Although figs (genus Ficus) are a diverse and a common component of tropical and subtropical ecosystems, individual species often occur at extraordinarily low population densities, creating a challenge for successful fig wasp dispersal between hosts and pollination. Working in Central America, we have used genetic markers to show that, despite their small size and short life spans, fig wasp pollinators routinely disperse >10 km between host trees and that interbreeding populations of figs comprise many hundreds to thousands of individuals (Nason et al. 1996; Nason & Hamrick 1997; Nason et al. 1998). We have also used molecular methods to show that the rate of pollinator-mediated gene flow greatly exceeds that of bat- and bird-mediated seed gene flow (Yu et al. 2010), and that dispersal of fig seed, not pollen, has been the limiting factor in post-Pleistocene range expansion (Yu & Nason 2012). At the latitudinal limits of this mutualism, we have shown that within-crown reproductive asynchrony does not promote wasp cycling and self-pollination within trees, as previously though, but instead promotes reproductive overlap and outbreeding between trees (Gates & Nason 2012). In addition to pollinating wasps (mutualists), figs also host a diversity of non-pollinating fig wasps (antagonists). Recent work in our lab has focused on mechanisms promoting the coexistence of multiple mutualist and antagonist taxa completing their life cycles in patchy, ephemeral resources, like figs (Duthie et al. 2014). We are also investigating how spatial variation in coexistence is influenced by host density (Duthie & Nason, submitted) and fecundity-dispersal trade-offs in fig wasps (Duthie et al. in press.).