Category: Ecology

Summary: Testing whether increasing low-rank mutualistic structure first spreads out and then re-localizes reactive amplification in heterogeneous May community matrices.

Classic May-type ecological models explain when random interactions destabilize large communities, but real communities may also contain coherent mutualistic substructures. This experiment asks how a low-rank mutualistic core changes the region where the system remains linearly stable yet highly reactive to perturbations.

The model adds structured mutualistic coherence to otherwise heterogeneous community matrices and tracks how the dominant reactive mode changes. The key hypothesis is non-monotonic: weak coherence should smear out amplification, while stronger coherence should create an outlier-like mode that concentrates reactivity again.

That makes the experiment less about a single critical point and more about how structure reshapes vulnerability. It links localization ideas to transient amplification in ecological random-matrix theory.

Method: Finite-size sweeps of structured May community matrices, measuring reactivity and mode localization as low-rank mutualistic coherence varies.

What is measured: Reactive-window width, localization of the reactive mode, non-monotonic response to mutualistic coherence, and related stability indicators.