Questions: Pattern Formation and Turing Instability

Question 1

A system of two chemicals has a stable equilibrium when well-mixed. When the chemicals are allowed to diffuse in space, spots and stripes appear. This seems paradoxical because diffusion should smooth things out. What resolves the paradox?

• A. Diffusion always creates patterns — the well-mixed state was an artifact of stirring
• B. The two chemicals diffuse at very different rates. The inhibitor diffuses faster, spreading out and suppressing the activator at long range while the activator amplifies itself locally. This mismatch creates a local activation / long-range inhibition dynamic that destabilizes the uniform state for specific spatial wavelengths.
• C. The chemicals react with the container walls, creating patterns at the boundaries
• D. Numerical errors in the simulation create spurious patterns

Answer: (B)

Turing's insight (1952) was that differential diffusion — different species moving at different rates — can destabilize a uniform equilibrium. The activator amplifies itself and the inhibitor, but diffuses slowly. The inhibitor suppresses both, but diffuses rapidly. The inhibitor 'runs away' from a local fluctuation, leaving the activator free to grow locally but suppressed at a distance. The result: regularly spaced peaks of activator separated by inhibitor-dominated valleys. The wavelength of the pattern is selected by the diffusion ratio and the reaction kinetics.

Question 2

Turing instability requires that the two species diffuse at different rates. If both diffuse at the same rate, can patterns still form?

• T. True
• F. False

Answer: False

Equal diffusion rates cannot produce Turing instability. The mathematical condition requires D_inhibitor/D_activator > some threshold (typically much greater than 1). If diffusion rates are equal, the diffusion operator acts as a scalar multiple of the Laplacian on the vector of concentrations, and if the well-mixed steady state is stable, adding equal diffusion only makes it more stable (diffusion can't destabilize what reaction kinetics already stabilized). The differential diffusion rate is essential — it creates the spatial scale separation between local activation and long-range inhibition.

Question 3

Turing's reaction-diffusion mechanism has been proposed to explain the stripe and spot patterns on animal skins. A key prediction is that the type of pattern (spots vs. stripes) depends on the geometry of the domain. What does this mean?

• A. The same chemical parameters produce spots on a wide body and stripes on a thin tail or leg, because the geometry constrains which spatial modes (wavelengths) can fit
• B. The patterns are painted on by genes, not by chemical reactions
• C. The geometry has no effect — spots and stripes are determined entirely by chemical concentrations
• D. Stripes only form on flat surfaces, spots only on curved surfaces

Answer: (A)

On a wide surface (like a torso), many wavelengths fit in both directions, and the interaction between 2D modes produces spots or labyrinths. On a narrow domain (like a tail), only one mode fits across the width, forcing stripes along the length. This prediction, confirmed in many species, is striking: a leopard's spots on its body become stripes on its tail, consistent with the same reaction-diffusion parameters on different geometries. Murray's famous quip: 'a spotted animal can have a striped tail, but a striped animal cannot have a spotted tail' follows from this geometry dependence.

Question 4

How does the Turing instability relate to the bifurcation theory you studied earlier?

Model answer: Turing instability is a bifurcation in an infinite-dimensional system (a PDE). The control parameter is typically the diffusion ratio or a reaction rate. At the critical value, a spatially uniform fixed point loses stability to a perturbation with a specific wave number k_c — this is analogous to a pitchfork bifurcation (the pattern breaks the spatial symmetry). The amplitude of the pattern grows from zero (supercritical) or jumps to finite amplitude (subcritical), just as in finite-dimensional bifurcations. The selected wavelength λ = 2π/k_c is determined by the dispersion relation — the wavenumber at which the growth rate first becomes positive.

The connection to bifurcation theory is deep and systematic. Near the onset of Turing instability, amplitude equations (like the Ginzburg-Landau equation) describe the slow evolution of the pattern envelope, reducing the infinite-dimensional PDE to a finite-dimensional bifurcation problem. The type of bifurcation (supercritical vs. subcritical) determines whether the pattern appears gradually or suddenly. Multiple-scale analysis and symmetry considerations (which patterns — rolls, hexagons, squares — are favored) connect directly to the equivariant bifurcation theory of systems with spatial symmetry.