Patterns of introduced species interactions affect multiple aspects of network structure in plant-pollinator communities

Species introductions have the potential to affect the functionality and stability of ecological communities, but because little is known about how introduced species form novel interactions, these impacts are difficult to predict. We quantified the impacts of species introductions on species interaction networks using five different model scenarios of how a novel species might form plant-pollinator interactions. The network structure was based on experimental manipulations on a community of plants and pollinators and shows that the community was more diverse, ordered, and compartmentalized, but less complex when an invasive plant generalist was present. Our models of species introductions reliably predicted several aspects of novel network structure in the field study. We found that introduced species that become incorporated into the community as generalists (both in the number and frequency of their interactions) have a much larger impact on the structure of plant-pollinator communities than introduced species that integrate into the community with few interactions. Average degree is strongly affected by the number of interactions the novel species forms and whether it competes for interactions, whereas connectance is affected by whether the novel species competes for interactions or adds new interaction partners. The number and size of compartments in the network change only when the novel species adds new interaction partners, while modularity and nestedness respond most to the number of interactions formed by the novel species. We provide a new approach for understanding the impacts of introduced and invasive species on plant-pollinator communities and demonstrate that it is critical to evaluate multiple structural characters simultaneously, as large changes in the fundamental structure of the community may be disguised.