65. Panicum L. (panic grass)

Plants with C₃ or C₄ photosynthesis, annual or perennial, rarely with stolons, forming tufts, clumps, or rarely colonies. Flowering stems erect to ascending or rarely spreading in the basal portion and rooting at the lower nodes, glabrous or hairy, sometimes with a dense ring (beard) of downwardly pointing hairs at the nodes. Leaves often differentiated into a dense basal rosette of shorter overwintering leaves and fewer, longer leaves on the flowering stems. Leaf sheaths glabrous or hairy, the ligule a line or band of hairs or a membrane, this glabrous or more commonly with a fringe of hairs along the margin. Leaf blades usually flat, the base rounded or abruptly narrowed, glabrous or hairy. Inflorescences dense or more commonly open panicles with at least some of the branches rebranched 1 or more times (except in P. obtusum), the spikelets densely or loosely spaced along the branches, these not appearing 1‑sided (except sometimes in P. obtusum), the axis with a spikelet at the tip. Spikelets not subtended by bristles or spines, without a cuplike ring or knoblike disk at the base. Lower glume much shorter than to less commonly about as long as the rest of the spikelet, wrapped around the spikelet base, awnless, glabrous. Upper glume about as long as the rest of the spikelet, not inflated or saclike at the base, rounded to sharply pointed at the tip, awnless. Lowermost floret staminate or sterile, the lemma about as long as the rest of the spikelet, awnless. Fertile (perfect) floret with the lemma shorter than that of the staminate or sterile floret, rounded at the tip, awnless, nerveless or obscurely nerved, glabrous, shiny or dull, thickened and hard (usually somewhat bonelike) at maturity, the margins also thick, wrapped around the palea and fruit, including the tip (after flowering). Paleas glabrous, shiny or dull, thickened and hard (usually somewhat bonelike) at maturity. Fruits narrowly to broadly oblong‑elliptic in outline. About 500 species, nearly worldwide, mostly in tropical and subtropical regions.

Perhaps no question of generic limits in grasses is more controversial than whether to segregate Dichanthelium from Panicum or to treat it as a subgenus within the latter genus. There seems little doubt that the species of Dichanthelium form a natural group, which Gould (1974) first elevated to generic level. Characters used to differentiate Dichanthelium from the rest of Panicum include: a C₃ (vs. mostly C₄) life cycle and the associated anatomical characters; production of an overwintering rosette of basal leaves that are morphologically distinct (shorter and broader) from the leaves of the flowering stems; 2 flowering periods each year, with more prominent inflorescences from relatively unbranched flowering stems in the spring (vernal flowering) followed by the production of fascicles of short branches on these stems with smaller inflorescences (often with cleistogamous spikelets) usually partially hidden by the leaves in the autumn (autumnal flowering); anatomical differences in the growing tips of stems; and micromorphological differences in the surfaces of the paleas (see table of comparisons in Gould and Clark, 1978). In the present treatment, species 1, 3, 4, 6–8, 10, 13–17, 20, 22, 23, 25, and 26 are members of the Dichanthelium group, whereas species 2, 5, 9, 11, 12, 18, 19, 21, 24, and 27 belong to the rest of Panicum.

However, Gould and his coworkers appear to have been most familiar with North American members of the group. Subsequent analyses of South American species in the Dichanthelium group, particularly the anatomical studies of Zuloaga and his collaborators (see Zuloaga and Morrone, 1991), have shown that there exist species that are intermediate for the characters supposedly separating Dichanthelium from the rest of Panicum. Also, the group of species related to P. hians (subgenus Steinchisma (Raf.) Zuloaga) have long been known to possess anatomy and physiology somewhat intermediate between the C₃ and C₄ types. Recently, in a numerical study of 64 characters and 188 New World species, Zuloaga et al. (1993) found strong evidence that American Panicum in the broad sense consists of five major lineages (with a number of subgroups within each lineage), and that species of Dichanthelium are nested in a larger lineage containing species considered part of Panicum by most botanists. Thus, the more conservative approach of treating Panicum in the broad sense, including Dichanthelium, is followed provisionally in the present treatment, in agreement with many other authors (including Lelong, 1986; Webster, 1988; Crins, 1991; Gleason and Cronquist, 1991; Zuloaga et al., 1993).

Whatever the generic treatment of Panicum, most botanists currently agree that too many species were recognized by earlier authors (including Steyermark, 1963). More recent revisions of portions of the genus (Gould and Clark, 1978; Lelong, 1984, 1986; Hansen and Wunderlin, 1988) have treated numerous taxa as minor morphological variants of other species. This has resulted in lengthy lists of synonyms for some of the widespread taxa, as well as relatively numerous varieties accepted for some of these species. A few of the species complexes, particularly those centered on P. acuminatum and P. dichotomum, are in need of more detailed biosystematic studies. Ecologists and field biologists sometimes have argued that various entities placed in synonymy under common, widespread species occupy discrete ecological niches and deserve taxonomic recognition at some level. It is difficult without more detailed taxonomic study to evaluate whether such cases are based upon genuinely distinct biological entities or merely reflect morphological variability within a small part of the range of a widespread species.