Dichanthelium transiens (Swallen) Sánchez-Ken 2017

Sánchez-Ken, J. Gabriel, 2017, Dichanthelium multiglandulosum (Paniceae, Panicoideae, Poaceae), a new species from Jalisco, Mexico, and the reestablishment of Panicum transiens as Dichanthelium transiens, Phytotaxa 307 (1), pp. 23-35 : 30-33

Dichanthelium transiens (Swallen) Sánchez-Ken , comb. nov. ( Fig. 5 ).

Basionym: Panicum transiens Swallen (1931: 436) .

Type: — MEXICO. Tamaulipas: Mpio. San Carlos: Sierra de San Carlos, Mesa de Tierra , vicinity of San José, 3500 ft, 19 July 1930, H. H. Bartlett 10454 (holotype US-00148063!; isotypes MEXU-3882!, MICH-1108749!, NY-00381779!, US-00148064!) .

Dichanthelium macrospermum Gould (1980: 358) .

Type: — MEXICO. San Luis Potosi: 30 mi E San Luis Potosi, along highway 86 to Rio Verde, 6900 ft, 13 July 1963, R. L. McGregor, L. J. Harms, A. J. Robinson, R. del Rosario & R. Segal 618 (holotype US-00133537!).

Plants perennial, 30–100 cm tall, knotty rhizomatous, with a corm-like base (not always evident), basal rosette lacking, only one phase of growth recognized, culms cespitose, slightly cane-like, commonly erect, but rarely ascending and rooting at lower nodes, scarcely branching at the base and above, eglandular; internodes 5–9, 1.5–2.5 mm diameter, papillose-pilose becoming glabrescent, rarely glabrous; nodes densely short pilose or glabrescent. Leaf sheaths shorter or longer than the internodes, eglandular, papillose-hirsute becoming glabrescent, margins pilose; collar densely short pilose; auricles sometimes present, as long as the ligule; ligules 0.1–0.3 mm long, membranous-ciliate, with the hairs or cilia shorter than the membrane; leaf blades (6–) 9–17 cm × (8–) 9–20 mm, the lowermost reduced, lanceolate, flat, spreading, hirtellous on both surfaces, sometimes glabrescent, throat densely short pilose behind the ligule without forming a pseudoligule, margins scabrous, sparsely short to long pilose at the base, base cordate, apex attenuate, foliar dimorphism lacking. Synflorescence (4–) 10–20 cm long, lowermost branches 3–10 cm long, diffuse, terminal or lateral when the plant is branched, paniculate, branches alternating, spreading, ascending, sparsely hirtellous to glabrescent; peduncle 7–25 cm long, normally hirtellous rarely glabrous, eglandular; axis hirtellous glabrescent toward the apex; axils glabrous, yellowish to brown; pedicels 1–10 mm long, the terminal ones much longer to 2 cm long, scabrellous, capillary, appressed, rarely spreading eglandular. Spikelets (2.9–)3–4 × 1.2–1.5 mm, elliptic to slightly obovate, acuminate, base attenuate, light green, rarely purple tinged, glandular at the base, papillose hirtellous, always with few hairs, the hairs spreading to nearly glabrous; first glume 1.8–2.6 mm long, ca. half as long as the spikelets, ovate to lanceolate, 1–3-nerved, glabrescent, with at least 1–2 hairs towards the margins distally; second glume 2.8–3.9 mm long, leaving the fertile floret exposed at maturity, 9-nerved, shortly papillose-hirtellous, the hairs distributed all over or sparse; lower floret sterile; sterile lemma 2.8–3.9 mm long, 9-nerved, texture and pubescence similar to the second glume; sterile palea 1.4–1.8 mm long, ca. half as long as the fertile floret, ovate, membranous, 2-keeled on lower half, glabrous or puberulent; fertile lemma 2.9–3 × 1.1–1.2 mm, elliptic to slightly obovate because of the stipe, smooth to apparently slightly rugulose depending on age, lustrous, yellowish to stramineous, 7-nerved but appearing 5-veined because the two veins adjacent to the central one are incomplete and visible only towards the apex, with a germination flap, apex acute, umbonate with tiny antrorse hairs on and around the umbo, falling with the glumes; fertile palea smooth, lustrous, similar to the fertile lemma but the papillae more evident; stamens 3; anthers 1–1.1 mm long; caryopsis oblong, embryo ca. half as long as the caryopsis, hilum oblong, one third as long as the caryopsis.

Plants from southern locations are more robust and more glabrescent than northerly-distributed plants, and most have many-flowered synflorescences.

Additional specimens examined:— MEXICO. Nuevo León: Mpio. Galeana, Sierra Madre Oriental, Mesa de la Camisa, about 15 mi SW of Galeana, 22 July 1934, C. H. Mueller & M. T. Mueller 1171 ( F, MEXU) ; Mpio. General Zaragoza: Sierra Madre Oriental, 0.5 mi NE Dulces Nombres , and just E of border into Tamaulipas, 1850 m, 18 June 1948, F. G. Meyer & D. J. Rogers 2574 ( MO) ; Mpio. Monterrey: Sierra Madre , 7 April 1933, C. H. Mueller & M. T. Mueller 166, 408 ( MEXU) ; Mpio. San Pedro Garza García: Parque Ecológico Chipinque, vereda El Empalme, 1266–1291 m, 6 October 2005, J. A. Encina , F. J. Encina & I. Ramírez 1540 ( MEXU) ; Mpio. Santiago: delante de San Juan Bautista, rumbo a Laguna de Sánchez , 1600 m, 19 May 2004, E. Estrada et al. 16125 ( MEXU). Querétaro: Mpio. Arroyo Seco: 4–5 km WNW de La Florida, 1300 m, 29 July 1991, E. Carranza G. 3322 ( IEB) ; 2 km W de El Jardín, 1400–1500 m, 7 June 1991, H. Díaz B. & E. Carranza G. 6662 ( IEB) ; Mpio. Jalpan, 5–6 km NE de San Antonio Tancoyol , 1000–1200 m, 6 March 1991, E. Carranza G. 3073 ( IEB) ; 3 km SE de El Lobo , 1780 m, 7 December 1987, L. M. Chávez 139 ( IEB) ; 5 km al W de El Lobo, sobre el camino a Landa , 1600 m, 26 June 1959, J. Rzedowski 10925 ( ENCB) ; 4 km S de Valle Verde (La Parada), sobre el camino a Landa , 1200 m, 2 June 2005, V. W. Steinmann & E. Pérez C. 5092 ( IEB) ; ca 2 km from Valle Verde ( La Parada) along the trail to La Cercada, 1350 m, 24 September 2002, V. W. Steinmann & S. Zamudio 2798 ( IEB) ; Mpio. Landa, Hoyo del Lodo, 5 km N de Acatitlán de Zaragoza, 1780 m, 2 October 1989, E. González P. 1085 ( IEB) ; 2 km SW de El Madroño, 1900 m, 23 May 1990, E. Carranza G. 2499 ( IEB) ; 6–7 km NE de la Lagunita de San Diego, vertiente SE del Cerro Grande, 2300 m, 2 August 2000, G. Ocampo & E. Pérez C. 895 ( IEB) ; 1.5 km NE de El Sabinito , 1500, 16 June 1990, H. Rubio 1728 ( IEB) ; 1 km SW de El Lobo , 1600 m, 1 August 1987, J. Rzedowski 44033 ( IEB). San Luis Potosí: Mpio. El Naranjo, 21.1 mi W of SLP-Tamaulipas border on MEX 80 from Antiguo Morelos, 4000 ft, 14 August 1989, J. F. Utley & K. Utley 8470 ( MEXU) ; about 18 mi E of Ciudad del Maiz on Mex 70 near pueblo of El Naranjo, 4000 ft, 27 May 1974, G. McPherson , T. Duncan & R. Sanders 915 ( NY) ; Mpio. Xilititla, Potrerillos, 1400 m, 28 February 1959, J. Rzedowski 10020 ( ENCB). Tamaulipas: Mpio. Guémez , El Chiue , 1800 m, 19 August 1981, J. L. Ramos 18 ( MEXU) ; La Ventanita, 1250 m, 4 November 1982, G. Villegas 538 ( MEXU) ; Mpio. Hidalgo, 1 km al S de Puerto Purificación, rumbo a Conrado Castillo, 13 July 1988, F. González M., I. Díaz & D. Corona 17045 ( MEXU) ; 5 km al N de los Caballos, May 1984, F. González M., L. González, V. Juárez, D. Baro & L. Hernández 14029, 14042 ( MEXU) ; Aserradero Galindo, a 6 km de Puerto Purificación, 5 June 1990, F. González M., V. Juárez & P. Tenorio 17430 ( MEXU) ; Mpio. Jaumave , La Trementina, 1940 m, 21 April 1994, G. Galván 453 ( MEXU) ; Mpio. San Carlos: Sierra de San Carlos, La Vegonia, vicinity of San José, 975 m, 5 July 1930, H. H. Bartlett 10090 ( US) ; Mina La Homogénea San José, 900–1000 m, 28 September 1985, O. L. Briones V. 2048 ( MEXU) ; Mpio. Victoria, camino al Molino , 800 m, 10 November 1986, M. H. Cervera 191ª ( MEXU) ; Altas Cumbres , 1240 m, 19 October 1982, P. Moya 14 ( MEXU) .

Distribution, habitat, and ecology:— This species has a restricted distribution in Mexico in the Sierra Madre Oriental mountains in the states of Queretaro, San Luis Potosi, Nuevo León and Tamaulipas ( Fig. 3 ). Found in cloud forest, oak or oak-pine forest, or sometimes secondary grasslands, at elevations from 800 to 2300 m.

Phenology:— Flowering from February to December.

Diagnostic characters:— Plants with a corm-like base, slight cane-like habit, basal rosette lacking, scarcely branching below and above, eglandular, collar densely short-pilose, ligules 0.1–0.3 membranous-ciliate, and spikelets (2.9–) 3–4 mm, elliptic to slightly obovate, papillose-hirtellous, the hairs spreading to glabrescent but always with few hairs.

Spikelet micromorphology:— The glumes and sterile lemma are papillose-hirtellous and at the base of both glumes there is a wrinkled area, which may represent a gland. The fertile lemma is smooth and shiny all over, although at high magnification (120×) papillae are evident. The fertile lemma apex is acute and umbonate, with a few tiny antrorse hairs on and around it ( Fig. 3 ).

Photosynthetic pathway: — Leaf blade cross sections confirm a “non-Kranz” anatomy associated with a C 3 photosynthetic pathway ( Brown & Smith 1975, Zuloaga et al. 1993). The anatomy of the leaf blade of D. transiens is similar to D. multiglandulosum but the former has more cells between vascular bundles.

Related species and section affiliation:— Panicum transiens was described by Swallen (1931), who placed it in the Pedicellata group sensu Hitchcock & Chase (1910) with the two other members of the group, P. pedicellatum Vasey (1889: 28) and P. nodatum Hitchcock & Chase (1910: 293) . The inclusion of P. transiens in this group was based on the lack of a basal rosette, corm-like culms and attenuate papillose-pilose spikelets. It differs from the two other species by having longer spikelets (4 mm long), a longer subacute first glume, and larger plants, with culms scarcely branching below and above and much longer and wider blades without foliar dimorphism. In the original description the author stated “ligule densely ciliate, about 0.5 mm ”, but he was probably referring to a membranous-ciliate ligule. The isotype (Bartlett 10454, MEXU) has a membranous-ciliate ligule with the membranous portion longer than the cilia. On the other hand, D. pedicellatum and D. nodatum have pilose ligules sensu Hitchcock & Chase (1910), although in the section sensu Freckmann & Lelong (2003), the authors mentioned membranous-ciliate ligules for D. pedicellatum . Based on my observations, the latter authors are correct, although the membranous portion of the ligule is very short, which makes it look like a line of hairs, contrary to P. transiens in which the membranous portion is longer than the cilia or hairs. The specimens A.F. Skutch 524 (probably US) and J.A. Steyermark 43443 (F) were included under the name Panicum transiens in Flora of Guatemala by Swallen (1955), although he was doubtful about their identity and assigned them to this species with the possibility they might be a different species. According to the Swallen´s (1955) key, these specimens have glabrous spikelets. Most specimens of P. transiens from throughout its known distribution range have papillose-pilose spikelets, although some specimens have nearly glabrous spikelets, with a few scattered hairs. Later, Zuloaga et al. (1993) identified the specimen A.F. Skutch 524 as Panicum divergens Kunth in Bonpland et al. (1816: 102), a name now considered a synonym of D. commutatum , and the specimen J.A. Steyermark 43443 (F) was identified as D. pedicellatum by Davidse in Davidse et al. (1994).

In a treatment of the Mexican species of Dichanthelium, Gould (1980) placed P. transiens in the synonymy of D. pedicellatum without comment. Possibly Gould (1980) did not see the type of P. transiens , because he did not cite it, and he probably followed Swallen´s assumption that it belonged to the Pedicellata group sensu Hitchcock & Chase (1910). In a Mexican grasses checklist, Beetle (1987) recognized P. transiens as a distinct species, but later as a synonym of D. pedicellatum ( Beetle et al. 1999) . After reviewing an isotype of P. transiens and comparing it with specimens of D. pedicellatum and D. nodatum I concluded the name was wrongly synonymized by Gould (1980) and should be recognized as a species in the genus Dichanthelium .

The culms of D. transiens are sometimes scarcely branched above (at least in a few specimens collected from February to December) and the leaves are (6–) 9–17 cm × (8–) 9–20 mm and lanceolate with cordate bases, whereas in D. pedicellatum and D. nodatum the culms are usually much branched above (in specimens collected from April to August) and the leaf blades are 3–12 cm × 2–8 mm and linear with bases rounded, truncate or subcordate. The venation of the leaf blades of D. transiens comprises a central thick vein and, on each side of the midvein, a tandem array of (4–)5 areas of several tertiary veins divided by a secondary vein, contrary to the 3 to 4 tandem arrays on each side of the central vein in the Pedicellata species. In some northern specimens of D. transiens the synflorescences are fewflowered like those in D. pedicellatum , but in most plants the synflorescences in D. transiens are (4–) 10–20 cm long, with southern specimens multi-flowered.

In the same publication, Gould (1980) described D. macrospermum as a close ally of D. albomaculatum ( Scribner 1900: 2) Gould (1980: 357) (the latter name was later synonymized to D. commutatum ) from the group Commutata, currently recognized as section Macrocarpa ( Freckmann & Lelong 2002) . According to Gould (1980), D. albomaculatum differed from D. macrospermum by its smaller spikelets, smaller first glume and glabrous axis and branches of the synflorescence. Dichanthelium macrospermum has not been collected since its description and was considered a rare species, known only from the three specimens reported in the protologue. Beetle (1987) and Beetle et al. (1999) treated it in Panicum ; in the former treatment it was reported for three states of Mexico (Hidalgo, Queretaro and San Luis Potosi) and in the latter treatment only from Queretaro and San Luis Potosi. After reviewing types of D. macrospermum and Panicum transiens , I conclude they are the same species, and because the latter name was published first, the accepted name in Dichanthelium should be D. transiens , proposed here.

To determine the sectional placement for D. transiens all the sections with membranous-ciliate ligules and large spikelets were compared ( Table 1). The three sections sensu Freckmann & Lelong (2002, 2003) with these characters are Pedicellata , Clandestina and Macrocarpa . Dichanthelium transiens cannot be placed in the Pedicellata section, where Swallen (1931) included it, because of the differences among D. transiens , D. nodatum and D. pedicellatum discussed above. The section Clandestina is excluded from consideration by having thick rhizomes, a basal welldifferentiated rosette, dense culms branching above, and elliptic to ovate spikelets. The characters of D. transiens agree best with the circumscription of section Macrocarpa ( Freckmann & Lelong 2002) ; in fact, D. transiens is similar to D. commutatum and many specimens of D. transiens were wrongly identified as part of the latter species. The inclusion of D. transiens in D. commutatum probably altered the original circumscription of the latter in publications such as Flora Novo-Galiciana ( McVaugh 1983) and Las Gramíneas de México ( Beetle et al. 1999), which will need to be verified. More recently, the D. commutatum species complex has been suggested to be polyphyletic ( Majure et al. 2016, Neubig 2016). If D. transiens belongs in sect. Macrocarpa , the circumscription of the section will need to be further expanded to include species that have corm-like bases, in addition to the characters of D. multiglandulosum also unique in the section (described above), and corm-like bases would no longer be exclusive to sect. Pedicellata . Otherwise, D. transiens would remain unclassified until relationships within the genus are clarified. Dichanthelium transiens is easily distinguished from D. commutatum and D. multiglandulosum by the presence of corm-like bases of the culms, densely pilose collars, pubescence of the internodes, pubescence of the peduncle and axis of the synflorescence, as well as larger and papillose-hirtellous spikelets ( Table 2).

KEY TO THE MEXICAN SPECIES OF DICHANTHELIUM SECT. MACROCARPA

1. Plants cane-like, 120–140 cm tall; culms 2.7–5 mm diameter; spikelets 3.5–4.3 mm long, glabrous; glandular spots on internodes, sheaths, peduncles and synflorescence axis, branches and pedicels ................................................................ D. multiglandulosum

- Plants cespitose to slightly cane-like, up to 120 cm tall; culms <2.7 mm diameter; spikelets 2.1–4 mm long, pubescent, sometimes glabrescent with a few scattered hairs; glandular spots on some organs but never all of them ....................................................... 2

2. First glume 1.8–2.6 mm long; collar densely short pilose; peduncle, axils and branches of the synflorescence pubescent; spikelets green, rarely purple tinged, attenuate at the base ........................................................................................................... D. transiens

- First glume 0.8–1.8 mm long; collar glabrous; peduncle, axils and branches of the synflorescence glabrous; spikelets usually purple, rounded at the base ....................................................................................................................................... D. commutatum

Figures

FIGURE 3. Micromorphology of the spikelet and fertile lemma of Dichanthelium multiglandulosum (A. Spikelet second glume view. B. Fertile lemma. C. Fertile lemma apex) and D. transiens (D. Spikelet second glume view. E. Fertile lemma. F. Fertile lemma apex).

FIGURE 5. Dichanthelium transiens. A. Slender cane-like habit showing a multi-flowered synflorescence. B. Collar area. C. Ligule area. D. Few flowered synflorescence. E. Spikelet showing the first glume. F. Spikelet showing the second glume. G. Fertile lemma. (Based on Díaz & Carranza 6662 (IEB)).