Drimia edwardsii N.R.Crouch & Mart.

Crouch, Neil R. & Martínez-Azorín, Mario, 2015, Drimia edwardsii (Asparagaceae, Scilloideae), a new urgineoid species from the Mkhomazi River Valley of eastern South Africa, Phytotaxa 195 (2), pp. 135-144 : 137-142

publication ID	https://doi.org/ 10.11646/phytotaxa.195.2.2
persistent identifier	https://treatment.plazi.org/id/B26F87AE-FFF5-C077-46C4-FF56A4B4B997
treatment provided by	Felipe
scientific name	Drimia edwardsii N.R.Crouch & Mart.
status	sp. nov.

Treatment

Drimia edwardsii N.R.Crouch & Mart. View in CoL -Azorín, sp. nov. ( Figs 1–2 View FIGURE 1 View FIGURE 2 )

The new species differs from both D. delagoensis Baker and Urginea lydenburgensis R.A.Dyer in its bulb comprised of very loose, cucullate, pedicellate scales, rather than having tight-packed scales in at least its core. It differs from D. delagoensis in producing hypogeal solitary rather than semi-epigeal gregarious bulbs, and in producing a single filiform leaf rather than 3–8 broader linear ones. It differs from Urginea lydenburgensis in its narrower leaves, angular-ovoid rather than ovoid bulbs, and bulb scales that are greyish-white rather than pinkish.

Type: — SOUTH AFRICA. KwaZulu-Natal, Port Shepstone (3030 AA): middle Mkhomazi River, on north-facing shale rock ledge, 400 m, 21 September 2011, N. Crouch 1280 (holotype BNRH!, isotype HSMC!).

Herb. Bulbs solitary, hypogeal, angular-ovoid, 70 × 50 mm excluding neck; neck 10–35 mm, prominent; bulb scales 6–10, cucullate, pedicellate, very loose, greyish-white, apices obtuse; roots contractile, branched. Leaf solitary, synanthous, filiform, sometimes flexuous, shallowly grooved adaxially and single to multi-faceted abaxially, 70–360 × 0.6–2 mm, erect, dark green, punctulate, denticulate along facet edges; apex acute, sometimes browning through irregular abscission, leaf bases persistent. Inflorescences 1 per bulb, elongate racemose, 300–620 mm long, erect; peduncle 150–320 × 1.4–2 mm, ca. 18(–56) flowered, basal portion hirtellous, hairs erect to 0.9 mm long; pedicels 2.6–5.2 mm long at anthesis, erect spreading; lowermost bracts weakly spurred (spurs to 0.1 mm long), caducous, navicular, 0.7 mm long. Flowers campanulate, 1 to 4 open at a time, pale brownish with darker keels, opening late morning and fading by evening; tepals oblong, ca. 4.5–6 × 1 mm, canaliculate, pale brownish, keel green with purplebrown median stripe, margins white, apices obtuse, base fused for ca. 0.3 mm, biseriate with blades of outermost series overlapping inner, erect below forming a cup ca. 2–2.5 mm long, spreading above, never recurved. Stamens adnate to perianth for ca. 0.5 mm; filaments erect, subterete and tapering, ca. 2.5 mm long; anthers not erect, dorsifixed, dehiscing by longitudinal slits, to 0.5 mm long, yellow with yellow pollen. Ovary ellipsoid, 1.5 × 1.3 mm, truncate; style columnar, ca. 2.5 mm long, white, apically truncate; stigma minutely papillate. Capsule (immature) prolate, ca. 7.5 × 3 mm, pedicels not elongating (4–5 mm long) during frutescence. Seeds (immature) narrowly lanceolate, flattened, ca. 5 × 1.5 mm.

Biology:—Flowering occurs in early spring, August through October in southern Africa. Within single inflorescences, few (1–4) flowers open simultaneously ( Fig. 1A View FIGURE 1 ); the flowers are short-lived. Plants are highly inconspicuous, even when flowering.

Habitat:—The type locality is on a northern aspect at elevations of 380– 400 m. The steep terrain is rocky with exposed dolerite boulders punctuated by small shale outcrops, and plants occur predominantly in shallow clay-rich soils overlying shale, sometimes tightly seated in narrow crevices. The associated flora is dominated by Euphorbia tetragona Haworth (1827: 276) , Spirostachys africana Sonder (1850: 23) , Aloe candelabrum Berger (1906: 246) and Asparagus divaricatus Rottler ex Baker (1875b: 618) , in Eastern Valley Bushveld (SVs 6) within the Savanna Biome ( Rutherford et al. 2006).

Distribution:—Known so far only from the central Mkhomazi River Valley in southern KwaZulu-Natal ( Fig. 3 View FIGURE 3 ). Although presently appearing highly endemic, plants possibly occur in nearby tributaries of the Mkhomazi, as well as the adjacent Mzimkulu River system to the south.

Taxonomic relationships:—The new species shows affinities with U. lydenburgensis and D. delagoensis , both allopatric and readily distinguishable on vegetative characters alone. The most characteristic feature of D. edwardsii is its hypogeal bulb comprised of loose, pedicellate, cucullate greyish-white scales ( Figs 1B View FIGURE 1 , 2H View FIGURE 2 ). Urginea lydenburgensis is also hypogeal and 1–2-leaved ( Fig. 4D View FIGURE 4 ), but produces a pinkish ovoid bulb with smooth compact scales ( Fig. 4F View FIGURE 4 ), not pedicellate. Drimia delagoensis has bulbs without necks and that are mostly epigeal, with thick, fleshy scales somewhat loose-packed peripherally, not pedicellate. The exposed bulb scales are dark olive-green and typically wither at their apices, which turn reddish-brown ( Fig. 4B View FIGURE 4 ). Whereas the bulbs of D. delagoensis are characteristically clumpforming, those of U. lydenburgensis seldom do so ( Fig. 4D View FIGURE 4 ) and D. edwardsii is invariably solitary ( Fig. 1E View FIGURE 1 ). Drimia delagoensis usually presents 3–6(–8) leaves that are broad basally, deeply channeled above and rounded beneath ( Fig. 4B, 4C View FIGURE 4 ), unlike the unifoliate D. edwardsii , the leaves of which are filiform, shallowly grooved above and faceted beneath ( Fig. 2A, 2B View FIGURE 2 ). These and further distinguishing characteristics are given in Table 1. Whereas Jessop (1977) synonymized U. lydenburgensis under D. delagoensis , we concur with Reid (1993) that it is a distinct species. Besides morphological characters ( Table 1), field observations, ethnomedicinal trade research in Nelspruit (Mpumalanga) and phytochemical profiling ( Koorbanally et al. 2005, Crouch et al. 2006) support this distinction.

Although the bulb of D. edwardsii is remarkably reminiscent of that produced by D. haworthioides , a hysteranthous taxon from the Eastern Cape Province ( Jessop 1977), the latter is readily separable on account of its short lanceolate leaves with hairy margins, and perianth segments that reflex strongly rather than spread, corresponding well with the flower morphology of Drimia sensu stricto, the type species of which is D. elata Jacquin (1797: 38) .

The transfer to Sekanama of D. delagoensis was anomalous, for Speta (2001) seemingly overlooked bulb tunic and deciduous differences with S. sanguinea ( Fig. 4H View FIGURE 4 ). He concurred with Jessop (1977) that D. delagoensis “is closely related to D. sanguinea , with which it shares the property of being toxic”. However, D. delagoensis fide Baker (1897) is not poisonous to stock ( Dyer 1942b), an observation corroborated by Koorbanally et al. (2005) who unexpectedly isolated a homoisoflavanone rather than bufadienolides (cardiac glycosides), a class normally yielded by urgineoids ( Mulholland et al. 2013). By contrast, Urginea lydenburgensis is toxic to stock ( Dyer 1942a) and has yielded bufadienolides ( Crouch et al. 2006), which disrupt heart function. Accordingly, the ‘toxicity’ of D. delagoensis has been conferred only through the subsuming of U. lydenburgensis by Jessop (1977) in an expanded species concept.

D. edwardsii and its closest allies ( Table 1) all lack a prominent bulb tunic and share in common small, campanulate, pale-brownish flowers appearing together with the leaves. These characters readily set them apart from the proteranthous S. sanguinea , the tepal lobes of which are white, the anthers green, and the inflorescences dense and many-flowered ( Fig. 4G View FIGURE 4 ).

Etymology:—The specific epithet honours Trevor J. Edwards, in recognition of his mentoring of Hyacinthaceae taxonomists in southern Africa.