Diversidade e estrutura arbórea ao longo de um gradiente altitudinal na Floresta Tropical Pluvial Atlântica
Carregando...
Data
2020-10-09
Autores
Kamimura, Vitor de Andrade [UNESP]
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
A ecologia busca elucidar os processos e mecanismos responsáveis pelos padrões de diversidade, estrutura, montagem e funcionamento das comunidades. Neste trabalho, investigamos três questões fundamentais na ecologia de comunidades, utilizando um grande conjunto de dados da composição, demografia e atributos funcionais de comunidades arbóreas amostrados ao longo de um gradiente de elevação na Floresta Atlântica. No primeiro capítulo, utilizamos abordagens integrativas por meio de métricas da diversidade de espécies e filogenética sob diferentes arranjos de espécies (p.e. comunidade vs. famílias) para avaliar os processos ecológicos que conduzem as relações entre diversidade e elevação em florestas com mega diversidade. Encontramos um padrão de variação unimodal da diversidade da comunidade arbórea, emergindo da combinação dos diferentes padrões de diversidade de suas famílias com maior abundância. A diversidade filogenética local foi negativamente relacionada com a elevação e a diversidade regional (turnover) de espécies foi maior quanto maior a distância entre as comunidades. Estes resultados indicam a competição e a filtragem ambiental como principais processos nos mecanismos de montagem de comunidades ao longo de gradiente de elevação, e que a união das abordagens ecológica e evolutiva, sob diferentes arranjos de espécies, é fundamental para entender os processos que conduzem as relações entre diversidade e os gradientes ambientais. No segundo capítulo, além dos efeitos da variação de altitude, estudamos os efeitos da extração seletiva de madeira no padrão de diversidade e na estrutura filogenética de comunidades arbóreas, e avaliamos como a escala espacial pode alterar os resultados dessas análises. O corte seletivo e o filtro de altitude aumentaram a diversidade filogenética, apesar de conduzirem ao agrupamento filogenético. Os efeitos da altitude e da exploração madeireira foram mais fortes em larga escala, e os resultados foram alterados de acordo com os clados usados na análise. Dessa maneira, o filtro ambiental pode ofuscar o efeito da extração seletiva de madeira e, quando o filtro de elevação é excluído, os nichos abertos provavelmente por perturbações da exploração madeireira, levam a maior co-ocorrência de espécies filogeneticamente distantes. No terceiro capítulo, estudamos as relações entre os atributos funcionais e as taxas demográficas das espécies arbóreas considerando a importância de incluir interações entre os atributos, assim como o tipo florestal e o intervalo de tempo entre os censos (incluindo ou não um evento de seca), usando técnicas estatísticas inovadoras para melhorar o teste de premissas da análise funcional com um grande conjunto de dados. Aplicando o ‘aprendizado de máquina’ por meio de modelos de árvore de regressão impulsionados, descobrimos que a inclusão das interações entre atributos aumentou significativamente o poder dos modelos para prever as relações entre atributos e taxas demográficas, dependentes do tipo de vegetação e do intervalo de tempo. Assim, para avaliar a funcionalidade das comunidades vegetais, é importante considerar que pode haver múltiplos fenótipos com respostas demográficas semelhantes, resultados de interações entre atributos em diferentes habitats e intervalos de tempo. Concluindo, demonstramos que o uso de diferentes abordagens na ecologia de comunidades é fundamental para que possamos avaliar os processos ecológicos e mecanismos que conduzem a mega-diversidade de florestas tropicais e o funcionamento de suas comunidades.
A central point in ecology is to elucidate the processes and mechanisms responsible for the patterns of diversity, structure, assembly, and functioning of communities. Here, by using a large set of data on the composition, demography, and functional attributes of tree communities sampled along an elevation gradient in the Atlantic Forest, we investigated three fundamental questions across three chapters in the ecology of communities. In the first chapter, we used an integrated approach of species diversity and phylogenetic diversity under different species arrangements (e.g. community vs. families) to assess the ecological processes that drive the relationships between diversity and elevation in mega-diversity forests. We found a unimodal pattern of diversity variation for the tree community, emerging from the combination of different patterns of diversity of its most abundant families. The local phylogenetic diversity was negatively related to the elevation and the regional diversity (turnover) of species was greater as greater the elevation difference between communities. These results showed species competition and environmental filtering as the most important processes in the community's assembly mechanisms along an elevation gradient and that the union of ecological and evolutionary approaches, under different species arrangements, are fundamental to evaluate the processes that drive relationships between diversity and environmental gradients at each scale. In the second chapter, along with the effects of altitude variation, we studied the effects of selective logging on the diversity and phylogenetic structure of tree communities and evaluated how spatial scale change these results. Selective logging tends to increase phylogenetic diversity but leads to phylogenetic clustering. The effects of elevation and logging were stronger on large scale, and the results were changed according to the clades used in the analysis. Therefore, the environmental filter can overshadow the effects of selective logging, when the elevation-filter was excluded, it revealed open niche-spaces probably created by disturbance of logging, leading to the co-occurrences of distantly related species. In the third chapter, we studied the trait-demography rates relationships of tree species taking into account 16 trait-interactions, forest type, time-interval between censuses (including or not a drought event) under robust statistical techniques to improve testing functional ecology assumptions by means of large datasets. Using machine learning through boosted regression tree models, we found that the inclusion of the trait-interactions significantly increased the power of the models to predict the trait-demography rates relationships and that trait-interactions were dependent on vegetation type and on time interval. Thus, to assess the functionality of the attributes, it is necessary to incorporate the concept of multiple phenotypes with similar demographic responses, resulting from various trait-interactions in different habitats and time intervals. To sum up, we demonstrate that the use of different approaches over community ecology is crucial for assessing the ecological processes and mechanisms that lead to a mega-diversity of tropical forests and the functioning of their communities.
A central point in ecology is to elucidate the processes and mechanisms responsible for the patterns of diversity, structure, assembly, and functioning of communities. Here, by using a large set of data on the composition, demography, and functional attributes of tree communities sampled along an elevation gradient in the Atlantic Forest, we investigated three fundamental questions across three chapters in the ecology of communities. In the first chapter, we used an integrated approach of species diversity and phylogenetic diversity under different species arrangements (e.g. community vs. families) to assess the ecological processes that drive the relationships between diversity and elevation in mega-diversity forests. We found a unimodal pattern of diversity variation for the tree community, emerging from the combination of different patterns of diversity of its most abundant families. The local phylogenetic diversity was negatively related to the elevation and the regional diversity (turnover) of species was greater as greater the elevation difference between communities. These results showed species competition and environmental filtering as the most important processes in the community's assembly mechanisms along an elevation gradient and that the union of ecological and evolutionary approaches, under different species arrangements, are fundamental to evaluate the processes that drive relationships between diversity and environmental gradients at each scale. In the second chapter, along with the effects of altitude variation, we studied the effects of selective logging on the diversity and phylogenetic structure of tree communities and evaluated how spatial scale change these results. Selective logging tends to increase phylogenetic diversity but leads to phylogenetic clustering. The effects of elevation and logging were stronger on large scale, and the results were changed according to the clades used in the analysis. Therefore, the environmental filter can overshadow the effects of selective logging, when the elevation-filter was excluded, it revealed open niche-spaces probably created by disturbance of logging, leading to the co-occurrences of distantly related species. In the third chapter, we studied the trait-demography rates relationships of tree species taking into account 16 trait-interactions, forest type, time-interval between censuses (including or not a drought event) under robust statistical techniques to improve testing functional ecology assumptions by means of large datasets. Using machine learning through boosted regression tree models, we found that the inclusion of the trait-interactions significantly increased the power of the models to predict the trait-demography rates relationships and that trait-interactions were dependent on vegetation type and on time interval. Thus, to assess the functionality of the attributes, it is necessary to incorporate the concept of multiple phenotypes with similar demographic responses, resulting from various trait-interactions in different habitats and time intervals. To sum up, we demonstrate that the use of different approaches over community ecology is crucial for assessing the ecological processes and mechanisms that lead to a mega-diversity of tropical forests and the functioning of their communities.
Descrição
Palavras-chave
Aprendizado de máquinas, Árvores, Corte seletivo de madeira, Demografia, Diversidade, Ecologia de comunidades, Ecologia evolutiva, Ecologia funcional, Floresta Atlântica, Floresta tropical, Gradiente de elevação, Atlantic forest, Community ecology, Demography, Diversity, Elevation gradient, Evolutionary ecology, Functional ecology, Machine learning, Selective logging, Tree, Tropical forest