Arundinella goeringii Steud.

Veldkamp, J. F., 2015, Arundinella (Gramineae) in Malesia with notes on other taxa and on aluminium accumulation, Blumea 59 (3), pp. 167-179 : 172

3. Arundinella goeringii Steud.

Arundinella goeringii Steud. (1854a) 116. — Type: Goering II, 139 (holo P; from Java , not Japan).

[ Arundinella hispida Nees ex Buse var.‘ β angustior ’ Buse (Feb.1854) preprint: 19; 359,not a name, but a diagnose. — Vouchers: Junghuhn s.n. ‘Tjibogo’ (L, sh. 902.23-56, 908.83-1356, ‘Tjerimai’ (L, sh. 908.83-1332; fragm. in K), with label by? Nees].

Arundinella agrostoides Trin. var. intermedia Kuntze (1891) 761. — Type: Kuntze ‘ Java, Megamendong’ (NY not found; K).

Arundinella procumbens Janowski (in Janowski & Mez 1921) 84; (1922) 25. — Type: Warburg 1648 (holo B †), perhaps in A, BM, E, K, NY. Error for 16483 ?

Arundinella agrostoides Trin. var. ciliata auct. non Kuntze: Kuntze (1891) 761, pro Kuntze 4805 (K, NY) from Java.

Arundinella fuscata auct. non Buse.

Arundinella hispida auct. non Kuntze.

Arundinella nepalensis auct. non Trin.

Plants annual or perennial. Culms tufted or more or less solitary, erect, 0.15–0.6 m long. Nodes bearded. Sheaths sparsely pilose with bb hairs. Throat hairy. Ligule 0.3–1 mm high, pilose. Blades flaccid to rather stiff, flat, linear-lanceolate to linear, 4.5–25 cm by 4.5–17 mm, pilose on both sides, margins pilose. Peduncle under the panicle glabrous. Panicle contracted, often interrupted at base, 6.5–30 by 1–6 cm, lowermost branch 1–3 together, the longest one 2–4 cm long, branches angular, smooth to scaberulous, glabrous. Pedicels smooth to scaberulous. Spikelets 4.2–6 mm long, greenish to purple suffused. Glumes glabrous; lower glume 2.4–4 mm long, 0.6–0.86 times as long as the spikelet, distinctly shorter to slightly longer than the lower lemma, apex acuminate to caudate to mucronate, 3- or 5-nerved; upper glume 5-nerved, apex long-acuminate. Lower lemma male or bisexual, 2.8–3.5 mm long, nerves 5, rarely 3, not anastomosing, apex acuminate; upper lemma 1.8–2.3 mm long, whitish to brown or castaneous (at maturity), microscopically scaberulous (40 ×!), 0-nerved, callus hairs 0.4–0.7 mm long, 0.2–0.37 times as long as the lemma, apex acute, without lateral setae, awn distinctly exserted beyond the glumes, geniculate, 3.3–5 mm long when dry, column 1–1.9 mm long when dry. Anthers 3, 0.8–1 mm long.

Distribution — Malesia: Sumatra (W Coast), Java (widespread, especially in the W), Celebes (Malino), Philippines (Palawan),? Sumbawa (Ploem s.n., L sh. 926.140-174, sterile).

Habitat — Poor soil; 40–1600 m altitude.

Notes — Nees apparently around 1834 sent around lists of names and descriptions to a.o. Buse, Steudel, Trinius , and Wight, who attributing them to Nees used them in their publications. Only Wight (Cat. Indian Pl. (1834)) specifically acknowledged Nees’s contribution, so the names there are to be cited ‘Nees in Wight’. Unfortunately the majority are nomina nuda. In the other cases there is only external evidence of the source of this information. Thus Art. 46.6 applies and the names must be cited as ‘Nees ex ...’.

In Wight’s catalogue there appear several species of Arundinella , some of which have Acratherum as an infrageneric indication. These were cited by Steudel (1840: 21, 143) as synonyms of Arundinella spp. One of these is Arundinella (Acratherum) fuscata Nees (Wight Cat. 1667, probably the same as the Wight Herb. Propr. 183 cited by Janowski (1922: 26). A duplicate of this is in the Trinius Herbarium 405.1).

The validation of Arundinella fuscata Nees was by Buse (1854, preprint: 19). However, no specimens or provenance were given, only a var. β. ‘angustiore ’, which word from the typography is not a name, but a remark. The specimens mentioned for this (in L) are all marked β on their labels. One (sh. 908.83-1332) has the combination in what appears to be Nees’s handwriting. Steudel, very shortly afterwards (1854a: 114), used Arundinella fuscata again, merely stating ‘Penins. Ind. or.’, but because of the above obviously meaning the Wight collection.

Note that Buse’s and Steudel’s A. fuscata are not true isonyms, as they are heterotypic, being based on different specimens in different institutes, but according to Art. 9. Note 2(c) they would be.

Hooker (1896: 74) correctly attributed the combination to Buse, and now we see that Wight Cat. 1667 came from the Nilgiris of S India. Bor (1955: 406, 1960: 424, as A. purpurea Hochst. ex Steud. ), wrote that it is an endemic of that area.

Obviously, it has been thought that a name used in a Malesian account could not be applied to a Nilgiri endemic, hence Hooker’s use is cited as ‘non Nees ex Buse’, and another combination is used. I hope to have now shown that it is actually the other way around: Buse used a description for an Indian species and applied it to what he thought was an aberrant form from Java. Anyway, Buse validated Arundinella fuscata , whereby Hooker’s use of it for the Indian species was correct (as so often) and must replace the later A. purpurea .

Arundinella fuscata can therefore not be used for our Malesian endemic. The specimens of Buse’s β belong to a species for which the oldest name is A. goeringii Steud. , which has been rarely used (e.g. in Boldingh 1916: 94, Janowski 1922: 23, 26, Keng 1936: 5, 37, t. 17). There is often some confusion about the provenance of Goering collections: Japan or Java? But the type of this one is the Malesian species and came from Java.