Stickmannia hexandra (Aublet) Kuntze (1891: 721)

Aona, Lidyanne Yuriko Saleme, Bittrich, Volker & Amaral, Maria Do Carmo E., 2018, Taxonomic and nomenclatural notes on Brazilian Dichorisandra (Commelinaceae), Phytotaxa 348 (1), pp. 1-13 : 4-6

5. Dichorisandra hexandra (Aublet) Clarke (1902: 703) .

≡ Commelina hexandra Aublet (1775: 35) .

Type:— Habitat Caiennae & Guianea, prope ripas fluviorum, s.d., Aublet s.n. Lectotype (designated by Aona et al. 2014: 232): BM 000938194 ! (photo available at http://plants.jstor.org/stable/pdf/10.5555/al.ap.specimen.bm000938194); isolectotype P!).

≡ Stickmannia hexandra (Aublet) Kuntze (1891: 721) , syn. nov.

≡ Dichorisandra aubletiana Schultes f. (1830: 1181) , nom. illeg. (Art. 52.2, Ex. 2)

≡ Dichorisandra hexandra (Aublet) Standl. in Standl. & S. Calderón, Lista pl. Salvador: 48. 1925 (isonym).

≡ Dichorisandra hexandra (Aublet) Kuntze ex Handel-Mazzetti (1908: 203) (isonym).

= Convallaria diffusa Velloso (1827 [publ. 1831]: 160).

Lectotype (designated by Aona et al. 2014: 232):—[illustration] Tab. 160, Fl. Flum. Icon. Vol. 3, 1827 [publ. 1831], image available at https://www.biodiversitylibrary.org/item/15444#page/162/mode/1up

= Dichorisandra affinis Martius ex Schultes f. (1830: 1185) .

≡ Dichorisandra aubletiana Schultes f. var. affinis (Martius ex Schultes f. 1830: 1185) Clarke (1881: 274) . syn. nov.

Type:— BRAZIL. Pará : Habitat in sylvis udis parsim Provinciae Paraënsis et fl. Nigri. s.d., Martius s.n. (Lectotype [here designated]: M 0243709 !; Isolectotype: M 0243708 !; M 0243710 !) .

= Dichorisandra ovata Martius ex Schultes & Schultes f. var . α (1830: 1182).

≡ Dichorisandra aubletiana Schultes f. var. ovata (Martius ex Schultes f. (1830: 1182) Clarke (1881: 274). syn. nov.

Type:— BRAZIL. Minas Gerais: Habitat in sylvis primavaeris ad Mariana. Apr. Dr. Martius s.n. (Lectotype [here designated]: M 0086143 !; Isolectotype: M 0243705 !) .

= Dichorisandra inaequalis Presl (1830: 140) .

≡ Stickmannia inaequalis (C. Presl) Kuntze (1891: 721) , syn. nov.

Type:— COUNTRY UNKNOWN. s.d., s.l., (Lectotype [here designated]): PR 222583 A !; isolectotype: PR 222583 B!, PR 222583 C !).

= Dichorisandra mexicana Presl (1830: 140) .

≡ Stickmannia mexicana (C. Presl) Kuntze (1891: 721) , syn. nov.

Type:— MEXICO. s.d., s.l. (Lectotype [here designated]): PR222584 B !; isolectotype: PR 222584 A !).

= Dichorisandra ovalifolia Presl (1830: 140) .

≡ Stickmannia ovalifolia (C.Presl) Kuntze (1891: 721) , syn. nov.

Type: PANAMA. s.d., s.l. (holotype): PR 222540!).

= Dichorisandra intermedia Martius ex Schultes f. (1830: 1183) . syn. nov.

≡ Dichorisandra aubletiana Schultes f. var. intermedia Martius ex Schultes f. (1830: 1183) Clarke (1881: 274) . syn. nov.

Type:— BRAZIL. Habitat in sylvis Provinciae Paraënsis. s.d., Dr. Martius s.n. (Lectotype [here designated]: M 0086151 !) ; Isolectotypes: M 0243706 !, M 0243707 !)

= Dichorisandra ovata Martius ex Schultes f. (1830: 1182) var. ramosior Schultes f. (1830: 1182) . syn. nov.

Type:— BRAZIL. Minas Gerais. s.d., Martius s.n. (Lectotype [here designated] (or perhaps holotype): M 0086144!).

= Dichorisandra tenuior Martius ex Schultes f. (1830: 1183) . syn. nov.

≡ Stickmannia tenuior (Martius ex Schultes f.) Kuntze (1891: 721) , syn. nov.

Type:— BRAZIL. Brasilia. s.d., Dr. Martius s.n. (Lectotype [here designated]: M 0086146 !; Isolectotypes: M 0243701 !, M 0243702 !) .

= Dichorisandra aubletiana Schultes f. (1830: 1181) var. brasiliensis Schultes f. (1830: 1181) . syn. nov.

Type:— BRAZIL. Habitat in sylvis locis apricis. Provinciae Rio Negro. 1819. Dr. Martius s.n. (Lectotype [here designated] (or perhaps holotype): M 0086153!).

= Dichorisandra aubletiana Schultes f. (1830: 1181) var. persicariifolia (“ persicariaefolia”) Clarke (1881: 274). syn. nov.

Type:— BRAZIL. Amazonas : São Gabriel, capueirus. April 1852. R. Spruce 2202 (Lectotype [here designated]: K 000363226 ! (photo available at http://plants.jstor.org/stable/pdf/10.5555/al.ap.specimen.bm000884414); Isolectotypes: BM 000884414 !, NY!, RB!, G 00489469 !, P06233131 !, P06233133 !, B!, TCD 0008094 , F 0 BN010745 ) .

= Dichorisandra gracilis Nees & Martius (1823: 14) , syn. nov.

≡ Stickmannia gracilis (Nees & Martius) Kuntze (1891: 721) , syn. nov.

Type:— BRAZIL. Bahia: circa viam Felisbertiam, s.d, Wied s.n. Lectotype (designated here):—[illustration] tab. 1. fig. III in Nees & Martius (1823), available at https://www.biodiversitylibrary.org/page/37012360#page/111/mode/1up

Epitype (designated here):— BRAZIL. Bahia, Una , REBIO de Una, coletas efetuadas na estrada de acesso a Reserva. 12 May 1994. A. M. A. Amorim 3069 et al. ( UEC!; Isoepitype CEPEC!, SP!) .

Distribution: — D. hexandra is the most widely distributed species of the genus and occurs in Argentina, Bolivia, Colombia, Costa Rica, El Salvador, French Guiana, Guatemala, Panama, Paraguay, Suriname, Venezuela, and Brazil. In Brazil, it occurs in all regions of the country ( Barreto 1997, 2005, Aona & Leoni 2006).

Habitat:—Dense Ombrophilous Forest in the Amazon and on the Atlantic coast and Seasonal Semi-deciduous Forest. This species is more abundant in the Atlantic forest.

Notes:— D. hexandra generally has a scandent habit and flowers with purplish petals and six stamens. Other species of Dichorisandra also have these characteristics and have often been confused with D. hexandra . However, the material from the region of Cayenne, used by Aublet (1775) to describe Commelina hexandra (1775: 35), always has an inflorescence axis covered by very short trichomes, anthers dehiscent by two apical pores, and seeds with an orange to red aril. Within the species, the size and indument of the leaf blades and color of the stamens varies widely, but diagnostic characters, such as habit and anther dehiscence, can be used to identify herbarium material. Dichorisandra hexandra is the morphologically most variable and widely distributed species. Over the years, various collections with character states representing extremes of variation have been used to describe distinct species, without being clearly distinguishable from D. hexandra . Due to this, many published names are here synonymized under D. hexandra .

The lectotype of D. hexandra designated by Aona et al. (2014) was chosen based on features of habit and anther dehiscence, which are diagnostic characteristics. Clarke (1902: 703) made the new combination Dichorisandra hexandra in 1902. Ignoring this fact, Handel-Mazzetti (1908) also made the combination D. hexandra (1908: 203) and attributed it to Kuntze. In addition, Standley (1925) once again published the combination D. hexandra . The last two names are isonyms according to Art. 6.3 of the ICN ( McNeill et al. 2012).

Schultes f. transferred Commelina hexandra to Dichorisandra and in honor of Aublet called it D. aubletiana (1830: 1181), which is an illegitimate name according to Art. 52.2 of the ICN ( McNeill et al. 2012). Furthermore, Schultes f. described D. aubletiana var. β brasiliensis (1830: 1181), which is here synonymized under D. hexandra .

Schultes f. (1830) described many taxa similar to D. hexandra that differ in leaf morphology, degree of pubescence, or number of flowers per cincinni ( D. intermedia , D. ovata var. ramosior , D. tenuior and D. aubletiana var. brasiliensis ). However, due to the great morphological variation these taxa exhibit and the overlap between diagnostic characters states we believe maintaining these taxa as different entities is not justifiable.

The same is true for the names published by Clarke (1881), who reduced many species suggested by Martius and described by Schultes f. to the rank of varieties, e.g. D. aubletiana var. ovata , D. aubletiana var. intermedia , D. aubletiana var. affinis , and D. aubletiana var. persicariifolia (“ persicariaefolia”). Note we correct here the epithet persicariaefolia to persicariifolia, according Art. 60.8 of the ICN.In the present work, these varieties are also considered synonyms of D. hexandra . Despite D. aubletiana is an illegitimate name, the epithets of the new combinations made by Clarke (1881) for the varieties under this name are legitimate according to Art. 55.2of the ICN ( McNeill et al. 2012). Many specimens at M and K were labelled with these varietal names. We have chosen lectotypes from fertile material with flowers and fruits or with annotations by Martius or Clarke. In the case of D. aubletiana var. persicariifolia , the protologue in Clarke (1881) provides information about various collections, including specimens from Brazil (Spruce 2202, Spruce 4316, Sellow 1515), Peru (Spruce 4497), Colombia (“ New Granada ”) (L. Schlim 257) and Costa Rica (Hoffmann 638) (all syntypes, according to Art. 9.5). Spruce 2202 (K 000363226) was chosen as lectotype because it is fertile and duplicates exist in various herbaria. Still in relation to D. aubletiana var. persicariifolia , the specimen K 000363227 at Kew was not considered when choosing the lectotype. Although this specimen is clearly D. aubletiana var. persicariifolia , there are two labels on the specimen sheet. Both have the collectors number Spruce 2202 but one has a superscript added (i.e. Spruce 2202 x) and an annotation suggesting this is a mixed collection. For this sheet, only the part labelled Spruce 2202 can be considered an isolectotype.

We found no specimens of D. gracilis originating from cultivated plants or from Bahia at BR and M used by Nees & Martius (1823) to describe this species. Moraes et al. (2013), while studying the private herbarium of Prince Wied-Neuwied (designated as the “Wied Herbarium”), also failed to find any original specimens. The description in the protologue (“ Caulis longus, gracilis , subsimplex, angulosus, glaber, nodis remotis, geniculatus et ob laxitatem saepe decumbens ”), and the illustration in Nees & Martius that shows a flower with six stamens, a feature which overlaps with D. hexandra . Our analysis of material of D. hexandra collected in Bahia confirmed that there are no significant differences between these taxa. Based on these observations, it was not possible to morphologically separate D. gracilis from D. hexandra , so D. gracilis is here synonymized under D. hexandra . The illustration t.1 fig. III (a, b, c and d) of D. gracilis in Nees & Martius (1823) was chosen as the lectotype according to Art. 9.12 of the ICN ( McNeill et al. 2012). However, because this illustration only contains information about floral structures, additional material from Bahia was chosen as an epitype to allow the unambiguous application of the name.

Presl (1830) described three new species of Dichorisandra from Panama and Mexico. In the present work, only photographs of the types of D. inaequalis Presl (1830: 140) and D. mexicana Presl (1830: 140) were seen. Although the protologue of D. inaequalis contains information regarding the distribution of the species (“Hab. in Panama et in montanis Provinciae Huanoccensis Peruviae”), there is confusion regarding the geographic annotations of this material. After Kaspar Maria von Sternberg (1761–1838) and Carl Borivoj Presl (1794–1852) purchased material collected by Thaddaeus Peregrinus Xaverius Haenke (1762–1816), the specimens arrived at the National Museum in Prague (PR) in packages without individual labels but the annotations “Panamenses” and “Peruanae montanae guanoccenses”. It is likely that Sternberg transcribed these two locations into Haenke’s specimens. Consequently, C.B. Presl cited in the protologue both localities, but this should be rather read as a reflection of his doubts about the exact origin of the material (O. Sida, pers. com.).

Along with the photographs, the information in the protologue provides enough evidence to state that D. inaequalis , D. mexicana and D. ovalifolia Presl (1830: 140) are synonyms of D. hexandra . Grant (2000) reached the same conclusion when describing D. amabilis Grant , which is the only other Mesoamerican species of Dichorisandra . His position is accepted in the present work.