Optimal control of garlic mustard, a biennial weed with strong density dependence and limited dispersal

Background/Question/Methods

The invasive plant, garlic mustard (Alliaria petiolata) is an obligate biennial (rosettes and adults) and invasions are typically spatially structured with large, dense cores and small, isolated, satellite populations. We used experiments and modeling to explore population dynamics and their consequences for control of this species. We characterized negative density-dependent survival and fecundity using observational and experimental data to parameterize a population growth model. We measured seed dispersal distances to estimate invasion speed. We experimentally pulled adults in cores in three sites and in satellites in three sites to test management predictions of our population growth models.
Results/Conclusions

Population modeling showed that compensatory density-dependent survival and fecundity, in combination with stage-structure, result in stable, two-point cycling. High levels of rosette and adult mortality are required to significantly reduce population densities. Spatial modeling indicated that population growth is exceptionally high at the invasion front and that satellite populations resulting from long distance dispersal drive the rapid invasion speed. Experimental management indicated that pulling adult plants in satellite populations greatly reduces population growth and invasion speed. Based on a synthesis of our research program, our recommendations for best management practices of garlic mustard are to 1) induce 100% mortality in a small spatial area rather than to spread efforts widely, 2) focus resource efforts on finding and removing satellite populations rather than managing the invasion core. We incorporated outreach into our research by developing a program to involve high school students and teachers in experimental plant removals. Our research has direct management applications, and we have made our results and recommendations available to local county and state land managers.