Basque Centre for Climate Change Leioa, Pais Vasco, Spain
Abstract: Pollinator populations face serious declines, threatening the plant communities that depend on them as well. Because of this, the study of plant-pollinator interactions has gained significant traction in the scientific literature. In this sense, the use of network theory to analyze the complex links between plants and their pollinators has proved a tremendously powerful tool to describe the structure of these communities. However, in many cases this structure is analyzed in a static way, using data collected and aggregated across all the flowering season. While this approach provides consistent results for the structure at the macro-scale (e.g., connectance, nestedness or modularity), the dynamics of micro (e.g., individual species roles) or meso-scales (e.g., local patterns of indirect interactions) might not be captured using such aggregated approaches. Here, we use a recently-developed mechanistic multilayer framework based on the frequency of insect visits to plant individuals to assess pollinator foraging behaviors and to evaluate the potential conspecific and heterospecific pollen flows among plant individuals. We then relate these pollen flows to a number of compositional and configurational variables including plant species diversity, composition, the aggregation of floral resources in space or the competition of a focal pollinator species with other species through shared resources. We further investigate how pollinator foraging decisions ultimately affect the pollination service they provide to plant species by assessing the impact of network structure on plant reproductive success for 10 plant species. We apply this framework to a set of plant-pollinator networks gathered within 5 sites at 8 different time periods for 2 consecutive years within a mountain grassland habitat. We find that while a greater plant species diversity at the community level leads to a greater probability of heterospecific pollen flow, when we analyze patterns at the individual level, our results show that the aggregation of floral resources and the competition between pollinator species significantly affect pollen flows. Further, these two variables show significant within-year variations which directly affect pollinator roles, leading to changes in the efficiency of different pollinator species through time. These findings suggest great complexity of the pollinator responses to dynamic changes of the community over a flowering season, potentially with great implications for plant reproductive success and for the whole plant-pollinator network functionality.
