Protecting standing Amazon forests beats restoration for biodiversity and carbon, study finds

As the 2026 El Niño intensifies across the Amazon basin, a landmark study published in Science has delivered one of the most rigorous answers yet to a question that has long divided conservation science: when money is limited, is it better to protect the forests that still stand, or to replant the ones that have already fallen? The research, led by scientists at Lancaster University with colleagues from the United Kingdom, Brazil, Sweden and Australia, concludes that safeguarding existing forests from disturbances such as selective logging and wildfire often delivers the biggest return on investment, a finding with immediate relevance as longer, hotter dry seasons push the world’s largest rainforest closer to its ecological limits.

The study represents the first systematic comparison of its kind, pitting strategies that protect standing forests against those that restore cleared land on agricultural properties. Rather than relying on simple correlations, the team built a counterfactual framework: using observed land-use changes between 2010 and 2020, they modelled how each conservation strategy would have affected biodiversity and carbon stocks compared with the changes that actually took place. This approach allowed the researchers to isolate the genuine contribution of each intervention, rather than crediting or blaming policies for outcomes that might have occurred anyway.

The scale of the analysis is itself remarkable. The researchers covered more than three million hectares across two regions of the Brazilian state of Pará, Santarém and Paragominas, both situated along the so-called arc of deforestation where agricultural expansion presses hardest against primary rainforest. High-resolution satellite imagery was combined with field data gathered from 381 survey transects distributed across 38 catchments and 499 rural properties. To translate habitat condition into biological terms, the team modelled the distributions and responses of 160 bird species and 426 tree species, capturing two major dimensions of tropical biodiversity: the mobile fauna of the forest canopy and midstory, exemplified by species such as the pygmy antwren recorded in the Santarém region, and the slow-growing, long-lived trees that anchor the ecosystem’s carbon storage.

Crucially, the researchers did not stop at measuring benefits. They also estimated the costs associated with each intervention, allowing the gains delivered per unit of investment to be compared within a single common framework. This cost-effectiveness lens is what gives the study its policy punch, because conservation budgets are finite and choices between competing strategies carry real consequences for how much biodiversity and carbon can be secured with the resources available.

The headline result is unambiguous. Avoiding forest disturbances, chiefly wildfires and selective logging, generated the greatest benefits for forest biodiversity of any single strategy examined, and was almost as important as preventing outright deforestation for protecting carbon stocks. In other words, a forest that remains standing on satellite imagery is not necessarily a forest that is safe: degradation through fire and timber extraction erodes its ecological value and its carbon even when the canopy appears intact. The authors argue that policies and resources to prevent and tackle the wildfires that disturb and degrade tropical forests are urgently needed.

Senior author Professor Jos Barlow of Lancaster University stressed how sharply the stakes rise in an El Niño year. These climatic events, combined with ongoing climate change, produce longer, hotter and more intense dry seasons across the arc of deforestation, greatly increasing both the extent and the severity of forest fires. Lead author Dr Leonardo Miranda, also of Lancaster University, cautioned that no single strategy is a silver bullet, but said the results provide clear evidence that conservation should begin by prioritising the forests that remain, and that protection must extend beyond simply avoiding deforestation to include preventing the disturbances that quietly degrade what is left.

Restoration, by contrast, delivered important but conditional benefits. On properties where most of the original forest had already been lost, replanting and assisted regeneration proved valuable, offering biodiversity and carbon gains that protection alone could not supply because there was little forest left to protect. Across the study area as a whole, however, restoration implemented on its own was less effective, and it was also the most expensive of the strategies examined. The implication is that conservation resources should be directed according to local landscape conditions: restoration where forest cover has collapsed, protection where forest persists but is under threat from fire and logging.

Perhaps the most sobering finding is that no single intervention was sufficient to reverse biodiversity and carbon losses across the entire three-million-hectare study area. Net gains in biodiversity and carbon storage were achieved only when both avoidance interventions, preventing deforestation and preventing disturbances, were implemented together, and combining all three interventions, including restoration, delivered even greater gains. Dr Miranda noted that combined interventions will become even more important as climate change progresses, particularly because young, regrowing forests are especially vulnerable to wildfires, meaning that restoration investments can be wiped out in a single bad fire season if protection is neglected.

The timing of the findings could hardly be more pointed. Forest restoration features prominently in global commitments and in Brazilian federal and state policies, yet the study warns that investing in new forests makes little sense if existing forests are simultaneously being degraded and lost. Dr Joice Ferreira, senior researcher at Embrapa and principal investigator of the Capoeira research programme on restoration, described the problem as pouring money into a leaky pot: restoring forests in one place while others are being degraded and lost elsewhere. What is urgently needed, she argued, is the integration of these policy agendas, ensuring that forest restoration goes hand in hand with the protection of the forests that still exist.

Although the research focused on two Amazonian regions, the researchers believe the findings are likely applicable to tropical deforestation frontiers around the world and carry relevance for policies well beyond the tropics. Dr Toby Gardner, co-author and Senior Research Fellow at the Stockholm Environment Institute and co-founder of the supply chain sustainability initiative Trase, highlighted the responsibility of the companies and markets that profit from internationally traded commodities such as soy and beef, arguing that they must answer not only for deforestation but also for the degradation of remaining forests that is inextricably linked to these production systems. The research was convened by the Sustainable Amazon Network, Rede Amazonia Sustentavel, and funded by sources including the United Kingdom’s Global Centre on Biodiversity for Climate and Brazil’s National Council for Scientific and Technological Development. The paper, titled Protecting tropical forests is more cost-effective for biodiversity and climate than restoration, was published in Science on 17 September 2026, and its central message is already resonating: in an era of intensifying fire seasons and tightening budgets, the cheapest and most effective conservation strategy may be the most obvious one, keeping the forests we still have standing, intact and undisturbed.

Subject of Research: Cost-effectiveness of forest protection versus restoration strategies for biodiversity and carbon conservation in the Brazilian Amazon

Article Title: Amazonian study identifies most effective ways of conserving biodiversity and carbon

Article References: Amazonian study identifies most effective ways of conserving biodiversity and carbon. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: Amazon, biodiversity, carbon stocks, forest restoration, selective logging, wildfire, deforestation, El Niño, Brazil, Pará, conservation policy, Science journal