Revisiting functional diversity from a network perspective: regular equivalence-based approach

Species are embedded in a network of trophic interactions known as a food web. We propose to measure functional diversity of an ecosystem from a network perspective. Our approach is based on the concept of regular equivalence (REGE), which can quantify the similarity in species trophic roles in a food web. Using the outcome of REGE analysis, we define three REGE-based functional diversity indices. First is the average dissimilarity in species trophic roles, second is the optimal number of trophic role groups, and the third is the degree of evenness that species are partitioned into trophic role groups. We quantified these three REGE-based functional diversity indices for 92 aquatic food webs and compared the results with that from random food webs. We found that most food webs had higher average trophic role dissimilarity and more trophic role groups than random food webs. However, real food webs and random food webs showed no difference in the evenness of species partition into trophic role groups. We also determined the relationship between REGE-based functional diversity and the structural properties of food webs. We found that food webs with high functional diversity were those with sparse network organization, highly modular structure, or low network cohesion. Furthermore, we also investigated the relationship between our REGE-based functional diversity indices and their trait-based and interaction-based counterparts; their weak-to-moderate correlation suggests our REGE-based approach offers a complementary view to functional diversity of an ecosystem.