Context-specific interactions among propagule pressure, traits, and environment prevent generalization of mechanisms regulating invasion establishment

Background/Question/Methods

There has been little consensus on the relative importance of three major determinants of invasion success – propagule pressure, community susceptibility to invasion (invasibility), and invader traits – given the difficulty simultaneously testing individual and interactive impacts on invader establishment. We used a fully factorial 0.5 ha seed addition experiment (n=1280) within a restored tallgrass prairie and naturally regenerated old field (Cambridge, Ontario, Canada, to differentiate the relative effects of propagule pressure (three levels of seed density – low, medium, high) and community invasibility (combinations of community diversity, perturbation (Mowing), eutrophication (Nitrogen addition), and small mammal granivory (Fencing)) on the establishment of ten plant species (including several pernicious invaders of North American grasslands) that were absent from the field site prior to the experiment. We periodically measured light, soil moisture and soil nitrogen to determine resource association with community invasability treatments as a mechanistic proxy of suitability for recruitment. In order to incorporate traits as a testable component in our study, we used a complimentary greenhouse experiment to quantify traits expected to be associated with colonization among the ten species used in the field experiment.

Results/Conclusions

By grouping treatments and species together, we find that greater recruitment counts were associated with higher propagule pressure (i.e., more seeds resulted in more established plants). However, proportional establishment rates (i.e., seeds established/seeds added) were greatest at low levels of propagule pressure, suggesting intraspecific competition was strong at high propagule pressure levels. We also find that plant community perturbation (mowing) was associated with greater establishment rates. Furthermore, perturbation had a stronger and longer lasting influence on light availability in a species-rich (tallgrass prairie) as compared to a species-poor grassland (old field). The most uniformly successful seeded species were regionally wide-spread agronomic perennial grasses, such as Smooth brome (Bromus inermis). Despite these generalized results, we find that establishment of each species was sensitive to different combinations of imposed effects.  For example, one species was regulated solely by granivory (the entire population eliminated), whereas granivory was unimportant for most other species. Consequently, our results show that no consistent mechanism best predicts establishment rates for each species. Instead, propagule pressure, environmental conditions (community invasability) and traits interact in species-specific ways to regulate ideal conditions for establishment. These findings may help to explain difficulties with deriving generalizations of invasion in many biological systems.