Phenotypic and optical plant trait variation across space and time in the seasonal tropics: patterns, drivers and consequences
Carregando...
Data
2018-12-14
Autores
Streher, Annia Susin [UNESP]
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
O objetivo geral dessa tese foi explorar os conceitos da ecologia baseada em atributos a fim de entender como o ambiente modula a variação de atributos funcionais em diferentes escalas espaciais e temporais, combinando sensoriamento remoto e ecologia vegetal. No segundo capítulo da tese, avaliou-se os padrões de fenologia remota do mosaico de vegetação na montanha do Espinhaço, investigando quais os agentes ambientais são responsáveis pelos padrões observados. O algoritmo TIMESAT foi utilizado para extrair os indicadores fenológicos de uma série temporal de 14 anos de imagens de satélite MODIS/NDVI. A disponibilidade de água e luz, modulada pela topografia, foram os principais responsáveis pelas respostas da fenologia remota na região, determinando o início, fim e comprimento da estação de crescimento. A temperatura teve um papel importante na determinação das taxas de desenvolvimento das folhas e na força da sazonalidade da vegetação. No capítulo três, testou-se a generalidade dos padrões globais de estrutura e função da vegetação na escala da paisagem, na porção sul do Espinhaço, conhecida como Serra do Cipó em Minas Gerais. Também buscou-se determinar a relação dos atributos funcionais que representam as dimensões de estrutura e função da vegetação com os gradientes de elevação e topo-edáficos encontrados na região. Foram coletadas características funcionais (LMA, LDMC – relacionados à economia foliar, área foliar e altura da planta – relacionados com a estrutura da vegetação) de 1650 indivíduos, compreendendo todas as formas de vida, em cinco locais ao longo de um gradiente de elevação, abarcando os diferentes tipos de vegetação encontrados em cada elevação. A organização fenotípica na estrutura e função das plantas encontrada em escalas globais foi análoga à encontrada entre as espécies que co-ocorrem localmente nos campos rupestres. A dimensão relacionada a estrutura da vegetação apresentou variação ao longo dos gradientes ambientais, porém as relações alométricas foram igualmente importantes para explicar as variações encontradas nessa dimensão funcional. A dimensão fenotípica relacionada a função foliar não apresentou variação relacionada a nenhum dos gradientes ambientais avaliados, indicando que os atributos funcionais, LMA e LDMC, não estão relacionados com as estratégias de aquisição e uso dos recursos em ambientes sazonalmente secos. No capítulo quatro, utilizou-se os dados funcionais de folhas coletados no capítulo três, para testar a capacidade da espectroscopia de solo em estimar as características funcionais foliares e diferenciar plantas com diferentes formas de crescimento. Os espectros de reflectância foliar foram capazes de predizer com precisão as características funcionais de folhas da vegetação independente da forma de crescimento, porém os modelos apresentaram imprecisão em torno dos valores maiores de LMA. Este resultado aponta para uma limitação ou do método da espectroscopia e/ou do método de modelagem utilizado neste estudo. Gramíneas e plantas lenhosas apresentaram respostas espectrais mais dissimilares, enquanto as herbáceas representam um tipo espectral intermediário, parcialmente semelhantes as gramíneas (no visível) e parcialmente semelhantes às plantas lenhosas no infravermelho médio. No capítulo cinco foi investigada a influência dos níveis taxonômicos na relação entre a diversidade espectral e funcional em um subconjunto de plantas coletadas no capítulo três. A variação interespecífica foi maior que a variação intraespecífica para todas as características funcionais e espectrais da vegetação, mas o tamanho da influência intraespecífica foi uma resposta específica de cada espécie. Os resultados indicaram também que a idade foliar pode estar contribuindo mais que o esperado na variabilidade espectral intraespecífica e assim, dificultando o delineamento de um paralelo com os processos reconhecidos pela ecologia baseada em atributos que geram a variação intraespecífica (i.e., plasticidade). A partição da variância mostrou que tanto os atributos funcionais quanto os atributos espectrais variaram principalmente no nível de família, indicando que ambos são conservados evolutivamente. Este estudo contribui para a construção de teorias ligando a diversidade espectral com a diversidade funcional e taxonômica, os quais são muitas vezes difíceis de quantificar nos trópicos, auxiliando a impulsionar um sistema de monitoramento da biodiversidade baseado em sensoriamento remoto hiperespectral.
Here, I explore trait-based ecology to understand how the environment shapes plant trait variation at multiple scales, combining remote sensing technologies and plant ecology. In the second chapter, the patterns and drivers of land surface phenology were assessed for the Meridional Espinhaço Range in Brazil. The TIMESAT algorithm was used to extract the phenological indicators from a14-year time series of MODIS/NDVI satellite images. Water and light availability, modulated by topography, are the most likely drivers of land surface phenology in the region, determining the start, end, and length of the growing season, while temperature had an important role in determining the rates of leaf development and the strength of vegetation seasonality. In chapter three, I tested if the generalities of global patterns of plant form and function dimension held in finer scales in the seasonally dry tropics, and its relation with the an elevational and topo-edaphic environmental gradient. Leaf functional data (LMA, LDMC and leaf area) and plant height from 1650 individual comprising all life-forms of locally co-occurring plants was gathered, at five sites along an elevational gradient, sampling all vegetation types found within each elevation, at the southern portion of the Espinhaço range, known as Serra do Cipó, in Minas Gerais. The phenotypic organization of plant form and function found at global scales was similar to the one found among locally co-occurring species in campo rupestre. The whole-plant size dimension varied along the elevational gradient, however, the leaf economics dimension (LES) behaved differently than plant size, and no variation of key LES traits along environmental gradients was found. In chapter four, the ability of spectroscopy to estimate leaf functional traits and to differentiate plants comprising different growth forms was investigated. Leaf reflectance spectra was able to accurately predict leaf functional traits from different growth forms, but the models lost precision towards higher LMA values, pointing out a saturation point from spectroscopy and/or a limitation from the modelling approach adopted in this study. Grasses and woody plants were the most spectrally dissimilar, while forbs represented an intermediary spectral group. In chapter five the influence of taxonomic levels on the relationship between spectral and functional diversity was investigated. Interspecific variation was greater than intraspecific variation for functional and spectral traits, but the amount of intraspecific variation was a specific response of each species. Leaf age may be contributing more than expected to intraspecific spectral variability, hampering the delineation of a full parallel with trait-based ecology. The variance partition showed that both, functional and spectral traits varied mainly at the family level, indicating that both are evolutionary conserved. This study contributes to the construction of theories linking spectral to functional and taxonomic diversity, helping to build a biodiversity monitoring system based on remote sensing.
Here, I explore trait-based ecology to understand how the environment shapes plant trait variation at multiple scales, combining remote sensing technologies and plant ecology. In the second chapter, the patterns and drivers of land surface phenology were assessed for the Meridional Espinhaço Range in Brazil. The TIMESAT algorithm was used to extract the phenological indicators from a14-year time series of MODIS/NDVI satellite images. Water and light availability, modulated by topography, are the most likely drivers of land surface phenology in the region, determining the start, end, and length of the growing season, while temperature had an important role in determining the rates of leaf development and the strength of vegetation seasonality. In chapter three, I tested if the generalities of global patterns of plant form and function dimension held in finer scales in the seasonally dry tropics, and its relation with the an elevational and topo-edaphic environmental gradient. Leaf functional data (LMA, LDMC and leaf area) and plant height from 1650 individual comprising all life-forms of locally co-occurring plants was gathered, at five sites along an elevational gradient, sampling all vegetation types found within each elevation, at the southern portion of the Espinhaço range, known as Serra do Cipó, in Minas Gerais. The phenotypic organization of plant form and function found at global scales was similar to the one found among locally co-occurring species in campo rupestre. The whole-plant size dimension varied along the elevational gradient, however, the leaf economics dimension (LES) behaved differently than plant size, and no variation of key LES traits along environmental gradients was found. In chapter four, the ability of spectroscopy to estimate leaf functional traits and to differentiate plants comprising different growth forms was investigated. Leaf reflectance spectra was able to accurately predict leaf functional traits from different growth forms, but the models lost precision towards higher LMA values, pointing out a saturation point from spectroscopy and/or a limitation from the modelling approach adopted in this study. Grasses and woody plants were the most spectrally dissimilar, while forbs represented an intermediary spectral group. In chapter five the influence of taxonomic levels on the relationship between spectral and functional diversity was investigated. Interspecific variation was greater than intraspecific variation for functional and spectral traits, but the amount of intraspecific variation was a specific response of each species. Leaf age may be contributing more than expected to intraspecific spectral variability, hampering the delineation of a full parallel with trait-based ecology. The variance partition showed that both, functional and spectral traits varied mainly at the family level, indicating that both are evolutionary conserved. This study contributes to the construction of theories linking spectral to functional and taxonomic diversity, helping to build a biodiversity monitoring system based on remote sensing.
Descrição
Palavras-chave
Ecologia baseada em atributos, Sensoriamento remoto, Espectroscopia, Fenologia, Atributos funcionais foliares, Campo rupestre, Trait based ecology, Remote sensing, Spectroscopy, Phenology, Functional traits,