Impact of interaction strength heterogeneity on the distinctness of community types.

<p>A total of <i>q</i> = 500 local communities, each with <i>p</i> = 80 species randomly drawn from a universal pool of <i>n</i> = 100 species. The nominal components were drawn from , the interaction heterogeneity matrix elements were taken from and <i>α</i> is varied with the set of values {7, 3, 2, 1.6, 1.01} for each column in the figure. The topology component <b>G</b> has all elements equal to 1, giving a complete graph. The scaling factor was set at <i>s</i> = 0.07. (a) Histogram of the diagonal elements of the heterogeneity matrix <b>H</b>. (b) Visualization of the universal interaction matrix <b>A</b> as a weighted adjacency matrix of a digraph. (c) Principle coordinate analyses of the normalized steady state for each local community using the Jensen-Shannon distance. The Silhouette Index and optimal number of clusters are denoted. Further details can be found in <a href="http://www.ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1004688#sec007" target="_blank">Materials and Methods</a>.