For about a decade, the Bronstein lab and many collaborators (notably Goggy Davidowitz) have (with the help of NSF) been studying an interaction between the perennial desert herb Datura wrightii and Manduca sexta. This insect is both the plant’s primary pollinator and its most damaging herbivore. We’re quantifying costs and benefits of the interaction to both partners, linkages between costs and benefits to each partner, and cross-linkages between costs to one partner and benefits to its mutualist. The results consistently surprise us. For example, we’re finding that the plants don’t need the plants nearly as much as we expected (they are highly self-compatible), while the same can be said about the moths (they rely heavily on a secondary nectar resource, Agave). At the same time, the costs aren’t as high as we expected either: for example, along with our collaborator Andy McCall, we are finding that the plants are remarkably tolerant to being consumed! The Datura/Manduca system is relatively simple biologically and easily manipulated. It’s therefore ideal for hypothesis-testing studies at levels of biological organization ranging from the individual to the ecosystem, using techniques taken from ecology (Alarcón et al. 2008, Bronstein et al. 2009), plant physiology (Bronstein et al. 2006, Barron-Gafford et al. 2012), insect physiology (Contreras et al. 2013) and even neurobiology (Riffell et al. 2008). Most of my major research questions – the costs of mutualism, the benefits of cheating, the role of community context, and the ecological and evolutionary dynamics of mutualism – can be explored using this fascinating interaction. As such, it’s a particularly attractive system for integrative studies that involve large numbers of undergraduate, graduate, and postdoctoral students.