Investigating populationsAQA A-Level Biology: Revision notes

It is usually impossible to count every individual in an area, so ecologists sample it and scale up. A sample must be random and large enough, otherwise it is biased and unrepresentative.

Quadrats (a square frame of known area) suit non-motile or slow-moving organisms such as plants, limpets or snails, because they stay in place while counted.

To estimate a population:

1. Count the organisms in each quadrat, or estimate percentage cover.
2. Calculate the mean number per quadrat.
3. Divide by the quadrat area to give the density, then multiply by the total area.

Worked example: 130 plants in 20 quadrats of 0.25 m² gives a density of 130 ÷ 5 = 26 per m². In a field of 2000 m² the estimate is 26 × 2000 = 52 000.

To avoid bias, quadrat positions are chosen randomly. Lay out two tape measures at right angles to form a grid over the area, then use a random number generator to give pairs of coordinates and place a quadrat at each.

A larger number of quadrats gives a more reliable mean, but takes longer. Using the same size of quadrat each time keeps counts comparable.

Random sampling is used to compare the abundance of a species in two areas or to estimate the total number in one area.

Where conditions change gradually, for example up a shore, across a sand dune or away from a path, a belt transect shows how the distribution of species changes across an environmental gradient.

A tape measure is laid along the gradient and quadrats are placed at regular intervals along it. The abundance or percentage cover of each species is recorded at each position, together with the abiotic factors there. The data can then be compared to see whether a change in the factor matches a change in the species.

A transect uses systematic sampling, so it is not random, but it is the best way of showing the change along a gradient.

For motile animals, quadrats do not work because the animals move. Instead the mark-release-recapture method is used.

1. Capture a sample of animals, count them (n₁) and mark them in a way that does not harm them or make them more visible to predators.
2. Release them and allow time for them to mix with the population.
3. Capture a second sample (n₂) and count how many are marked (m).

Estimated population = (n₁ × n₂) ÷ m.

Worked example: 50 marked, then a second sample of 40 containing 8 marked gives (50 × 40) ÷ 8 = 250.

The method gives a valid estimate only if these assumptions hold:

• The marked animals have mixed randomly with the rest of the population.
• There has been no immigration, emigration, birth or death between the samples, so the population is unchanged.
• The mark is not harmful, does not make the animal more likely to be eaten or caught, and does not fade or rub off.
• The samples are large enough to be representative.

If marked animals are more likely to be eaten, fewer are recaptured, so m is too small and the population is overestimated. If marks rub off, m is again too small and the estimate is too large.

In this practical you investigate how a named abiotic factor, such as light, soil moisture or pH, affects the distribution of a named species.

• Choose a gradient and use a belt transect, or compare two areas with random quadrats.
• Use the same size of quadrat each time, and count or estimate percentage cover.
• Measure the abiotic factor at each quadrat position, for example with a light meter, moisture meter or pH probe.
• Control other abiotic factors and take repeats to calculate means.
• Analyse with a correlation test (for example Spearman's rank), comparing the coefficient with the critical value at p = 0.05.

Correlation does not prove causation: other factors may change along the gradient too. Follow safety and ecological care: avoid damaging the habitat and wash hands after soil contact.