Research
I am interested in a variety of problems in theoretical ecology and evolution that are closely related at least in my own head if no where else. I enjoy questions of “Pasteur’s Quadrant”, basic science with clear policy relevance, and am fascinated with how we can better confront theoretical models with a rapidly expanding array of real world data. These interests couple to a long standing passion for mathematical biology, where I have always been interested in the role of nonlinear and stochastic dynamics – two themes that run throughout my work.
Though I began in more esoteric problems in stochastic population dynamics, that exploration slowly led me to the concept of regime shifts, which I have explored in both macro-evolutionary context through my work in comparative phylogenetics, and more recently in ecosystem shifts through my work in early warning signals.
The idea of predicting and avoiding ecosystem regime shifts has since led me more broadly into questions of ecological managment through the lense of decision theory. While decision theory has a long history as a quantitative approach to handling decision-making under uncertainty, the classical simplifications required rarely reflect either the real complexity of ecosystems or make best use of the data available. My postdoctoral work explores this in several projects, including nonparametric Bayesian approaches to management, quantifying the value of information under multiple uncertainty, and my part in a NIMBioS working group on Pretty Darn Good Control, seeking to replace optimal solutions of toy problems with merely decent solutions for more realistic scenarios.
Project Landing Pages
The menu to the right links to abstracts for most of my recent reseach projects. These pages also include automated feeds from my workflow, indicating active issues or next steps, recent notebook entries, current literature I’m reading on the topic, etc.
My Toolbox
- Analytic methods: Dynamical systems, stochastic calculus, probability, master equation/linear noise approximation, optimal control
- Numerical methods: High-performance computing, numerical differential equations, approximate likelihood methods, machine learning, stochastic dynamic programming.
- Data Science methods: Automating data extraction and data-mining from large data repositories (XML, regexp, semantic/linked data). Software development practices, authoring R packages, data management.
I also enjoy exposure to fieldwork whenever I can convince colleagues to take me along. Though my own research has no field component, I’m thankful for recent opportunities from surveying pitfall traps for tiger salamanders to helping excavate icthysaur fossils in the Nevada desert. Have a field site that needs an extra pair of hands? I have over eight years of experience as an alpine guide and can carry a shovel.