An annotated list of the species of Ensete, Musa and Musella.
Sectional arrangement of Musa
Historically the genus Musa has been divided into sections. The sections are merely convenient divisions of the genus, they do not have any formal taxonomic significance, are not definitive, nor do they have equal rank. They are however, useful in discussing the relationships between species. Traditionally, Musa sections were circumscribed by chromosome number and phenotypic characters. Recently the discovery of new species with anomalous characters and the application of genome analysis has revealed deficiencies in the traditional sections such that they may not survive much longer. For example, there now appear to be species that "link" the sections. Musa flaviflora in section Eumusa (Musa) 1 and Musa velutina in section Rhodochlamys appear to link these two 10-chromosome sections. This link is exemplified by Musa ornata that seems to be a hybrid between M. flaviflora and M. velutina. Similarly, Musa lawitiensis var. suratii has recently been interpreted as intermediate between the 11-chromosome sections Callimusa and Australimusa suggesting those sections cannot be maintained as distinct. These examples show a blurring of the distinctness of the traditional sections. Then again there are phenotypically similar species such as Musa paracoccinea and the un-named Chuoi Rung Hoa Sen with strange nail-headed seeds that do not neatly fit into the traditional sections.
Studies by Carol Wong and colleagues in Singapore have revealed that genetic differences between each section in the same chromosome group are smaller than those within each section. This means that the traditional separation of the sections can no longer be substantiated. The studies of Wong et al do, however, maintain the separation between the 20 and 22 chromosome species. The 2n = 14 Musa ingens remains in its own section. The morphological differences that once supported the separation of the sections are no longer considered important in determining sectional status. The species sorted into the sections as interpreted by Wong et al are here.
The placement of some Musa species in the sections is provisional, see notes.
NOTE This table will need to be modified to take into account the proposed 'acuminata' and 'ornata' groups within Sect. Musa, and the 'coccinea' and 'textilis' groups within sect. Callimusa as defined by Wong et al Gardens' Bulletin (Singapore), 55 (1) : 97-111. (2003).
Genus Chromosome number
(x = 1n)
but see notes.
(genus Musa only)
Distribution Species Main uses
Africa to Papua New Guinea. E. gilletii
E. wilsonii [see note]
E. Banta, Thailand [see note]
Food staple, fibre, thatch, construction, beads, medicinal & ornamental.
Ingentimusa Papua New Guinea M. ingens [see note] not known
Australimusa Queensland, Indonesia, Papua New Guinea, Philippines. M. alinsanaya
M. beccarii [see note]
M. boman [see note]
M. fitzalanii [extinct]
M. insularimontana [see note]
M. lawitiensis [see note]
M. monticola [see note]
Fibre (M. textilis), fruit (fe'i bananas).
Callimusa Indo-China to Indonesia
M. campestris [see note]
M. coccinea [see note]
M. exotica [see note]
M. hirta [see note]
M. lokok [see note]
M. paracoccinea [see note]
M. pigmaea [see note]
M. splendida [see note]
India to Samoa. M. acuminata
M. banksii [see note]
M. flaviflora [see note]
M. griersonii [see note]
M. itinerans [see note]
M. siamea [see note]
M. thomsonii [see note]
M. sp. 'Burmese Blue' [see note]
Fruit, vegetable, wrapping, ornamental.
India to Samoa. M. balbisiana
M. sikkimensis [see note]
Fruit, fibre, vegetable, wrapping, ornamental.
Rhodochlamys India to Indo-China. M. angcorensis [see note]
M. lutea [see note]
M. ornata [see note]
M. rosea [see note]
M. velutina [see note]
M. viridis [see note]
China (Yunnan & Guixhou) Vietnam, Laos. M. lasiocarpa
M. splendida [see note]
Animal fodder, vegetable, medicinal & ornamental.
Based on a Table 1.1 in Stover & Simmonds 1987 but modified substantially by information from Simmonds 1960, Hotta 1967, Argent 1976, Simmonds & Weatherup 1990, Jong & Argent 2001, Wong et al 2001.
Ensete wilsonii (Tutcher) Cheesman should probably be reduced under Ensete glaucum (Roxb.) Cheesman.
Ensete Banta, Thailand. There are reports that John Banta has found a new Ensete sp. in Thailand.
The section name Eumusa is very commonly used in the literature but following article 32 of the the ICBN it is illegal and the section should properly be named section Musa. Eumusa is widely used to prevent confusion between the genus Musa and the section Musa. By convention taxa above the genus level are not commonly italicised..
Musa angcorensis Gagnep. is poorly known and may not be a good species. Simmonds 1960 very tentatively placed it in section Callimusa but, if it exists at all, it is more likely to be section Rhodochlamys.
Musa banksii F. Muell. seems to be confirmed as a species in section Musa (1) by Simmonds & Weatherup 1990 supporting the conclusion of Argent 1976 although Shepherd 1990 disagrees.
Musa beccarii Simmonds was "confidently placed within section Australimusa" by Wong et al 2001 based on seed structure, chromosome number and AFLP results. However, Shepherd 1999 gives the chromosome number as x (= 1n) = 9 so its position still seems somewhat anomalous and Jong & Argent 2001 maintain M. beccarii as species incertae sedis on this basis. It exists as two varieties, var. beccarii and var. hottana.
Musa boman Argent is placed in section Australimusa as determined by Argent 1976 although it was transferred to section Musa (2) by Simmonds and Weatherup 1990 despite the fact that it had there an "anomalous" chromosome number of x (= 1n) = 10.
Musa campestris Beccari exists as six varieties, var. campestris, var. lawasensis, var. limbangensis, var. miriensis, var. sabahensis and var. sarwakensis.
Musa coccinea Andrews has been confirmed by Liu et al 2002 and by Argent and Kiew 2002 as the correct name for the plant often called Musa uranoscopos Lour. in the literature.
Musa exotica Valmayor is described from Vietnam where it is known as Chuoi Rung Hoa Do. It has previously been considered to belong in the Rhodochlamys but the fruits perpendicular to the rachis exclude it from this section as described by Cheesman. Valmayor places it in the Callimusa. It has the correct chromosome number but the diagnostic seed description is lacking.
Musa flaviflora Simmonds (one of the "parents" of M. ornata according to Shepherd 1990) should perhaps be treated as a sub-species of Musa acuminata Colla according to Simmonds & Weatherup 1990.
Musa griersonii Noltie is a good species and from its synonymy it must belong to section Musa. I have tentatively placed it in section Musa (1) pending further information.
Musa hirta Becc. is a good species although somewhat enigmatic and its affinities with M. beccarii suggest it belongs to section Callimusa to which the latter was transferred by Simmonds and Weatherup 1990. However, it should be noted that Jong & Argent 2001 maintain M. beccarii as incertae sedis and this must cast doubt also on the position of M. hirta.
Musa lawitiensis Nasution and Supardiyono exists as 4 varieties: var. lawitiensis, var. suratii, var. sarawakensis & var. kapitensis.
Musa lokok Geri & Ng is placed here speculatively.
Musa insularimontana Hayata endemic to a single island off Taiwan is poorly known. It is close to M. textilis and is perhaps vulnerable to reduction.
Musa ingens Simmonds was placed in a new section Ingentimusa by Argent 1976 and is maintained there by Jong & Argent 2001. Simmonds and Weatherup 1990 placed M. ingens in section Musa (2) although it has there an "anomalous" chromosome number x (= 1 n) = 7.
Musa monticola (Hotta ex) Argent was placed tentatively in section Australimusa by Argent 1976 and confirmed there by Jong & Argent 2001 and Wong et al 2001.
Musa lutea Valmayor et al is placed here tentatively because of its upright inflorescence.
Musa ornata Roxb. seems to be a "secondary species" according to Shepherd 1990, a relic of a hybrid swarm between M. flaviflora and M. velutina.
Musa paracoccinea A. Z Liu & D. Z. Li is placed in section Callimusa by the authors but not entirely convincingly.
Musa pigmaea Hotta (nomen nudum as yet) is close to M. beccarii according to Hotta so presumably belongs in section Callimusa in which the latter is placed by Simmonds and Weatherup. However, it should be noted that Jong & Argent 2001 maintain M. beccarii as incertae sedis and this must cast doubt also on the position of M. pigmaea.
Musa rosea sensu Baker is more or less unknown. It may be M. ornata.
Musa sakainana Meekiong et al is placed here speculatively.
Musa siamea (comb. nov) is likely to be elevated from M. acuminata subsp. siamea if Wong et al's genetic results are confirmed.
Musa sikkimensis Kurz is a good species allied to Musa nagensium Prain and I have placed it with the latter in section Musa (2) pending further information.
Musa splendida A. Chev. has been re-discovered in Vietnam. It has obvious phenotypic affinities with M. exotica and M. paracoccinea and while neither of these sit comfortably in the Callimusa it seems reasonable for the time being to include them all in that section.
Musa thomsonii Noltie is a good species and from its synonymy it must belong to section Musa. I have placed it in section Musa (1) pending further information.
Musa velutina Wendl. & Drude may be the same as Musa dasycarpa Kurz in which case the latter would have priority.
Musa viridis Valmayor et al is placed here tentatively because of its upright inflorescence.
Musa sp. 'Burmese Blue' is placed here deliberately provocatively. This plant has been describe as a form of M. acuminata and M. balbisiana but having grown the plant it seems to me that it is neither and is possibly an undescribed species.
Musella splendida seems to me doubtfully distinct from Musella lasiocarpa.
home back top
last updated 29/11/2007