Spotlight #56 – Chaos, Ecology & Neuroscience

Chaos theory is the study of systems that are highly sensitive to initial conditions — small differences in starting state grow exponentially, making long-term prediction impossible. Yet these systems are not random: they follow exact deterministic equations, and their trajectories are confined to a geometrically beautiful object: the strange attractor.

The butterfly effect

Edward Lorenz discovered deterministic chaos in 1963 while running a simplified atmospheric convection model. He noticed that re-running a simulation from a printout (rounded to 3 decimal places instead of 6) produced a completely different forecast after a short time. The three equations that encode this are:

ẋ = σ(y − x)
ẏ = x(ρ − z) − y
ż = xy − βz

At σ=10, ρ=28, β=8/3 the trajectory traces the famous double-lobed butterfly, never repeating, spiraling around two unstable fixed points forever. The largest Lyapunov exponent is λ≈0.905, meaning nearby trajectories diverge by a factor of e each unit of time.

Fractal geometry

The Lorenz attractor has a Hausdorff dimension of approximately 2.06 — more than a surface but less than a solid. This fractional dimension is the geometric signature of chaos: the attractor is folded infinitely but occupies zero volume. Thomas' attractor at b≈0.19 shows a striking 3-fold symmetric structure best explored by dragging to view from different angles.

Explore in the library:

🌀 Strange Attractors 🌿 Bifurcation Diagram 🎲 Chaos Game 🦋 Lorenz System

Population ecology studies how species interact, compete, and co-evolve within ecosystems. The Lotka-Volterra equations (independently derived by Alfred Lotka in 1925 and Vito Volterra in 1926) capture the essence of predator-prey dynamics:

dH/dt = rH − aHL        (prey: grow at rate r, eaten at rate a per predator)
dL/dt = baHL − mL      (predators: reproduce from prey, die at rate m)

The solution is periodic oscillations: prey peak, predators lag behind and peak, prey collapse, predators starve and collapse, prey recover. The period and amplitude depend on initial conditions — unlike the Kuramoto model, there is no fixed attractor.

Why agent models matter

The differential equation version assumes infinite, perfectly mixed populations. Real ecosystems are finite and spatial. The wolf-sheep simulation shows that spatial structure creates refugia — patches where prey survive wolf waves — which can stabilize what would otherwise be diverging oscillations or extinction cascades. Stochastic demographic noise (individual random deaths and births) also produces qualitatively different behaviour than smooth ODEs.

Try this experiment

1. Start with default parameters and observe 2-3 population cycles.
2. Increase Wolf Food Gain to 70. Watch wolves dominate and sheep go extinct.
3. Use More Sheep to reset to a sheep-heavy initial condition and compare the transients.
4. Reduce Grass Regrowth to 1. Can the ecosystem sustain itself?

Explore in the library:

🐺 Wolf-Sheep Predation 🐦 Boids Flocking 🪸 Coral Reef 🦠 Bacteria Colony

The brain is full of oscillating neurons: gamma oscillations (40 Hz) underlie attention and consciousness, theta rhythms (4–8 Hz) are linked to memory and navigation, alpha waves (8–12 Hz) reflect idle-state cortical synchrony. How does a population of neurons with diverse intrinsic frequencies settle into collective oscillation? The Kuramoto model provides a mathematical answer: a phase transition.

From individual oscillators to collective rhythm

In the simulation, each oscillator represents a neuron (or neural population) with its own natural firing rate ωᵢ. The coupling term (K/N)Σsin(φⱼ−φᵢ) acts like synaptic excitation: neurons try to fire in phase with their neighbours. Below K_c the network is asynchronous — each neuron fires at its own rate, and the population average cancels out. Above K_c a macroscopic oscillation emerges that is not present in any individual neuron's program.

Biological manifestations

• Circadian clock — Neurons in the suprachiasmatic nucleus (SCN) synchronize to produce the 24-hour body clock. Jet lag is a desynchronization event.
• Cardiac arrhythmia — Pacemaker cells lose synchrony during atrial fibrillation; defibrillation is a forced re-synchronization.
• Parkinson's disease — Pathological over-synchrony in dopamine circuits (beta-band locking) is now a target for deep brain stimulation.
• Fireflies — Pteroptyx malaccae synchronize their flashing across entire trees, a macroscopic Kuramoto phenomenon visible to the naked eye.

Explore in the library:

🔄 Kuramoto Synchronization 🧠 Brainwave Oscillations ❤️ Cardiac Action Potential 🕐 Circadian Rhythm