Introduction

Lindau (1913) described the monotypic genus Linocalix from Bare in Cameroon as part of his ongoing studies on the African Acanthaceae, placing it near the genus Gatesia A. Gray (= Yeatesia Small fide Scotland & Vollesen 2000). Linocalix has since largely disappeared from the taxonomic literature, not having been mentioned in the relevant publications on West African Acanthaceae (most notably Heine 1963, 1966) nor in checklists of the tropical African flora (Lebrun & Stork 1997; Klopper et al. 2006) or the vascular plants of Cameroon (Onana 2011). In their classification of Acanthaceae, Scotland & Vollesen (2000) listed Linocalix (as Linocalyx) as a synonym of Justicia L. but without further elaboration or justification. With a treatment of the Acanthaceae for the Flore du Cameroun series yet to be produced, the status of Linocalix has remained unconfirmed. Thus Linocalix was one of the last genera of vascular plants to remain unresolved in the World Checklist of Selected Plant Families database (R. Govaerts, pers. data) and on The Plant List (2013). The specimen of the single species, L. albus Lindau (Ledermann 6140) at Berlin is believed to have been destroyed by the bombing of the Berlin herbarium during World War II. Fortunately, a duplicate of this collection is held at the BR herbarium and is available via the JSTOR Global Plants website (under Linocalyx albus; JSTOR 2016). Whilst depauperate, this specimen does include a flowering spike with one young corolla visible, which is a close match for the rare species Chlamydocardia subrhomboidea Lindau. This placement is confirmed by Lindau’s description of L. albus in the protologue, which is a good match for C. subrhomboidea. Linocalix is therefore synonymised within Chlamydocardia here.

Chlamydocardia was placed in subtribe Justiciinae of tribe Ruellieae of subfamily Acanthoideae in the classification of the Acanthaceae by Scotland & Vollesen (2000). More recent molecular evidence places the genus as the basal-most clade of the “Tetramerium lineage” (Daniel et al. 2008), one of the four major lineages in the tribe Justiceae sensu McDade et al. (2000). Chlamydocardia is recognised by having the combination of (1) a small (c. 9 – 12 mm long), corolla with a narrowly cylindrical tube much longer than the reflexed limb (the lower lip fully divided into three lobes, the upper lip comprising two almost completely fused lobes); (2) anthers with unequal-sized thecae but these held at ± the same height (i.e. not offset) and lacking appendages; and (3) broadly prolate to subglobose, 6-zonocolporate pollen with a mesocolpium of numerous verruculae often arranged in ± well defined longitudinal lines (Daniel et al. 2008; Darbyshire 2010; Fig. 1). This pollen type has been proposed as a synapomorphy for the genus (Daniel et al. 2008; see also Scotland & Vollesen 2000: fig. 39 A & B). Lindau’s description of the pollen of Linocalix albus is a close match for this unique pollen type Two species of Chlamydocardia are currently recognised; these are summarised in the taxonomic account below and an identification key is presented.

Fig. 1
figure 1

SEM images of the pollen of Chlamydocardia buettneri. A whole grain, equatorial view (scale bar = 50 μm); B detail of aperture and sculpturing (scale bar = 10 μm). Images from Brenan 9317 ex Cameroon.

Taxonomic account

Chlamydocardia Lindau (1894: 39); Clarke (1900: 234); Heine (1963: 423; 1966: 184); Darbyshire (2010: 648). Type species: Chlamydocardia buettneri Lindau.

Linocalix Lindau (1913: 39), synon. nov. Type species: Linocalix albus Lindau.

Genus nov. sensu Vollesen & Darbyshire (2004: 231).

Bracts variously obovate, linear-oblanceolate or spathulate-obcordiform, widest in the distal half, 0.7 – 7 mm wide.......................................................................C. buettneri

Bracts ovate, elliptic or subrhombic, widest in the proximal half or at the midpoint, 6 – 9.5 mm wide.......................................................C. subrhomboidea

Chlamydocardia buettneri Lindau (1894: 39); Clarke (1900: 234); Heine (1963: 423; 1966: 185, pl. 38: 1 – 9); Vollesen (1998: 2); Vollesen & Darbyshire (2004: 224); Hawthorne & Jongkind (2006: 438, 446, 447); Sosef (2006: 37); Darbyshire (2010: 650, fig. 83); Onana (2011: 30). Types: Gabon, Sibange Farm, fl. & fr. Sept. 1884, Büttner 469 (B† holotype); Gabon, forêt de Mondah, N of Libreville, fl. 13 Aug. 1992, Dibata 1054 (K! neotype, selected here; MO, WAG isoneotypes). Fig. 2.

Fig. 2
figure 2

Chlamydocardia buettneri. A habit; B detail of stem indumentum; C bract, outer surface; D bract, outer surface, narrow extreme; E bract, outer surface, “spathulate-obcordiform” form; F bracteoles and calyx; G corolla with stamens and pistil; H dissected corolla with stamens; J mature capsule with seeds; K mature seed. A – C & H – K from G. Taylor 3312 ex Uganda; D from Eilu 450 ex Uganda; E from Cheek 7162 ex Cameroon; F & G from Brenan 8455 ex Nigeria. Reproduced from F.T.E.A. Acanthaceae Part 2 (Beentje 2010: fig. 83), with permission of the Trustees of the Royal Botanic Gardens, Kew. drawn by juliet williamson.

Chlamydocardia lanciformis Lindau (1911: 300). Types: D.R. Congo, Beni, Muera, fl. & fr. Jan. 1908, Mildbraed 2225 (B† syntype); pl. 32 G – J in Wiss. Ergebn. Deut. Zentr. Afr. Exped., Bot. 2, 1911 (lectotype, selected here).

distribution. Widespread in the wet Guineo-Congolian forests of West and Central Africa: Ivory Coast, Nigeria, Cameroon, Equatorial Guinea, Gabon, Congo (-Brazzaville), D.R. Congo, Uganda.

conservation status. This widespread species was assessed as of Least Concern (LC) by Darbyshire (2010).

notes. Chlamydocardia buettneri is variable in terms of bract shape, ranging from broadly obovate (e.g. Casas 12100 from Bioko), to spathulate-obcordiform, i.e. with a narrowly cuneate or attenuate base and with two expanded rounded lobes towards the apex, usually also with a short acumen (e.g. Cable 841 from Cameroon, Keay in FHI 28278 from Nigeria; Fig. 2E), or linear-oblanceolate (e.g. Eilu 450 from Uganda, Fig. 2D). The lattermost form was previously separated as C. lanciformis but there are plants intermediate between this and the “spathulate-obcordiform” bract form across its range, notably in Uganda. What unites all these bract forms is that they are widest in the distal half, with a long-attenuate or -cuneate base and (except for linear-oblanceolate forms) with a short but marked acumen. Variants on the “spathulate-obcordiform” bract shape are the most common and widespread form – these bracts are quite unlike anything seen elsewhere in African Acanthaceae.

The original specimen of Chlamydocardia buettneri is believed to have been destroyed by the bombing of the Berlin herbarium during World War II and no duplicates have so far been traced. Dibata 1054 is therefore chosen as a neotype as it is an informative specimen and the collecting locality is not too distant from the original type locality. Likewise, the original specimen of Chlamydocardia lanciformis is believed to have been lost in the same way; the illustration in the protologue is therefore selected as a lectotype here.

Chlamydocardia subrhomboidea Lindau (1895: 119); Clarke (1900: 235); Heine (1966: 185, in notes); Onana (2011: 30). Types: Cameroon, between Victoria [Limbe] and Bimbia, fl. 10 July 1894, Preuss 1309 (B† holotype); Cameroon, Moliki, fl. & fr. Nov. 1938, Jacques-Félix 2534 (P 02894780! neotype, selected here; K!, P 02894779!, P 00536453! isoneotypes). Fig. 3.

Fig. 3
figure 3

Chlamydocardia subrhomboidea, collected as Osborne 299 ex Cameroon. photo: j. osborne.

Linocalix albus Lindau (1913: 408). Type: Cameroon, Bare, fl. 15 Nov. 1909, Ledermann 6140 (B† isotype; BR 629 315! lectotype, chosen here), synon. nov.

Chlamydocardia buettneri sensu Heine (1963: 423) pro parte, non Lindau.

Genus nov. sensu Vollesen & Darbyshire (2004: 231).

distribution. Cameroon, D.R. Congo.

conservation status. Assessed as Endangered EN B2ab(iii) by Darbyshire & Cheek (in Onana & Cheek 2011: 70) as it is known from only five locations and is threatened by conversion of its forest habitat to agricultural land in at least two of its known sites.

notes. Chlamydocardia subrhomboidea is usually easily separated from C. buettneri by its larger, ovate or at most elliptic bracts that lack an acumen, the apex being acute or at most shortly attenuate. These large bracts result in a more markedly strobilate flowering spike than in C. buettneri (see Fig. 3). In addition, C. subrhomboidea is typically a more robust plant than C. buettneri with larger leaves, but there is some overlap in these vegetative characters. A small number of specimens of C. buettneri from Cameroon (notably Brenan 9317 and Cable 1216, both K!) with particularly broad obovate bracts begin to approach those of C. subrhomboidea. It is therefore conceivable that C. subrhomboidea could be an extreme form of C. buettneri but this would be best investigated using molecular methods and it would be premature to reduce C. subrhomboidea to synonymy without such a study. Interestingly, Heine (1963) initially synonymised C. subrhomboidea within C. buettneri but later resurrected it as a clearly distinct species, after having studied specimens from Cameroon and D.R. Congo (Heine 1966).

The holotype of Chlamydocardia subrhomboidea is believed to have been destroyed by the bombing of the Berlin herbarium during World War II and no duplicates have so far been traced. A neotype is therefore selected here; one of the three P sheets of Jacques-Félix 2534 is chosen as this is an informative specimen, with both flowers and mature fruits, and the collection site is not too distant from the type locality; this specimen is also duplicated at K.

The identity of Chlamydocardia nuda

Chlamydocardia nuda C. B. Clarke (1900: 235). Type: Gabon, Como R., 60 miles from Gaboon, fl. July 1896, Bates 494 (K! holotype; BM!, G-DC!, L!, P! isotypes).

notes. This rare species, endemic to Gabon, was originally placed with some doubt in Chlamydocardia by Clarke (1900), but rejected from that genus by Lindau (1911). Heine (1966) subsequently transferred this species to Justicia L. with a new name J. tigrina Heine as the epithet nuda was already occupied in Justicia by J. nuda (Nees) O. Schwartz (Schwartz 1939: 259). Heine’s J. tigrina remains the accepted name in current works (see e.g. Sosef 2006). However, this placement also appears to be in doubt and the status of this species is currently under investigation by the first author [I.D.] together with colleagues at Rancho Santa Ana Botanic Garden and the California Academy of Sciences. In any case, it certainly does not belong in Chlamydocardia, having a very different floral morphology to that genus: the corolla is markedly 2-lipped with only a shortly divided lower lip, the limb is approximately equal in length to the tube and the anther thecae are puberulous and markedly offset, each with a flattened basal “proto-appendage” with small conical protrusions (see Heine 1966: pl. 38, 10 – 13).