Category: Ecology

Summary: Comparing how different ecological interaction structures change the width of the stable-but-reactive regime, not just asymptotic stability alone.

Ecological communities can be linearly stable in the long run yet still strongly amplify perturbations over short times. This experiment asks whether network structures that look similar under classical May-style stability criteria can differ substantially in the size of that stable-but-reactive window.

The simulation builds random, modular, nested, and predator-prey community matrices, then measures both the asymptotic stability threshold and the onset of transient reactivity across repeated trials. By tracking the gap between those thresholds, it estimates how much room each structural class leaves for perturbations to grow before eventually decaying.

That distinction matters because ecosystems are affected by transient shocks, not only by infinite-time limits. The experiment therefore compares structure-dependent stability and reactivity in one unified framework instead of treating them as separate questions.

Method: Repeated May-style community-matrix simulations across multiple ecological ensemble types, measuring both stability and reactivity thresholds and their separation.

What is measured: Stability threshold, reactivity threshold, stable-but-reactive window width and ratio, reactive fraction, nonnormality, and leading-eigenvalue diagnostics by ecological structure.