Boosia Speta, 2001

3. Boosia Speta View in CoL View at ENA

in Stapfia View in CoL 75: 168 (2001) ( Figs 13–15 View FIGURE 13 View FIGURE 14 View FIGURE 15 ). Typus generis:— Boosia macrocentra (Baker) Speta View in CoL (holotype).

= Drimia sect. Macrocentrae J.C.Manning & Goldblatt in Strelitzia 40: 19 (2018) pro parte. Typus sectionis:— Drimia macrocentra View in CoL (Baker) Jessop (holotype).

Description:—Bulbous geophyte. Bulb hypogeal or partially epigeal. Roots thickened and branched. Leaves 1 to several per bulb, hysteranthous or synanthous, narrowly-linear and elongated, terete, glabrous, smooth, 5‒60 cm long and 2‒15 mm wide, usually purple-red at base. Inflorescence usually multiflowered long and narrow raceme, rarely short and subcorymbose, axes of inflorescence straight or zig-zag; peduncle terete commonly wiry, sinuous at base and purplish, smooth, and glabrous; pedicels usually short and curved, to 25 mm long. Bracts lanceolate, acute, up to 8 mm long, lowermost with characteristic very long basal spur, up to 5 cm long, usually bifid and clasping stem; bracteoles absent. Flowers stellate, erect-patent, diurnal. Tepals 6, biseriate, 3.5‒8.0 mm long, free from base, outer tepals ovate-lanceolate, inner tepals oblong-lanceolate to obovate; adaxial side of tepals white; abaxial side of tepals white with narrow central and longitudinal band commonly purplish-green. Stamens 6, erect to spreading, not connivent to style; filaments slightly flattened, free, smooth; anthers dorsifixed, dehiscing by longitudinal slits along their whole length. Ovary ovoid to oblong, trigonous, green, sometimes with white apical or lateral markings, truncate to style. Style white, narrow, erect, straight. Stigma small, papillose, trigonous. Capsule from narrowly ovoid to subellipsoidal, 3‒11 mm long, triloculate, loculicide, valves splitting to base, with withered perigone segments circumscissile below and forming an apical cap, pedicels of dehiscent dry capsules long lasting, green and photosynthetic. Seeds numerous, narrowly ovate-hemidiscoid or fusiform, 2.0‒ 6.5 mm long, commonly elongate, about 2 to 5 times longer than wide, testa black, glossy, longitudinally striate, with commonly elongate cells.

Number of species and distribution:— Boosia includes 11 species, occurring in southern and east Africa, which are restricted to the Uzambara-Zululand Region and the Southern and Eastern Sections of the Zambezian Subregion (sensu Takhtajan 1986 and Martínez-Azorín et al. 2023a). The highest diversity in the genus occurs in the southeastern regions of South Africa, in subtropical areas in the KwaZulu-Natal Province ( Fig. 8 View FIGURE 8 ), with some species extending north to East Africa.Although we preliminarily accept most described taxa in the group as different species, the taxonomy of this genus is in urgent need of revision to explore their morphological variability across its whole distribution range. For further information on Boosia species see Manning & Goldblatt (2018).

Karyology:—2n=20 ( De Wet 1957, as Urginea tenella Baker ); 2n=40 ( De Wet 1957, as Urginea pretoriensis ); 2n=40 plus 2 fragments ( De Wet 1957, as Urginea rubella ).

History, diagnostic characters, and taxonomic relationships:— Speta (2001) placed Urginea macrocentra ( Figs 13.1 View FIGURE 13 , 14.2 View FIGURE 14 ) into the monotypic Boosia , based on its single, terete, thick, and corky proteranthous leaf, short bracts with very long, coloured, flattened spurs up to 3.8 cm long, elongated raceme with stellate flowers, tepals very shortly connate at their base, and flattened seeds. He also suggested that Drimia modesta (Baker) Jessop represented the sister group and questioned the suitability of applying the name Urgineopsis to this group. The phylogenetic analyses of Pfosser & Speta (2001, 2004) included two “ Boosia ” samples, one labelled “H840 B. macrocentra ” from Barkly Pass and the other “H847 Boosia sp. ” from Swellendam, and revealed their sister relationship, with strong support. Those latter two samples were sister to a clade including two samples labelled “H852 Urgineopsis cf. modesta ” from Nieuwoudtville, South Africa, and “H924 Urgineopsis sp. aff. tenellum ” from cultivated material. As explained by Martínez-Azorín et al. (2023a), neither of the two samples named Urgineopsis in their analyses correspond to the concept applied to that genus in the present work.

The recent phylogenetic studies by Martínez-Azorín et al. (2023a) covered 16 samples related to Boosia macrocentra and found those to form a strongly supported clade, which is sister to Geschollia . The 16 samples were divided into two subclades that can be interpreted in biogeographic terms. One of the clades comprises 12 samples from eastern South Africa (mostly KwaZulu-Natal) and Lesotho that includes species such as Urginea saniensis Hilliard & Burtt (1985: 253) , U. modesta Baker (1892: 6) , U. natalensis Baker (1897: 468) , U. tenella Baker (1897: 464) , and Drimia flagellaris Edwards et al. (2005: 122) ; these share with U. macrocentra terete leaves, usually long-spurred basal bracts, and flattened, elongated seeds. This group of species is here accepted as the genus Boosia , which occur along eastern South Africa to East Africa and can be recognised by a combination of the following characters: terete leaves; usually long racemose inflorescence with sinuous and purplish peduncle base; lowermost bracts commonly with long spurs, usually bifid and clasping the stem; bracteoles absent; flowers stellate with free tepals; spreading stamens; pedicels of dehiscent, dry, yellow capsules and inflorescence peduncle green, long lasting and photosynthetic; withered tepals persisting as a cap at the top of the developing capsules; and the commonly elongated and striate seeds. Boosia is sister to Geschollia , a lineage that differs in the usually single leaf, pedicels drying simultaneoulsy with capsules, comparatively very small capsules and seeds, and in its centre of diversity being located in the Eastern Cape Province of South Africa ( Martínez-Azorín et al. 2019d).

The other sister subclade to Boosia in Martínez-Azorín et al. (2023a) includes samples from western South Africa and comprises two samples of Fusifilum magicum M̧ller-Doblies et al. (2001: 491), the Pfosser and Speta sample “H847 Boosia sp. ” from Swellendam, and a sample from Betty’s Bay, South Africa of Urginea revoluta . The identity and taxonomy of this latter morphologically heterogeneous subclade remains uncertain until further samples and additional molecular markers are included in the phylogenetic analyses.

Accepted species and required new combinations:—

Boosia flagellaris (T.J.Edwards, D.Styles & N.R.Crouch) Mart. - Azorín, N.R. Crouch, M.B.Crespo & M.Á.Alonso comb. nov. ≡ Drimia flagellaris T.J.Edwards, D.Styles & N.R.Crouch View in CoL in S. African J. Bot. 71(1): 122 (2005), basionym ( Figs 1.9 View FIGURE 1 , 13.2 View FIGURE 13 , 14.1 View FIGURE 14 ). Type:— SOUTH AFRICA. KwaZulu-Natal. Pietermaritzburg (2930): Krantzkloof , in cliff faces, (–DD), 28 July 2003, N.R. Crouch 1023 (NU0016039! holo.; K000400681!, NH!, PRE! iso.).

Boosia macrocentra (Baker) Speta View in CoL in Stapfia View in CoL 75: 169 (2001) ≡ Urginea macrocentra Baker View in CoL in Gard. Chron. 1887, 1: 702 (1887), basionym ( Figs 1.10 View FIGURE 1 , 13.1 View FIGURE 13 , 14.2 View FIGURE 14 ). Type:— SOUTH AFRICA. Eastern Cape, Transkei, without date, Barber 895 (K000400573! holo.).

= Urginea lilacina Baker, Fl. Cap. (Harvey) View in CoL 6: 469 (1897). Type:— SOUTH AFRICA. KwaZulu-Natal. Pietermaritzburg (2930): Inanda , (–DB), September 1881, J. Medley Wood 642 (K000257362! lecto. designated by Jessop in J. S. African Bot. 43: 292. 1977; BM000911789!, NH 0005786-0!, SAM! isolecto.).

= Urginea schlechteri Baker View in CoL in Bull. Herb. Boissier ser. 2, 4: 1000 (1904). Type:— SOUTH AFRICA. KwaZulu-Natal. Stanger (2931): Claremont , (–CC), 1 August 1893, Schlechter 3155 (BOL! lecto. designated here; GRA!, K000257372!, PRE0049746-0!, Z000027554! isolecto.).

Boosia mandalensis (Baker) Mart. - Azorín, N.R. Crouch, M.B.Crespo & M.Á.Alonso comb. nov. ≡ Urginea mandalensis Baker, Fl. Trop. Afr. [Oliver et al.] 7(3): 536 (1898), basionym. Type:— MALAWI. [formerly Nyasaland], Ndirandi Mountain near Blantyre, Scott-Elliot 8476 (K000400582! holo.).

Boosia modesta (Baker) Mart. -Azorín, N.R. Crouch, M.B.Crespo & M.Á.Alonso comb. nov. ≡ Urginea modesta Baker in Bot. Jahrb. Syst. 15(3, Beibl. 35): 6 (1892), basionym ≡ Urgineopsis modesta (Baker) Speta in Linzer Biol. Beitr. 12(1): 205, 230 (1980), comb. inval. (lacking basionym reference) ( Figs 1.11 View FIGURE 1 , 14.3 View FIGURE 14 ). Type:— SOUTH AFRICA. Eastern Cape, Pondoland, Bachmann 273 (K000099182! holo.).

Boosia natalensis (Baker) Mart.- Azorín, N.R. Crouch, M.B.Crespo & M.Á.Alonso comb. nov. ≡ Urginea natalensis Baker, Fl. Cap. (Harvey) 6(3): 468 (1897), basionym ( Figs 1.12 View FIGURE 1 , 14.4 View FIGURE 14 ). Type:— SOUTH AFRICA. KwaZulu-Natal. Pietermaritzburg (2930): Inanda, (–DB), June [18]79, J. Medley Wood 277 (K000257347! lecto. designated by Jessop in J. S. African Bot. 43: 303. 1977; NH0004740- 0! isolecto.).

Boosia nyasae (Baker) Mart. - Azorín, N.R. Crouch, M.B.Crespo & M.Á.Alonso comb. nov. ≡ Urginea nyasae Rendle in Trans. Linn. Soc. London, Bot. 4(1): 50 (1894), basionym ( Figs 1.13 View FIGURE 1 , 15.1 View FIGURE 15 ). Type:— MALAWI. Milanji , elev. 6000 ft., October 1891, A. Whyte s.n. (BM000911782! lecto. designated by Manning & Goldblatt in Bothalia 43(1): 76. 2013).

Boosia pretoriensis (Baker) Mart.- Azorín, N.R. Crouch, M.B.Crespo & M.Á.Alonso comb. nov. ≡ Urginea pretoriensis Baker in Bull. Herb. Boissier ser. 2, 1: 786 (1901), basionym. Type:— SOUTH AFRICA. Transvaal [Gauteng], Pretoria, Colle supra Aapies [Apies] River, A. Rehmann 4307 (Z000027551! holo.).

Boosia rubella (Baker) Mart. - Azorín, N.R. Crouch, M.B.Crespo & M.Á.Alonso comb. nov. ≡ Urginea rubella Baker, Fl. Cap. (Harvey) 6(3): 467 (1897), basionym ( Figs 1.14 View FIGURE 1 , 15.2 View FIGURE 15 ). Type:— SOUTH AFRICA. KwaZulu-Natal. Pietermaritzburg (2930): near the Mooi River , (–AA), elev. 4000–5000 ft., 22 October 1894, J. Medley Wood 5723 (K000099183! lecto. designated as “holo.” by Jessop in J. S. African Bot. 43: 303. 1977; BM001122598!, BOL140323!, SAM0023220-0!, PRE0048621-0! isolecto.).

Boosia saniensis (Hilliard & B.L.Burtt) Mart. -Azorín, N.R. Crouch, M.B.Crespo & M.Á.Alonso comb. nov. ≡ Urginea saniensis Hilliard & B.L.Burtt View in CoL in Notes Roy. Bot. Gard. Edinburgh 42(2): 253 (1985), basionym ≡ Drimia saniensis (Hilliard & B.L.Burtt) J.C.Manning & Goldblatt View in CoL in Bothalia 33(1): 111 (2003) ≡ Urginea saniensis Hilliard & B.L.Burtt View in CoL in Taxon 63: 1332 (2014), nom. superfl. ( Figs 1.15 View FIGURE 1 , 15.3 View FIGURE 15 ). Type:— LESOTHO. Underberg (2929): top of Sani Pass, (–CA), 6 November 1973, Hilliard & Burtt 7102 (E00193964! lecto. designated here: the eight bulbs with both flowers and leaves enclosed into the envelope placed in the central part of the sheet; NU0015641! isolecto.).

Comments:—In the light of the recentmost proposal by Mosyakin & McNeill (2022) to amend Art. 8 of the ICN, after which the information in the protologue will rule, and considering that the protologue in Hilliard & Burtt (1985) only indicates a single collecting date (06 November 1973) in the holotype designation, a lectotypification is needed from the mixed type collection E00193964. Based on the new considerations, the description made by Martínez-Azorín & Crespo (2014: 1332) is therefore nomenclaturally inoperative, since it is either superfluous or not valid.

Boosia tenella (Baker) Mart. - Azorín, N.R. Crouch, M.B.Crespo & M.Á.Alonso comb. nov. ≡ Urginea tenella Baker, Fl. Cap. (Harvey) 6(3): 464 (1897), basionym ( Figs 1.16 View FIGURE 1 , 15.4 View FIGURE 15 ). Type:— SOUTH AFRICA. KwaZulu-Natal. Harrismith (2829): Van Reenen, (–AD), elev. 5000 ft., 17 December 1891, J. Medley Wood 4562 (K000257340! lecto. designated as “ type ” by Hilliard & Burtt in Notes Roy. Bot. Gard. Edinburgh 40: 286. 1982; NH0006437-0! isolecto.).

Boosia umgeniensis (Poelln.) Mart. -Azorín, N.R. Crouch, M.B.Crespo & M.Á.Alonso comb. nov. ≡ Urginea umgeniensis Poelln. View in CoL in Ber. Deutsch. Bot. Ges. 61: 209 (1944), basionym ≡ Urginea pauciflora Baker in Bull. Herb. Boissier ser. 2, 1: 786 (1901), nom. illeg. [non Baker, Fl. Trop. Afr. [Oliver et al.] 7(3): 539 (1898)]. Type:— SOUTH AFRICA. KwaZulu-Natal. Pietermaritzburg (2930): Umgeni Falls, (–AC), A. Rehmann 7455 (Z s.n., first-step lecto. designated as “holo.” by Manning & Goldblatt in Strelitzia 40: 36. 2018; second-step lecto. designated here:—Z000027550! lecto.; Z000027549! isolecto.). Note:—A second step lectotypification is needed due to the presence of two herbarium vouchers with different barcode numbers under the same collection.