A grain that ripens in six weeks and grows in sand. Milling it wastes 40 per cent.

Fonio is a cereal with a very short list of requirements. It will grow in dry sandy soil that will not carry much else, and it can be ready to harvest in six to eight weeks.

Its lineage runs back more than 5,000 years, which may make it the oldest cultivated cereal in Africa. West African farmers grow around 700,000 tonnes of it a year.

Most people outside the region have never eaten it.

Why the speed matters

A six-to-eight-week crop is not just convenient. In the Sahel it fills a specific and dangerous gap.

The period before the main harvest comes in is when household stores run lowest — the season sometimes called the hungry gap. A cereal that can be sown and harvested inside two months lands in exactly that window.

Add the soil tolerance and it is a crop suited to precisely the conditions that are spreading: hotter, drier, thinner ground.

What is in it

Fonio is a good source of protein, and specifically rich in sulphur-containing amino acids — methionine and cystine, which are the ones most cereals are short of. It supplies iron and zinc, has a low glycaemic index, and contains no gluten.

That combination is why it appears on shelves in Europe and North America as a health food, at prices that have very little to do with what a farmer in Mopti receives for it.

The bottleneck is a machine

The obstacle to fonio is not the field. It is what happens after the field.

The grains are extremely small, and hulling them is difficult. According to reporting by the Global Landscapes Forum, current milling technology wastes up to 40 per cent of the harvest and produces a product that fails international quality standards.

That figure is the whole economics of the crop. Two out of every five grains a farmer grows are lost between harvest and market, and what survives cannot reliably be exported.

Traditionally the work was done by hand, mostly by women, and it is laborious enough to cap how much any household plants. Researchers at Cornell's Climate-Resilient Farming Systems programme and at World Agroforestry have worked on the problem alongside groups including 3-A Sahel and Bioversity International.

Who decides what gets fixed

Rice, wheat and maize have had a century of engineering attention. Threshers, hullers, dryers and mills for them are mature technology, refined across decades and continents.

Fonio has had a fraction of that, for the same reason enset has: the crops that receive agricultural research money are the crops eaten by the people who allocate it.

A 40 per cent processing loss is not a natural property of the grain. It is a solvable engineering problem that has not been solved, because of who eats it.