Evaluating fire-induced dieback of Amazonian rainforest

Lead Research Organisation: University of Oxford
Department Name: Geography - SoGE


Wildfires have become the new norm in many parts of the Amazonian humid forest, an ecosystem that did not co-evolve with this stressor. Large areas of previously undisturbed and human-modified forests are catching fire, jeopardising the future of the largest and most biodiverse tropical rainforest in the world, and potentially acting as a feedback that would further increase regional and global climate change. In recent dry years the extent of fire has greatly exceeded the rate of deforestation in the Brazilian Amazon. These fires result in in about half of the trees dying, and open up the forests to make them more vulnerable to subsequent fires.

Despite the growing prevalence of Amazonian wildfires, we still have a very limited understanding of why these low intensity understorey fires cause such high rates of tree mortality, which species functional traits predict vulnerability or survival to these fires, what are the impacts of wildfires on the forest carbon balance and what are the patterns of taxonomic and functional recovery following a fire event.

We propose a research plan to achieve major advances in our understanding of such wildfire impacts, including of the underlying mechanisms that cause both short-term and longer-term tree mortality. This work will be based at a field site in Santarem in Eastern Brazilian Amazonia, where we have collected several years of measurements of detailed vegetation ecology and carbon cycle tracking over a range of plots. These include a number of plots which experienced fire during the 2015/16 El Niño.

We will implement this project by combining a state-of-the-art forest burn experiment with continued monitoring of a unique set of long-term sampling plots, some of which we have tracked through the 2015-16 wildfire event associated with a strong El Niño. We are uniquely placed to address these fundamental questions given our network of burned and unburned forest plots that is already in place, the strong partnerships whave forged with park managers and federal agencies, and the numerous past datasets that we can use as baseline information.

The fire experiment will involve setting fire to limited patch of forest (with the close cooperation of local fire brigades), tracking fire intensity and tracking the physiology and mortality responses of individual trees in the fire plots, including trees that have their root mats or their bark insulated from fire damage. We will also experiment with different fire break methodologies to explore the most effective way to stop such fires. With the intensive carbon cycle studies we will track the carbon cycle responses for up to seven years after the 2015 fires, giving us novel insight into the longer term carbon cycle responses and the ecosystem responses and recovery after a fire event.

As well as advancing scientific knowledge about a pervasive and increasing threat to the future of tropical forests in the Anthropocene, our co-designed pathways to impact ensures we will also inform and improve approaches to minimise risk of fire-induced dieback of humid Amazonian forests. We will work closely with local fire managers, and engage with state and national policymakers, to draft recommendations on how to manage forest reserves and forest-agriculture mixed landscapes to minimise the risk of fire spread. If applied at a large scale, these fire prevention strategies are a crucial tool that can help minimise the risk of extensive fire-induced dieback within Amazonian forests.

Planned Impact

DIEBACK - Pathways to Impact

The DIEBACK project links cutting-edge science with a central issue for the management of Amazonian forests: wildfires. Objective/Question 5 "Can we develop a better understanding of fire transmission within forests that allows the design of more effective fire prevention strategies" is entirely focused on developing a landscape-scale understanding of forest flammability and designing more effective fire prevention strategies in forests that did not co-evolve with fire as a natural stressor. As such, this project will directly benefit policy makers and key stakeholders at the local, regional and federal scales in Brazil, providing guidance on how to better tackle this increasingly widespread issue and therefore helping to protect the world's largest and richest tropical forest.

Below we describe the different ways this project aims to have real-world impact:

Local scale - we will produce booklets and short films to showcase our findings. We will continue our extensive outreach work in the Santarem region by arranging meetings with the managers of the Floresta Nacional do Tapajos, the small-farmers union, the large-farmers union, and indigenous leaders. During these meetings we will hand the booklets, show the films and discuss our findings on the most effective ways to build forest firebreaks. Booklets will also be handled in rural schools. These actions have been highlighted by different stakeholders as key for disseminating scientific results and will build on the large rapport we have already built with the local stakeholders.

Regional and Federal scales - we will produce a high-profile policy brief summarising our results and experiences related to forest flammability, fire prevention and efficient combat. This will be handed in an end of the project workshop to be held in Brasilia, to which relevant regional and federal stakeholders will be invited, including the environment secretaries of Amazonian states, the head of the Brazilian forest fire brigade, representatives of the Forest Service, representatives of ICMBIO (the agency that manages all Brazilian reserves), representatives of the environment ministry, and representatives of the agriculture ministry.

Brazilian researchers and practitioners. By working closely together with regional and federal agencies, we will also engage in transfer knowledge through the research process itself. This is evident in the integration of practitioners in project co-design. Co-authorship with project partners will ensure this integration leads through to project outputs.

Media - in order to reach the general public, we will ensure that our papers are well covered by both the Brazilian and the international media. We will explore having elements of the media (e.g. National Geographic) accompany our fire experiment. All project members will use social media to promote the project's findings to both the scientific community and the general public.

The DIEBACK project will likely have real-world impacts beyond its duration, leaving a scientific legacy that will help to inform environmental policies well after its end.


10 25 50