If the name “H.enigma” applies to the plant (or plants) from east of Riversdale at Komserante, it is a name that I really do not advise to be taken seriously from a botanical point of view. It is useful at population level and to demonstrate the nature of classification difficulties but it is a minor problem in so far as those difficulties extend. The plants were first shown to me by J. Dekenah on the same day that he also showed me ”H. magnifica” in the Nature Reserve just south of Riversdale that is less than 3km away. My impression then was that it was the same element even if it did look a bit different. The plants are quite large (to 70mm diameter), fairly tubercled and often with lines in the upper retused area of the leaf face. While I originally classified “H. maraisii” under “H. magnifica”, I later separated them because it seemed so incongruous to include all the variants of the western “H. maraisii” with the few populations of “H. magnifica” then known. Also, as Essie Esterhuizen pointed out, “H atrofusca” as a variant of “H. magnifica”, seemed to be more dominant than had been realized. There were several other complications largely due to ignorance. Since my revision I have done so much more exploration and turned up so much new material that I have been forced to the conclusion that there is really one main element involved and that is H. mirabilis. This is where I believe the Komserante plants belong and the difference from the Nature Reserve population is due to a degree of infusion of H. retusa.

I revisited the site with Kobus Venter many years ago but did not look at a reported second population higher up the hill, taking it to be a little different based on plants I saw in Kobus’ collection. What was on my mind while we were recently exploring the area further east to examine the possible connection of H. mirabilis “magnifica” to “splendens” (and which we confirmed), was the fact I had never seen Kobus’ plants from Kruis Rivier northeast of Riversdale other than in Kobus’ collection. The plants I saw were also generally more robust than “H. magnifica” and more evenly tubercled. Kobus kindly took me to that Kruis River locality and much to my surprise the plants were in flower late October (see JDV92/65 Figs1). This is quite wrong for H. mirabilis, which is essentially a summer flowering species. I later went again to explore Komserante more thoroughly and to look at both the “magnifica” populations to which I believe the name “H. enigma” has been applied. The populations are in fact no more than 75m apart and cannot be considered to be genetically discrete at all (see MBB7778 Figs 2, and MBB7779 Figs 3). While it is true that the habitats are slightly different, this is reflected in the plants that at the upper slope of the hillside are vegetatively more robust and even clump forming, while those lower down in a bushier grassier habitat tend to be solitary and more withdrawn into the soil. These plants flower in summer and it is evident to me that there must have been some genetic exchange with H. retusa that grows approximately 200m away on the same hillside.

1. JDV92/65 H. mirabilis Kruisrivier.

2. MBB7778 H. mirabilis. Komserante.

3. MBB7779 H. mirabilis. Komserante.

Habitat preferences are very strict and it is curious to observe that there are four small buttresses forming the western slope of the low hillside. On the northern one we have the two H. mirabilis populations. On the second is H. retusa ’geraldii’ (see MBB7780 Figs 4) that is very clump forming, the third has no haworthias and on the fourth is a less clump forming H. retusa that is often referred to as ‘fouchei’  (see MBB7781 Figs 5) because the leaves are slightly longer and more upright than in ’geraldii’.  This is not quite correct. Mr. Dekenah showed me what he took to be the true H. retusa ‘fouchei’, further south off the hillside and along the stream at Komserante. This habitat has since been cleared for farmland. The plants were very large and the leaves considerably more erect than is the case for the plants on the hillside. Curiously H. minima is present on the first and last buttresses. Flowering time for these H. retusa is spring but we did see vicarious flowering in summer so that definitely it is possible and probably a regular occurrence for out-of-season flowering to occur with the possibility of hybridization. I noted such hybrids between H. retusa and H. mirabilis in our exploration further east from Riversdale and I did think some of the clones in the Komserante plants evidenced this as well.

4. MBB7780 H. retusa. Komserante.

5. MBB7781 H. retusa cf. ‘foucherii’. Komserante.

The reality is that the Kruisriver and Komserante “magnifica” may flower at different times, but they are vegetatively very similar indeed. I consider that they belong in the same system as H. mirabilis and this conservative view of “species” as systems is the most sensible way to interpret and understand the plants. The approach can be taken further, as I tried to do in an earlier article concerning H. mutica ‘nigra’. In that article I dealt with a whole series of populations in and around Heidelberg, to show that there are three main elements viz. H. retusa (including H. turgida), H. mirabilis (including H. magnifica) and H. floribunda. I did not deal specifically with, nor resolve, the issue of H. mutica var nigra. I also noted populations along the lower Duiwenhoks river and suggested that H. floribunda is absorbed into H. mirabilis virtually south of the N2 east west highway, re-emerging on the northwestern corner of the Potberg in the south.

My most recent exploration was to check the flowering times of the original population of H. mutica var nigra at its origin at Kransriviermond and what I assumed to be the same thing on rather limited examination further north at Morning Star. Both these populations are along the Duiwenhoks river southeast of Heidelberg and in a continuum of four populations of H. retusa extending from just east of Heidelberg to still further east at an eastern Diepkloof locality (I think it has been referred to as Droekloof and Droerivier too, and the name ‘chromatica’ attached to the plants – see MBB7794 Figs 6).  Less than 10km away is a population of H. retusa at Pienaarsriver (see MBB7776 figs in Chapter 3) that is very near to the whole array of populations of Kiewietskraal that I discussed in my writings about H. mutica var nigra.  Two of those were distinctly H. retusa and so is the Pienaarsriver population that is different in that the plants are mostly quite dark in colour.

6. MBB7794 h. retusa ‘chromatica’. SE Heidelberg.

In those same writings I mentioned several populations of H. mirabilis along the Duiwenhoks river as well as populations of H. retusa var turgida. Here I am unashamedly relating “turgida” to “retusa” as I believe that when one calls all the populations now known to mind, as well as the variants they contain, this is the correct position of the plants in relation to a sensible and rational species concept. It is evident that H. floribunda is merged in H. mirabilis south of Heidelberg and the same thing occurs south of Swellendam. Thus I see no problem in now dealing with the population at Kransriviermond to which I applied the name H. mutica var nigra. I attached the varietal name to H. mutica because this is what G.G.Smith had done, and I believed that it was the link between H. retusa in the east and H. mutica in the west, when there were still huge gaps in the known distribution records for both species.

I have mentioned a population of ostensibly H. mutica var nigra at Morning Star, which is about 5km north of the Kransriviermond origin of the plants so named. This Morning Star population (see MBB7221 Figs 7) is very significant because while it includes virtually the same range of variants as the Kransriviermond plants, it flowers in summer as opposed to spring for Kransriviermond. Not only that, but it is 300meters away from a population of H. mirabilis (previously this would have been H. magnifica) that flowers and seeds at the same time. The habitat is marine plane ferricrete that overlies Bokkeveld shale but there are subtle differences between the two habitats.

7. MBB7221 H. retusa ‘nigra’. Morning Star.

The Kransriviermond plants (see MBB7804 Figs 8) are down in the Duiwenhoks valley in exposed Bokkeveld shale and I consider that the dark colour of the plants in the two populations is due to infusion of H. mirabilis. The rounded leaf tip would be derived from the “atrofuscoid” element of H. mirabilis while some plants also have the tuberculate roughness that that element often has.

8. MBB7804 H. retusa ‘nigra’. Kransriviermond.

Further south and west is the Slangrivier valley where there is also a suite of populations that includes smaller forms of H. mirabilis (H. heidelbergensis var. toonensis) as well a H. retusa var turgida also in an unusual form that supports the taxonomic position I now adopt.

The conclusion is that the Morning Star and Kransriviermond ‘nigra’ can be regarded as H. retusa within the same broader concept that includes ‘turgida’. The problem now arises from still another new discovery. This is from along the Buffeljags River (see MBB7801 Figs 9) about 10km southeast of Swellendam and 12km from the nearest obvious H. retusa at Goedverwagting further east; and from the less obvious H. retusa (that I have elsewhere noted as H. mirabilis) south of the Tradouw Pass. The nearest H. mutica are within about 15km to the west at Napky and Luiperdskop. The plants are the now customary “variable”. The leaf tips are very rounded. Some plain, some lined, some with curious submersed pinkish dots. The surfaces can be quite smoothish or else the surface cells terminate in minute spines that give the surfaces a glistening appearance – evident in H. mutica too. These plants flower in summer whereas both H. mutica and H. retusa are spring flowering. Note needs to be taken of the habitat. It is an extensive and old river gravel embankment and terrace, and therefore another variant on the interface of recent ferricrete deposits and the older Bokkeveld shales. The vegetation is grassy Karooid bushveld, unlike the dry fynbos to the northwest or the renosterveld (mainly cultivated) to the east.

9. MBB7801 H. mutica. ‘Buffeljags’.

Kobus Venter saw the photographs and suggested that the plants might be H. mirabilis’  and reminded him of ‘atrofusca’, perhaps also by virtue of the flowering time. This is of course what my “enigma” is all about. It is compounded in this case because these Buffeljags ‘muticas’ are rather odd in that the leaf ends tend to be flattened almost as though H. floribunda has also entered the fray. Kobus was much happier with the H. mutica connection when he saw the plants in the field.

The story here closes with H. mutica MBB7741 from Dankbaar about 12km west of the Breede River bridge at Swellendam. The plants here also have very glistening surfaces and thus seem to confirm the very original perception of G.G. Smith’s that the Duiwenhoks plants and the plants west of the Breede River were connected. It is certain that he never envisaged the complexity of the interrelationships between three main elements viz. H. mirabilis, H. floribunda and H. retusa.

Footnote. It is just assumed that perhaps the name “H. enigma” has been used for the Komserante H. mirabilis, because that population is indeed enigmatic as this article explains. Gerhard Marx informs that “enigma” applies to a H. mirabilis population further west where the more atrofusca variants occur. Since writing this chapter yet another similar population to the Kruis River and Komserante populations has been found about 15km SE Riversdale. ♦