A new University of Exeter whole forest model sharpens, rather than breaks, the case for mangroves as climate infrastructure by showing which coasts are most at risk.
The promise of mangroves as climate infrastructure rests on a startling ratio: these coastal forests cover less than 1% of Earth's surface yet hold roughly 15% of all carbon stored in the ocean, according to new University of Exeter research published in Earth's Future on 3 June 2026.[^1] The whole-forest model described in that paper does not retract the number. It shows that whether a mangrove stand behaves as a carbon sink or a carbon source depends less on the trees themselves than on the coastline behind them. Where rising seas can drown the landward edge and erode the soil, even a vigorous forest can lose carbon faster than it grows.
[^1]: Iwantoro, A. P., Urrego, D. H., Xie, D., Nicholas, A. P., Hapsari, K. A., Rodríguez-Rodríguez, J. A., et al. (2026). The importance of scale in the future of mangrove blue carbon under sea-level rise. Earth's Future. https://doi.org/10.1029/2025EF006984
The mechanism is plain. Saltwater pushed inland by storm surge and chronic flooding kills the aerial roots, the snorkel-like structures that keep the trees breathing when tides rise. Dead roots collapse into oxygen-poor mud, where microbes break them down and release stored carbon. Soil that took centuries to build washes out to sea. In a young stand or at the eroding edge, that loss can outweigh the new growth in surviving trees. The Exeter model is the first to put that whole-forest balance sheet on one page, rather than extrapolating from a few healthy cores.
It is also, crucially, a forward-looking projection, not a measured present-day reversal. The research paper is explicit that today's global mangrove estate still stores more carbon than it releases, and that the carbon-source outcome is a model output for a future sea-level trajectory rather than a current measurement.[^1] The timing depends on the emissions path and the coast in question, and the authors do not pin either down with a single year.
That distinction matters for the people who decide where to spend coastal-protection money. A sink on a trajectory to become a source later this century is a planning problem now, but it is not a present-day crisis. Field studies have repeatedly shown extraordinary carbon densities in mature mangrove stands, and that point still stands. What the Exeter model adds is the view from above the stand: when you stitch the high-density core together with the eroding edge, the forest's net carbon balance is more sensitive to sea-level rise than single-plot work suggested. For carbon-credit markets, that sensitivity is the real headline. For coastal engineers, it is a siting instruction.
Policy levers already exist, and the new model tells planners which to pull first. Protecting existing stands from clearing and from the canals that drain their soil remains the cheapest carbon action on the coast, because a hectare of mature mangrove holds several times more carbon than the same hectare of restored forest will accumulate in a human lifetime.[^1] Where the seaward edge is already eroding, the priority is enabling landward migration: removing seawalls, buying out doomed coastal properties, and rezoning low-lying agricultural land back to wetland so the forest can roll uphill. Where neither is possible, hydrology restoration — rebuilding tidal flow that has been blocked by roads and aquaculture ponds — can extend the productive life of a stand.
The Exeter model also has something to say to blue-carbon finance. Credits issued against a forest that is on a trajectory to lose soil carbon are not worth what credits issued against a forest whose landward edge is zoned wetland. The new model gives certifiers a way to discount, or refuse, projects where the coastline is closing in. That is a quieter headline than "mangroves turn into carbon sources," and it is the one more likely to change which projects get built.
What to watch next: the primary paper names the sea-level-rise scenarios tested (several IPCC trajectories), covers mangroves globally, and gives a net-source transition timeline that varies by coastline geometry and subsidence rate. Independent groups will re-run the model with different coastline and sediment data. Restoration siting guides from agencies that fund coastal climate work will be tested against this whole-forest view, or they should be. The headline that matters is not that mangroves have stopped being climate heroes. It is that being a climate hero, for a coastal forest, requires a coastline that is allowed to move.