technical report Distribution of node centrality indices in food webs


Network analysis examines the role of different species in ecological communities. This can be achieved by several indices designed to quantify the topological position of a given species, each index emphasizing a different aspect of node centrality. Comparison of the rank order of species produced by these indices can help in understanding community organization, and also in revealing the relationships among indices. However, the rank orders disregard absolute differences between the importance values of species and therefore ranking may not completely reveal the quantitative information contained within food web data. To evaluate the consequences of differences between indices, we compared the rank order of species obtained by different node centrality indices and also the distributions of these indices, using data from six different food webs. Robust patterns emerged from the examination of similarities in rank orders, but distributions changed greatly with food web structure. This implies that the combined analysis of rank orders and distributions of centrality indices provides more information on species importance and could better serve ecologists in looking for topological keystone species.

Paper Details


B. Bauer,  F. Jordan,  J. Podani