Ricerc@Sapienza

Classical biological control of weeds depends on finding agents that are highly host-specific. This requires not only correctly understanding the identity of the target plant, sometimes to subspecific levels, in order to find suitable agents, but also identifying agents that are sufficiently specific to be safe and effective. Behavioral experiments and molecular genetic tools have revealed that some arthropod species previously thought to be polyphagous really consist of multiple cryptic species, host races or biotypes, some of which are more host-specific than others. Whereas true species are reproductively isolated, individuals from subspecific populations may potentially interbreed with those of other populations if they should encounter them. Furthermore, biotypes may consist of individuals sharing a genotype that is not fixed within a monophyletic group, and thus may not be evolutionarily stable. This raises the question of how such populations should be classified, and how to confirm the identity of live arthropods before releasing them as classical biological control agents. The existence of host races or cryptic species may greatly increase the number of prospective biological control agents available. However, it may also create new challenges for governmental regulation. These issues are discussed using pertinent examples, mainly from North America.