Analisando as interações e combinando redes: plantas e polinizadores em uma área de Cerrado

• Ano de defesa: 2019
• Autor(a) principal: Rosa, Bianca Bonami
• 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: Análise de rede; Ecologia; Redes de polinização; Comunidade e ecossistema; Modularidade; Polinização; Ecossistemas; Network analyses; Pollination networks; Community and ecosystem; Modularity; CNPQ::CIENCIAS BIOLOGICAS::ECOLOGIA::ECOLOGIA DE ECOSSISTEMAS
• Link de acesso: https://repositorio.ufu.br/handle/123456789/27022; http://dx.doi.org/10.14393/ufu.di.2019.1335

The biodiversity loss for a long time was evaluated throught the loss of individual species, but community and ecossystem studies guided us to a broader viewpoint about environmental functions and the related species. The pollination is a function of extreme importance to the environment, and a more profound inquiry about the inherent interactions is important in order to understand the organization of the ecossystem. The research on interaction networks allows the preview and analyses of communities, but this type of research is not sampled enough outside determined functional groups, being the “construction” of complete networks a rather recente proposal, although focused on frugivory. The objective of this study was to reproduce this methodology to a pollination network in a Cerrado area, the Estação Ecológica do Panga (EEP), and importante research ground for Universidade Federal de Uberlândia, MG. Utilizing eight works about pollination and related interactions done in there, an adjacent matrix was built and used as a base to form the network. The analysed metrics were modularity, that by itself can demonstrate the complexity of the environment, comparing the main network to three smaler ones of the most relevant pollinators: Apidae, Throchilidae e Sphyngidae, and connectance. The R Languange was the one used to evaluate the modularity and Pajek Program was the one used to assemble the network. From the eight works, 247 species were compilated, of wich 109 are plant species while 138 are pollinators.The adjacente matrix that contains all the species harbors a modularity value of 0.507, describing it as partialy modulated network that subdivides as 6 modules, each one with a dominant group of pollinators and a niche overlap value of 0,078 and 0,083. The three separeted groups had a Q<0,350 and therefore are lowly or non modulated. Most species presented little interaction outside their own modules, however 15 conector species were identified in the network: the plants Caryocar brasiliense, Vochysia cinnamomea, Byrsonima intermedia, Vernonia polyanthes and Bidens gardneri, and the pollinizators Centris nitens, Aellopus fadus, Xanthepicharis bicolor, Megachile paulistana, four species of the genus Xylocopa and two from the genus Triepicharis. Through this work, it’s possible to visualize how a network analyses allows identifying the species with ecological restoration potential, and also that, albeit having na abundance of pollinators, the Panga Ecological Estation possess little niche overlap, and therefore, little functional redundance.