Re-evaluation of the genus Englerodendron (Leguminosae–Detarioideae), including Isomacrolobium and Pseudomacrolobium

Additional keywords: Africa, Anthonotha, classification, generic delimitation, Macrolobium, monophyly.

The tribe Amherstieae Benth. is the most diverse group within subfamily Detarioideae of the Leguminosae, with 50 genera and ~570 species being currently recognised (Legume Phylogeny Working Group 2013, 2017; de la Estrella et al. 2018). Within this tribe, floral morphology is extremely variable and, in many taxa, highly modified, making classification of the group very challenging (Mackinder 2005; de la Estrella et al. 2018). Flowers vary from bilaterally to radially symmetrical, with sepals sometimes entirely lacking but most commonly having 4–5 (rarely up to 10) sepals. Petal number is also extremely variable. Five petals are found in many species and this seems to be the ancestral state (Ojeda et al. 2019); however, several reduction series exist, often resulting in one large adaxial petal with the other four being reduced or absent. In a few genera, all petals have disappeared. In Englerodendron usambarense Harms, an additional one or two petals are present, resulting in six- or seven-merous flowers (Breteler 2008). Stamen numbers in tribe Amherstieae follow trends similar to those seen in the petal number. The ancestral value is likely to be 10, which is reduced in several lineages, often resulting in three fertile stamens and several staminodes, but these staminodes may disappear in a later step (Ojeda et al. 2019). In some cases, the stamen number is increased, being, for example, ~20 in some species of Hymenostegia Harms, and up to 80 in Maniltoa Scheff. Reduction in stamen number, including transformation to staminodes as an intermediate step towards complete loss, appears to be a general trend observed in several groups (Bruneau et al. 2014), potentially being correlated with flower size. Indeed, members of the genus with the largest flowers in the Amherstieae (Berlinia Sol. ex Hook.f.) have 10 stamens and no or partial reduction in the numbers of sepals and petals. However, not all lineages with small flowers have a reduced stamen number. Some genera with small flowers, such as Michelsonia Hauman, Bikinia Wieringa and Zenkerella Taub., have 10 stamens (Cowan and Polhill 1981; Wieringa 1999). One consequence of this variability is that the number of stamens and staminodes appears to be highly homoplasious and does not provide a good basis for classification. Some of the most extreme variation in floral traits within Amherstieae is found within the Berlinia clade (sensu Bruneau et al. 2008), where great plasticity in the number of petals is observed within the same species and among flowers of the same individual (Breteler 2006, 2008, 2010, 2011; Ojeda et al. 2019). For example, Breteler (2008) observed that the corolla in Isomacrolobium Aubrév. & Pellegr. is extremely labile, which hampered any attempt to structure the observed variability into distinct taxonomic entities. The only distinction Breteler (l.c.) could make was between an actinomorphic and a zygomorphic corolla.

The Berlinia clade was first described by Bruneau et al. (2000, 2001) as the ‘Macrolobieae’ clade, and further delimited to include 18 exclusively African genera (Bruneau et al. 2008). The ancestral biome of this clade was reconstructed as primary evergreen rainforest in tropical Africa (de la Estrella et al. 2017), where Detarioideae, and especially members of the Berlinia clade, form the most dominant group of trees in western central Africa (Wieringa 1999; Newbery et al. 2013).

Within the Berlinia clade, several genera have been the focus of recent taxonomic treatments, increasing our knowledge of their diversity and morphological complexity (Wieringa 1999; Mackinder and Pennington 2011; de la Estrella et al. 2012a, 2012b; de la Estrella and Devesa 2014; van der Burgt et al. 2015). Anthonotha P.Beauv., Isomacrolobium and Englerodendron Harms, which show particularly complex floral ontogeny, were recently revised by Breteler (2006, 2008, 2010, 2011).

Anthonotha was first described by Palisot de Beauvois (1806) to accommodate a new species from West Africa, but subsequently described African taxa were initially placed within a pantropical Macrolobium Schreb. (see de la Estrella and Devesa (2014), for a history review of the group). Although it is the most speciose genus within Detarioideae and was previously posited as being non-monophyletic (Mackinder 2005), Macrolobium, as currently circumscribed, is an American genus recently demonstrated to be monophyletic (Murphy et al. 2018). The African taxa previously treated under Macrolobium had already been accommodated by Léonard (1952, 1954, 1955) within a reinstated Anthonotha and the newly published Gilbertiodendron J.Léonard, Paramacrolobium J.Léonard and Pellegriniodendron J.Léonard. Isomacrolobium and Triplisomeris Aubrév. & Pellegr. were subsequently split off from Anthonotha (Aubréville and Pellegrin 1958 and Aubréville 1959 respectively) and Aubréville (1968) added a further segregate genus, namely, Leonardendron Aubrév. Breteler (2006, 2008, 2010, 2011), accepted Anthonotha sensu Aubréville & Pellegrin (Aubréville 1959) but rearranged the other species into Isomacrolobium (zygomorphic flowers) and Englerodendron (non-zygomorphic flowers). Gilbertiodendron and related genera have been the subject of further taxonomic and phylogenetic studies (de la Estrella et al. 2012a, 2014), but relationships among other Berlinia clade genera, especially those of Anthonotha and related genera, have remained obscure (de la Estrella et al. 2017, 2018).

With the aim to resolve relationships within Detarioideae, Ojeda et al. (2019) developed a target capture-sequence bait set for the subfamily, by selecting orthologues shared among four representative transcriptomes from the group. Target capture sequence (Hybseq) has proven to be an effective strategy for phylogenetic reconstruction across different taxonomic levels (Vatanparast et al. 2018; Villaverde et al. 2018). With this Detarioideae bait set (ver. 1.0, MYbaits, Arbor Biosciences, Ann Arbor, MI, USA), Ojeda et al. (2019) reconstructed a phylogeny for Anthonotha and closely related genera, which included near-complete species sampling of Anthonotha, Isomacrolobium (except I. sargosii (Pellegr.) Aubrév. & Pellegr. and I. hallei Aubrév.) and Englerodendron, the three genera that had been previously recognised within the group (Bruneau et al. 2000, 2001, 2008; de la Estrella et al. 2018). With near-complete genus-level sampling across the subfamily, the previously unplaced Pseudomacrolobium Hauman, a monospecific African endemic genus from Congo (Kinshasa), was shown to be closely related to Englerodendron and Isomacrolobium (Ojeda et al. 2019). The Ojeda et al. (2019) phylogeny (Fig. 1) shows more robust support across the clades than does any previous study. Two main clades are resolved, one including all Anthonotha species and the other containing the three remaining genera of Englerodendron, Isomacrolobium and Pseudomacrolobium. However, neither Englerodendron nor Isomacrolobium is monophyletic in this phylogeny (Fig. 1), leaving several options for generic re-delimitation.

Fig. 1.  Phylogeny of the Berlinia clade, showing the monophyly of Anthonotha sensu stricto and of Englerodendron sensu lato. This tree was obtained from Ojeda et al. (2019), and is based on a concatenated analyses of 922 exonic regions (clusters) and a matrix size of 239 334 base pairs (74.06% taxon occupancy) using maximum likelihood (RAxML, ver. 8.2.9, https://github.com/stamatak/standard-RAxML, the GTRCAT model, 1000 bootstrap replicates and default settings, Stamatakis 2014). Full analytical details are available in Ojeda et al. (2019). Values above branches are bootstrap values.

The Englerodendron clade as shown in Fig. 1 could be split into different recircumscribed genera (Englerodendron, Isomacrolobium, Pseudomacrolobium); however, this would require reinstatement of genera currently placed in synonymy under Isomacrolobium, and the resulting genera would not be diagnosable morphologically. Breteler (2006, 2008, 2011) previously recognised that the members of this group exhibit extensive floral variation, and the phylogenetic evidence (Fig. 1) supports the option, chosen here, of combining these three genera within an enlarged Englerodendron (Ojeda et al. 2019).

Most of the morphological differences given by Breteler (2008) to differentiate Anthonotha from Isomacrolobium apply equally to the newly circumscribed Englerodendron, even though, in that paper, Breteler did not consider Englerodendron because he assumed it to be more closely related to Oddoniodendron De Wild. and Isoberlinia Craib & Stapf ex Holland (Breteler 2006). However, these characters apply to his Englerodendron species as well, just as they do for Pseudomacrolobium, which, because of its many stamens (10–13), has never previously been considered as being related to this clade. Anthonotha is characterised by the presence of a single large petal in the adaxial position, which has a well-developed, inrolled, gutter-shaped claw that is at least half as long as the lamina and the lamina is always bilobed or divided into two. The other four petals may be reduced in various stages, or absent. Englerodendron in this new circumscription has (1–)2–5 large petals, whereas the other petals may be reduced in various stages or be absent. Where only a single petal remains (possible in E. vignei (Hoyle) Estrella & Ojeda), it is not located in the adaxial position. Petals may be clawed, but the claw is often short and not inrolled into a solid support for the laminae, and the lamina of the adaxial petal may be lobed or split, although not as strongly as it is in Anthonotha. In addition to these floral differences, Anthonotha can be distinguished by its dense, usually persistent (except in A. trunciflora (Harms) J.Léonard), appressed indumentum on the lower leaflet surfaces, whereas the leaflets of Englerodendron are sparsely hairy to glabrous, the indumentum not obscuring the lower leaflet surfaces.

Type species: Englerodendron usambarense Harms.

Pseudomacrolobium Hauman, Bull. Séances Inst. Roy. Colon. Belge 23: 477 (1952), syn. nov.

Type species: Pseudomacrolobium mengei (De Wild.) Hauman.

Isomacrolobium Aubrév. & Pellegr., Bull. Soc. Bot. France 104: 497 (1958), syn. nov.

Type species: Isomacrolobium leptorrhachis (Harms) Aubrév. & Pellegr.

Triplisomeris Aubrév. & Pellegr., Bull. Soc. Bot. France 104: 497 (1958), syn. nov. [previously synonymised under Isomacrolobium Aubrév. & Pellegr. (Breteler 2008)].

Type species: Triplisomeris explicans (Baill.) Aubrév. & Pellegr.

Leonardendron Aubrév., Adansonia sér. 2, 8: 167 (1968).

Type species: Leonardendron gabunense (J.Léonard) Aubrév.

Trees or shrubs, up to 36 m tall and 85-cm diameter at breast height. Stipules free or united at base. Leaves paripinnate, 1–7-jugate, leaflets opposite to subopposite, glabrous to sparsely hairy below. Inflorescence of most species a compound raceme up to 1.25 m long, but, in some species, shorter and not pendulous. Bracteoles covering the flower in bud. Flowers 5-merous (6- or 7-merous in E. usambarense). Sepals 5(–7), or 4 when the two adaxial ones are united. Petals (1–)2–5(–7), base often narrowed into a claw, but claw not stiff or inrolled. Fertile stamens 3–7 or 10(–13), supplemented by up to 7 staminodes. Ovary 2–8 ovulate. Pods leathery to woody, (explosively) dehiscent or only dehiscent once on the forest floor, laterally flattened, in general oblong, depauperate pods also (ob)ovate.

Seventeen species, distributed in rain forest of tropical Africa from Guinea to Congo (Kinshasa) and one disjunct species in the Usambara Mountains of Tanzania (Englerodendron usambarense) (Robyns 1952; Breteler 2006, 2008, 2011; van der Burgt et al. 2007).

Isomacrolobium brachyrhachis Breteler, Pl. Ecol. Evol. 144 (1): 66 (2011).

Type: Gabon, 50 km south-east of Achouka, A.M.Louis, Breteler & de Bruijn 729 (holo: WAG [3 sheets]!; iso: BR!, K!, LBV!, MA!, MO!, P!, PRE!).

Macrolobium conchyliophorum Pellegr., Bull. Soc. Bot. France 88: 503 (1941); Anthonotha conchyliophora (Pellegr.) J.Léonard, Mém. Cl. Sci. Acad. Roy. Sci. Belgique (8vo) sér. 2, 30(2): 223 (1957); Isomacrolobium conchyliophorum (Pellegr.) Aubrév. & Pellegr., Bull. Soc. Bot. France 104: 498 (1958).

Type: Gabon, near Lastoursville, Micouma, 23 Nov. 1929, M.G. Le Testu 7680 (lecto: P[2 sheets]!; isolecto: BM, BR [2 sheets]!, WAG [3 sheets]!), fide F.Pellegrin, Mém. Inst. Étud. Centrafr. 1: 45 (1949) and, subsequently, J.Léonard, Mém. Cl. Sci. Acad. Roy. Sci. Belgique (8vo) sér. 2, 30(2): 224 (1957)).

Vouapa explicans Baill., Adansonia 6: 181, adnot. (oct. 1865); Macrolobium explicans (Baill.) Keay, Kew Bull. 8(4): 490 (1953); Anthonotha explicans (Baill.) J.Léonard, Bull. Jard. Bot. État. Bruxelles 25: 202 (1955); Triplisomeris explicans (Baill.) Aubrév. & Pellegr., Bull. Soc. Bot. France 104: 497 (1958); Isomacrolobium explicans (Baill.) Breteler, Syst. & Geogr. Pl. 78(2): 143 (2008).

Type: Guinea, Fouta Djallon, 1837, Heudelot 738 (holo: P [3 sheets]!; iso: BM!, BR!, HBG, K [2 sheets]!, OXF!, WAG [2 sheets]!).

In the first part of his paper (Adansonia 5: 361), Baillon stated that he was studying the herbarium of the ‘Musée des Colonies Françaises’; this herbarium is now part of P.

Macrolobium heudelotii Benth., Trans. Linn. Soc. 25: 308. (Nov. 1865) nom. superfl.

Type: Guinea, Fouta Djallon, 1837, Heudelot 738 (syn-: BM!, BR!, K [2 sheets]!, P [3 sheets]!, OXF!, WAG [2 sheets]!).

Bentham referred to the sheet in the Hooker herbarium (then still private, now part of K), but another sheet was part of his own herbarium that was incorporated in K before publication. Both got annotated with observations and were clearly used for the description; hence, all Heudelot 738 sheets should be regarded as syntypes.

Anthonotha gabunensis J.Léonard, Mém. Cl. Sci. Acad. Roy. Sci. Belgique (8vo) sér. 2, 30(2): 224 (1957); Isomacrolobium gabunense (J.Léonard) Aubrév. & Pellegr., Bull. Soc. Bot. France 104: 498 (1958); Leonardendron gabunense (J.Léonard) Aubrév., Adansonia sér. 2, 8: 167 (1968).

Type: Gabon, Roungassa, 21 Dec.1929, G.M.P.C. Le Testu 7808 (holo: BR!; iso: BM, BR!, K, LISC, P!, WAG!). The isotype at BR arrived in BR only after publication of the basionym; hence, the other BR sheet is the holotype as indicated in the protologue.

Macrolobium graciliflorum Harms, Bot. Jahrb. Syst. 45(2): 302 (1910); Anthonotha graciliflora (Harms) J.Léonard, Bull. Jard. Bot. État. Bruxelles 25: 202 (1955); Isomacrolobium graciliflorum (Harms) Aubrév. & Pellegr., Bull. Soc. Bot. France 104: 498. 1958.

Type: Equatorial Guinea, Nkolentangan, Dec. 1907, Tessmann B.57 (lecto: K, isolecto: BR (fragment from K), B (destroyed) fide F.J.Breteler, Pl. Ecol. Evol. 144 (1): 71 (2011)).

Isomacrolobium hallei Aubrév., Fl. Gabon 15: 362 (1968); Anthonotha hallei (Aubrév.) J.Léonard in L.J.G. van der Maesen et al. (eds), Biodiversity Afr. Pl. (Proc. XIV AETFAT Congr.) 446 (1996).

Type: Gabon, Abanga, chantier C.E.F.A., bord riv. Lano, Jun. 1963, N.Hallé 2195 (holo: P!).

Macrolobium isopetalum Harms, Bot. Jahrb. Syst. 40(1): 25–26 (1907); Anthonotha isopetala (Harms) J.Léonard, Bull. Jard. Bot. État. Bruxelles 25: 202 (1955); Isomacrolobium isopetalum (Harms) Aubrév. & Pellegr., Bull. Soc. Bot. France 104: 498 (1958).

Type: Cameroon, Bipindi, 1905, G.Zenker 3384 (lecto: BR!; isolecto: BM!, E, G!, HBG, GOET, K, L!, M, MO!, P!, S, US, fide J.Léonard, Mém. Cl. Sci. Acad. Roy. Sci. Belgique (8vo) sér. 2, 30(2): 224 (1957)).

Type: Cameroon, South-west Province, Korup National Park, north-western plot near P transect, Subplot 44LN, 6 Apr. 2005, X.M. van der Burgt 741 (holo: WAG!; iso: BR, G, K!, MO, P, SCA, WAG!, YA).

Macrolobium lebrunii J.Léonard, Bull. Jard. Bot. État. Bruxelles 22: 186 (1952); Anthonotha lebrunii (J.Léonard) J.Léonard, Mém. Cl. Sci. Acad. Roy. Sci. Belgique (8vo) sér. 2, 30(2): 226 (1957); Isomacrolobium lebrunii (J.Léonard) Aubrév. & Pellegr. ex Breteler, Syst. & Geogr. Pl. 78 (2): 143 (2008).

Type: D.R.Congo, between Dekese and Bumbuli, Oct. 1932, J-P.A.Lebrun 6497 (holo: BR[3 sheets]!; iso: K!, WAG!, YBI).

Macrolobium leptorrhachis Harms, Bot. Jahrb. Syst. 33(1): 157 (1902); Anthonotha leptorrhachis (Harms) J.Léonard, Bull. Jard. Bot. État. Bruxelles 25: 202 (1955); Isomacrolobium leptorrhachis (Harms) Aubrév. & Pellegr., Bull. Soc. Bot. France 104: 498 (1958).

Type: Cameroon, Bipindi, Sep. 1901, G.A,Zenker 2445 (lecto: P!; isolecto: B (destroyed), BR!, G!, GOET, K!, L!, MO!, WAG!, fide F.J.Breteler, Pl. Ecol. Evol. 144 (1): 73 (2011)).

Berlinia mengei De Wild., Pl. Bequaert. 3(1): 143 (1925); Pseudomacrolobium mengei (De Wild.) Hauman, Bull. Séances Inst. Roy. Colon. Belge 23: 477 (1952).

Type: Congo (Kinshasa), Djombo, 5 Mar. 1913, A.Mengé 88 (holo: BR).

Macrolobium leptorrhachis var. nigericum Baker f., in A.B.Rendle, E.G.Baker, H.F.Wernham and S.Moore, Cat. Pl. Oban 29 (1913); Macrolobium nigericum (Baker f.) J.Léonard, Fl. Congo Belge & Ruanda–Urundi 3: 416 (1952); Anthonotha nigerica (Baker f.) J.Léonard, Bull. Jard. Bot. État. Bruxelles 25: 202 (1955); Isomacrolobium nigericum (Baker f.) Aubrév. & Pellegr., Bull. Soc. Bot. France 104: 498 (1958).

Type: Nigeria, Oban, 1911, Talbot 582 (lecto: BM[2 sheets]!; isolecto: K!), fide F.Pellegrin, Bull. Soc. Bot. France 104: 498 (1958))

Macrolobium obanense Baker f. in A.B.Rendle, E.G.Baker, H.F.Wernham and S.Moore, Cat. Pl. Oban 28 (1913); Anthonotha obanensis (Baker f.) J.Léonard, Bull. Jard. Bot. État. Bruxelles 25: 203 (1955); Isomacrolobium obanense (Baker f.) Aubrév. & Pellegr., Bull. Soc. Bot. France 104: 498 (1958).

Type: Nigeria, Oban, 1911–1912, Mr. & Mrs. Talbot 1428 (lecto-: BM!,; isolecto- BR! (fragment), K!) fide F.Pellegrin, Bull. Soc. Bot. France 104: 498 (1958)).

Macrolobium elongatum Hutch., Bull. Misc. Inform. Kew 1916(9): 229 (1916); Anthonotha elongata (Hutch.) J.Léonard, Bull. Jard. Bot. État. Bruxelles 25: 202 (1955); Isomacrolobium elongatum (Hutch.) Aubrév. & Pellegr., Bull. Soc. Bot. France 104: 498 (1958).

Type: Sierra Leone, Pujehun, 16 Feb. 1914, C.E.Lane-Poole 161 (holo: K[2 sheets]!; iso: BR).

Macrolobium ernae Dinkl., Repert. Spec. Nov. Regni Veg. 42: 157 (1937); Anthonotha ernae (Dinkl.) J.Léonard, Bull. Jard. Bot. État. Bruxelles 25: 202 (1955); Triplisomeris ernae (Dinkl.) Aubrév. & Pellegr., Bull. Soc. Bot. France 104: 497 (1958).

Type: Liberia, Monrovia, 5 Feb. 1922, Dinklage 2805 (syn-: B (destroyed), HBG, P! [two sheets]). Breteler (2011) selected a neotype (Jongkind 5706, WAG), only to discover later that some isotypes were extant.

Macrolobium sargosii (Pellegr.) Pellegr., Bull. Soc. Bot. France 77: 666 (1931); Anthonotha sargosii (Pellegr.) J.Léonard, Mém. Cl. Sci. Acad. Roy. Sci. Belgique (8vo) sér. 2, 30(2): 225 (1957); Isomacrolobium sargosii (Pellegr.) Aubrév. & Pellegr., Bull. Soc. Bot. France 104: 498 (1958).

Type: D.R. Congo, Kouilou River, R.Sargos 101 (holo: P!; iso: BR!, WAG!). Both BR and WAG are fragments from the holotype obtained after the publication of the protologue.

Macrolobium triplisomere Pellegr., Bull. Soc. Bot. France 88: 508 (1941); Anthonotha triplisomeris (Pellegr.) J.Léonard, Mém. Cl. Sci. Acad. Roy. Sci. Belgique (8vo) sér. 2, 30(2): 226 (1957); Triplisomeris triplisomeris Aubrév. & Pellegr., Bull. Soc. Bot. France 104: 497 (1958), nom. inval. [tautonym]; Isomacrolobium triplisomere (Pellegr.) Breteler, Syst. & Geogr. Pl. 78(2): 143 (2008).

Type: Gabon, Nyanga Mount, between l’Onoij and Mouila, 26 Nov. 1925, G.M.P.C. Le Testu 5792 (lecto-: P!; isolecto: BM!, WAG!), fide J.Léonard, Mém. Cl. Sci. Acad. Roy. Sci. Belgique (8vo) sér. 2, 30(2): 226 (1957)).

Type: Tanzania, Lushoto District, between Amani and Bomole, Oct. 1905, H.G.A.Engler 3436 (syn- B(destroyed), B! [in Herb. A.Peter 48011, photo: K!]). Harms also mentioned a second gathering (W. Busse 2210) that is assigned to the species with doubt (‘Offenbar’), rendering it unclear whether this should be considered syntype material as well.

Macrolobium vignei Hoyle, Bull. Misc. Inform. Kew 1933(4): 171 (1933); Anthonotha vignei (Hoyle) J.Léonard, Bull. Jard. Bot. État. Bruxelles 25: 203 (1955); Isomacrolobium vignei (Hoyle) Aubrév. & Pellegr., Bull. Soc. Bot. France 104: 498 (1958).

Type: Ghana, Dompim, May 1930, C.Vigne 1968 (lecto-: K!; isolecto: BR, FHO!, WIS). fide A.Aubréville & F.Pellegrin, Bull. Soc. Bot. France 104: 498 (1958)). Although Hoyle marked the K sheet as holotype, the protologue cites both Kew and Oxford, rendering them syntypes.

Adapted from previous treatments for the included genera (Cowan and Polhill 1981; Breteler 2006, 2008, 2010, 2011; van der Burgt et al. 2007).

The authors declare that they have no conflicts of interest.

Manuel de la Estrella was funded by the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement number 659152 (GLDAFRICA).