Optimal Channel Networks accurately model ecologically-relevant geomorphological features of branching river networks

River networks have universal fractal structure with hydrological and biological consequences. While Optimal Channel Networks (OCNs) capture rivers’ scaling features, ecology studies often use random networks that ignore real scaling. We show that branching probability of random networks is scale dependent, varying with observation resolution, and thus is not a robust ecological driver. Random networks bias estimates of population stability and persistence, whereas OCNs yield estimates comparable to real rivers. We advocate OCNs as model landscapes for realistic and generalizable ecohydrological projections in riverine networks.

Luca Carraro, Florian Altermatt