Ant diversity patterns at local, regional and global scales
Biogeography; Communities; Tree richness; Disturbance; Habitat loss
Biodiversity patterns are affected by factors such as climate, human disturbances and biotic interactions, which operate at different spatial scales. Understanding diversity patterns and their determinants is essential for effective conservation efforts. However, many regions and taxonomic groups are often overlooked due to a lack of adequate data. In particular, invertebrates and arid/semiarid regions receive less attention. For example, ants play a fundamental role in ecosystem functioning, but there are knowledge gaps regarding their global biodiversity patterns, their response to anthropogenic disturbances and their relationship with plant composition and diversity. Similarly, the Caatinga, a species-rich dry forest, lacks clear knowledge about its animal diversity patterns. Therefore, in this thesis we set out to investigate the determinants of ant diversity at three different scales: (i) global, testing the effect of climate and topography, (ii) regional, testing the effect of anthropogenic disturbances and (iii) local, testing the effect of tree species richness and their facilitation capacities. In addition, (iv) we examined the determinants of animal diversity in the Caatinga and the correlation of diversity distribution between different animal taxa. In the first chapter we tested the relative strength of abiotic gradients in determining spatial patterns of ant diversity around the world and between biogeographic realms. We used multiple regressions to relate global ant diversity data to climatic and topographic data. Our findings suggest that ant diversity is higher in tropical regions, especially in areas with higher annual rainfall and higher average temperature. However, the abiotic determinants between biogeographic realms varied greatly, which challenges the generality of the global pattern. This highlights the importance of considering regional historical and ecological contexts when investigating biodiversity patterns. In the second chapter we tested the influence of chronic anthropogenic disturbances (e.g. logging and grazing) and habitat loss on ant species richness (total, specialist and generalist species). We then used multiple regressions to examine the impact of disturbance vectors on ant diversity patterns. We found that habitat loss reduces species richness at all levels, including specialist and generalist species, while chronic disturbance has a particularly strong impact on species overall and on specialist species. Our findings highlight that it is necessary to include the drivers of chronic disturbance in order to measure our real impact on biodiversity. The third chapter examines the impact of tree richness, tree facilitation and the presence of plants with extrafloral nectaries on ant diversity patterns, including species richness, phylogenetic diversity and abundance. We collected ants in all 155 plots of the BrazilDry project, a biodiversity experiment established in the Açu National Forest. The plots contain one, two, four, eight and 16 plant species with varying compositions and a gradient of facilitation. Our results suggest that plant diversity is the main mechanism responsible for the richness and phylogenetic diversity of ants in the BrazilDry experiment. Possibly the result comes from a bottom-up effect in the trophic chain, in which plants provide essential resources for the ants, increasing their diversity. In chapter four, we examine the correlation between the distribution of animal diversity in the Caatinga among five taxonomic groups: birds, reptiles, amphibians, mammals and ants. We also tested the influence of abiotic factors on the distribution of diversity within these groups. To do this, we retrieved species distribution maps and climate variables from online databases. Pearson's correlations were used to test the overlap of diversity between the groups, and multiple linear models were used to test the relative importance of abiotic factors in organizing the diversity of these groups. Our findings indicate that animal biodiversity in the Caatinga largely overlaps, but the response to abiotic conditions varies according to the taxonomic group analyzed. With this thesis, we have filled important gaps in ant diversity, distribution, abiotic determinants and responses to tree richness. In addition, we describe the abiotic determinants and the degree of overlap of animal diversity in the Caatinga.