Who's breeding bananas?

Anne Vézina Thursday, 07 August 2014

Bananas have a reputation for being difficult to breed; a reputation that seemed richly deserved when after more than 60 years of trying, breeders had yet to release an improved banana hybrid. No wonder you used to be able to count the number of banana breeding programmes on the fingers of one hand1 . Nowadays, however, you would not have enough of two hands. According to a survey done for the CGIAR Research Program on Roots, Tubers and Bananas (RTB) in preparation for a workshop on banana genetic improvement that took place last fall, at least 12 organizations have a banana breeding programme2 .

Africa, Asia and the Americas each have four breeding programmes based on their soil. Africa is represented by the Centre Africain de Recherches sur Bananiers et Plantain (CARBAP) in Cameroon, the Centre National De Recherche Agronomique (CNRA) in Côte d’Ivoire, the International Institute of Tropical Agriculture (IITA) in Nigeria and Uganda, and the National Agricultural Research Organisation (NARO) in Uganda; Asia by the Chinese Academy of Tropical Agricultural Sciences (CATAS) and the Guangdong Academy of Agricultural Sciences (GDASS) in China, the Indonesian Fruits Research Institute (ITFRI) in Indonesia and the National Research Centre for Banana (NRCB) in India; and the Americas by the Armstrong Atlantic State University (AASU) in the USA, the French agricultural research centre CIRAD in Guadeloupe, the Empresa Brasileira de Pesquisa Agropecuária (Embrapa) in Brazil and the Fundación Hondureña de Investigación Agrícola (FHIA) in Honduras (see Update for more centers).

The survey results also show that a minimum of 76 improved hybrids representing a range of cultivar groups have been released so far, and that the strategies for coaxing bananas into exchanging genes with each other have diversified over the years (see table below and click here to download the PDF of the summary table). Some of the breeding programmes also have a genetic engineering component.

The evolution of banana breeding

Banana breeding started in the early 1920s, when the British government set up programmes in Trinidad and Jamaica to make the then export banana, the cultivar ‘Gros Michel’, resistant to Fusarium wilt. Not surprisingly, breeders struggled to obtain seeds from the highly sterile cultivar. For example, out of the 123 seeds obtained by hand-fertilizing 20,000 flowers with the pollen of a wild banana species, only 18 germinated (at the time breeders were dependent on seeds germinating naturally, whereas today they ‘rescue’ the embryo in the seed and culture it). To top it off, several of the seedlings died before flowering, which usually happens when hybrids are missing one or more chromosomes (aneuploids) or have more than 4 copies of each chromosome (hyperploids).

Looking for seeds
Looking for seeds

Scientists knew that breeding seedless bananas would be anything but simple. They were also handicapped by a lack of basic knowledge about the plant’s genetics and cell biology, not to mention a nomenclature system to classify cultivated bananas that was nothing if not confusing. As Ernest Cheesman noted in 19323 , they weren’t even sure of the identity of the wild species they were using as male parent.

With their crosses, breeders were trying to combine the agronomic qualities of the cultivar with the particular disease resistance of the wild banana. The problem was that few of the resulting hybrids had both the fruit quality of the female parent (the cultivar) and the disease resistance of the male one. The solution was to create ‘elite’ male parental lines that were disease resistant and had good fruit qualities, and to cross them with the cultivar needing improvement. This strategy became known as the 3x/2x breeding scheme. The numbers represent the ploidy of the parents. The parent that has three copies of each chromosome (3x) is called a triploid. The majority of banana cultivars, including ‘Gros Michel’, are triploids4 . The parent with two sets of chromosomes (2x) is called a diploid. 

To improve the agronomic characteristics of the diploid male parent, breeders started crossing diploids with each other (2x/2x). Diploid improvement programmes  were set up in Trinidad and Jamaica, and later on in Honduras, where the United Fruit Company (nowadays Chiquita) started its own banana breeding programme in 1959. In 1984, United Fruit donated its programme to FHIA, which became the first organization to release tetraploid hybrids (four chromosome sets) by using the 3x/2x breeding scheme.

Tetraploid hybrid with seeds
Tetraploid hybrid with seeds

Since the triploid parent (a cultivar with female residual fertility) will contribute anything from 1 to 3 chromosome sets because of its incapacity to undergo normal meiosis, the hybrids resulting from a 3x/2x cross will either be diploid, triploid or tetraploid. Breeders, however, only evaluate the tetraploid progeny to introduce to farmers because these hybrids are the only ones that have all the genes of the cultivar. However, since tetraploid hybrids tend to be pollen fertile, and as a result may occasionally produce fruits containing seeds, breeders started crossing them with improved diploids to obtain 'secondary' triploid hybrids in what is called the 4x/2x breeding scheme.

According to the information provided by the respondents to the survey, the 3x/2x breeding scheme combined with a diploid improvement programme is still the most widely used strategy. It is followed by the 4x/2x breeding scheme, which should not be confused with the 4x/2x breeding scheme that calls for ploidy manipulation. In the latter, the 4x parent is a diploid whose chromosome numbers have been doubled using colchicine. These doubled-diploids are then crossed with diploids to produce triploids. Robert H. Stover and Ivan W. Buddenhagen first proposed the breeding scheme in a 1986 paper. It is further described in a profile on Frederic Bakry, the first breeder to use it.

The complete results of the survey, along with a review paper prepared for the RTB workshop, provide details on the various molecular tools now available to breeders. They have yet to speed up banana breeding, which still remains a challenge, but if progress continues to be made, the next review may reveal that even more organizations have given banana breeding a try.

In India, the Kerala Agricultural University and the Tamil Nadu Agricultural University also breed bananas.

1 According to a 1994/1995 review sponsored by the World Bank, 4 organizations were breeding bananas at the time: CIRAD, Embrapa, FHIA and IITA.
2 The list does not include programmes that try to improve bananas by means other than cross-fertilization. The list may also not be comprehensive. See Update below for the names of two banana breeding organizations that were added after the blog post was put online.
3 Cheesman, E.E. 1932. Genetic and cytological studies of Musa. 1. Certain hybrids of the Gros Michel banana. Journal of Genetics 26(3):291-312.
4 Triploid bananas are the product of a mating between two diploid cultivars, one of which passed on its entire genome when its chromosomes failed to segregate during meiosis, a phenomenon called meiotic restitution.
Related stories
Further reading
Review paper by R. Ortiz and R. Swennen. From crossbreeding to biotechnology-facilitated improvement of banana and plantain, published in Biotechnology Advances 32 (2014) 158-169.
PDF of the results of the survey on banana breeding