Investigating Neutral and Niche Contributions to Species Dominance and Rarity in the Amazon Basin, With a Focus on the Brazil Nut Clade (Lecythidaceae)

Abstract

Why are certain species common and others rare? This simple question motivates extensive research in the field of ecology. The universal pattern of few common and many rare species in a community is especially notable among Amazonian trees, where only 1.4% of species make up half of all individuals. These abundant species are known as “hyperdominant.” There are likely various ecological and evolutionary causes of their demographic success depending on spatial, temporal, and taxonomic scale of interest, however in many cases these causes remain unknown. This dissertation investigates patterns of species commonness and rarity at multiple spatial scales in the Amazon, with a focus on hyperdominance. After introducing the research in chapter one, chapter two provides a thorough investigation of species-level demographic differences within the environmentally, economically, and culturally important Lecythidaceae (Brazil nut) family. Lecythidaceae contains a disproportionately high number of hyperdominant species given its richness and contains some of the most dominant species in the Amazon. I generated demographic data with a large forest census and asked how species-specific demographic rates and biomass are influenced by species- and individual-level traits and habitat, and quantified species topographic niche. I document very slow growth rates and high mortality, but stable plot biomass. High-moisture habitats experienced high mortality and a decrease in biomass, suggesting that flooding is a greater risk to individuals than is drought. There were no overlapping habitat associations among sister species, suggesting habitat associations provide a niche partitioning mechanism among Lecythidaceae. In chapter three, I test the ability of a null model – Hubbell’s Neutral Theory – to account for hyperdominance. I ask whether Amazon-wide dominance is truly “hyper,” or could be explained by demographic drift given random birth, death, and speciation. The results suggest that abundances of hyperdominant species and the strength of hyperdominance cannot be explained by neutral processes alone. I verified that the global dispersal neutral model provided a conservative test of hyperdominance and demonstrated the results to be robust to taxonomic error in the empirical data. Together, these results strongly suggest deterministic processes play a significant role in Amazonian hyperdominance. Chapter four considers a non-neutral explanation for Lecythidaceae hyperdominance at the regional scale. I investigate how demographic traits contribute to dominance, and ask whether dominant Lecythidaceae deviate from conventional demographic trade-offs, maximize multiple vital rates, or exhibit a particular plant functional type, or. I find that hyperdominant species maximize recruitment and growth at the expense of long-term survival, and they demonstrate a higher average rank across all demographic rates than non-dominant species. They do not, on average, deviate from either of the two classic demographic trade-offs, although one particularly dominant species deviates from both. This dissertation demonstrates that hyperdominance is not a function of demographic drift, and that there are strong associations between demographic strategy and Lecythidaceae dominance. Chapter five concludes with a discussion of unexamined causes of hyperdominance and future directions, notably consideration of mechanistic drivers of demographic differences. Hyperdominant species contribute disproportionately to tropical ecosystem functioning, thus understanding their ecology is integral in the face of unprecedented human-induced change to the Amazon and the world.