Modelagem da relação entre fenologia foliar e hidráulica das plantas na floresta Amazônica: uma abordagem trait-based dos efeitos da precipitação reduzida
Carregando...
Arquivos
Data
2021-04-23
Autores
Sophia, Gabriela Martins
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
Modelos dinâmicos de vegetação global (DGVMs) vêm sendo desenvolvidos para melhor entender as respostas da vegetação às mudanças climáticas. No entanto, os DGVMs geram respostas variáveis uma vez que a vegetação é representada por poucos tipos funcionais de plantas. Modelos baseados na variabilidade dos atributos funcionais surgem como alternativa para representar melhor as estratégias de vida adotadas pelas plantas. Incluir a fenologia foliar nesses modelos é de suma importância pois a mesma é capaz de controlar a sazonalidade dos fluxos de carbono, água e energia. Em florestas tropicais como a Amazônia, a fenologia é impulsionada principalmente pela disponibilidade de água no solo. Portanto, simular os impactos de um clima mais seco requer a conexão entre fenologia e estratégias hidráulicas das plantas. Este trabalho contribuiu para o desenvolvimento do modelo CAETÊ com a implementação de um módulo de fenologia foliar vinculado ao sistema hidráulico das plantas. O desenvolvimento foi aplicado para todo o bioma Amazônico e teve como objetivo principal melhorar a representação da sazonalidade do índice de área foliar (LAI) com consequente melhoria no ciclo do carbono e água, a fim de avaliar os impactos da mudança do clima sobre ambos. Para tal, dois atributos funcionais foram utilizados como variantes: WD (densidade da madeira - determinando a hidráulica) e τfolhas (tempo de residência do carbono nas folhas - determinando a idade foliar). Equações já existentes em outros modelos de vegetação foram usadas e adaptadas ao modelo CAETÊ. O modelo foi aplicado com uma redução de 30% da precipitação. Como resultados, a inclusão do módulo de fenologia melhorou o desempenho do modelo em relação à sazonalidade do LAI nas regiões em que a água no solo é um fator limitante; bem como melhorou a representação da produtividade primária bruta do ecossistema (GPP) e proporcionou uma variabilidade espacial mais demarcada da região Amazônica. A nível de ciclagem biogeoquímica, uma redução na precipitação em 30% levou à uma redução na provisão de água no solo, na GPP, LAI e evapotranspiração (ET) para regiões menos úmidas da Amazônia; enquanto que para áreas mais úmidas não houve impactos evidentes, coerente às observações de que nessas regiões uma redução no regime de chuva não é suficiente para determinar os processos biogeoquímicos e LAI. Por fim, conclui-se de modo geral que a inclusão da fenologia foliar melhora o desempenho do modelo, mas que é necessário ajustes no novo módulo para que represente a sazonalidade também em regiões que não são determinadas pela água no solo.
Dynamic global vegetation models (DGVMs) have been developed to better understand the vegetation's response to climate changes. However, DGVMs generate variable responses since it is represented by a small set of plant functional types. Models based on the variability of functional traits appear as an alternative to better represent the plant life strategies. Including leaf phenology in these models is of paramount importance because it plays a role in controlling the seasonality of carbon, water, and energy fluxes. In tropical ecosystems, such as in the Amazon, phenology is mainly driven by soil water availability, so simulating the impacts of a predicted drier climate requires the representation of the connection between phenology and the hydraulic strategies of plants. This work contributed to the development of the CAETÊ trait-based model through the implementation of a leaf phenology module linked to plant hydraulic system. The development was applied to the entire Amazon biome and its main objective was to improve the representation of the seasonality of the leaf area index (LAI) with consequent improvement in the carbon and water cycle, and therefore to assess the impacts of climate changes on it. For this, two functional traits are being used as variants: WD (wood density - determining the hydraulic system) and τleaves (leaf carbon residence time - determining the leaf age). Equations already existing in other vegetation models were used and adapted to the CAETÊ model. The model was applied under a 30% reduction of precipitation. As a result, the inclusion of the phenology module improved the model's performance in relation to LAI seasonality in regions where soil water is a limiting factor; as well as improving the representation of the gross primary productivity of the ecosystem (GPP) and providing a more demarcated spatial variability of the Amazon region. At the level of biogeochemical cycling, a 30% reduction in precipitation led to a reduction in soil water availability, in GPP, LAI and evapotranspiration (ET) for less humid regions of the Amazon; while for more humid areas there were no evident impacts, consistent with the observations that in these regions a reduction in the rainfall regime is not sufficient to determine the biogeochemical and LAI processes. Finally, it is generally concluded that the inclusion of leaf phenology improves the performance of the model, but that adjustments to the new module are necessary so that it also represents seasonality in regions that are not determined by water in the soil.
Dynamic global vegetation models (DGVMs) have been developed to better understand the vegetation's response to climate changes. However, DGVMs generate variable responses since it is represented by a small set of plant functional types. Models based on the variability of functional traits appear as an alternative to better represent the plant life strategies. Including leaf phenology in these models is of paramount importance because it plays a role in controlling the seasonality of carbon, water, and energy fluxes. In tropical ecosystems, such as in the Amazon, phenology is mainly driven by soil water availability, so simulating the impacts of a predicted drier climate requires the representation of the connection between phenology and the hydraulic strategies of plants. This work contributed to the development of the CAETÊ trait-based model through the implementation of a leaf phenology module linked to plant hydraulic system. The development was applied to the entire Amazon biome and its main objective was to improve the representation of the seasonality of the leaf area index (LAI) with consequent improvement in the carbon and water cycle, and therefore to assess the impacts of climate changes on it. For this, two functional traits are being used as variants: WD (wood density - determining the hydraulic system) and τleaves (leaf carbon residence time - determining the leaf age). Equations already existing in other vegetation models were used and adapted to the CAETÊ model. The model was applied under a 30% reduction of precipitation. As a result, the inclusion of the phenology module improved the model's performance in relation to LAI seasonality in regions where soil water is a limiting factor; as well as improving the representation of the gross primary productivity of the ecosystem (GPP) and providing a more demarcated spatial variability of the Amazon region. At the level of biogeochemical cycling, a 30% reduction in precipitation led to a reduction in soil water availability, in GPP, LAI and evapotranspiration (ET) for less humid regions of the Amazon; while for more humid areas there were no evident impacts, consistent with the observations that in these regions a reduction in the rainfall regime is not sufficient to determine the biogeochemical and LAI processes. Finally, it is generally concluded that the inclusion of leaf phenology improves the performance of the model, but that adjustments to the new module are necessary so that it also represents seasonality in regions that are not determined by water in the soil.
Descrição
Palavras-chave
Ecologia, Mudanças climáticas, Biologia Modelos matemáticos, Fenologia vegetal, Ciclos biogeoquímicos, Ecology, Climatic changes, Biology Mathematical models, Cycles, Fenologia foliar, Hidráulica, Leaf phenology, Hydraulics, Biogeochemical cycles