After the Flood, the Forest

On planting bananas in the warming Gulf Coast

Written by Hadley Cels


There is a narrow band along the coast of the Gulf of Mexico in which banana cultivation is possible. And it’s growing, moving northward with climate change, despite occasional havoc by arctic vortices. The Gulf South region of the US is becoming subtropical. At some point in our lifetimes, large parts of Texas, Louisiana, Mississippi, Alabama, and Georgia will experience freezing temperatures for the last time.

The consequences for plant communities will be drastic. Even in 2011, a NASA computer model predicted that by 2100, 40% of land ecosystems would shift “from one major ecological community type—such as forest, grassland, or tundra—toward another,” while essentially all land ecosystems will undergo significant changes in plant communities. As models have consistently underestimated the rate of change, the catastrophes that populate the imagined “2100” of climate futurity appear closer every year.

Climate change will shift ecosystems to the right on this graph, with increasing temperature, and either up or down according to its impacts on precipitation. But, critically, the precipitation in a given place is not determined solely by global weather patterns. As is well known, the high rainfall in the Amazon basin is due to evapotranspiration from the exhalations of the rainforest. In other words, the difference between a subtropical desert and a subtropical rainforest can be nothing more than the trees themselves.


Plants do migrate, but at varied rates. Without assistance, the slow ones, particularly trees, will disappear instead. In this catastrophic era, the category of invasive species will be less relevant than that of refugee species. The US Gulf Coast is badly positioned for natural migration. Compared to those emerging subtropics of the Anthropocene which are contiguous with existing subtropical forests (such as parts of the Chinese provinces of Jiangsu, Anhui, Henan, Shaanxi, and Gansu), the region’s most diverse subtropical forests grow on islands in the Caribbean or on the other side of a border wall and an expanding desert, facing their own disruptions by drought & storms.

We can imagine future solidarities that might develop between regions. People in the mountains of a Caribbean island rush to gather seeds and cuttings from the junque before the season of mudslides and hurricanes begins, sending them in care of migrants who sail past abandoned deepwater platforms in the Gulf and are welcomed at the shore by the keepers of another doomed forest. Mangroves sinking beneath the waves in the Yucatán go to the salt-sick ancient cypress swamps of the Atchafalaya Basin of Louisiana. Cypress seedlings are sent up the Mississippi to populate former cornfields in its expanded floodplain.

The metric of invasiveness may be increasingly useless in this epoch, but we still have a certain moral responsibility toward any living ecosystem. As with ourselves, the fact that the cypress are doomed does not make it okay to cut them down a day sooner. At the same time, so much destruction has already occurred that migrating species need not displace any ancient forests or wild ecosystems. There are about half a million acres of sugarcane in Louisiana alone (doubled from 265,000 in 1981), averaging 1400 acres per plantation. Nothing is more invasive than a sugarcane plantation—not because sugarcane originated in Africa, but because biodiversity is so low in a sugarcane field that it’s closer to factory than ecosystem. Invasiveness is best defined not by a species’ geographic origin, but by its behavior and impact in an ecosystem. Native species can become invasive in disturbed ecosystems under certain conditions. The pine bark beetles destroying forests in the western United States are not foreigners; they are locals exploiting disturbances wrought by mild winters.

In any case, the dream of assisted migration of entire subtropical ecosystems remains far away, not only blocked by borders but obscured by our ignorance of these systems. Even more important than a deepening of ecological study will be a widening of what is already known. It is hard to imagine a society in which only a tiny percentage of the population has any personal connection to plants or the soil succeeding in such a project. The process of planting the forest is also the process of becoming the people who live in the forest.

Here in so-called Louisiana, we have participated in motions in this direction. We share two of these efforts, one rural and one urban, to invite collaboration and elaboration.

To be clear, neither project holds subtropicalization as a goal to be worked toward, but simply as an imminent shift in the terrain of our struggle to live well and in good relation to the Earth and those we share it with. We are planting many temperate native trees as well, like cypress, tupelo, live oak, and sycamore. Our task is not to reconstruct the ancient forest nor to "design" the forests of the next century, but rather to support both endemic and migratory species as they adopt their own strategies for navigating this shift. The thriving forests of this century will be complex communities including migrating plants and endemic plants, each evolving to meet an ever-changing situation which includes human habitation.

Our rural example is situated on a dozen acres surrounded by agro-industrial wastelands: petrochemical infrastructure crisscrossing commercial sugarcane, rice, and crawfish production. It is an effort to transform a monoculture field into a multi-layered food forest, free nursery, and propagation hub for bananas, fruit trees and other useful plants.

Now is probably a good time to mention that banana trees are not trees, botanically speaking. They’re the world’s largest non-woody plant, basically a big perennial grass. This is one reason why bananas are useful in this transition. Since they sprout from the root even if they’re mowed over or the aerial parts die from frost, and because each stalk dies after fruiting anyway, an arctic vortex or hurricane-force winds doesn't represent a major setback. They grow fast, able to either shade out grasses or reach tree canopy height to access sun in a forest. Ecologically speaking, we hope they will act as a mid-successional species in the transition from either temperate forest or grassland to subtropical forest.

Each stalk produces fruit after 10-15 months without a hard freeze. In other words, the narrow band on the map where it’s easy to grow bananas is accompanied by a much larger area in which they grow well but don’t fruit. This is another reason why bananas are an ideal messenger species for the coming subtropical forest; in the hills of Georgia or the Piney Woods of east Texas and western Louisiana, banana plants can thrive and multiply but don’t fruit due to freezing weather. Increasingly popular as ornamentals, their eventual fruits will bear undeniable truths about climate change.

Destitute Dole

Cloning a banana plant is less of an operation than it sounds like. The banana is busy cloning itself without our help. Left to its devices, one banana plant will form a circular clump, spreading year after year. Baby bananas plants are called pups. A sharp shovel is used to separate them from the larger clump. As long as one or two roots stay connected to the pup, success rates are very high. A little math: start with ten bananas. Take five clones or "pups" from each tree per year. At the end of year five, you’ve got…77,760 banana trees.

It’s worth asking, before anyone tries to grow seventy-seven thousand banana trees: where will they be planted? The simplest option would be to find about eighty acres somewhere and plant as many as possible. Commercial banana plantations average 800-1000 plants per acre. We could fit all seventy-seven thousand on just five percent the land area of one average sugarcane plantation. To explain why we won't be doing that—and to understand the system we hope to desert—let's take a brief look at life under the regime of a modern banana plantation.

In an article in Feral Atlas, Alyssa Paredes, author of Plantation Peripheries: The Multiple Makings of Asia’s Banana Republic, describes visiting a woman living adjacent to a multinational-owned banana operation in the Philippines, in Mindanao: “Geronima had built a hiding place on the bottom level for her and her children to run to when the crop dusters came, but she found that it provided little respite. In the middle of our meal, she disappeared into the garden time and time again, picking up leaves covered in white spots as proof that chemical drift had invaded her property. It made the moringa plants curl, the cacao trees stop bearing fruit, and the chicken in the yard drop dead, she mentioned—midway through our lunch of rice and tinola, a gingery soup made with, well, chicken and moringa.”

The chemical mixture is always changing as growers lose an arms race with fungal evolution. As such, the exact composition of the poison dropped from a particular airplane, drifting onto one’s home, is unknowable. From Feral Atlas again: “Filipinos living in the vicinity of banana plantations do not know the chemicals by name but instead refer to them only by color or with the generic term hilo or ‘poison.’”

The commercial banana plantation sows the seeds of its own destitution more blatantly than most institutions today, spurring the evolution of the fungal pathogens which threaten the entire industry. Paredes explains: ”Empowered by the conventions of plantation agriculture, Sigatoka’s causal pathogen, Mycosphaerella fijiensis Morelet, becomes a formidable foe. The densely planted, highly susceptible Cavendish variety provides it with a convenient setting to hop from one host to another, while strong wind currents over the plantation’s manicured, low-lying canopy disperse its spores over long distances."

Pathogens thrive in a banana plantation for the same reasons they spread in prisons. And as with prisons, this is not our central critique. Prisons which are built to prevent the spread of disease are not what we want. Similarly, even if a universal anti-fungal treatment were developed tomorrow, we have other reasons to avoid recreating the industrial model.

Monoculture is only efficient if your goal is to produce as much fruit with as few people as possible, centralizing production in order to control the harvest. Our goal is different: to produce as many human-banana-plant interactions as possible and decentralize production so that it's within everyone's reach and cannot be controlled.

As such, many of the fruit trees grown at the rural project are being sent to New Orleans to be planted along sidewalks, in lawns, churchyards, and empty lots. This effort is coordinated by Lobelia Commons, an open collective of gardeners which formed early in the pandemic. They deliver free vegetable seedlings, build micro-nursery stands to distribute free plants, help people grow edible mushrooms, and generally try to build a living food commons in the city.

The Flood

Part of the joy of planting a fruit tree is knowing that the tree may outlive you, providing fruit for many generations. In New Orleans, we content ourselves with the other joy of planting a fruit tree—the part about eating and sharing the fruit during your lifetime. If anything, the looming exodus from the city is motivation to learn collectively how to live and eat well in this region. When the waters finally come to claim New Orleans, the resulting migration will carry with it whatever practices and ideas of communal life are widespread in the city, dissolving the rural-urban divide and determining whether the wider region's process of subtropicalization leads to forests and grasslands which are lush, diverse and densely inhabited, home to a thriving people, or if it will lead to deserts and wastelands, a deeper alienation of people from the land and its continued domination by extractive agriculture and industry.

In the meantime, we are leaving the planting of trees that only fruit after 10-15 years to our friends at higher elevations, and we are propagating banana plants, which fruit in a year or two.