Mostrar registro simples

dc.contributor.advisorMelo, Gabriel A. R. (Gabriel Augusto Rodrigues de)
dc.contributor.authorAmaral Neto, Laércio Peixoto do
dc.contributor.otherVarassin, Isabela Galarda
dc.contributor.otherUniversidade Federal do Paraná. Setor de Ciências Biológicas. Programa de Pós-Graduação em Ciências Biológicas (Entomologia).
dc.date.accessioned2015-12-04T16:13:00Z
dc.date.available2015-12-04T16:13:00Z
dc.date.issued2015
dc.identifier.urihttp://hdl.handle.net/1884/40489
dc.descriptionOrientador : Prof. Dr. Gabriel A. R. Melo
dc.descriptionCo-orientadora : Profa. Dra. Isabela G. Varassin
dc.descriptionTese (doutorado) - Universidade Federal do Paraná, Setor de Ciências Biológicas, Programa de Pós-Graduação em Ciências Biológicas (Entomologia). Defesa: Curitiba, 20/03/2015
dc.descriptionInclui referências : f.87-91
dc.descriptionÁrea de concentração : Entomologia
dc.description.abstractResumo: A análise de redes mutual ísticas é uma ferramenta extremamente útil para a compreensão do nível de interação entre os di ferentes subsistemas existentes nos sistemas ecológicos e provê ferramentas que possibilitam a avaliação conjunta de um número grande de variáveis, permitem extrapolar propriedades da rede a sistemas ecológicos e possibili tam extrapolações e predições. Apesar de suas vantagens, alguns trabalhos alertam sobre alguns entraves que podem causar distorções na interpretação dos padrões observados nas redes, entre eles a presença de pilhadores, a variação temporal das interações e a premissa de que todas as interações são igualmente possíveis de acontecer. O presente trabalho explora o efeito destes entraves na estrutura e propriedades de uma rede de interações abelha-planta. As análises foram feitas a partir dos dados coletados ao longo de dois anos em uma área de campo natural no Parque Estadual de Vila Velha, Ponta Grossa, Paraná. A área foi visitada duas vezes por mês onde foram observadas as interações entre abelhas e plantas atentando-se para os seguintes aspectos: (1) recurso floral utilizado; (2) se o visitante tocava os órgãos reprodutivos da flor; (3) se o mecanismo floral era ativado e; (4) se havia congruência entre o comportamento do visitante e a morfologia da flor. Os visitantes foram então classificados como pilhadores ou polinizadores. Foram registradas também características das flores tais quais: tipo morfológico, cor, simetria, tamanho da flor (profundidade do recurso, altura), recursos disponíveis e coletados e características referentes ao tamanho e comportamento dos visitantes. Esses dados foram usados para confeccionar matrizes baseando-se em diferentes regras explicadas com detalhes em cada capítulo. O trabalho consta com três capítulos. O primeiro aborda como os polinizadores e pilhadores compartilham os recursos florais. Uma vez que é esperado que os pilhadores diminuam o recurso disponível para os polinizadores e que estes sejam competidores, o objetivo deste capítulo é avaliar se de fato ocorre uma sobreposição de nich o entre as duas funções de visitantes, tanto no sentido das espécies visitadas como nas características das flores. Esta sobreposição de nicho foi avaliada também temporalmente (em cada janela temporal). Nossa hipótese é que quando as visitas de pilhadores e polinizadores são observadas em uma perspectiva de redes, estes dois grupos de visitantes não apresentam uma grande sobreposição de nicho devido à disponibilidade de outros recursos. Os resultados mostraram que polinizadores e pilhadores não apresentam uma grande sobreposição de nicho, sobretudo quanto ao tipo floral e recursos utilizados pelos visitantes . Quando o aspecto temporal é considerado nota-se que apenas uma pequena proporção das espécies de plantas disponíveis é compartilhada. Foi avaliada ainda at ravés de modelos a capacidade de certas combinações de características florais preverem a estrutura da rede baseada na função dos visitantes florais (uma rede onde as espécies de abelhas foram substituídas por sua função, polinizador ou pilhador). Nenhum dos modelos testados foi capaz de predizer a estrutura desta rede função-planta. O segundo capítulo t raz de maneira mais di reta uma análise sobre os efeitos em propriedades da rede quando se excluem das análises as visitas de abelhas que não prestam serviços de polinização, ou seja, as espécies pilhadoras. Para isso foram construídas redes de interação com e sem pilhadores para cada janela temporal em que se observou a presença de abelhas que interagiam sem prestar serviços de polinização. Frente aos resultados do primei ro capítulo (que os pilhadores e polinizadores não compartilham muitos dos recursos disponíveis), partiu-se da hipótese de que a retirada dos pilhadores teria pouco efeito nas propriedades da rede. Apenas uma pequena proporção de todas as interações observadas é devida às interações entre plantas e pilhadores (<10%). Entretanto, a retirada dos pilhadores resulta em redes com um menor número de espécies, tanto de plantas como de abelhas, menor valor de aninhamento e menos especialização média (d') das abelhas. A compartimentalização e índice de especialização das redes (H2') não variaram significativamente, mas foram detectados diferentes efeitos quando cada janela temporal foi analisada separadamente. Esse efeito está relacionado com a posição ocupada pelas interações dos pilhadores na rede. O terceiro capítulo indaga se a complementaridade fenotípica e as interações proibidas são capazes de predizer as propriedades da rede. Para essa avaliação foram usados modelos onde: (1) as interações entre as plantas e polinizadores eram apenas limitadas pelo encaixe fenológico e morfológico entre as características das flores e das abelhas (nesse caso, um desacoplamento em qualquer das características impede a formação de uma interação e, por isso, essas interações são chamadas "proi bidas") ou; (2) pela complementaridade fenot ípica (nesse caso eram levadas em consideração as preferências das abelhas por certas características das flores e a falta de complementaridade não impede que a interação ocorra) . Uma vez que os grupos funcionais de visitantes florais são definidos baseando-se em várias características, assim como o são as síndromes de polinização complementares a estes grupos funcionais, esperou-se que quanto mais características fossem usadas, mais precisas seriam as predições dos modelos. Baseado nos resultados dos outros capítulos, esperou-se que algumas combinações de características fossem mais efetivas na predição da estrutura das redes, como os recursos coletados, tipo floral e cores. Os resultados mostram que modelos de complementaridade fenotípica são melhores que os modelos que se utilizam de interações proibidas. Também foi observado que combinações de características sempre possuem um melhor desempenho que características isoladas. Modelos que levam em conta apenas a complementaridade fenológica são ruins, embora esta característica aumente o poder de predição dos outros modelos quando usada conjuntamente. De manei ra geral essa tese apresenta alguns efeitos e traz discussões sobre dois aspectos que podem interferir na interpretação dos resultados em estudos de redes: A presença de pilhadores e as interações proibidas. Apesar dos estudos de redes possuírem a vantagem de permitir generalizações sobre um sistema e a análise de grandes conjuntos de dados, o estudo mais detalhado das interações envolvidas pode t razer respostas em que apenas os padrões e mét ricas da rede não seriam suficientes. No que se refere aos pilhadores, este foi o primeiro trabalho com redes em um aspecto mais amplo para testar seu efeito. Trabalhos prévios focam observações centradas em uma espécie de planta ou uma família de plantas. Foi demonstrado que, embora estas representem apenas uma pequena proporção de interações e que compartilhem poucas espécies com os polinizadores, a exclusão de interações entre pilhadores e flores altera algumas propriedades da rede. Mesmo as métricas que parecem não ser afetadas pela exclusão dos pilhadores sofrem interferências variáveis quando analisadas em menores escalas temporais, mostrando a importância da distinção deste tipo de comportamento nos estudos de redes. O efeito das interações proibidas é um tema recorrente em trabalhos de redes atuais onde se avalia a complementaridade fenológica e morfológica das espécies que interagem como determinantes dos padrões observados. No entanto o presente estudo tem como principal novidade o uso de uma quantidade comparativamente maior de características morfológicas de ambos os lados, abelhas e plantas, além de inclui r a complementaridade fenotípica. Concluiu-se que as características morfológicas podem ter uma importância maior quando mais características são incluídas, o que possivelmente deve-se ao fato de proporcionar uma maior descrição do encaixe funcional entre as espécies. Palavras-Chave: Melitofilia; acoplamento fenológico; interações mutual ísticas; métricas de rede; pilhadore; polinizadores; interações proibidas.
dc.description.abstractAbstract: The analysis of mutualistic networks is an extremely useful tool for understanding the level of interaction between existing subsystems of ecological systems and provides tools that enable evaluation of a large number of variables, extrapolation the network properties to ecological systems and predictions about evolution of the ecological systems. Despite its advantages, some studies warn of some barriers that may distort the interpretation of patterns observed in the networks, including the presence of robbers, the temporal variation of the interactions and the assumption that all interactions are possible to happen. This paper explores the effect of these barriers in the structure and properties of a network of bee-plant interactions. Analyses were made from data collected over two years in a natural field area in the Vila Velha State Park, Ponta Grossa, Paraná. The area was visited twice a month where they were observing the interactions between bees and plants paying attention to the following aspects: (1) used floral resources; (2) if the visitor played the reproductive organs of the flower; (3) if the mechanism was activated and floral; (4) whether there was coherence between visitor behavior and morphology of the flower. The visitors were then classified as robbers or pollinators. It was also recorded characteristics of such flowers which: morphological type, color, symmetry, flower size (depth of resource, height), available resources and collected and characteristics relating to the size and behavior of visitors. These data were used to produce matrices based on different rules explained in detail in each chapter. This work comprises three chapters. The first addresses how pollinators and robbers share the floral resources. Since it is expected that the robbers reduce the resource available to pollinators and they are considered competitors, the purpose of this chapter is to assess whether in fact there is a niche overlap between the two functions of visitors, both in the sense of species visited as in characteristics of flowers. This niche overlap was also evaluated temporally (for each temporal window). Our hypothesis is that when the robbers' and pollinators' visits are observed in a network perspective, these two groups of visitors do not have a large niche overlap because of the availability of other resources. The results showed that, especially as the floral type and resources used by visitors, pollinators and robbers do not have a large niche overlap. When the temporal aspect is considered noted that only a small proportion of species of plants available is shared. It was also assessed through models the ability of certain combinations of floral traits predict the structure of the network based on the function of floral visitors (a network where the bee species were replaced by their function, pollinator or nectar thief). None of the models tested was able to predict the structure of this function-plant network. The second chapter brings more directly an analysis of the effects on the network properties when it excludes from the analysis the visits of robbers. For this, interaction networks were built with and without robbers for each time window in which it was observed the presence of bees interacting without paying pollination services. Based on the results of the first chapter (the robbers and pollinators do not share many of the available resources), it started with the assumption that the withdrawal of the robbers would have little effect on the network properties. Only a small proportion of all the interactions observed are due to interactions between plants and robbers (<10%). However, the removal of the robbers results in networks with a smaller number of species, both plant as bees, lower nesting less-skilled and (d ') of bees. The compartmentalization and networks specialization index (H2 ') have not varied significantly, but were detected different effects when each time window was analyzed separately. This effect is related to the position occupied by the interactions of the robbers on the network. The third chapter asks whether the trait complementarity and forbidden links are able to predict the properties of the network. For this evaluation were used models where: (1) the interactions between plants and pollinators were only limited by phenological and morphological fit between the characteristics of flowers and bees (in this case a decoupling in any of the characteristics prevents the formation of an interaction and therefore, these interactions are called " forbidden links "), or; (2) the trait complementarity (in this case are taken into account the preference for certain characteristics of the bees from flowers and the lack of complementarity does not prevent the interaction occurs). Since the functional groups of flower visitors are defined based on various characteristics, as are the complementary pollination syndromes to these functional groups, it was expected that the more characteristics were used, they would be the most accurate predictions of the model. Based on the results of the other chapters, it was expected that some characteristics combinations were more effective in predicting the structure of the networks, as the collected resources, floral type and colors. The results show that trait complementarity models that are better than forbidden links models. It was also noted that combinations of characteristics always have a better performance than isolated features. Models that take into account only the phenological complementarity are the worst, although this feature increases increases the predictive power of other models. Overall this thesis presents some effects and shed some light on two aspects that can interfere with the interpretation of the results in network studies: The presence of robbers and prohibited interactions. Although networks has the advantage of allowing a system and generalizations about the analysis of large data sets studies, more detailed study of the interactions involved can provide responses that only patterns and network metrics would not be enough to respond. With regard to robbers, this was the first work with networks in a broader aspect to test its effect. Previous studies focused observations centered on a species of plant or plant family. It has been shown that although these represent only a small proportion of interactions and share few species with pollinators, excluding interactions between robbers and flowers amending certain network properties. Even metrics that do not seem to be affected by the exclusion of robbers suffer variable interference when analyzed in smaller time scales, showing the importance of distinguishing this type of behavior in studies of networks. The effect of forbidden links is a recurring theme in interaction networks where it is evaluated the phenological and morphological complementarity of species that interact as determinants of the observed patterns. However, the present study has as main novelty the use of a comparatively larger amount of the morphological characteristics of both sides, bees and plants, and includes trait complementarity. It was concluded that the morphological characteristics may have an increased importance as more features are included, which is probably due to the fact that a further description of the functional fit between species. Keywords: melittophily; phenological coupling; mutualistic interactions; network metrics; robbers; pollinators; forbidden links.
dc.format.extent120 f. : il., tabs.
dc.format.mimetypeapplication/pdf
dc.languagePortuguês
dc.relationDisponível em formato digital
dc.subjectEntomologia e malacologia de parasitas e vetores
dc.subjectPolinização por abelhas
dc.subjectInteração animal-planta
dc.titleInferências sobre uma rede de interações abelha-planta : investigando o papel de polinizadores e pilhadores e a influência da complementaridade fenotípica e de interações proibidas sobre a estrutura e propriedades da rede
dc.typeTese


Arquivos deste item

Thumbnail

Este item aparece na(s) seguinte(s) coleção(s)

Mostrar registro simples