EART60061 Chap.8 Exponential and Logistic Growth Models
Exponential and Logistic Growth Models
Exponential and Logistic Growth develops one of the simplest models with an analytical solution. If the growth rate depends only on population size, growth is exponential, which cannot last because resources run out. Multiplying by the moderating function (K minus N) divided by K keeps early growth exponential and stops it at the carrying capacity K.
Integrating the logistic equation gives the population at any time, which is what a client wants. The starting value and K are read from data, and r is the gradient of the straight early part of a plot of ln N against time. In the rainforest regrowth exercise, calibrating r and K validates the model at one site but not at the cultivated site, where linear early growth points back to the assumption of unlimited resources.
The same exponential mathematics reappears in later chapters, where Fick's law and donor-controlled fluxes make a quantity change in proportion to itself. Throughout, the chapter separates reading parameters from data, which is approximate, from adjusting them during calibration, which is a deliberate modelling step.
What this chapter covers
- 01
Exponential growth and its limits
- 02
The logistic moderating function
- 03
Carrying capacity
- 04
Estimating r from a semi-log plot
- 05
Calibrating r and K
- 06
Validation at two regrowth sites
- 07
Exponential change as a recurring model
Worked example · free
Fit and use a logistic model for seedling regrowth
- 1Read N0 = 8 and K = 400 from the data.
- 2r = (ln 20 - ln 8) / 4 = 0.916 / 4 = 0.229 per month.
- 2Logistic: N = 400 times 8 / (8 + 392 times e to the power -2.29) = 3,200 / 47.7 = 67 seedlings.
- 1Exponential: 8 times e to the power 2.29 = 79 seedlings, larger because nothing limits it.
Key terms
- Specific growth rate
- The constant r linking the rate of growth to population size in exponential and logistic models.
- Logistic function
- The factor (K minus N) divided by K, which equals one for small populations and zero at the carrying capacity.
- Exponential growth
- Growth in which the rate of increase is proportional to the current size, so the quantity multiplies by a constant factor each period.
- Semi-log plot
- A graph of the logarithm of y against untransformed x, on which exponential growth appears as a straight line.
Exponential and Logistic Growth Models FAQ
Why is exponential growth inevitably unrealistic?
If the growth rate depends only on population size, the population multiplies by the same factor in each period without limit. Resources eventually become limiting, so unchecked exponential growth can last only briefly, as the doubling grains of rice on a chessboard illustrate.
How is r found from regrowth data?
While the population is far below the carrying capacity the logistic term is close to one, so growth is nearly exponential. A plot of the natural logarithm of N against time is then straight, and its gradient over that early period is r.
What does it mean when the model cannot be validated at a site?
If calibrating r and K still leaves a poor match, the structure of the model is wrong for that site. Returning to the assumptions, the linear early growth at the cultivated site suggests resources were limited from the start, plausibly by soil condition after cultivation.
Why is the logistic equation an analytical solution?
The logistic differential equation can be integrated exactly, giving the population at any time directly from N0, K and r. That is unusual for environmental models; more realistic ones cannot be integrated and must be solved numerically in time steps, which then need verification.
How do I choose the window for estimating r?
Use only the early part of the record, where the plot of the natural logarithm of N against time is close to a straight line. Different windows give slightly different gradients, which is why adjusting r afterwards during calibration is justified rather than arbitrary.
Exam move
Derive the logistic differential equation yourself by asking what factor must equal one for small N and zero at K. Practise estimating r from two early points and from a gradient, then predict with the integrated equation. Rebuild the regrowth spreadsheet with parameters in separate cells and practise changing r or K and reporting the new prediction quickly.
Then compare the model with real data by plotting both on one graph, and write the calibration and validation outcome in a single sentence for each site.