LB 33-279 - Macroanalysis of Apis mellifera invaded plant-pollinator networks

The western honey bee (Apis mellifera) is nearly ubiquitous in human altered landscapes. At the same time, concerns are rising about the prevalence and impacts of competition between honey bees and native pollinators. As such the potential of this species to have negative consequences for both plants and pollinators in invaded ecosystems should not be overlooked.The goals of this study were to determine the impact of A. mellifera generalization, defined as the proportion of realized interactions to all possible plant partners (i.e. normalized degree), on characteristics of global plant-pollinator networks. We predict the A. mellifera normalized degree will be positively correlated with network metrics, connectance (C) and nestedness (NODF), and negatively correlated with modularity (Newman’s Q). This description has the potential to inform future predictions about the effects of the massively introduced species such as honey bees, as well as conservation efforts.Our approach to testing these predictions involved using an open-access repository of 185 aggregated plant-pollinator networks from around the world. Qualitative (binary) bipartite networks were constructed, and the network metrics mentioned above were extracted. In addition, available information on sampling effort and location of each network was recorded. Generalized linear regressions were used to test the predictions of each metrics response to honey bee generalization while accounting for differences in sampling effort and spatial autocorrelation. Here we show that of the networks in the dataset, A. mellifera was present in 132, comprising about 71%. Of networks where A. mellifera interactions were present, its normalized degree ranged between 0.1 and 1.0 indicating a large range in generalization (mean = 0.23, sd = 0.24). Our findings supported our prediction, as connectance and nestedness were positively correlated with A. mellifera, whereas modularity was negatively correlated.In conclusion, A. mellifera is frequent in worldwide networks and typically interacts with several plant species, forming more links than most other pollinators in the network. This in turn plays a central role in determining network structure. These findings are limited in that they do not account for interaction rewiring in the presence/absence of A. mellifera, which is needed to better understand the effects of its introduction. As plant-pollinator networks highly dynamic, it is plausible native communities can rewire if A. mellifera is removed from the ecosystem—if the community is resilient enough.