Four Kilometres Down

On the abyssal plain of the eastern Pacific, four to six kilometres beneath the surface, the seafloor is scattered with dark lumps roughly the size and shape of potatoes. They lie loose on the sediment, sometimes half-buried, in densities of ten to thirty kilograms per square metre across an area of several million square kilometres.

They are polymetallic nodules, and they are metal ore of exceptional grade: manganese, nickel, copper and cobalt, precipitated from seawater around a nucleus — a shark's tooth, a fragment of shell — at a rate of a few millimetres per million years. They are the slowest-growing geological formation known. On any timescale relevant to human institutions, they are not renewable.

The question of whether to collect them is now the most consequential unresolved environmental decision in international law, and it is genuinely difficult. What follows attempts to state each case at its strongest.

The case for collection

The metals are required, and they must come from somewhere. Electrification demands nickel, copper and manganese in quantities that current supply cannot meet. Refusing seabed extraction does not reduce demand; it relocates supply. The relevant comparison is therefore not between mining and not mining, but between mining in one place and mining in another.

That comparison is not obviously favourable to land. Terrestrial extraction of the same metals requires removing overburden, processing ore of far lower grade, constructing tailings dams that occasionally fail catastrophically, and operating in regions where a substantial share of global cobalt is produced under conditions that include artisanal workings with documented child labour. Nodules require no blasting, no overburden removal and no tailings dam, and they contain four payable metals in one object at grades exceeding most terrestrial deposits.

Nobody lives there. The abyssal plain supports no fisheries, no communities and no displacement of populations. Operations occur in international waters under a regulator that predates the industry, with a royalty regime intended to distribute benefit to states that have no seabed of their own.

The alternative is being decided by default. Land-based expansion continues regardless. A refusal to permit seabed collection is itself a decision with environmental consequences, taken without anyone counting them.

The case against collection

The ecosystem is not empty; it is slow. Abyssal biomass per square metre is very low, which is often reported as though it meant the habitat were barren. The measure that matters for conservation is not biomass but diversity and endemism, and by that measure the region is remarkable: surveys consistently find that a majority of species recovered are new to science, and many appear restricted to small areas. What is scarce is not life but energy — the food supply is a thin rain of organic particles from the surface — and everything that follows is a consequence of that. Growth is slow, reproduction is infrequent, and recovery from disturbance is correspondingly protracted.

The nodules are themselves the habitat. A significant fraction of local fauna is sessile and requires hard substrate for attachment. The nodules are the only hard substrate on an otherwise soft plain. Removing them does not disturb the habitat; for those organisms it eliminates it, and it cannot be restored on any timescale, because the nodules take millions of years to form.

The empirical record is not encouraging. In 1989 a research team ploughed a small area of abyssal seafloor to simulate disturbance and has returned periodically since. Decades later the plough tracks remain plainly visible and faunal communities within them have not recovered to surrounding levels. That experiment was smaller and gentler than a commercial operation.

The plume problem is unresolved. Collection lifts sediment that has settled undisturbed for millennia. Part forms a near-bottom cloud extending well beyond the collection track; part returns as a discharge from the surface vessel, at a depth that has been the subject of considerable dispute, because midwater ecosystems are poorly characterised and connected to fisheries above them. Plume behaviour at commercial scale has never been observed, and models disagree by more than an order of magnitude.

Where the argument actually turns

Two considerations are decisive, and neither is ecological.

The first is that the demand case has weakened during the period the debate has run. Battery chemistry has shifted substantially towards iron phosphate, which uses neither nickel nor cobalt, precisely because manufacturers wished to escape the cost and supply concentration of those metals. A resource justified principally as necessary for electrification is a resource whose necessity depends on a technological forecast that is already changing. This does not eliminate the case; it weakens the urgency that has been used to argue against delay.

The second is procedural. The regulator has issued exploration contracts for decades while the rules governing exploitation remain unfinished. A provision permitting an applicant to force consideration of a plan within a fixed period has been invoked, which raises the possibility of commercial extraction proceeding under provisional standards. Whatever one's view of the activity, deciding a novel and irreversible question under a procedural deadline is a poor way to arrive at a durable answer.

A position, stated with its conditions

The honest conclusion is conditional rather than absolute.

The disturbance is real, unusually persistent, and imperfectly bounded. The alternative is also real, also damaging, and borne disproportionately by people rather than by species nobody has named. Anyone certain of the answer is not weighing both.

What would make collection defensible is specifiable, and the list is short: baseline surveys sufficient to establish what is present before it is removed; demonstration that the midwater plume can be constrained to a depth with characterised consequences; a regional plan reserving connected, representative areas from extraction permanently rather than by exception; independent monitoring paid for by the operator but not reporting to it; and an explicit accounting showing that the seabed tonne substitutes for a terrestrial tonne rather than adding to it.

None of that is impossible. None of it is close to being in place. Until it is, the burden of proof properly sits with the party proposing an irreversible intervention in a system it has not yet described — and that burden is a matter of ordinary evidential practice, not of environmental sentiment.