5. Saxicolella marginalis (G.Taylor) Cheek

(Cheek et al. 2000: 152; Onana & Cheek 2011: 252; Onana 2011: 116; 2012: 137).

Type: Nigeria, “ Ogoja Province, River Butum, Utanga, about 2 km north of Bagga, on smooth granite rocks, just below, at, and just above water-level in fast-flowing stream.”, fl. 25 Dec. 1938, Keay, Savory & Russell in FHI 25152 (holotype BM 000910316!).

Butumia marginalis G. Taylor (1952: 55; 1953: 123 – 125).

Annual herb. Root rosette-like, 5 – 12 cm diam., radiating from a central point, the distal ends ribbon-like, 1.5 – 5 (– 7.5) mm wide, rarely bifurcating. Stems of sessile leaf rosettes, numerous, spaced along the margins of the roots, (0.9 –) 1.2 – 1.5 (– 1.7) mm apart, 1.2 – 22 mm long (including the single terminal flowers), phyllotaxy spiral. Leaves (3 –) 7 per stem, dimorphic, outermost (proximal), subulate, 0.25 mm long, lacking stipules; innermost (distal) 0.6 – 1 (– 2) mm long, stipulate, stipules paired, marginal, triangular, 0.2 mm from base. Spathellum bud (pre-dehiscence) ellipsoid, 1.1 – 1.25 mm long, 0.6 mm wide, apex mucronate, dehiscing irregularly into two halves. Flower erect, included within spathellum at anthesis. Pedicel 0.9 – 1 mm long at anthesis. Tepals subulate, 0.3 – 0.5 mm long. Stamen exceeding gynoecium, 1.7 mm long, filament dorsiventrally flattened, 1.2 mm long, anther oblong, 0.5 mm long, cells superposed. Gynophore 0.1 – 0.2 mm long. Ovary olive green, ellipsoid, 0.9 – 1 mm long, 0.6 – 0.75 mm wide, in transverse section orbicular, unilocular, longitudinal ribs 8, including two commissural ribs. Stigmas purple, 2, elliptic, complanate, 0.25 mm long, 0.15 mm wide. Fruit ellipsoid, 1 mm long, dehiscing into two persistent valves. Seeds ellipsoid, 0.2 mm long. (Fig. 5).

DISTRIBUTION. SE Nigeria and NW Region Cameroon, one site in each.

SPECIMENS EXAMINED. CAMEROON. North West Region, Fundong, Touristic Hotel, 20 m fall on Chumni R., 1300 m, fl. fr. 22 Nov. 1996, Cheek 8730 (K!, SCA!, YA!)

HABITAT. Waterfalls and rapids in forested or formerly forested mountainous areas, on rounded boulders of granite or basalt, in Nigeria with Ledermanniella aloides

(Taylor 1953: 127 re Keay in FHI 25152); 300 – 1300 m alt.

CONSERVATION STATUS. Ouedraogo (2010) listed this species as Critically Endangered B1ab(iii)+2ab(iii), stating that it had only been recorded from Nigeria and Cameroon and that there is a continuing decline of its habitat quality due to water pollution and its populations are severely fragmented. He cited an earlier assessment of CR in 2000. This was probably that of Cheek in Cheek et al. (2000: 69) where the species was assessed as CR B1+2c due to the threat from laundry operations at the top of the fall housing this species in the town of Fundong, Cameroon. Nevertheless, following the discovery of the species at the site in 1996, monitoring by the first author in December 1998 and November 1999 showed that approximately the same number of plants, several hundred, were present (Cheek et al. 2000: 69). This publication also formally transferred the species from Butumia to Sapicolella. Ouedrago (2010) further states that “the species may also be present in Ghana and Niger but this needs to be confirmed”. No evidence has been found to support this belief and it seems unlikely. Several other extremely rare, range-restricted and threatened species also occur at high altitude at Mt Oku (Cheek et al. 1997, 2000) e.g. Scleria cheekii Bauters (Bauters et al. 2018), Deinbollia onanae Cheek (Cheek et al. 2021b).

Sapicolella marginalis is genuinely rare in Cameroon, since the first author has targeted searches for it at other falls and rapids in the Mt Oku area and not found it, although two other species of Podostemaceae have been found (Cheek et al. 2000). Nearcomprehensive botanical surveys in other locations S, Wand Eof Kilum-Ijim have failed to find additional locations although they brought to light several other species of Podostemaceae (see references above under Sapicolella ijim). The status of the subpopulation at the type locality in Nigeria is unknown. The Red Data Book of Cameroon Plants (Onana & Cheek 2011: 252) reassessed the species as EN B2ab(iii) since two locations (above) are recorded. Kuetegue et al. (2019) also assess the species as EN B2ab(iii) citing no new data.

PHENOLOGY. Flowering and fruiting in late November to late December.

ETYMOLOGY. The specific epithet derives from the rows of sessile shoots that line the margins of the roots.

VERNACULAR NAMES. None known.

NOTES. In describing this species as a new genus, Butumia, Taylor (1952: 57) stated: “Amongst African genera the plant is most closely related to Sapicolella and Pohliella, in each of which the flowers is unistaminate and erect within the spathella, but it differs from these genera in having entire rosulate leaves, much more shortly pedicellate flowers and complanate stigmas”. At that time both genera were only known from their type species and their circumscription was incompletely known.

It is remarkable that the only known Cameroonian site for this species is only a few kilometres distant and at similar altitude (c. 1300 m), to the only known site for Sapicolella ijim (see under that species). However, this submontane altitudinal band of the Cameroon Highlands is immensely rich in endemic rangerestricted species. Other examples include Coffea montekupensis Stoff. (Stoffelen et al. 1997) and Impatiens etindensis Cheek & Eb.Fischer (Cheek & Fischer 1999).

The Cameroonian location is about 100 km SSE of the type and only other known location in Nigeria. The species most closely similar in morphology to Sapicolella marginalis is S. deniseae of the Guinea Highlands in Guinea, far to the west. Both species share the unusual character of sessile, rosette-like shoots arranged along the margins of the radiating ribbon-like roots. Several species of the Cameroon Highlands do occur disjunctly in the highlands of Guinea e.g Dorstenia astyanactis Aké Assi (Couch et al. 2019), Brachystephanus oreacanthus Champl. (Champluvier & Darbyshire 2009) and Isoglossa dispersa I.Darbysh. & L.J.Pearce (Darbyshire et al. 2011), so long-distance dispersal, perhaps by birds, is credible as an explanation.