Haworthia History – Zurich (1997)


The history of this succulent genus is an interesting one as it has its roots in the earliest botanical exploration of the Cape.  Haworthia were included in the first collections of plants when the main concern of botanical collections was medicinal and culinary use of plants.  Haworthia were depicted by several different artists in the late 17th and early 18th century.  These early illustrations were surprisingly good and can be interpreted according to present understanding.  The fact that they have been the object of much confusion is really attributable to the inherent difficulty in understanding the genus and its component species, rather than to poor quality of the illustration.

The genus is not easily separated into separate and easily distinguishable species and the history is correspondingly confused.  Although there are some earlier illustrations, it is the illustrations in Caspar Commelin’s ‘Horti Medici Amstelodamensis in 1701 that form the basis of the taxonomic record.

By the time that Linnaeus published his celebrated Species Plantarum in 1753, eight species were already illustrated, of which only five were accommodated in Linnaeus’ classification.  Unfortunately these were also really regarded as only varieties of one single species of Aloe.  The English naturalist and botanist, Adrian Hardy Haworth, in 1804, seems to have provided the first real impetus to the study of Haworthia when he classified 21 species under a section Parviflorae of the genus Aloe.  These included six species which were later recognised in the genus Astroloba.  Haworth further described about 20 more species, and Baker also some 17 species in 1880.  Baker revised the whole genus in Flora Capensis Vol.6 of 1896.  Here he recognised 64 species.  Berger again revised the genus in 1908 in Engler’s Das Pflanzenreich.

Joseph Karl von Poellnitz became interested in the genus and published extensively from 1928 to about 1940.  He was killed in an Allied bombing raid towards the end of the second World War.  He was followed by Gerald Graham Smith, a South African, who was the first in a position to really see and experience the genus as it occurs in nature.  A.J.A. Uitewaal, at Amsterdam, was a european contempory and these two did communicate briefly.  However, much acrimony arose between Flavio Resende and Smith, which led to Smith abandoning his interest in the genus.  J.R. Brown and J.W. Dodson, in the United States, maintained a deep interest in Haworthia until about 1957.  However, both were handicapped by lack of field experience.  In 1957, Brown listed 160 species and 210 varieties, and this listing was virtually repeated in the Handbook of Succulent plants by Herman Jacobsen published in 1954.

In 1962, Bayer described a new variety of H. limifolia but only started a more serious and comprehensive review of the genus in 1969.  Colonel Charles Scott was simultaneously engaged in studying the genus and these two authors did not agree on a classification system or in fact on the application of many of the available names.  Bayer published an illustrated check-list in 1972 which was followed by a revised edition in 1982.  Col Scott’s revision appeared in 1975.  While there has been a general failed attempt to reconcile the works of the two authors, Bayer’s work has generally been more widely accepted.  Neither author followed the botanically correct route of typification of the botanical names and this was done by Breuer and Metzing (1996).  Their work tends to replicate the failure of european workers who are really not familiar with the plants they are attempting to classify.  It is a common mistake to think that a few, even extended, excursions into the Cape countryside can substitute for proper familiarity with the scale of the countryside, the geographic features, the general flora and vegetation, and the way in which Haworthia species relate to these and also to each other.

The classification of Haworthia species is intimately related to the problem of genera in the family Aloaceae.  Many authors have expressed opinions on how the family should be arranged into genera.  Whatever these opinions are, they have all suffered from the same weakness, which is the failure to properly recognise the characters on which existing genera are based.  Furthermore, they have failed to examine the variability of the characters of the species which constitute those genera.

It is quite incorrect to assume the general point of view that Haworthia is a homogenous taxonomic unit equivalent of Astroloba.  On a larger scale, the genus Aloe is also assumed (by those proposing generic rearrangements) to be an homogenous unit.  This is not true for either of the three genera mentioned.  In the case of Aloe, there are a number of species which may fit no more comfortably in Aloe than does, say, Chortolirion.  An example is the group of grass-aloes or Leptaloeae.  In the case of Haworthia there are three very distinct groups recognised originally by Uitewaal and formalised as subgenera by Bayer.  Dr.M. Hayashi (unpublished communication) has gone further to suggest that these three subgenera are also divisible.  Astroloba similarly can be shown, on the basis of flower structure, to comprise two groups.  However, mention must also be made of the many attempts to divide Haworthia into sections.  These have without exceptions, proved fallacious to the extent that members of the same species have been assigned to different sections.

While the debate about generic states is unresolved, the question of species is even more complex and debateable.  Bayer has put forward the view that the species, in their respective sub-genera, can only be recognised on the basis of geographic continuity and co-occurrence.

The available reference works on Haworthia are:

  • Haworthia and Astroloba – A collector guide’ by J.W. Pilbeam (1978).
  • The genus Haworthia, a taxonomic revision’ by C.L. Scott (1985).
  • The new Haworthia Handbook’ by M.B. Bayer (1975).

These are all fully illustrated.


Haworthia are confined to Southern Africa and almost entirely South Africa.  One species, H. venosa subsp. tessellata extends into southern Namibia, and another, H. limifolia occurs in Swaziland, Mocambique and South Africa.  Most of the species occur in the Cape Province of South Africa, and again these are found either in the Little Karoo, Southern Cape or Eastern Cape.

Their distribution is very close associated with river drainage systems and topographical features.  Some species are very localised while others such as H. viscosa and H. nigra are more widely spread.  The distribution of the species in the subgenera Haworthia and Hexangulares is very much the same, but the species in the Robustipedunculares are restricted to the Southern Cape and barely occurring in the Little Karoo and southwestern Great Karoo.

Geology and associated edaphic factors are very important to Haworthia species.  Generally they do not occur in the nutrient poor soils (and rocks) of the Cape sandstome formations.  Where they do, there is invariably variation associated with any shift to soils derived from shale sediments or from igneous rocks.  This ecotypic variation can complicate the identification and classification of the species.


Ecology and adaptations
Haworthia are all small succulent plants presumed to be adapted to withstand long periodic droughts.  Because of their size, they do not compete with more vigourous species nor in dense plant communities.  They enhabit rocky or dry situations where the soils are shallow and offering poor support to larger more vigorous plant species.  Where they occur in grassland, it is also in rocky situations which offer them protection against vegetation competition as well as agaist fire and grazing animals.  H. limifolia is occasionally an exception and may occasionally be found in the litter of dry forest floors.

They do not seem to have any particular physical defences against grazing animals and when exposed are readily damaged by such grazing.  Even the hard, sharp leaves of Astroloba species does not prevent them from being damaged by either goats or sheep.  Damage is often done to plants simply by the trampling of passing animals and this particularly applies to those species which form above-ground rosettes.

The saving grace of the plants is their choice of rocky habitats.  This may be either the steep rocky sides of cliffs and river-side rocks, or in the cracks and crevices of other less prominently exposed rocks.  Several species are able to withdraw into the soil by the action of contractile roots, and only the leaf-tips are exposed.  Many of these again, are quite cryptic and are very difficult to find.  Species such as H. cooperi and H. truncata may only be visible during the active growing period or when they flower.

The root systems are significant and many species have large fleshy and even fusiform roots which may serve as a moisture reserve against dry periods.

Spination is a feature of several species and it is assumed that this character offers both protection against sunlight radiation as well as in gaining moisture from dew.  Leaf necrosis is also a feature in at least two species, namely H. semiviva and H. lockwoodii.  Although pronounced in the former species which derives its name from the die-back of the leaf-tips, it is H. lockwoodii which is quite spectacular in the way the dry leaf-tips enclose the plant like dry, white onion-skins.

Protection against irradiation seems to be a major requirement for Haworthia, which has to be balanced against the plants need to store water and yet also obtain energy for photosynthesis.  Thus there are many adaptations of the leaf surfaces.  Translucence, which allows light to enter directly into the inner leaf-tissue is a common feature.  This translucence may be represented by small translucent dots or by translucent areas between leaf-veins.  In several species, the leaves are truncated into a flat-end-area which may be modified to either limit or enhance light penetration.  Tubercles on the leaf-surfaces are also a common feature in several species and it is difficult to suggest what their function actually is.  H. maxima is dark coloured and has prominent white tubercles, while H. marginata inpractically the same kind of habitat, is light coloured and has none.  Similarly in H. scabra, the typical variety is dark and has many concolourous tubercles, while the var. starkiana is smooth surfaced but light coloured and shiny.  The tubercles may thus play a role in heat and light regulation.

Coloration is also strongly affected by exposure to sunlight and a response to exposure is usually the loss of green coloring (loss of chlorophyll) and the development of bright colours in the red-range.

It is seldom that the plants stand alone and exposed to the elements.  Nurse plants and protection against the elements is a vital need for all species.