The Argument Over The Abyssal Plain

AScattered across the abyssal plains of the central Pacific, at depths of between four and six kilometres, lie potato-sized lumps of rock known as polymetallic nodules. Each has grown around a shark's tooth or a fragment of shell at a rate of perhaps a centimetre every million years, gathering manganese, nickel, cobalt and copper as it does so. The commercial interest is plain enough: the metals concerned are those that battery manufacturers have been competing for since the mid-2000s, and the nodules sit loose on the sediment rather than locked inside bedrock, so that collecting them requires no blasting and no shafts. The ecological difficulty is equally plain, though it is less often stated precisely. A machine that lifts nodules lifts the top layer of sediment with them, and the creatures that grow on the nodules have nowhere else to go, because the nodules are the only hard surface for hundreds of kilometres in any direction. Whether the resulting damage is a price worth paying, and against what it ought to be compared, is the argument the industry has yet to settle.

BHelena Vestergaard, a benthic ecologist at the Kalmar Institute of Ocean Science, has built her case on a single accident of history. In 1978 a consortium dragged a collector across a patch of the Perenna Basin measuring 2.4 square kilometres, and then abandoned the site. Vestergaard's team resurveyed it in 2014 using towed cameras, logging 31,000 images along the old tracks and along undisturbed ground beside them. Mobile scavengers, the shrimps and sea cucumbers that wander over the sediment, had returned to 84 per cent of the density recorded outside the tracks. Sessile filter feeders anchored to nodules stood at 11 per cent. The tracks themselves were still legible in the imagery, their ridges as sharp as on the day the machine passed. Vestergaard's conclusion is not that the abyssal seabed cannot recover, but that recovery is uneven: what moves comes back, what is fixed in one place does not, and thirty-six years is long enough to establish the difference.

CThe account yields a prediction. If it is the loss of hard surface rather than the churning of sediment that keeps the fixed animals away, then adding artificial surface to a stripped area ought to bring some of them back even where the sediment is still disturbed. Vestergaard tested this in 2019 by cementing ceramic blocks onto three cleared plots. Within four years the blocks carried sponges and corals at roughly a third of natural density, while the bare sediment a metre away carried almost none. She is careful about what the result shows. The colonists were the fast-growing species, she notes, and none of the slow-growing corals that dominate untouched nodule fields appeared at all; the experiment establishes that surface is the limiting factor, not that a mined field can be put back together.

DIdris Okonkwo, a resource analyst at the Bergen School of Industrial Ecology, does not dispute the survey but rejects what is drawn from it. A figure for damage, he argues, means nothing until it is set beside the damage that would otherwise occur, since the metals will be taken from somewhere. His group's 2021 comparison put the greenhouse emissions of nodule collection at about 60 per cent of those from an equivalent tonnage of nickel won from tropical laterite deposits, and noted that the land route displaces settlements and clears forest, neither of which has an obvious counterpart at five kilometres down. Okonkwo concedes two things. The comparison assumes that seabed production replaces land production rather than being added to it, an assumption for which he has no evidence beyond the industry's own projections; and his emissions figures exclude the processing of the ore, which happens on land whichever route is taken and which he expects would narrow the gap.

EThe dispute has long since left the seminar room. In 2021 a small Pacific state notified the licensing body of its intention to begin commercial extraction, triggering a clause that obliged the body either to finish its mining code within two years or to consider applications under whatever rules happened to exist. The deadline passed with no code agreed. More than thirty governments have since called for a moratorium, and several manufacturers of electric vehicles have undertaken not to buy seabed metal until the science is clearer. Okonkwo observes, awkwardly for his own argument, that setting one harm beside another cannot by itself tell anyone how much harm is tolerable, since that is a question about values rather than tonnes. Vestergaard replies that the distinction is cleaner on paper than in practice, because negotiators read any published ratio as a licence to proceed. Both agree that the shortage of baseline surveys is the real obstacle to regulation.

FA partial settlement is now visible, in which the two lines of work are read as answers to different questions rather than ranked against each other. Where the question is how quickly a disturbed patch comes to resemble its surroundings, the sediment-dwelling community supplies a measurable and moderately encouraging answer; where it is whether the assemblage that grew on the nodules can exist without them, the answer appears to be no, and no amount of waiting changes it. What neither approach can currently reach is how far the affected species range beyond the areas under licence, since a species found only within one licence area is lost along with it, while one distributed across the whole basin is not, and the sampling needed to tell these cases apart has never been done. Both researchers reject the assumption behind the earliest environmental statements, namely that a seabed holding few animals per square metre must be a seabed where little is at stake.