Pollination is a critical mutualistic interaction that helps support biodiversity around the world. A large proportion of pollination services are carried out by bees (Hymenoptera). The sum of the interactions between plants and pollinators in a community is represented as a network where plant and bee species are ‘nodes’ and their interactions are ‘links’ in the network. Plant-pollinator network structure should reflect community-level responses to environmental stressors as individual species find novel interaction partners, are lost due to stress, competition, or the loss of a critical partner(s). We compare plant-pollinator network structure between two California habitats with different baselines levels of temperature stress. Further, we use historical and resurvey data to compare network structure over time as temperature stress and species invasions have acted on each habitat. Specifically we ask whether habitats with a strong habitat filter are more flexible overall and are thus less sensitive to temperature and water stress compared to communities adapted to less extreme climates.
Results/Conclusions
Grassland (lower stress) plant-pollinator networks are more robust against species loss than sandhill chaparall networks (higher stress). However, temporal comparisons reveal greater changes in grassland networks over time than in sandhill networks. Habitat-level comparisons alone over-predict the resilience of grassland networks relative to sandhills. Temperature stress and species invasions have advanced in each habitat over time and grassland plant species richness fell by 66%, while bee species richness remained stable implying host switching for bees but not plants. Sandhill plant and bee species richness were stable over time. In addition, species in grassland networks critically depend on non-native species such as Apis mellifera (Hymenoptera) and Hypochaeris radicata (Asteraceae). Invasive species have become critical resources in grassland habitats. The short and long term impacts of non-native species on pollination network structure, and the community-level implications of controlling (removal/mitigation) them should be carefully considered. In sandhills, species’ adaptations to higher baseline levels of temperature stress may, at least partially, buffer against increasing environmental stress. In addition, strong habitat filtering may limits species invasion into habitats with higher baseline levels of stress.