Category: Ecology

Summary: Testing whether adaptive harvesting in a noisy predator-prey system improves yield at short delays but increases collapse risk once observational delay becomes too large.

Harvesting a fluctuating ecological system requires balancing yield against the risk of driving the population into collapse. This experiment studies a Lotka-Volterra-style predator-prey model and asks whether adaptive harvesting outperforms a fixed rule only when observations are fresh enough.

The code compares adaptive and fixed policies across delay and noise conditions, measuring total yield, prey and predator means, effort variability, collapse events, and collapse timing. It is designed to detect a delay inversion where adaptive harvesting helps at small lag but loses its value when control responds too late.

That makes the result relevant to fisheries and renewable-resource management. The central issue is not merely maximizing output, but identifying when delayed feedback becomes destabilizing.

Method: Predator-prey simulations with adaptive or fixed harvesting under observation delay and environmental noise, summarized across repeated trials.

What is measured: Total yield, prey and predator abundance, effort variability, collapse rate, collapse time, adaptive-minus-fixed yield, and delay-driven inversion signal.