[Last modification/edition: March 4th, 2024]
The classification system referred to at iNaturalist is based upon widely recognized taxonomies such as "Plants Of The World Online" (http://www.plantsoftheworldonline.org/) or "World Flora Online" (http://www.worldfloraonline.org/), formerly "The Plant List". However, for bananas, either cultivated or wild, it is frequently outdated and banana scientists do not rely on them today. Working with Bioversity International, we are in the process of making things evolve. Read it here and there on the Promusa website. In the meantime, here is a short focus on the most common types of cultivated bananas found around the world, and the diversity of wild types. And in the last chapter, we explain how to best deal with the correct botanical classification of bananas inside iNaturalist...
1- Bananas...
1-1 Bananas and plantains? Everything is bananas!
One of the most common mistakes is to set bananas against plantains. These two words are common names that have no hierarchical or classification value. This error is based on the idea that bananas would be dessert/sweet fruits, and plantains would be cooking fruits or vegetables. In fact, all plants in the genus Musa are bananas. Some are cooked, some are sweeter and are eaten raw as fruit, and many can be used in either way! The second error is that plants of composition only acuminata are bananas, and plants of hybrid nature acuminata × balbisiana are plantains. Again, this is not true, as all types of fruits are found in each of these monospecific or interspecific types. Finally, depending on the language, the word banana or plantain is used in different ways, which further complicates understanding and interpretation. In short, from a technical point of view, we should therefore call all these fruits bananas, and reserve the word plantain for the specific group they represent (see below). And above all, not to mix common language and scientific language!
1-2 Monospecific acuminata bananas
Both wild types and cultivated clones belong to Musa acuminata. Wild types are diploid and cultivated clones are diploid (AA) or triploid (AAA), rarely tetraploid. Cultivated forms are often classified into Groups. A Group of varieties of bananas is issued from a single sexual event, followed by a more or less intensive vegetative multiplication across the ages. The intensity of this multiplication depends on the success of the Group. And the higher it is, the most likely it is that natural variations occur, leading to the arousal of numerous varieties within the Group. This has been the case for two important Groups of AAA genomic composition. The most famous is Cavendish, representing nearly all export dessert/sweet bananas around the world (notice the "nearly"...) - also widely used for local consumption. It is also the case of the Mutika Group (East African Highlands beer and cooking bananas), whose varieties are the most important staple crop in this region. Other Groups of interest are Gros-Michel and Red (dessert types) for instance. It is also important to note that numerous types of bananas are not included in any Group. This is the case for less popular types, with a weaker geographical distribution, and which are in fact, the only representative of their own group! This phenomenon is true for any known genomic combination in bananas. Especially, for cultivated diploids, the vast majority are not structured into Groups. [Some rare Groups exist, however: Sucrier, Mchare, and Pisang Jari Buaya]
Hopefully, you may understand now that setting 'Cavendish' as the only common name for all Musa acuminata bananas is absolutely not reflecting the reality and the diversity this name covers at iNaturalist.
Sub-levels of Musa acuminata (Colla)
Different subspecies have been defined within the seminiferous species Musa acuminata (Colla). Specialists sometimes disagree upon the correct classification level of these types. The most well-known ones are:
-
Musa acuminata subsp. banksii (F.Muell.) N.W.Simmonds, sometimes referred to as a species (and it is the case at iNaturalist): Musa banksii F.Muell.
-
Musa acuminata subsp. zebrina (Van Houtte ex Planch.) Nasution
-
Musa acuminata subsp. malaccensis (Ridl.) N.W.Simmonds
-
Musa acuminata subsp. siamea N.W.Simmonds
-
Musa acuminata subsp. burmannica N.W.Simmonds
1-3 Interspecific acuminata × balbisiana bananas
Musa × paradisiaca is the official denomination of all M. acuminata × M. balbisiana hybrids [I personally don't like this name since it's confusing with the old and "should never be used anymore" Musa paradisiaca, but that's a fact.]. All these hybrids are cultivated clones, mainly triploid. Musa × paradisiaca is in no way equivalent to plantain! Two hybrid formulae are defined: AAB and ABB, depending upon the relative importance of M. acuminata and M. balbisiana in these types.
Within AAB, the most common Group is Plantain, which refers only to this particular cooking type, mainly found in Africa. Other examples of Groups are Silk and Prata, which are dessert types originating from India, well-distributed also in South America. Popoulou and Maia Maoli are cooking types spread across the Pacific, sometimes wrongly referred to as Pacific Plantains, which adds to the confusion.
Within ABB, most of the cooking types belong to the Bluggoe Group (widely distributed around the world, often named Orinoco), and dessert types belong to the Pisang Awak Group. Other known Groups are for instance Saba, Ney Mannan, or Monthan.
Of course, many other Groups exist, with lower importance or distribution.
It must be added that in India, there are numerous diploid types, hybrid between Musa acuminata and Musa balbisiana. These cultivated AB bananas are included within Musa × paradisiaca. The main known types are Ney Poovan and Kunnan, but a highly wider diversity exists here, and is often still to be clearly categorized.
Recent discoveries suggest that the reality is in fact even more complicated (which is not really a surprise). Particularly, a significant number of natural tetraploid plants are known today, sometimes linked to already existing groups (Pisang Awak for instance), and sometimes not.
1-4 Other species
Musa balbisiana is only a diploid “wild” species. It is a seedy banana, like the “wild” Musa acuminata types, but far more vigorous. I put “wild” between quotes since it is also a resource-plant, sometimes cultivated for its fibers, for instance in the Philippines, sometimes used for cattle feeding, and whose unripe fruits are also sometimes eaten. Although a wide diversity exist, no subspecies are defined within M. balbisiana, whereas subspecies exist within M. acuminata (malaccensis, banksii...).
A bunch of other species exists within the genus Musa, that did not participate – or very marginally – in the elaboration of the cultivated banana species: Musa schizocarpa, Musa itinerans, Musa basjoo, etc. As we gain insight into this marginal diversity, it appears that new combinations of species exist, that are not yet well documented. It is then very likely that new hybrid formulae will flourish in a near future for these plants.
All the plants described so far (seminiferous and cultivated) belong to the ‘Eumusa’ section. Ornamental types exist within the ‘Rhodochlamys’ section: Musa ornata, Musa velutina, and Musa laterita for instance. Recently, it was suggested to gather these two sections in a single one called ‘Musa’ (notice the originality…) but it is still a matter of discussion between specialists.
In the Pacific area, cultivated bananas with a distinctive erect bunch are called Fehi (or Fe’i). They are sometimes referred to as Musa troglodytarum. These plants belong to the ‘Australimusa’ section, which includes also various seminiferous species: Musa textilis and other less known species (Musa boman, Musa jackeyi...).
Here again, Latin names should not be used to identify Fe’i bananas. The old Musa uranoscopus, the often still used Musa troglodytarum, and even sometimes Musa fehi, are names to ban definitely. We may preferably use a Fe’i Group to gather these cultivated types from the Pacific. Moreover, as new studies about these types are arising, it seems that their classification is even more complicated than previously thought. For instance, we now know that triploid types exist also here. It is also known today that natural intersectional hybrids between Eumusa and Australimusa are found, for instance in Papua New-Guinea and which reveal to be tetraploid. This classification is then for sure, on the move!
Another section, ‘Callimusa’, regroups ornamental types such as Musa coccinea, Musa beccarii, or Musa campestris for instance. And here again, a fusion between ‘Australimusa’ and ‘Callimusa’ has been proposed, within a single new ‘Callimusa’ section.
1-5 Genus Ensete
Within the Musaceae family, there is another important genus, Ensete. Plants in this genus are fertile, diploid, and don't produce naturally suckers. They only reproduce the sexual way. Several species have been described, but three of them are the most important. Ensete ventricosum is mainly found in Africa and America. It is even a very important staple crop in Ethiopia, where it is intensively cultivated, and several varieties are described. Ensete glaucum grows essentially in South East Asia and may also be found in India. In this latter area, you may also find Ensete superbum, growing naturally on the West side of Kerala. Other less important species are Ensete giletii in West and Central Africa, the small Ensete homblei, at the crossroads of Zambia, Congo and Angola, and Ensete perrieri (native to Madagascar). And the differences between these last species are often tough to describe.
1-6 Genus Musella
One last species is a botanical matter of discussion among specialists: Musella lasiocarpa. A third genus, Musella has been created especially for it. This small plant, giving erected yellow or orange buds, and known to endure cold weather and even snow, has also sometimes been classified as Ensete lasiocarpum. The discussion is still going on, but the official present naming is Musella.
2- Systematic
2-1 Some naming examples and how to deal with the real botanical classification inside iNaturalist
The main goal of iNaturalist is to record and classify wild organisms, whether they are plants or animals. Classification then relies on Latin binomial specifications with concepts of the genus, species, subspecies, and other infraspecific groupings. Recorded plants should be wild ones, living in natural conditions. But when it comes to bananas, most of the observed types are cultivated ones. They are sterile and only reproduce through vegetative multiplying. In this case, the used scheme of classification should be a horticultural one, as developed by the International Code for the Nomenclature for Cultivated Plants (ICNCP - https://www.ishs.org/news/icncp-international-code-nomenclature-cultivated-plants-9th-edition), but it is not implemented at iNaturalist. Trying to apply, for instance, the concept of species, to a sterile plant, is nonsense. This is why the classification and naming of bananas at iNaturalist is generally non-satisfactory and represents a compromise between what should be done and what is effectively possible. According to the ICNCP, which itself derives from the ICN, here are some naming examples for bananas, following the iNaturalist taxonomy scheme.
- The Plantain cultivar ‘French Clair’: Musa × paradisiaca, Plantain Group, ‘French Clair’ (AAB)
- The Cavendish cultivar ‘Grande Naine’: Musa acuminata, Cavendish Group ‘Grande Naine’ (AAA)
- The ‘Mlali Angaia’ Mchare diploid clone: Musa acuminata (Gp. Mchare) ‘Mlali Angaia’ (AA)
- The ‘Ney Poovan’ diploid hybrid clone: Musa × paradisiaca ‘Ney Poovan’ (AB)
- The wild balbisiana variety ‘Klue Tani’: Musa balbisiana var. klue tani OR Musa balbisiana ‘Klue Tani’ (if cultivated)
- The wild ‘Pahang’ acuminata variety: Musa acuminata subsp. malaccensis var. pahang (wild)
Please remind that this is a (non-satisfactory) compromise between what should be written and how iNaturalist deals with classification. Species names should not appear when speaking of cultivated forms. So here is the same list, strictly following ICNCP standards:
- The Plantain cultivar ‘French Clair’: Musa, Plantain Group, ‘French Clair’ (AAB)
- The Cavendish cultivar ‘Grande Naine’: Musa, Cavendish Group ‘Grande Naine’ (AAA)
- The ‘Mlali Angaia’ Mchare diploid clone: Musa (Gp. Mchare) ‘Mlali Angaia’ (AA)
- The ‘Ney Poovan’ diploid hybrid clone: Musa ‘Ney Poovan’ (AB)
- The wild balbisiana variety ‘Klue Tani’: Musa balbisiana var. klue tani OR Musa ‘Klue Tani’ (BB) -if cultivated
- The wild ‘Pahang’ acuminata variety: Musa acuminata subsp. malaccensis var. pahang (wild)
2-2 Some names, not to be used anymore… at all… (list in progress)
Many Latin binomial names can be found in literature and on websites. But in reality, very few of these names are actually accepted as official names for Musa. Many old names, dating from a time when Musa taxonomy was still hesitant, are still used here and there, although classification articles have since clarified their synonymy, redundancy or uselessness. What's more, in more recent times, a number of careless horticulturalists have not hesitated to invent new Latin names, without any basis for doing so, purely for commercial purposes. The result is a blossoming of eccentric names that have no justification whatsoever. Here is a list of names to be avoided, which I will amend as I go along.
I would remind you that for a species' Latin name to be valid, it must be the subject of an official and accepted publication describing the plant, its distribution, its origin and its differences from other closely related species. For example, the journal Adansonia, published by the Muséum d'Histoire Naturelle in France, is one of the systematics publications for vascular plants.
- Musa paradisiaca
-
Musa × paradisiaca var anything: should only be used alone
- Musa sapientum
- Musa Cavendishii