Dissertation Research: Disentangling drivers of community structure and composition of a tropical forest

Princeton University, Princeton NJ

Investigators

Abstract

This project seeks to disentangle and characterize drivers of plant and animal community structure and composition in one of the world's largest regenerating tropical forests. Determining why species are able to exist in a particular time and place (particularly the role of chance vs. inevitability) is simultaneously one of the most central tasks in ecology and a deeply pressing question for conservation in a rapidly changing world. Understanding these forces in regenerating tropical forest is a particularly timely given (1) the incredible biodiversity of tropical forests, (2) that the proportion of remaining forest classified as regenerating is high and ever-rising, and (3) such knowledge is a prerequisite for deploying ecological restoration as an effective tool for preserving threatened species and ecosystems. The project team will assess three ultimate drivers of community composition and structure: soil quality, landscape composition, and initial vegetative conditions. Using a network of vegetative plots, species composition of terrestrial mammal, bat, bird, and singing-insect assemblages will be assessed. Fundamental to this task is developing a set of tools that will allow for the rapid, cheap, and non-invasive assessment of these animal groups. To this end, ultrasonic and audible-frequency recorders will be deployed, and the research team is developing techniques that will allow for the semi-automated analysis of recordings using machine learning. To further assess vegetative structure and species interactions as proximate drivers of composition, arrays of synced recorders will be used to determine how key forest structural features and the distribution of some species affects the presence of others.