Fieldwork and sampling techniquesIB MYP Biology: Revision notes

It is impossible to count every organism in a large habitat. Instead, ecologists take samples and use them to estimate the whole population.

For a good estimate the sample must be:

• random, so every part of the habitat has an equal chance of being sampled, which avoids bias
• large enough, so it is representative. A bigger sample (more quadrats) gives a more reliable estimate.

A quadrat is a square frame of known area (often 1 m²) used to sample plants or slow-moving animals.

1. Mark out the area with two tape measures at right angles.
2. Use a random number generator to pick coordinates.
3. Place the quadrat at each coordinate and count the organisms.
4. Calculate the mean per quadrat: total count ÷ number of quadrats.
5. Population estimate = mean per m² × total area (m²).

Worked example: a field of 1000 m² is sampled with 1 m² quadrats and counts of 4, 7, 5, 6 and 3 are recorded. Mean = 25 ÷ 5 = 5 per m². Estimate = 5 × 1000 = 5000.

Some plants, such as grass or moss, cannot be counted as individuals. Instead, we estimate percentage cover: the proportion of the ground in the quadrat that the species covers.

With a grid quadrat of 100 squares, count the squares that are mostly covered. If 42 squares are covered, the percentage cover is 42%.

Percentage cover is quick, but estimating by eye is subjective, so use the same person each time.

A transect is a line across a habitat along which samples are taken. It is used to study how the distribution of organisms changes as an environmental factor changes, e.g. from the seashore into a sand dune, or from a path into a woodland.

1. Lay a tape measure in a straight line across the habitat.
2. Place a quadrat at regular intervals (e.g. every 5 m) and record the species or percentage cover.
3. Compare the results at each distance.

Use random quadrats to estimate a population, but a transect to show a change along a gradient.

This method is used for animals that move, e.g. beetles or fish.

1. Catch a first sample, mark each animal harmlessly and release it.
2. Wait for the animals to mix with the population.
3. Catch a second sample and count how many are marked.

Lincoln index: population = (first sample × second sample) ÷ number marked in the second sample

Worked example: 40 marked; second sample of 36 contains 9 marked. Population = 40 × 36 ÷ 9 = 160.

Assumptions: marks are harmless and do not fade, marked animals mix evenly, and there is little birth, death or movement between samples.

• Too few quadrats: results are not representative. Fix: use more and calculate a mean.
• Non-random placing: choosing good-looking areas causes bias. Fix: use random coordinates.
• Misidentifying species or miscounting: use a key and the same person.
• Subjective estimates of cover: count the squares of a grid.
• Mark-release-recapture: marks that harm animals or fade, animals avoiding traps, or populations changing between samples.

To be reliable, repeat the sampling and compare the means.