Abstract
The enemy release hypothesis—that the success of invading species is due to release from natural enemies that occur in their home range but not in the new range—can be an important explanation for successful invasions. Testing this hypothesis is difficult, however, because testing requires documented cases of not only successful but also unsuccessful invasions. Therefore, observational data on after-the-fact establishment following unintentional introductions is insufficient, because they only include successful invasions. Here, we investigate the role of enemy release for parasitoid biological control agents and their hyperparasitoid natural enemies. We created a dataset of all parasitoid wasp species attacking aphids that were introduced into North America, including data on their hyperparasitoid species in the native and new range. In total, information on 29 species of primary parasitoids and 54 species of hyperparasitoids was obtained, combining to create 259 parasitoid-hyperparasitoid species associations. Results show that introduced parasitoids experience partial enemy release, but the degree of enemy release does not affect the chance of successful establishment or successful control. This lack of effect of enemy release might be due to the broad geographical range of the hyperparasitoids (perhaps from unintentional introductions) which reduces the degree of enemy release. In addition, the life-history traits of the hyperparasitoid communities in the new range were different from the native range, with relatively more generalist hyperparasitoid species in the new range. These results show that introduced natural enemies can experience enemy release, but that this does not necessarily help to predict their successful establishment.