Breaking down Complex Saproxylic Communities: Understanding Sub-Networks Structure and Implications to Network Robustness

Anna Dornhaus,Eduardo Galante,Cecilia Díaz-Castelazo,Javier Quinto,Víctor Rico-Gray,Estefanía Micó,Hervé Brustel,Ma. Ángeles Marcos-García

doi:10.1371/journal.pone.0045062

Abstract

Saproxylic insect communities inhabiting tree hollow microhabitats correspond with large food webs which simultaneously are constituted by multiple types of plant-animal and animal-animal interactions, according to the use of trophic resources (wood- and insect-dependent sub-networks), or to trophic habits or interaction types (xylophagous, saprophagous, xylomycetophagous, predators and commensals). We quantitatively assessed which properties of specialised networks were present in a complex networks involving different interacting types such as saproxylic community, and how they can be organised in trophic food webs. The architecture, interacting patterns and food web composition were evaluated along sub-networks, analysing their implications to network robustness from random and directed extinction simulations. A structure of large and cohesive modules with weakly connected nodes was observed throughout saproxylic sub-networks, composing the main food webs constituting this community. Insect-dependent sub-networks were more modular than wood-dependent sub-networks. Wood-dependent sub-networks presented higher species degree, connectance, links, linkage density, interaction strength, and were less specialised and more aggregated than insect-dependent sub-networks. These attributes defined high network robustness in wood-dependent sub-networks. Finally, our results emphasise the relevance of modularity, differences among interacting types and interrelations among them in modelling the structure of saproxylic communities and in determining their stability.

Highlights

Network analysis is a valuable tool for studying the diversity of species and interactions in large trophic networks [1]
Research on ecological communities has been dominated by small-scale studies [9], and restricted to a single type of interaction [4], while only recently, spatio-temporal scales of ecological communities [10,11,12] or complex networks with different types of interaction [13,14] have been addressed with network analysis
Among the Coleoptera and Diptera (Syrphidae) saproxylic species coexisting in tree hollows, we considered three levels of interaction: 1) complete network, 2) sub-networks defined according to the use or not of woody resources, and 3) sub-networks according to their feeding guild

Summary

Introduction

Network analysis is a valuable tool for studying the diversity of species and interactions in large trophic networks [1]. A high number of ecological communities have been studied under this perspective, discovering specialised interacting patterns as nestedness in mutualistic networks [2,3] or modularity in antagonistic networks [4], providing insight into the function and evolution of the components of the system [5]. Ecological network studies are largely focussed on qualitative data, assuming that all interacting species are important [15]. Specialised network patterns are best defined at quantitative scale in both mutualistic and antagonistic communities [16], and the relative abundances of the components of the networks influences structural patterns as asymmetry [17]

Methods

Results

Conclusion