Figure 1.

An illustration of the interaction matrix of a perfectly nested network as compared to the interaction matrix of a non-nested network (Nakaikemi marsh pollination network) composed of the same number of nodes and links. In a perfectly nested network (left panel), one can define a line (marked in red) that perfectly partitions the matrix into a filled region (i.e., the region above the line) and an empty region (i.e., the region below the line). The same feature does not hold for a non-nested network (right panel).