Sorghum caucasicum (Trin.) Griseb.
Tufted perennial; culms 50-150 cm high, erect from a shortly rhizomatous base, often robust. Leaf-blades 10-55 cm long, 2-12 mm wide. Inflorescence with a central axis 4-20 cm long, the racemes numerous and borne loosely or densely in irregular whorls upon simple or branched peduncles; racemes 2-5 cm long (the lowest shorter than the central axis), pubescent. Sessile spikelet narrowly elliptic, 3-4 mm long; lower glume chartaceous, ± hairy below the middle, usually not glossy, with or without a pit; awn 10-25 mm long. Pedicelled spikelet glabrous, with 0-3 pits.
Type: China, Bladh (LD).
Distribution: Pakistan (Punjab, N. W. F. P, Gilgit & Kashmir); Old World tropics.
Much of the taxonomic difficulty in this genus arises from the promiscuous habits of Bothriochloa bladhii whose rapacious introgression has vastly increased its own variability, blurred the boundaries with adjacent species to the point of extinction, and led to the emergence of new races from among the introgression products (DeWet & Harlan in Am. J. Bot. 53: 94-98. 1966; and Faruqi in Phyton (Austria) 13:285-303. 1969). The species is here treated in a wide sense to include all specimens with an elongated inflorescence. The only exception is the hybrid with Dichanthium annulatum which is sufficiently distinct from Bothriochloa to warrant a hybrid formula.
In some accounts (e.g. Bor, Grasses Burma Ceyl. Ind. Pak. and in Fl. Iran; Rozhev & Shishkin in Fl. URSS) Bothriochloa caucasica is recognised at species level. Its ancestry is complicated, but according to DeWet & Harlan (in Taxon 19:339. 1970) it probably arose from tetraploid races of Bothriochloa intermedia and Capillipedium parviflorum hybridising and then backcrossing to Capillipedium parviflorum, Although it does form a small, reasonably compact subpopulation of Bothriochloa bladhii, with the lower lemma of the sessile spikelet only about half as long as the lower glume, the distinction is not sufficient for it to warrant species rank. It is a better species from the cytogeneticist’s point of view than it is from the morphologist’s. It is not in fact a caucasian endemic, but occurs at least as far east as Chitral and doubtless could arise spontaneously wherever its parents grow together.
The various hybrids of Bothriochloa intermedia, according to DeWet & Harlan (1970), are as follows:
Bothriochloa intermedia x Bothriochloa ischaemum = Bothriochloa taiwanensis (included in Bothriochloa bladhii)
Bothriochloa intermedia x Dichanthium annulatum = " Dichanthium grahamii” (here kept apart from Bothriochloa)
Bothriochloa intermedia x Capillipedium parviflorum = Bothriochloa glabra (included in Bothriochloa bladhii)
(Bothriochloa intermedia x Capillipedium praviflorum ) x Capillipedium parviflorum = Bothriochloa caucasica (included in Bothriochloa hladhii)