The publication of a recent scientific article on our study in the journal PNAS (Proceedings of the National Academy of Sciences) has brought important visibility to research that has been built up over two decades by IPAM scientists and partners. However, some of the interpretations presented deserve clarification, especially as hasty conclusions can lead to confusion about what the study actually shows.
For twenty years, our work monitored experimental plots in the Amazon-Cerrado transition in Mato Grosso, one of the driest and most pressured regions of the Amazon. There we evaluated, on a scale rare for ecological studies, how the forest responds to repeated fires, severe droughts and windstorms, testing whether these disturbances would lead to “savannization”. Understanding that savannization would include the favoring of Cerrado species.
The main conclusion after all these years is that forests subjected to severe disturbances can show significant capacity for recovery when protected from new impacts. This is an important finding. But this recovery does not mean the absence of damage, nor does it eliminate concerns about the biome’s vulnerabilities.
On the contrary, the study shows that recovery occurs with changes in the structure, composition and functioning of the forest, often resulting in ecosystems that are more simplified and potentially more vulnerable to new disturbances. This result in no way invalidates scientific warnings about forest degradation and hotspots in the Amazon. It helps to qualify them.
It is also important to clarify that the debate often summarized under the term “savannization” is more complex than an abrupt conversion of forest into savannah. Our study showed that, after these disturbances, there was no establishment of grass or woody species typical of the Cerrado. In the first few years, the most impacted areas were occupied by exotic grasses associated with pastures, such as brachiaria and andropogon, which favored the spread of fire. Over time, these areas began to be reoccupied by forest species, which are mainly more generalist and potentially more vulnerable. In other words, we didn’t observe a conversion to savannah, but a degraded, simplified and more flammable forest. And it is precisely in this context that our results should be interpreted.
Another central aspect of the study is that the observed recovery occurred because, after the first six years of the fire experiment, the fire was stopped. We have used the analogy of the “hammer effect” to illustrate this point. When successive disturbances continue to affect the forest, its ability to recover decreases. In the experiment, the removal of this factor made room for regeneration.
This condition, however, does not represent the reality of much of the Amazon. Furthermore, this resilience operates within a climate envelope that is being progressively narrowed by global warming. Each new extreme drought and each new major fire reduces the conditions in which this recovery can still occur. For this reason, the study’s main message is not that the forest is “safe”, but that its resilience depends on protection against new disturbances.
It is also important to avoid undue extrapolations. The experiment was carried out in a particularly resilient forest in the Amazon-Cerrado transition and cannot, in isolation, be generalized to the entire Amazon basin, which is made up of forests with different levels of vulnerability and ecological responses.
One of the most relevant findings is that burned and unburned areas are becoming more similar at the edge, for example, which indicates that even apparently intact forests are undergoing degradation processes. This result reinforces the urgency of conservation.
Our study does not invalidate this body of knowledge about the vulnerability of the Amazon rainforest to disturbances. On the contrary, it contributes to qualifying it, showing greater ecological complexity and, at the same time, reaffirming that the resilience observed depends on the conditions in which it is found, in relation to the pressures it has been suffering.
In this sense, the results should not be read as a reason to relativize forest protection policies or reduce attention to climate and ecological risks. Rather, they reinforce the importance of preventing fires, containing degradation, implementing integrated fire management, strengthening protected areas and confronting global warming, which narrows the conditions in which this resilience still operates.
The Amazon may be more resilient than sometimes assumed. This is good news. But resilience is not invulnerability. The study’s central contribution is to show that there is still a window for recovery, provided that the forest has the right time, protection and climatic conditions. This is precisely, without fear of exaggeration, an argument in favor of extending, not reducing, the commitment to its protection.