Digitaria tenuis

12. DIGITARIA TENUIS View in CoL (Nees) Henrard, Blumea 1: 99. 1934. Trichachne tenuis Nees, Fl. Bras. Enum. Pl. 2 (1): 89. 1829. — TYPE: BRAZIL. Brasilia: Piauhy, Martius s. n. (holotype: M; isotypes: BAA–648!, K!).

Iconography — Vega and Rúgolo de Agrasar (2005: 263).

Geographical Distribution and Ecology — Caribbean (St. Thomas Island), Venezuela, Guyana ( Judziewicz 1990), and Brazil. Extends from 100– 850 m. Grows in dry sandy scrubs and grassy banks from sea level to 850 m.

Excluded Species — Digitaria sellowii (Müll. Hal.) Henrard

Trichachne sellowii was described by C. Müller in 1861. The diagnosis was based on a single specimen originally identified as Anatherum brasiliense Spreng. , deposited in Sprengel herbarium (first at HAL and then at B, where it was destroyed), and collected in “Brasilia, ubi forsan in Montevideo legit Sellow ”. In the diagnosis the lower glume is not mentioned, the upper glume is described as the superior lower glume and the lower lemma as the major lower glume. Dimensions of the spikelets are not included and upper floret is described as having membranous shortly ciliate margins, a character that is not present in Digitaria . For these reasons, Parodi (in litt.) and Rúgolo de Agrasar (1974) suggested that the specimen belongs to a species of Leptocoryphium or Axonopus . Another possible interpretation is that the diagnosis refers to a species of Digitaria but has omissions and mistakes.

Henrard (1950) published the combination Digitaria sellowii (Müll. Hal.) Henrard and stated that the type specimen of T. sellowii was destroyed during the Second World War. He designated a neotype (Kneucker 368) and provided an illustration of the spikelet and upper floret based on the same specimen ( Henrard, 1950: 674). This specimen was previously mentioned by Henrard (1950: 642) in the synonymy of D. sacchariflora (Raddi) Henrard and then later identified as D. swalleniana Henrard by Rúgolo de Agrasar (1974: 91). On the basis of a lost type specimen of T. sellowii , omissions and possible mistakes in the original diagnosis, and a neotype that can be applied to another species, Rúgolo de Agrasar (1974: 156) considered D. sellowii as a dubious taxon.

Later Canto-Dorow (2001) reported that she had found the type specimen of T. sellowii at B, restored D. sellowii (Müll. Hal.) Henrard , and included D. similis Beetle ex Gould as a synonym. Canto-Dorow (2014) applied this taxonomical criterion in regional floras from Brazil. Our analysis of high quality images from the putative type specimen “Brasilia, Sellow s. n. ” deposited at B is in accordance with the diagnostic characters of D. similis Beetle ex Gould. The label is not in full concordance with the protologue, and the specimen differs by many characters from the T. sellowii diagnosis ( Müller 1861). On the basis of Parodi (in litt.) and Rúgolo de Agrasar (1974) criteria, as well as the characteristics pointed in previous paragraphs, this specimen cannot be unequivocally assigned to the type of T. sellowii . Even though, if Sellow s. n. would have been found at B, it had been identified as D. similis . On this basis, the identity of D. sellowii is unresolved and we excluded it from D. sect. Trichachne .

Placement of Megaloprotachne albescens and Nomenclatural Consequences — The genus Megaloprotachne C. E. Hubb. once included two species, M. albescens C. E. Hubb. and M. glabrescens Roiv. , both native to southern Africa. The latter species was synonymized under M. albescens ( Figueiredo and Smith 2008) . After studying the type specimens and original diagnosis of both species, the main difference is the lack of villous hairs on collars and leaf blades in M. glabrescens , and their distribution is limited to the lower parts of the lower leaf sheaths. In M. albescens , villous hairs are abundant in leaf sheaths, nodes, collars, on the blades of lower leaves, especially on the ventral surfaces. These observations are in agreement with the taxonomic criteria of Figueiredo and Smith (2008).

Taking into account all our findings, the transfer of M. albescens to Digitaria is provided: