The problem of scale and variation of biological processes remain two fundamental hurdles to ecological understanding. Because many biological processes are variable and non-linear, our ability to explain ecological processes quickly breaks down when we scale up. While the development of scale transition theory in the past two decades has provided a powerful mathematical framework to unite disparate understandings at different biological scales (Chesson 2012, Denny & Benedetti-Cecchi 2012), other less understood emergent phenomena still cause theoretical predictions and empirical data to contradict.
Plant defense against herbivory has historically been dominated by mean-field theories, until a growing number of recent studies demonstrating the protective effect of trait variation (Herrera 2009, Wetzel et al. 2016). Yet, much remains unknown about some of the most basic biology before we can scale up to understand individual performance and population dynamics: How much variation is there? How does variation propagate across scales? What is the mean-variance relationship? What are the shapes of important functions (herbivore performance curves, herbivory-plant fitness curves)? How do herbivores sample plant trait variation? What is the extent of temporal and spatial autocorrelation? What is the relevant scale of different biological processes?
I investigate some of these questions using a combination of data synthesis, experiments, and mathematical modeling. More recent experimental work involves painting exogenous glucosinolate on Arabidopsis to examine the effect of plant toxin variation at different scales on plant fitness, herbivory distribution, and herbivore performance. I am also looking into using artificial diet to manipulate the variation and spatial arrangements of different food qualities. Other modeling work include exploring the emergent macro-patterns of phytochemical diversity, borrowing ideas from community ecology, statistical physics, and macroecology.