Sorghum arundinaceum (Desv.) Stapf

The Sorghum arundinaceum complex has been studied in detail by Snowden (in J. Linn. Soc., Bot. 55:191-260. 1955) who recognised 13 species, but it is now appreciated that these are freely interfertile with one another and with Sorghum bicolor. DeWet, Harlan & Price (in Am. J. Bot. 57: 704-707. 1970) and Doggett (Sorghum, 1970) point out that there is no cytogenetic justification for recognising more than one species, though Doggett prefers, as a matter of practical convenience, to retain a separate binomial for the crop plant.

Sorghum arundinaceum is extremely variable, a variability enhanced by human selection of grain races which introgress with the wild species. Variation is continuous and it is doubtful whether formal infraspecific categories are of much value. The species is questionably native in Asia, probably being indirectly introduced as genetic throwbacks from grain Sorghums.

Very little material from our area has been seen, but since Sorghums are so widely cultivated in Pakistan there can be no doubt that this species, as well as Sorghum x drummondii, will, with a little searching, eventually be found in greater quantity.

Annual or short-lived perennial without rhizomes; culms 0.3-4 m high often robust, the nodes glabrous or pubescent. Leaf-blades variable, often large, 5-75 cm long, 5-70 mm wide. Panicle linear to broadly spreading, 10-60 cm long; primary branches compound, ultimately bearing racemes of 2-7 spikelet pairs. Sessile spikelet lanceolate to narrowly ovate, 4-9 mm long, glabrescent to white pubescent, sometimes tomentose or fulvously pubescent, awnless or more often with an awn 5-30 mm long. Pedicelled spikelet linear to lanceolate, male or barren, smaller than the sessile. 2n=20.

Distribution: Kashmir; throughout Africa, extending eastwards to Australia; India; introduced to tropical America.