Topology regulates the distribution pattern of excitations in excitable dynamics on graphs.

We study the average excitation density in a simple model of excitable dynamics on graphs and find that this density strongly depends on certain topological features of the graph, namely connectivity and degree correlations, but to a lesser extent on the degree distribution. Remarkably, the average excitation density is changed via the distribution pattern of excitations: An increase in connectivity induces a transition from globally to locally organized excitations and, as a result, leads to an increase in the excitation density. A similar transition can be induced by increasing the rate of spontaneous excitations while keeping the graph architecture constant.