What can make me feel less guilty about buying bananas? Science.

I am genuinely curious about the size of the fraction of carbon in my two-year-old that is derived from bananas. When we have bananas in the house, which is most of the time, she eats at least part of one every day. She loves them peeled, in smoothies, dried, in banana bread, or in these banana-rich cookies, which sound like they shouldn’t be good but are totally amazing. Bananas are inexpensive and delicious, and making nutritious food with them gives me a sense of parental accomplishment. Nonetheless, always I feel a niggling sense of guilt whenever I plunk a bunch of bananas into the shopping cart. Prosaic though it may be, most of this is contrition inspired by the “local food” movement. I know that very little is benign about the process responsible for bringing these highly perishable tropical fruits to my table for less than a dollar a pound. The remainder of my remorse is conviction that bananas should not be taken for granted. Not only is banana history and biology interesting, but the banana variety in our grocery stores, the Cavendish, is in danger of commercial extinction. There isn’t an easy solution to the problem or an obvious candidate for a replacement variety. The history of the Cavendish’s rise, and the biology behind its current peril, makes for a good story.

Banana basics

Before we launch into banana history and breeding, I want to make sure we’re all on the same page about basic banana botany. Bananas are in the monocot genus Musa (family Musaceae, order Zingiberales). There isn’t an exact count on the number of species in the genus because hybrids are rampant, which we will see below is a very significant detail, but there are probably around four or five dozen species, all native to tropical Asia (Heslop-Harrison and Schwarzacher 2007). The genus is split into three sections that differ from one another in their number of chromosomes. The vast majority of the hundreds of domesticated banana cultivars, including all the varieties that make it to the U.S. or Europe, are hybrids or varietals of two wild species from the section Musa: Musa acuminata and M. balbisiana. The genomes (all of an organism’s genetic material) of particular hybrids are derived entirely or mostly from one species or the other. Domesticated varieties likely spontaneously arose in nature (first in Southeast Asia and later in the East African highlands) from multiple hybridization events and were identified by early farmers (Heslop-Harrison and Schwarzacher 2007). Another group of bananas, called Fei bananas, which have flavorful vitamin-A rich fruit, were independently domesticated, probably in New Guinea around 6500 years ago, from a species in the Australimusa section. And by “banana” here I refer to both the sweet “dessert” banana varieties that are usually consumed raw (like Cavendish), and the starchier “plantain” or “cooking” varieties that are usually cooked, often as a savory dish.

If you’ve never seen a banana plant, try to find one someday. They are impressive. Topping out at around 3m in height, some banana plants might look like trees, but trees they are not. Banana plants are very large herbs, the biggest in the world. Trees have persistent woody trunks and branches, but all aboveground parts of the banana plant die after fruit production and are not woody. After the aboveground plant portions die after fruiting, a new shoot arises from the perennial belowground stem, called the rhizome. The rhizome is the only true stem of the plant.

The sturdy aboveground trunk-like structure is a pseudostem, formed of overlapping leaf sheaths, developing leaves, and, eventually, the flower stalk (inflorescence), which grows up through the middle of the pseudostem. The banana leaf sheath is a structurally reinforced region of the petiole, which is the stalk that connects the photosynthetic leaf blade to the stem (rhizome, in this case). The strength of the pseudostem is impressive, given the massive size of banana leaves (useful for making packets for steaming everything from fish to tamales or an attractive serving dish accessory), which can top 2m in length and nearly a meter in width, and fruit clusters that can weigh over a hundred pounds.

With the exception of one group of bananas called “Fei” bananas, the inflorescence droops downward after emerging from the psuedostem. Large, often colorful bracts (modified leaves, also prominent in artichokes, asparagus, and bamboo shoots) enclose the developing whorls of flowers. The large outermost bracts lift open first and hang like awnings over the first whorl of flowers. Below them the main axis of the inflorescence (rachis) grows downward, supporting the bud, the pendulous cluster of overlapping bracts and flowers. Bracts and flowers open sequentially along the rachis. The first few flower whorls consist exclusively of “female” (pistillate) flowers, lacking pollen-bearing “male” anthers.

Each ovary of these flowers develops into a banana.  The number of female flower whorls depends on variety and growing conditions. Male flower whorls are the last to develop. Colloquially in the U.S., we give the name “bunch” to the cluster of half-a-dozenish bananas that we buy in the grocery store. Folks in the banana business, though, use “bunch” to refer to all the bananas growing on a rachis. Each whorl of fruit is somewhat morbidly referred to as a “hand,” and individual bananas are “fingers.”

The banana fruit itself is structurally pretty straightforward. Richard Spjut, the authority on fruit type classification, calls banana fruit a pepo, a type of berry with a tough rind. Botanically, a “berry” is a simple, fleshy fruit that develops from a single ovary from a single flower. Other pepos include squashes (Cucurbita sp., family Cucurbitaceae) and papaya (Carica sp., family Caricaceae). Like a squash, the banana rind consists of tissue from the mother banana plant that supported the flower (hypanthium) fused to the ovary wall. Internally, the banana ovary typically has 3 chambers, called locules, which are sometimes discernable in ripe Cavendish bananas.

The locules in wild bananas are almost entirely filled up with large seeds, which are padded with only a small amount of soft, edible tissue (endocarp, the innermost layer of the ripe ovary wall). The cultivated bananas all fail to develop seeds (see discussion below), and the shriveled ovules are visible as tiny dark specks in the middle of the banana amidst the prolific amount of sweet, fleshy endocarp (the part we eat) that occupies the entire locule. The bitter strings between the peel and the edible endocarp are vascular bundles, clusters of xylem and phloem conduits that transport water, nutrients, sugar and phytochemicals to the developing fruit. The vascular bundles usually pull away from the ripe fruit, leaving behind vertical ridges in the soft endocarp.

How we got stuck with the Cavendish and its current peril

Discomfort with the extraordinarily cheap price of bananas grown thousands of miles away is arguably justified. It took low pricing, aggressive corporate marketing and technological advances in fruit transport starting in the late nineteenth century to launch bananas to the top of the list of America’s most popular fruits (Koeppel 2007). The fast-growing popularity of bananas in the United States fueled the rapid conversion of Latin American rainforest to banana plantations on a vast scale. The high human and environmental cost of the plantations was often accepted by tropical government officials in the thrall of fruit corporations, a situation for which the phrase “Banana Republic” was coined (Koeppel 2007). These Latin American plantations still supply nearly all bananas sold in the U.S., and the banana export crop is economically important for the nations that produce it. Today Americans consume more bananas than oranges and apples combined. Most of those bananas are non-organically grown, and non-organic bananas are one of the most pesticide-intensive crops.

Globally, though, most bananas grown around the world don’t travel very far. There are at least five hundred varieties of edible seedless bananas grown around the tropics. Most bananas are produced for local or regional consumption and occupy a substantial fraction of daily calories consumed in many tropical regions. The bananas feeding the tropical population are mostly not the same varieties sent to the U.S. and Europe. Or, rather, I should say variety.  At least 95% of the bananas sold in the U.S. are largely genetically invariant clones of a single banana cultivar: the Cavendish. In 2010, Cavendish bananas comprised 40% of the world’s banana production, of which about 14% was shipped internationally. So, within banana-producing latitudes, about a quarter of bananas consumed are Cavendish. A few other dessert banana and plantain varieties are beginning to make inroads into American grocery stores, but my kid has only consumed Cavendish. And this apparently is a shame, as connoisseurs of the world’s edible bananas describe the Cavendish as relatively unremarkable in flavor, and it lacks the vitamin A found in some other varieties.

Bananas from the Gros Michel cultivar group, reportedly superior to the Cavendish group in flavor, dominated the international banana trade from the 1880s to the 1950s. Gros Michel plants were easy to grow, and the large fruits had tough skins that would withstand late 19^th-century shipping cargo holds, and the fruit had a long enough ripening time to guarantee that a profitable number would make it to market before becoming overripe (Koeppel 2007). In the 1960s Cavendish cultivars replaced Gros Michel as the banana of choice to ship to grocers in cold parts of the world after Fusarium wilt (or “Panama disease”) decimated Gros Michel plantations in Latin America. Like the genetically invariant potatoes that got wiped out in one fell pathogenic swoop during the Irish Potato Famine (see our nightshade post for some info on that), the massive scale of the Gros Michel plantations did nothing to protect the susceptible plants, all sterile clones produced from cuttings, from Fusarium wilt.

Cavendish cultivars are immune to the strain of Fusarium wilt (tropical race 1) that commercially destroyed Gros Michel, but they are susceptible to another strain (tropical race 4) that is currently making its way around the world. It’s just a matter of time, then, before Fusarium wilt race 4 makes its way to the gigantic Latin American monocultures of Cavendish banana clones. Random genetic mutations do arise even in these clonal populations, some of which produce noticeably useful new varieties (Heslop-Harrison and Schwarzacher 2007), so the vast banana plantations aren’t 100% genetically uniform, but they’re awfully close. Fusarium wilt cannot be controlled with known fungicides, and once it’s in the soil, Fusarium wilt is all but impossible to eradicate. Therefore, once a particular Fusarium wilt is entrenched in a banana field, cultivars sensitive to it will never again be commercially viable at that location. There are other diseases that affect bananas (hence the high pesticide use on banana plantations), but none of them is an absolute show stopper like Fusarium wilt.

Inferior to the Gros Michel though the Cavendish may be, given no or very little alternative in supermarkets, the banana has lost no popularity in the Cavendish era. We’re currently stuck with the Cavendish in the U.S. because, in the tropics, the Cavendish is also relatively easy to grow at the mega-plantation scale, and the banana industry has now standardized itself to the Cavendish. Many of the other hundreds of varieties grown around the world, though, are likely less susceptible to disease than the Cavendish and Gros Michel, so diversifying varieties grown in banana plantations for export may be part of the solution to mitigate the impact of disease and keeping the U.S. and Europe supplied with bananas. Diversifying farms, though, must be accompanied by banana crop improvement, through traditional breeding and/or genetic modification.

The trouble with banana breeding

You might ask, why not just use traditional plant breeding to mix up the banana gene pool to get disease resistance, good flavor, higher yields, marketability, tolerance to environmental stresses, and high nutrition value in a commercially viable set of bananas? After all, harnessing various combinations of disease resistance and edibility from the highly diverse potato lineages in the Andes saved the international potato market. The roadblock to doing this is seedlessness in edible bananas. Traditional plant breeding relies on sexual reproduction (pollinating flowers) to create novel genetic combinations, which are conveniently packaged in viable seeds. Domesticated bananas are all but entirely sterile and seedless. They produce fruit parthenocarpically (in the absence of fertilization or pollination, a phenomenon we’ve previously talked about in our posts about figs and persimmons) Cultivated bananas are propagated asexually (“vegetatively”). Farmers can do this by dividing up the rhizome into chunks, each of which can sprout a banana plant clone, or, more common these days on big farms, new clones can grow in a lab from a small amount of tissue (Heslop-Harrison and Schwarzacher 2007), a method that is less likely to spread diseases. Once established, a rhizome sends up multiple shoots (“suckers”) in succession, one of which is allowed to grow up to replace the previous shoot after the bananas are harvested.

The parthenocarpic edible banana cultivars are a bit genomically quirky, which contributes to their sterility (or near-sterility). Most of them are triploid, meaning that each of their somatic (non-gamete) cells has three copies of the ancestral genomes, instead of the usual two (Heslop-Harrison and Schwarzacher 2007, Ortiz and Swennen 2014). That is, M. acuminata and M. balbisiana are like humans in that their somatic cells have two copies of the genome, one from each parent. Plants are prone to occasionally making gametes with two (or more) genome copies, and very occasionally these unusual gametes successfully fuse with a normal gamete with just one genome copy, resulting in offspring with three copies. These triploid individuals rarely produce gametes that result in viable offspring, but it happens (Ortiz and Swennen 2014).

So it’s at best very difficult to mix up the genomes of the current set of domesticated bananas using traditional plant breeding because sexual reproduction is impossible to rare. Nonetheless, starting with wild species and the few seeds occasionally produced by hybrid cultivars, banana breeding is a thing (Heslop-Harrison and Schwarzacher 2007, Ortiz and Swennen 2014), and new insights into the genetic history of the M. acuminata–M. balbisiana hybridization events may provide guidance for future successful traditional banana breeding efforts. Given the protracted timeline of traditional banana breeding, though, using highly-targeted laboratory-mediated genetic modification to relatively quickly produce seedless bananas with desirable qualities may be a fairly uncontroversial application of this approach. Several research groups are using genetic modification to achieve various banana improvements, including overcoming susceptibility to Fusarium wilt and other diseases. One particularly clever and promising solution to the latter is to disable the mechanism in banana cells that Fusarium wilt uses to overcome the plant without affecting banana growth or fruit quality. Other groups have successfully used genetic engineering to enhance banana micronutrient and vitamin content, which could improve the nutrition status of millions of impoverished people for whom bananas and plantains are a staple food crop (Ortiz and Swennen 2014).

Researchers are rapidly accumulating information about the genetic basis of desirable traits in Musa, genetic diversity of wild and cultivated varieties, and the structure of the Musa genome. The genomes of both M. acuminata and M. balbisiana have now been sequenced. This genetic and genomic knowledge should expedite the production of better banana varieties by focusing traditional breeding efforts and increasing the options for genetic transformation (Heslop-Harrison and Schwarzacher 2007, Ortiz and Swennen 2014). In addition to preventing another major crash of the banana crop due to disease, if these banana improvement efforts are successful, they should also help reduce the voluminous chemical inputs inherent to modern banana farming and otherwise make life easier for banana farmers (Ortiz and Swennen 2014). This should help ameliorate some (but not all) of the eco-guilt associated with buying bananas in the temperate world. So the next time you toss a bunch of bananas in your grocery cart, spare a thought for the diligent folks trying to keep the world eating better and better bananas.

References

Heslop-Harrison, J.S.,  and T. Schwarzacher. 2007. Domestication, genomics, and the future for banana. Annals of Botany 100: 1073-1084.

Koeppel, Dan. 2007. Banana: The fate of the fruit that changed the world. Hudson Street Press.

Ortiz, R., and R. Swennen. 2014. From crossbreeding to biotechnology-facilitated improvement of banana and plantain. Biotechnology Advances 32: 158-169.

Many of the links in this post and much of the botanical information about the banana plant structure are from the fantastic website of the ProMusa banana research consortium. Special thanks to several ProMusa members who promptly provided useful answers to my banana questions.