The structure and ecological function of the interactions between plants and arbuscular mycorrhizal fungi through multilayer networks

Abstract

Arbuscular mycorrhizas are one of the most frequent mutualisms in terrestrial ecosystems. Although studies on plant mutualistic interaction networks suggest that they may leave their imprint on plant community structure and dynamics, this has not been explicitly assessed. Thus, in the context of plant-fungi interactions, studies explicitly linking plant-mycorrhizal fungi interaction networks with key ecological functions of plant communities, such as recruitment, are lacking. In this study, we analyse, in two Mediterranean forest communities of southern Iberian Peninsula, how plant-arbuscular mycorrhizal fungi (AMF) networks modulate plant-plant recruitment interaction networks. We use a new approach integrating plant-AMF and plant recruitment networks into a single multilayer structure. We also develop a new metric (Interlayer Node Neighbourhood Integration, INNI) to explore the impact of a given node on the structure across layers. The similarity of plant species in their AMF communities is positively related to the observed frequency of recruitment interactions in the field. Results reveal that properties of plant-AMF networks, such as plant degree and centrality, can explain about the properties of plant recruitment network, such as in- and out-degree (i.e. sapling bank and canopy service) and its modular structure. However, these relationships differed between the two forest communities. Finally, we identify particular AMF that contribute to integrate the neighbourhood of recruitment interactions between plants. This multilayer network approach is useful to explore the role of plant-AMF interactions on recruitment, a key ecosystem function enhanced by fungi. Results provide evidence that the complex structure of plant-AMF interactions impacts functional and structurally plant-plant interactions, which in turn may potentially influence plant community dynamics, through their effects on the structure of the recruitment network.