reprinted from the
San Francisco Succulent and Cactus Society Newsletter

For connoisseurs of the succulent Aizoaceae, or "mesembs," there are several compellingly odd, but difficult to cultivate species and genera that have taken on an almost mythical quality. These vegetable Maltese Falcons include Muiria hortenseae, Conophytum burgeri and Didymaotus lapidiformis, but surely the strangest and most horticulturally recalcitrant of the bunch is Dactylopsis digitata, the finger mesemb. So, it came as something of a shock to the community of succulent-plant enthusiasts when Dactylopsis was subsumed into Phyllobolus (Gerbaulet, 1997). Many of us never did quite get around to changing our labels to Phyllobolus digitatus, though, and in this article I will defend the continued use of Dactylopsis, on grounds other than that of simple inertia.

Before getting into my defense of Dactylopsis, I should summarize the arguments that were used to justify the transfer to Phyllobolus. The reasoning was based on the discipline of cladistics, which is a method of discovering relationships among organisms, and reconstructing evolutionary trees (cladograms). In cladistics, species are grouped together if they have synapomorphies, which are shared derived characteristics not present in other, putatively related groups. One goal of a cladistic analysis is to find "monophyletic" groups – groups of species that include all of the descendants of a common ancestor, and exclude representatives from all other evolutionary lineages. It is the over-whelming consensus among modern biologists that cladistics is the best way to determine the evolutionary history of living things, and that classification schemes should only recognize taxa that are monophyletic.

No-one, including Gerbaulet, contests the idea that Dactylopsis is a monophyletic entity. Depending upon whom you talk to, the group contains either two species (D. digitata, the large-leaved southern form, and D. littlewoodii, the small-leaved northern form)(van Jaarsveld and Pienaar, 2000), or a single, variable species (Gerbaulet, 2001). The two forms are mainly distinguished by size, and share synapomorphies not seen in any other genus in their subfamily (the Mesembryanthemoideae): large, hyper-succulent cylindrical leaves, and tiny white flowers with filamentous petals that tend to keep their form as they dry out. The reason why Dactylopsis was sunk by Gerbaulet is that she claimed that it had other synapomorphies in common with Phyllobolus, in particular the dwarf, tuberous relatives of P. resurgens. Although Gerbaulet did not present a formal cladistic analysis, she contended that Dactylopsis was nested within the evolutionary branch (clade) containing P. resurgens, and thus that Phyllobolus would not be a monophyletic genus if Dactylopsis was maintained as a separate entity (Gerbaulet, 1997).

Based upon gross morphology, it is difficult to see what characters could possibly link Dactylopsis to Phyllobolus: the succulent, persistent green stems, smooth seeds, and smooth, waxy leaves of Dactylopsis contrast strongly with the ephemeral stems, rough seeds and warty, bladder-cell covered leaves of the P. resurgens group. The strength of Gerbaulet’s (1997) argument lies instead with anatomical characteristics. The stems and roots of Dactylopsis, and the tubers of geophytic Phyllobolus species, thicken via the production of successive rings of vascular tissue that consists mostly of spongy, water-storing parenchyma cells. These organs also become covered by a well-developed periderm (an outer layer of protective cork cells) in both genera. However, the status of these characters as synapomorphies unique to these two genera is dubious. All perennial Mesembryanthemoideae that I have examined thicken their stems and roots via successive cambia giving rise to concentric rings of vascular tissue, which tends to be mostly parenchyma in organs that do not need to support themselves mechanically, such as the stems of a cushion plant like Dactylopsis or the subterranean tubers of Phyllobolus. A periderm, too, seems to be developed to a greater or lesser extent in all of the genera of perennial mesembs that show secondary thickening growth in roots or stems. I would suggest that the anatomical characters used to sink Dactylopsis into Phyllobolus are not synapomorphies of these genera at all, but ancestral characters (symplesiomorphies) present in a wide variety of genera in the subfamily, and thus uninformative for determining evolutionary relationships.

If the link between Dactylopsis and Phyllobolus is tenuous, what are the closest relatives (sister groups) to Dactylopsis? The extreme morphological reduction of dwarf succulents like Dactylopsis can make the task of identifying synapomorphies difficult. However, I have found several characters that indicate a strong link to Aspazoma, an obscure, monotypic genus of small shrubs (Opel, 2002). A relationship between Dactylopsis and Aspazoma was first suggested almost half a century ago by Schwantes (1957), who placed both genera in the same subtribe.

The most suggestive characteristics linking Dactylopsis and Aspazoma involve the structure of the leaves. In both genera, the epidermal bladder cells that are present in most Mesembryanthemoideae have been lost, and the leaves are quite smooth and waxy. Both genera have sheathing leaf bases that surround the stem, a character that does not occur in any other genus in the subfamily. Moreover, the function of the sheathing leaf bases is the same in both genera: the leaves dry up to form a protective tunic that covers the succulent stems during the summer dormant period. This mode of dormancy is unique in the family Aizoaceae, though many genera (e.g. Conophytum) show an analogous strategy, in which succulent leaves (rather than stems) are protected by a papery tunic of old leaves during the dry season.

Another character that suggests that Dactylopsis is related to Aspazoma, and not Phyllobolus, is their succulent stems, which have a layer of green chlorenchyma in their cortex, and prominent bladder cells on their epidermis. The stems remain succulent for two growing seasons, and then lose their green cortex and develop a periderm. Nothing similar is known from Phyllobolus, though the same type of stem morphology and development occurs in the genera that are commonly considered to be related to Aspazoma: Psilocaulon and Brownanthus (secondary growth in the stems of these shrubby genera is woody, consisting mainly of fibers and tracheids, whereas the cambia of Dactylopsis mostly produce soft parenchyma). In addition, Dactylopsis and Aspazoma both have smooth-coated seeds, while seed coats in Phyllobolus tend to be rough. In terms of overall shape, the elongated pyriform seeds of Dactylopsis have little in common with either the robust, D-shaped seeds of Phyllobolus or the large, flattened seeds of Aspazoma.

Given the available morphological evidence, it seems most reasonable to maintain Dactylopsis as a separate genus, closely related to Aspazoma, but easily distinguished by its extreme succulence and unusual floral structure. Further research is required to determine how the hypothesized Dactylopsis/Aspazoma clade falls out with respect to other allied genera, such as Brownanthus, and it is probably safe to assume that more rearrangements will need to be made before the taxonomy of this group settles into a stable state.