The last years have witnessed the emergence of research in the field of complex networks. Complex networks describe a considerable amount of natural and social systems with a large, irregular, and changing in time structure, build up of thousands of nodes and ties between them. Certainly the continuous increase of computational power has supported the analysis of the wide databases of real networks (e.g. among others World Wide Web or social, neural, metabolic networks), and has created the basis for the identification of the unifying laws and statistical properties common to most of these networks (Albert and Barábasi, 2002). Even though at the beginning it was supposed that complex networks should be treated as random graphs, it was rapidly recognised that this was not always the case and that other types of structures exist, such as small-world (Watts and Strogatz, 1998) and scale free networks (Barabási and Albert, 1999). This is important, since the properties of complex networks are related and/or encoded in their topology. For this reason tools and measurements to capture in quantitative terms these aspects have been developed. Boccaletti et al. (2006) have recently reviewed the indices and structural properties (e.g. size, density, degree, clustering, diameter, etc.) used normally to define network topology, whereas Melián and Bascompte (2004) have shown, for food web networks, how the network structure correlates with its robustness and its response to external perturbations, showing that the cohesive organization of a network in dense sub webs increases the resistance against fragmentation. A food web constitutes a special description of a biological community with focus on trophic interactions between consumers and resources (de Ruiter et al., 2005). Therefore, food webs are deeply interrelated with ecosystem processes and functioning since the trophic interactions represent the transfer rates of energy and matter within the ecosystem. In addition, the study of ecological network structure and stability provides an important tool in the assessment of the impact of perturbations in the ecosystem itself. In particular it is known that trophic webs are not randomly assembled, but are the result of the interaction of different cohesive subgroups. Therefore, identifying the tightly connected groups within a network is an important tool for understanding the main energy flows of the network itself, as well as for defining a hierarchy of nodes and connections within a complex structure (Boccaletti et al., 2006). For this reason a considerable effort in ecosystems theory has been devoted to understand how food webs are structured and how this structure influence ecosystem processes.
In addition the introduction of dynamics, through bionergetic-based models, in food webs has allowed the development of explicit dynamic network models of shared nutrient consumption, including competition among produces for multiple resources, as well as effects of anthropogenic pressures. This has also allowed to include population dynamics within this framework (Martinez et al., 2006) and to extend the type of network parameters and analysis one is able to calculate.
However, despite the ubiquity of complex ecosystems and the vast numbers of interaction between different species, a higher food web complexity has been demonstrated not correlated with a higher stability (May, 1972, 1974). In recent studies on adaptive networks (Garcia- Domingo, Saldana, in press), different configurations of trophic networks and nonlinear bioenergetic dynamics (Dunne et al., 2004; Martinez et al., 2005; Boccaletti et al., 2006) are helping to elucidate the relation between ecological stability and complexity in food webs. In particular, it has been observed that little cohesive groups of nodes in the food webs represent groups of key species that make the entire network more resistant to external perturbation, decreasing the probability of network fragmentation when species are removed (Melián and Bascompte, 2004). Several definitions of cohesive sub groups or clusters have been proposed to analyse this effect (de Nooy et al., 2005), such as K-cores, cliques, components.
Our purpose is to define terrestrial and aquatic food web networks at selected European sites prototypical of European ecosystems and then analyze the network structure, properties and composition of cohesive sub webs. The analysis will focus in particular on the link between terrestrial and aquatic ecosystem. In a second step, we plan to add the spatio-temporal dynamics in those food webs and then to examine their changing properties. Finally, we would like to assess the network persistence when subjected to anthropogenic pressures linked to the application of several environmental EU policies.
In this work, we present the preliminary results for two food webs and develop the general strategy. It is clear that, in order to carry out a comparative analysis, we need to extend the spatial coverage including at least one representative site for each European ecosystem. Our work is continuing across these lines.