Contradictory patterns in diversity-invasibility relationships across communities and scales suggest that fully taking into account multiple mechanisms and drivers may be more productive in understanding invasion dynamics than searching for a single unifying theory. We conducted a large plant community assembly experiment with wetland species to test different assumptions about diversity-invasibility relationships. Our field experiment was designed to simulate a situation where species would reassemble into communities showing different levels of functional diversity after disturbance. We then used diversity-interaction modeling based on structural equations to test multiple hypotheses about the role of species interaction mechanisms in biotic resistance to a set of invaders (unsown species in our case). The model selection approach was based on functional group identity, species or functional diversity, and interactions between and within functional groups. Our models also included the influence of environmental heterogeneity (flooding area) on invasion dynamics.
Results/Conclusions
We found that functionally diverse communities are less invasible to unsown species including exotic Phragmites australis and Salix alba (evidence ratio of complementarity diversity effect vs selection effect = 2.5:1). Model results suggest interactions between certain pairs of functional groups mainly drive the negative diversity-invasibilty relationship. We propose life history trade-off (annual-perennial), with species occupying different phenological niches, and resource partitioning via functional complementary as primary mechanisms underlying the negative diversity-invasibility in these communities. This study also shows invader-specific responses to diversity-driven biotic resistance, interacting with environmental heterogeneity and propagule pressures. Each functional group in resident communities plays different roles in invasion dynamics and therefore diversity-invasibility relationships are inherently complex and dynamic even at the community scale.