2020 ESA Annual Meeting (August 3 - 6)

COS 63 Abstract - Large wildfire impacts on riparian forests in a tropical savanna biome

Bernardo Flores1, Marina Hirota2, Michele de Sá Dechoum2, Isabel B. Schmidt3 and Rafael Oliveira1, (1)University of Campinas, Brazil, (2)Federal University of Santa Catarina, Brazil, (3)Department of Ecology, University of Brasilia
Background/Question/Methods

Tropical savanna biomes are increasingly under stress. Climate change, biological invasions and homogenized landscapes are reducing the resilience of these systems, while fire-suppression policies are increasing their vulnerability to large wildfires. In 2017, Brazil suffered one of its worst fire years on record. One particular case called huge media attention; a large wildfire spread across the iconic Chapada dos Veadeiros National Park (CVNP), burning all types of ecosystems including the fire-sensitive riparian forests. The CVNP is a UNESCO World Heritage Site, due to its unique biodiversity, geological and spiritual values to humanity. Using field data from 36 different forest sites, randomly spread across five streams affected by the wildfire, we measured fire impacts on the structure of trees and herbaceous plants, as well as on top soil variables. We hypothesized that because these ecosystems are not adapted to fire, impacts would be severe and potentially catastrophic.

Results/Conclusions

Our findings indicate a striking sensitivity of riparian forests in the region. We found that, on average, the wildfire killed 50 % of the adult trees and 88 % of the saplings. It destroyed 52 % of the soil root mat, exposing 28 % of bare soil and causing available P concentration to increase 12-fold. Only six months after the wildfire, our assessments reveal that, on average, the cover of opportunistic vines already doubled, and invasive grasses increased by 3 %. Flooded riparian forests were shown to be much more sensitive than non-flooded ones, possibly due to the relatively thinner barks of their trees. Within the framework of adaptive cycles, and grounded on paleoecological evidence from other studies, we hypothesize three plausible scenarios for these ecosystems. In the coming decades, disturbed forests may (1) recover, (2) become trapped in a degraded state with native opportunistic and non-native invasive plants, or (3) shift into an alternative state of palm swamp savanna. In view of the rising wildfire risk in tropical savannas, our findings stress the need for an integrated ecosystem management that considers riparian forests as critical parts of the system.