Direct interactions between invasive plants and native pollinators: evidence, impacts and approaches.
Invasive non-native plants form interactions with native species and have the potential to cause direct and indirect impacts on those species, as well as the functioning of invaded ecosystems. Many entomophilous invasive plants form interactions with resident pollinators; sometimes, these interactions are necessary for the reproductive success of the invader. However, the direct role native pollinators play in plant invasion is not well understood and varies according to invasive plant traits, including breeding system and pollination syndrome. The majority of studies addressing impacts on plant-pollinator mutualisms have focussed on the indirect impacts of plant invasion for native plant pollination. Fewer studies have focussed on the direct effects of invasive plants on native flower visitors. Impacts of invasive plants on native pollinators can occur at a range of scales: from the individual flower visitors (in terms of nutrition, health and fitness), to populations (size, density and growth rates), communities (richness, diversity and composition) and community-level interactions (insect-flower interaction networks). Most research to date has focussed on community-level impacts, with almost nothing known about the effects of invaders on native flower visitor individuals or populations. Invasive plant traits, including reward quantity and quality, spatial and temporal availability and accessibility, modulate effects on native flower visitors, and thus, different plant species have different impacts. Similarly, flower visitors do not all respond in the same way to invasive plants. Thus, generalizations are difficult to make, but understanding impacts at the individual and population level for different visitor taxa is key to explaining community-level impacts. There have been varied approaches to determining impacts, with most studies attempting to compare invaded vs. non-invaded habitats. The pros and cons of different approaches are discussed. Since it is impractical to study every invasive plant in every ecological context in which it occurs, we recommend a better understanding of relevant individual-level traits to predict direct interactions between invasive plants and native pollinators.