The Fertiliser That Arrives On The Wind

AEvery year the world's deserts lift something like two billion tonnes of mineral particles into the atmosphere, and the Sahara alone exports about 180 million tonnes of that load westward across the Atlantic. Roughly 28 million tonnes of it settles over the Amazon basin. Because desert dust carries iron and phosphorus, two elements in short supply both in the surface waters of the open ocean and in the ancient soils of the tropics, it has been described since the early 1990s as a fertiliser distributed by the wind. The image is attractive and, in outline, correct. What it conceals is that the journey from a dry lake bed to a living cell passes through three separate stages — lifting, transport and uptake — and that a large fraction of the material is lost, or rendered useless, at each one. The tonnage that leaves a desert is therefore the least informative number in the sequence, though it is the one most often quoted.

BIngrid Solberg, an atmospheric chemist at the Kestrel Institute, has spent much of her career on the second of those stages. Iron leaves the desert locked in oxides that seawater barely touches: of the iron in freshly lifted dust, only one or two per cent will dissolve. Acids formed from industrial and volcanic emissions attack the surfaces of the particles while the dust is still aloft, and this weathering releases more. Between 2011 and 2019 her group sampled 63 plumes over the Atlantic and the Southern Ocean and found a median soluble fraction of 3.4 per cent, a figure that climbed above 7 per cent only in plumes whose back-trajectories had carried them across industrial regions. The most nourishing dust, in other words, is the dust that has been through the dirtiest air. Solberg is careful to present this as an objection to a habit of measurement rather than to the idea that dust feeds anything at all.

CThe third stage is the hardest to observe. Solberg's group paired each deposition event with satellite records of chlorophyll, the pigment that stands in for the mass of drifting plants. In 41 of the 63 events the increase over the following ten days stayed under 15 per cent, which is within the range that the same patch of water produces of its own accord. The unmistakable responses were confined to what oceanographers call high-nutrient, low-chlorophyll water — the Southern Ocean, the sub-arctic Pacific and a belt along the equator — where nitrate is plentiful and iron is the single element in short supply. Elsewhere, adding iron changes little, because iron was not what the plants were waiting for. The lesson is not that dust fails to fertilise the sea, but that it fertilises a minority of it, and only where the local shortage happens to match what the dust is carrying.

DOn land the accounting has been done most fully by Tomás Iriarte, who directs the Marabá Centre for Basin Studies. Amazonian soils are among the oldest on earth and have been leached by rain for millions of years, so the phosphorus that the rivers carry out to sea must be replaced from somewhere. About half the dust reaching the basin originates in a single place, the Bodélé Depression in Chad, the bed of a lake that dried out some 5,000 years ago and whose sediment is rich in the silica shells of diatoms. Taking a phosphorus content of 0.08 per cent, Iriarte's group put the annual delivery at roughly 22,000 tonnes, close to the quantity the rivers remove. He insists on two qualifications. The figure is an average for the basin as a whole, and the north-eastern margin receives several times as much per hectare as the western interior; and the phosphorus content itself rests on a small number of samples.

EThe argument has since been carried into a field neither researcher chose. If dust feeds the ocean, iron might be spread on it deliberately, and in 2019 a consortium released 12 tonnes of iron sulphate over 300 square kilometres of the Southern Ocean. Chlorophyll tripled within twelve days. The sediment traps moored beneath the patch, however, caught only a modest addition to the flux of sinking material, most of the new growth having been eaten or having decayed in the upper layers. Iriarte observes, awkwardly for his own emphasis on dust as a supplier of nutrients, that a bloom and a store of carbon are not the same thing, and adds that his phosphorus figure is routinely quoted as though the element arrived ready for roots, when much of it binds to iron and aluminium oxides in the soil before any plant reaches it. Solberg replies that this is precisely her case moved onto land: the most conspicuous part of a result is the least informative part of it.

FWhat the two accounts amount to is a pair of conditions rather than a verdict. Where the receiving system is short of exactly the element the dust is carrying, and where the mineral has been weathered enough in transit to give that element up, deposition produces an effect large enough to measure; where either condition fails, tonnage predicts nothing. Neither researcher still holds the position each was assigned a decade ago, when the exchange was reported as a quarrel over whether dust mattered. The question that stays open is whether the carbon fixed after a dust-fed bloom sinks past the depth at which winter mixing would return it to the surface, and it will not be settled quickly, because sediment traps are moored in too few places and record for periods too short to catch events that arrive on the wind.