The Forest That Drowns

The second possibility is drawing fresh attention beyond saltwater shores. Reporting in New Scientist on October 9, Chris Simms describes research mapping more than 260 million dead trees near US ocean coasts and the Great Lakes. The striking finding is not merely the count: freshwater landscapes also contained hotspots of forest mortality associated with rising water.

The study, led by Henry Chi Hang Yeung at the University of Virginia, is a bioRxiv preprint posted September 22, not a peer-reviewed finding. Its authors report that mortality rates doubled within a decade around expanding freshwater lakes and in inland wetlands experiencing intensifying precipitation. In low-lying coastal forests, where salinization compounded inundation, rates tripled.

A familiar image—skeletal trees standing where a coastal forest once lived—needs a wider map. Salt is not a prerequisite for a ghost forest.

A map, not 260 million verdicts

The researchers used roughly one-metre-resolution aerial imagery and deep learning to track dead trees from 2012 to 2023. According to Simms, the images came from the National Agricultural Imagery Program. The survey covered areas within 10 kilometres of US Pacific, Gulf and Atlantic coasts and US Great Lakes shorelines.

That boundary matters. This was not an inventory of every inland forest in the country. Within the mapped area, more than half of the mortality was concentrated in just 10 per cent of forests, with hotspots in inland wetlands as well as low-lying coastal forests.

For the Great Lakes, Simms reports that high precipitation and runoff drove record-high water levels between 2013 and 2020. The researchers suggest that sustained waterlogging increased the likelihood of nearby tree deaths. The distinction from the coastal mechanism is useful: freshwater inundation can be damaging without the additional stress of salt.

But a photograph of a dead tree is not its death certificate. Arya Iwantoro of the University of Plymouth tells New Scientist that the changing hydrological conditions offer convincing evidence of an important contribution to mortality, while cautioning against attributing every death directly to inundation or salinity. Storms, drought, pests, drainage and soil conditions can interact.

The 260-million figure is therefore a mapped count of dead trees, not a count proven to have drowned. The pattern deserves attention without being promoted into a single explanation for every trunk.

What grows after the ghost

The carbon implications are similarly a question about transitions, not just subtraction. Trees die and decompose; a forest that fails to recover may lose its capacity to take up carbon. Yet the wetland vegetation that replaces it may also accumulate carbon, Iwantoro notes. The relevant accounting follows the changing landscape, rather than ending when the canopy disappears.

The coherenceist lesson here is practical: let the process redraw the category. If attention stops at the ocean shoreline, freshwater inundation can fall outside the familiar picture of water-driven forest loss. If attention stops at dead trees, the ecosystem that follows can fall outside the carbon account. Neither omission is repaired by pretending every place has the same trajectory.

Widening the picture means tracking water, local conditions and replacement vegetation together. It does not make the lost forest less real. It gives the people studying and caring for these places a more faithful starting point.

The trees cannot move uphill. Our understanding has rather fewer roots.