Leão, Tarciso2021-01-132021-01-132018-10https://hdl.handle.net/11299/217786University of Minnesota Ph.D. dissertation. 2018. Major: Conservation Biology. Advisor: Peter Reich. 1 computer file (PDF); 165 pages.The understanding the process of plant species extinction in tropical species-rich regions is substantially incomplete. What makes a species vulnerable to extinction? Are there recognizable species attributes that can help us to predict vulnerability? What are the underlying reasons for high vulnerability? What are the relative contributions of the main drivers of extinction such as land use change and climate change? This doctorate research addresses these questions focusing on the Atlantic Forest region of Brazil. The Atlantic Forest combines an extraordinarily high plant diversity, high human presence, and a relatively well-known flora, making it a region where species extinction is likely to occur and possible to monitor. Chapter 1 seeks for correlates of extinction risk and insights on potential mechanisms underlying vulnerability to extinction. The chapter provides a first look on the broad ecological patterns in the extinction risk of plants in the Atlantic Forest. It helps to identify which plant groups are more threatened. It uses decision tree, standard regression, and phylogenetic regression to explore the relationships between species attributes and extinction risk. As expected, geographic range size was the main correlate of species extinction risk. Epiphytes (plants growing on other plants, commonly observed in the Atlantic Forest) and species restricted to rocky outcrops and sandy coastal plains (‘restinga’) were associated with highest extinction risk, and related to extinction risk even after controlling for the effects of geographic range size. Those findings support the notion that slow growth and adaptation to resource-limited environments is a reason for high vulnerability to extinction. Chapter 1 also shows that extinction risk depends on phylogenetic relationships (i.e. closely related species tend to have similar extinction risk). Therefore, phylogenetically related factors not included in the analysis were likely important to explaining further differences in extinction risk, raising interest to test emerging hypotheses that diversification rate and lineage age explains variation in species’ geographic range sizes and extinction risks. Chapter 2 focused on understanding the variation in species geographic range sizes, which is the single most important predictor of extinction risk. This chapter includes a larger and improved dataset relative to Chapter 1, including evolutionary attributes, a higher resolution phylogeny, double the number of herbarium records, double the number of species (13,785 species, i.e. 90% of angiosperms of the Atlantic Forest), and more complete information on ecological attributes. This dataset allowed to investigate the existence of evolutionary and ecological patterns in the variation of species’ geographic range sizes. The chapter shows that lineages with more species tend to have species with smaller range sizes. This trend was particularly strong when focusing on the proportion of species with vulnerably small range size. A two-fold increase in the number of species at the genus level was associated with an average increase of 40% in the number of species with vulnerably small range sizes. The evolutionary histories of the lineages, in particular the rates of speciation and extinction, seem to underlie these observed patterns. Chapter 2 also addresses associated issues that have been inconclusive in the scientific literature. It shows how the relationship between life form and range size depends on the vegetation type, which seems to clarify a long-held inconsistency in the literature. It challenges the notion that epiphytes tend to have larger range sizes than terrestrial plants, and suggests that the attributes and mechanisms leading to the rarity of epiphytes may apply to the lineages in which epiphytism evolved as a whole rather than to epiphytic species alone. Chapter 3 investigates the projected effects of two of the main drivers of plant habitat loss: land use change and climate change. Using habitat suitability modelling, the chapter shows that land use change is likely to have a larger and consistently negative impact on species area of suitable habitat, while climate change will tend to have a more variable and selective impact – varying from largely negative to largely positive depending on the species. Chapter 3 also led to the conclusion that small-ranged and epiphytic species are at particularly high risk of losing area of suitable habitat in the near future. Finally, this dissertation in its totality provides evidence and inference for recognizing patterns and understanding mechanisms behind the variation in plant species geographic range sizes and extinction risks. I hope the knowledge provided here contributes to improve our understanding of the problem of plant extinction in species-rich tropical forest regions.enThesis or Dissertation