January 11, 2022

Banana, Fusarium, and American singer-songwriter Anastacia

Written by: Jamie Pike.  Over the past week, I have had several conversations about classification and taxonomy and something I have become quite familiar with throughout my PhD are the challenges surrounding banana and Fusarium classification. My PhD focuses on developing diagnostic tools for the disease Fusarium wilt, caused by Fusarium oxysporum f. sp. cubense (Foc) (syn. Fusarium odoratissimum), in one of the world’s top agricultural commodities, the banana (Musa spp.). In only one sentence I have described both the plant and the fungus in a few different ways; banana and Musa spp. as well as Fusarium oxysporum f. sp. cubense and Fusarium odoratissimum.

 So, what would you call this plant?

Figure 1: Name that plant!

I quickly asked some of my friends and colleagues to name the plant in figure 1 and we had banana, pisang, banane, banaan, saging and muz. “Ah, but Jamie, that’s why we use the binomial nomenclature. Linnaeus solved that problem way back in the 1750s! Just call it Musa and be done with it.” But Musa what?

Bananas have a complex genetic history, involving hybridisation between species and polyploidy (more than two complete sets of chromosomes). Due to domestication, most of the varieties which are cultivated today are triploid (three sets of chromosomes) and descend from different subspecies of Musa acuminata, or when the latter hybridised with Musa balbisiana. Consequently, most domesticated banana varieties are sterile and seedless, with their genomes comprised of acuminata and/or balbisiana.

 Therefore, when naming banana plants, varieties are frequently assembled into subgroups (based on similarity to other varieties) and a cultivar name given as well as genome group score. The group scoring system, developed by Norman Simmonds and Kenneth Shepherd in the 1950s, classifies bananas into genome groups based on ploidy contribution of the wild species; Musa acuminata (A) and Musa balbisiana (B). An example of the current system used to classify banana would be “Cavendish banana (Musa spp. AAA group) cv ‘Grand Nain’”. It indicates the variety’s sub-group (Cavendish), genus (Musa), genome group scoring (AAA), and cultivar (Grand Nain). But this is not like the traditional binomial formula we’re used to. This system seems to be pretty well established in banana circles, but there are still discussions about it (see: Nomenclature system for edible bananas).

For Foc, however, the matter of classification is certainly not settled. A paper published by Maryani et al., in 2019 attempted to reclassify the pathogen into the Fusarium of Banana Species Complex, proposing nine new species. This year, a paper by Torres Bedoya et al., argued that the revision was premature and encouraged the use of the traditional system.

Foc belongs to the Fusarium oxysporum species complex. Fusarium oxysporum is comprised of endophytes, pathogens, and potentially some saprophytes. Most endophytes within the Fusarium oxysporum species complex are just classified as Fusarium oxysporum, but the pathogens are further classified into special forms or formae speciales (f. sp.), of which there are at least 106, and potentially more than 200! “But Jamie, why are they divided into special forms?” Good question, I’m glad you asked. Well, they’re only able to infect specific host species or groups of plants. For instance, Fusarium oxysporum f. sp. lycopersici infects tomato, Fusarium oxysporum f. sp. lactucae infects lettuce, and Fusarium oxysporum f. sp. cubense (Foc) only infects banana (Figure 2).

Within most special forms there are races. The races are determined by the susceptibility of cultivars within the host species. In Foc, there are three races; Races 1, 2 and 4. Race 1 can infect banana varieties like ‘Gros Michel’ but not those within the Cavendish subgroup, Race 2 can infect most cooking bananas and Race 4 can infect the hosts of Race 1 and Race 2 and Cavendish varieties. Race 4 is classified even further into Tropical Race 4 and Subtropical Race 4, but we won’t get into that!

Figure 2: Classification of Fusarium oxysporum pathogens. Pathogenic forms of Fusarium oxysporum are further classified into formae speciales or ‘special forms’ based on their host specificity. Each special form can be further divided into race, dependent upon their ability to infect specific plant cultivars.

This system is a little convoluted, but when it was first proposed it seemed to solve the Fusarium oxysporum pathogen classification problem. However, under the pressure of phylogenetic analysis which has employed new sequencing technologies, the special form and race classification system has started to fall apart. Short sections of DNA from a specific gene or genes that have variation between species but little variation within species so you can classify an organism, known as genetic barcodes, have become standard for phylogenetic analysis of Fusarium oxysporum isolates. When these analyses are performed, isolates from one race of a special form frequently appear to be more closely related to races from other special forms than races within their special form. Adding more barcodes doesn’t solve the problem either.

Why? Well, we’re pretty sure it’s because Fusarium oxysporum can transfer chromosomes between isolates, and that these chromosomes are responsible for its ability to infect. So, the genome of an isolate can be divided into two parts: the first containing all the essential genes for survival, the “core genome” and the second, the “accessory genome”, which contains the gained chromosome(s) that enables the special form of Fusarium oxysporum to infect specific plant species. Barcoding typically classifies isolates using genes found in the core genome, and the special form and race classification system uses genes in the accessory genome.

Think of it in terms of celebrities; if you were going to classify Anastacia, Cameron Diaz and Tom Hanks and we said hair colour is a core genome trait and acting ability is accessory. For the core genome trait, Cameron Diaz and Anastacia would group together and Tom Hanks would be left on island alone screaming “WILSOOOON!”. If you grouped based on the accessory genome trait, acting ability (or for the special forms of Fusarium oxysporum, the ability to infect a specific plant species), Cameron Diaz and Tom Hanks would group together, and Anastacia would be left outside alone.

Neither of these methods for grouping Fusarium oxysporum – or indeed celebrities – is wrong or right, but it highlights the challenges presented by the classification system historically used for Fusarium oxysporum. And, like the banana, conveys some of the challenges faced by modern-day taxonomy and classification. That being said, I have no useful solutions to the classification challenges presented by banana or Foc and will continue to spend hours reading papers where it is debated. Except now, you will all know my struggle.


Maryani, N., Lombard, L., Poerba, Y.S., Subandiyah, S., Crous, P.W. and Kema, G.H.J., 2019. Phylogeny and genetic diversity of the banana Fusarium wilt pathogen Fusarium oxysporum f. sp. cubense in the Indonesian centre of origin. Studies in Mycology92, pp.155-194.

Torres Bedoya, E., Bebber, D. and Studholme, D.J., 2021. Taxonomic revision of the banana Fusarium wilt TR4 pathogen is premature. Phytopathology, (ja).