A taxonomic synopsis of unifoliolate continental African Vepris (Rutaceae)

Descriptions and illustrations are presented for three new species to science, Vepris udzungwa Cheek, V. lukei Cheek (both Udzungwa Mts, Tanzania), and V. robertsoniae Q.Luke & Cheek (SE Kenyan kaya forests), in the context of a synoptic taxonomic revision and with an identification key to all the known unifoliolate taxa of Vepris in continental Africa. The new species are described and illustrated, while the remaining 10 taxa are given skeletal taxonomic treatments (lacking descriptions). One widespread species in montane eastern Africa is renamed as V. simplex Cheek because its previous name, V. simplicifolia (Engl.) Mziray, is predated by V. simplicifolia Endl. (basionym of Sarcomelicope simplicifolia (Endl.) T.G.Hartley, a widespread species of Australia, Lord Howe and Norfolk Islands, and of New Caledonia). The new combination Vepris hanangensis var. unifoliolata (Kokwaro) Cheek is made. Full or provisional conservation assessments are presented for all species using the IUCN 2012 standard. Of the 13 taxa, nine are considered threatened or near threatened, of which one is Near Threatened, three are Vulnerable, two Endangered, and three Critically Endangered, with two of the latter, both of the Uluguru Mts of Tanzania, possibly extinct globally, although not yet assessed on the IUCN Red List.


Introduction
Vepris Comm.ex A.Juss.(Rutaceae-Toddalieae) is a genus with 93 accepted species, 23 in Madagascar and the Comores, and 69 in Continental Africa with one species extending to Arabia, and another species endemic to India (Plants of the World Online (continuously updated)).The genus was last revised for tropical Africa by Verdoorn (1926).Founded on the Flore du Cameroun account of Letouzey (1963), nine new species were recently described from Cameroon (Onana & Chevillotte 2015;Cheek et al. 2018a;Onana et al. 2019;Cheek & Onana 2021;Cheek et al. 2022a), taking the total in Cameroon to 24 species, the highest number for any country globally, followed by Tanzania (16 species).The greatest concentrations of Vepris species in Cameroon are within the Cross-Sanaga Interval (Cheek et al. 2001), with 15 species of Vepris of which nine are endemic to the Interval.The Cross-Sanaga has the highest species and generic diversity per degree square in tropical Africa, including endemic genera such as Medusandra Brenan (Peridiscaceae, Breteler et al. 2015;Soltis et al. 2007;Barthlott et al. 1996;Dagallier et al. 2020).Much of this diversity is associated with the Cameroon Highland areas, different highlands each having a species of a given genus, e.g. as in Kupeantha Cheek (Rubiaceae, Cheek et al. 2018b).By comparison, neighbouring Gabon has just seven species of Vepris (Sosef et al. 2006), and just one species, V. lecomteana (Pierre) Cheek & T.Heller, is listed for Congo-Brazzaville (Plants of the World Online (continuously updated)).
In continental Africa, Vepris species are easily recognised.They differ from all other Rutaceae because they have digitately (1 -) 3 (-5)-foliolate (not pinnate) leaves, and unarmed (not spiny) stems.The genus consists of evergreen shrubs and trees, predominantly of tropical lowland evergreen forest, but with some species extending into submontane forests and some into drier forests and woodland.Vepris species are often indicators of good quality, relatively undisturbed evergreen forest, since they are not pioneers.
Species of Vepris in continental Africa extend from South Africa, e.g.V. natalensis (Sond.)Mziray, to the Guinean woodland in the fringes of the Sahara Desert (V.heterophylla (Engl.)Letouzey).Mziray (1992) subsumed the genera Araliopsis Engl., Diphasia Pierre, Diphasiopsis Mendonça, Oricia Pierre, Oriciopsis Engl., Teclea Delile, and Toddaliopsis Engl.into Vepris, although several species were only formally transferred subsequently (e.g.Harris 2000;Gereau 2001;Cheek et al. 2009;Onana & Chevillotte 2015).Mziray's conclusions were largely confirmed by the molecular phylogenetic studies of Morton (2017), but Morton's sampling was limited, identifications appeared problematic (several species appear simultaneously in different parts of the phylogenetic trees), and more molecular work would be desirable.Morton studied about 14 taxa of Vepris, all from eastern Africa.More recently Appelhans & Wen (2020), focussing on Rutaceae of Madagascar, have found that the genus Ivodea Capuron is sister to Vepris and that a Malagasy Vepris is sister to those of Africa.However, the vast majority of the continental African species, including all those of West and Congolian Africa, remain unsampled, leaving the possibility open for changes to the topology of the phylogenetic tree when this is addressed.
Characteristics of some of the formerly recognised genera are useful today in grouping species.The "araliopsoid" species have hard, non-fleshy, subglobose, 4-locular fruit with 4 external grooves; the "oriciopsoid" soft, fleshy, 4-locular syncarpous fruit; "oricioid" species are 4-locular and apocarpous in fruit; the fruits of "diphasioid" species are laterally compressed in one plane, bilocular and bilobed at the apex; while "tecleoid" species are unilocular in fruit and 1-seeded, lacking external lobes or grooves.There is limited support for these groupings in Morton's study, Due to the essential oils distributed in their leaves and the alkaloids and terpenoids distributed in their roots, bark and leaves, several species of Vepris have traditional medicinal value (Burkill 1997).Burkill (1997) details the uses, essential oils and alkaloids known from five species in west Africa: V. hiernii Gereau (as Diphasia klaineana Pierre), V. suaveolens (Engl.)Mziray (as Teclea suaveolens Engl.), V. afzelii (Engl.)Mziray (as T. afzelii Engl.), V. heterophylla (as T. sudanica A.Chev.), and V. verdoorniana (Exell & Mendonça) Mziray (as T. verdoorniana Exell & Mendonça) (Burkill 1997: 651 -653).Research into the characterisation and anti-microbial and anti-malarial applications of alkaloid and limonoid compounds in Vepris is active and ongoing (e.g., Atangana et al. 2017), although sometimes published under generic names no longer in current use, e.g.Wansi et al. (2008).Uses include applications as synergists for insecticides (Langat 2011).Cheplogoi et al. (2008) and Imbenzi et al. (2014) respectively list 14 and 15 species of Vepris that have been studied for such compounds.A review of ethnomedicinal uses, phytochemistry, and pharmacology of the genus Vepris was recently published by Ombito et al. (2020), listing 213 different secondary compounds, mainly alkaloids and furo-and pyroquinolines, isolated from 32 species of the genus, although the identification of several of the species listed needs checking.However, few of these compounds have been screened for any of their potential applications.Recently, Langat et al. (2021) have published three new acridones and reported multilayered synergistic anti-microbial activity from V. gossweileri (I.Verd.)Mziray, recently renamed as V. africana (Hook.fex Benth.)O. Lachenaud & Onana (Lachenaud & Onana 2021).There is no doubt that new compounds will continue to be discovered as chemical investigation of Vepris species continues.
Three new species of Vepris are described in the context of a synoptic treatment of the African unifoliolate species.The research was supported by preparation for a taxonomic revision of African Vepris by the first author, and of floristic work for conservation prioritisation in the surviving forests of Kenya and Tanzania by the second author.The paper builds on the foundation laid for the three western African unifoliolate species by Lachenaud & Onana (2021), and increases the number of described unifoliolate Vepris species for continental Africa from 10 to 13. Unifoliolate Vepris species are likely not a natural group, but may have arisen more than once from ancestral trifoliolate species.Yet without a well-sampled phylogeny it is difficult to be certain.

Materials and Methods
This taxonomic study is based on herbarium specimens predominantly held at BM, EA and K, field observations in Guinea and Republic of Congo by the first author, and field observations of live material in Kenya and Tanzania by the second author.All specimens seen are indicated "!".The specimens were mainly collected using the patrol method as indicated e.g. in Cheek & Cable (1997).Herbarium citations follow Index Herbariorum (Thiers continuously updated), nomenclature follows Turland et al. (2018), and binomial authorities follow IPNI (continuously updated).Material of the new species was compared morphologically with material of all other African Vepris, principally at K, but also using material and images from BM, BR, EA, FHO, G, GC, HNG, P and YA.Herbarium material was examined with a Leica Wild M8 dissecting binocular microscope fitted with an eyepiece graticule measuring in units of 0.025 mm at maximum magnification.The drawing was made with the same equipment using a Leica 308700 camera lucida attachment.The description was made following the format of Cheek et al. (2022a, b, c) using terms from Beentje & Cheek (2003).Specimen location data are given as on the specimen labels, understanding that the political units formerly termed Districts in Kenya are currently termed Counties.
For the extinction risk assessment, points were georeferenced using locality information from herbarium specimens.The conservation assessment was made using the categories and criteria of IUCN (2012); EOO was calculated with GeoCAT (Bachman et al. 2011).
The elevational range is 80 -624 m.In the field notes of several specimens it is described as being found in 'forest relicts', suggesting that it only occurs in patches of intact 'primary' forest and is absent from secondary forest.Plants occur as scattered individuals at low density; they appear to be dioecious, with female flowers larger and fewer than male flowers.Pollinators are unknown.Fruits are 1 -2 cm diam., 4-lobed orange berries, probably primate-dispersed (Cheek 2017).CONSERVATION STATUS.Vepris laurifolia was only known from 18 individuals in Sierra Leone, but five of these are known to have been destroyed in recent years (hydroelectric dam flooding), with two probably destroyed (due to agriculture in the area), and six more are due to be lost in the next 1 -10 years due to infrastructure developments (hydroelectric dam, transport corridor).Although only 11 -13 surviving individuals are documented, it is possible that as many as 50 -100 individuals may be found elsewhere, but intact forest habitat for this species only occurs as scattered remnants and is threatened with clearance for agriculture.Even in these scattered islands, the species is mostly absent (M. Cheek pers. obs. 2012-2016, Guinea;X. van der Burgt pers. obs. 2009-2016, Sierra Leone) Breteler (1995) based on the Liberian specimen (J.F. F. E. de Wilde 3754, type specimen), the specimens collected in Guinea in 1937 near Mt Benna (Jacques-Felix 2096 for whom the species is named), and another specimen collected in 1954 near Mt Kakoulima (Schnell 7568).Since the species was not named until 1995, it did not feature in the Rutaceae of Flora of West Tropical Africa (Keay 1958) and so is not mentioned in either Flore de la République de Guinée (Lisowski 2009) nor Flore de la Côte-D'Ivoire (Ake Assi 2001), both of which are based on the Flora of West Tropical Africa.Lachenaud & Onana (2021) recently discovered that Garcinia laurifolia was an earlier and unexpected synonym of V. felicis.The name G. laurifolia was originally published in Clusiaceae, probably due to the opposite leaves and poor state of the type collection (Scott-Elliott 4806 from Ninia, Talla Hills, Guinea).Lachenaud & Onana (2021) have compared the types and there is no doubt that G. laurifolia is identical to V. felicis, and since the epithet laurifolia is earlier and still available in Vepris it takes precedence.The specimens we refer to above are additional to those given in Lachenaud & Onana (2021).
Vepris laurifolia is unusual among unifoliolate Vepris in that at the apex of flowering stems, the leaves are opposite (not alternate) and the leaflet is not articulated with the petiole; further the fruits are 4-locular (in other unifoliolate species they are 1 or 2-locular), with four widely separated style bases.This shrub is so unusual in its genus that flowering specimens in the field have been misidentified as Rinorea Aubl.(Violaceae).However, the translucid spots usual in Rutaceae can be found using a lens, young leaves, and bright light (Cheek 2017).This species should be a priority for molecular phylogenetic analysis because it is so morphologically anomalous.http:// www.ipni.org/ urn: lsid: ipni.org: names: 77319 773-1
RECOGNITION.Vepris udzungwa differs from V. lukei Cheek in the lateral nerves of the leaf-blade 7 -8 (-10) on each side of the midrib (not 22 -28), leaf apex rounded (not acuminate), and from all other Tropical East African (Uganda, Kenya, Tanzania) unifoliolate species except V. mildbraediana G.M.Schulze in the hairy stems (not glabrous) and in the fruit asymmetric at both apex and base.DISTRIBUTION.Tanzania, endemic to the upper elevations of the Udzungwa Mts.
SPECIMENS EXAMINED.TANZANIA.UDZUNGWA MOUN-TAINS NATIONAL PARK, Camp 357 -pt 358, 7.68°S, 36.62°E, 1980m, fr. 12 Oct. 2002, Luke et   or plausible threats makes it impossible to define any threat-defined location, and the species must be assessed as Least Concern, LC, with the caveat that this status is dependent on continuing, effective conservation measures.PHENOLOGY.Immature fruits in June (end wet season), mature fruits and growth pulse (flush) in Oct. (start wet season).ETYMOLOGY.Taking the name of the mountain range and National Park in which the species was discovered and to which it appears to be unique.VERNACULAR NAMES & USES.None are known.NOTES.Vepris udzungwa, apart from V. mildbraediana, is unique among all Tropical East African (Uganda, Kenya, Tanzania) unifoliolate Vepris species in the hairy stems (Fig. 2B), and also in the fruits which are not only asymmetric at the base (as in e.g.V. hanaganensis), but also at the apex, the style being subapical (Fig. 2G).In addition, alone among these species it has longitudinally irregularly ribbed wrinkled fruits (Fig. 2G).
Luke et al. 6895 had previously been determined as "?" and numbers 8639 and 9109 as "Vepris sp.(=Luke 6895)" by Kaj Vollesen, indicating that in annotating the three specimens Vepris sp.cf.eugeniifolia, he recognised them to represent a possibly distinct, unplaced taxon.
An unusual feature of Vepris udzungwa is the very densely lenticellate stems.Parts of the older stems can be more than 50% covered in lenticels, while in most other Vepris species the older stems are only sparsely lenticellate.It is possible that V. udzungwa shares a recent common ancestor with V. lukei Cheek, which occurs in the same mountain range at a lower altitudinal band.However, these two species are morphologically unlikely to be confused (see under the latter species, below).
Numerous other species have been relatively recently discovered and are restricted or are largely restricted to the Udzungwa Mts, e.g., Polyceratocarpus askhambryan-iringae A.J. Marshall & D.M.Johnson (Annonaceae, Marshall et al. 2016), Trichila lovettii Cheek (Meliaceae, Cheek 1989).Ancistrocladus tanzaniensis Cheek & Frim.-Møll.(Ancistrocladaceae, Cheek et al. 2000;Cheek 2000;Taylor et al. 2005) Mendonça (1961: 84;1963: 204) HABITAT.This small puberulous shrub c. 0.7 m tall is found in the deep shade of mixed evergreen forest, sometimes associated with rivers and gulleys, at low to mid-elevations, 300 -1,600 m.CONSERVATION STATUS.Vepris drummondii is known from only 11 collections, collected between 1955 and 2015, and is listed as Vulnerable, B1ab(iii)+2ab(iii), with an EOO of 69 km 2 and an AOO of 32 km 2 based on known occurrence data (Darbyshire et al. 2017b).Although this may be a slight under-estimate, total AOO is unlikely to exceed 100 km 2 .Several of the localities for this species are within Forest Reserves, for example Glencoe Forest Land in Zimbabwe, although these are managed for commercial forest production rather than for biodiversity and so do not guarantee protection.At Maronga in Mozambique much of the lowland forest has been cleared outside of the core zone of the National Park zone and there is also significant artisanal gold mining activity along the Mussapo River west of Maronga, which has almost certainly resulted in riverine forest loss there.However, there is still intact forest suitable for this species within the core protected area at Maronga (Darbyshire et al. 2017a(Darbyshire et al. , 2017b)).The species is considered secure at both the Chimanimani National Park (Timberlake et al. 2016) and at Glencoe in Zimbabwe.PHENOLOGY.Flowering in November (early wet season), fruiting in April (start of dry season).

Vepris drummondii
ETYMOLOGY.Named for the late Robert (Bob) Drummond (1924Drummond ( -2008)), a life-long African botanist and botanical collector who collected the type specimen of this species and who was curator and a stalwart of the SRGH herbarium in Harare until the end of his life (Timberlake et al. 2017).VERNACULAR NAMES & USES.None are recorded.NOTES.Vepris drummondii is unlikely to be confused with any other species, since it is the only unifoliolate Vepris in the Flora Zambesiaca area (Mozambique, Malawi, Zambia, Zimbabwe, Botswana, Caprivi strip of Namibia).It is similar to V. mildbraediana of the Uluguru Mts of Tanzania, but that species has longer and denser hairs on the axes, and the partial-peduncles are only c. 2 mm long and few-flowered, while in V. drummondii they are much more fully developed.Nonetheless, these two may be sister species to each other.

Vepris mildbraediana G.M.Schulze
DISTRIBUTION.Tanzania, Uluguru Mts.SPECIMENS EXAMINED.Only known from the type specimen.HABITAT.Submontane forest; 1860 m alt.CONSERVATION STATUS.Vepris mildbraediana does not appear on the IUCN Red List (iucnredlist.org).Since the collector of the type specimen stated that it was "isolated", we can deduce that only a single plant was observed when the species was last seen 88 years ago.Given subsequent losses and threats to habitats in the Uluguru Mts (Ndang'ang'a et al. 2007) and the record of only a single individual (and certainly fewer than 50) despite multiple surveys for plants (but not targeting this species so far as we know), we provisionally assess this species as CR (PE) B2ab(iii), D, that is Critically Endangered (Possible Extinct).PHENOLOGY.Only known in flower in November (late dry season).Fruits unknown.ETYMOLOGY.Named for Gottfried Wilhelm Johannes Mildbraed (1879Mildbraed ( -1954)).He was an heroic botanist.Despite being captured in then German Kamerun in the First World War, losing all his specimens as spoils of war to the British (they were sent to K), and being imprisoned in France (1914France ( -1919)), he continued collecting specimens in tropical Africa (1907 -1928).As a taxonomist, he identified and published his discoveries and those of others.He collected in Cameroon, Tanzania, Burundi, and Rwanda, among other places.
VERNACULAR NAMES.None are known.NOTES.Kokwaro (1982) treated Vepris mildbraediana as an "Insufficiently known species", stating that he had not seen the type nor any other specimens so named, and that Bruce 510, which he described as Vepris sp.A, "has many similar characters (and is from the same locality) except for its paniculate inflorescence.On the other hand, V. mildbraediana may be a synonym of V. ngamensis if its inflorescence is a raceme as stated."Happily, thanks to JStor Global Plants and the African Plants Initiative (http:// apps.kew.org/ herbc at/ gotoA pi.do), while the holotype at B is destroyed, an isotype at BR was detected and is available as a high quality image.https:// plants.jstor.org/ stable/ 10. 5555/ al. ap.speci men.br000 00062 73002?searc hUri= genus% 3DVep ris% 26spe cies% 3Dmil dbrae diana It shows that the type specimen is densely covered in long, patent, yellow-brown hairs, persistent on the stem for 5 -6 nodes, and also on the petioles, abaxial midrib and inflorescence axis.No other unifoliolate East African species described has such dense indumentum.Vepris drummondii and V. udzungwa are the only other African unifoliolate Vepris described that have hairy stems, but those are only present at the first internode, and the hairs are white, appressed, sparse (c.10% coverage of the surface), and minute (0.05 mm long).Vepris mildbraediana has a panicle, but the partial-peduncles are only c. 2 mm long, unlike the raceme reported for V. ngamensis.Vepris sp.A of FTEA, apart from being glabrous.has more slender inflorescence axes and the stamens are twice as long as the petals (in V. mildbraediana the axes are stout and the petals are as long as the anthers).There is no doubt that V. mildbraediana is a distinct species.Evergreen tree 2 -5 m tall, when dried smelling of dried fish.Leafy stems glabrous, drying black, glossy, terete, 2 -5 mm diam., internodes 4 -26 mm long, increasing in length from the beginning of the flush (growth pulse), the main axis with 7 -11 nodes per flush, growth of different flushes separated by 1 -7 cm of naked stem, with fine longitudinal lines; lenticels lacking.Leaves alternate, glabrous, thinly coriaceous, ± concolorous, drying grey-green, glossy, narrowly oblong-elliptic, 6.5 -13.7 × 1.8 -3.7 cm; acumen 0.4 -1.1 cm long; base broadly acute; margin undulate, slightly revolute; secondary nerves 22 -28 on each side of the midrib, arising at c. 80° from the midrib, brochidodromous, forming a looping inframarginal nerve c. 2 mm from the margin; intersecondary nerves well developed; tertiary and quaternary nerves raised, forming a conspicuous reticulum on the lower surface; gland dots sparse and barely detectable in transmitted light, only slightly translucent; concolorous with blade and inconspicuous in reflected light except as minute raised spots on the abaxial surface.Petiole articulated at apex, 2.5 -48 mm long, variable in length; those first produced in a season longest, becoming successively shorter at successive nodes, plano-convex, c. 1 × 1 mm in section, the adaxial surface flat, the margins with a slender wing c.0.5 mm wide, held at c. 45° from the vertical plane of the petiole axis and bearing orbicular crater-like glands 0. Vepris lukei is known only from the two specimens cited above, both shown on Google Earth as within the Udzungwa Mts National Park, separated by c. 13 km.The protection level of this National Park is high currently, and so threats do not exist for this species at present (Q.Luke pers. obs. 2002Luke pers. obs. -2004)).expanding leaves, and also the leaves from the previous season's growth.These show a progressive reduction in length of the petiole during a season's growth.The first formed petiole is 48 mm long, the second formed 43 mm long, then 38 mm, 30 mm, 25 mm, 22 mm, 14 mm, 10 mm, 5 mm, and finally at the end of the growth pulse, before dormancy, 2.5 mm long.

Vepris lukei
Vepris lukei is unusual amongst E African unifoliolate Vepris species in possessing winged petioles.All other species have canaliculate or terete petioles except V. udzungwa (margins with minute patent wings c. 0.4 mm wide).It is also unusual in the extremely high number of secondary nerves, 22 -28 on each side of the midrib -resembling a Calophyllum (Calophyllaceae).Further, it is unique in this group in the highly reduced female inflorescences which appear to be 1-flowered.Examination of immature fruiting specimens gives no indication that they bore more than one flower.This species remains known from only two collections, and male and female flowers at anthesis remain to be obtained.
Vepris lukei has many similarities with V. robertsoniae and for this reason they may share a recent common ancestor and may well be sister species.Both species smell of fish when dried, have numerous parallel secondary nerves on the leaf blade and crater glands on the petiole, are glabrous except for the sepal margins, and lack paniculate inflorescences and nectar discs.For these reasons it is logical that material of V. lukei was formerly named as V. robertsoniae.However apart from ecology and geography, the two species differ in several key morphological characters (Table 1) and there is no doubt that they are distinct.
The geographical and ecological disjunction between the two very similar and probably sister species, one at low altitude in the coastal forests of SE Kenya, the other at high altitude in the Eastern Arc Mountains of Tanzania, is seen in several other genera.RECOGNITION.Similar to Vepris eugeniifolia, differing in the elliptic (rarely slightly ovate-elliptic) leaf-blades (vs ovate); flowers single along the rhachis in the inflorescences (vs in glomerules along the rhachis); fruits ovoid-ellipsoid or ellipsoid, apex acute or slightly rostrate (vs globose, apex rounded).Additional diagnostic characters are given below in the notes and in (Araceae), Angraecum teres Summerh., and Calyptrochilum christyanum (Rchb.f.) Summerh.(Orchidaceae).CONSERVATION STATUS.Vepris robertsoniae is known from nine locations with an extent of occurrence of 7825 km 2 and area of occupancy of 88 km 2 .Threats include degradation of habitat, such as by tourism infrastructure development, conversion to agriculture, pole cutting and logging.Therefore the species has been assessed as Vulnerable, VU B1ab(i,ii,iii,iv,v)+2ab(i,ii,ii i,iv,v) (Musili et al. 2020).PHENOLOGY.Flowering Nov. -June (-Aug.),fruiting in March -June (through the early dry season).ETYMOLOGY.Named for Mrs Anne Robertson of Kenya, pioneering collector of plants and early champion of the conservation of Kenya's coastal forests.Her studies have resulted in the discovery of plant species several others of which, apart from Vepris robertsoniae, are named for her, including Barleria robertsoniae I.Darbysh.9)) on each side of the midrib.The base of the blade is broadly acute or rounded, the margin at that point convex or straight, while in V. eugeniifolia it is usually concave.In dried specimens of the latter species the blade folds along the midrib, exposing the abaxial surface in old leaves on a sheet, while those of V. robertsoniae remain flat.Vepris robertsoniae has also been confused with V. lukei.See under the latter species for a discussion of their affinities and for diagnostic characters separating them (Table 1).
Numerous other taxa with a similar range to Vepris robertsoniae, also threatened and restricted to a set of Kenyan coastal Kaya forests, have been steadily documented in recent decades.Vepris welwitschii is a tree to 6 m tall that is most likely to be confused, and indeed has been, with V. africana, the only other unifoliolate species of the genus that occurs in Angola.The two can be distinguished using the characters cited below under the latter species and in the key to species.Most notably V. welwitschii has black fruit, not orange or red as is usual in the genus.
Lectotypification, synonymy and delimitation of this species was expertly performed by Lachenaud & Onana (2021).However, they opted to choose as lectotype a syntype at LISU for which there is no evidence that Bentham, credited author of the name, had seen.The syntype at BM does not have this deficiency.They also point out that this species remains incompletely known e.g.open flowers are not available.

V. sp. A of FTEA sensu
DISTRIBUTION.Tanzania, Uluguri Mts.SPECIMEN EXAMINED.TANZANIA, Uluguru Mts, Bunduki, fl. 10 Jan. 1935, Bruce 510.HABITAT.Submontane forest c. 1700 m alt.CONSERVATION STATUS.Vepris sp.A of FTEA has not been formally named and therefore does not appear on the IUCN Red List (iucnredlist.org).Provisionally it should be regarded as Critically Endangered (Possibly Extinct), CR(PE) D, since only a single plant was known at a site that has threats (Ndang'ang'a et al. 2007).Forest loss at Uluguru Mts, to which it seems to have been endemic, has been concentrated in the habitat of Vepris sp.A of FTEA (see discussion).Since to the best of our knowledge the species has not been seen for 87 years and its habitat has been largely cleared, there is a high probability that it is extinct.PHENOLOGY.Flowering in January (end wet season, start dry season), fruits unknown.VERNACULAR NAMES.None are recorded.NOTES.Kokwaro (1982: 18) first recognised this entity and stated "The specimen is somewhat similar to Teclea amaniensis except the stamens are clearly 8.It is, however, inadequate to formally describe a new species.It is also close to Vepris ngamensis but here the inflorescence is a panicle.See also V. mildbraediana, p. 23".It was treated by Mziray (1992: 78) as an "Insufficiently known taxon".This entity appears to be a most distinct and yet undescribed species.
10. Vepris amaniensis (Engl.)Mziray (pro parte 1992: 70).Types: Tanzania, Amani, Engler 565 (syntype, B destroyed); Warnecke 516 (syntype B, destroyed); neotype proposed here Tanzania "Tanganyika Terr., Amani", 5 April 1922, Salmon 171 neotype K barcode K000593352!; isoneotype EA).Teclea amaniensis Engl.(Engler 1905: 244;Kokwaro 1982: 24 pro parte).Vepris ngamensis I.Verd.(Verdoorn 1926: 399); Kokwaro (1982: 18).Type: Tanzania, E Usambara Mts, Amani, Engler 565 (holotype B destroyed; neotype selected here: Tanzania "Tanganyika terr., Amani, 4 April 1919, Salamani bin Kilwa G6172 (neotype K barcode K000593351!; isoneotype EA)).synon.nov.Meaning "from Amani", referring to the origin of the original specimens which were collected at or near Amani in the Usambara Mts of then German East Africa, Tanganyika, now Tanzania.VERNACULAR NAMES.None are recorded.NOTES.While finalising the key and skeletal species accounts for this paper, the first author found that the specimens assigned to this species at K, although concordant as a whole with the description in Kokwaro (1982), contained more than one species.Most of the material was not in agreement with the description of Teclea amaniensis in the original protologue of Engler (1905), nor the description of that species by Verdoorn (1926), which appears based on Engler's description (although is less precise).It seems that between the time of Verdoorn (1926), who only cited Warnecke 516K, and Kokwaro (1982), numerous additional specimens of at least one other unifoliolate shrub were collected in the Usambaras and adjoining areas, including Kenya.It was erroneously attributed to V. amaniensis, although accommodated in the expanded description of the species in FTEA.Most of this material has the apex of the petiole winged, hairy stems, an inflorescence shorter than the petioles, and often an occasional trifoliolate leaf among the predominantly unifoliate ones.These seem to represent a further new species that will be the subject of a future paper.None of the specimens of the putative new species were collected in Amani.In contrast, only seven surviving specimens (see specimens examined above) represent the species that fits the descriptions of Engler (1905) and of Verdoorn (1926).These have thin papery, elliptic leaflets with a length: breadth ratio of c. 2.5:1, glabrous stems, petioles which are terete at the base and canaliculate at the apex, inflorescences 0.9 -4 (-5) cm long, far exceeding (usually) the petioles, and leaves which are uniformly unifoliolate.All the specimens are from Amani except two from nearby Bulwa.A neotype has been selected from among them that matches the original description, as all the original material of Teclea amaniensis (the syntypes Engler 565 and Warnecke 516k in Herb.Amani) have been destroyed or lost.Although Mziray (1992) states that the latter is at K, there is an old annotation on a species cover that this specimen is "not here".In addition, the label of Salmon G 6171 (E African Agricultural research station, Amani, 5 April 1922) states in script contemporary with the original label, "The type is not in herb.Amani".This suggests that no duplicates were left by Engler's team in the Amani Herbarium (so they could not have been transferred to EA with the rest of that herbarium).
Vepris ngamensis is here formally added to the synonymy of the earlier published V. amaniensis.Treated by Mziray (1992: 78) as an "Insufficiently known taxon", V. ngamensis is only known from certainty from the type, Engler 565, also collected at Amani, but destroyed at Berlin.Although Kokwaro also attributed Drummond & Hemsley 3349 (not found, presumed missing) to V. ngamensis, he had not actually seen the original material.When Verdoorn described V. ngamensis in 1926 from material that had been annotated by Engler as Teclea ngamensis (Verdoorn 1926;Kokwaro 1982), she presumably missed the fact that this same specimen is one of the two syntypes of T. amaniensis.Comparing the original descriptions of V. ngamensis (Verdoorn 1926) with that of T. amaniensis Engler (1905) shows no point of morphological difference except in the number of stamens.The former having four (hence assigned to the genus Teclea) and the latter seven (so ascribed to Vepris).While some specimens cited above have four stamens, another (Salamani bin Kilwa) is annotated "Stamens 5 -6!".Although stamen number was formerly used to assign species to different genera, and has value as a species character, Mziray (1992) cited the range in variation of stamens from 4 -8 (sometimes on the same plant) in V. heterophylla as evidence that this is not in itself a reliable character for generic separation, nor even in some cases for separating species.We neotypify V. ngamensis above, in the absence of any original material, choosing material from the type location that matches its protologue most closely.
Vepris gossweileri (I.Verd.)Mziray nom.illeg.[non V. gossweileri I.Verd.(Verdoorn 1926: 399)].Mziray (1992: 72;Figueiredo & Smith (2008: 155); Langat et al. (2021: 1 -11).(Langat et al. 2021).In the Republic of Congo, the species is only known from a distinctive coastal thicket on white sand where it can grow with V. teva (Gosline & Cheek 2014;Cheek et al. 2014;Langat et al. 2022).Lachenaud & Onana (2021) report that V. africana is unusual for the genus in being androdioecious; pollen is produced in both flower types but male flowers have pistillodes only.This feature separates it from the similar but dioecious Comorian V. unifoliolata (Baill.)Labat, M.Pignal & O.Pascal.They attribute the presence of the species on São Tomé, a volcanic oceanic island, as probably resulting from dispersal by frugivorous birds or possibly by marine currents.The specimens cited above are additional to those reported in Lachenaud & Onana (2021), but do not increase the range of the species.
Vepris africana has been confused with V. welwitschii in Angola where they both occur, and these species are superficially very similar.However, the first has subsessile flowers (pedicels 0 -0.5 mm long), unilocular ovaries, 4 stamens, and orange fruits; the second has pedicels 1 -2.5 mm long, bilocular ovaries, 8 stamens, and black fruits.
Vepris hanangensis var.unifoliolata in leaf might be confused with V. simplex, which also occurs at this altitude.However, V. hanangensis var.unifoliolata, as in the typical variety, has long cylindrical fruits held in large persistent panicles, unlike the globose fruits on reduced racemes of V. simplex.
The collectors of the type stated that the tree grew up to 150 feet (= 45 m) tall.This would make it by the far the tallest growing of the African unifoliolate Vepris species.However, this is an error since the tallest tree in Karura Forest is no more than 15 m tall (QL pers.obs.2022).Only the two specimens cited are known to us.
DISTRIBUTION.Vepris simplex occurs from the Ethiopian Highlands in the vicinity of Addis Abeba southwards along the E African rift mountains through the highlands of Kenya and Tanzania (Kokwaro 1982;Friis 1992;Beentje 1994), reaching the Mafinga Mts of northern Malawi (White et al. 2001: 515) Vepris simplex is by far the most collected species of unifoliolate Vepris in tropical Africa, with 317 specimen records on gbif.org.The majority of the specimens were collected in the 1400 -2300 m altitudinal band in Kenya, extending to the Ethiopian Highlands, and southwards into the high mountains of northern Tanzania: e.g.Kilimanjaro, Mt Hanang, Mt Meru.It is morphologically uniform through much of this range, although flowers in Ethiopia are smaller than those in Kenya.In the Arusha area, the leaves are much longer than the norm, oblong and the length: breadth ratio is about 3: 1, leaves measuring c. 14.5 × 4.5 cm e.g.Willan 514 (K!).In the main part of its range the species often occurs on volcanic rocks such as lava flows and in at least some locations it is "exceedingly common" (Mt Kulal, Kenya, Bally 5582, K).Further investigation is needed of specimens from lower altitude evergreen forest areas around Morogoro, Tanzania, and in SE Kenya, e.g.Magogo & Glover 693 (Mrima Hill, K!), Brenan et al. 14519 (Shaitani Forest near Diani, K!), which are discordant with specimens in the rest of the range.They have large acuminate, papery leaves, with length: breadth ratio exceeding 2:1 (longer than the norm), with minute, green flowers.These might represent a separate variety or subspecies.
Hermaphrodite flowers with functional ovaries and only two stamens were observed on the otherwise male Bally 2578 (K!), by John Hutchinson (specimen annotation).
Trees are predated by elephants (Grimshaw 93341, 94409 both K!), which favour this species, and birds eat the fruits (Grimshaw 93341, K!), presumably dispersing the seed.This species has the smallest fruits (3 -5 mm diam.), borne in the greatest numbers per stem than any other unifoliolate Vepris species, which may be an adaptation to bird dispersal and contribute to its having the largest range and being the most frequent of any of the species.White et al. (2001: 516) classify the species as a Sub-Afromontane endemic while Friis (1992) suggests it is an Afromontane endemic (Timberlake 2021c).
Vepris simplex is confused with V. eugeniifolia.See diagnostic characters under the last species.ETYMOLOGY.Originally named Toddalia simplicifolia by Engler for the unifoliolate (simple not compound) leaves, now known not to be a diagnostic specific character as it must have seemed when first published.Vepris simplex, the new name, coined here, is intended as a convenient, shorter alternative name which is needed since the epithet simplicifolia is not available for this species (see Notes, below).(Turkana), and the use of wood for roof beams and other artefacts is reported (Beentje 1994).NOTES.The only surviving original material of Engler (1895) that has been located is the isotype at EA.It is sterile, although the protologue is based on flowering material.Therefore, an epitype is needed to fix the application of the name to the species and remove ambiguity.There seems some uncertainty about the locality.On the EA isotype label an undecipherable word followed by "Hochwald" (high forest) is written.The location has been inferred or interpreted at a later date by a note in pencil on the label as "Usambara Mts".However, in the protologue "Usambara Mts" are not mentioned, although an altitudinal range of 1300 -1600 m is given.The only place name given on the label is Bagamoyo, which was the capital of what was then German East Africa and is a historic coastal port town relatively near to the Usambaras.Polhill & Polhill (2015: 199) give an itinerary for Holst in 1892, the year in which he collected the original specimen.He was entirely in FTEA T3 (the botanical province containing the Usambara Mts).Given these facts, an epitype has been selected of a fertile specimen representative of the species, also from the T3 area; this is Parry 222 (EA, K), chosen because it is of good quality, is in flower showing the representative large male flowers with four stamens, and with the thick, ovate-elliptic leaflets with rounded apices that together unambiguously indicate this species.
When Mziray (1992) made the combination Vepris simplicifolia, he was probably unaware that the name was already occupied by V. simplicifolia Endl.(Endlicher 1833), making Mziray's combination an illegitimate later homonym under the Code (Turland et al. 2018).The last name was coined for a plant from Norfolk Island in the western Pacific.It is the basionym for Sarcomelicope simplicifolia (Endl.)T.G.Hartley (1982: 369) of Australia and New Caledonia, which has many local names and uses (Hartley 1982).Therefore, a new name is needed for the African taxon, which is addressed above.Mziray (1992), when making the combination Vepris simplicifolia, incorrectly gave the authorship as (I.Verd.)Mziray, mistakenly attributing authorship of the basionym to Verdoorn.However, Verdoorn had made it clear that she was making a combination based on Engler's (1895) Toddalia simplicifolia (Verdoorn 1926).Therefore, the correct authorship of her name is Teclea simplicifolia (Engl.)I.Verd.and that of Mziray's is V. simplicifolia (Engl.)Mziray.

A note on the Eastern Arc Mountains and Coastal Forests of East Africa
The three new species published in this paper are restricted to the Eastern Arc Mountains and Coastal Forests (EACF) of East Africa, in Tanzania and Kenya.The EACF form an archipelago-like phytogeographical unit well-known for high levels of species endemism in many groups of organisms (Gereau et al. 2016).Among the better-known mountain blocks are the Nguru Mts, the Udzungwa Mts, the Uluguru Mts, and the Usambara Mts.Supported by moist air currents from the Indian Ocean, the Eastern Arc Mountains alone have been said to have 223 species of endemic tree (Lovett 1998), and are variously stated to have 800 (Tanzania Forest Conservation Group (undated), or as many as 1500 species (Skarbek 2008) of endemic vascular plants, although the most reliable current total is 552 (R.E.Gereau, pers.comm.).In herbaceous groups such as the Gesneriaceae, over 50% of the taxa (23 endemic species and a further nine endemic infraspecific taxa) for Tropical East Africa (Uganda, Kenya and Tanzania) are endemic to the Eastern Arc Mts (Darbyshire 2006), and in the Acanthaceae, there are numerous endemic species in multiple genera endemic to the Eastern Arc Mts, e.g.Stenandrium (= Stenandriopsis) warneckei (S.Moore) Vollesen, Isoglossa bondwaensis I.Darbysh.,I. asystasioides I.Darbysh.& Ensermu, and Sclerochiton uluguruensis Vollesen (Darbyshire 2009;Darbyshire et al. 2010;Darbyshire & Ensermu 2007).In terms of documented plant species diversity per degree square, the Eastern Arc Mts are second in tropical Africa only to Southwest Cameroon in the Cross-Sanaga Interval of West-Central Africa (Barthlott et al. 1996;Cheek et al. 2001).Several forest genera have disjunct distributions, being found only in the Cross-Sanaga Interval and in the EACF and not in between, e.g.Zenkerella Taub.and Kupea Cheek & S.A. Williams (Cheek et al. 2003;Cheek 2004b).The EACF include the sole representatives of plant groups otherwise restricted on the continent to the forests of Guineo-Congolian Africa, e.g.Afrothismia Schltr.and Ancistrocladus (Cheek & Jannerup 2006;Cheek et al. 2000;Taylor et al. 2005).Extensive forest clearance in the EACF within the last 100 -150 years has removed forest from some mountains entirely, and reduced forest extent greatly in others.Since the 1970s more than 12% of these forests have been cleared (Tanzania Forest Conservation Group (undated).However, forest clearance has appeared to stabilise in the last ten years (Tanzania Forest Conservation Group (undated) in many but not all areas important for plant conservation, giving hope that species extinctions can be avoided, or at least kept to a minimum.

Conclusions
The published and provisional extinction risk assessments of the 13 unifoliolate continental African Vepris taxa treated in this synopsis indicate that all but two are threatened or Near Threatened.Thankfully, the three new species to science published in this paper are all at the lower level of extinction risk, as a result of the higher levels of protection in the Udzungwa Mountains National Park of Tanzania (V.lukei and V. udzungwa), and the local community protection of the indigenous people of SE Kenya of their Kaya forests (V.robertsoniae).However, the future for the three species indicated as Critically Endangered seems fragile or even non-existent.The forest habitat of V. laurifolia in western Africa (Guinea to Ivory Coast) is steadily being reduced by development projects of multiple sorts including mining and hydropower, and by clearance of the last scraps for agriculture.The two species restricted to the Tanzanian Uluguru mountain forests are of highest concern because they were last seen nearly 100 years ago, each from a single plant (so far as we are aware), and their forest habitat has seen massive clearance.According to Ndang'ang 'a et al. (2007), the Ulugurus had the highest losses of forest of all Tanzanian EACF areas from the 1970s to 2000, with about 12% loss.All forest is considered to have been lost below 1800 m alt.(Eastern African Plant Red List Authority pers.comm.to first author).It may be that both species are already extinct, in the case of Vepris sp.A of FTEA (only recorded below 1800 m alt.), even before it has a scientific name or a formal IUCN conservation assessment published.Until species are scientifically named, it is difficult for an IUCN conservation assessment to be published (Cheek et al. 2020b, although there are exceptions, as in V. robertsoniae of this paper).Most new species to science published today, such as those in this paper, are rangerestricted, meaning that they are almost always automatically threatened, although there are exceptions, such as the widespread V. occidentalis Cheek & Onana (Cheek et al. 2019a).Documented extinctions of plant species are increasing (Humphreys et al. 2019) and recent estimates suggest that as many as two fifths of the world's plant species are now threatened with extinction (Nic Lughadha et al. 2020).Global extinctions of African plant species continue apace.At the foot of the Udzungwa Mts, the achlorophyllous mycotrophs Kihansia lovettii Cheek and Kupea jonii Cheek (Triuridaceae, Cheek 2004b) are likely extinct as a result of the placement of the Kihansi hydroelectric dam, not having been seen since its construction in 1994 (28 years ago), despite targeted searches.Although not directly threatened by development, another mycotroph, this time in one of the forest fragments of SE Kenya, Afrothismia baerae (Thismiaceae, Cheek 2004a), has also not been found despite monitoring in the last 10 years.Global extinctions have also been reported in Guinea, such as Inversodicraea pygmaea G.Taylor, and in 2022 after first collection in 2018, Saxicolella deniseae Cheek (both Podostemaceae, both extinct due to hydropower construction, Couch et al. 2019, Cheek et al. 2017;Cheek et al. 2022c).New extinctions have recently been reliably reported from Gabon (Moxon-Holt & Cheek 2021;Cheek et al. 2021) andCameroon (Cheek &Williams 1999;Cheek et al. 2018c, Cheek et al. 2019b), including species of Vepris (Cheek et al. 2018a).If future extinctions are to be avoided, improved conservation prioritisation exercises are needed, such as Important Plant Area programmes (Darbyshire et al. 2017a), supported by greater completion of Red List assessments, although this can be slow and problematic (Bachman et al. 2019), and, globally, only 21 -26% of plant species have conservation assessments (Bachman et al. 2018).Where possible, as an insurance policy, seed banking and cultivation of threatened species in dedicated nurseries are urgent.Above all, completion of botanical taxonomic inventories is needed to feed into these exercises; otherwise, we will continue to lose species before they are even discovered for science, and certainly before they can be investigated for their potential for beneficial applications.New compounds to science with high potential for humanity are being discovered in Vepris species each year (e.g.potent antimicrobial compounds in Vepris africana, Langat et al. 2021).Such discoveries will not be possible if species extinctions are allowed to continue.

Fig. 2 .
Fig. 2. Vepris udzungwa.A habit, fruiting branch; B young stem showing hairs; C older stem, showing dense lenticels and persistent hairs; D leaf, adaxial surface; E abaxial leaf surface showing conspicuous black oil gland dots; F stem node and immature infructescence with leaf, showing winged petiole; G fruit, side view, showing basal and apical asymmetry (stigma arrowed); H transverse section of fruit showing seed.A -E, G & H from Luke et al. 9109; F from Luke et al. 8639.drawn by andrew brown.

Fig. 3 .
Fig. 3. Vepris lukei.A habit, fruiting branch; B large leaf and attachment to stem; C base of leaf-blade, articulation and winged petiole showing gland, together with transverse section of petiole to show wings; D abaxial surface of leaf-blade showing reticulate quaternary nerves and inconspicuous oil glands; E mature fruit showing raised black oil glands on surface.A & D from Luke et al. 9166; B, C, & E from Luke & Luke et al. 10343.drawn by andrew brown.

Fig. 5 .
Fig. 5. Vepris robertsoniae.A habit, flowering branch; B base of leaf-blade showing articulation with canaliculate petiole and glands; C abaxial surface of leaf-blade showing surface with black oil glands; D male inflorescence; E male flower bud; F male flower with stamen and petal removed to show pistillode; G female inflorescence; H female flower; J longitudinal section of female flower showing unilocular ovary with a single pendulous ovule; K stem node; L fruit.A, B, D -F from Luke et al. 2679; C from Luke et al. 1670; G -J from Robertson 6852; K, L Luke et al. 1771.drawn by andrew brown.
CONSERVATION STATUS.Vepris simplex is listed as Least Concern in view of its vast range and numerous locations, and few specific threats.It is in numerous protected areas in Kenya and Tanzania (Timberlake 2021c).PHENOLOGY.Flowering May -July (dry season), fruiting Sept. -Nov.(dry season, Ethiopia); flowering June -Feb.,fruiting June -Jan.(dry and wet season, Kenya), and flowering Nov. -June, fruiting Feb. -June, Aug. (mainly wet season, Tanzania).
. It is also absent, or extremely rare, from most of Liberia, where most of the surviving forest in West Africa remains.Botanical inventory work there over many years by C. Jongkind has not discovered this species (C.Jongkind pers.comm.to M. Cheek 2014).None of the large national parks (e.g.Tai National Park, Gola Rainforest National Park) are known to support it, despite botanical inventory effort.The species was assessed as Critically Endangered (CR) under criterion C2a(i) since fewer than 250 mature individuals are thought to exist and there is a continuing decline in the number of mature individuals, with fewer than 50 individuals in each subpopulation (Cheek 2017).In Guinea the species is included in two Tropical Important Plant Areas, Kounounkan and Ziama (Couch et al. 2019).PHENOLOGY.Flowers mainly in April & May (start of wet season).Fruit in Sept. (end wet season).ETYMOLOGY.Named for the resemblance of the leaves to those of the genus Laurus (Lauraceae).

Table 1 .
Diagnostic characters separating Vepris robertsoniae from Vepris lukei.Named by Martin Cheek for William Richard Quentin Luke, better known as Quentin Luke (1952 -), lead collector of all known specimens of Vepris lukei, and the most prolific living field botanist in East Africa.He is a Kenyan botanist and is Research Associ- ETYMOLOGY.ate of the East African herbarium (EA).Full biographical and bibliographical information can be found inPolhill & Polhill (2015: 276 -277).He has brought to light previously unknown species from across Africa e.g., in eastern Democratic Republic of Congo: Keetia namoyae O.Lachenaud & Q.Luke (Lachenaud et al. 2017) and from Mali and Guinea the only endemic African Calophyllum, C. africanum Cheek & Q.Luke (Cheek & Luke 2016; Couch et al. 2019).He has discovered numerous new species of plants especially in Kenya and Tanzania, such as the incredible Tanzanian tree acanth Barleria mirabilis I.Darbysh.&Q.Luke(Darbyshire &  Luke 2016).He has also collected and described many other novel plant species from Tanzania and Kenya.More than ten species are named for him, e.g.Keetia lukeiBridson (Rubiaceae, Bridson 1994), including also the Tanzanian species Cola quentinii Cheek (Cheek & Dorr 2007) and Cola lukei Cheek (Cheek 2002).Most recently Lukea Gosline & Cheek, a new genus to science has been named in his honour (Cheek et al. 2022b).VERNACULAR NAMES & USES.None are known.NOTES.Luke et al. 10343 had previously been identified as Vepris robertsoniae ined., and Luke & Luke 9166 as "Vepris sp., not matched" by Kaj Vollesen in 2004.Luke et al. 10343 has new shoots with
robertsoniae: the male flowers have four stamens.It is not known whether the stamens in female flowers are non-functional, not producing viable pollen (staminodes) or functional, but the last is suspected.Characters separating the two species are given in Table2.The two are easily separated vegetatively since V. robertsoniae has a canaliculate (not terete) petiole and the blade has 8 -15 lateral nerves (not6 -8 (- (Hooker 1863)Onana 2021) fruits in September, mature fruits in October(Lachenaud & Onana 2021), i.e. late wet season, early dry season.ETYMOLOGY.Named for the Austrian Friedrich Welwitsch (1806 -1872), the most famous botanical collector of specimens in Angola, who collected the original specimens from which the species was described.He is also commemorated by the genus Welwitschia Hook.f.(Hooker 1863).
Verdcourt 122), Cephalosphaera usambarensis (Warb.)Warb.(Myristicaceae),Anisophylleaobtusifolia Engl.&Brehmer (Anisophylleaceae, Lovett et al. 869); 900 -1000 m alt.CONSERVATION STATUS.Timberlake (2021a)in assessing the extinction risk of Vepris amaniensis states: "Some of the forests from which Vepris amaniensis is recorded, particularly in Tanzania, are under threat of clearance for small-scale and subsistence agriculture.The extent of occurrence (EOO) is calculated at 210,887 km 2 and the minimum area of occupancy (AOO) is 104 km 2 .As there are only nine recorded locations . . . the species is assessed as Vulnerable" VU B2ab(ii,iii,v).However, it has recently been discovered (see Notes below) that this species is restricted to near Amani and Bulwa in the Usambara Mts, with a far smaller AOO and EOO, and so will merit re-assessment, likely as EN.
transferred T. hanangensis to Vepris, but did not make a new combination in Vepris for T. hanangensis var.unifoliolata.Young plants of this variety frequently have some 3-foliolate, some 2-foliolate and a majority of 1-foliolate leaves.Unifoliolate leaves from young plants are exceptionally large, up to 30 × 12 cm (White et al. 2001)en forest, riverine thicket, evergreen rocky bushland, drier types of upland forest and woodland with Juniperus and Acacia, extending into the understorey of Podocarpus forest(Friis 1992), and in Malawi in montane thicket(White et al. 2001).