Heads of state, agriculture and finance ministers and investors meet in Kigali from August 31 for the Africa Food Systems Forum, the continent's largest agricultural gathering. They arrive with an opportunity: to build a food system that holds up against what is coming – climate change, biodiversity loss and declining soil health, alongside our dependence on imported food, fuel and fertiliser.

After the closure of the Strait of Hormuz, a waterway carrying around a third of the world's seaborne fertiliser trade, urea doubled to more than $850 a tonne by April. Prices have since eased, but the World Bank still expects fertiliser to average over 30 per cent higher across 2026, with relief arriving only in 2027.

Science tells us where a system is weak long before it breaks. It often takes a crisis for that to sink in, and this year the lesson arrived painfully. African nations cannot afford more crises of that kind.

Governments did the right thing and protected farmers, largely through subsidies, even as aid budgets were shrinking. But support of this kind is simple to extend and hard to withdraw, so each shock leaves the commitment larger and the room for manoeuvre smaller.

Farmers need that support, and they need mineral fertiliser, nothing replaces it quickly at the scale Africa requires. But a subsidy that rises with an import price buys protection without changing what makes the farm vulnerable. Our research points the same way each season: where soils are degraded or acidified, a large share of applied nitrogen is never taken up by the crop, so the farmer pays for the whole bag and harvests a fraction of it. The subsidy absorbs the price. The soil still loses the nutrient.

The most useful debate in Kigali is how to spend the same money on making farms less vulnerable in the first place.

The obstacle has never been the science. Most delegates in Kigali could list the alternatives to an input subsidy without notes. What has been missing is the room to act on them: input support is visible, popular and long-established, so in an ordinary year redesigning it means asking farmers to accept change before they see the benefit.

This year changes the terms. Governments are paying far more for the same protection, and production has not risen to match. The choice is no longer reform or the status quo. It is paying more each year for the same result or spending the same budget on what each farmer's land actually needs. Farmers keep their support. The money buys more.

The audience has changed too. A decade of advocacy struggled to make the link between a distant energy market and a domestic food bill feel real to a treasury; six months of disruption did it. Kigali is one of the few gatherings where finance and agriculture ministries sit in the same room.

The institutional timing is equally short. The Kampala CAADP Declaration, which moves Africa from agriculture-led growth to a whole agrifood systems approach, came into force on January 1. The Nairobi Declaration commits governments to reversing degradation on 30 per cent of the continent's degraded soils by 2034. Africa is in year one of both, the year they become national plans and budgets and Kigali is the last continental gathering before those budgets close. Three shifts would make the difference.

Healthy soil holds nutrients where roots can reach them, and precision does much of the rest: applying the right amount in the right place raises the share of nitrogen a crop takes up. Alliance advisory service has helped 72,000 Ethiopian farmers get more from the same bag.

Crop choice matters too. Legumes fix nitrogen from the air and leave it for whatever crop follows: across sub-Saharan Africa, maize after a legume consistently outyields maize after maize, and in trials without nitrogen fertiliser the returns more than doubled. Millet, sorghum and pulses thrive on little or no nitrogen fertiliser, which keeps their cost base at home.

Through the Pan-Africa Bean Research Alliance, which we host, more than 650 improved bean varieties have reached over 40 million farmers in 31 countries, while community seedbanks we support in Kenya and Uganda safeguard more than 600 varieties across 72 indigenous crop and tree species.

These approaches cut emissions and create jobs as well. The core argument is financial: costs set at home, around local and diversified crops rather than abroad.

Public money shapes farming twice over, through what governments subsidise, and through what they buy.

Subsidies work best when they follow the farmer rather than the product, widening what the same money buys: soil amendments, lime, better seed. Zambia has already done it. Its electronic voucher, now running in all 116 districts, lets more than a million farmers choose inputs suited to their own soils instead of taking a uniform package. Farmers keep their support; the money buys what each field needs.

Domestically-based, nationally-funded procurement is the other half, and Africa has barely used it. School meals, hospitals and public canteens create steady demand, and anchoring it to diverse local production builds markets for crops that need fewer imported inputs while improving the diets of children and their communities.

Nigeria's Dangote refinery has been exporting fuel and urea to Côte d'Ivoire, Cameroon, Tanzania, Ghana and Togo, softening the disruption for its neighbours. Morocco supplies more than half of Africa's phosphate, and gas reserves in Nigeria, Mozambique, Tanzania and Senegal could supply more. Regional production, blending and trade turn one supply route into several.

The fastest gain comes after the harvest. Africa loses a substantial share of what it grows before it reaches market, which is why halving post-harvest losses is written into the Kampala plan. Better storage, roads and processing recover food already paid for in fertiliser, labour and water.

In 2025 our scientists traced the road from a warming climate to conflict in Nigeria and found that it runs through the food system. Heat reduces harvests, incomes fall, and children are the first to show it, in rising rates of wasting, the acute malnutrition measured when a child's weight falls too low for their height. Where wasting rose, violence became more likely to follow. It proved to be an early warning signal for conflict that had yet to begin.

This year's price shock works the same way. When food costs rise, households buy cheaper and starchier food, and diets narrow. Prices recover, but a child carries the developmental cost for life. Four of the ten countries holding two-thirds of the world's acute hunger are African, and a working food system is a foundation for stability.

The choice in Kigali

Africa holds the assets, diverse landscapes and carbon sinks, a young and growing population, expanding regional trade, and more of the world's uncultivated arable land than any other continent. It also has one in five people undernourished and soils eroding beneath the crops meant to feed them.

Prices will settle, and that is the risk. Once they do, the case becomes abstract again, the system stays as it is, and the next disruption costs more than this one. The frameworks are in force and the evidence is clear. What is open, for a few more months, is the space to act and Kigali is where it all comes together.

The writer is Director General of Alliance of Bioversity International and CIAT, part of CGIAR, a global research partnership for a food-secure future. Its Africa hub is in Nairobi, with offices including Bukavu, Kigali and Addis Ababa.