Long-term consequences of biocontrol agents: Demographic variation among Lythrum salicaria populations 16 years after herbivore introduction

Background/Question/Methods

Introductions of non-native species represent opportunities to test ecological and evolutionary theory about the consequences of sudden and sustained changes to community structure. Biological control programs that introduce native herbivores to manage populations of invasive species lacking specialist enemies can initiate similar sudden and sustained shifts in community structure. An often untested assumption of these programs is that biocontrol agents establish top-down control, reducing the vigor and abundance of invasive species long-term. Purple loosestrife (Lythrum salicaria), is an invasive wetland plant in the United States. In 1992 a biocontrol program introduced two species of specialist leaf feeding beetles to the United States from Germany. Some purple loosestrife populations have been routinely subject to 90-100% defoliation and others experienced little to no herbivory by the biocontrol agents. We censused three sites that were consistently subject to high levels of herbivory as well as three sites with historically low levels of herbivory by specialist herbivores. Vegetation population dynamics in invaded wetlands were tracked for two years to test the assumption that biological control agents reduce the abundance and vigor of invasive purple loosestrife to levels common in the native range.

Results/Conclusions

Herbivore damage to purple loosestrife increased throughout the growing season, while plant health, leaf chlorophyll content, declined correspondingly at most sites. In two populations with historically high herbivory, L. salicaria was less dominant and reproduced less than in populations lacking established biocontrol. Additionally the dominance of the invader had been reduced to levels at or below most populations in the native range. Two other populations, one with historically high and one with historically low herbivory, were still heavily dominated by L. salicaria.  At these sites L. salicaria was taller, accumulated greater biomass and comprised a greater proportion of the total vegetation than the species did in most native populations measured in this study.  These results indicate that biological control may continue to reduce the abundance and vigor of L. salicaria where the biological control agents have established, but biological control alone is not likely to result in landscape wide reduction in the vigor and dominance of this wetland invader. Overall this study showed that, 16 years after the initiation of biological control of L. salicaria in Minnesota, the program continues to be a success reducing vigor and abundance of the invader in some areas.