Estruturação de redes de interação planta-planta em gradientes de severidade abiótico

Abstract

Ecological interactions are important mechanisms in the composition and dynamics of communities. A community is composed of species that interact with each other and can be graphically represented as a network. Through the structural properties of the network: nesting, modularity, and connectivity, it is possible to detect patterns of organization of interactions, as well as verify how they change in response to variations in environmental factors. However, such metrics have been used mainly to detect patterns of community structuring between different trophic levels. According to the Stress-Gradient Hypothesis: SGH, positive interactions predominate in the extreme of the gradient of greater severity, while negative interactions predominate in the extreme of the gradient of lesser abiotic severity. In more severe environments, facilitation is the predominant positive interaction, while in milder abiotic conditions, competition predominates. However, there are some controversies on the generalization of GHS, especially those regarding the type of gradient analyzed. Therefore, we applied unipartite network metrics to verify how the patterns of interspecific plant-plant interactions vary in six locations distributed along an altitudinal gradient in the Brazilian tropical semi-arid domain. We analyzed the structural properties of plant-plant interaction networks: modularity, nesting, and connectivity and the relationship with climatic and edaphic variables through multiple regression analysis and minimal model choice. We then analyzed how the functional properties of species vary: degree, centrality, connectivity within and between modules, and module and network hubs along the gradient. In areas with a lower aridity index (drier), the net showed greater nesting, indicating that facilitation is the predominant mechanism in the structuring of the community. While in places with a higher aridity index (more humid), the network structure showed greater modularity, indicative of the predominance of competitive interactions. Nesting in places of greater water stress and modularity under conditions of greater water availability confirms SGH predictions. However, if we add the soil fertility component, we do not confirm the SGH: organic carbon and bases. There is a higher organic carbon content in the wetter places, provided by litter decomposition due to higher humidity, but a lower sum of bases due to greater soil leaching. Therefore, if we use the aridity index, the structure is explained by SGH, but if we analyze only based on soil fertility data, we refute SGH. Based on this, we confirm that the controversies in the literature about SGH are due to the type of gradient analyzed. Considering that in a semi-arid tropical climate, the primary abiotic filter is water availability, even to solubilize soil nutrients thatplants absorb, our results confirm the SGH. Furthermore, we verified that each network's composition and dominant species play distinct functional roles as climatic and edaphic factors change. In areas of lower water availability, there is lesser richness, and generalist species predominate. As water availability increases, species richness increases, and species tend to be peripheral, with narrower niches. Based on these results, we can infer that the anthropogenic climate crisis in the Brazilian semi-arid region, whose forecasts are for an increase in temperature and a 30% reduction in precipitation by the end of the 21st century, will result in a reduction in the diversity of peripheral species and a predominance of generalist species. In very severe gradients, research has shown that the competition mechanism prevails at both extremes, with low richness and a predominance of competition in the extremes of aridity and high richness and a predominance of competition in the extreme with the greatest supply of resources. As we do not have very humid climates in the Brazilian semi-arid region, potentially, the current wetter areas where facilitation predominates will lose species, and communities will be structured by competition, with a predominance of peripheral species. The current wetter areas and structured by competition will become drier, will lose peripheral species, and will start to be structured by generalist species, where the facilitation mechanism will predominate, with broader niches. Future research on the flower and fruit resources that will potentially be lost will indicate the potential harm to flower visitors and frugivores/dispersers. The thesis was structured in 4 chapters: 1) Introduction/theoretical review on the theme of the thesis; 2) Does the stress gradient hypothesis explain the plant-plant interaction networks in an edaphicclimatic gradient? 3) Variations in the ecological function of species in a resource gradient and 4) Summary and general conclusions of the thesis.