Since resistance can be costly and many herbivore resistance traits are genetically controlled, improved competitive ability of introduced plants in novel habitats through a shift in allocation from defence to growth has been predicted by the Evolution of Improved Competitive Ability or EICA hypothesis (Blossey & Nötzold 1995).

Introduced plants are often free from specialized native herbivores, but can not be free from all of herbivores, attacked by native polyphagous species or exotic specialized herbivores. When testing the EICA hypothesis, it is necessary to incorporate fundamental differences between native and exotic herbivores and between generalist and specialist herbivores.

To test the EICA, we transplanted 10 genotypes of tall goldenrods collected from introduced and original habitats and conducted a manipulative experiment inoculating a specialist exotic aphid, a generalist exotic lace bug, and a native generalist moth caterpillar. As a result, the EICA hypothesis was supported in exotic herbivores, but unsupported in native herbivores. This suggests that even if introduced plants can be free from co-evolved herbivores, introduced plants may not lose protection against native generalists because attacks by polyphagous herbivores are inevitable in novel habitats.