(The article below was published in Alsterworthia International Vol 14 Issue 1, 16-27 March 2014)

Haworthia obserata sp. nov. J. G. Marx

Gerhard Marx

marx.gerhard@gmail.com

Abstract: A new species of Haworthia (Asphodelaceae, Aloaceae) subgenus Haworthia, is described, known only from a small area on the farm Brandrivier east of Barrydale in the Little Karoo. Although its habitat is north of the Langeberg Mountains and it shares a few superficial similarities with the geographically nearby Haworthia breueri Hayashi, it is closer related to the summer-flowering elements like Haworthia magnifica, H. mirabilis and H. groenewaldii which are found south of the mountain range.

Type: South Africa, Western Cape Province, 3321CC Brandrivier, between Barrydale and Garcia’s Pass.  Martin Scott -WMS 102 (holotype, GRA, ex cult).

Description:

Plant a strictly solitary rosette, flattened bell-shaped from side view, up to 7 cm in diameter and to 2,5 cm deep; leaves thick and fleshy, to 20 mm wide and 32 mm in length; average 6 per plant in habitat but up to 14 in cultivation; windowed upper surface area triangular from above, surface subtly scabrid with numerous minute sharp papillae, flattened but slightly convex, up to 18 mm wide and 22 mm long with pointed tip ending in sharply acuminate and slightly sideways twisted tip; dark grey-green in colour with few off-white parallel facial lines, occasionally some white soft-edged dots in-between towards the narrowing end area. Facial lines consist of densely grouped white longitudinal flecks, occasionally with few broken opaque dark green to brown islands in the centre, a few stretching almost to the leaf tip, others about a third as long. Lower leaf keeled only near the tip for about 4 mm, keel always with slight sideways twist; margins smooth from the base for half the leaf length, thereafter bearing numerous minute teeth up to the leaf-tip, lower leaf surface smooth, dark brown-grey gradually blending into light green towards the base. 

Roots numerous, up to 18 , thick and fleshy, up to 9 mm tick near the base and narrowing towards the tips, up to 10 cm long ( root measurements apply to plants in cultivation).

Inflorescence a solitary slender raceme up to 42 cm long including peduncle, 3 mm wide at base with up to 15 sterile bracts, 3mm wide and 5 mm in length with dark brown median line; raceme to 130 mm long bearing 12 to 15 spirally arranged flowers. Pedicels to 1,5 mm long. Fertile bracts 3mm wide and to 6 mm long, deltoid, acute, with dark brown to maroon dusky central line with subtle pink cloudy patches near the recurved lobe tips. Floret to 16 mm long, white with each lobe having a dark green to brown soft-edged median line on both sides,  wider stained on the inside ; perianth to 4 mm thick, free portions of upper lobes strongly recurved at the tips, lower lobes curving downward for a third to half their length.

Ovary 3 to 3,5 mm long, 1,5 mm in diameter, dark green. Style 1 mm long, curving upward. Stamens up to 6 mm long.

Fruit to 16 mm long, 4 mm in diameter, smooth, bluish green. Seeds to 2,5 mm in length, 1 mm wide, with flattened lip along angles, dark grey-brown, tuberculate.

Add c

·         The typical appearance of H. obserata in habitat, small and rarely with more than seven leaves.

aption

The area immediately north of the Langeberg and Outeniqua mountains appears to be a transitional zone where some Haworthia elements that are mostly confined to the area between the coast and the mountains seem to have ‘bounced’ across the mountains to blend with the elements that are limited to the Little Karoo areas. These ‘intermediate’ populations are difficult to place as they may contain features of both the coastal as well as the Little Karoo residents. Examples of such cases are H. indigoa and H. truteriorum occurring immediately north of the Outeniqua mountains and both showing some shared leaf features with the Little Karoo-confined H. bayeri, while both are much closer linked to the coastal H. mirabilis-H. magnifica groups in terms of flower characters and flowering time. The intimidating presence of the dividing mountain range would cause a geographically-based species concept to force these elements into close association with the Little Karoo species while a biological species perception would emphasize the close relationship suggested by the flower features and flowering time to the coastal elements to the south-west. As a result of the latter conflict it may be best to regard these transitional elements currently as separate species in their own right as the differences to both their northern and southern relatives are significant and any attempts at lumping are dependent upon forceful artificiality.

The new species described here is another case of such an intermediate element occurring on the farm Brandrivier on the northern foothills of the Langeberg mountain range.

·         Despite being fully mature these two plants of H. obserata have only four leaves each.

The chosen name for this distinct component within the genus Haworthia refers to the fact that its existence has been concealed by both artificial incarceration and various other obscuring factors. It grows extremely well hidden in the close company of the large and remarkably unique and equally rare Haworthia opalina. The attention and attraction generated by the well-known H. opalina amongst succulent enthusiasts have not been welcomed by the landowners and for the past decade or more succulent tourists as well as botanical researchers, regardless of how impressive their credentials, have all strictly and unapologetically been denied to visit the locality.

In addition, the Brandrivier area falls within a rather peculiarly puzzling and somewhat illogical juxtaposition of Haworthia populations. At first glance the new Haworthia obserata sp nov can be easily mistaken to represent a form of Haworthia breueri, particularly if the observation is largely influenced by the fact that H. breueri occurs relatively nearby towards the east.

Another loosely similar retusoid Haworthia grows even closer as hardly more than one kilometre away Haworthia multifolia (H. emelyae var multifolia Bayer) can be found. Both H. breueri and H. multifolia are linked to the H. emelyae/ picta group in terms of shared flower features and early spring flowering habits.

The flower features and flowering time of H. obserata links it intimately to the summer-flowering H. mirabilis- H.magnifica- H. maraisii group. H. obserata flowers rather late in summer, mainly during March and early April.

On the neighbouring farm Klein Doornrivier and only three kilometres to the west of the Brandrivier locality of H. obserata grows a form of H. maraisii that also shares the same flower characters and flowering time as H. obserata. The Klein Doornrivier plants differ substantially in terms of leaf colour and shape as well as the number of leaves per rosette from H. obserata but the flower features indicate a much closer relationship between these Klein Doornrivier plants and H. obserata, than to the H. emelyae/picta-related elements to the east of Brandrivier. 

Photo on the left shows the largest plant seen in habitat having longer and sharply pointed leaves with sparse and small rough tubercles on the upper surfaces. Photo on right shows another plant of H. obserata hiding amongst thatch-reed.

As mentioned, if one’s observations are influenced by the geographic situation of H. obserata, then one could easily be tempted upon first impression and brief superficial observation to link it to the nearby-occurring H. breueri. The fact that the flowers and different flowering season indicate that in reality there is not a close relationship with H. breueri is yet another example of the precariously deceptive nature of Haworthia elements. The challenge of studying this variable genus lies in the fact that frequently the features that may appear to the indiscriminating observer like similarities are in fact deceptive disguises concealing rather significant differences.

In general, however, also the morphological features of H. obserata var nov compare quite correspondingly to H. magnifica-related elements occurring to the south of the Langeberg range. It can look rather similar to some forms of H. magnifica found to the east of Riversdale but also comes surprisingly close in appearance to the variety of H. groenewaldii growing along the eastern border of the Bontebok National Park. The latter H. groenewaldii – linked component differs to such an extent from the typical form of the species growing further east near Mullersrus, that it probably deserves full varietal status at least.

·         Haworthia obserata in cultivation. Compare the slim peduncle buds with the robust peduncle (typical of the H. emelyae group) in the photo of H. multifolia. Photo on the right shows a close-up view of the leaf surface texture and colour in cultivation.

For comparison – left photo shows H. multifolia growing less than 2 km to the east of the habitat of H. obserata . Note the robust flower peduncle. In the  centre is the well-known H. breueri on the farm Waterval, 30 km east of Brandrivier. Note the more roughly toothed, longer and narrower leaves and rosette consisting of well over 20 leaves. On the right is the form of H. groenewaldii on Rotterdam farm along the eastern border of the Bontebok National Park, which also show a few superficial similarities to H. obserata in terms of leaf colour but a more important affinity is in the shared flowering time.

Left and centre  the flower peduncle and flowers of H. obserata. Compare with the flower peduncle of H. breueri ( on the right) to see that the flowers are less densely arranged and slightly smaller in size.

History:

It was mentioned above that the chosen name of the new species partially refers to the artificial isolation measures applied by the current landowners. In view of the latter, the habitat photos published herewith as well as the origin of the Type specimen need some explanation:

On 11 August 2008 I was fortunate enough to visit the locality of H. opalina accompanied by Ernst van Jaarsveld and Bruce and Daphne Bayer. The visit was the result of months of diplomatic negotiations and arrangements done by Ernst van Jaarsveld. In addition to Ernst’s unique diplomatic skills and kind personality as well as being such a well-known succulent celebrity known from numerous TV appearances it also helped significantly that Ernst was well-known and trusted by the original owner of Brandrivier farm who is the father of the two brothers who are the current owners.

The main aim of the visit was to see and photograph H. opalina in habitat.

On a mild sunny day in August we all met at the Brandrivier homestead and were then accompanied by Mr George Nel to the H. opalina locality on top of a low hill not very far from the house.

It was quite a thrill to see H. opalina in the flesh and we all made sure to appreciate every moment of the extremely rare opportunity. After having taken more than enough habitat pictures of H. opalina, I started searching a little distance away from the H. opalina spot to see if I could not find H. multifolia ( = H. emelyae var multifolia sensu Bayer) which was reported to be growing not far from H. opalina by Kobus Venter. Kobus visited the locality during 1994 when the more hospitable Mr Nel senior was still in command on the farm and Kobus reported having found H. multifolia nearby H. opalina.

I searched thoroughly literally under every bush and finally encountered a very well-camouflaged dull grey coloured retusoid Haworthia growing there. It did not look like H. multifolia at all but I vaguely remembered Kobus Venter’s comments that in the Brandrivier population the plants are more robust. So at the time I assumed that it must be how drastically different H. multifolia looks at this locality and so too did Bruce Bayer when he saw the plants I had just found. Afterwards in his Haworthia Update Vol. 6 he published photos of these plants under the caption “ MBB 7846. Haworthia emelyae ‘multifolia’ Brandrivier” (p.30, 31)

Back home I immediately looked up the published habitat photo taken by Kobus Venter of ‘JDV 94/32, H. emelyae var multifolia, Brandrivier’ ( see photo on p. 71 of Haworthia Revisited by M.B. Bayer). The green erect-leaved rosettes of the plants on the latter photo really look nothing at all like the flattened dull grey-green plants I found near the H. opalina population on Brandrivier. In fact, the plants I observed there have mostly only 5 to 7 leaves per rosette while the multifolia plants on Kobus’ photos showed plants with over 20 leaves per rosette. In addition to the differences in leaf shape, size and colour, I became convinced that what I saw on Brandrivier was not H. emelyae var multifolia at all. Even more importantly, the time of our Brandrivier visit (mid-August) is the start of the flowering season for H. breueri, H. multifolia, H. emelyae and relatives while these Brandrivier plants showed no sign of developing buds (as can be seen on the habitat pictures herewith).

A few months later I discussed the Brandrivier visit with Martin Scott since he too, like Kobus Venter, had been allowed to visit the H. opalina population way back during the early 1990’s and I wanted to know whether Martin had perhaps also seen the strange dull-grey retusoid Haworthia there. Yes, indeed, he said, but not as close to the H. opalina population as I had found the plants; Martin found a population of them on the adjacent low hill immediately to the east. Martin was allowed to collect some leaves of a few clones (WMS 102) and these produced plants which he still has in cultivation. Despite regular successful propagation from leaves, Martin has not been able to propagate it in such good quantities from seed as one can easily do with H. emelyae relatives like H. multifolia and H. breueri.  Most summer-flowering Haworthias (for example H. splendens, H. maraisii, H. marxii, H. archeri, H. dimorpha, H. wittebergensis , etc.) are considerably more difficult to propagate from seed than spring-flowering species. Martin was kind enough to give some of his original material of it to me to study under cultivation and to see if they would be more willing to produce seed for me in my cooler and well ventilated greenhouse situated high on a hill slope outside Oudtshoorn.

One of these WMS 102 samples also became the Type Specimen, sent to GRA.

During the past four years I have now also consistently observed the late summer flowering time of this species (March) as well as its reluctance to produce abundant fruits despite careful hand-pollination and the resulting seedlings are very slow in development, much slower than any H. emelyae/ picta relatives.  

·         The fynbos habitat on top of a low hill on Brandrivier farm, habitat of H. opalina and H. obserata sp nov.