The native and naturalised species of Peltocalathos and Ranunculus ( Ranunculaceae : Ranunculeae ) in southern Africa

The five native and naturalised species of Ranunculus L. and Peltocalathos Tamaru (Ranunculaceae: Ranunculeae) from southern Africa are reviewed, with full descriptions and nomenclature, taxonomic history, ecological notes, and distribution data. All species are illustrated. Peltocalathos, a monotypic genus based on R. baurii MacOwan, has been overlooked in the regional literature until now. The new name R. dregei J.C.Manning & Goldblatt is proposed as a replacement name for the later homonym R. meyeri Harv., non Lowe. We conclude that the poorly known R. capensis Thunb. represents juvenile forms of the widespread and variable R. multifidus Forssk. and we synonymise the name, along with the three trivial epithets R. pinnatus var. hermannii DC., R. pubescens var. glabrescens Burtt Davy and R. pubescens var. harveyanus Burtt Davy, which have been overlooked in the recent literature. Based on number (mostly < 40 achenes per head) and size (mostly 1.2–1.8 mm long) of achenes, we conclude that the southern African material of the aquatic species recently re-identified as R. rionii Lagger is more appropriately treated as R. trichophyllus Chaix. 1 Compton Herbarium, South African National Biodiversity Institute, Private Bag X7, Claremont 7735, South Africa. E-mail: J.Manning@ sanbi.org.za. 2 Research Centre for Plant Growth and Development, School of Life Sciences, University of KwaZulu-Natal, Pietermaritzburg, Private Bag X01, Scottsville 3209, South Africa. 3 B.A. Krukoff Curator of African Botany, Missouri Botanical Garden, P.O. Box 299, St. Louis, Missouri 63166, USA. E-mail: peter.goldblatt@mobot.org. 180 Bothalia 43,2 (2013) MATERIALS AND METHODS All relevant types were examined, as well as all herbarium material from BOL, GRA, NBG, PRE and SAM (acronyms after Holmgren et al. 1990), the primary collections of southern African material. All species with the exception of Ranunculus muricatus were also studied in the field.


INTRODUCTION
Ranuculaceae are a large family of ± 2 150 spp. in ± 60 genera distributed worldwide, showing a diversity of floral organisations and fruit types (Tamaru 1993).Several classifications of the family have been proposed over the years, most recently by Tamaru (1995), who recognised three subfamilies and eleven tribes, based largely on chromosome base number and carpel and fruit types, including placentation.Tribe Ranuculeae DC. of subfamily Ranuculoideae Hutch.includes ± 650 spp.occurring on all continents, and is diagnosed by carpels with a solitary, unitegmic and usually ascending ovule, and petals with at least one nectary near the base (Tamaru 1993, Emadzade et al. 2010).Molecular phylogenetic studies suggest that the tribe is monophyletic but that parallel, adaptive and convergent evolution of morphological characters is rife, hindering the development of a morphology-based classification of the tribe.The recent generic classification of Ranunculeae derived by Emadzade et al. (2010) from their molecular analysis largely corroborates the narrow generic circumscriptions proposed by Tamaru (1995) [exceptions are the inclusion of Aphanostemma A.St.-Hil.and Gamsoceras Steven in Ranunculus but the exclusion of Coptidium (Prantl.)Beurl.ex Rydb.and Ficaria Guett.as separate genera], ultimately recognizing 18 genera in the tribe, of which Ranunculus L. (500-550 spp.) is by far the largest.This phylogeny also establishes that the aquatic members of R. sect.Batrachium DC. are deeply nested in the genus in a clade that includes wetland species such as R. repens L., affirming that their segregation at the higher taxonomic levels of genus (Gray 1821) or subgenus (Gray 1886) is not justified.
Rancunculeae are poorly represented in sub-Saharan Africa, with only a handful of species of Ranunculus recorded as native, mainly from the temperate eastern highlands of the continent (Oliver 1868;Milne-Redhead & Turrill 1952;Exell & Milne-Redhead 1960).The generic diversity is higher in southern Africa, with two indigenous genera.In addition to Ranunculus itself, the subcontinent is home to the recently described endemic monotypic genus Peltocalathos Tamaru (1992), based on R. baurii MacOwan and separated from Ranunculus by its peltate leaves and longitudinally ribbed achenes lacking a sclerenchyma layer in the carpel wall, the ribs corresponding to the vascularisation of the carpel walls.In addition to the native taxa, the European Myosurus minimus L. is naturalised in parts of Western and Northern Cape.The last regional revision of Ranunculus in southern Africa is that of Harvey (1860), in which six species were recognised.The application of three of these names is now known to have been incorrect.Current checklists for the subcontinent list seven species (including R. baurii, now P. baurii), with R. capensis Thunb.regarded as a doubtful species and R. plebeius R.Br.ex DC. as a doubtful determination (Winter 2006).The description of the monotypic Peltocalathos for R. baurii has been overlooked in all of the regional literature.

Peltocalathos baurii (MacOwan)
Distribution and ecology: a montane species endemic to the southern African Drakensberg, ranging from the Katberg in Eastern Cape through KwaZulu-Natal, Lesotho, Swaziland and Mpumalanga, with a single collection from the headwaters of the Olifants River in Limpopo (Figure 2).
Peltocalathos baurii is restricted to damp gullies, streamsides and the base of wet cliffs, from 1 000-3 000 m.
Diagnosis: a distinctive species recognised by its peltate leaves with hollow petioles and scapose inflorescence of moderately large flowers with numerous petals and caducous sepals.The leaf blades are suborbicular Plants in sheltered situations tend to be much larger than those growing in shallower soils or in more exposed conditions on banks or in alpine seepages.Although the two extremes are striking, we have encountered a complete range of variants on wet banks and seepages on Naude's Nek in the southern Drakensberg, ranging from small, solitary-flowered plants with leaves less than 30 mm in diameter to well-grown individuals over 300 mm high with multiple flowers and leaves to 80 mm diam.
Plants collected early in the season at high altitudes around 3 000 m in the Drakensberg Mtns, most commonly on the summit plateau of Mt aux Sources, are exceptionally dwarfed, with solitary (rarely two) flowers on scapes 30-70 mm long and leaves either just emergent or with reduced blades 5-20 mm diam.Tölken (1965), who also remarked on this form, was uncertain how to treat it.Although these populations may represent a distinct alpine ecotype, the species as a whole exhibits such marked phenotypic plasticity that they are more likely merely extreme morphotypes, especially since early-flowering individuals from lower altitudes elsewhere may be similarly diminutive.
History: The species was described in 1881 by Peter MacOwan (1830MacOwan ( -1909)), principal of Shaw College in Grahamstown and expert on the regional flora, from a collection made near Baziya, west of Umtata, by the missionary Leopold Richard Baur (1825-1889).The collection is undated but was presumably made after 1873 since it was during this year that Port Elizabeth businessman and amateur botanist Russell Hallack (1824Hallack ( -1903) ) visited Baur and encouraged him to botanise the area and send his collections to MacOwan.MacOwan associated Baur's plants with a collection from Lesotho made in 1861 by the professional plant collector Thomas Cooper (1815Cooper ( -1913)), who had recognised that his find represented a new species.This collection (BOL, SAM) is the voucher for seeds and rhizomes that Cooper despatched to his employer, W.W. Saunders of Reigate, for cultivation, where they were successfully grown and flowered.A brief description by Kew botanist D. Oliver of the herbarium specimens pressed from these cultivated plants formed a footnote to MacOwan's description of R. baurii.
With the arrival of a duplicate of Baur's collection at Kew in 1885, Oliver lost no time in having both it and Cooper's specimens lithographed for Hooker's Icones plantarum (Oliver 1886a, b), and in describing the latter under the name R. cooperii.The differences between the two were slight, 'chiefly in dimensions and number of flowers' (Oliver 1886a) and were followed by the caveat, or presentiment, that the two taxa might later have to be reduced to a single species as it was 'not improbable that connecting forms may be found'.Both names [the former sometimes misspelled as 'bauerii'] continued to be applied to plants from KwaZulu-Natal Ranunculus L. in Species plantarum 1: 548 (1753).Type species: Ranunculus acris L. (Jarvis 2007).
Annual or perennial herbs, sometimes aquatic, glabrous or pubescent.Roots fibrous to fleshy, usually without a taproot.Stems erect or creeping, sometimes stoloniferous.Leaves spiral, cauline and/or basal, petiolate, stipules adnate to base of petiole and sheathing; blades simple and usually palmately lobed or pinnatisect (rarely peltate in R. lyallii from New Zealand), or compound, trifoliolate or imparipinnate, in aquatic species often decompound into filiform or linear segments.Flowers solitary and terminal or in cymes, bisexual, actinomorphic, all parts spiral.Sepals usually 5(3), spreading or reflexed.Petals [also termed honey-leaves] 5 or 6 (-10), yellow or white, often glossy adaxially, narrowed at base into short claw, with nectar-secreting pit on adaxial surface at top of claw often covered by laterally adnate scale or flap forming nectary-pocket.Stamens numerous, rarely 5 or less.Carpels free, usually numerous, with 1 basal ovule.Achenes smooth, tuberculate or transversely rugose, glabrous or pubescent, with or without persistent style-beak; pericarp with inner sclerenchyma layer present.500-550 spp., cosmopolitan but mainly temperate Northern Hemisphere; 4 spp. in southern Africa, with 3 spp.native and 1 sp.naturalised.Some species are reported to be poisonous to livestock when fresh (Eichler & Walsh 2007) but Burtt Davy (1926) reported that the leaves of R. multifidus are eaten in spring and at other times when grass is scarce.Diagnosis: readily distinguished by the mostly subglabrous, simple or 3-5-lobed, orbicular leaves and echinate achenes, 5-8 mm long, with a stout, blade-like or hooked beak 2-3 mm long.It is possible that Harvey's (1860) reference to Ranunculus sceleratus L. is relevant here.In his account of the genus in southern Africa, he reported the occurrence of R. sceleratus in ditches near Cape Town but neglected to preserve any specimens.Native to Europe, R. sclereratus is a subglabrous annual with lobed and crenate basal leaves, deflexed sepals, and elongate-conical receptacles bearing numerous small, ± 1 mm long, ovoid, faintly rugose achenes (Cook 1964).No plants of R. sceleratus have ever been collected in southern Africa and the identity of Harvey's plants remains unclear.A collection by Ecklon and Zeyher from the banks of the Zwartkops River listed by them under this name is in fact R. multifidus and although it is possible that Harvey (1860) made a similar error we consider this to be unlikely given his familiarity with that species on the Cape Peninsula.R. sceleratus is vegetatively much more similar to R. muricatus than to R. multifidus and if confusion occurred, it is more likely to have been between the former two species, especially without fruits.[See Perrier de la Bâthie (1950) and Exell & Milne Readhead (1960) for additional synonyms from tropical Africa] Tufted perennial, mostly 100-500 but much dwarfed in exposed situations and up to 1 000 mm in protected places, developing new shoots at base; flowering stems erect or sprawling, simple or well-branched in upper parts, densely or sparsely patent-or adpressed-hirsute or -pilose below with hairs to 2 mm long and adpressedpubescent above, rarely glabrescent or subglabrous.Roots several to many, thong-like.Leaves basal and cauline, petioles of basal leaves 20-150(-300) mm long with stipules adnate and sheathing for 5-30(-40) mm, usually pilose, sometimes subglabrous, blade ovate in outline, ternate or ternately pinnate to bipinnatisect (rarely tripinnatisect) with 1 or 2 lateral pairs of leaflets, (20-)30-90(-130) × (15-)20-70(-100) mm, ultimate segments coarsely and irregularly toothed, with callus at tip of each serration, usually adpressed-pilose on both surfaces, rarely glabrescent, cauline leaves progressively smaller, ultimately subsessile and lanceolate; juvenile leaves suborbicular and lobed to ternate.Flowers in simple or compound cymes, 10-18 mm diam.Sepals 5, strongly reflexed, ovate, cucullate, 2.5-5.0 mm long, hispid beneath, yellowish.Petals 5, obovate, 4-9 × 3-7 mm, varnished bright yellow above, dull and paler beneath, narrowed into short claw in basal ± 1 mm, obtuse, veins branched, nectary pocket-like.Stamens ± 30 in 2 or 3 series, filaments 1-3 mm long, anthers ± 1.5 mm long.Carpels 30-50(-70), glabrous.Achenes in an ovoid head, broadly obovoid or subcircular, 2.0-2.5 mm diam., usually verrucose or tuberculate but sometimes ± smooth, margin keeled and grooved, shortly beaked, beak acute, ± 0.5-0.8mm long; receptacle 2-5mm long, pubescent.Flowering time: Aug.-Jan.Figure 5.

Additional specimens seen
Distribution and ecology: widespread through southern and east tropical Africa to Ethiopia, Arabia and Madagascar.In southern Africa, Ranunculus multifidus is largely restricted to the more mesic southern and eastern parts (Figure 6), from near sea level to 3 000 m, and is largely absent from the western half of the subcontinent apart from some isolated populations in higher or locally moist situations.
Plants typically occur in moist or damp, open grassy habitats and floodplains, along streams and bogs, sometimes partially submerged, and in open patches in temperate forest; it is sometimes ruderal in gardens or along ditches and around dams.The species is self-fertile; producing numerous fruits, and has a tendency to become weedy in damp waste places, suburban gardens and parklands, especially in the southern and southwestern Cape, where it is commonly associated with human activity.Local variants are readily established.
Diagnosis: a widespread and variable species, especially in the degree of dissection of the leaves, which range from ternate to tripinnatisect, development and density of the vestiture on the stems and leaves, size of the flowers, and development of tubercles on the fruits, which vary from scarcely evident to almost spinulate.Plants from high altitudes and exposed or trampled situations are typically dwarfed and compact or almost prostrate and those from sheltered locations larger and more lush, but the species is always readily distinguished from other sub-Saharan species by its ± appressed pubescent, ternately pinnate or bipinnate leaves with one or two pairs of lateral leaflets, and ± tuberculate achenes.The flowers are small to moderately sized, with reflexed, hispid sepals and 5 obovate petals, 4-9 mm long.

History:
The correct identification of the South African material of this species took some time to establish.The first collections of the species on the subcontinent were made in the late eighteenth century by Carl Thunberg (1743-1828), who described them under the names R. capensis Thunb.(1800) and R. pubescens Thunb.(1800).Subsequently Harvey (1860), in his account of the genus for Flora capensis, segregated the material available to him among three different names: typical plants with pinnate or bipinnate leaves, including R. pubescens Thunb., were treated by him, and later by Oliver (1868), under the name R. pinnatus Poir.; smaller plants with subsimple radical leaves were treated as R. capensis Thunb.; and a collection from the Sneeuberg with ternate leaves and smooth achenes was referred to the Australian species R. plebeius DC., a mistake that was later corrected by Exell & Milne-Redhead (1960).The priority of the name R. multifidus Forssk.(1775) over R. pinnatus Poir.(Lamarck 1804), under which name the species was better known in southern Africa, was established by Milne-Redhead & Turrill (1952).
Collections of subglabrous plants from Mpumalanga were described as var.glabrescens by Burtt Davy (1921).The protologue clearly designates Burtt Davy 17682 (K) as the [holo] type, with Burtt Davy 17164 included among the paratypes, but this was somewhat confused when Burtt Davy (1926) later identified both collections as types [syntypes].This is incorrect.He also segregated Burke's Sneeuberg collection [previously treated as R. plebeius by Harvey ( 1860)] plus some others as his new var.harveyanus Burtt Davy (1926).Both variants fall within the range of variation of Ranunculus multifidus and we see no justification for upholding them.We have seen no authentic material of R. plebeius from southern Africa.
The identity of Ranunculus capensis Thunb.( 1800), based on a collection of three small plants-two of them just coming into flower and the third sterile-deserves further consideration and it was treated as a doubtful species by Winter (2006).The name was associated by Harvey ( 1860) with two or three additional collections from the Cape Peninsula and Clanwilliam, notably Ecklon & Zeyh. En. No. 13 (SAM) from Green Point and the latter is indeed a perfect match with Thunberg's type.After close study we conclude that the species represents a seedling variant of R. multifidus with juvenile, ternate leaves.This conclusion was also reached by Adamson (1950), who suggested that the taxon was a dwarf state of the variable R. pubescens (now R. multifidus).The pubescent stems and petioles, the hirsute leaf blades, the relatively small flowers with reflexed, hispid sepals, and the pubescent receptacle with small, compressed achenes with short beak are all consistent with R. multifidus.Additional collections from Green Point (Ecklon s.n.SAM13992 and Thode s.n.SAM9260) are perfectly intermediate with more typical R. multifidus and we accordingly formally synonymise the name here.
Distribution and ecology: distributed mainly along the eastern escarpment, from Gaika's Kop in the Amatola Mtns in Eastern Cape through KwaZulu-Natal and Lesotho to Standerton and Vaalbank in Mpumalanga and the Leolo Mountains in Limpopo (Figure 4), extending north of South Africa into the eastern highlands of Zimbabwe and Mozambique (Exell & Milne-Redhead 1960), with a remarkable southern outlying station near the summit of the Swartberg Mtns in Western Cape.Although there are early records further inland, from Parys in the Free State, Pretoria and Johannesburg in Gauteng, and Potchefstroom in North-West, the species has not been collected there since the first half of the twentieth century and is presumably extinct here.
Ranunculus dregei restricted to peaty seeps or boggy places at higher altitudes, 1 500-3 000m, where it forms mats in open spaces.
Diagnosis: a distinctive species, readily recognised by its dwarf, creeping habit, simple leaves with solitary flower per leaf tuft, and narrow, ligulate petals with unbranched veins.
The species was first recorded from north of the Limpopo River Valley, in the highlands of Mozambique andZimbabwe, in 1956 (Exell &Milne-Redhead 1960) and annotations on some herbarium specimens by T. Duncan indicate an intention to reduce it to a variety of the tropical African Ranunculus volkensii Engl.Although the two taxa are similar in habit, R. volkensii has noticeably thicker stolons, ± 2-3 mm diam., often lanceolate leaves, and more critically, obovate petals with branching venation, quite unlike the ligulate, 3-nerved petals of R. dregei.It also has mostly ± twice as many carpels as R. dregei.History: first collected in Eastern Cape by J.F. Drège (1794Drège ( -1881) ) sometime in the early summer of 1832, the specimens themselves are not localised, Ranunculus dregei was listed as being collected on the Katberg in Drège's Zwei pflanzengeographische Documente.These collections were annotated with the manuscript name Ficaria radicans by Ernst Meyer, who contributed the botanical text to the Documente, but the species was only formally described by Harvey ( 1860), who named it after Meyer.Harvey (1860) unfortunately overlooked the fact that the name R. meyeri had already been proposed by R.T. Lowe (1857) as a replacement name for the East-Central Asian R. grandifolius E.Mey.( 1830) under the misapprehension that the latter was synchronous with his R. grandifolius Lowe (1831) 1887), but the name was synonymised by Exell & Milne-Redhead (1960).We have been unable to trace the type specimen in the Schinz herbarium in Zurich (H.P. Linder, pers. com. 3 Jan. 2013) Villars: 335 (1786);Cook: 126 (1966).Batrachium trichophyllus (Chaix.)Bosch: 5 (1850); Eichler & Walsh: 312 (2007).Type: 'Frequentissimus in rivulis quietis, fossisque aqua plenis ', Haller 1162(G, neo., designated by Lourteig (1951), not seen).R. drouetii F.W.Schultz ex Godron: 10 (1842).Type: Flora Galliae et Germaniea exsicattae n. 404 (not seen).
Distribution and ecology: distributed widely through Eurasia and North Africa, North America, Australia, Tasmania, and southern (but not tropical) Africa, where it is recorded so far only from Lesotho and South Africa (Figure 9), and reportedly also Botswana (Cook 2004) but not Namibia (Craven 1999).In South Africa the species is scattered through the southern and eastern interior of the county along the Vaal/Orange Rivers and tributaries, and along the southern escarpment from Calvinia to Queenstown, with a few coastal stations in the southwestern and southern Cape around Velddrif, Cape Town, Bredasdorp and Uitenhage.
Plants grow submerged in freshwater seasonal pools and ponds and in slow, intermittent or perennial streams, furrows or irrigation ditches up to 50 cm deep, persisting only briefly along the muddy edges as the water dries up.It has been recorded from sea level to well over 2 000 m.In southern Africa R. trichophyllus has occasionally been regarded as troublesome in the past, viz.as a 'serious impediment in the irrigation canals' near Cradock in the 1950s [Barnes s.n.(GRA)].
The natural occurrence of the species in the Southern Hemisphere is worth closer examination given the large disjunction to Eurasia and North America.The early collection of the species in pristine habitats in Australia suggests that it is probably native there (Eichler & Walsh 2007), and similar considerations apply in southern Africa, where it was first documented (under the name Ranunculus aquatilis L.) in the early 1800s from the Eastern Cape, around Uitenhage and Graaff Reinet, by C.F. Ecklon (1795-1868), C.L. Zeyher (1799-1858) and I.L. Drège (1853Drège ( -1921)).Documentation of R. trichophyllus on the Cape Peninsula only dates from the end of the 19 th century, which is unusually late for such a distinctive species.
Diagnosis: the only aquatic species in the subregion, distinguished at once by its habit, finely dissected leaves, white petals with yellow claws, and small, ovoid, transversely rugose achenes, usually with a few, scattered bristles along the lower keel below the remains of the style but sometimes glabrous or ± densely bristly.The transverse folds on the mature achenes are due to the presence of styloid crystals in the carpel walls (Cook 1966).
History: The recent determination of the southern African material as Ranunculus rionii (Glen 2003;Winter 2006) follows the identifications of the southern African material by C.D.K. Cook (1966Cook ( , 2004)), a specialist in the taxonomy of Ranunculus sect.Batrachium.Prior to this, the South African collections had been identified as R. trichophyllus (Adamson 1950) or its synonym R. drouetii (Burtt Davy 1926).The two species R. rionii and R. trichophyllus are sympatric in Europe and differ solely in the number and size of their achenes, with R. rionii characterised by smaller, more numerous achenes (Cook 1964(Cook , 1966(Cook , 1993;;Pizarro 1995).The number of achenes per fruiting head ranges from 60-90 in R. rionii and from 16-35 in R. trichophyllus (Cook 1966).Dimensions for the achenes in R. rionii have been variously given as up to 1 mm long (Cook 1964(Cook , 1966(Cook , 1993) ) or 1.0-1.2mm long (Pizarro 1995) and for R. trichophyllus as more than 2 mm long (Cook 1964), more than 1.5 mm long (Cook 1966, or 1.20-1.75 mm long (Pizarro 1995).R. trichophyllus is a very widespread species, occurring throughout Eurasia, including the Mediterranean Basin, North America and Australia, where it is probably also native, but R. rionii has a more restricted distribution in Central and Eastern Europe and the eastern Mediterranean into Asia Minor (Cook 1966).
The illustration of the fruiting head of Ranunculus rionii given by Cook (2004) in his guide to Aquatic and wetland plants of southern Africa certainly conforms to that species in its numerous achenes, as does the illustration of the individual achene, with a length of less than 1 mm, but the source of the original material for these drawings is not given.The majority of the southern African material that we have examined has fewer than 40 achenes per fruiting head, with individual mature achenes measuring 1.2-1.8mm long, excluding the stylar remnants (immature achenes are smaller), thus consistent with R. trichophyllus and not R. rionii.Some specimens, however, including Comins 800 (BOL) from Tarkastad and others from along the Vaal River, have heads of 40-60 achenes, each achene ± 1 mm long, thus potentially R. rionii.Although the difference between the two taxa in achene number and size as initially treated by Cook (1963) was absolute, the ranges of variation in the two characters have gradually converged (Cook 1966(Cook , 1993;;Pizarro 1995) to the stage where it is difficult to assign 'intermediate' specimens.It may also be relevant here that R. rionii has been treated as conspecific with R. trichophyllus in the past at subspecific or varietal rank (Cook 1963).The size of the achenes may bear relation to the number of (developing) carpels, as seen in Wilman 21494 (PRE), in which heads with over 40 ripened carpels have the individual achenes ± 1 mm long whereas those with few fertilised carpels have achenes ± 1.3 mm long.Significantly, no authors have as yet proposed that more than one aquatic species occurs in the subcontinent, although it is not impossible that there has been more than one introduction from Europe.The matter deserves attention but until then we adopt a conservative approach and provisionally treat all of the southern African material under the older name R. trichophyllus.This name has also been applied to the material from Ethiopia (Teketay & Edwards 2000).
Distribution and ecology: native to the Mediterranean region but adventive in the southwestern Cape, where it has been recorded from the Cape Peninsula to Kleinmond and Swellendam (Figure4), in marshy ground and riverbanks.It has been collected so far only near settlements and has not been recorded in undisturbed vegetation.The species is widely naturalised in Australia and parts of the continental United States, mainly as a weed of damp waste places and croplands but occasionally along lake and stream margins in native vegetation(DiTomaso & Healy 2007;Eichler & Walsh 2007).
Aluka images!].[The SAM collection is chosen as lectotype as being most likely the specimen on which MacOwan based his description and in displaying insect damage (to the foliage) as reported in the protologue].
from Madeira.In fact it is R. grandifolius Lowe that is the illegtimate later homonym and thus requires a new name, and R. meyeri Lowe becomes an illegitimate superfluous name for R. grandifolius E.Mey.This renders R. meyeri Harv.an illegitimate later homonym requiring a new name, for which we propose R. dregei in honour of the original collector of the species.A collection by Anton Rehmann (1840-1917) from Hennopsrivier, west of Pretoria, with glabrous petioles was distinguished as var.transvaalensis by Szyszylowicz ( but have no reason to question this decision.Other collections from Gauteng, viz.Repton 3430 from Rietvlei Reserve near Pretoria and Gilfillan 162 from near Johannesburg probably represent the same variant, being almost entirely glabrous, with scattered hairs only on the sheathing leaf bases.