Drought-induced Amazonian wildfires instigate a decadal-scale disruption of forest carbon dynamics

By Camila V.J. Silva

I always remember an impressive image when I first stepped in a fire-disturbed humid forest: giant trees standing dead. That scene was choking and also scaring as I walked nearby those trees. In contrast to the massive disturbance caused, the ability of these forests to regenerate its understorey up to 5 to 6 years after fire was surprising. Big trees that were missing opened huge gaps in the canopy favouring the establishment of many herbaceous plants and pioneer species. After the fire, we immediately noticed that these forests stored much less biomass than before, so we asked for how long would these humid fire-affected forests take to recover to pre-disturbance levels?

Figure 1: Increase in mortality rates of large and dense wood trees after drought-induced wildfires in Central Amazonia

Everything we knew already was based on studies focusing in short-term responses of vegetation to fire (<10 years) and data sets of single-off forest inventories. To overcome the lack of knowledge on long-term responses of Amazonian vegetation to fire, a network of permanent plots in burned sites (FATE) was created. Since 2009, we have been walking in the interior of the Brazilian Amazon measuring and monitoring more than 9.000 trees in burned and unburned forests. Within a variety of fire-affected forests in different regeneration stages, we established 35 permanent plots to create a chronosequence of 31 years since last fire. To compare with the pre-fire, we simultaneously installed and measured 29 permanent plots in undisturbed primary forests adjacent to the sampled burned areas (Figure 2), which allowed us to investigate shifts in biomass stocks and forest dynamics over time.

Figure 2: Intact tropical forest. Photo by Yadvinder Malhi
Figure 3: Burned tropical forest. Photo by Yadvinder Malhi

Using this long-term forest inventory data, in a recent study published in Phil Trans B special edition on ‘The impact of the 2015/2016 El Niño on the terrestrial tropical carbon cycle: patterns, mechanisms and implications’ we found an increase in mortality rates of large and dense wood trees after drought-induced wildfires in Central Amazonia. In our study, we show that 30 years after the fire-disturbance, ‘apparently’ recovered forests store 25% less biomass than the adjacent undisturbed primary forest (Figure 4). The wildfires enhanced mortality rates of large and high wood density trees, where the largest amount of biomass in old-growth forests is stored. We found that for decades, the pronounced biomass loss was not sufficiently compensated by forest growth and recruitment (wood productivity). We show that the post-fire growth response was predominantly associated with groups that contribute relatively little to overall aboveground carbon stocks: small-medium sized trees represented in light or intermediate wood density classes.



Figure 4: Differences in total aboveground biomass (TAGB) and Net TAGB balance (mortality and wood productivity) in relation to unburned forest: (Silva et al. 2018)

Our chronosequence show that in the long-term, the balance between mortality and wood productivity was close to undisturbed forests equilibrium (Figure 4). However, in order to fully recover, gains in biomass would need to be higher than losses, and net biomass balance should have more positive values when compared to undisturbed forests. Humid tropical forests are not fire-adapted and their recovery to fire-disturbance can be a very slow process that persists for many decades, which is beyond the time frame of our study.

Worryingly, climate models predict that drought-induced wildfires will become more frequent and fire-affected forests in the Amazon may increase. In turn, our study shows that wildfires slow down or stall the post-fire recovery of Amazonian forests, highlighting the urgent need to avoid fires in humid tropical forests.

Read more here