Food Web Simulation

Analytical and Visual Representation of Ecosystem Dynamics

This project is an ecosystem simulation designed around food web interactions, representing biological dynamics in a local ecosystem. Built for the Simulation and Modelling course, the project explores population balance, predator-prey relationships, and trophic-level distributions.

The goal was to simulate an ecosystem with multiple intertwined food webs and test if it could reach sustainable levels resembling natural ecological processes. Outcomes included analyzing predator-prey waveforms, reproductive survival based on carrying capacity, and biomass distributions across trophic levels.

Inspirations came from a shared appreciation of nature, as well as interactive media like Equilinox. The food web logic was based on BJYU'sexplanation of trophic levels, including producers (grass), primary consumers (rabbits, mice, grasshoppers), secondary consumers (small birds, frogs, snakes), and tertiary consumers (hawks, foxes, owls).

The simulation used AI decision trees to drive organism behavior. Animals prioritized needs in order of importance (sleep, water, food, reproduction), with threshold values determining when resources were sought. Differential equations were applied to model population changes and predation:

• Ordinary differential equation of time to track ecosystem progression.
• Lotka-Volterra predator-prey equations to simulate population cycles.
• Velocity-based equations for animal position during movement or hunting.

Tests confirmed realistic predator-prey cycles, carrying capacity stabilization, and energy flow between trophic levels. For example, rabbit populations grew rapidly when grass was abundant, then stabilized at a carrying capacity of ~600. Biomass distribution across trophic levels aligned with expected ecological patterns.

Limitations included small environment sizes and simplified AI awareness (animals always know resource locations). While this helped test ecosystems efficiently, it limited scalability. Extinction events were also not implemented, with grass respawning infinitely as a baseline resource.

The project successfully demonstrated ecological interactions through code. By fragmenting development into small, testable food chains, the team built a robust system of trophic levels, biomass flows, and population dynamics. This project reflects both ecological principles and computational problem-solving in ecosystem simulation.

Developed Skills: Simulation modelling, ecological algorithms, AI decision trees, differential equations (Lotka-Volterra), data visualization, and system-level testing.