Research Article |
Corresponding author: Abraham E. van Wyk ( braamvanwyk@gmail.com ) Academic editor: Brecht Verstraete
© 2024 Abraham E. van Wyk, Richard G.C. Boon.
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Citation:
van Wyk AE, Boon RG (2024) Marijordaania (Celastraceae: Cassinoideae), a new monotypic genus from South Africa. Plant Ecology and Evolution 157(1): 100-112. https://doi.org/10.5091/plecevo.116544
|
Background and aims – The species name Gymnosporia filiformis, validly published in a 1927 revision of South African Celastraceae, was based on syntypes representing both Maytenus cordata and a new species. Subsequent studies supported the distinction of Gymnosporia and Maytenus as separate genera. From 1984, this new species was provisionally referred to by the designation “Maytenus sp. A”. Maytenus in Africa comprises a diverse group of species not closely related to the New World members of Maytenus. Here, we clarify the taxonomic identity and generic placement of G. filiformis, and compare it with species with which it has historically been confused.
Material and methods – Descriptions and observations are based on extensive field work, supported by conventional taxonomic methods, including study of relevant literature and herbarium collections.
Key results – The name Gymnosporia filiformis is lectotypified using a syntype representing “Maytenus sp. A”. We describe a new monotypic genus, Marijordaania, to accommodate “Maytenus sp. A”, and propose the new combination, Marijordaania filiformis. An amended description is provided for the species. It is morphologically quite different from Maytenus cordata, but also from Maytenus acuminata and Maytenus abbottii, species with which it has been confused in the past. Marijordaania filiformis grows in the understorey of evergreen Scarp Forest and is relatively rare, known with certainty only from KwaZulu-Natal and the Eastern Cape provinces, South Africa. A preliminary conservation assessment suggests that M. filiformis should be classified as Endangered.
Conclusion – Marijordaania filiformis is a taxonomically isolated species with no apparent close relatives among African Celastraceae. While the generic status of other African members of Maytenus still needs to be resolved, it would not impact the taxonomic status of Marijordaania. In a recent key for Celastraceae genera based on morphology and geography, M. filiformis ran to the genus Euonymus, but this association lacks support from available molecular evidence.
Celastreae, Celastroideae, gutta-percha, lectotypification, Maputaland-Pondoland-Albany Hotspot, new genus, Pondoland Centre, Scarp Forest, taxonomy
In her taxonomic revision of South African Celastraceae,
In the early 1980s, Anthony Thomas Dixon (Tony) Abbott (1936–2013), a farmer and amateur botanist, undertook in-depth explorations of the flora in southern KwaZulu-Natal and Pondoland, South Africa. Based on field observations, Abbott questioned the validity of botanists applying, at the time, the name Maytenus acuminata to what he perceived as four distinct species growing sympatrically in the Umtamvuna Nature Reserve. In response to Abbott’s claims, one of us (AEvW) re-evaluated
More recently, extensive taxonomic studies on Celastraceae in southern Africa provided robust evidence supporting the distinction of Gymnosporia and Maytenus as separate genera, a perspective previously expressed by
Comprehensive analyses considering both morphological and molecular characters supported the recognition of a New World Maytenus and a primarily Old World Gymnosporia (
Descriptions and observations in the present paper are based on extensive field work conducted over many years by both authors in the natural habitat of plants, including members of the Celastraceae, primarily focusing on KwaZulu-Natal and the Eastern Cape regions of South Africa. This was supplemented by study of relevant literature and herbarium collections. One of us (RGCB) visited Krantzkloof Nature Reserve in Durban and Umtamvuna Nature Reserve near Port Edward to conduct fieldwork and morphological observations on specifically Maytenus sp. A, as well as associated members of Maytenus.
The following herbaria were visited to examine specimens: NH, NU, PCE (now incorporated in NU), PRE, PRU, and UDW (now incorporated in NU). Photographs of selected specimens were obtained from SAM and BOL. Also consulted were high-resolution images of type material available in K on the Internet through
In the section “Additional material examined”, locality citations were reproduced as per the specimen labels. In some cases, the spelling of the locality name was either corrected or an alternative spelling was provided. The correction or alternative spelling is included in square brackets. Specimens are arranged according to the Degree Reference System proposed by
A preliminary conservation assessment was conducted using the standard procedures based on IUCN guidelines (
In this paper, we follow the recently proposed more natural classification of the Celastrales by
When
In the protologue, Davison did not explicitly explain her concept of the taxon, nor did she give an indication of its morphologically most similar relatives. However, the choice of specific epithet and the key provide clues. In the description, the peduncles (technically “pedicels”) are described as “...filiformi, circa 2 cm. longi...”. In her identification key to the infrageneric taxa of Gymnosporia, a primary distinction is made between two groups of taxa based on inflorescence type: “Fasciculatae”, for those with flowers in fascicles (probably representing very condensed cymose inflorescences), and “Cymosae”, for those with flowers in cymes. Gymnosporia filiformis appears in the key to the “Cymosae”. In the first couplet of this key, Gymnosporia cordata is differentiated by its leaves being sessile and cordate at the base, whereas G. filiformis conforms to the contrasting statement with leaves petiolate and not cordate at the base. The key then progresses through several statements to a final couplet, which differentiates between G. filiformis and G. acuminata. Here, the lead for G. filiformis states: “Pedicels about 2 cm. long or more; leaves rounded at the base”, whereas the contrasting statement for G. acuminata reads: “Pedicels up to 0.5 cm. long; leaves cuneate at the base”. From all of this, it is clear that Davison recognised G. filiformis as a taxon distinct from both Maytenus cordata and Maytenus acuminata based on its leaves being petiolate without a cordate base, and flowers with slender and relatively long pedicels (about 20 mm or more).
A weakness of Davison’s key is her rather narrow concept of Maytenus cordata. While “typical” forms of this species are easily recognised by essentially sessile leaves with a cordate lamina base, other forms are also frequently encountered with short, but distinctly petiolate leaves with a rounded or cuneate lamina base. To add to the potential confusion, Maytenus cordata occasionally has flowers with relatively long pedicels, though these rarely exceed 20 mm in length.
A study of the four syntypes of Gymnosporia filiformis showed that three of the four gatherings, namely Wood 8257 (NH9205), Wood 9987 (NH10543), and Wood 11608 (SAM0007363-9), represent Maytenus cordata. Moreover, the illustration provided for Gymnosporia filiformis in the protologue was based on Wood 11608, thus it depicts Maytenus cordata. Davison herself was also confused by the similarity between these two species, seeing that a duplicate of one of the syntypes, Wood 9987 (K), was cited under Maytenus cordata (
In the original description of Gymnosporia filiformis (
According to Art. 9.3 of the International Code of Nomenclature for algae, fungi, and plants (
The delineation of genera within Celastraceae has historically faced difficulties due to the limited presence of distinctive morphological characters in both vegetative and reproductive features. This is further complicated by high levels of morphological homoplasy. However, recent advancements in molecular studies have proven informative in recognising specific clades, thereby aiding in the establishment of more natural generic and higher-level taxonomic groupings (e.g.
In a subsequent comprehensive phylogenetic analysis of the Celastrales conducted by
Prominent morphological differences between Marijordaania filiformis and the three species of Maytenus with which it has been confused in the past. The true generic status of these three African species of Maytenus still needs to be assessed.
Character | Marijordaania filiformis | Maytenus cordata | Maytenus acuminata | Maytenus abbottii |
Habit | Semi-scandent shrub or slender tree to 6 m | Erect shrub or slender tree to 4 m | Erect shrub or tree to 15 m | Erect shrub, occasionally semi-scandent, rarely a slender tree to 4 m |
Elastic threads (gutta-percha; trans-1,4-polyisoprene) | Absent | Present | Present | Present |
Inflorescences | Peduncles with 1- or 2(3)-flowered cymes | Peduncles with 2–10-flowered cymes | Peduncles with 2–20-flowered cymes; rarely fasciculate | Peduncles with 1–4 abbreviated racemes |
Pedicels (length) | (20–)35–40(–45) mm | 5–12(–20) mm | 5–10(–14) mm | 5–10(–12) mm |
Pedicels (articulation) | Ca 0.5 mm above the base | 1–2.5 mm above the base | 2–4 mm above the base | At the base |
Flowers (merosity) | 5-merous | 5-merous | 5-merous | 4-merous |
Flowers (predominant colour) | Pink-red | Pale green | Cream or white, often tinged pink, rarely pink-red | White or cream, occasionally tinged pale pink |
Disc (presence) | Present | Present | Present | Obscure (absent?) |
Disc (morphology) | Fleshy, divided into 5 discoid subunits (lobes), the latter convex below filaments | Fleshy, discoid, concave, not subdivided | Fleshy, discoid, concave, not subdivided | No distinct disc; perhaps represented by a narrow strip of tissue around the base of the style |
Stamens (insertion) | Filaments arising from centre of disc lobes | Under margin of disc | Under margin of disc | Filaments flattened and widening towards base to form a staminal tube |
Ovary (number of locules) | 5 | (1–)3 | 3 | 2–4 |
Seeds (postchalazal vascular bundles) | Absent | Absent | Absent | Present |
Aril (colour) | White | Orange | Orange | White |
Aril (surface) | Glabrous, sparsely puberulent towards base | Glabrous | Glabrous | Puberulent |
In
The identification of diagnostic morphological characters for the new genus proved quite difficult, given the prevalent trend of subtle morphological variation among currently acknowledged genera in the family. In plant taxonomy, two types of diagnoses can be employed: one providing essential characters and another offering differential characters (
Marijordaania filiformis (Davison) A.E.van Wyk & R.G.C.Boon.
A member of Celastraceae, subfamily Cassinoideae (sensu
As for the species.
The generic name commemorates Dr Marie Jordaan (née Prins) [1948–], in recognition of her considerable contributions towards the taxonomy of the southern African flora in general, and the Celastraceae in particular.
Gymnosporia filiformis
Davison, Bothalia 2: 311. 1927. (
“Maytenus sp.” (
“Maytenus sp. A” sensu
Marijordaania filiformis. A. Flowering branchlet. B. Inflorescence, showing short peduncle with bracts and a single flower in bud (just before anthesis). C. Flower, viewed from above. D. Flower, viewed from below. E. Transverse section of ovary, showing five locules, each with two ovules. F. Single stamen, inserted on swollen, discoid, lobe of disc. G. Selection of almost mature, intact fruit, plus one old, dehisced and empty one; shape varies depending on the number of ovules that develops into seed. H. Seeds, three with the aril intact, one with aril removed. Scale bars: A, G = 20 mm; B–D, H = 5 mm; E, F = 1 mm. A based on Jordaan 394 (NH), B–F from Abbott 1798 (PRU; spirit collection), and G–H from Abbott 6082 (PRU; spirit collection). Drawn by Daleen Roodt.
SOUTH AFRICA • Natal [KwaZulu-Natal], Inanda [2830DB]; recd. Jan. 1880 (K), Dec. 1881 (BOL); Wood 734; lectotype (designated here): K [K000035898]; isolectotypes: K [K000035899], BOL, PRE [PRE0259491-0].
Marijordaania filiformis. A. Flowering branchlet, viewed from below; note distichous arrangement of leaves. B. Branchlet showing flowers borne on long, slender, and pendant pedicels; also present is a young developing fruit. C. Flowers; note each stamen inserted on a swollen discoid subunit of the disc. D. Flower viewed obliquely from the side. E. Flower, viewed from below. F. Fruit, intact and almost mature. G. Fruit, dehisced, showing a dark-coloured seed partly covered by a white aril. Photographs A–C by Sharon Louw; D–G by Richard Boon.
Evergreen, slender, lax, semi-scandent, sometimes ascending, shrub or rarely a small tree up to 6(–8) m tall, unarmed, glabrous, leaves and bark lack elastic threads on breaking; new growth reddish, reddish brown, or bronze. Trunk usually multi-stemmed from ground level, up to ca 10 stems, usually fewer, possibly producing suckers from subterranean roots, individual stems up to ca 40 mm diam. near ground level. Bark smooth, grey-brown, much-covered in crustose lichen. Branches slender, flexuose, initially reddish brown, soon changing to green, 4-lined due to 2 narrow wings extending down the stem from the leaf base, with pale, moderately dense lenticels when young, finally terete, grey, lenticels obscure. Stipules marcescent, free, paired, lateral, lanceolate, 0.4–0.5 mm long. Leaves simple, alternate, distichous, ovate or elliptic, (15–)40–65(–92) × (10–)20–35(–45) mm; base rounded; apex acute to narrowly acute or acuminate, rarely obtuse, mucronulate, gland-tipped; margin flat, glandular-denticulate to glandular-serrulate mainly in the distal half to two-thirds, with up to 18 irregularly- and well-spaced, dark teeth on each margin, usually fewer; chartaceous to thinly coriaceous; mature leaves dark green above, some leaves with ca circular patches of crustose, grey lichen, paler green below; midrib on adaxial surface prominent, raised and angled, raised in the proximal half of the abaxial surface and rounded, becoming plane in the distal half, brochidodromous; secondary (principal lateral) veins ca 6–10 pairs, raised on adaxial surface, obscure on abaxial surface; tertiary veins coarsely reticulate, distinct on adaxial surface, indistinct below. Petiole ca 1.5–2 mm long, canaliculate above, decurrent, forming raised lines on branchlets. Inflorescences cymose, solitary, axillary, 1- or 2(3)-flowered, peduncle 2–5 mm long, tipped by persistent lanceolate bracts 1–2 mm long, ca 1 mm wide. Flowers bisexual, actinomorphic, pentamerous, ca 10 mm in diam.; pedicel filiform, pendant, (20–)35–40(–45) mm long, articulated ca 0.5 mm above the base, reddish green. Sepals free, unequal with 2 smaller outer ones, 3 larger inner ones, green, pinkish green to pink-red with margin entire, whitish, the outer ones ca sub-orbicular, ca 1 × 1.5 mm; margin entire; apex rounded and shortly mucronate, the inner ones sub-orbicular, ca 1.5 × 2.5 mm, apex rounded or obtuse, shortly mucronate. Petals sub-orbicular, ca 4.5–5 × 3.5 mm; base cuneate; apex rounded; margin undulate, entire, central part pink-red, fading distally into a white margin. Stamens with filaments arising from the centre of the disc lobes, erect, ca straight, terete, gradually widening to the base, ca 0.5 mm long; anthers dorsifixed, directed upwards, dehiscing by longitudinal slits; pollen pale yellow. Disc prominent, ca 3 mm in diam., comprising 5 discoid subunits, each ca 1 mm in diam., fleshy, green to pale yellowish green or yellowish. Ovary almost entirely included in and adnate to the disc, 5-locular with 2 erect collateral ovules in each locule; style ca 0.5 mm long; stigma capitate. Fruit capsular, 1–5-locular, flattened globose and lobed, ca globose when 1-locular, dehiscing loculicidally to the base, ca 8–12 × 15–17 mm in dry specimens, tipped by the persistent style and stigma, smooth, light yellowish green, calyx persistent, sepals revolute. Seed red-brown, 7–8 × 5 mm; aril white enveloping more than half the seed, margin irregularly lobed, smooth-surfaced but sparsely puberulent towards the base.
Floristically, the range of Marijordaania filiformis (Fig.
Topographical map showing the known distribution of Marijordaania filiformis. Solid circles indicate localities supported by herbarium specimens; open circles with small central dots depict sight and other records in need of vouchers (details in text). The insert shows a map of southern Africa with names of countries; the grey rectangle indicates the area depicted by the topographical map.
Additionally, there are four other possible or probable localities for the species (Fig.
The third potential locality for the species is the Ozwatini Plateau north of Ndwedwe in central KwaZulu-Natal, South Africa. A specimen from here (Scott-Shaw 2057 in CPF, now incorporated in NU) might be M. filiformis, but the flowers are too young to be certain. The fourth possible locality is the Majuleni Forest located along a small tributary of the Mtentu River near Baleni, Pondoland, Eastern Cape, South Africa. This locality is almost certainly correct, but it is based on a photographic record with no voucher specimen having been preserved (Dorothy McIntyre pers. comm. 4 Mar. 2019; Sinegugu Zukulu pers. comm. 28 Jul. 2023).
Marijordaania filiformis occurs in the understorey of sub-tropical Scarp Forest (vegetation type FOz5 of
The geology of the parent rock where the species is found can be igneous, metamorphic or sedimentary (
Essentially, very little is known about animals associated with M. filiformis. Whitefly pupae of the family Aleyrodidae were found on the leaf undersides at the Krantzkloof Nature Reserve, Durban. It is noteworthy that similar whitefly pupae were observed in a photograph, most likely of our species, in the Majuleni Forest, Pondoland.
Fertile material is infrequently encountered and, when present, there are usually few flowers and fruit. Flowering and fruiting are apparently not linked to seasons as fertile material has been collected in most months of the year.
English vernacular names in use are “red-flowered false silkybark” and “red-flowered silkybark”, and Afrikaans names are “rooiblomvalssybas” and “rooiblomsybas” (
Endangered: (EN) C2a(i). The EOO of M. filiformis was estimated at 6865 km2. The AOO calculated was 56 km2 using the recommended 2 km cell width. The areas were determined using the localities where the species has been collected. Only one herbarium collection included geographic coordinates, therefore the localities used in the assessment are not precise. A 2 km cell width seems reasonable given that the species occurs in relatively small, discrete forest patches surrounded by unsuitable habitat. A 2 km cell drawn around the estimated localities includes most of the forest at that location.
Marijordaania filiformis occurs in 15 subpopulations as defined by the
Most subpopulations are separated from the nearest subpopulation by large distances and unsuitable habitat, thus genetic or demographic exchange seems unlikely. In three cases, two subpopulations are about 6 km apart and there are forest links in between. In these instances, it could alternatively have been assumed that plants at the two localities belong to a single larger population. However, M. filiformis has specific habitat requirements, is rare within suitable habitat, and produces few flowers and fruits, which means that subpopulations separated by even fairly small distances are probably functionally disconnected.
Several of the subpopulations are protected in statutory conservation areas. Subpopulations that are inadequately protected are threatened by degradation of their forest habitat. Marijordaania filiformis is not known to be used for traditional purposes. The species occupies rocky habitats, which affords it protection from wildfire. Individuals that grow adjacent to rivers are probably at risk of being killed or damaged by floods. Both wildfire and flooding are likely to increase across the species’ range due to climate change. The species probably reproduces mainly clonally, which means that genets in the sub-populations are genetically identical.
There is no evidence for a large reduction in population size, which is required for threatened Category A. The EOO and AOO distribution thresholds of Vulnerable and Endangered respectively are met for Category B, but only one of the three conditions is met (B2(b)), therefore the taxon does not qualify in this category. It is possible, perhaps even likely, that the estimate of 1500 mature individuals is too high and the species qualifies for Vulnerable in Category D1, but unfortunately there is no evidence to support this, and there are insufficient data to do the quantitative analysis required to qualify for Category E.
While the size and rate are unknown, the species’ population size is inferred to be declining due to apparent ongoing infrequent recruitment and establishment of new plants, and habitat degradation at some localities. With an estimated population size of < 2500 individuals and < 250 mature individuals in each sub-population, the species should be classified as Endangered (EN) C2a(i). There is a need to survey the subpopulations more fully to get a better understanding of threats and population size and trends.
SOUTH AFRICA – KwaZulu-Natal • Zululand, Ngoye [Ongoye Forest Nature Reserve], [2831DC]; 18 Aug. 1985; Abbott 2682; PCE, PRU • Mtunzini District, left hand branch of Ngoya Forest Road [Ongoye Forest Nature Reserve], [2831DC]; 20 Feb. 1961; Wells & Edwards 95; NU, PRE • Krantzkloof Nature Reserve, (2930DD); 12 Jul. 2014; Boon 69; NH • Krantzkloof Nature Reserve, (2930DD); 29 Jan. 1984; Jordaan 304; NH • Kloof Nature Reserve [Krantzkloof Nature Reserve], Pinetown District, [2930DD]; 16 Feb. 1966; Moll 3021; NU, PRE • Durban, Krantzkloof Nature Reserve, (2930DD); 20 Jan. 1988; Van Wyk 8243; PRU • Krantzkloof Nature Reserve, north of Kloof Falls Road, (2930DD); 20 Jan. 1988; Williams 158; NH • Ntimbankulu [Ntimbankulu Nature Reserve], (3030CB); 21 Jun. 1984; Abbott 2002; PCE, PRU • Oribi Gorge [Oribi Gorge Nature Reserve], [3030CB]; Sep. 1996; Edwards 1438; NU • Umtamvuna Nature Reserve, Smedmore Forest, (3130AA) [3030CC]; 22 Feb. 1986; Abbott 3018; NH, NU, PRU • Umtamvuna Nature Reserve, Smedmore Forest, (3030CC); 27 Jul. 1986; Abbott 3217; PRU • Umtamvuna Nature Reserve, Smedmore Forest, (3030CC); 23 Apr. 1989; Abbott 4366; PRU • Umtamvuna Nature Reserve, Smedmore Forest, (3030CC); 11 May 1992; Abbott 5768; PRU • Umtamvuna Nature Reserve, Smedmore Forest, (3030CC); 9 Jul. 2008; Abbott 8901; PRU • Umtamvuna-natuurrreservaat [Umtamvuna Nature Reserve], Smedmore-woud [Smedmore Forest], (3030CC); 10 Jun. 1982; Abbott s.n.; PRU • Umtamvuna Nature Reserve, Smedmore Forest, (3030CC); 14 Jul. 1983; Jordaan 264; NH • Smedmore [Smedmore Forest, Umtamvuna Nature Reserve], [3030CC]; 10 Jun. 1982; Nicholson 2259; NU, PRE • Umtamvuna Nature Reserve, Smedmore Forest, (3030CC); 4 Sep. 1994; Van Wyk BSA 2576; PRU • Umtamvuna Nature Reserve, Smedmore Forest, (3030CC); 15 Oct. 1984; Van Wyk & Lowrey 6826; PRU • Umtamvuna Nature Reserve, Smedmore [Smedmore Forest], (3030CC); 15 Oct. 1984; Van Wyk & Lowrey 6827; PRU • Umtamvuna Nature Reserve, Smedmore [Smedmore Forest], (3030CC); 15 Oct. 1984; Van Wyk & Lowrey 6831; PRU • Umtamvuna Gorge, Umtamvuna Valley bottom, [3130AA]; 10 Nov. 1994; Meter 100; NU. – Eastern Cape • Mbotyi, Ntsubane Forest, (3129BC); 21 Aug. 1989; Abbott 4416; PRU • Mkambati Nature Reserve, Daza Riverine Forest, (3129BD); 11 Nov. 1984; Jordaan 329; NH • Mkambati Nature Reserve, (3129BD); 11 Dec. 1986; Jordaan 1016; NH • Lwandile north of Umtata River mouth, Nkwintyini Forest, (3129CC/DD) [3129CC]; 7 Apr. 1990; Cooper 253; NH, PRE • Upper Bulawa Forest near Mpande, (3129CD); Cooper 262; NH • Port St Johns, Mount Sullivan, (3129BC) [3129DA]; 2 Oct. 1988; Abbott 4153; PCE, PRU • Mount Sullivan, Nenga Forest, (3129DA); 11 Jan. 1989; Abbott 4206; PCE, PRU • Port St. Johns, Silaka Nature Reserve, (3129DA); 4 Aug. 1988; Van Wyk 8407; PRU • Port St. Johns, Mtambala Forest Station, (3129DA); 5 Aug. 1988; Van Wyk 8429; PRU • The Haven [The Haven Hotel], Elliotdale District, [3228BB]; 13 Feb. 1967; Gordon-Gray 1347; NU • Transkei, Dwesa Nature Reserve, S. of Mendu River Mouth, (3228BD); 1 Aug. 1988; Van Wyk 8334; PRU.
In nature, sterile, fresh material of Marijordaania filiformis can potentially be confused primarily with Maytenus cordata (Fig.
Maytenus cordata. A. Flowering branchlet; note cordate leaf bases and flowers with relatively short pedicels. B. Branchlet with mature, dehisced fruit, showing seed with orange-yellow aril; insert shows a dehisced capsule with seeds completely enclosed by an orange aril. Upper right hand leaf blade broken in half and gently pulled apart to show the presence of silvery elastic threads of gutta-percha. C. Inflorescences with flowers, borne on relatively short, erect or spreading pedicels. Photographs by Richard Boon.
A particularly reliable field character for the identification of some southern African members of the Celastraceae is to test for the presence of elastic (rubbery) threads associated with the phloem tissue of, among others, the leaf venation and bark (
The long, pendant, filiform pedicels of Marijordaania filiformis are usually 35–40 mm long, whereas in the other species they are mostly shorter than 10 mm during flowering. The pedicels may elongate when in fruit, but never match the length of those in M. filiformis. Marijordaania filiformis has pink-red petals with white margins (Fig.
Additional differences include Maytenus abbottii having tetramerous flowers, while the other species have pentamerous flowers. Maytenus acuminata reaches tree proportions and often grows in exposed positions such as forest margins and around rock outcrops. On the other hand, the other species are slender shrubs to small trees, predominantly occupying forest interior habitats, often in deep shade. Maytenus cordata is a more erect plant with glossier leaves, and its new growth is green, in contrast to the bronze to reddish green new growth observed in M. filiformis.
Unlike the situation in nature, dried material of Marijordaania filiformis can easily be confused with that of Maytenus cordata in herbaria, especially those forms of the latter with shortly petiolate leaves and a rounded or cuneate lamina base. Unfortunately, testing for elastic threads in herbarium material is unreliable, and especially in older material the threads may no longer be detectable. As explained earlier,
In herbaria, Marijordaania filiformis can, in addition to the difference in the number of locules, also be distinguished from Maytenus cordata by its flowers being larger, ca 10 mm in diam. [vs smaller, ca 6 mm], cymes 1- or 2(3)-flowered [vs more floriferous, usually 6–8-flowered, these opening successively over time], pedicels (20–)35–40(–45) mm long, pendant [vs 4–10(–20) mm, erect or spreading], and articulated ca 0.5 mm above the base [vs (1.5) 2.0–4.0(–8) mm]. The part of the pedicel below the articulation is persistent, hence providing a useful taxonomic character for distinguishing between the two species, especially in herbarium specimens of which the open flowers have already been shed.
We wish to thank Hester Steyn and Elizma Fouché for the distribution map, Daleen Roodt for the line drawings, Sharon Louw for photographs, and Hugh Glen for discussions on plant nomenclature. Others who helped were: Benny Bytebier, Christina Potgieter, Cornelia Klak, Craig Mulqueeny, David Styles, Dean Phillips, Dorothy McIntyre, Eberhard Fischer, Elize Cloete, Geoff Nichols, Graham Grieve, Ivan Savinov, Kenneth Oberlander, Linda Loffler, Michelle Smith, Ron Uken, Sinegugu Zukulu, Talia Krishanlal, Tony Dold, Waldo Bekker, and Yashica Singh. The curators/staff of the following herbaria are thanked for permission to consult specimens or for providing images of specimens: BOL, NH, NU, PCE, PRE, PRU, SAM, and UDW. Financial and institutional support from the University of KwaZulu-Natal and University of Pretoria is gratefully acknowledged. The study was also funded by a South African National Research Foundation grant (NRF 132105) awarded to AEvW.