The setting sun appears as a bright red circle with the black silhouette of burnt tree against it

Sunset in the Brazilian Amazon, south of Porto Velho. Researchers project that the Amazon carbon sink is on course to disappear by 2035.Credit: Carl de Souza/AFP/Getty

As tropical forests grow, they pull carbon dioxide out of the atmosphere — one of their many services to humanity and the planet.

Decades of measurements in hundreds of plots in Africa and South America show how tropical trees such as Brazil nut (Bertholletia excelsa) and kapok (Ceiba pentandra) absorbed as much as 4.4 billion tonnes of carbon dioxide annually in the 1990s and early 2000s. That’s enough to more than offset the European Union’s carbon emissions during the same period.

This effect is baked into many of the climate models that researchers use to project future global-warming scenarios. However, a study published in Nature this week suggests that the benefits from this tropical carbon ‘sink’ might be fleeting (W. Hubau et al. Nature 579, 80–87; 2020). And that could mean the international community will need to pledge yet faster emissions reductions if the world is to limit global warming to below 2 °C, in line with the 2015 Paris climate agreement.

An international team led by geographers from the University of Leeds, UK, reports on page 80 that the Amazon rainforest has been absorbing less atmospheric carbon each year since the early 1990s. Forests in Africa have also been absorbing less atmospheric carbon since around 2015. This is due in large part to rising tree mortality.

Trees are dying, the researchers found, because temperatures are rising and drought is increasing, a trend that is likely to continue as greenhouse gases build up. A decade from now, Africa’s carbon sink will be 14% lower compared with 2010–15. The Amazonian carbon sink is on course to disappear completely by 2035. If that happens it will result in more carbon dioxide in the atmosphere, and therefore more global warming.

As we reported in a Feature last week, the Amazon’s 5 million square kilometres look more precarious than ever (see Nature 578, 505–507; 2020). Average temperatures in this rainforest, which spans nine countries, have risen by 1–1.5 °C over the past century; there have been three severe droughts since 2005 and tree clearing has shrunk the forest by 15% since the 1970s. Brazil, once praised for its efforts in slowing deforestation, lost 10,000 square kilometres last year — the largest drop for a decade. A ten-year ban on planting sugar cane in the Amazon was lifted last November; and a bill to regulate oil and mining exploration is making its way to the national congress, Brazil’s parliament.

In September, independent researchers from the region formed a science panel to propose what needs to be done to conserve the Amazon. The panel hasn’t yet completed its report, but its overarching message cannot be in doubt: Brazil and other tropical nations need to halt deforestation and promote new forests in degraded — and often abandoned — lands.

At November’s summit of the United Nations Framework Convention on Climate Change in Glasgow, UK, participating countries will be expected to redouble their pledges to meet the Paris climate agreement’s goals. If tropical carbon sinks can no longer be fully relied upon to help reach that target, it means more ambitious decarbonization will be needed.

At the same time, the lesson for governments around the world is clear enough: tropical forests are working for humanity — and for countless other creatures. To protect them, humanity must halt both deforestation and global warming.