Quantitative outcomes of insect herbivore-plant interactions typically vary spatially and temporally. Prediction of where and when insects will impose significant damage, or mediate indirect interactions among plants is elusive, but important for management of exotic species. We examine two novel insect -plant interactions to address: 1) relative effects of novel vs. co-evolved interactions on plant fitness and population dynamics; and, 2) roles of phenology and spatial variation in driving those effects. In case 1, interaction between Cirsium canescens and Rhinocyllus conicus is direct, and the insect is exotic. In case 2, the interaction between C. vulgare and insect herbivores is indirect, and mediated by proximity to its native conger, C. altissimum. In both cases, we found that the novel interactions had larger effects on plant survival, growth and fecundity than did similar, co-evolved ones. For both, spatial variation in the abundance of alterative host plants at site and landscape scales were important in explaining variation in plant damage. Further, in case 1, C. canescens geographic location influenced the magnitude of floral herbivory; and, degree of phenological overlap between insect and plant was critical to explaining variation in weevil oviposition loads. Also in case 2, the proximity of exotic C. vulgare to native C. altissimum determined its native insect load. We conclude that key processes driving variation in novel interactions are similar to those driving coevolved interactions. However, broader consideration of such processes is needed in assessing management options.