Leandra glabrata (Bunbury) Cogniaux (1886: 172)

Reginato, Marcelo & Goldenberg, Renato, 2018, Taxonomic notes on Leandra (Melastomataceae, Miconieae) - II, Phytotaxa 371 (2), pp. 84-92 : 86-87

4. Leandra glabrata (Bunbury) Cogniaux (1886: 172) . ≡ Clidemia glabrata Bunbury (1841: 108) . Type:— BRAZIL.

Minas Gerais: “Hab. propre Gongo Soco”, C.J.F. Bunbury s.n. (lectotype BR-5190331! designated here).

= Leandra sessiliflora Cogniaux (1886: 610) . Type:— BRAZIL. Minas Gerais: “Gandarela”, Glaziou 14774 (lectotype LE designated here, not seen, isolectotypes B destroyed [F type photo 16979], BR-5195213!, K-535689!, P-117056!). Syn. nov.

Specimens examined:— BRAZIL. Espírito Santo: Alfredo Chaves, P. Fiaschi 3555 (NY); Domingos Martins, G.G. Hatschbach 50775 (CEPEC); Itaguaçu, A.P. Fontana 831 (UPCB); Marechal Floriano, G.G. Hatschbach 60102 (CAS, CEPEC); Vargem Alta, R. Goldenberg 2139 (NY). Minas Gerais: Caldas, A.F. Regnell 545 ( US); Catas Altas, R. Goldenberg 1498 (NY, UPCB); Itabira, A.F.M. Glaziou 17537 (R); Lima Duarte, A.C. Mezzonato 17 (UPCB); Ouro Preto, M. Reginato 1166 (NY, UPCB); Rio Preto, F.R.G. Salimena 1173 (UPCB); Sabará, L. Riedel 2703 (NY). Paraná: Adrianópolis, J.M. Silva 3177 (UPCB); Tunas do Paraná, R. Goldenberg 742 (NY, UPCB).

Leandra sessiliflora was described in “Flora Brasiliensis” “addenda et emendata”, and placed it right after L. glabrata (Cogniaux 1888) , which means that the author was aware that both species should be similar. It seems that he was mislead by the depauperate, nearly sterile specimens by Glaziou, since none of the characters listed by him in L. sessiliflora ’s description are absent in our present day sampling for L. glabrata .

The choice for Bunbury’s type for L. glabrata is tentative, since no duplicate has been found in other herbaria. The original text by Foster mentioned (p.110) that Bunbury’s specimens were donated to the Linnean Society, now housed by the British Museum (for details on the Linnean Herbarium see Linnean Society 2016), but Desmond (1994) mentioned that Bunbury’s Brazilian plants were deposited at Kew. We had Leandra glabrata ’s folders checked both at BM and K, and Bunbury’s specimen was not found.As for L. sessiliflora, Cogniaux (1888) mentioned B and LE; since the specimen at B was destroyed, the specimen at LE should be chosen as the lectotype.

5. Leandra multiplinervis (Naudin) Cogniaux (1886: 180) . ≡ Clidemia multiplinervis Naudin (1851: 358) . Type:— MEXICO. “prope Zacualpan”, J.J. Linden s.n. (lectotype P-00116850! designated here, isolectotype BR-5192014!). = Leandra longisetosa Cogniaux (1886: 181) . Type:— BRAZIL. Rio de Janeiro: “Serra do Coelho”, A.F.M. Glaziou 8691 (lectotype BR-

5192946! designated here, isolectotypes G-353582!, K-535680!, R-10054!, P-117338!, P-117339!). Syn. nov.

Specimens examined:— MEXICO. Chiapas: Pueblo Nuevo Solistahuacan, T.B. Croat 65168 (NY); Rayón, D.E. Breedlove 32334 (NY); Tenejapa, A.S. Ton 790 (NY). GUATEMALA. Alta Verapaz: Coban, H. Tuerckheim 1819 (GH, NY, R); San Juan Chamelco, H. Foerther 10406 (M); Tactic, H. Tuerckheim 8388 (GH, M, NY). Baja Verapaz: Coban, H. Foerther 10871 (M); El Progresso, H. Foerther 2671 (M); La Union Barrios, R.M. King 7131 (NY); Purulha, J.J. Castillo 995 (NY, RB). EL SALVADOR. Santa Ana: A. Molina 16700 (NY), A. Molina 12542 (NY). BRAZIL. Espírito Santo: Castelo, R. Goldenberg 1179 (RB); Conceição do Castelo, G.G. Hatschbach 49910 (CEPEC). Minas Gerais: Bocaina de Minas, F. Markgraf 10380 (RB); Camanducaia, L.D. Meireles 856 (RB); Fervedouro, J.A. Lombardi 3128 (UPCB); Itamonte, P. Schwacke 5253 (RB); Passa Quatro, A.C. Brade 19066 (RB). Paraná: Campina Grande do Sul, G.G. Hatschbach 4067 (UPCB); Morretes, M. Reginato 761 (UPCB); Piraquara, R. Goldenberg 790 (NY). Rio de Janeiro: Itatiaia, A.C. Brade 17320 (RB); Nova Friburgo, F. Almeda 8809 (UPCB); Petrópolis, D. Sucre 4160 (RB); Resende, G. Eiten 6690 (NY); Santa Maria Madalena, M. Reginato 1216 (NY, UPCB); Teresópolis, M. Reginato 1110 (UPCB). São Paulo: Bananal, A.C. Brade 20103 (NY, RB); Campos do Jordão, P.C. Porto 3171 (RB); Cunha, P.G. Windisch 5036 (SP); Mogi das Cruzes, M. Kirizawa 1036 (SP); Salesópolis, J. Mattos 12455 (NY, SP); São Paulo, M. Kuhlmann s.n. (SP 40047).

Leandra multiplinervis is long known to have a disjunct distribution between eastern Brazil and Mesoamerica. Its type specimen is from Mexico, but specimens from eastern Brazil have been assigned to this species ever since Cogniaux (1888). Interestingly, no specimens are known from the Andes or other intermediary areas. Cogniaux (1888) described L. longisetosa and placed it close to L. multiplinervis in his key. Overall, the types of both species are similar, except for the very conspicuous setose indumentum on the branches of L. longisetosa vs. branches glabrous or slightly pubescent in L. multiplinervis , which is the only difference pointed out by Cogniaux (1888). Nonetheless, some variation is observed in the amount and size of trichomes across different specimens of L. multiplinervis . It is noteworthy that some eastern Brazilian specimens are more similar to the Mesoamericans than to some other specimens from Brazil. Thus, treating the eastern Brazilian specimens as L. longisetosa (type from eastern Brazil), while keeping the Mesoamerican specimens under L. multiplinervis (type from Mexico), does not seem to be reasonable based on morphology. Therefore, we propose to recognize a single species, which is polymorphic for branch indumentum in eastern Brazil.