Multilayer networks facilitate the integration of multiple interaction types by quantifying the per capita effects of species on each other, thereby advancing our understanding of the intrinsic complexity of natural food webs. Using a multilayer network framework, this study provides the first food web data for a small oceanic island (Montaña Clara, Canary Islands) and compares it with available data from another small island of continental origin in the context of island biogeography theory. During the two most contrasting seasons, we collected data on interactions between plants and their pollinators, herbivores, seed dispersers, and saprotrophic, symbiotic and pathogenic root-associated-fungi. Pollination, herbivory and seed dispersal were sampled via flower visitation census, direct observation, and faecal analysis, respectively, while fungi were identified using DNA metabarcoding. We identified 64 animal species and 367 fungal amplicon sequence variants (ASVs) interacting with the 13-plant species sampled. Five plant species (38%) showed the highest values of multilayer versatility (> 0.5), indicating they are the most important for the structure of the ecosystem. A total of 23 modules were detected, more than half of which were restricted to a single type of interaction, and approximately 73% of the species switched modules between pairs of interaction types. Comparing the results with those from the continental island, the oceanic shows a simpler network, higher overall versatility and a less modular structure. These results are consistent with ecological and evolutionary predictions from island biogeography theory. Future studies involving different interaction types and conducted on islands with different traits, such as taxonomical/functional disharmony, density trade-offs, or interaction release, will allow for the assessment of the generality of the observed patterns.

Peer reviewed