---

ON THE SHORES of Lake Naivasha, 90 kilometres north-west of Nairobi, the greenhouses begin almost at the water’s edge. Acre upon acre of translucent polythene stretches inland from the lakeside, encasing roses so immaculate they seem almost artificial. Within 48 hours of cutting, those blooms will be changing hands at the flower auction in Amsterdam, or wilting gently in a vase in Frankfurt or London. What remains behind at the lake is rather less pretty.

The water level of Lake Naivasha has fallen by roughly four metres since commercial flower farming took root in the 1980s. More than 60 farms line its shores and account for a substantial share of all water withdrawn from the lake. The once-crystalline water has turned turbid; the lake’s fish stocks have dwindled. The bloom of water hyacinth, which thrives on nutrient-rich runoff from fertiliser-laden farm effluent, has periodically choked the shallows. In December 2009 the lake hit its lowest recorded level in living memory.

This is not merely a Kenyan story. From the high plateau surrounding Bogotá to the Rift Valley of Ethiopia, the global cut-flower industry has planted itself in developing countries with reliable sunshine, cheap labour, and—crucially—abundant fresh water. As the industry has boomed over the past four decades, that water has grown measurably scarcer. The question that governments, development economists, and guilt-stricken European consumers struggle to answer is whether the economic gains are worth the hydrological cost.

---

The Thirst of a Single Rose

Start with the numbers. A single rose stem requires somewhere between seven and 13 litres of water during its production life, according to research by Mekonnen, Hoekstra, and Becht. The figure sounds manageable until it is scaled to industrial production. Ethiopia alone was estimated to consume 60,000 litres of water per hectare per day at peak growing periods. In Colombia, a hectare of flowers draws roughly 150,000 litres per week. The total virtual water embodied in cut flowers exported from the Lake Naivasha basin amounted to 16 million cubic metres per year between 1996 and 2005—water that left Kenya embedded in roses destined for European windowsills.

Greenhouses are particularly water-hungry. They demand irrigation, pesticide application, cooling systems, root-zone preparation, and constant cleaning. Unlike open-field crops, which at least return some moisture to the local water table through percolation, greenhouse production frequently extracts groundwater through boreholes and returns contaminated effluent. The net effect on local hydrology is almost uniformly negative.

The industry concentrates in regions that are already water-stressed because those same climatic conditions—reliable equatorial sunlight, stable temperatures, minimal frost—produce the strongest blooms. This is the central irony of the cut-flower trade: the places best suited to growing flowers are often the places that can least afford to spare the water. Research published in BioScience in 2025 found that many African countries currently leading in floriculture show some of the highest potential for local conflict risk associated with water stress globally—meaning the industry’s water demands carry implications not just for ecosystems, but for political stability.

---

Kenya: The Lakeside Bargain

Kenya is the world’s fourth-largest exporter of cut flowers and Africa’s largest, a distinction that has reshaped the economy of the Rift Valley in ways both remarkable and troubling. The industry generates well over $800 million in foreign exchange annually, representing roughly a quarter of Kenya’s agricultural export earnings. Flowers have made Kenya’s horticultural sector the single largest source of foreign currency after tea. More than two million Kenyans depend on the industry, directly or indirectly, for their livelihoods. In Kenya, women make up 60 to 70 percent of the flower workforce—providing formal employment for women who might otherwise have few options.

The Naivasha basin has been the engine of this success and its primary victim. Landsat satellite images from 1973 onwards document the lake’s gradual retreat as greenhouse coverage expanded along its shores. Local fishermen, once visible, have been progressively displaced. Hippopotamus populations, whose habitat has shrunk and been converted to farmland, have declined sharply. The farms pump directly from the lake and from boreholes, returning pesticide-laced wastewater in a cycle that has visibly degraded water quality.

Even so, the economic argument for the industry remains powerful—and the environmental record has, belatedly, begun to improve. The Kenya Flower Council has tightened its Code of Practice, requiring member farms to meet standards on water abstraction, chemical use, and worker welfare. Drip irrigation systems, which can reduce water consumption by 50 to 75 percent compared to overhead sprinklers, have spread. Water resource user associations, supported by organisations including WWF Kenya, have brought hundreds of abstraction points into regulatory compliance. Whether these gains can outpace the industry’s continued expansion remains uncertain—but the trajectory, at least, has turned.

---

Ethiopia: Lessons Unlearned

Ethiopia’s entry into the global flower market is more recent and more dramatic. From almost nothing in the early 2000s, the country became Africa’s second-largest flower exporter and the world’s fifth-largest, trailing only the Netherlands, Colombia, Ecuador, and Kenya. By 2020, floriculture had become the second most important export commodity after coffee, accounting for more than 14 percent of the nation’s total export earnings. Between 1999 and 2020, the industry earned more than $2.8 billion for the country. Some 84 active flower farms now operate, creating tens of thousands of formal-sector jobs—predominantly for women—in an economy where such work is scarce.

The environmental record is considerably less admirable. Research published in Frontiers in Environmental Science found that water-use efficiency on Ethiopian flower farms is generally very low, with major concerns regarding resource depletion and persistent conflicts over water rights. Water levels in the Rift Valley lakes and rivers—the primary farming region—have declined measurably. In the Sululta district north of Addis Ababa, communities reported that the Aleltu River began drying during the dry season after floriculture firms sank multiple boreholes nearby. Land that had been designated by the government for youth irrigation activities was fenced off by flower companies in 2019.

These conflicts are not abstract. Land and water previously accessible to subsistence and smallholder farmers have been enclosed by commercial operations, displacing families from riverside resources they depended on for drinking water and small-scale cultivation. The Ethiopian government has prioritised the industry’s expansion—floriculture fits neatly into an export-led development strategy—while enforcement of environmental standards has remained weak.

The Ethiopian Horticulture Producers and Exporters Association has made genuine efforts: 36 wastewater treatment plants have been constructed, treating close to 60 million cubic metres of used water. Ethiopia now has one of the highest concentrations of constructed wetland treatment systems in the region. These are meaningful achievements. But critics note that treatment capacity lags behind the volume of water being extracted, that monitoring by government agencies is inadequate, and that the smallholder farmers absorbing the costs of industry expansion have no seat at the table when policy is set.

---

Colombia: Old Problems, Familiar Patterns

Colombia’s Sabana de Bogotá, the high plateau surrounding the capital, is one of the oldest flower-farming regions in the developing world. Colombian floriculture dates to the 1960s, and the country became the world’s second-largest flower exporter by the 1990s—a position it holds today. The industry exported an estimated 660 million stems in 2020 alone. In 2013, some 8,000 hectares were under cultivation, 7,000 of them under greenhouse cover, with 73 percent of farms concentrated on the Bogotá savanna.

Water use is intensive and long-documented. Flower farms draw from aquifers and rivers already under pressure from the expanding Bogotá metropolitan area. The industry creates what researchers call an economic lock-in: in towns on the Sabana, so much employment depends on floriculture that communities become dependent on it even when it causes harm. Short-term contracts have proliferated, making it harder for workers—among them roughly 100,000 women, many of them single mothers—to organise without risking their livelihoods.

The news is not uniformly grim. Colombia’s flower sector has, over decades, developed more sophisticated environmental management than most. Today, more than 60 percent of the water used in Colombian flower production comes from harvested rainwater, stored in reservoirs that have themselves become habitats for aquatic birds. Many farms have established closed-loop irrigation systems that capture and recycle runoff, reducing freshwater consumption by 40 to 60 percent while preventing nutrient-laden water from entering natural waterways. The industry’s Florverde Sustainable Flowers certification, developed by Asocolflores, provides a credible domestic standard. Colombia’s path suggests that with time, regulation, and commercial pressure, the sector can be made substantially cleaner—though the legacy damage to the Sabana’s soils and watershed remains.

---

Ecuador: A Smaller Stage, Larger Proportions

Ecuador’s flower industry is concentrated in the highlands around Quito—particularly in the provinces of Pichincha and Cotopaxi—where the country produces some of the world’s most prized long-stemmed roses. The industry has made Ecuador the world’s third-largest flower exporter, and roses make up the dominant share of its output.

The fragile páramo ecosystems that supply fresh water to Quito and other highland cities are the same environments in which flower cultivation has expanded. These high-altitude wetlands act as vast sponges, absorbing rain and releasing it gradually into rivers below. They have no substitute. Pressures on them from agriculture, mining, and urban expansion are substantial; flower farming has historically pushed toward these sensitive watersheds.

Ecuador also illustrates the limits of worker protection in export-led agriculture. Fairtrade research found that workers in Ecuador and Ethiopia are paid between 50 and 65 percent of a living wage; problems with freedom of association are common; and Ecuador has a documented history of union suppression in the sector. Researchers have found that children living near floricultural greenhouses—where pesticides were applied intensively in the weeks before Valentine’s Day or Mother’s Day—showed altered short-term brain activity, suggesting pesticide exposure through contaminated clothing and tools brought home by parents. They are currently investigating whether agricultural chemicals are contaminating water downstream from the farms.

---

The Economic Calculus

The case for tolerating the environmental costs of floriculture rests on a series of economic arguments, some stronger than others.

The strongest is employment. Flower farms in Kenya, Ethiopia, Colombia, and Ecuador create formal-sector jobs—with wages, contracts, and at least nominal labour protections—in economies where such employment is scarce. The jobs disproportionately employ women, providing income and independence in societies where both are hard to come by. In Colombia, flower production is the largest employer of women in rural areas. In Kenya and Ethiopia, it has lifted significant numbers of households out of agricultural subsistence.

The foreign exchange argument is similarly real. These are small, open economies that depend on hard-currency earnings to service debt, import fuel, and finance public investment. When Kenya earns more than $800 million from flowers, or when Ethiopia earns more than 14 percent of its export revenues from a single new crop, these are not marginal contributions. For governments trying to industrialise, floriculture offers one of the fastest routes from subsistence agriculture to modern manufacturing-style production, with short feedback loops and global market integration.

The weaker part of the economic case concerns how the gains are distributed. Where the benefits flow primarily to large commercial operators—many of them foreign-owned—while the costs fall on local communities who lose access to water, land, and clean rivers, the development rationale becomes harder to defend. Research in Ethiopia found that smallholder farmers who had previously cultivated edible crops for domestic food security were displaced into daily wage labour, reducing both their incomes and the region’s food self-sufficiency. The floriculture economy, in those cases, creates new employment while destroying older, more resilient livelihoods.

---

The Virtual Water Problem

There is a specific way in which the flower trade makes water scarcity worse that is sometimes missed in aggregate economic assessments: virtual water export.

When a country exports flowers, it exports the water used to produce them. That water does not return. In the case of Kenya—a country that sits in an arid and semi-arid belt where per-capita freshwater availability is below the global stress threshold—16 million cubic metres of virtual water left the Lake Naivasha basin every year embedded in cut flowers during the period studied. That is water abstracted from a stressed ecosystem and dispatched to wealthy consumers who live in countries where water is, by comparison, abundant. From a hydrological perspective, the global trade in cut flowers is a mechanism for transferring scarce water from poor, dry countries to rich, wet ones—a form of resource extraction whose environmental costs are borne locally while the aesthetic benefits are enjoyed elsewhere.

This does not mean the trade is irrational for the producing country. Water has no export value if it stays in the lake; flowers convert it into foreign exchange. But it does suggest that the environmental accounting used to assess the industry’s sustainability is often incomplete. The question is not merely whether water is being used efficiently on a given farm, but whether the total extraction is compatible with the long-term viability of the ecosystem from which it is drawn.

---

A Path Through the Thicket

The question of whether flower farming in developing countries represents good development policy does not admit a clean answer. The honest position is that it depends—on how farms are managed, on how revenue is distributed, on how seriously governments enforce environmental rules, and on whether the ecosystems being drawn upon can sustain the extraction over time.

The worst cases—Lake Naivasha at its most degraded, the Aleltu River running dry in Sululta, Ecuadorian children exposed to pesticide contamination—represent genuine failures of governance, where economic incentives outran environmental safeguards and where communities without political voice absorbed costs they had no power to refuse. These cases are not representative of the entire industry, but they are common enough to constitute a pattern.

The better cases—Colombia’s rainwater harvesting reservoirs, Kenya’s improving abstraction compliance, Ethiopia’s constructed wetlands—demonstrate that the industry is capable of managing itself more responsibly when pushed by certification schemes, consumer pressure, and effective regulation. The technologies exist: drip irrigation can cut water use by 50 to 75 percent; hydroponics can cut it by up to 90 percent; closed-loop water recycling systems are commercially viable and spreading. The constraint is not technical. It is political.

For governments in producing countries, the temptation to prioritise export revenues over environmental enforcement is understandable, particularly when foreign exchange is scarce and unemployment is high. For European and American consumers, the temptation to regard a bunch of roses as a simple transaction—beauty exchanged for money—is equally understandable. Both temptations lead to the same place: more flowers, less water, and communities absorbing the difference.

The fair trade and sustainability certification schemes that have proliferated across the industry address part of the problem. They are imperfect instruments—standards vary, auditing is incomplete, and the poorest farmers are least able to bear the certification costs—but they represent a genuine mechanism for internalising environmental costs that would otherwise remain invisible. Consumers in wealthy countries who choose certified flowers are making a small but non-trivial difference.

Ultimately, though, the solution lies in governance rather than in consumer choices. Countries that have developed credible, enforced environmental regulations—and that have given communities affected by water extraction a meaningful voice in how resources are allocated—have produced better outcomes. Countries that have treated floriculture purely as an export-revenue machine have produced worse ones. The blooms may be beautiful. The policy question is whether they are grown in water that anyone can afford to spare.

---

Florist