of names is so difficult because even in a single population the plants can be so different. Plant geography is interesting here in that Ceropegia (Stapeliopsis) stayneri is only found in its lower reaches. Recent developments in the taxonomy of the asclepiads and the disappearance of many genera is also worth noting.

9. 2019.05.15 – It gets exciting for whatever direction one goes. This is to Haarwegskloof where Jannie Groenewald found this namesake. Here we also have H. mutica and H. mirabilis. So, I am trying to explain that they are actually the same species!!! Three posts a day and all my pictures; by the years end you too will concede that they are.

Haarwegskloof is west of the Breede River and it is important to consider the nature of rivers and river systems as they affect speciation. Rivers follow fault lines and other geological features, while haworthias thrive on exposed eroded surfaces and low biomass sites. So rapidly and deeply eroded formations provide different habitats to shallow surface erosion. The immediate river environment results in stepper cliffs and rawer rock. Thus we have the very deeply cut Gouritz river and fairly deeply cut Goukou, Duiwenhoks, Breede and Slang Rivers. West of the Breede we have curious rather shallow valley drainage systems with some water courses barely reaching the sea.

10. 2019.05.15 – Still going further west from Haarwegskloof we come to Rooivlei, N Bredasdorp to the most SW known retusa and we have entanglement supreme. We have various forms of mirabilis, and perhaps rossouwii or variegata while mutica lurks not far away. These last two locations do make it difficult to accept that we are dealing with one species when in fact it suggests that the situation may be far more complicated. But really if one takes the various nodes on river or drainage systems, the situation becomes much more obvious.

Steven Molteno: What I don’t fully understand is how Rooivlei and Haarwegskloof connect to the major river systems to the east. Breede river tributaries maybe?

Bruce Bayer: No, they do not and I don’t think so much significance needs to be attached to this if any. The occurrences of the populations are attached to geographic and geological features so drainage systems are just a weak underlying kind of thing. It is like the role of the Langeberg separating the karoo from the S Cape that also relates to the greater retusa complex and how emelyae emerges as a consequence.

Steven Molteno: Okay thanks Bruce, yes I didn’t see an obvious riverine corridor going there. But then there are several interesting parallels between that area north of Bredasdorp, and (specifically fluvial!) terrains further east. The relevant Haworthias seem to extend here of course, as you point out. In addition, Tulista marginata – usually known from high, eroded fluvial terraces further east (e.g. Die Plotte, Riversdale) or north (Bontebok, Drew, and maybe … overlying the Ashton Enon Formation!) … T. marginata is also on the flat-topped hills N. Bredasdorp, which intrigues me. I even tried to look up whether perhaps a long-dead primeval river might have run past there, before the glacial cycles played havoc with the sea levels! It doesn’t seem so, but who knows.

11. 2019.5.16 – Where does one go from here? These pictures all date from after the 1999 revision so forgetting all the pictorial information to that date I have only given a peek at what was “turgida”. I still have 5-6 folders for that, 24 for retusa itself and 14 for “nigra”. Then many more for mirabilis, mutica, pygmaea and emelyae. There are 6-8 threads among these that trace various continuities and I think next I will just go via “nigra” on the way to mutica?

What I find extraordinary is the way in which amateur succulent journals resist any effort to rationalise names in accordance with some sort of scientific intent. The Latin binomial seems to be all significant and it has made life in Haworthia absurd. All it needs is a very simple adaptation and the adoption of a convention where the species name is taken as generic and is so understood in the community?

Cohan Fulford: taxonomy is seen as a dirty word by many horticulturists, with what is perceived as constant changes – often to well known and cherished entities – and dissent among taxonomists taken as meaning the whole field is arbitrary.

Bruce Bayer: Yes Cohan and even qualified scientists in other disciplines, including botany, do not actually understand the processes.

True “retusa” is already posted from this same locality on the Duiwenhoks and just across the river is a small form of mirabilis that could involve floribunda! When I first saw a picture of this I speculated that it might prove to be involved with H. mutica. I cannot remember the exact history and why it came to be known as nigra var mutica – but as a matter of interest I quote from the New Haworthia Handbook with reference to mutica and retusa…”if the two species do meet it can be suggested to be in the area between Heidelberg and the lower Breede River”. How true this has proved to be. But no meeting at all – just a dramatic flow.

Cohan Fulford: Every one beautiful! Fascinating to get a glimpse into how the various taxa interact.

Bruce Bayer: Better understood as largely interaction within one taxon? There is or are levels above this doing similar things. Why are taxonomists not examining these weird happenings?

12. 2019.5.17 – Here are more pictures of the potential mutica viz a mirabiloid/retusa with floribunda infused. Between Kransriviermond and Heidelberg.

Steven Molteno: It seems a lot of the complexity arises from repeated radiations and confluences – in certain directions. The directions interest me because the southern cape’s been subject to repeated pulses of arid – wet climate (driven by the ice caps’ cycles apparently). So the suitable “bands” of habitat would’ve moved back and forth over the millennia, inhaling and exhaling Haworthia radiations across the Overberg! Perhaps especially along the river valleys, which sound like plant highways!

Bruce Bayer: Yes – I think one of the strongest points to make is that the model has been that an imagined model based on this sort of thinking has been predictability e.g. the fact that nigra is continuous with mutica.

Irving Navarro: Steven Molteno, I always wondered how bruynsii came to be, it’s yellow roots always amazed me

Bruce Bayer: Irving that IS an interesting observation. Bruynsii may have its origins in or near sordida but yellow in the roots may be an aloid character?

13. 2019.5.17 – There is also a very strong north/south influence in the distribution of the elements. Floribunda is playing a much stronger role in my narrative than I anticipated. It covers virtually the same range as the 6 elements of retusa. But to the east it does not occur south of the principal road from Cape Town to Mossel Bay. It is only at Swellendam where it does come south of that and it emerges from mirabilis as a wholly separate element far further south at the Potberg where it is partly confounded with H. variegata. It has an incredibly close association with the atrofusca forms of the mirabiloids. These here are from 3 populations west of Heidelberg where the preceding muticoid influence is very strong as well as the input from floribunda. As a general comment, the mirabiloids to the west are generally more dominant in the southern parts.

Bob Hunter: Bruce – I am finding this series of posts very.interesting, and am glad that they are going to be collated, edited, and published. I hope that, given the problems with the traditionally used language here, that published collation can in some way accompanied by some sort of ‘graphical’ illustration of these insights – a hard ask, I know, but it would help a lot to make people think more about this issue.

Bruce Bayer: Thank you Bob, there is already a great deal published in HaworthiaUpdates.org as edited by Lawrence Loucka, but the whole issue is bigger than myself and it would be nice to grow it as a community of interested people like yourself. Steven Molteno is addressing a big part of the story and we are going to learn a lot more.

14. 2019.5.17 – I have not pre-planned this narrative of the haworthia species mystery in the belief that there is a natural coherence to it despite the very broken nature of the available evidence. Human habitation and agricultural development have changed the landscape dramatically while exploration itself is not complete. I am finding a flow to the story that comes out of the evidence and not out of personal prejudice. Here I post some pictures for a population a bit south of the previous one. The retusoid and muticoid influence is dropping out and it is changing to mirabiloid dominance with an unquestionable infusion of florabundoid.

The language of “oids” is proving very meaningful and from here I will pick up again on the mirabiloid retusoid and the muticoid mirabiloids. I asked Lawrence to edit into the introduction I bit I wrote about the nature of the plants as they relate to geographic space, geological formation, slope and aspect. But also the question of a spring, as opposed to a late summer – early autumn flowering period being a major driver of diversity of this single species complex.

From the population illustrated, the mirabiloids burst southward and westward right up northward into the Robertson karoo and through the Tradouw Pass into the Little Karoo. From there they extend eastwards and merge into the emelyaeoids as other retusoids spread northward via Tradouw Pass, Garcia Pass, and Gouritz Gorge.

15. 2019.5.19 – Language – I have been and am, using the suffix “-oid” to mean belonging to a group of similar things. So far I have not shown that mirabilis and retusa are the same group of things. In fact I may not have done so for turgida and retusa either and tend to treat this latter as an accepted fact. What I have shown is my perception that a mutica-like group originates from retusa and mirabilis. So muticoid arose from retusoid and mirabiloid. So far my posts are following these groups westwards from the Duiwenhoks river and Heidelberg. My last two posts were mirabiloid with a floribunda infusion and these flower autumn. The next post is a little further west of a retusoid – I will call it muticoid retusoid because it does have a mirabilis infusion and it will lead to mutica. As will another group I will call muticoid mirabilis. Bear in mind that flowering time is important and not always available. These pictures now posted, were taken in September and thus spring flowering.

Incidental but very important is that nearby is one of the few floribunda populations South of the main e/w highway. These elements/oids are not sharing habitat!!

2019.5.16 – I need to digress – again. Here at Kransriviermond we have two things viz mirabiloid and retusoid. They hybridise to produce this element we call muticoid (“nigra”). The word “hybridise” normally implies interbreeding of different things (species). This is not the case here at Kransriviermond where it is recombination of diverging elements of the SAME species. In my next set of pictures from further north you will need to understand the dramatic role of H. floribunda. This is because this species is intrinsic in the mirabiloid at Kransriviermond. Note this carefully as it becomes very significant both eastwards and westwards and the issue will re-occur. Another thing very important to note is that the popular succulent journals service commercialism and popular interest; and not science. Some years ago there was a very well-written and eloquent article in one of the journals that virtually called for dumping of any discussion of the meaning of names in respect of species. All that was important was for a set of binomials to satisfy the collector frenzy and make it easy to label pictures? Thankfully Lawrence Loucka is editing and compiling all this stuff I am now posting, for Updates. I trust that one day there will be a true record untainted by the poor record objectivity has in plant classification. I will work on a map.

Please be sure you have read that digression carefully. This population is between Kransriviermond and Heidelberg. A significantly small version of H. rossouwii is nearby, a turgidaoid green form of retusa is not that far away and there is a mirabiloid very close by that is distinctly floribundoid. Haworthia floribunda has a characteristic rounded leaf tip without a terminal point – is this what appears across the board from pygmaea to mutica? I need to do a separate series on floribunda to explain its curious distribution and how it is absorbed in, or reacts with, the greater retusa complex. Sadly for me this sort of stuff is not popular with the Haworthia community and botanists have proven seriously disinterested – including those that have pretended to explore the DNA of the group.