Numerous observational studies demonstrate the invasion of exotic plant species on native communities. In contrast, long-term experiments and demographic analyses of invader impacts on native species are rare and the population-level impacts of invaders remain largely assumed. Even when invaders are known to have population-level impacts, the importance of invader effects relative to other species interactions such as herbivory are rarely tested. It has been argued that in forests of eastern North America invaders might be less important than the impacts of overabundant white-tailed deer.
In this study we compared the demographic impacts of the common invader Alliaria petiolata (garlic mustard) to the effects white-tailed deer using a decade of longitudinal data from a crossed deer exclusion/Alliaria removal experiment. We analyzed these data using multi-level models to determine whether Alliaria removal (Alli-) causes vital rates of three common forest herbs (Maianthemum racemosum, Polygonatum biflorum, Trillium erectum) to diverge over time from plants in paired control plots and compared the impact of the invader to that of deer herbivory. We used the resulting vital rates to parameterize matrix models using data from 2013-2014 to determine the impact of our treatments on plant fitness and population growth for Maianthemum and Trillium.
Each native species had at least one key vital rate benefit from Alliaria removal. Maianthemum flowering (Alli-xtime p = 0.002) and Polygonatum abundance (Alli-xtime p = 0.05) were impacted as much by Alliaria removal as by release from deer herbivory. Trillium seedling recruitment benefitted from Alliaria removal (Alli-xtime p = 0.01), but overall this species’ vital rate responses were complex, likely due to higher intra- and possibly inter-specific competition in Alli-and deer exclusion plots.
Matrix models built from these vital rates indicate that population growth for Maianthemum is high when both deer are excluded and Alliaria removed (lambda=1.08), yet removal of Alliaria alone results in a similar rate (lambda=1.06). Trillium population lambda are currently ~identical across all treatments (lambda=1), likely damped by competition in the deer exclusion plots. Sensitivity analyses indicates when Alliaria is present population stability is almost entirely due to stasis of adult plants, while Alliaria removal increases the role of recruitment to population stability. Prior work demonstrated that Alliaria’s negative effects on native vital rates are mediated by disruption of the AMF-native plant mutualism, not competition. To our knowledge this is the first demonstration of the long-term impacts of Alliaria on native plant demography.