Chimarrhis

  • Authority

    Delprete, Piero G. 1999. Rondeletieae (Rubiaceae). Part I. Fl. Neotrop. Monogr. 77: 1-226. (Published by NYBG Press)

  • Family

    Rubiaceae

  • Scientific Name

    Chimarrhis

  • Type

    Type species. Chimarrhis cymosa Jacquin.

  • Synonyms

    Pseudochimarrhis, Pseudochimarrhis turbinata (DC.) Ducke, Chimarrhis cymosa Jacq.

  • Description

    Genus Description - Trees, often tall canopy trees, bole irregular, with or without buttresses; buttresses sometimes very large, especially in the Amazonian species; bark smooth to deeply fissured; wood yellow, very hard. Stipules interpetiolar, above leaf attachment, free or connate at base, adnate to the petioles, persistent or caducous, leaving a scar encircling the stem above the node. Leaves elliptic to ovate to obovate, usually acuminate at apex; not pellucid punctate; petioles often thickened at base but not pulvinate; domatia a tuft of hairs (tuft-pit domatia in C. jamaicensis) or absent. Inflorescences subterminal (axillary on the terminal nodes), corymbose, with opposite to subopposite lateral branches terminating in cymules. Flowers protogynous, sessile to pedicellate, very fragrant; hypanthium obconical to turbinate. Calyx extremely reduced, truncate or with barely distinguishable lobes, persistent; in few species (e.g., C. brevipes and C. gentryana) in some of the flowers one calyx lobe expanded into a white to greenish white calycophyll. Corolla funnelform, deeply lobed, with reflexed lobes, white, greenish white or cream-white; tube funnelform to ciatiform; with a ring of pubescent hairs inside on the tube or on the base of the lobes; lobes (4-)5(-6); aestivation narrowly imbricate, superfidaily resembling valvate aestivation. Stamens 5, alternate to the petals, totally or partially exserted; filaments attached on the upper part of the tube, slender, basally flattened, basally pilose to villous; anthers elliptic, dorsifixed, dehiscing by lateral slits. Pollen tricolporate, exine densely thick-reticulate. Style exserted; style branches protruding above corolla before anthesis, very short, rounded to ovate, reflexed at maturity. Ovary 2-celled, turbinate to obovate, placentation peltate on the septum; ovules many in each locule, vertically inserted; immature fruits green and semi-carnose when fresh. Capsules woody, globose, oblong or obovoid; dehiscing septicidally, disk-loculicidal dehiscence present in old capsules of some species. Seeds suborbicular, with central hilum and narrow concentric wing, margin undulate.

  • Discussion

    Chimarrhis is unique in the Rubiaceae in its axillary-subterminal inflorescences, capsules with peltate placentation, and minute seeds umbilically attached and with minute concentric wings. The seeds of at least one species (C. barbata) have a tendency to develop bipolar wings, resembling those of the tribes Cinchoneae and Calicophylleae. All the species of Chimarrhis are trees, varying from mid-canopy to upper-canopy levels. Mature individuals of the genus usually develop buttresses, especially those growing in the Amazon basin (see Figs. 68A, 71 A). The bark of these trees is smooth and grayish in young individuals; and fibrous, deeply fissured, reddish, and falling off in small irregular pieces in mature individuals (see Figs. 67C,D, 69C,D, 76B).

    The tree architecture of this genus has been studied in only a few species. Chimarrhis microcarpa (Halle et al., 1978), C. glabriflora (pers, obs.), and C. hookeri (pers, obs.; Fig. 68A) display the "Aubreville Model (Terminalia branding)," whereas C. cymosa (Halle et al., 1978) develops the "Petit’s Model," according to the architectural classification of Halle et al. (1978).

    Leaf domatia were found to be a reliable taxonomic character, variable within certain species or consistent within others (Fig. 58A-L; see Domatia, above). Chimarrhis jamaicensis and C. cubensis can be recognized solely by their characteristic domatia (Fig. 58B,C), while in the species with glabrous blades (C. latifolia, C. brevipes, C. gentry ana, and C. barbata) domatia were not found.

    Several species have been reported to be either heterostylous or dioecious (Grisebach, 1862; Urban, 1899), but all Chimarrhis species have strongly protogynous flowers, which are neither dioecious nor heterostylous. Flower anthesis in this genus starts with the protrusion of the style above the still-closed corolla, followed by the opening of the stigmas (resembling two lips exserted beyond the corolla, Fig. 4J), which then become receptive. Shortly after the corolla opens, the style is abscissed, the filaments elongate, and the anthers are exserted well above the corolla (Fig. 4K), becoming functional and opening by lateral slits. The flowers of this genus are minute (a few millimeters long), cream-white, and sweet-fragrant.

    The corollas of Chimarrhis have been described as having valvate aestivation, and for this reason this genus was placed in the Condamineeae sensu Hooker. Close observation of the corollas of Chimarrhis revealed that the aestivation is instead narrowly imbricate, superficially resembling valvate aestivation. More specifically, the tips of the lobes do not overlap (valvate), but the lateral margins of the petals obviously do overlap (imbricate) in bud and in early stages of anthesis.

    Chimarrhis is often confused with Calycophyllum (Calycophylleae), but the latter has terminal inflorescences (vs. lateral subterminal), seeds with strongly bipolar wings, smooth thin bark exfoliating in long vertical strips, and white wood (vs. yellow in Chimarrhis). Chimarrhis is also often confused with Bathysa (Rondeletieae) and Simira (Rondeletieae), but the latter two differ from the previous in having terminal inflorescences and different seed morphology. Simira differs from Chimarrhis in having larger seeds, horizontally inserted, laterally attached, with well-developed lateral-orbicular wings, and reddish wood. Bathysa is similar to Chimarrhis in having small cream-white fragrant flowers, small fruits, minute seeds, and yellowish wood; the previous differing from the latter in having floral aestivation strongly imbricate, roundish petals with irregular margins (vs. narrowly triangular with entire margins in Chimarrhis), and horizontal seeds, laterally attached, and minute lateral wing (vs. vertical seeds, ventrally attached, minute concentric wing peltate to a central hilum).

    After careful observation, I concluded that the species of chimarrhis can be divided into two easily recognizable and probably monophyletic groups, using the persistence and general morphology of their stipules, and their geographic distribution: subgen. Chimarrhis, with readily caducous stipules; and subgen. Pseudochimarrhis (Ducke) Delprete, with persistent stipules. The second subgenus is a new status for the genus Pseudochimarrhis Ducke (1925), which he separated from Chimarrhis and transferred to the Cinchoneae. To the two species already included in subgen. Pseudochimarrhis (C. turbinata and C. barbata) I add C. duckeana (described below), C. brevipes, and C. gentryana, all with persistent deltoid stipules. Subgenus Pseudochimarrhis occurs mostly in the lowlands of the Amazon basin, while the subgen. Chimarrhis is distributed in the Antilles, in Central America, the Andean slopes, and western Amazonian rain forests.

    Chimarrhis is here recognized as a genus of 14 species (9 in subgen. Chimarrhis, and 5 in subgen. Pseudochimarrhis).

    Taxonomic History

    Chimarrhis was founded by Jacquin (1763) with the description of C. cymosa from Martinique. Jacquin reported it to be locally called “Bois de riviere,” and derived its Latin name from the Greek [Greek letters] (chimarro = torrent), because it was originally encountered in the proximity of torrents. He also reported Chimarrhis as having one seed per locule, an error that was reproduced by Poiret (1811).

    Achille Richard (1830) placed Chimarrhis cymosa (as "Chimarrhis corymbosa [hort, var.] Jacquin") in synonymy under Macrocnemum, but admitted that the genera were poorly known to him. He correctly noted that the capsules of Chimarrhis are polyspermous and septicidal, as in M. candidissimum (= Calycophyllum candidissimum), M. jamaicense, and M. tetrandrum. Because he erroneously reported C. cymosa as C. corymbosa, when including this species into Macrocnemum, the combination M. corymbosum (= Condaminea corymbosa) already existed, and he had to establish the new name M. longifolium (based on Jacquin’s collection from Martinique).

    De Candolle (1830) recognized Chimarrhis as a good genus, placing Macrocnemum longifolium (and Chimarrhis corymbosa) under C. cymosa, and describing C. turbinata on the basis of material collected by Patris in Cayenne (French Guiana). He placed Chimarrhis in the tribe Hedyotideae, subtribe Rondeletiinae (as "Rondeletieae").

    Joseph Dalton Hooker (1873) attributed four species to Chimarrhis: two from the West Indies (but only one species was described at that time!), the third species C. turbinata, and the fourth from a Spruce’s collection (Spruce 4930) from Tarapoto, Peru, but without giving it a name. He placed Chimarrhis in his newly founded tribe Condamineeae, subtribe Condamineinae (as "Eucondamineae").

    Baillon (1880) placed Chimarrhis in his "Cinchona Series" (or Cinchoneae). He reported it "exceptional in this group as it would be in any other in which it might be placed," later adding that its “stamens are of two kinds according as we examine this or that flower with filaments short or sometimes long” (probably suggesting dimorphic flowers, which do not exist in any Chimarrhis species), and erroneously reduced Sprucea (= Simira) and Sickingia (= Simira) to synonymy under Chimarrhis.

    Schumann (1889,1891) reported Chimarrhis to be a genus of three species, recognizing C. cymosa, C. turbinata DC., and describing C. hookeri (Schumann, 1889) from Spruce 4930, previously pointed out by Hooker (1873). Schumann (1889, 1891) placed Chimarrhis in the Condamineeae.

    Urban (1899) subdivided C. cymosa into three subspecies: subsp, genuina from the Lesser Antilles; subsp, jamaicensis from material collected in Jamaica; and subsp, microcarpa from material collected in Cuba (Wright 1262).

    Schumann and Krause (1908a) (four years after Schumann’s death) described Chimarrhis dioica as a species exceptional to Chimarrhis in having dioecious flowers. This species was later chosen by Steyermark (1963) as the type species of his new genus Dioicodendron.

    Standley (1918) did not recognized Urban’s (1899) subspecies and reduced them to synonymy (along with Macrocnemum longifolium) under C. cymosa. In addition, he transferred Rustia ferruginea Standi. (= Bathysa sp.) to Chimarrhis. Later, Standley (1926) described C. microcarpa based on specimens collected in Trinidad, and C. pittieri (Standley, 1930b) based on specimens collected in Venezuela. Standley (1931c, 1936) merely listed C. dioica (= Dioicodendron dioicum) and C. pittieri (= Bathysa pittieri), maintaining Chimarrhis in the Condamineeae (sensu Hooker).

    Adolpho Ducke (1876-1958) was veiy influential in understanding the phenology and clarifying the taxonomy of the Amazonian species of Chimarrhis, primarily due to his 50 years of accurate fieldwork in the Brazilian Amazon. He often collected one tree in flowering stage and returned to the same individual several months later to collect its fruiting material, mounting both collections on the same herbarium sheet. Ducke (1922) separated Pseudochimarrhis (with C. turbinata as type species) from Chimarrhis, stating that the latter genus differed from the former in having narrower, longer flowers, and vertically arranged seeds, characters that would place it in the Cinchoneae. Three years later, Ducke (1925) added a second species to Pseudochimarrhis (P. barbata).

    Benoist (1920) described Bathysa difformis (= C. turbinata), which he later transferred to Pseudochimarrhis (Benoist, 1933).

    Bremekamp (1934) stated that the separation of Pseudochimarrhis from Chimarrhis, proposed by Ducke (1922, 1925) was unacceptable, reporting: "The seeds of C. cymosa Jacq., the type species of the genus, however are arranged in exactly the same way as those of C. turbinata, and the longer and narrower flower is a character of so little importance that it is in itself not sufficient to justify the creation of a new genus. Pseudochimarrhis turbinata (DC.) Ducke is therefore referred back to Chimarrhis and P. barbata Ducke became Chimarrhis barbata (Ducke) Brem. comb, n." In the same work Bremekamp described C. longistipulata, which is here treated as synonymous with C. microcarpa (see discussion under this species). Pseudochimarrhis difformis has never been positioned in Chimarrhis, but this species is synonymous with C. turbinata.

    Rizzini (1947) treated Chimarrhis turbinata as a dubious taxon, proposing C. duckei Rizz. (nom. superfl.) as a substitute name for Pseudochimarrhis turbinata (DC.) Ducke.

    duced Pseudochimarrhis to Chimarrhis and synonymized the other species accordingly. He maintained Chimarrhis barbata and C. turbinata under Chimarrhis; described C. brevipes and C. bathysoides [= Bathysa bathysoides ((Steyerm.) Delprete]; raised C. cymosa subsp, jamaicensis Urb. and subsp, microcarpa to specific rank; maintained C. williamsii, C. hookeri, C. glabriflora, and C. parviflora as previously described; and divided C. microcarpa into var. microcarpa and var. speciosa (which are here raised to specific rank). Finally, in the list of excluded species, Steyermark (1965) reduced C. dioica and C. venezuelensis to synonymy under Dioicodendron dioicum; treated C. decurrens as synonymous with Allenanthus erythrocarpa Standi.; transferred C. goudotii, C. pisoniaeformis, and C. pittieri to Sickingia [= Simira]', and returned C. ferruginea to Rustia.

    Dwyer (1980) and Burger and Taylor (1993) recognized Chimarrhis as a genus of 14 species ranging from Costa Rica to South America and the West Indies.

    Steyermark (1965) published the only monograph of Chimarrhis, where he recognized 12 species, and with only the new species fully described. Following Bremekamp’s (1934) observations, Steyermark reduced Pseudochimarrhis to Chimarrhis and synonymized the other species accordingly. He maintained Chimarrhis barbata and C. turbinata under Chimarrhis; described C. brevipes and C. bathysoides [= Bathysa bathysoides ((Steyerm.) Delprete]; raised C. cymosa subsp. jamaicensis Urb. and subsp. microcarpa to specific rank; maintained C. williamsii, C.hookeri, C.glabriflora, and C. parviflora as previously described; and divided C. parviflora as previously described; and dibided c. microcarpa into var. microcarpa and var. speciosa (which are here raised to specific rank). Finally, in the list of excluded species, Steyermakr (1965) reduced C. dioica and C. venezuelensis to synonymy under Dioicodendron dioicum; treated C. decurrens as synonymous with Allenanthus erythrocarpus Standle; transferred C. goudotii, C. pisoniaeformis, and C. pittieri to Sickingia [=Simira]; and returned C. ferruginea to Rustia