In standard dynamic models for biological populations and communities (Lotka-Volterra, Kermack-McKendrick, etc.) the state variables are the numbers of individuals in different populations or life stages. Parameters are constant, or externally forced (e.g., periodic). In the last few decades, ecologist have learned that in fact Darwinian evolution of the parameters governing ecological interactions is often fast, with substantial change within a few generations (or less). My talk will center on laboratory experiments using microorganisms as a model system to study rapid evolution in predator-prey interactions. Low-dimensional models for coupled ecological and evolutionary dynamics have been successful at explaining the main features of the dynamics, and at making predictions that were confirmed by subsequent experiments. Relaxation oscillations and canards are common in the models, and much can be learned by treating evolution as the fast variable in a slow-fast system. This is joint work with Nelson Hairston, Jr. (E&EB) and many students and postdocs.