Mangrove forests hold some of the highest densities of carbon recorded in any ecosystem, but have experienced widespread deforestation through conversion to aquaculture and agriculture. Alongside deforestation, mangroves have shown simultaneous natural expansion in some parts of the world, and considerable investments have been made into restoration programmes. Here we estimate net changes in the global mangrove carbon stock due to land cover change between 1996 and 2016, using data on mangrove deforestation and forestation, and proportional changes in carbon stock during processes of mangrove loss and gain. The global mangrove carbon stock declined by 158.4 Mt (95% CI = −156.8–525.9 Mt); a reduction of 1.8% of the stock present in 1996. Efforts to conserve and restore mangroves appear to have had some success, and - along with natural forestation - have contributed to relatively low net losses of mangrove carbon stocks over two decades.

Tropical peatlands store around one-sixth of the global peatland carbon pool (105 gigatonnes), equivalent to 30% of the carbon held in rainforest vegetation. Deforestation, drainage, fire and conversion to agricultural land threaten these ecosystems and their role in carbon sequestration. In this Review, we discuss the biogeochemistry of tropical peatlands and the impacts of ongoing anthropogenic modifications. Extensive peatlands are found in Southeast Asia, the Congo Basin and Amazonia, but their total global area remains unknown owing to inadequate data. Anthropogenic transformations result in high carbon loss and reduced carbon storage, increased greenhouse gas emissions, loss of hydrological integrity and peat subsidence accompanied by an enhanced risk of flooding. Moreover, the resulting nutrient storage and cycling changes necessitate fertilizer inputs to sustain crop production, further disturbing the ecosystem and increasing greenhouse gas emissions. Under a warming climate, these impacts are likely to intensify, with both disturbed and intact peat swamps at risk of losing 20% of current carbon stocks by 2100. Improved measurement and observation of carbon pools and fluxes, along with process-based biogeochemical knowledge, is needed to support management strategies, protect tropical peatland carbon stocks and mitigate greenhouse gas emissions.