Pseudorhipsalis, Britton & Rose, 1923

Korotkova, Nadja, Borsch, Thomas & Arias, Salvador, 2017, A phylogenetic framework for the Hylocereeae (Cactaceae) and implications for the circumscription of the genera, Phytotaxa 327 (1), pp. 1-46 : 12-13

The Pseudorhipsalis clade and the position of P. ramulosa

All species and subspecies were sampled in our study and Pseudorhipsalis is found as polyphyletic. A Pseudorhipsalis clade including the nomenclatural type P.alata (Swartz) Britton & Rose receives maximum support in all analyses, but P. ramulosa is not part of that clade, its exact position within the phyllocactoid clade remains unresolved. Pseudorhipsalis amazonica (K. Schumann) Ralf Bauer is resolved as part of the Pseudorhipsalis clade with maximal support. This species is unusual as it is the only cactus to have blue tepals. Schumann regarded this as so unique that he placed it in an own monotypic genus Wittia K. Schumann ( Schumann 1903) . A new generic name Wittiocactus Rauschert (1982: 558) was published later to replace Schumann’s illegitimate name Wittia because Wittia Pantocsek (1889: 110) is a genus of Bacillariophyta. Wittiocactus was then included in Disocactus ( Hunt & Taylor 1990) , yet still unresolved. Bauer (2002) moved it to Pseudorhipsalis because of many common characters, e.g. seedling morphology. The clade of P. lankesteri (Kimnach) Barthlott , P. himantoclada (Roland-Gosselin) Britton & Rose and P. alata (Swartz) Britton & Rose is supported by 1 PP but not found by the MP analysis. These three species are unique by forming pollen tetrads, a character found in no other Cactaceae species ( Barthlott 1975, Leuenberger 1976).

Pseudorhipsalis was first described by Britton & Rose (1923: 213) with P. alata and P. himantoclada , two species that had been previously placed in Rhipsalis , and remained in this circumscription until the 1990ies. Subsequently two different generic concepts were suggested: Kimnach (1993) included Pseudorhipsalis in Disocactus , while Barthlott (1991) treated Pseudorhipsalis as distinct genus and expanded it to include six further Central American species showing small flowers. Barthlott’s concept is largely confirmed by the plastid tree obtained here, while our data show no indication for merging Pseudorhipsalis into Disocactus .

Pseudorhipsalis ramulosa is resolved outside the Pseudorhipsalis clade and found in a polytomy with Disocactus and Epiphyllum . We tested alternative topologies for P. ramulosa being sister to either Epiphyllum , Disocactus or the rest of Pseudorhipsalis . The tree with P. ramulosa unresolved was significantly favoured over the alternative topologies tested (online supplement S3). The affinities of P. ramulosa were unclear for a long time. It was originally described as Cereus ramulosus Salm-Dyck (1834: 340) , then treated as Rhipsalis ramulosa (Salm-Dyck) Pfeiffer , and still included in Rhipsalis by Britton & Rose (1920) after they had established the genus Pseudorhipsalis . They mentioned its similarity with Pseudorhipsalis but did not include it therein ( Britton & Rose 1920). Kimnach (1961) included P. ramulosa in Disocactus whereas Barthlott (1991) combined it into Pseudorhipsalis . This taxon is generally similar to other Pseudorhipsalis species in vegetative and floral morphology, but there are also some differences. The floral tube is less pronounced in contrast to the other Pseudorhipsalis species. P. ramulosa sets fruits without pollination, while pollination is necessary in the other species of Pseudorhipsalis ( Kimnach 1961) . It is also widely distributed in tropical America, from Mexico and Central America to western South America and Antilles ( Haiti, Jamaica) while the other Pseudorhipsalis species are much more restricted in their distribution, most are found only in Costa Rica and Panama ( Barthlott et al. 2015). It has been suggested that the self-fertility, and also the berry-like sticky fruits adapted for bird-dispersal could explain the large distribution of P. ramulosa ( Kimnach 1961) . It could be a lineage within Hylocereeae that was especially successful in colonizing a comparatively large area by adapting to selfing and bird-dispersal. This is very similar to R. baccifera (Solander) Stearn , an epiphytic cactus with likewise small white flowers, self-pollination and sticky, bird-dispersed fruits that has the largest distribution of all cacti ( Barthlott 1983, Barthlott et al. 2015). This independently evolved evolutionary strategy of R. baccifera and P. ramulosa would be another example for marked convergences in cacti.

To continue in recognizing Pseudorhipsalis ramulosa under Pseudorhipsalis and thus accepting a polyphyletic Pseudorhipsalis is not a good option in our opinion. We therefore propose to establish a new monotypic genus Kimnachia gen. nov. for P. ramulosa (see the Taxonomic synopsis).