Taxonomic notes and new species of the southern African genus Babiana ( Iridaceae : Crocoideae )

A member of Iridaceae subfamily Crocoideae. Babiana Ker Gawl. comprises some 80 species from southern Africa. Field studies have shown the need for several taxonomic and nomenclatural changes, w hile a number of new species have been discovered. The type of B flabellifolia Harv. ex Klatt is a short-tubed plant that matches B. truncata GJ.Lewis. and that name falls into synonymy. The name B. flabellifolia sensu GJ.Lewis (1959) has been misapplied to long-tubed plants from the western Karoo which are now renamed B. praem orsa sp. nov. In addition. B. truncata, as originally circumscribed, included two species, one short-tubed and a second with a longer tube, which we describe here as B. cuneata sp. nov. The type off? hypogaea Burch, has also been misinterpreted and matches the species described as B.flavida. which thus falls into the synonymy of B hypogaea. A second species. B.falcata GJ.Lewis. closely matches this species and is also reduced to synonymy. The widespread southern African species long known as B hypogaea (hypogea sensu GJ.Lewis) matches the type ol B. bainesii Baker and must now be known by that name. Babiana stricta var. erectifolia (GJ.Lewis) GJ.Lewis is appropriately included in typical B. stricta (Aiton) Ker Gawl. However, var. regia GJ.Lewis is a very different plant and is treated as a separate species. B. regia comb, et stat. nov.. as is long-tubed var. grandiflora GJ.Lewis. which is described as the new species, B. longiflora sp. nov. Plants included in B. stricta var. sulphurea sensu GJ.Lewis are also included in var. stricta. We are unable to match the type of Gladiolus sulphureus Jacq.. basionym of var. sulphurea (Jacq.) Baker, w ith any known species and the name is thus excluded. Lastly, the name B. disticha Ker Gaw l., type of the genus, is superfluous for Gladiolus fragrans Jacq. and the new combination B fragrans comb. nov. is made, reducing B. disticha to synonymy. Babiana fragrans Eckl., which was thought to prevent the transfer of G. fragrans. is a nomen nudum and thus invalid and cannot be taken into account in considerations of nomenclatural priority.


INTRODUCTION
When last revised in 1959 by the South African botanist.G J. Lewis, the genus Babiana Ker Gawl.(Iridaceae: sub family Crocoideae) was considered to comprise 60 species in southern Africa and one on the Indian Ocean island of Socotra.Since the publication of this monograph, two species.B. lewisiana B.Nord.(Nordenstam 1970) and B. virginea Goldblatt (Goldblatt 1979) have been described, and a further two.Antholyza plicata Thunb.and A. ringens L. have been restored to Babiana as B thunbergii Ker Gawl.and B. ringens (L.) Ker Gawl.respectively (Goldblatt 1990).The latter two species were included in Babiana by most 19th century botanists, notably Ker Gawler (1805) and Baker (1892;1896), although not by Lewis (1959).who fol lowed Brown (1932) in maintaining Antholyza L. for B. rin gens, and Anaclanthe N.E.Br.for B. thunbergii and its syn onym. A. namaquensis N.E.Br.The Socotran species, B socotrana Hook.f., is now known to be allied to Lapeirousia Pourr.and Savannosiphon Goldblatt & Marais.and has been referred to the new genus Cyanixia Goldblatt & J.C.Manning.asC.socotrana (Hookf.)Goldblatt & J.C .Manning (Goldblatt et al. 2004).Currently then.Babiana includes 64 species, all from southern Africa, mainly the w inter rainfall zone of the south and west of the subcontinent.
Field research, especially since 1995.often conducted in conjunction with our studies of the pollination biology of Babiana and other genera of the Iridaceae.has led to the realization that there are several shortcomings in Lewis's (1959) account of the genus as well as a number of undescribed species.In this paper we deal with changes to the taxonomy of the B. flabellifolia, B. hypogaea and B. stricta species complexes, describe two new species, and make two new combinations resulting from these changes.
Examination of the type specimen of Babiana flabelli folia Harv.ex Klatt.most likely collected in northern Namaqualand.reveals that it does not match the western Karoo plant that Lewis (1959) associated with this name.The latter plant, which is distinguished by a long perianth tube and short fdaments.is renamed B. praemorsa sp.nov.Accordingly, the short-tubed Namaqualand species B. truncata GJ.Lewis.which closely matches B. flabellifo lia.assumes this earlier epithet.Our studies also show that plants referred by Lewis to B. truncata represent two species, one largely from Namaqualand with a perianth tube 18-36 mm long, and a second, described here as B. cuneata sp.nov., with a perianth tube 40-60 mm long.Similarly, the type of Babiana hypogaea Burch, (spelled hypogea by Lewis), collected by William Burchell in 1812.closely matches the species described as B.flavida by Lewis in general appearance and perianth dimensions, and differs significantly from the widespread southern African plant currently called B. hypogea sensu Lewis (1959).The latter must now assume the name B. bainesii Baker, and B. flavida GJ.Lewis falls into the synonymy of B hypogaea.The revised taxonomy accords with the treatment by Baker ( 1896) in Flora capensis.
In addition, taxonomic changes are needed for the southwestern Cape species.Babiana stricta (Aiton) Ker Gawl., a complex in which Lewis (1959) recognized five varieties.Some of these closely resemble the type of B. stricta but others differ strikingly from that species.No single exclusive feature links the five varieties, except for the more or less subequal tepals and, with the excep tion of var.grandiflora G.J.Lewis, the relatively narrow, rigid leaves with a short rough pubescence.Neither of these features is unique to this complex.Lastly, Gladiolus fragrans Jacq. is an earlier name for Babiana disticha Ker Gawl., and we make the nomenclatural cor rection in the new combination B. fragrans.The nomen clatural and taxonomic consequences of these conclu sions are dealt with formally below.
A more extensive account of Babiana is planned for a future paper in which we will describe additional new species and present a revised infrageneric classification of the genus.Including the new species described below and those in manuscript, Babiana comprises 80 species.This makes it one of the larger genera of southern African Iridaceae, after Gladiolus L. (165 species in southern Africa) and Moraea Mill.(147 species), and slightly larger than Romulea Maratti (78 species) and Hesperantha Ker Gawl.(77 species).

The Babiana flabellifolia complex
Defined by its unusual, abruptly truncated leaves and by its large floral bracts with the inner notched at the tips, and so provisionally considered a clade by us, the B. fla bellifolia complex currently includes three species.These are the central and northern Namaqualand species, B. pubescens (Lam.)G.J.Lewis, which has an elongate perianth tube ± 50 mm long and curved near the apex, and a dorsal tepal ± 2 0 mm long; a second long-tubed species from the western Karoo, identified as B. flabelli folia by Lewis (1959); and a widespread, relatively short-tubed species, named B. truncata by Lewis (1959).
The type of Babiana flabellifolia was collected by the Rev. Henry Whitehead at an unrecorded locality in Namaqualand.Whitehead, an Anglican clergyman, was sta tioned at Modderfontein near Springbok, and his collections are assumed to have been made nearby in northern Namaqualand (Gunn & Codd 1981).The type of B. flabelli folia, however, exactly matches B truncata G J Lewis (1959) in its short tube, except for conspicuous pubescence on the leaves and bracts, and is from Namaqualand where B trun cata is common, and where the long-tubed western Karoo species does not occur.B. truncata is typically glabrescent.The latter taxon extends from Steinkopf in northern Namaqualand to the Bokkeveld Plateau near Nieuwoudtville in the south and its range thus includes that of B. flabellifolia.We suggest that the glabrescent and pubescent plants with perianth tubes of similar length represent a single species, which must carry the earlier name.The conspicuous pubes cence of the type specimen is certainly unusual but sparse pubescence is occasionally encountered in plants matching B. truncata, especially in the immediate vicinity of Springbok.We do not regard the hairiness of the type specimen as sig nificant enough to treat it as a separate taxon.We rename the western Karoo plant B. praenwrsa here.
In addition to a name change for the short-tubed Namaqualand species, which must now be called Babiana flabellifolia, a change to its circumscription seems necessary.Some plants from the south of the range of the species have a perianth tube 40-60 mm long and large subequal tepals.We have concluded that they rep resent a separate species, which we here describe as B. cuneata.This species extends from Lokenberg in the Bokkeveld Mountains to the Cold Bokkeveld and Swartruggens near Ceres in the south, and along the Roggeveld Escarpment to near Laingsburg in the east.Plants 80-150 mm high including leaves.Stem short, hairless.Leaves crowded at base, up to 6 , oblonc, (6-)8-15 x 8-18(-30) mm.leathery, plicate, abruptly truncate, hairless or finely velvety hairy.Spike compact, 2-5(-7)-flowered, flowers borne at ground level: bracts 25-50 mm long, green with brown attenuate tips, smooth or sparsely hairy, inner about as long as outer and forked apically, basal margins sometimes fused around ovary for ± 2 mm.Flowers zygomorphic.pale bluish to violet, lower lateral tepals with white median marks outlined with dark violet to purple below, unscented or sweetly scented; perianth tube 40-60 mm long, cylindrical, straight, widening slightly in upper 8 mm: tepals spread ing, subequal or three upper slightly larger.26-30(-40) x 6-12 mm.upper laterals fused to lower for ± 2 mm.Stamens unilateral: filaments 15-18 mm long, exserted 9-10 mm; anthers 6-8(-11) mm long, pale yellow to cream-coloured.Ovary smooth, shortly stipitate in lower flowers; style dividing opposite anther bases or apices, branches ± 6 mm long, arching outward between or over anthers.Flowering time: August to late September.Figure 1.
Distribution: dry rocky sandstone or dolerite slopes and flats in arid fynbos or renosterveld, from the Bokkeveld Mountains to the interior Cold Bokkeveld as far south as Karoopoort.and eastward on the Roggeveld Escarpment south to the foot of the Witteberg near Laingsburg (Figure 2).
Diagnosis and relationships: Babiana cuneata was previously included by Lewis (1959) in the species that she called Babiana truncata (now B flabellifolia) but comparison of live plants examined in the field shows that the populations from the south of the range of that species represent a distinct species, geographically and edaphically separated from B. truncata.They are recog nized by their blue to blue-grey flowers (either fragrant or apparently unscented) with an elongate perianth tube 40-60 mm long; ± equal, ascending tepals 26-30(-40) mm long; and openly displayed stamens with filaments 15-18 mm long and exserted 9-10 mm from the tube.The point of division of the style varies.In the type popu lation from Katbakkies Pass, the branches separate oppo site the base of the anthers but in plants to the south, from below the Roggeveld Escarpment, the style reaches the anther apices.Flowering occurs from early August to late September.Plants here included in B. flabellifolia have in contrast strongly scented, bilabiate flowers with an arched dorsal tepal more or less concealing the stamens, a perianth tube 18-36 mm long, and often dry.submembranous, or occasionally green, leafy bracts.In speci mens where it is possible to see floral details, the style divides opposite the anther apices.Flowering in B. fla bellifolia occurs from late June through July, rarely in early August in seasons when the rains are late or condi tions are particularly cold.
The flowers of Babiana cuneata closely resemble those of B. sambucina.which is widespread across dry parts of the Cape Floral Region, but does not occur in the Roggeveld and typically has lanceolate, acute or acumi nate leaves.
History: Babiana cuneata was evidently first collect ed by Rudolf Schlechter in August 1897 at Papelfontein (sic.)[Papkuilsfontein].some 16 km south of Nieuwoudt ville at the northern end of its range.In 1941 R.H. Compton collected the species nearby, at Lokenburg in the Bokkeveld Mountains, a short distance to the south.The range of B. cuneata falls partly within the arid inte rior of the Cape Floral Region but there are also several recent records from the Roggeveld Escarpment and Klein Roggeveld to the east.Most collections are from rocky habitats, with the substrate rock either sandstone or dolerite.Plants at the southern end of the range, in the Klein Roggeveld south of Komsberg Pass, have unusual ly large flowers with the dorsal tepal ± 40 mm long and anthers 9-11 mm long, and the style is also unusual in dividing opposite the anther apices rather than opposite their bases.These plants are also distinctive in their glau cous leaves, which are often 20-30 mm wide, which is exceptionally broad for the species.Despite these differ ences, however, we see no reason at present to recognize these plants as a separate taxon.Plantae 50-150 mm altae, caule brevi raro usque ad 20 mm supra terram extenso, foliis oblongo-cuneatis plicatis minute villosis, spica 5-vel 6-flora congesta, bracteis viridis minute villosis 25-50 mm longis, interiore parum breviore, exteriore ad apicem furcata, floribus zygomorphis atroviolaceis maculis lanceolatis albis vel cremeis notatis, tubo perianthii 40-60 mm longo cylindrico, tepalis subaequalibus usitate 18-22 x 3.5-5.0mm horizontaliter extensis, filamentis unilateralibus suberec- tis 8-9 mm longis, antheris 5-6 mm longis, ovario glabro.
Distribution: dolerite outcrops in the Calvinia District of the western Karoo, from the Hantamsberg to Bloukrans Pass, often growing in rock crevices in dolerite pavement where the corms are secure from predation.
Diagnosis and relationships: in her account of Babiana, Lewis (1959) treated B. praemorsa as B. fla bellifolia, a species described by F.W. Klatt in 1867-68, based on a specimen, from 'Namaqualand Minor'.This is certainly not the same as the plants from the Calvinia District with which she associated the name.The type specimen (a single plant) at the Kew Herbarium has flowers with a short, obliquely funnel-shaped perianth tube, ± 18 mm long with the narrow part ± 10 mm long, and broadly cuneate leaves that are densely covered with long hairs on the major veins and margins.Except for the prominently hairy leaves and slightly shorter tube, this specimen corresponds exactly to the Namaqualand plant that Lewis called B. truncata, now reduced to synonymy in B. flabellifolia.
One of four species of the Babiana flabellifolia com plex, B. praemorsa is recognized by its flowers with a straight, cylindrical perianth tube 40-60 mm long, and firm, narrow, spreading tepals, 18-22 mm long.The peri anth is deep violet rather than blue as in the related longtubed species of the alliance and they are perhaps most like those of the central Namaqualand B. pubescens, with which it is easily confused.Babiana praemorsa can be distinguished from this and another long-tubed species, B. cuneata from the dry interior of northern Western Cape, by its short filaments, 8-9 mm long, exserted 4-5 mm from the perianth tube, and the tepals spreading at right angles to a nearly straight tube of uniform diameter.
As well as occurring in the Calvinia District, Babiana praemorsa was cited by Lewis as growing at 'Upington, Namaqualand' (Orpen s.n., NBG), and at Whitehill (Compton 13389, NBG) south of Laingsburg.Whitehill was the site of the first Karoo Botanical Garden, no longer extant.The latter plant is B. cuneata from the southern limit of its range.The Upington record is, how ever, almost certainly an error in labelling the prove nance of the plant, as Babiana praemorsa is evidently a narrow endemic of dolerite outcrops in the Calvinia District.Since it was treated by Lewis, B. praemorsa has been collected at several localities not known to her, extending the known range of the species.

The Babiana hypogaea complex
Recognized by the partly underground spike, the flowers thus with a subterranean ovary, the slender, lin ear to falcate leaves, and, according to Lewis (1959), a stipitate ovary, the Babiana hypogaea complex included three species in Lewis's (1959) Lewis (1959) and later authors (e.g.Soldi 1969; Goldblatt 1993), must be called B. bainesii.Confusingly, in the protologue of B. hypogaea, Burchell (1824) described the flower colour as blue (icaeruleis), possibly an error or the result of basing the description on his pressed speci mens in which the tepals have a dull mauve cast, as do all dry specimens of the species.This is a result of the purple pigmentation on the reverse of the tepals, masking the dull yellow colouring of the inside of the tepals.Our revised taxonomy corresponds to the treatment of the complex by Baker (1896) in Flora capensis.
The so-called stipitate ovary described by Lewis (1959) in Babiana hypogaea is misleading.Species of Iridaceae with an inflorescence (typically a spike) borne at or below ground level often have shortly stipitate lower flowers and we suggest that this is simply a devel opmental aberration of the inflorescence due to its posi tion.This feature is associated with the underground inflorescence in other Iridaceae, including Crocus L.f Duthiastrum M.P.deVos and Romulea, and even some other Babiana species with spikes borne at ground level (e.g.B. cuneata and B. sambucina (Jacq.)Ker Gawl.) and does not in itself mark the complex as sharing a synapomorphy other than the underground spike.
Babiana bainesii and B. hypogaea are morphological ly similar in their underground spikes with long-tubed flowers arising below the ground, and in their narrow, linear leaves.The significant features of Babiana hypogaea are the pale yellow to buff flowers with a tube 30-40 mm long, and the shorter, often sparsely hairy to almost smooth leaves, sometimes with inclined to pros trate blades.The blades are often visibly constricted and apparently flexible at the base, which suggests an inclined to prostrate orientation even when they appear upright in pressed specimens.In contrast, B. bainesii has larger flowers with a tube (40-)50-60(-70) mm long, blue-mauve to violet flowers with prominent white markings edged in darker blue on the lower lateral tepals, and densely hairy leaves, mostly 15-25 mm long.Misunderstanding the identity of the type collection, Lewis (1959: 112), united B. bainesii with B. hxpogaea and described B. flavida for pale yellow-to buff-flow ered plants largely of Bushmanland.that mostly flower in the winter and early spring.This species matches closely Burchell's type collection in its smaller flower size, colour and other details, as well as in its habitat and winter to spring flowering.
Babiana hypogaea occurs west and south of the range of B. bainesii, in the northwestern Great Karoo, Bushmanland, and adjacent southeastern Namibia.It mostly flowers in winter and early spring, June to September (rarely as late as December, according to the type of B. flavida), in contrast to February to April or May for B. bainesii.The flowering time and ranges of both species overlap to some extent in the Kimberley-Vryburg area of Northern Cape and North-West but there is no indication that they converge in floral morphology along the area of overlap.Flowering in Babiana hypo gaea appears to be related to rainfall timing, normally coming into bloom in winter and early spring in response to autumn showers but blooming in summer when occa sional good early summer rains fall across Bushmanland and the Upper Karoo.
While Babiana hypogaea and the plant described by Lewis (1959) as B. flavida are essentially identical, another species of Bushmanland with similar flowers.B. falcata G.J.Lewis was distinguished by its short leaves, 25-70 mm long and 1.5-3.0mm wide, that are falcate, firm, strongly plicate, and somewhat pungent.Lewis contrasted the species with B. flavida (i.e.B. hypogaea), which she described in her key to the species as having leaves erect or decumbent but not firm or falcate.The range of specimens now available make the distinction untenable.Leaf shape ranges without any clear division from the short, falcate, and firm type in the west to the suberect.more prominently hairy type in the east.We therefore, include B. falcata in B. hypogaea.Some col lections even show a range of leaf shapes from linear and erect, to sword-shaped, to falcate.Babiana hypogaea Burch.. Travels 2: 415 (1824).Ant holy za hypogaea (Burch.)Klatt in Abhandlungen der Naturforschenden Gesellschaft zu Halle 15: 345 (1882).Type: South Africa, [Northern Cape).Pellat Plains near Takun (or Litakun), 24 August 1812, WJ.Burchell 2241 (K.lecto.!, here designated; B. P. iso.).
Babiana flavida G J. Lewis: 117 (1959).syn.nov.Type: South Africa.Plants acaulescent, 50-80 mm high, often growing in tufts with leaves up to 150 mm long.Stem subterranean, simple or with short or vestigial lateral branches.Leaves linear to falcate or sword-shaped, often bent near base, sometimes inclined toward ground or evidently prostrate, 25-150 x 1.5-3.0mm, slightly pleated, sparsely hairy to finely long-haired, somewhat pungent.Spike arising below ground level, with 1 or 2 flowers per branch, seemingly in a 2-6-flowered spike; bracts smooth, more or less membranous below, green in upper third where they reach above ground, 25-50 mm long, inner as long as or slightly exceeding outer and forked apically with attenuate tips.Flowers zygomorphic, greenish yellow to buff, flushed with pale brown or mauve outside, tepals darker in midline especially when dry, lower tepals with pale nectar guides edged with reddish arrow-shaped marks near base and with dark reddish streaks in throat, strongly scented; perianth tube emerging from below ground, cylindric, widening at throat, 30-40 mm long; tepals subequal, dorsal 35-42 mm long, lower tepals joined to upper laterals for up to 6 mm, forming a promi nent lip.lower tepals 30-35 mm long.Stamens unilater al; filaments 15-18 mm long; anthers 8-11 mm long.Ovary smooth, shortly stipitate in lower flowers; style dividing close to anther apices.
Plants mostly 150-250 mm high (leaves only), some times growing in tufts.Stem underground, often branched.Leaves linear to sword-shaped, pleated, much exceeding flowers, 15-25 x 3-10 mm.densely hairy, scabrid, or virtually smooth.Semi-spike compact.2-8flowered.borne below ground level, lateral flowers usu ally with short stalks; bracts membranous to papery with dry, rusty tips or dry and rusty throughout, sparsely hairy, mostly 35-60 mm long, inner slightly shorter than outer and forked at tips.Flow ers zygomorphic, shades of blue to violet or mauve with tepals paler (rarely w hite) toward edges, lower lateral tepals with white markings often edged in dark blue, usually sweetly scented; perianth tube mostly 40-60(-70) mm long, cylindric with expanded throat; dorsal tepal 40-50 mm long, upper lat erals joined to lower for ± 6 mm to form a lip.lower tepals 25^4-0 mm long.Stamens unilateral; filaments 10-15 mm long; anthers 8-10(-12) mm long.Ovary smooth, shortly stipitate in lower flowers; style usually dividing more or less opposite anther apices, branches ± 6 mm long.Flow ering time: mainly in summer.February to April, occasionally in December or January; southern Karoo populations flower in August or September.
Distribution: stony or sandy slopes and flats in dry grassland and bush, across summer-rainfall southern Africa from Murraysburg and Carnarvon in the Upper Karoo to Zimbabwe and southern Zambia in the north east, and to northern Namibia in the northwest.
The most widespread species of the genus, Babiana bainesii extends from the central Upper Karoo to south ern Zambia (Lewis 1959: Goldblatt 1993).thus having a range greater than the rest of the genus.It grows in grass land on soils ranging from deep Kalahari sands to stony slopes and flats, and typically flowers from late summer until late autumn, the exact timing in dry areas probably depending on rainfall in areas of low annual rainfall.With their narrow, erect leaves, plants are difficult to see among the grass where they grow, but w hen in bloom can often be located by the strong, sweet scent of the bluish to violet flowers.Lewis (1959) recognized two additional varieties of B hypogea: var.ensifolia GJ.Lewis from the Upper Karoo, and var.longituba G J.Lewis from the mountains of Limpopo.The latter was distinguished by its densely hairy leaves and bracts, and elongate perianth tube up to 70 mm long.The long tube in these plants is only the extreme of a range that includes many plants w ith a tube 55-65 mm long (the type of B. bainesii from the Witwatersrand has a tube nearly 60 mm long).Near Haenertsburg in Limpopo, close to the type locality of var.longituba.plants have a perianth tube 35-59 mm long (n = 15) (P.J.D. Winter pers.comm.).Information now available indicates that var.longituba does not merit taxonomic recognition and we reduce it to synonymy in B. bainesii.Babiana hypogea var.ensifolia has comparatively short, broad leaves.50-140 x 10-14 mm.unlike the nar row, sparsely to densely hairy, narrowly sword-shaped to linear leaves, mostly 150-250 x 3-10 mm, of typical B. bainesii.It approaches in general appearance the largely southern Cape species.B. sambucina but has the entirely subterranean inflorescence, short lateral branchlets.and rusty bracts of B. bainesii.These plants flower in early spring.August and September.Var.ensifolia is poorly documented and requires additional study before its sta tus can be established.It may be recognized or not at the discretion of the reader but we hesitate to make the new combination in B. bainesii until more is learned about these plants from the central parts of the Upper Karoo.
Babiana schlechteri.described by J.G. Baker in 1901, and based on a collection from Witbank.east of Johannesburg, does not differ in any significant way from the type of B. bainesii from near Johannesburg.
3. The Babiana stricta complex Lewis (1959) recognized five varieties in the south western Cape species.Babiana stricta.all with the inner bracts divided to the base and a densely hairy ovary, fea tures present in many other species.Listed in the se quence presented by Lewis, the varieties and their sig nificant taxonomic features are as follows: 1. var.stricta-leaves stiff, narrow, with short rough pubescence; stem suberect: flowers zygomporphic.blue; tepals subequal and spreading, lower three with white markings; anthers unilateral [incorrectly said by Lewis to be symmetrically arranged], blue, arrow-shaped with connective wider toward base; style mostly dividing opposite the anther bases.
2. var.erectifolia -similar to var.stricta but flowers white to cream-coloured, also with spreading tepals, lower with pale yellow markings: anthers unilateral, dark blue-black, w ith connective broader below.
3. var.sulphurea-similar to var.stricta but flowers yellow with unilateral dark blue anthers w ith connective broader toward the base; 4. var.regia-strikingly different from the above three varieties-leaves stiff, narrow, with short pubes cence; stem spreading horizontally; flowers radially symmetric, blue with dark red centre: tepals cupped: anthers linear, dark brown, without visible connective: style short, dividing below' bases of anthers. 5. var.grandiflora-leaves broad w ith soft pubescence: stem suberect: flowers subactinomorphic.blue-mauve: tepals spreading, perianth tube 22-32 mm long; anthers unilateral.darkblue, linear with narrow connective visible: style dividing opposite middle and apex of anthers.
We see no merit in this taxonomy, in which Babiana stricta has a w ider circumscription than any other species of the genus and seems to comprise disparate elements.Var.grandiflora.which we have re-collected at its type locality north of Piketberg.evidently represents a distinct species.Its broad, soft-textured leaves, and faintly scent ed flower with relatively long perianth tube, hollow to the base, suggests that it is more closely related to B. disticha, the type of the genus.Likewise, the features of var.regia are so distinctive that it too must represent a separate species, not obviously immediately related to B. stricta.
In contrast to these two examples, we see no reason to separate Babiana stricta var.erectifolia from var. stricta.and thus unite them.Lastly, plants assigned to var.sulphurea seem to us merely cream-coloured to pale yellow-flowered variants of B. stricta but we cannot identify the type of Gladiolus sulphureus, the basionym of var.sulphurea.
Babiana stricta var.erectifolia (GJ.Lewis) GJ.Lewis: 43 (1959).B. erectifolia GJ.Lewis: 3 (1938) The complete synonymy of Babiana stricta is present ed by Lewis (1959: 40) and is not repeated here.Note that we also include B. stricta var.sulphurea sensu GJ.Lewis (based on Gladiolus sulphureus Jacq.) in B. stricta but do not regard the type, an illustration in Jacquin's leones plantarum rariorum (1793), as matching B. stricta and are in fact unable to recognize the plant illustrated.
As here circumscribed, Babiana stricta comprises plants with narrow, fairly stiff, roughly hairy leaves and a zygomorphic flower with unilateral stamens and anthers slightly sagittate with the connective expanded in the lower half.Lewis described the flower as actinomorphic or almost so, but we have seen no plants in the field that have such flowers, nor does the type illustration show an actinomorphic flower.The tepals are subequal in length but the dorsal is held somewhat apart from the others and the lower three bear contrasting markings, either white to cream-coloured, edged in a darker colour when the peri anth is predominantly blue or pink, or yellow when the perianth is otherwise cream-coloured.The perianth tube is 12-18 mm long and almost filiform, the walls tightly enveloping the enclosed style so that nectar is forced into the slightly wider distal portion of the tube.The stamens are erect, but clearly unilateral, with anthers facing the lower tepals and the style is held against the back of the anthers.Plants with symmetrically disposed stamens as illustrated by Lewis evidently exist but are not usual.
Babiana regia (G J.Lew is) Goldblatt <£ J.C M an ning, comb, et stat.nov.As outlined above, Babiana regia has strikingly marked tepals, deep blue with a red base.Although known since at least the early 19th century, when it was collected by J.F. Drêge (Lewis 1959), it was long confused with B. rubrocyanea (Jacq.)Ker Gawl., which has a similarly coloured flower but unilateral stamens and unusual broad style branches.Lewis realized that these plants were different species and recognized B. stricta var.regia for plants with symmetrically disposed stamens and short, slender style branches.However, it sits uncomfortably in B. stricta because the flower is radially symmetric, the tepals nearly equal and cupped even when fully open, and the brown anthers are linear, lacking the expanded connective of B. stricta.Vegetatively the two also differ, for B. regia has an arching stem nearly horizontal above, whereas the stem of B. stricta is suberect.
Babiana regia is restricted to the Western Cape fore lands between Malmesbury and Stellenbosch and favours sandy soils.As far as we are aware, it persists today at just one site, adjacent to the Farm Joostenbergkloof, on the northwest foothills of the Joostenberg, between Durbanville and Paarl.It co-occurs with other rare species, notably Geissorhiza purpurascens and Hesperantha spicata subsp.fistulosa in a small area heavily infested with Australian Acacia species.We suspect that all three species will soon be extinct unless some action can be taken to conserve them.Babiana stricta var.grandiflora GJ.Lewis: 46 (1959).Type: South Africa, [Western Cape).18 miles north of Piketberg.28 July 1950, G J. Lewis 2197 (SAM, holo.!, iso.!).
Distribution: sandstone outcrops in transitional fynbos-renosterveld.B. longifllora is known only from rocky flats at the foot of the Piketberg and Porterville Mountains of Western Cape.
Diagnosis and relationships: Babiana longiflora has been rarely collected, mostly at the same locality at the foot of the Piketberg or close by, and we consider its affinities to be have been misunderstood.It seems to us not to be closely allied to B. stricta nor to any of its four varieties that were recognized by Lewis.In particular, it has fairly broad leaves that are only slightly plicate and with soft pubescence, unlike the fairly rigid, narrow and deeply pleated, ± erect and shortly hairy leaves of B. stricta.Plants have long-tubed, pale purple flowers with subequal, ascending tepals.and long, rather blunt outer bracts.We suspect that it may be most closely related to B. fragrans, which has comparable, soft-textured.fairly broad leaves with long hairs, and large floral bracts rusty only at the tips.B. longiflora can be recognized among the species of section Babiana by the short floral bracts, the inner divided to the base, a perianth tube 25-30 mm long (relatively long for the section) and dark violet sta mens reaching to about the middle of the dorsal tepal.Other long-tubed species with inner bracts divided to the base.B. ecklonii and B. latifolia.are unlike B. longiflora in their dark violet perianth, the tube 30-47 mm long, curved at the apex, and the lower tepals joined to one another for at least 4 mm (5-7 mm in B. ecklonii), thus forming a strongly bilabiate perianth with horizontally extended lower tepals and the dorsal erect.Similar in vegetative features, B. fragrans has strongly scented flowers with a perianth tube 18-20 mm long, broader, usually pale blue (rarely pale yellow) tepals, and the style dividing opposite the lower third of the anthers.
We include plants from the foot of the Porterville Mountains at Twenty Four Rivers, some 40 km east of the type locality here.These plants have dark blue flowers but otherwise closely resemble other specimens of Babiana longiflora.notably in their straight perianth tube ± 30 mm long, hollow to the base and containing nectar.In Lewis's (1959) account of what she called Babiana plicata, the type species of Babiana.she cited as syno nyms both B. disticha Ker Gawl.and Gladiolus fragrans Jacq.Bullock (1961) in his review of Lewis's mono graph of the genus pointed out that B. plicata is a super fluous name for Gladiolus fragrans and suggested that the correct name for the plant was B. disticha.The earli er name G. fragrans was not considered by either Lewis or Bullock as available for transfer to Babiana because of the hom onym .fi./ragra/ii Eckl.(1827), a synonym of B. nana (Andrews) Spreng.Babiana fragrans Eckl., however, is a nomen nudum and thus is invalid and can not be taken into consideration in questions of priority of species epithets.The name for the type species of Babiana thus becomes B. fragrans.We have examined the type of G. fragrans, an illustration in the Hortus schoenbrunnensis, and concur with Lewis that this species and B. disticha are synonyms.
A relatively unspecialized species.Babiana fragrans is recognized by the subequal, spreading tepals with the dorsal only slightly longer than the lower, and small pale markings edged in darker blue or purple on the lower tepals.The perianth tube is about as long as or slightly longer than the dorsal tepal and the erect, usually branched stem bears spikes of up to 10 flowers.The suberect stamens are unilateral and the anthers parallel and contiguous, and usually dark blue.The soft-textured, hair>' leaves are weakly pleated and often oblong rather than the usual sword shape of most species of the genus.Both Ker Gawler (1802), when describing the synonym B. plicata.and Lewis (1959) remarked on the strong, pleasing fragrance, likened by Lewis to that of a carna tion.Somewhat variable across its range.B. fragrans from the interior southwestern Cape has a more strongly bilabiate flower with the dorsal tepal up to 5 mm longer than the lower and the upper lateral tepals are united for a short distance w ith the lower, thus forming a more pro nounced lip than is evident in plants from the Cape Peninsula and nearby.
The confused pre-1800 taxonomic history of Babiana plicata was outlined in detail by Lewis (1959) and is not repeated here.
account of Babiana.These were B.falcata, B.flavida and B. hypogaea (for no apparent reason spelled hypogea by Lewis).Examination of type material of B. hypogaea, collected by William Burchell in August 1812, and described by him in 1824, shows the slender leaves and pale flower with tepals darkly pigmented along the midline that closely corre spond to what Lewis called B. flavida.That plant must therefore assume the earlier name B. hypogaea, whereas plants with longer leaves and blue to violet flowers in which the lower tepals are marked with white, and which were associated with the name B. hypogea by [North-West | , Warren ton.December 1922.C G Adams 819 (BOL.holo.!).Babiana falcata GJ.Lewis: 115 (1959), syn.nov.Type: South Africa, [Northern Cape], 20 miles east of Springbok, red sand flats, 8 September 1950.W.F. Barker 6671 (NBG.holo.!).
Flowering time: mainly June to September, occasionally December to May.Distribution: flats on red sand plains in the Upper Karoo, Bushmanland.and southeastern Namibia.Additional collections made since the publication of Lewis's monograph represent range extensions into south ern Namibia where B. hypogaea in its current sense was previously known only from a collection near Graspoort.The new record from near Kimberley (Leistner 2635), identified by Lewis as B. falcata subsequent to the publi cation of her monograph of Babiana, is a particularly close match to the type collection of B. hypogaea.
WESTERN CAPE -3218 (Clanwilliam): 18 miles north of Piket berg, (-CC), 28 July 1950.Barker 6371 (NBG).3318 (Cape Town): Twenty Four Rivers, sandy alluvium.(-BB), 1 September 1992.Goldblatt & Manning 9363 (MO.NBG).I he new combination, Babiana fragrans for the type of the genus, B. disticha B. cuneata has cupped tepals 26-40 mm long and filaments 15-18 mm long and is thus readily distinguished from the long-tubed western Karoo species B. praemorsa (called by Lewis B. flabellifolia', see Lewis 1959; plate 21), which has a straight perianth tube 40-60 mm long, horizontally spreading tepals 18-22 mm long, and rela tively short filaments, 8-9 mm long.The filaments in the other long-tubed species in the complex, B. pubescens, are 14-15 mm long.A revised description is provided below for Babiana flabellifolia (syn.B. truncata), as well as formal descrip tions of the new species B. cuneata and B. praemorsa.We also lectotypify B. cuneifolia which was treated as a synonym of B. pubescens by Lewis.Most specimens of the type, Drêge 2627, actually have a short perianth tube and accord closely with B. flabellifolia!truncata.Babiana pubescens is not known to occur in southern Namaqualand where the type collection of B. cuneifolia was made and we suspect that specimens of B. pubescens from another site were later mislabelled with the Drêge collection number.