Flow Lab

Syllabus / Module 15: Applications: Biology, Chemistry & Society

Lesson 15.2

Predator–Prey & Competing Species

Lotka–Volterra, conserved quantity, functional responses, Rosenzweig–MacArthur limit cycles, competition models.

Written byClaude Sonnet

🌊
Why this matters for simulation: Conserved quantities in LV mirror energy conservation in inviscid flow.

By the end of this lesson you can…

  • Derive the conserved quantity of the Lotka-Volterra equations, explain why its orbits are closed, and say why that property is structurally fragile.
  • Classify the fixed points of a planar system from the trace and determinant of the Jacobian, and say when linearisation is not enough.
  • Build the Rosenzweig-MacArthur model from a saturating Holling type II response and locate its Hopf bifurcation at $K = 1 + 2x^*$.
  • Explain the paradox of enrichment from the size of the limit cycle and the depth of the prey minimum.
  • Decide among coexistence, competitive exclusion and bistability for two competing species from nullclines and invasion conditions.
  • Write and test a numerical Jacobian and a trace-determinant classifier, and use them to find a Hopf point.
On this page
·
  1. Predators as a dynamic harvest
  2. Lotka-Volterra and its conserved quantity
  3. Structural fragility and the numerics that expose it
  4. Fixed points, Jacobians and the classification
  5. A saturating predator: Holling type II
  6. The limit cycle and the paradox of enrichment
  7. Competing species
  8. The exercise: classify fixed points numerically
  9. Concept checks
  10. Problems
  11. What comes next
  12. Key Takeaways

15.1 Population Models ended with a harvest: a constant effort EE removes ENE N and the stock settles at K(1−E/r)K(1 - E/r). Now let the harvester be alive. Its numbers yy grow only by eating the prey xx, so the “effort” is a variable that depends on how much prey there is. The state is a point in the plane, and the tools are those of 7.3 Phase Portraits & the Trace–Determinant Plane and 7.7 Nonlinear Systems & Linearization: nullclines, fixed points, a Jacobian at each and a trace-determinant classification. Three models test them: one with a conservation law, one with a limit cycle, and one that decides which of two species survives.

▶ Video Lotka-Volterra orbits and the conserved quantity, why Euler spirals outward, the Holling response and the Hopf bifurcation of Rosenzweig-MacArthur, enrichment, and the four outcomes of competitioncaptions only
Transcript
  1. A predator is a harvester whose effort is itself a variable. Scale the Lotka-Volterra model so that x prime is x times one minus y, and y prime is alpha y times x minus 1. The nullclines are x equals 1 and y equals 1, and the fixed points are 0 0, a saddle, and 1 1, where the trace is 0, the determinant is alpha, and the eigenvalues are plus or minus i root alpha. That is a centre of the linearisation, with period 2 pi over root alpha, which cannot say whether orbits close or spiral.
  2. Dividing the two equations separates the variables, and integrating gives H equals alpha times x minus ln x, plus y minus ln y. This is constant along every solution, so the orbits are closed level curves around 1 1. With alpha equal to 1 and start 0.5, 0.8, H is 2.2163. Prey and predators both swing between 0.4779 and 1.8091, with period 6.5117, longer than the small orbit period of 2 pi, 6.2832. Each time average is exactly 1.
  3. But this is structurally fragile. Add a little prey crowding, epsilon, to the prey equation. The fixed point moves, the trace becomes minus epsilon, and the determinant is alpha times one minus epsilon. For epsilon equal to 0.1 the eigenvalues are minus 0.05 plus or minus 0.9474 i: a stable spiral. From 0.5, 0.8 the run reaches 0.9957, 0.8996 at t equal to 100. The conservation law was an accident of the exact form.
  4. The same fragility lives in the numerical method. Forward Euler multiplies the distance from a centre by the square root of 1 plus omega squared h squared at every step, so the orbit spirals outward. Up to time 40, Euler with step 0.01 ends with H up by 4.65 percent, and with step 0.001 by 0.39 percent, first order in h. Fourth order Runge-Kutta with step 0.02 ends at minus 9.4 times 10 to the minus 11. Euler turns a conservative system into an unstable one.
  5. A predator spends time handling each prey, so it eats a x over 1 plus a T h x per unit time: a Holling type 2 response, saturating at 1 over T h. Put that into logistic prey and you get Rosenzweig-MacArthur. With e equal to 1 and m equal to one half, the predator nullcline is the vertical line x star equals 1, and the coexistence point is x star equals 1, y star equals 2 times 1 minus 1 over K. The determinant is positive, so the trace decides everything: tau equals one half minus 3 over 2 K.
  6. The trace is negative for K below 3, zero at K equals 3, and positive beyond. At K equal to 3 the eigenvalues are plus or minus 0.4082 i, a Hopf bifurcation with onset period 15.39. At 2.7 trajectories spiral into the point. At 3.3 the fixed point repels, but the flow is trapped, so by the Poincare-Bendixson theorem a limit cycle must exist. It is isolated and attracting, so its amplitude belongs to the model, not to the start, and grows like the square root of K minus 3: a supercritical Hopf bifurcation.
  7. Now enrich the system by raising K. The prey maximum rises towards K, but the minimum collapses. At K equal to 5, prey run from 0.0357 to 4.4778, with period 21.59. At K equal to 10 the minimum is 2.5 times 10 to the minus 5, a drop by a factor of about 1400, and the predator minimum falls from 0.159 to 0.0036. Such troughs invite chance extinction. This is Rosenzweig's paradox of enrichment: more food, a less stable community.
  8. Now two species that share a resource, each hurting the other through coefficients alpha 12 and alpha 21. At the corner where species 1 sits at capacity, the eigenvalue for species 2 is r 2 times 1 minus alpha 21 K 1 over K 2, and positive means species 2 can invade. With 0.5 and 0.7 both can invade: coexistence at 0.7692, 0.4615, a stable node. With 1.5 and 1.3 neither can: both corners are stable and the interior point is a saddle, so whoever starts ahead wins. With 0.5 and 1.5, or 1.5 and 0.5, one species excludes the other. The interior point is stable exactly when alpha 12 times alpha 21 is below 1.
  9. Next, epidemics are another coupled nonlinear system, with the same mass action product S times I. Dividing dI by dS gives an exact conserved quantity, I plus S minus gamma over beta times ln S, like H here. But it does not oscillate: susceptibles are only used up, so the orbit runs to a final state on an axis. What replaces the Hopf bifurcation is a threshold: I first grows only if R 0, beta S 0 over gamma, exceeds 1, a close relative of the invasion tests above.

Predators as a dynamic harvest

Keep the prey growth x p(x)x\,p(x) and add a loss g(x) x yg(x)\,x\,y, where gg is the functional response (prey eaten per predator per unit time, per available prey). The predator converts eaten prey into births at rate ee and dies at rate mm:

x′=x p(x)−g(x) x y,y′=y(e g(x) x−m).x' = x\,p(x) - g(x)\,x\,y, \qquad y' = y\left(e\,g(x)\,x - m\right).

Constant pp and gg give the Lotka-Volterra model (1925 and 1926, from memory), solved by one clever integral. A logistic pp with a saturating gg gives Rosenzweig-MacArthur (1963, from memory), solved by geometry and a bifurcation.

Lotka-Volterra and its conserved quantity

Scale time and the two populations so that the equations are

x′=x(1−y),y′=α y (x−1).x' = x(1 - y), \qquad y' = \alpha\,y\,(x - 1).

Prey grow when predators are fewer than 11, predators grow when prey exceed 11, and α\alpha is the ratio of the two time scales. The nullclines are x=0x = 0, y=1y = 1 for x′x' and y=0y = 0, x=1x = 1 for y′y', so the fixed points are (0,0)(0, 0) and (1,1)(1, 1). The Jacobian is

J=(1−y−xαyα(x−1)),J(0,0)=(100−α),J(1,1)=(0−1α0).J = \begin{pmatrix} 1 - y & -x \\ \alpha y & \alpha(x - 1)\end{pmatrix}, \qquad J(0,0) = \begin{pmatrix}1 & 0\\0 & -\alpha\end{pmatrix}, \qquad J(1,1) = \begin{pmatrix}0 & -1\\ \alpha & 0\end{pmatrix}.

The origin is a saddle (eigenvalues 11 and −α-\alpha). At (1,1)(1,1) the trace is 00, the determinant α\alpha and the eigenvalues ±iα\pm i\sqrt\alpha: a centre of the linearisation, with period 2π/α2\pi/\sqrt\alpha. A centre is the borderline case where the linearisation is silent, since a tiny nonlinear term can tip it into a spiral. Here nothing does, because of a conservation law.

Derivation. Divide the equations:

dydx=αy(x−1)x(1−y)⟹1−yy dy=α x−1x dx.\frac{dy}{dx} = \frac{\alpha y (x - 1)}{x(1 - y)} \quad\Longrightarrow\quad \frac{1 - y}{y}\,dy = \alpha\,\frac{x - 1}{x}\,dx.

The variables separate. Integrating, ln⁡y−y=α(x−ln⁡x)+const\ln y - y = \alpha(x - \ln x) + \text{const}, so

H(x,y)=α (x−ln⁡x)+(y−ln⁡y)H(x, y) = \alpha\,(x - \ln x) + (y - \ln y)

is constant along every solution. (Check: H′=α(1−1/x) x(1−y)+(1−1/y) αy(x−1)=α(x−1)(1−y)+α(y−1)(x−1)=0H' = \alpha(1 - 1/x)\,x(1 - y) + (1 - 1/y)\,\alpha y(x - 1) = \alpha(x - 1)(1 - y) + \alpha(y - 1)(x - 1) = 0.) Each of x−ln⁡xx - \ln x and y−ln⁡yy - \ln y has a single minimum, at 11, and grows without bound as the variable goes to 00 or ∞\infty. So the level sets of HH are closed curves around (1,1)(1,1): a smooth bowl, and the solutions are the contour lines. The centre is a true nonlinear centre. This is the energy argument of 7.8 Conservative Systems & Lyapunov Functions in a new setting: HH plays the role of energy, with a surface that is not quadratic.

Numbers. With α=1\alpha = 1 and start (0.5,0.8)(0.5, 0.8), H=2.2163H = 2.2163. The orbit swings between 0.47790.4779 and 1.80911.8091 in both variables with period 6.51176.5117, longer than the small-orbit period 2π=6.28322\pi = 6.2832: unlike the harmonic oscillator, the period depends on the energy. A separate scipy run (tolerance 10−1210^{-12}) agrees with the sim to four decimals.

Volterra’s averages. Since x′/x=1−yx'/x = 1 - y and ln⁡x\ln x returns to its starting value over a period, ∫0P(1−y) dt=0\int_0^P (1 - y)\,dt = 0: the time average of yy is 11. Likewise y′/y=α(x−1)y'/y = \alpha(x - 1) gives an average prey level of 11. The averages equal the fixed point whatever the amplitude, and for the orbit above both are 1.0001.000 (within 10−310^{-3} in the sim). In the unscaled model x′=ax−bxyx' = ax - bxy, y′=−cy+dxyy' = -cy + dxy the averages are c/dc/d and a/ba/b; a per-capita kill rate qq on both species replaces a→a−qa \to a - q, c→c+qc \to c + q, so the average prey goes up and the average predator goes down (Volterra’s principle, from memory).

Structural fragility and the numerics that expose it

Lotka-Volterra is structurally fragile: a small change in the equations changes the phase portrait. Add a little crowding, x′=x(1−εx−y)x' = x(1 - \varepsilon x - y). The fixed point moves to (1,1−ε)(1, 1 - \varepsilon) and the Jacobian is

J=(−ε−1α(1−ε)0),tr⁡J=−ε,det⁡J=α(1−ε).J = \begin{pmatrix} -\varepsilon & -1 \\ \alpha(1 - \varepsilon) & 0\end{pmatrix}, \qquad \operatorname{tr}J = -\varepsilon, \quad \det J = \alpha(1 - \varepsilon).

The trace is no longer 00. For ε=0.1\varepsilon = 0.1, α=1\alpha = 1 the eigenvalues are −0.05±0.9474i-0.05 \pm 0.9474i, a stable spiral decaying at rate ε/2\varepsilon/2. From (0.5,0.8)(0.5, 0.8) the run reaches (0.9957,0.8996)(0.9957, 0.8996) at t=100t = 100 (RK4 and scipy agree). The conserved quantity was an accident of the exact form, and a neutral orbit, which remembers its start forever, is a poor model of a population cycle anyway.

The same fragility lives in the method. Forward Euler multiplies the distance from a centre by 1+ω2h2>1\sqrt{1 + \omega^2 h^2} \gt 1 at every step. With α=1\alpha = 1, start (0.5,0.8)(0.5, 0.8) and T=40T = 40, Euler with h=0.01h = 0.01 ends at H/H0−1=+4.65%H/H_0 - 1 = +4.65\% and with h=0.001h = 0.001 at +0.39%+0.39\% (first order in hh), while RK4 with h=0.02h = 0.02 gives −9.4×10−11-9.4\times10^{-11}. Euler turns a conservative system into an unstable one, see 4.7 Energy & Symplectic Integrators.

◆ Interactive Lotka-Volterra: click to set a start, switch between RK4 and Euler, then add prey crowding eps and watch the closed orbit become a spiral

Loading simulation…

Fixed points, Jacobians and the classification

Every planar model in this lesson is handled by one routine, which is the planar case of 7.7 Nonlinear Systems & Linearization.

  1. Find the nullclines of x′=fx' = f and y′=gy' = g, and intersect them to get the fixed points.
  2. At each, form J=(fxfygxgy)J = \begin{pmatrix}f_x & f_y\\ g_x & g_y\end{pmatrix} and compute τ=tr⁡J\tau = \operatorname{tr}J, Δ=det⁡J\Delta = \det J.
  3. Classify. Δ<0\Delta \lt 0: saddle. Δ>0\Delta \gt 0 and τ2>4Δ\tau^2 \gt 4\Delta: node, stable if τ<0\tau \lt 0. Δ>0\Delta \gt 0 and τ2<4Δ\tau^2 \lt 4\Delta: spiral, stable if τ<0\tau \lt 0. τ=0\tau = 0 with Δ>0\Delta \gt 0: linear centre.
  4. Trust the result except on the borderline (τ=0\tau = 0, Δ=0\Delta = 0 or repeated roots), where the nonlinear terms decide. (Hartman and Grobman, from memory: near a hyperbolic fixed point the nonlinear flow is topologically a copy of the linear one.)

A saturating predator: Holling type II

A predator spends time handling each prey. With attack rate aa and handling time ThT_h, it eats per unit time (Holling’s disc equation, 1959, from memory)

a x1+aTh x.\frac{a\,x}{1 + a T_h\,x}.

For small xx this is linear (the Lotka-Volterra term), and for large xx it saturates at 1/Th1/T_h. Scale so that the half-saturation is 11 and take logistic prey. The Rosenzweig-MacArthur model is

x′=x(1−xK)−x y1+x,y′=y(e x1+x−m).x' = x\left(1 - \frac xK\right) - \frac{x\,y}{1 + x}, \qquad y' = y\left(\frac{e\,x}{1 + x} - m\right).

We use e=1e = 1 and m=12m = \tfrac12. The predator nullcline is the vertical line x=x∗=m/(e−m)=1x = x^* = m/(e - m) = 1. The prey nullcline is the hump y=(1−x/K)(1+x)y = (1 - x/K)(1 + x): it starts at y=1y = 1, peaks at x=(K−1)/2x = (K - 1)/2 and vanishes at x=Kx = K. At (K,0)(K, 0) the predator growth rate is eK/(1+K)−meK/(1 + K) - m, positive iff K>1=x∗K \gt 1 = x^*, so predators invade only then, and the nullclines meet at

x∗=1,y∗=2(1−1K).x^* = 1, \qquad y^* = 2\left(1 - \frac1K\right).

Jacobian at the coexistence point. Using ∂x[x/(1+x)]=1/(1+x)2\partial_x[x/(1 + x)] = 1/(1 + x)^2:

J=(1−2x∗K−y∗(1+x∗)2−x∗1+x∗y∗(1+x∗)20).J = \begin{pmatrix} 1 - \dfrac{2x^*}{K} - \dfrac{y^*}{(1 + x^*)^2} & -\dfrac{x^*}{1 + x^*} \\[2mm] \dfrac{y^*}{(1 + x^*)^2} & 0 \end{pmatrix}.

The determinant is Δ=12⋅14 y∗=14(1−1/K)>0\Delta = \tfrac12 \cdot \tfrac14\, y^* = \tfrac14(1 - 1/K) \gt 0, so the point is never a saddle, and the sign of the trace decides everything. Simplifying with y∗=(1−x∗/K)(1+x∗)y^* = (1 - x^*/K)(1 + x^*),

τ=x∗ (K−1−2x∗)K (1+x∗)=12−32Kfor x∗=1.\tau = \frac{x^*\,(K - 1 - 2x^*)}{K\,(1 + x^*)} = \frac12 - \frac{3}{2K}\quad\text{for } x^* = 1.

It is negative for K<1+2x∗=3K \lt 1 + 2x^* = 3, zero at K=3K = 3, positive beyond. Geometrically, the point is stable when the predator nullcline cuts the hump to the right of its top, and unstable when it cuts the rising side. At K=3K = 3 the eigenvalues are ±i1/6=±0.4082i\pm i\sqrt{1/6} = \pm 0.4082i: a Hopf bifurcation (3.9 Bifurcations, 7.9 Limit Cycles & Hopf Bifurcations) with onset period 2π/0.4082=15.392\pi/0.4082 = 15.39.

KKτ\tauΔ\Deltaeigenvaluesfixed point
2.72.7−0.0556-0.05560.15740.1574−0.0278±0.3958i-0.0278 \pm 0.3958istable spiral
33000.16670.1667±0.4082i\pm 0.4082iborderline
3.33.3+0.0455+0.04550.17420.17420.0227±0.4168i0.0227 \pm 0.4168iunstable spiral
55+0.2+0.20.20.20.1±0.4359i0.1 \pm 0.4359iunstable spiral

The sim’s numerical Jacobian and numpy agree on every digit shown.

The limit cycle and the paradox of enrichment

Above K=3K = 3 the fixed point repels, but the flow is trapped in a bounded region (the prey self-limit, the predator starves without prey). The planar Poincaré-Bendixson theorem (from memory: a bounded trajectory that does not approach a fixed point approaches a periodic orbit) forces a limit cycle. Unlike the Lotka-Volterra orbits it is isolated and attracting, so amplitude and period belong to the model, not to the start.

Long runs (RK4 in the sim, scipy DOP853 at tolerance 10−1210^{-12} for comparison) give the cycle:

KKprey min – maxpredator min – maxperiod
3.13.10.6153−1.51770.6153 - 1.51770.9803−1.71060.9803 - 1.710615.6915.69
3.23.20.4894−1.77600.4894 - 1.77600.8431−1.87600.8431 - 1.876015.9815.98
3.33.30.4048−1.99240.4048 - 1.99240.7428−2.00850.7428 - 2.008516.2816.28
440.1401−3.15660.1401 - 3.15660.3678−2.70400.3678 - 2.704018.4418.44
550.0357−4.47780.0357 - 4.47780.1593−3.56920.1593 - 3.569221.5921.59
885.13×10−4−7.85065.13\times10^{-4} - 7.85060.0166−6.15380.0166 - 6.153831.3831.38
10102.50×10−5−9.93732.50\times10^{-5} - 9.93730.0036−7.90130.0036 - 7.901338.1438.14

Near the bifurcation, the half-width of the prey swing divided by K−3\sqrt{K - 3} is 1.431.43, 1.441.44, 1.451.45 for the first three rows (computed from the table): the cycle grows like K−3\sqrt{K - 3}, the mark of a supercritical Hopf bifurcation. The period starts at 15.3915.39 and lengthens.

Now look at enrichment. Raising KK lifts the prey maximum towards KK but the minimum collapses: from K=5K = 5 to 1010 it falls from 0.03570.0357 to 2.5×10−52.5\times10^{-5}, a factor of about 14001400, and the predator minimum from 0.1590.159 to 0.00360.0036. Real troughs like these are a handful of individuals, and chance extinction follows. This is Rosenzweig’s paradox of enrichment (1971, from memory): more food makes the community less stable.

◆ Interactive Rosenzweig-MacArthur: slide K across 3 and then up to 10; the right panel plots the prey minimum and maximum of the attractor against K

Loading simulation…

Competing species

Now two species that do not eat each other but share a resource, each hurting the other’s growth. The Lotka-Volterra competition model is

x1′=r1x1(1−x1+α12 x2K1),x2′=r2x2(1−x2+α21 x1K2).x_1' = r_1 x_1\left(1 - \frac{x_1 + \alpha_{12}\,x_2}{K_1}\right), \qquad x_2' = r_2 x_2\left(1 - \frac{x_2 + \alpha_{21}\,x_1}{K_2}\right).

Here α12\alpha_{12} counts how many individuals of species 11 one individual of species 22 replaces. The nullclines are x1=0x_1 = 0 or x1+α12x2=K1x_1 + \alpha_{12}x_2 = K_1, and x2=0x_2 = 0 or x2+α21x1=K2x_2 + \alpha_{21}x_1 = K_2. Take K1=r1=1K_1 = r_1 = 1. There are three fixed points on the axes, (0,0)(0,0), (K1,0)(K_1, 0), (0,K2)(0, K_2), and a fourth where the two lines cross, inside the quadrant or outside it.

Corners. At the origin J=diag⁡(r1,r2)J = \operatorname{diag}(r_1, r_2): an unstable node, because tiny populations of both grow. At (K1,0)(K_1, 0) the matrix is triangular, so the eigenvalues are its diagonal entries:

λ=−r1andλ=r2(1−α21K1K2).\lambda = -r_1 \quad\text{and}\quad \lambda = r_2\left(1 - \frac{\alpha_{21}K_1}{K_2}\right).

The first is species 11 at its capacity. The second is the growth rate of a rare species 22 among species 11: positive means species 22 can invade. So species 11 invades at (0,K2)(0, K_2) iff α12K2<K1\alpha_{12}K_2 \lt K_1, and species 22 invades at (K1,0)(K_1, 0) iff α21K1<K2\alpha_{21}K_1 \lt K_2.

Four cases. With K1=K2=1K_1 = K_2 = 1, r1=r2=1r_1 = r_2 = 1 and four choices of the pair (α12,α21)(\alpha_{12}, \alpha_{21}):

(α12,α21)(\alpha_{12}, \alpha_{21})invadesoutcomefixed-point types
(0.5,0.7)(0.5, 0.7)bothcoexistencecorners saddles, interior (0.7692,0.4615)(0.7692, 0.4615) stable node
(1.5,1.3)(1.5, 1.3)neitherbistableboth corners stable nodes, interior (0.5263,0.3158)(0.5263, 0.3158) saddle
(0.5,1.5)(0.5, 1.5)species 11 onlyspecies 11 wins(1,0)(1,0) stable node, (0,1)(0,1) saddle, no interior point
(1.5,0.5)(1.5, 0.5)species 22 onlyspecies 22 wins(0,1)(0,1) stable node, (1,0)(1,0) saddle, no interior point

Worked example (coexistence). For (0.5,0.7)(0.5, 0.7) the lines x1+0.5x2=1x_1 + 0.5x_2 = 1 and x2+0.7x1=1x_2 + 0.7x_1 = 1 cross where x1=(1−0.5)/(1−0.35)=10/13=0.7692x_1 = (1 - 0.5)/(1 - 0.35) = 10/13 = 0.7692 and x2=(1−0.7)/0.65=6/13=0.4615x_2 = (1 - 0.7)/0.65 = 6/13 = 0.4615. At the crossing the brackets in the equations vanish, so only the “outer” derivatives survive:

J=−(x1∗α12 x1∗α21 x2∗x2∗)=−(0.76920.38460.32310.4615).J = -\begin{pmatrix} x_1^* & \alpha_{12}\,x_1^* \\ \alpha_{21}\,x_2^* & x_2^*\end{pmatrix} = -\begin{pmatrix} 0.7692 & 0.3846\\ 0.3231 & 0.4615\end{pmatrix}.

The trace is −1.2308-1.2308 and the determinant x1∗x2∗(1−α12α21)=0.3550×0.65=0.2308x_1^*x_2^*(1 - \alpha_{12}\alpha_{21}) = 0.3550 \times 0.65 = 0.2308, so τ2−4Δ=0.592>0\tau^2 - 4\Delta = 0.592 \gt 0: a stable node, eigenvalues −0.2308-0.2308 and −1.0000-1.0000 (numpy agrees). In general Δ∝1−α12α21\Delta \propto 1 - \alpha_{12}\alpha_{21}, so the interior point is stable iff α12α21<1\alpha_{12}\alpha_{21} \lt 1 and a saddle otherwise. For (1.5,1.3)(1.5, 1.3) the determinant is −0.1579-0.1579 and the eigenvalues +0.1579+0.1579 and −1-1.

Bistability. The stable manifold of the saddle (0.5263,0.3158)(0.5263, 0.3158) separates the basins of the two corners. In the sim the start (0.1,0.8)(0.1, 0.8) ends at (0,1)(0, 1) and (0.8,0.1)(0.8, 0.1) at (1,0)(1, 0): whoever is ahead wins. Only a start exactly on that manifold reaches the interior point.

Competitive exclusion. In rows three and four the winner’s line lies beyond the other’s, there is no interior point, and every trajectory ends at the winner’s corner. The slogan (Gause, 1934, experiments on Paramecium, from memory) is that two species cannot share one limiting resource. The content is the invasion test: coexistence needs each species to limit itself more than its competitor.

◆ Interactive Competition: click the phase plane, then use the four presets and read the nullclines, the eigenvalues at each fixed point and the case name

Loading simulation…

The exercise: classify fixed points numerically

In the exercise you write a central-difference Jacobian, a trace-determinant classifier, and a bisection for the Hopf value of KK, using the facts above. The tests are the table entries.

🐍 PythonEuler against RK4 on the Lotka-Volterra orbit, and the Rosenzweig-MacArthur cycle
⌘/Ctrl + Enter
🧪 Coding exerciseExercise: Jacobian, classifier and the Hopf value of K
⌘/Ctrl + Enter

Concept checks

✅ Concept check

Scaled Lotka-Volterra with $\alpha = 0.25$. What is the period of very small orbits around the fixed point $(1, 1)$?

✅ Concept check

Add a tiny prey crowding term, $x' = x(1 - \varepsilon x - y)$ with $\varepsilon \gt 0$, to the Lotka-Volterra model. What happens near the fixed point?

✅ Concept check

In the Rosenzweig-MacArthur model of this lesson, $K$ is raised from $5$ to $10$. What happens to the attractor?

✅ Concept check

Two competitors with $K_1 = K_2 = 1$ and $\alpha_{12} = 1.5$, $\alpha_{21} = 1.3$. Which statement is correct?

✅ Concept checkselect all that apply

Select all true statements about classifying a planar fixed point from its Jacobian.

Problems

✏️ Pen & paperGeneral Lotka-Volterra●○○

For x′=ax−bxyx' = ax - bxy, y′=−cy+dxyy' = -cy + dxy with positive constants, find the interior fixed point, the eigenvalues of the Jacobian there, and a conserved quantity. For a=1a = 1, b=0.5b = 0.5, c=0.75c = 0.75, d=0.25d = 0.25 give the fixed point and the period of small orbits.

Show worked solution

Setting x′=0x' = 0 and y′=0y' = 0 gives (c/d,a/b)(c/d, a/b). The Jacobian there is (0−bc/dda/b0)\begin{pmatrix}0 & -bc/d\\ da/b & 0\end{pmatrix}, with trace 00 and determinant acac, so λ=±iac\lambda = \pm i\sqrt{ac}. Separating dy/dxdy/dx and integrating gives H=dx−cln⁡x+by−aln⁡yH = dx - c\ln x + by - a\ln y. Check: H′=(d−c/x) x(a−by)+(b−a/y) y(dx−c)=(dx−c)(a−by)+(by−a)(dx−c)=0H' = (d - c/x)\,x(a - by) + (b - a/y)\,y(dx - c) = (dx - c)(a - by) + (by - a)(dx - c) = 0. With the numbers: fixed point (0.75/0.25,1/0.5)=(3,2)(0.75/0.25, 1/0.5) = (3, 2), ac=0.75=0.8660\sqrt{ac} = \sqrt{0.75} = 0.8660, period 2π/0.8660=7.2552\pi/0.8660 = 7.255.

✏️ Pen & paperThe Hopf value for another predator●●○

Keep e=1e = 1 but take m=1/3m = 1/3 in the Rosenzweig-MacArthur model. Find x∗x^*, the Hopf value of KK, and for K=4K = 4 the coexistence point, the trace, the determinant and the eigenvalues.

Show worked solution

x∗=m/(e−m)=1/2x^* = m/(e - m) = 1/2, and τ=x∗(K−1−2x∗)/(K(1+x∗))\tau = x^*(K - 1 - 2x^*)/(K(1 + x^*)) vanishes at K=1+2x∗=2K = 1 + 2x^* = 2. For K=4K = 4: y∗=(7/8)(3/2)=21/16y^* = (7/8)(3/2) = 21/16, τ=12⋅2/(4⋅32)=1/6\tau = \tfrac12\cdot 2/(4\cdot\tfrac32) = 1/6, the off-diagonal entries are −1/3-1/3 and (21/16)/(9/4)=7/12(21/16)/(9/4) = 7/12, so Δ=7/36=0.1944\Delta = 7/36 = 0.1944. The eigenvalues are 1/12±i27/12=0.0833±0.4330i1/12 \pm i\sqrt{27}/12 = 0.0833 \pm 0.4330i: an unstable spiral, surrounded by a limit cycle.

✏️ Pen & paperInterior point of competition●●●

Take K1=K2=1K_1 = K_2 = 1, r1=r2=1r_1 = r_2 = 1 and α12=0.8\alpha_{12} = 0.8, α21=0.9\alpha_{21} = 0.9. Show that the interior fixed point exists, find it, and classify it with the trace and the determinant. What does this say about coexistence when competition is almost as strong between species as within a species?

Show worked solution

The lines meet at D=1−0.72=0.28D = 1 - 0.72 = 0.28, x1∗=0.2/0.28=0.7143x_1^* = 0.2/0.28 = 0.7143, x2∗=0.1/0.28=0.3571x_2^* = 0.1/0.28 = 0.3571. With J=−(x1∗0.8x1∗0.9x2∗x2∗)J = -\begin{pmatrix}x_1^* & 0.8x_1^*\\ 0.9x_2^* & x_2^*\end{pmatrix}, τ=−1.0714\tau = -1.0714 and Δ=x1∗x2∗⋅0.28=0.0714\Delta = x_1^*x_2^*\cdot 0.28 = 0.0714, so τ2−4Δ=0.862>0\tau^2 - 4\Delta = 0.862 \gt 0: a stable node with eigenvalues −0.0714-0.0714 and −1.0000-1.0000. The slow rate 0.07140.0714 means a decay time of 1414: as α12α21→1\alpha_{12}\alpha_{21} \to 1 the determinant vanishes and coexistence becomes weak and slow.

What comes next

Coupled equations gave us a fragile conservation law, a robust Hopf bifurcation and a competition settled by a determinant. 15.3 Epidemic Models is another coupled nonlinear system. The SIR model has the same mass-action product S IS\,I, and dividing dI/dS=−1+γ/(βS)dI/dS = -1 + \gamma/(\beta S) and integrating gives an exact conserved quantity, I+S−(γ/β)ln⁡SI + S - (\gamma/\beta)\ln S, like HH here. But it does not oscillate: susceptibles are only used up, so the orbit runs to a final state on an axis. What replaces the Hopf bifurcation is a threshold: II first grows iff R0=βS0/γ>1R_0 = \beta S_0/\gamma \gt 1, a close relative of the invasion tests above.

Key Takeaways

  • A predator is a harvester whose effort is itself a variable: nullclines, fixed points and Jacobians describe the plane.
  • Lotka-Volterra conserves H=α(x−ln⁡x)+(y−ln⁡y)H = \alpha(x - \ln x) + (y - \ln y): closed orbits, period growing with amplitude (6.51176.5117 against 2π2\pi), time averages equal to the fixed point.
  • That conservation is structurally fragile: prey crowding ε\varepsilon gives trace −ε-\varepsilon, and forward Euler spirals outward (+4.65%+4.65\% in HH at h=0.01h = 0.01, RK4 −9×10−11-9\times10^{-11}).
  • Classify a planar fixed point by Δ<0\Delta \lt 0 (saddle) and by the signs of τ\tau and τ2−4Δ\tau^2 - 4\Delta; on the borderline the nonlinear terms decide.
  • Rosenzweig-MacArthur has trace 12−32K\tfrac12 - \tfrac{3}{2K}, a Hopf bifurcation at K=1+2x∗=3K = 1 + 2x^* = 3 and a supercritical limit cycle.
  • Enrichment destabilises: from K=5K = 5 to 1010 the prey minimum falls from 0.03570.0357 to 2.5×10−52.5\times10^{-5}.
  • Competition has four outcomes set by the invasion conditions; the interior point is stable iff α12α21<1\alpha_{12}\alpha_{21} \lt 1.
  • SIR shares the coupling and a conserved quantity but has a threshold R0R_0 and no cycles.
Finished this lesson?

Mark it complete to track your progress (saved in this browser). Reading every section does this for you.