What plant-pollinator network structure tells us about the mechanisms underlying the bidirectional biodiversity productivity relationship?

The Biodiversity - Ecosystem Functioning (B-EF) relationship remains a topic of ongoing debate with most studies focusing on primary productivity, and documenting that this relationship takes many forms. It remains unclear if biodiversity drives productivity or productivity shapes biodiversity or the relationship is bidirectional. B-EF studies explore almost exclusively the relationship between species richness and ecosystem functioning, while the role of biotic interactions, a key component of ecosystem functioning, has been neglected. Here, using data of 80 local plant-pollinator networks on 20 Aegean islands, and of gross primary productivity (GPP) from the MODIS satellite, we explored the bidirectional relationship between interaction network structure (nestedness and specialization), species richness (plants and pollinators) and mean and inter-annual variability of GPP. We found that nestedness and specialisation of plant-pollinator networks is driven by mean GPP. However, specialisation alone was a significant predictor of mean GPP, implying that networks tend to be more specialised in low-productivity areas. Pollinator species richness exerted a strong effect on mean GPP with the remaining factors playing a minor role, while the effect of mean GPP on pollinator species richness was weaker. Furthermore, the nestedness of plant-pollinator networks drives inter-annual variability of GPP with more nested networks displaying less variability, which is in accordance with the predictions of the insurance hypothesis. Plant and pollinator species richness were also associated with inter-annual variability of GPP. (C) 2022 The Authors. Published by Elsevier GmbH on behalf of Gesellschaft fur Okologie.

Automated summary

This study explores the bidirectional relationship between interaction network structure, species richness, and gross primary productivity (GPP) based on data from plant-pollinator networks and satellite measurements. The results suggest that both the structure of the networks and the richness of pollinator species have significant effects on GPP. Additionally, the network structure influences the inter-annual variability of GPP. The findings highlight the importance of considering biotic interactions in understanding the B-EF relationship.