Students of speciation debate the role of performance trade-offs across different environments early in speciation. We tested for early performance trade-offs with a host shift experiment using a member of the Enchenopa binotata species complex of treehoppers (Hemiptera: Membracidae). In this clade of plant-feeding insects, different species live on different host plants and exhibit strong behavioral and physiological host specialization. After five generations, the experimental host shifts resulted either in no adaptation or in adaptation without specialization. The latter result was more likely in sympatry; in allopatry, populations on novel host plants were more likely to go extinct. We conclude that in the early stages of speciation, adaptation to novel host plants does not necessarily bring about performance trade-offs on ancestral environments. Adaptation may be facilitated rather than hindered by gene flow, which prevents extinction. Additional causes of specialization and assortative mating may be required if colonization of novel environments is to result in speciation.

"Quasinatural" selection experiment followed by reciprocal transplant rearing experiment; fecundity measured as # nymphs eclosing from eggs