Determinantes bióticos e abióticos de atributos do material combustível em campos e savanas do Cerrado

• Ano de defesa: 2023
• Autor(a) principal: Manzan, Waleska Barbosa Ferreira
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Uberlândia; Brasil; Programa de Pós-graduação em Ecologia e Conservação de Recursos Naturais
• Palavras-chave em Português: Inflamabilidade; Fogo; Combustível; Savana; Queimadas; Cerrado; Flammability; fuel traits; savanna; cerrado; CNPQ::CIENCIAS BIOLOGICAS::ECOLOGIA::ECOLOGIA DE ECOSSISTEMAS; Ecologia; Resíduos como combustível; Queimada - Savanas - Cerrados; Fogo e ecologia
• Link de acesso: https://repositorio.ufu.br/handle/123456789/39158; http://doi.org/10.14393/ufu.di.2023.8087

Fire is a key driver of the structure and composition of plant communities in tropical savannas. Surface fuel traits are important determinants of fire behavior in these communities and variation in these traits is influenced by both biotic and abiotic drivers. Fuel type, biomass, bulk density, height and moisture are essencial fuel traits determining fire regimes and their impacts on savanna ecosystems. In this study, we analyzed the spacial variation in fuel traits, how they are intercorrelated and respond to environmental conditions and fire history. We also estudied how they contribute to explain fire frequency in grasslands and savannas. We tested the hypothesis that fuel traits are determinated by biotic factors (fuel traits are intercorrelated) and abiotic factors (vary in response to soil texture, vegetation cover, climate and fire history). We also hypothesized that spacial variation in fuel traits helps us to understand fire history in savanna landscapes. The present study was conducted in the Serra da Canastra National Park (SCNP), Brazil. We collected samples in 31 plots in 5 different soil classes and analyzed the following fuel traits: type (ground-layer cover composition), mean height, continuity, biomass, bulk density and moisture. In each plot, soil samples were collected for physical analysis. We also collected fire history, climate and topography data using the geographical coordinates of the plots. To test for correlations among fuel traits, we used a Pearson correlation test and a principal component analysis, and for abiotic factors we used generalized linear models. Fuel traits, such as bulk density, biomass, continuity, moisture content and ground-layer cover by dicotyledons (including leaf litter) were positively intercorrated. The variation in fuel height and continuity was explained by soil clay content and time since the last fire (positive correlation). This last fator also predicted fuel moisture content. Tree cover was positively correlated to dicotyledons and leaf litter ground-layer cover and was negatively correlated to fuel height. Mean anual precipitation showed a positive correlation to fuel continuity whereas rainfall seasonality showed a negative correlation to fuel height. Fire frequency was explained by fuel bulk density and mean annual precipitation (both associations were positive), explaining 83% of fire frequency variation in these plots. Our results demonstrated that fuel traits and environmental conditions show strong control over fire regimes and vary predictability in space.