Turbinicarpus, Rapicactus and Kadenicarpus. Three evolutionary lineages

A shared name in collection history

On older labels, Turbinicarpus subterraneus stands beside T. valdezianus, while T. pseudomacrochele belongs to the same small group of tuberculate cacti. In the current treatment, these are Rapicactus subterraneus, Turbinicarpus valdezianus and Kadenicarpus pseudomacrochele, respectively. The change separates the historical genus into lineages with distinct origins. An old label remains useful as an indication of the name used in a publication or collection record. [Vázquez-Sánchez et al. 2019, pp. 407–409, 412, 416]

In POWO as checked on 10 October 2026, Turbinicarpus [2-WWW] comprises 14 species and an additional three hybrids, Rapicactus [2-WWW] five species, and Kadenicarpus [2-WWW] three species. The 2019 study and the contemporary list differ in scope. The current list also contains combinations accepted later. [2-WWW]

DNA and eighteen structural traits

Vázquez-Sánchez and colleagues compared 66 taxa, including 29 outside historical Turbinicarpus, used to place the group studied on a broader Cacteae tree. They used five plastid DNA regions, matK, rbcL, psbA-trnH, rpl16 and trnL-F. They added 18 structural traits, comprising 13 morphological and five anatomical traits. These data described both externally visible organs and their tissues. [Vázquez-Sánchez et al. 2019, pp. 407–409, Table 2–3]

The tree was reconstructed using Bayesian inference. Rapicactus and Kadenicarpus received support of PP = 1, and Turbinicarpus in the narrow sense PP = 0,89. PP is the posterior probability obtained for a branch under the model and dataset used. Individual relationships within the genus had weaker support. A large group of Turbinicarpus species remained partly unresolved. [Vázquez-Sánchez et al. 2019, pp. 408–409, Fig. 1]

In this tree, Rapicactus was close to Acharagma. Kadenicarpus was the sister lineage to a group containing Ariocarpus and Turbinicarpus. Tree position concerns the 2019 analysis and its sample of plastid regions and structural traits. A shared globose habit did not define a single identical pattern of relationships. [Vázquez-Sánchez et al. 2019, pp. 409, 412]

The seed coat retains small differences

Seed microsculpture describes the surface relief of seed-coat cells. In the Rapicactus studied it was papillate, in Turbinicarpus striate and in Kadenicarpus smooth. These differences become clear at high magnification. Additional information comes from the region of the funicular scar and micropyle, the opening in the ovule integuments. In Rapicactus, these sites are separate, while in the other two genera they form a common complex. [Vázquez-Sánchez et al. 2019, pp. 409, 416–417]

The authors also checked the distribution of these traits on the tree. Papillate microsculpture occurred in Rapicactus and Acharagma but, according to the reconstruction, was acquired independently. Longitudinal fruit opening was shared with further genera. This shows the evolution of individual traits within a broader context. A group can have a recognisable combination of traits even when none is exclusive on its own. [Vázquez-Sánchez et al. 2019, p. 412]

Crystals and fibres in a precisely defined location

De la Rosa-Tilapa and colleagues compared the anatomy of 35 taxa, analysing lower, middle and apical stem regions. They used sections, polarised light and electron microscopy. This allowed the layer containing a particular crystal or fibre to be identified. The hypodermis beneath the stem epidermis contained concentric druses in Rapicactus and prismatic crystals in Turbinicarpus and Kadenicarpus. [De la Rosa-Tilapa et al. 2019, online version, pp. 1–2]

A druse is a cluster of many crystals, while the prismatic form has the elongated geometry of an individual crystal. In Kadenicarpus, xylem fibres appeared in bundles passing through cortical parenchyma. At the same time, the main stem wood of these plants was fibre-free. Rapicactus subterraneus was distinguished by a few fibres in the main wood itself. These two findings refer to different vascular zones. [De la Rosa-Tilapa et al. 2019, online version, pp. 2–4]

Prismatic crystals in the hypodermis are shared by two lineages, so their presence alone does not separate Turbinicarpus and Kadenicarpus. Fibre location and seeds provide additional information. Crystal arrangement in the pith of individual species can differ from that in the hypodermis. Druses were found there in K. pseudomacrochele subsp. minimus. Naming the layer is therefore part of correctly describing a trait. [De la Rosa-Tilapa et al. 2019, De la Rosa-Tilapa et al., online version, p. 2. Vázquez-Sánchez et al., p. 412] [Vázquez-Sánchez et al. 2019, De la Rosa-Tilapa et al., online version, p. 2. Vázquez-Sánchez et al., p. 412]

Shared anatomy and species-specific exceptions

Wood in most of the taxa compared had vessels surrounded by wide-band tracheids. Their walls form broad annular or spiral thickenings. Xylem rays remained unlignified, while the cortex contained storage parenchyma. These traits describe the organisation of a succulent stem in which vascular tissues coexist with a large volume of parenchyma. [De la Rosa-Tilapa et al. 2019, online version, pp. 2–4]

Exceptions crossed genus boundaries. In Turbinicarpus alonsoi, T. lophophoroides and Rapicactus subterraneus, small cortical bundles were amphicribral, with xylem surrounded by phloem. Most other taxa had collateral bundles in this zone, with xylem and phloem side by side. This comparison shows the range of anatomical variation rather than a single pattern across the entire former group. [De la Rosa-Tilapa et al. 2019, online version, p. 2, Fig. 3B]

Epidermis and cork in Rapicactus

In the 2019 anatomical sample, the outer wall of Rapicactus epidermal cells was on average thicker than in the other two lineages compared. The mean for the material studied was 3,06 ± 1,39 µm, compared with 1,66 ± 0,69 µm in Turbinicarpus and 1,17 ± 0,46 µm in Kadenicarpus. This is a feature of the outer epidermal wall, separate from cuticle thickness. A smooth cuticle and elongated epidermal cells were shown in Rapicactus mandragora (Turbinicarpus mandragora in the source). [De la Rosa-Tilapa et al. 2019, online version, p. 2, Fig. 1B, 2B]

Periderm formed on older portions, a secondary covering produced by phellogen originating from the epidermis. In the Rapicactus studied, thin-walled and thick-walled cork cells formed alternating layers, except in R. zaragozae. The same species was distinguished by a single hypodermal layer. R. beguinii, in turn, had small vascular bundles in the pith, a structure recorded in this study in only two taxa of historical Turbinicarpus s.l. Species differences complement traits shared by the genus. [De la Rosa-Tilapa et al. 2019, online version, pp. 2–4]

Two lineages in a mountain landscape

Donati and colleagues collected 743 population records for 47 taxa, combining data from 12 herbaria, field documentation and field-number databases. Their 2017 analysis distinguished Rapicactus, with 152 records, from Turbinicarpus as then circumscribed, with 591. The second group also included plants later separated as Kadenicarpus. Distribution data had georeferencing uncertainty below 5 km. Site environments were compared using regional climate, soil and terrain maps. [Donati et al. 2017, Donati et al., online version, pp. 2–3. Vázquez-Sánchez et al., pp. 407, 416–420][Vázquez-Sánchez et al. 2019, Donati et al., online version, pp. 2–3. Vázquez-Sánchez et al., pp. 407, 416–420]

The diameter comparison covered 230 populations, 59 of which belonged to Rapicactus. Measurements were taken at the end of the dry season, excluding spines. In Rapicactus, diameter increased northwards. The relationship with latitude had p = 1,47 × 10⁻⁷. Site elevation, mean annual temperature, total rainfall and evapotranspiration were not significant in this model. The full model explained about half the variation in diameter, R² = 0,5048. [Donati et al. 2017, online version, pp. 3, 7, Table 3]

The authors described R. beguinii sites in cooler valleys, beneath pines, beside watercourses and on gypsum slopes with shrubs. They proposed that slower evaporation and greater water availability in these microenvironments may help explain the greater diameter in some populations. Diameter and location measurements demonstrated a geographic relationship. Explanation through water balance remains a hypothesis complementing the field data. [Donati et al. 2017, online version, p. 8]

The stem-structure study, DNA analysis and habitat comparison therefore provide information at different levels. The first describes tissue locations, the second their distribution on the phylogenetic tree, and the third current relationships in the landscape. Site elevation alone does not determine a safe overwintering temperature in a pot. Such a threshold would require data on cooling duration, moisture, acclimation and plant origin, which this geographic analysis did not investigate. [Vázquez-Sánchez et al. 2019, scope of the methods of the three studies discussed][De la Rosa-Tilapa et al. 2019, scope of the methods of the three studies discussed][Donati et al. 2017, scope of the methods of the three studies discussed]

Species boundaries remain a further question

The Turbinicarpus schmiedickeanus, T. saueri, T. viereckii and Kadenicarpus pseudomacrochele complexes did not form groups consistent with all subspecies boundaries recognised at the time. Some tree connections had weak support, and the material covered a limited number of individuals and populations. The authors identified a need for further sampling. Separating genera and resolving relationships within a species complex therefore involve different levels of detail. [Vázquez-Sánchez et al. 2019, pp. 414–415]

When describing an older experiment, retaining the authors' name connects the result with its material. A micropropagation result published for Turbinicarpus subterraneus therefore concerns today's Rapicactus subterraneus. In turn, spine traits described under Turbinicarpus flaviflorus now belong to T. schmiedickeanus subsp. flaviflorus. The current name, historical identification in parentheses and sample origin allow this literature to be read alongside today's atlas. [Vázquez-Sánchez et al. 2019] [Mosco 2009] [2-WWW]