Diversity of functional trade-offs enhances survival after fire in Neotropical savanna species

Prepared by Marina Corrêa Scalon, Fabricius Maia Chaves Bicalho Domingos, Wesley Jonatar Alves da Cruz, Ben Hur Marimon Júnior, Beatriz Schwantes Marimon & Imma Oliveras

Figure 1. Photos representing the four resprouting categories of Cerrado trees after fire: (a) aerial, (b) stem, (c) basal, and (d) underground. Nova Xavantina, Mato Grosso, Brazil. Photo credit: Marina C. Scalon

Trees in the Central Brazilian savanna have different strategies to survive and recover after fire, which naturally occurs in this habitat. After an accidental human-induced fire event in a protected area that had all trees already tagged and identified at the species level, we followed up their recovery and could identify how different tree species responded to fire. We checked whether individual trees resprouted from the canopy, from the main stem, from the base of the trunk and/or from underground organs (Figure 1). We also identified those that died or had all their above-ground stems killed. We then investigated how bark, growth rates and the protection of buds related to these recovery strategies and to species survival. It is important to emphasize that bark is a complex tissue, composed of an inner living part, usually associated with water storage, and an outer dead part, that can be corky, papery, smooth, among others. Bark also plays a very important role in protecting dormant buds until new shoots may develop. This bud protection can be classified into four different categories, according to how deep the buds are embedded in the bark, going from 0 for completely exposed buds; 1 for emerging buds but protected meristem (plant growth cells); 2 for buds allocated inside depressions in the bark; and 3 for buds completely protected underneath the bark (Figure 2).

Epicormic buds of Cerrado species divided into bud protection categories: 0 for completely exposed buds (Dipteryx alata); 1 for emerging buds but protected meristem (Hymenaea stignocarpa); 2 for buds allocated inside depressions in the bark (Byrsonima coccolobifolia); and 3 for buds completely protected under the bark (Kielmeyera rubriflora). Scales 5 mm. Red arrows are pointing to the bud. Nova Xavantina, Mato Grosso, Brazil. Photo credit: Marina C. Scalon

We found that species with more exposed buds had a higher risk of dying, while species that had buds deeply embedded in the bark promptly recovered, resprouting both from the canopy and the base of the trunk. We identified three main strategies that plant species adopt to survive after fire: (1) growing tall and wide, and thus escaping the flame height; (2) having a thick inner bark and high bark moisture, which are associated with preventing high temperatures in the internal tissues; or (3) having a thick outer bark and high bud protection, which are important to isolate buds from heat. Most importantly, species adopting different strategies have similar mortality rates and similar recovery, indicating that there is not a single “most effective” strategy and all three strategies are equally efficient to warrant species persistence.

Among the 24 species we studied, some completely lose their leaves during the dry season (deciduous species) to save more water by avoiding leaf transpiration, while others are capable of maintaining a full canopy (evergreen species). We took the opportunity to also test if these groups would respond differently after fire. Our results suggest that human-induced fire indeed can be more harmful for species that have flushed new leaves and may be short on energy to resprout (deciduous), not being able to fully recover.

Our study shows different strategies savanna trees evolved to endure fire, by combining bark properties, bud protection and growth rates to survive and recover after fire. It also revealed the effect human-induced fire may have on different groups of plants, which can be detrimental to the diversity of the Cerrado in the long-term, with the increase in fire frequency and intensity.

This is a plain language summary for the paper of Scalon et al. published in the Journal of Vegetation Science (https://doi.org/10.1111/jvs.12823).