Although several successful examples of biocontrol using fungal pathogens exist, situations in which there is no discernable impact remain unacceptably high. This response may result from targeting inappropriate host life stages, failure of the control agent to establish and proliferate because of poor matching of host and pathogen life histories, or from broad environmental mismatching. At a genetic level, interactions between host and pathogen may determine the ultimate success of control programs. Here, we use a simulation model to illustrate that the optimal level of pathogen virulence is influenced by factors such as host longevity, as well as whether pathogen impacts on targets are through reduced fecundity or increased mortality. Importantly, the optimal pathogen in terms of effective long-term control of a target population is not necessarily the most virulent strain. In part, this is because the imposition of selection by a biocontrol agent may induce rapid increases in the frequency of resistance (if present) in host populations, ultimately leading to decreased effectiveness. Thus, appropriate matching of host and pathogen characteristics is an important consideration in trying to maximize impact on an invasive weed while minimizing the evolution of resistance.

Additional index words: Coevolution, host–pathogen dynamics, optimal biocontrol, resistance, virulence.