What are the Lotka-Volterra equations?

The Lotka-Volterra equations are a pair of first-order nonlinear ODEs that describe the dynamics of biological systems in which two species interact as predator and prey. Prey grow at rate α but are removed by predation at rate β per predator: dN/dt = αN − βNP. Predators grow by consuming prey at efficiency δ but die at natural rate γ: dP/dt = δNP − γP. Solutions are closed periodic orbits in phase space — both populations oscillate indefinitely out of phase, with predator cycles lagging behind prey cycles.

Why do predator and prey populations oscillate?

When prey are abundant, predators have plentiful food and their numbers increase. The growing predator population then over-harvests prey, causing prey numbers to crash. With little food, predator numbers then decline. The reduced predation pressure allows prey to recover, and the cycle begins again. This negative feedback loop produces perpetual oscillations. The Lotka-Volterra model predicts that these oscillations are undamped — the trajectory is a closed loop in phase space — a property unique to this model and not typically seen in real ecosystems which usually have damping.

What is the equilibrium point of the Lotka-Volterra model?

The non-trivial equilibrium (coexistence point) is N* = γ/δ and P* = α/β. At these population levels both dN/dt = 0 and dP/dt = 0 simultaneously — the system is stationary. However this equilibrium is neutrally stable (a centre, not an attractor): any perturbation causes the system to orbit around it at a different amplitude forever, neither spiralling in nor out. This contrasts with stable equilibria where perturbations decay. In the phase portrait, the equilibrium appears as a fixed point surrounded by a family of closed orbits.

How do Lotka-Volterra equations compare to real predator-prey data?

The classic real-world dataset is the Hudson's Bay Company records of Canadian lynx and snowshoe hare pelts from 1845–1935, which show oscillations with a ~10-year period. The Lotka-Volterra model captures the qualitative behaviour (out-of-phase oscillations, predator lags behind prey) but real systems show damped oscilations, more irregular cycles, and phase lags that differ from the model's predictions. More realistic models add logistic prey growth (carrying capacity K), functional response curves (Type II/III Holling), age structure, and spatial heterogeneity — all extensions of the basic Lotka-Volterra framework.

Home▸Physics & Mechanics▸Predator-Prey Simulator | Lotka-Volterra Population Dynamics

🧪 Predator-Prey Simulator | Lotka-Volterra Population Dynamics

Interactive predator-prey simulator solving the Lotka-Volterra equations in real time. Adjust prey birth rate, predation rate, predator efficiency, and death rate to explore population cycles, coexistence equilibria, and phase portraits. GCSE, A-Level, IB Biology and Mathematics.

Physics & Mechanics2DModerate60 FPS

The Mathematics of Predator-Prey Dynamics

🌿 The Lotka-Volterra Model

Independently proposed by Alfred Lotka (1920) and Vito Volterra (1926), the Lotka-Volterra equations model the interaction between a prey species (N) and a predator species (P). Prey reproduce at rate α but are removed by predator encounters at rate βNP. Predators grow by consuming prey at efficiency δ but die naturally at rate γP. Together these give a system of two coupled nonlinear ODEs whose solutions are periodic orbits in phase space.

📈 Coexistence Equilibrium

Setting both derivatives to zero reveals the coexistence equilibrium: N* = γ/δ and P* = α/β. Strikingly, N* depends only on predator parameters (γ, δ) and P* only on prey parameters (α, β). This means increasing the prey birth rate α does not increase the equilibrium prey level — it only increases the equilibrium predator level. This counter-intuitive result, known as the Volterra effect, has real consequences for pest control.

🔄 Neutral Stability

The coexistence equilibrium is neutrally stable — a centre, not an attractor. Any initial condition (N₀, P₀) near but not at the equilibrium produces a closed orbit that repeats indefinitely: the populations oscillate forever at constant amplitude. This distinguishes Lotka-Volterra from models with carrying capacity (Rosenzweig-MacArthur), where the equilibrium is a spiral attractor and perturbations decay. The phase portrait shows a family of concentric closed loops around the equilibrium point.

🔡 Phase Lag

In the time-series plot, predator peaks follow prey peaks by approximately one-quarter cycle. This phase lag is a direct consequence of the dynamics: prey must first become abundant before predators have enough food to reproduce. Biologists use this lag as a qualitative signature of predator-prey coupling in field data — it was observed in the famous Hudson's Bay Company lynx–hare time series (1845–1935), which prompted Volterra to refine his model.

Key Equations

Equation / Symbol	Meaning
dN/dt = αN − βNP	Rate of change of prey: birth − predation losses
dP/dt = δNP − γP	Rate of change of predators: prey-fuelled growth − natural death
N* = γ/δ	Coexistence equilibrium prey level (depends only on predator parameters)
P* = α/β	Coexistence equilibrium predator level (depends only on prey parameters)
α	Prey intrinsic growth rate (yr⁻¹)
β	Predation rate coefficient (per predator per prey)
δ	Predator reproduction efficiency per prey consumed
γ	Predator natural death rate (yr⁻¹)
V = δN − γln(N) + βP − αln(P)	Lyapunov / conserved quantity — constant on each orbit

Ecological Scenarios

🐇 Snowshoe Hare & Lynx

The most-studied real predator-prey system. Hudson’s Bay Company pelt records show ~10-year cycles with hare numbers ranging 1,000-fold and lynx following with a lag of 1–2 years. The Lotka-Volterra model explains the basic oscillation but not the precise period or amplitude. Additional mechanisms include hare-vegetation dynamics and delayed density dependence.

🦊 Wolves & Elk (Yellowstone)

After wolves were reintroduced to Yellowstone in 1995 (removed in 1926), elk populations stabilised and riparian vegetation recovered — a trophic cascade. This shows how predators regulate prey beyond simple numbers: the fear of predation changes prey behaviour (foraging patterns, habitat use) in ways the basic Lotka-Volterra model cannot capture. Top-down trophic effects are a major area of modern ecology.

🐟 Plankton Dynamics

Marine phytoplankton (prey) and zooplankton (predators) exhibit rapid oscillations on timescales of days to weeks. Their dynamics are crucial for ocean carbon cycling — phytoplankton fix CO₂ via photosynthesis. Lotka-Volterra-type models adapted for nutrient limitation and light availability form the basis of ocean biogeochemical models used in climate science.

🍋 Pest Control

The Volterra effect predicts that indiscriminate pesticide use, which kills both pest (prey) and natural predator, results in higher pest equilibrium levels after treatment than before. This phenomenon was originally observed by Volterra analysing fishing data from the Adriatic Sea during WWI. It remains a key argument for biologically integrated pest management over broad-spectrum pesticide application.

Model Extensions

🔄 Logistic Prey Growth

Adding a carrying capacity K to the prey: dN/dt = αN(1 − N/K) − βNP. This is the Rosenzweig-MacArthur model. The equilibrium becomes a stable spiral attractor — perturbations now dampen rather than orbit indefinitely. For sufficiently large K, a Hopf bifurcation occurs: the stable spiral loses stability and a limit cycle appears (“paradox of enrichment”).

📈 Holling Functional Response

Real predators do not consume prey at a constant rate βN — at high prey density they become satiated. Type II Holling response: f(N) = βN/(1 + hβN), where h is handling time. Type III: sigmoid response, common in generalist predators. These functional responses replace the linear predation term and produce richer dynamics including stable limit cycles and bistability.

🗻 Spatial Models

When prey and predators are distributed in space with diffusion (or dispersal), the Lotka-Volterra PDEs can produce travelling waves — invasions of prey followed by predator pursuit fronts, spiral waves, and spatiotemporal chaos. These patterns are analogous to Turing patterns in the reaction-diffusion model and are observed in real ecosystems (vole–weasel cycles in Scandinavia).

👥 Multi-Species Food Webs

Generalised Lotka-Volterra (GLV) models extend to N species: dxᵢ/dt = xᵢ(rᵢ + Σⱼ aᵢⱼxⱼ), where the interaction matrix aᵢⱼ encodes all pairwise interactions (predation, competition, mutualism). GLV models are the foundation of theoretical community ecology and are used to study the relationship between food-web complexity, stability, and biodiversity.

Curriculum Alignment

Level	Topic	Concepts
GCSE Biology	Ecology; Interdependence	Food chains, predator-prey relationships, population size
A-Level Biology	Ecology; Population Dynamics	Interspecific competition, predation graphs, carrying capacity
A-Level Mathematics	Differential Equations	Coupled ODEs, phase plane analysis, equilibrium stability
IB Biology SL/HL	Ecology; Topic 4 / Option C	Population models, trophic levels, ecological efficiency
IB Maths AA HL	Differential Equations	Systems of ODEs, phase portraits, Euler/RK methods
AP Biology	Free Energy & Ecosystems	Population dynamics, species interactions, food web
University Year 1–3	Mathematical Biology; Ecology	Lotka-Volterra, Rosenzweig-MacArthur, Holling responses, GLV, stability analysis

Worked Example

Finding the equilibrium and conserved quantity

Parameters: α = 0.4, β = 0.01, δ = 0.005, γ = 0.2, starting at N₀ = 40, P₀ = 9.

Equilibrium: N* = γ/δ = 0.2/0.005 = 40. P* = α/β = 0.4/0.01 = 40.

Conserved quantity V = δN − γln(N) + βP − αln(P)
At N=40, P=9: V = 0.005(40) − 0.2 ln(40) + 0.01(9) − 0.4 ln(9) ≈ 0.2 − 0.738 + 0.09 − 0.878 = −1.326

Since N₀ = N* = 40 but P₀ = 9 ≠ P* = 40, the system starts off-equilibrium and will oscillate. Had we started exactly at (40, 40) it would remain there forever.

Prediction: Prey will peak before predators. Since α = γ = 0.2 × 2, the cycle period is roughly T ≈ 2π/√(αγ) ≈ 2π/√(0.08) ≈ 22 years. Verify this in the simulator using the Long Cycles preset (similar parameters).

Frequently Asked Questions

Can both populations go extinct?

In the pure Lotka-Volterra model (no carrying capacity), if either population reaches exactly zero the dynamics halt. In practice, the RK4 integrator clamps populations at zero to prevent negative values. In real ecosystems, stochastic effects can drive populations to extinction even when the deterministic Lotka-Volterra model predicts oscillation — this is why conservation biologists worry about small population bottlenecks during cycle troughs.

Why is the equilibrium neutrally stable rather than attracting?

The standard Lotka-Volterra system conserves a Lyapunov function V = δN − γln(N) + βP − αln(P), which is constant on every orbit. Because V is conserved, no energy is dissipated and trajectories cannot spiral inward. Mathematically, the eigenvalues of the Jacobian at the equilibrium are purely imaginary (±i√(αγ)), giving a centre rather than a spiral. Adding any form of density dependence or realistic functional response breaks this conservation and produces a true attractor.

How does this relate to the SIR epidemic model?

Both the Lotka-Volterra and SIR models are systems of coupled nonlinear ODEs describing population flows between compartments. They share a mathematical structure — nonlinear interaction terms (βNP and βSI respectively) — and both exhibit threshold behaviour. The SIR model has a single epidemic and then reaches a disease-free equilibrium; Lotka-Volterra oscillates indefinitely (neutrally stable) rather than settling. Both can be explored on MySimulator.uk.

What is the Volterra effect and why does it matter for farming?

The Volterra effect states that proportional harvesting of both predator and prey (as with a non-selective pesticide) raises the average prey population and lowers the average predator population. Mathematically the equilibrium shifts: increasing the “death term” for both species changes N* = γ/δ upward and P* = α/β downward. In agriculture this means that broad-spectrum insecticides killing both pest insects (prey) and their natural predator parasites can lead to a pesticide rebound — a worse pest problem after spraying than before.

Can I use this simulator for my IB extended essay?

Yes. Good IB EE research questions include: "How does changing the predation efficiency δ affect the period and amplitude of population oscillations?"; "Compare population dynamics under Lotka-Volterra vs a logistic prey model (Rosenzweig-MacArthur)"; or "How does the initial N/P ratio affect oscillation amplitude?" All three can be investigated systematically in this simulator by varying one parameter at a time and recording the phase portrait and period.

🧪 Predator-Prey Simulator | Lotka-Volterra Population Dynamics

The Mathematics of Predator-Prey Dynamics

🌿 The Lotka-Volterra Model

📈 Coexistence Equilibrium

🔄 Neutral Stability

🔡 Phase Lag

Key Equations

Ecological Scenarios

🐇 Snowshoe Hare & Lynx

🦊 Wolves & Elk (Yellowstone)

🐟 Plankton Dynamics

🍋 Pest Control

Model Extensions

🔄 Logistic Prey Growth

📈 Holling Functional Response

🗻 Spatial Models

👥 Multi-Species Food Webs

Curriculum Alignment

Worked Example

Finding the equilibrium and conserved quantity

Frequently Asked Questions

Further Reading

🆕 Unlock All 33 Simulations

What did you find?

🧪 Predator-Prey Simulator | Lotka-Volterra Population Dynamics

The Mathematics of Predator-Prey Dynamics

🌿 The Lotka-Volterra Model

📈 Coexistence Equilibrium

🔄 Neutral Stability

🔡 Phase Lag

Key Equations

Ecological Scenarios

🐇 Snowshoe Hare & Lynx

🦊 Wolves & Elk (Yellowstone)

🐟 Plankton Dynamics

🍋 Pest Control

Model Extensions

🔄 Logistic Prey Growth

📈 Holling Functional Response

🗻 Spatial Models

👥 Multi-Species Food Webs

Curriculum Alignment

Worked Example

Finding the equilibrium and conserved quantity

Frequently Asked Questions

Related Simulations

Further Reading

🆕 Unlock All 33 Simulations

Share this Simulation