Calculus › Section G · Differential equations › Topic G.4

The logistic equation and population models

Exponential growth is a runaway train: it never slows down. Real populations hit limits set by food, space and rivals. The logistic equation adds that brake, and together with predator-prey and epidemic models it forms the basic toolkit for modelling living populations.

Why exponential growth cannot last

The equation \(\frac{dP}{dt} = kP\) predicts that a population multiplies forever. For a few hours in a fresh flask of nutrients, bacteria really do grow that way. But no environment keeps giving: food runs out, space fills up, waste builds up, and the growth rate must eventually fall. A realistic model should behave exponentially while the population is small and then level off as it approaches the largest population the environment can sustain. The logistic equation is the simplest modification with exactly that behaviour, and it describes many laboratory cultures of microorganisms reasonably well.

The logistic equation

The model multiplies exponential growth by a braking factor:

\frac{dP}{dt} = rP\left(1 - \frac{P}{K}\right)

Here \(r\) is the intrinsic growth rate, the relative (per-individual) rate \(k\) the population would have with no limits, and \(K\) is the carrying capacity, the largest population the environment can sustain in the long run. When \(P\) is small compared with \(K\), the factor \(1 - \frac{P}{K}\) is close to 1 and growth is nearly exponential. As \(P\) approaches \(K\), the factor approaches 0 and growth grinds to a halt. If a population somehow exceeds \(K\), the factor is negative and the population shrinks back. The model assumes \(r\) and \(K\) stay constant and the population is closed.

Equilibria: where growth stops

A constant (equilibrium) solution requires \(\frac{dP}{dt} = 0\), so \(P = 0\) and \(P = K\) are the two equilibria. Their stability differs. For \(0 < P < K\) the right side is positive, so the population rises; for \(P > K\) it is negative, so the population falls. Solutions starting near \(K\), on either side, therefore move toward \(K\): it is a stable equilibrium. But a small positive population grows away from 0, so \(P = 0\) is unstable.

Where growth is fastest

How fast is the population growing at size \(P\)? The rate \(g(P) = rP(1 - P/K)\) is a downward-opening parabola in \(P\), with zeros at \(P = 0\) and \(P = K\). Its vertex is halfway between, at \(P = \frac{K}{2}\), where the growth rate equals \(\frac{rK}{4}\). So a logistic population grows fastest not at the start and not near capacity, but at exactly half the carrying capacity. On the graph of \(P\) against \(t\), this is the inflection point of the S-shaped curve: the place where the curve stops bending upward and starts levelling off.

Worked example: solving the logistic equation

The logistic equation is separable. Rewrite it as \(\frac{dP}{dt} = \frac{rP(K - P)}{K}\) and separate: \(\frac{K\,dP}{P(K - P)} = r\,dt\). The fraction on the left splits into two simple pieces, as you can check by recombining them:

\frac{K}{P(K - P)} = \frac{1}{P} + \frac{1}{K - P}

Integrating both sides gives \(\ln|P| - \ln|K - P| = rt + C\), so \(\ln\frac{P}{K - P} = rt + C\) for a population between 0 and \(K\). Exponentiate: \(\frac{P}{K - P} = Be^{rt}\). Solving for \(P\) and writing the constant in terms of \(P_0\) produces the standard solution

P(t) = \frac{K}{1 + Ae^{-rt}}, \qquad A = \frac{K - P_0}{P_0}

As \(t \to \infty\), \(e^{-rt} \to 0\) and \(P \to K\), confirming the levelling off.

Worked example: a stocked lake

A lake can sustain 800 fish, and it is stocked with 200, so \(K = 800\), \(P_0 = 200\) and \(r = 0.6\) per year. The solution formula gives \(A = \frac{K - P_0}{P_0} = \frac{600}{200} = 3\), so

P(t) = \frac{800}{1 + 3e^{-0.6t}}

After 2 years, \(P(2) = \frac{800}{1 + 3e^{-1.2}} \approx \frac{800}{1.904} \approx 420\) fish. It passes 400, the point of fastest growth, when \(1 + 3e^{-0.6t} = 2\), that is at \(t = \frac{\ln 3}{0.6} \approx 1.8\) years, where it is briefly adding \(\frac{rK}{4} = 120\) fish per year. After that the pace slackens as \(P\) coasts toward 800.

Harvesting and sustainable yield

Suppose fish are caught at a constant \(H\) per year. The model becomes \(\frac{dP}{dt} = rP(1 - \frac{P}{K}) - H\). Harvesting pushes the stable equilibrium below \(K\), but only up to a point: the natural growth rate never exceeds \(\frac{rK}{4}\), so a larger harvest has nothing to feed on. If \(H > \frac{rK}{4}\), the equation has no positive equilibrium at all and the population falls to zero no matter where it starts. For the lake above, \(\frac{rK}{4} = 120\) fish per year is the maximum sustainable harvest. This calculation is the mathematics behind fishing quotas: take less than the peak growth rate and the stock survives; take more and it collapses.

Two species: the predator-prey model

A single equation cannot describe two species that eat each other's numbers. The Lotka-Volterra model couples two equations, one per species, for prey \(x\) (say rabbits) and predators \(y\) (say foxes):

\frac{dx}{dt} = ax - bxy, \qquad \frac{dy}{dt} = -cy + dxy

Prey alone would grow exponentially at rate \(a\), but encounters with predators remove \(bxy\) of them. Predators alone would starve at rate \(c\), but the same encounters, counted as \(dxy\), feed them. The \(xy\) terms make the system genuinely two-way: each population's fate depends on the other's size. As a concrete example take \(x' = 2x - 0.5xy\) and \(y' = -3y + xy\). Competition models are built the same way for two species sharing one resource; depending on the constants, one species wins or both coexist.

Why predator and prey cycle

Set both right sides to zero to find the equilibrium: \(x(2 - 0.5y) = 0\) gives \(y = 4\), and \(y(x - 3) = 0\) gives \(x = 3\), so the equilibrium is \((3, 4)\). In general it sits at \(\left(\frac{c}{d}, \frac{a}{b}\right)\). Away from it, solutions loop: abundant rabbits feed a growing fox population; many foxes drive rabbits down; hungry foxes then decline; relaxed pressure lets rabbits rebound. Plotting \(y\) against \(x\) traces closed orbits around the equilibrium, and plotting either population against time shows out-of-phase oscillations. Fur-trading records of Canadian lynx and hares show roughly ten-year cycles, a famous real-world echo of the model.

Epidemics: the SIR model

Population thinking also models disease. The SIR model sorts people into susceptible \(S\), infected \(I\) and recovered \(R\), with

S' = -\beta SI, \qquad I' = \beta SI - \gamma I, \qquad R' = \gamma I

Infection moves people from \(S\) to \(I\) at a rate proportional to encounters \(\beta SI\); recovery moves them from \(I\) to \(R\) at rate \(\gamma I\). The key number is the basic reproduction number \(R_0 = \frac{\beta S_0}{\gamma}\): at the start, each infected person passes the disease to \(R_0\) others on average. If \(R_0 > 1\) the outbreak grows; if \(R_0 < 1\) it dies out. This is why vaccination targets push \(S_0\) down far enough to bring \(R_0\) below 1.

Common mistakes to avoid

The most common error is placing the fastest growth at the start or at the carrying capacity; it happens at \(\frac{K}{2}\), where the growth-rate parabola peaks. Second, swapping the stabilities: \(P = K\) is stable and \(P = 0\) is unstable, not the other way round. Third, mislabelling the predator-prey terms: predators decline alone (the \(-cy\)) and grow through encounters (the \(dxy\)), so the minus sign belongs to the death term. Fourth, treating \(r\), \(K\) and encounter rates as exact constants; they are estimates, and the model is only as trustworthy as they are.

Key takeaways

The logistic equation \(\frac{dP}{dt} = rP(1 - \frac{P}{K})\) grows exponentially when small and levels off at the carrying capacity \(K\), a stable equilibrium, while \(P = 0\) is unstable. Its solution is \(P = \frac{K}{1 + Ae^{-rt}}\) with \(A = \frac{K - P_0}{P_0}\), an S-curve whose inflection is at \(\frac{K}{2}\), where growth peaks at \(\frac{rK}{4}\). Harvesting above \(\frac{rK}{4}\) collapses the stock. Predator-prey systems cycle around \(\left(\frac{c}{d}, \frac{a}{b}\right)\), and the SIR epidemic model grows exactly when \(R_0 = \frac{\beta S_0}{\gamma}\) exceeds 1.

Flashcards

Flashcard

Carrying capacity

K: the largest population the environment can sustain

The stable equilibrium of the logistic model.

Test your understanding

Question 1 of 5

For dP/dt = 0.4P(1 − P/600), at what population is growth fastest?

Question 2 of 5

A lake has carrying capacity K = 800 and is stocked with P₀ = 200 fish. In the solution P = K/(1 + Ae^(−rt)), what is A?

Question 3 of 5

A fishery harvests a logistic stock with r = 0.6 per year and K = 800. Above what yearly harvest does the population collapse?

Question 4 of 5

In the system x′ = 2x − 0.5xy and y′ = −3y + xy, what is the equilibrium (x, y)?

Question 5 of 5

In the SIR model, when does an outbreak grow from the start?

Standards: 7 (BC)

← Previous topicNext topic →