Information Technology Reference
In-Depth Information
Fig. 14.3. Example of a bipartite plant-animal mutualistic network. Solid and empty nodes rep-
resent plant species, and their seed dispersers, respectively. These mutually benecial interactions
form networks of dependence involving dozens of species.
interactions and worked on small groups of interacting species [Waser et al. (1996);
Thompson (2005)]. More recently, these studies have been extended by scaling all
the way up from small to entire networks and by providing a rational framework to
characterize their complexity.
14.3.1. The structure of mutualistic networks
The rst round of papers on mutualistic networks were, not surprisingly, descrip-
tions of their structure, following the tradition in food webs as described in the
previous section. The rst pattern analyzed was the connectivity distribution, and
ecologists found a common structural pattern dened by truncated power-law con-
nectivity distributions. This put mutualistic networks in a similar context as other
networks because they exhibit heterogeneous connectivity distributions. While the
bulk of species had only one or a few interactions with other species, a few gener-
alists had far more interactions than would be expected under conventional models
[Jordano et al. (2003)]. Regardless of the type of mutualism (seed dispersal or polli-
nation) and other biological dierences, all communities had the same structure. A
small dierence, in relation to non-biological networks such as the Internet, was the
truncation for high numbers of interactions. This truncation can be explained by
several non-exclusive mechanisms such as forbidden links | some interactions be-
tween a plant and an animal can not occur due to, for example, size or phenological
constraints [Jordano et al. (2003)].
Mutualistic networks are, thus, heterogeneous. However, this property only
refers to a statistical description of the probability to interact with a given number
of species. It does not tell us anything about the identity of the interacting species.
The next step in the road to disentangle the structure of mutualistic networks was
describing its nested structure (Fig. 14.4). In a nested matrix, the number of inter-
actions per species are arranged in such a way that specialists interact with proper
subsets of the species with which generalists interact [Bascompte et al. (2003)]. This
generates a network with a core of interactions, generalist plants and generalist an-
imals interacting among themselves, and asymmetric tails, specialists interacting
with the most generalist species.
