Abstract
Understanding how ecological communities are structured is a major goal in ecology. Ecological networks representing interaction patterns among species have become a powerful tool to capture the mechanisms underlying plant-animal assemblages. However, these networks largely do not account for inter-individual variability and thus may be limiting our development of a clear mechanistic understanding of community structure. In this study, we develop a new individual-trait based approach to examine the importance of individual plant and pollinator functional size traits (pollinator thorax width and plant nectar holder depth) in mutualistic networks. We performed hierarchical cluster analyses to group interacting individuals into classes, according to their similarity in functional size. We then compared the structure of bee-flower networks where nodes represented either species identity or trait sets. The individual trait-based network was almost twice as nested as its species-based equivalent and it had a more symmetric linkage pattern resulting from of a high degree of size-matching. In conclusion, we show that by constructing individual trait-based networks we can reveal important patterns otherwise difficult to observe in species-based networks and thus improve our understanding of community structure. We therefore recommend using both trait-based and species-based approaches together to develop a clearer understanding of the properties of ecological networks.
Highlights
During recent decades, ecological networks have become an increasingly useful tool to capture the mechanisms underlying plant-animal assemblages (Bascompte & Jordano, 2007; Reiss et al, 2009; Heleno et al, 2014; Ings & Hawes, 2018)
Nectar holder depth of sampled flowers (n = 131 flowers from the six out of 10 plant species with measurable nectar holder depths) was positively correlated with maximum floral display size for most of the plant species subsampled for trait measurement (Fig. S2A)
Intertegular distance was positively correlated (R2 = 0.817) with proboscis length across the subset of individuals
Summary
Ecological networks have become an increasingly useful tool to capture the mechanisms underlying plant-animal assemblages (Bascompte & Jordano, 2007; Reiss et al, 2009; Heleno et al, 2014; Ings & Hawes, 2018). The vast majority of mutualistic networks published to date are composed of nodes representing plant and animal species that are connected by edges indicating the presence of interactions between them. This approach provides a holistic picture of the community structure, and allows the detection of patterns that cannot be inferred from the observations of the nodes (species). In such species-based networks, nodes consist of populations of conspecific individuals that may vary in many biological traits such as phenotype, phenology or behaviour. There is growing recognition that potentially important information is lost when averaging species data and ignoring inter-individual variation (Ings et al, 2009; Olesen et al, 2010; Tur et al, 2014; Kuppler et al, 2016; Ings & Hawes, 2018)
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call