Crepidorhopalon flavus (S.Moore) I.Darbysh. & Eb.Fisch., 2019

Crepidorhopalon flavus (S.Moore) I.Darbysh. & Eb.Fisch. View in CoL , comb. nov. — Fig. 1f–h View Fig , 2f View Fig

Lindernia flava S.Moore View in CoL in J. Linn. Soc., Bot. 40: 153 (1911). — Type: Swynnerton 1966 (lecto BM [ BM000930712 ], chosen here; isolecto K [ K000379639 ]), Zimbabwe, near Chirinda , fl. 14 June 1906 .

Lindernia whytei View in CoL sensu auct.: Philcox (1990) 62, p.p.; Philcox & Ghazanfar (2008) 70, p.p.

Perennial herb, trailing or scrambling, laxly branched; stems brittle, markedly 4-angular, winged along the angles, glabrous or with few short eglandular hairs above each node, distal internodes sometimes with few sessile glands; internodes of leafy stems (9–) 28–70 mm long when mature. Leaves sessile, broadly ovate, 9.5–22 by 6–20 mm, base rounded or shallowly cordate, margin serrate or shallowly so with 1–4 teeth per margin, apex acute or obtuse; primary venation palmate, usually with 5 main veins from base; glabrous except for few eglandular hairs on margin towards leaf base, surfaces also with sunken glands drying brown, most visible on lower surface. Flowers axillary, solitary, one flower per node, usually together forming a lax raceme; bracts either foliaceous or reducing in size and relative width distally, distal pairs of bracts often lanceolate-acuminate or ovate-acuminate, 2–6 by 0.7–4.5 mm, the pairs often markedly uneven in size, the bract subtending the flower smaller and narrower; bracteoles absent; pedicels 2–7.5(–10.5) mm long, glabrous, with sparse sessile glands. Calyx 5–9.5 mm long, lobes lanceolate, somewhat unequal in length, longest lobes 3.2–6.5 mm long, becoming divergent at maturity and with apices somewhat recurved, 5-veined, tubular portion hyaline between the veins, margins of lobes eglandular-pubescent towards base, elsewhere glabrous, with few sessile glands. Corolla 10–14 mm long, tube and upper lip yellow-brown or upper lip slightly purple-tinged, lower lip bright yellow, external surface glandular-puberulous, upper lip with pale longer eglandular hairs; tube 5.7–8.5 mm long, cylindrical towards base where 0.9–1.2 mm wide, widened towards mouth where 2–2.5 mm wide, internal surface with 4 lines of subsessile glands in proximal half; upper lip ovate, 2.3–3.5 by 2.3–3 mm, apex emarginate, internal surface with numerous glands; lower lip 4.3–6 mm long, 3-lobed, median lobe rounded-obovate, 2.5–3.5 mm long and wide, lateral lobes somewhat smaller, bosses on palate of lower lip with blunt-tipped multicellular hairs. Stamens 4, ventral stamens with clavate spurs, 0.5–0.8 mm long, papillate, filaments above spurred portion 2.7–3.5 mm long; posterior stamens with filaments 1.2–1.4 mm long; anthers of the two pairs of stamens adhering, thecae 0.6–0.9 mm long. Ovary 1.1–1.4 mm long, with sessile glands on surface; style 4– 6 mm long; stigma fan-shaped and minutely fringed. Capsule narrowly ellipsoid or more gradually tapered, brown, ± equal in length to calyx, 5.5–9.5 mm long; seeds ellipsoid, ± 0.5–0.6 by 0.45–0.5 mm, longitudinal ridges very shallow and inconspicuous.

Distribution & Ecology — This species is recorded in swamps, riverbanks and seasonally wet open grassland with sandrich peaty soils, including swampy clearances in forest, at 330–1130(–1340) m elevation.

Additional specimens seen. MOZAMBIQUE, Manica Prov., Mt Maruma , fl. & fr. 13 Sept. 1906 [ syntypes], Swynnerton 1922 ( BM, K) ; Dombe, entre Chis- santo e Zinesse , fl. & fr. 24 Nov. 1965, Pereira & Marques 876 ( BR, WAG *) ; E of Makurupini R., fl. 11 June 1971, Biegel 3583 ( WAG *) ; Sussendenga Dist., eastern foothills of Chimanimani Mts , c. 28 km W of Dombe, fl. & fr. 25 Apr. 1974, Pope & Müller 1295 ( K) ; Chimanimani Mts foothills, near Zomba Community, Magorogodo hill, fl. 27 Oct. 2013, Würsten & Dondeyne BE 875 ( BR) ; Chimanimani foothills, Maronga, Wandowani , W of Comeni’s compound, fl. 14 Nov. 2015, Darbyshire 895 ( K, LMA, Ndzou) ; Maronga, between Comeni’s compound and Murere River ,fl. 15 Nov. 2015, Darbyshire sight record. – ZIMBABWE, Manicaland Prov., Melsetter [Chimanimani], fl. 7 fr. 1 Oct. 1919, Walters 2728 ( K) ; ibid., Walters 2758 ( K) ; Umtali [Mutare], fr. Oct. 1937, Brain s.n. in SRGH 10828 About SRGH ( SRGH *) - see note; Tarka Forest Reserve, banks of Chisengu R., fl. & fr. May 1968, Goldsmith 72/68 ( K, P *) .

Conservation — Crepidorhopalon flavus is a highly range-restricted species, known only from the southern Manica Highlands of the Mozambique-Zimbabwe border region, where it is centred on the foothills of the Chimanimani Mountains. The EOO is 4 440 km 2, or only 646 km 2 if the Walters and Brain specimens are excluded in view of the fact that it is unclear as to whether the former was collected near Chimanimani town or in the Chimanimani Mountains and there is some uncertainty over the latter collection (see note). Six to eight locations are defined based on threats. Those sub-populations that lie within the formally protected areas of the Chimanimani Mts – the Chimanimani National Park in Zimbabwe and the Chimanimani National Reserve in Mozambique – are likely to be secure. The only activity recorded recently in the seasonal wetlands within the National Reserve was cutting of grass for thatching (I. Darbyshire, pers. obs.) and this is unlikely to impact the Crepidorhopalon significantly. However, the majority of locations lie outside of these protected areas and there has been much disturbance and destruction of natural habitats within its range due to high human population pressure and continued expansion of subsistence agriculture. This includes areas of seasonally wet soils which, away from areas of quartzite, are fertile and so favoured for cultivation (B. Wursten, pers. obs.). Some sub-populations may also have been impacted by artisa- nal gold mining activities along some of the rivers and streams in the southern foothills of the Chimanimani Mountains. With a continuing decline in extent and quality of habitat inferred, and with fewer than 10 locations known within an EOO of considerably less than 20 000 km 2, this species is assessed as Vulnerable under criterion B1 – VU B1ab(iii).

Notes — In view of the consistently different flower colour, coupled with the marked range (and hence genetic) disjunction and differing habitat requirements of the Zimbabwe / Mozambique plants, it is considered most appropriate to reinstate Lindernia flava as a good species and the new combination in Crepidorhopalon is made here. This species is otherwise very similar to C. whytei although the seeds differ slightly in having almost imperceptible longitudinal furrows, these being clearly visible in C. whytei (see Fig. 2 View Fig ).

In the specimen citations under Lindernia whytei, Philcox (1990) wrongly attributes Swynnerton 1922 as having been collect- ed from ‘serra do Gúruè, E of Picos Namuli, near source of R. Malema, c. 1800 m’; it was actually collected from Mt Maruma in Manica Province. The specimen Brain s.n. at SRGH does not have any flowers and there are no notes on flower colour; it is recorded from further north than the other collections and at a higher altitude (the maximum elevation recorded above). It is strange that this species has not been re-recorded from the well-botanised area around Mutare but there is no reason to doubt the provenance of this specimen. Further material from the Mutare area would be useful to confirm the identity of this specimen.

When describing Lindernia flava, Moore (1911) noted the close similarity to his L. gossweileri S. Moore (1907: 87) from near Capopa in Malange Province, Angola (Map 1), but said that they differ in L. gossweileri having calyces divided to the base and having a larger corolla with a longer tube and with the lips almost equal in size rather than markedly unequal. To our knowledge, L. gossweileri has never been recollected in Angola and the type specimen ( Gossweiler 1086, BM [BM000930701], K [K000379630]) is rather scanty, such that it is difficult to draw firm conclusions on its relationship to C. flavus . However, the differences noted by Moore seem to be rather minor and that of differing flower size and proportions may not be of great significance in view of the differences in flower size recorded in other species of Crepidorhopalon including C. whytei . Therefore, it is possible that these two yellow-flowered taxa represent a single species, although the extreme range disjunction seems improbable. More material from Angola is needed for confirmation, but if they do prove to be the same species then L. gossweileri would have nomenclatural priority and the epithet is unoccupied in Crepidorhopalon .

The wet grasslands in the foothills of the Chimanimani Mountains, the stronghold for this species, are of wider botanical interest as they hold several interesting species besides C. flavus . Of particular note is Mesanthemum africanum Moldenke which is endemic to the Chimanimani Mountains, occurring in damp sites on quartzitic sands both at higher altitudes and (at much lower abundance) in the foothills, probably a result of small populations being established from seed washed down along rivers from the high massif. During surveys of the Maronga area of Mozambique in 2015, C. flavus was found growing alongside two species of Xyris ( Xyridaceae) . The first of these proved to be X. angularis N.E.Br. ( Darbyshire 918), a widespread West African species that had not previously been recorded from Mozambique but is known from the Zimbabwe side of Chimanimani ( Lock 2010); the Chimanimani population is a significant outlier for this species. The second Xyris species collected ( Darbyshire 938) is an unmatched, potentially new species. The first Mozambican records of Edrastima angolensis (K.Schum.) Neupane & N.Wilkstr. ( Darbyshire 920) in the Rubiaceae and Fimbristylis aphylla Steud. in the Cyperaceae ( Darbyshire 926) were also made at the same wetland site ( Timberlake et al. 2016).