= Solution
Assume the decreasing functions are continuously differentiable so their local gains are defined. In the positive quadrant, the <nullclines> are $x=f(y)$ and $y=g(x)$, both decreasing. The <phase plane> arrows point right when $x<f(y)$ and left when $x>f(y)$; they point up when $y<g(x)$ and down when $y>g(x)$. Both axes point into the quadrant. Since $f,g$ are bounded, sufficiently large $x$ or $y$ decreases, giving a bounded forward-invariant rectangle containing all equilibria and eventually containing each trajectory. Thus <mutual repression> can have one attracting intersection or several intersections separating different <basins of attraction>.
An equilibrium obeys $x_*=f(y_*)$, $y_*=g(x_*)$. Its <Jacobian matrix> and eigenvalues are
$$
J=\begin{pmatrix}-1&f'(y_*)\\g'(x_*)&-1\end{pmatrix},\qquad
\rho_\pm=-1\pm\sqrt{f'(y_*)g'(x_*)}.
$$
The product is nonnegative. Hence an intersection is a <stable equilibrium> with two negative real eigenvalues when $f'g'<1$, a <saddle equilibrium> when $f'g'>1$, and nonhyperbolic when $f'g'=1$. This is the <mutual repression stability criterion>. To see why an intervening saddle normally accompanies <multistability>, eliminate $y$ and set $F(x)=f(g(x))-x$. Bounded positivity gives $F(0)>0$ and $F(x)<0$ for large $x$, while
$$
F'(x_*)=f'(g(x_*))g'(x_*)-1.
$$
Two downward, stable crossings require an upward crossing between them. For simple roots that middle crossing has $f'g'>1$ and is a <saddle equilibrium>. Its stable manifold separates the attracting states. Thus \b[robust multistability requires the loop gain to exceed one at an intervening equilibrium], with the stable equilibria themselves having gain below one. A tangency at gain one is the threshold where equilibria can merge; degenerate crossings require higher-order analysis. A gain exceeding one somewhere is necessary for multiple intersections, but by itself does not prove that a particular pair of nullclines has multiple stable intersections.
Define the logarithmic repression sensitivities
$$
\epsilon_f=\frac{\partial\log f}{\partial\log y}=\frac{yf'(y)}{f(y)},\qquad
\epsilon_g=\frac{\partial\log g}{\partial\log x}=\frac{xg'(x)}{g(x)}.
$$
Both are negative. At equilibrium $f(y_*)=x_*$ and $g(x_*)=y_*$, so their product is exactly $f'(y_*)g'(x_*)$. The condition can therefore be expressed as
$$
\boxed{\epsilon_f\epsilon_g<1\ \text{at stable intersections},\qquad
\epsilon_f\epsilon_g>1\ \text{at an intervening saddle}.}
$$
Equivalently, the product of the magnitudes of the two negative logarithmic gains must exceed one at that <saddle equilibrium>. The divergence of the <vector field> is identically $-2$, so the <Bendixson-Dulac criterion> excludes periodic orbits in the positive quadrant. The relevant stable long-time states in this bounded <phase plane> are thus equilibria, rather than attracting limit cycles. The following sketches illustrate the single-attractor and two-attractor possibilities; the saddle in the second panel lies on the diagonal, with its stable diagonal separating the two basins.
\Image[/past-exam-of-the-mathematics-course-of-the-university-of-cambridge/2007/iii/paper-74-mutual-repression.png]
{title=Mutual repression phase planes: one stable equilibrium and two stable equilibria separated by a <saddle equilibrium>}
Back to article page