Exostema in its current, narrowed circumscription (Paudyal et al., 2018) includes 8 species of shrubs and small trees. Most of these are characterized by a normally branched unarmed habit with chartaceous leaves, except Exostema spinosum has small leathery leaves clustered on short-shoots and spinescent stems. The flowers are generally 5-merous, nocturnal, and axillary on slender peduncles and pedicels, and solitary or in cymose or fasciculate groups of 2-5. The corollas are showy and white, with a slender tube that is about as long as the narrow lobes. The anthers are slender and exserted, and the stigmas are positioned among the stamens. The fruits are capsular and rather small, with numerous flattened, circular, marginally winged seeds. The flowers are fragrant at night and presumably pollinated by hawkmoths. The plants contain bitter compounds, and are locally called "quina" and used as a remedy for fevers. Exostema caribaeum is the most widepread and by far the most commonly collected species.
The history of the circumscription and systematics of Exostema are unusually complex. It is a relatively old genus that is similar to Cinchona, so many of its species were originally described (often more than once) in that genus. Soon after its description Exostema became circumscribed rather broadly, to include what are now separated as several distinct genera (Paudyal et al., 2018): Exostema (8 species), Motleyothamnus (1 species), Syringantha (1 species), Acuneaenthus (1 species), Adolphoduckea (1 species), and Solenandra. The current taxonomy separates most of the species of traditional Exostema into two genera, Exostema and Solenandra, with most of them in Solenandra.These two groups are supported by morphological and molecular characters, but their nomenclature has been problematic. The nomenclatural and taxonomic situation of Exostema is detailed below for reference.
Exostema includes a relatively large number of species with morphologically variant forms in the Antilles, especially Cuba, similarly to some other Antillean genera that inhabit dry vegetation on limestone and serpentine (e.g., Scolosanthus, Schmidtottia, Phialanathus). Most of the species are generally similar, with regularly developed, non-spiny stems and thin-textured medium-sized leaaves. Only one species is markedly xeromorphic: Exostema spinosum has stems bearing axillary short-shoots, regularly developed spinescent stems, and thick-textured small leaves clustered together.
The species-level taxonomy of Exostema and Solenandra are closely intertwined. Taxonomy of this group has varied from a more traditional view of rather broadly circumscribed, geographically coherent species based on extenstive field work (e.g., McDowell, 1995), to a more detailed view of numerous infraspecific taxa diagnosed by vegetative details that sometimes are variable (Borhidi, 1989; Borhidi et al., 2017, 2018; the overlapping variation in diagnostic features was noted by these authors themselves). Additionally, older taxonomies of these plants often differed markedly in species circumscription from each other (e.g., Grisebach's taxonomic scheme, as noted by Urban at various times) as well as from our modern view. Comparison among authorsi s further compliated for some modern taxonomic treatments by mixed nomenclatural citations, with formally published names cited together with names that are "sensu" literature references. In particular the treatments by Borhidi (1989) and Borhidi et al. (2018), contain several confusions and inaccurate citations, and some intended new combinations here are inaccurately based on names that were not actually published (e.g., "Exostema valenzuelae A. Rich.", which was not accepted in its original publication and so is invalid; "Exostema parviflorum A. Rich.", which was not published in this work and an explicit reference there to Bonpland's name). The species taxonomy here follows McDowell (1995).
Exostema is similar to Solenandra, which differs in its terminal inflorescences. Exostema is similar vegetatively to Coutarea, and these are often sympatric; Coutarea differs in its terminal inflorescences, funnelform pink corollas with the lobes much shorter than the tube, and larger capsules that are markedly flattened laterally. Some species of Scolosanthus also have branched and/or leaf-bearing spines and are similar to Exostema spinosa; Scolosanthus differs in its terminal inflorescences and drupaceous fruits. This genus has sometimes been incorrectly spelled "Exostemma".
McDowell (1995, 1996; McDowell & Bremer, 1998; McDowell et al., 2003) studied the systematics of Exostema in its traditional circumscription in excellent detail, including morphological and molecular analysis and extensive field work. Relatively early in the use of these techniques, McDowell & Bremer (1998) studied this group with cladistic methods and a combined morphological and molecular data and found three well supported clades and the two odd South American species, Exostema maynense and Exostema corymbosum, basal to the rest of Exostema; they included one outgroup, Coutarea hexandra. McDowell & Bremer regarded their three Exostema clades taxonomically as sections, Sect. Exostema, Sect. Pitonia, and Sect. Brachyantha, and documented repeated habitat shifts within each groups, in contrast to apparently a single characteristic pollination mode in each. McDowell (1996) also used morphological and molecular data to separate one species from Exostema into Syringantha. McDowell et al. (2003) then studied this group again with additional molecular sequence data and outgroups but more limited sampling of Exostema species, and found a more complex result, with Exostema paraphyletic with respect to several other genera [similarly to the results of Manns & Bremer (2010) and Paudyal et al. (2018)], and the three sectional clades no longer clearly supported. They considered their results preliminary in terms of making taxonomic changes due to limited their limited sampling and sequence data.
Based on the contemporaneous, unpublished molecular results of Johan Rova in his 1999 dissertation together with McDowell & Bremer's (1998) characterizations of their three Exostema clades, Borhidi (2002) separated Solenandra, which had long been synonymized with Exostema. Borhidi circumscribed Solenandra to comprise two of McDowell & Bremer's (1998) sections, and distinguished these two genera by various morphological features (2002: Table 1) including inflorescence position, flower size and biology (nocturnal and fragrant in Exostema vs. supposedly diurnal and lacking odor in Solenandra), ovules erect in Exostema vs. pendulous in Solenandra, stigma held above the stamens in Exostema vs. not above them in Solenandra, different pollen sizess, different capsule sizes, and some other features Borhidi (2010) separated Exostema and Solenandra in his Rubiaceae treatment for Mexico, but later again synonymized these, without comment, in his Rubiaceae treatment for Cuba (Borhidi, 2017). Shortly afterward Borhidi et al. (2018) again separated Solenandra and Exostema in Cuba based on Paudyal et al.'s (2018) results, and explained there that Solenandra had been separated by Borhidi earlier based on Borhidi (2002) Manns & Bremer (2010), but had then been synonymized following its synonymization by Lorence et al. (2012). [Note: The treatment of Lorence et al. was published in 2012, but preparation of the flora volume was a complex task that took significant time, and the Rubiaceae treatment was finalized in 2009-2010 before the publication of Manns & Bremer's analysis. - CMT]
More recently, Greuter & Rankin-Rodríguez (2021b) reviewed the systematics of this group again, using the analysis previously published by Paudyal et al. (2018), and reached a different conclusion. They recognized one broadly circumscribed genus that comprised what Paudyal et al. separated as Exostema, Solenandra, Hintonia, several new segregate genera, and Coutarea. Based on their view of this group as centered in Cuba and Exostema caribaeum as its presumed most well known species, Greuter & Rankin-Rodríguez (2021) proposed conservation of the name Exostema over the oldest name, Coutarea, for this entire group. However, Coutarea hexandra is much more widely distributed naturally and in cultivation, and this genus could be argued as a better name for this group is expanded in this new circumscription. If Exostema is eventually conserved against Coutarea, it will only have priority when these two genera are combined; when Paudyal et al.'s taxonomy is used, Coutarea and Exostema are separated so the conservation is not applicable.
Nomenclatural Discussion. The authorship and type species of Exostema have been cited differently by various authors, and has been variously confused but the correct authorship and typification are significant now that Solenandra has been separated.
Exostema was described in Humboldt & Bonpland's Plantae Aequinoctiales, and it has sometimes been attributed to both of these authors but the authorship of the names there is solely by Bonpland. Exostema was described there as based on generic characters and notes by L.C. Richard on herbarium specimens, and this name has sometimes been attributed to "L.C. Richard ex Bonpland"; however, the name "Exostema" was not attributed to Richard by Bonpland, only generic characters, so Richard cannot be credited with authorship this name. Exostema was described here with two species, Exostema parviflorum and Exostema peruvianum; these were there called "parviflora" and "peruviana" because Bonpland treated the genus as feminine, but it is today regarded as neuter with the corresponding change needed in the spelling of the species epithets [Art. 62.2 (c)]. At the end of his treatment Bonpland (p. 135) noted that his genus was the same as Cinchona subg. Exostema Pers., and he was raising Persoon's group to genus rank. This makes Persoon's group the original name of this genus, so Exostema was not a new group first recognized by Bonpland. In a footnote in French on this same page, Bonpland also noted that Candolle had recently presented a talk in Paris with the the same conclusion, treating Persoon's group as a genus with the same name. This footnote seems to have been added after Bonpland wrote his main treatment, in Latin, so Candolle does not seem to deserve any authorship credit for this genus name.
For a long time Exostema was regarded as created and solely authored by Bonpland, and Britton & Milllspaugh (1920) designated as the lectotype of this genus one of Bonpland's two species, Exostema parviflorum; Bonpland's other original species was Exostema corymbosum, iwhich s now separated as Motleyothamnus corymbosus. This authorship and leccttype were accepted by most subsequent authors (e.g., Hooker, 1873; Standley, 1938; Long & Lakela, 1976; Dwyer, 1980; Howard, 1989; Borhidi, 1989). In the mid-20th century. however, several other authors noted (with minimal or no explanation) the citation of Persoon's Cinchona subg. Exostema in the protologue, so that Persoon's name was actually the basionym of Exostema (Adams, 1975; Rogers, 1987; Correll & Correll, 1989; Wunderlin 1998; McDowell 1995, 1996; Lorence et al. 2012). The origin of this correction has not been found in this review.
The nomenclature of Exostema was next studied critically by Rogers (1987), who did expliclity cite and lectotypify Cinchona subg. Exostema. As the basionym of Exostema, the lectotype of Cinchona subg. Exostema is the type of Exostema. Persoon described his subgenus with eight species, and Rogers lectotypified it on Cinchona caribaeum. After Bonpland's description of his two new species. Persoon's species of Cinchona subg. Exostema were transferred into Exostema by Roemer & Schultes. The inaccurate citation of Exostema parviflorum as the type of Exostema by some authors has complicated the taxonomy of this group, and in particular the infrageneric classification because different sections have been considered the typical Sect. Exostema by different authors. In particular, Boihidi (1989) considerd Exostema parviflorum to be the type of the genus, so his "Sect. Exostema" now corresponds mostly to plants now included in Solenandra, and he inaccruately re-named the typical section of Exostema as Sect. Oligantha. Bothidi's "Sect. Exostema" mostly corresponds to McDowell's (1995) Sect. Brachyantha, but these are nomenclaturally distinct because McDowell lectotypified Sect. Brachyantha on Exostema corymbosum, so this section now is a synonym of Motleyothamnus.
The other named infrageneric groups, Sect. Ptionia, Sect. Polyphyllae, Sect. Longiflorae, and Sect. Floribundae (which has the same type as Sect. Pitonia) belong to Solendandra, except Sect. Pseudostemma included a heterogeneous group of species that are now classified in several other genera. The circumscription and identity of Candolle's Sect.Pitonia were emended by McDowell (1996), and this has created some later confusion. Borhidi et al. (2018: 303) noted accurately that Candolle originally included the type of Exostema in this section, making it the typical section and so properly called Sect. Exostema, but McDowell explicitly changed the circumscription of this section to exclude several of its species and was the first to lectotypify it. McDowell excluded from Sect. Pitonia both the nomenclatural type of the genus, Exostema caribaeum, so it no longer corresponds to the typical species, and he also exlcuced the mistaken type of the genus, Exostema parviflorum, so this can't be considered the typical section even when the type of the genus is confused. McDowell lectotypified Sect. Pitonia on Exostema sanctae-luciae, which was also the type of Borhidi's Sect. Floribundae. Further confusion arose when Borhidi et al. (2018) also stated that Candolle's Sect. Pitonia, described in 1830, included the type of Solenandra, but that is not accurate because Solenandra and its type species were not described until 1873 (this seems to be partly due to a confusion of a "sensu" citation, of the actual description of Solenandra ixoroides with the literature citation of Richard's inaccurate use in his 1845 flora of the name Exostema parviflorum for the Cuban plants that are now seprated in Solenandra ixofoides). Another persistent nomenclatural confusion in this group has been the name Exostema valenzuelae, which is an illegitimate name that has been incorrectly accepted by various authors; this name corresponds to Solenandra ixoroides, for more details see the dicussion on the web pages for that species.
Author: C.M. Taylor.
The content of this web page was last revised on 20 December 2021.
Taylor web page: http://www.mobot.org/MOBOT/Research/curators/taylor.shtml