Nobody counts every dandelion in a field. Ecologists count a small part of it and scale up. This page covers how to choose where to sample, how to use quadrats and transects, and how to estimate the size of a population of animals that will not stand still — plus the calculations that come with each.
📚 What you need to know
A population is everything you are interested in; a sample is the part of it you actually measure.
Random sampling uses random coordinates and avoids researcher bias.
Systematic sampling uses a regular pattern, and is used when conditions change across the area.
Quadrats are used for non-motile organisms; record abundance, percentage cover or percentage frequency.
Transects (line or belt) show how species change along an environmental gradient.
Capture–mark–release–recapture with the Lincoln index estimates the size of a mobile animal population.
Every method rests on assumptions, and you must be able to state them.
Populations and samples
Measuring the whole population is called a census. It gives accurate results and includes everything, but it is slow, expensive and produces a mountain of data. Sampling is quicker and cheaper, but it can be unrepresentative, especially if the sample is small or badly chosen.
Approach
Advantages
Disadvantages
Whole population (census)
Accurate; nothing is missed
Time-consuming, expensive, huge amount of data
Sample
Quicker, cheaper, easier to analyse
May be biased or unrepresentative; small samples are unreliable
The fix for an unreliable sample is almost always the same: take more samples and calculate a mean. Say that in any question asking how to improve reliability.
Random or systematic?
The two methods answer different questions, so choosing between them is a judgement, not a rule.
Random sampling protects against choosing the easy-looking patches. Systematic sampling is the right choice when you want to see change from one end of a site to the other.
🧩 Setting up a random sample
Lay two tape measures at right angles along the edges of the site to make a grid, for example 20 m by 20 m.
Generate random coordinates with a random number generator, for example 7 and 13.
Place the quadrat at that point and record what is inside it.
Repeat many times and calculate a mean per quadrat.
Scale up to the whole area using the total area and the quadrat area.
Transects: sampling along a gradient
When physical conditions change across a site — up a beach, up a hillside, away from a path — random sampling would hide the pattern. A transect follows the change instead.
Line transect: lay a tape in a straight line and record every organism that touches the line at set intervals, for example every 2 m.
Belt transect: place a quadrat at set intervals along the tape and record abundance or percentage cover in each one.
Pair the biological data with abiotic measurements at each point — light, soil pH, moisture, altitude — and you can suggest which factor is driving the change in species.
Quadrats: what to record
Measurement
How you record it
Best used when
Abundance (count)
Count every individual of the species inside the quadrat.
Individuals are easy to tell apart, such as daisies.
Percentage cover
Estimate what fraction of the quadrat area the species covers.
Individuals are impossible to separate, such as grass or moss.
Percentage frequency
Count how many of the small squares in the quadrat contain the species.
You want a quick, repeatable measure that is less subjective than cover.
Percentage frequency is more repeatable than percentage cover, because counting squares involves much less guesswork than judging an area by eye.
Counting animals that move
Quadrats are useless for beetles, fish or voles. For mobile animals, ecologists use capture–mark–release–recapture.
🧩 The method, step by step
Capture a large first sample and count it.
Mark each animal in a way that does not harm it or make it obvious to predators.
Release them and wait long enough for them to mix back in with the population.
Recapture a second large sample and count how many of them carry marks.
Calculate using the Lincoln index.
Lincoln index
population estimate = (M × N) ÷ R
Here M is the number marked and released in the first sample, N is the total caught in the second sample, and R is the number of marked animals found in that second sample.
Worked examples
WORKED EXAMPLE
Scaling up from quadrats
A student places eight 1 m² quadrats at random in a 400 m² meadow and counts the daisies in each: 5, 2, 0, 4, 3, 6, 1, 5. Estimate the total number of daisies in the meadow.
Step 1: total the counts5 + 2 + 0 + 4 + 3 + 6 + 1 + 5 = 26Step 2: mean per quadrat26 ÷ 8 = 3.25 daisies per m²Step 3: scale up to the whole meadow3.25 × 400 = 1300about 1300 daisieskeep the zero in the mean — dropping empty quadrats inflates the estimate
WORKED EXAMPLE
Using the Lincoln index
Ecologists catch 124 ground beetles in pitfall traps, mark them with a dot of non-toxic paint and release them. Two days later they catch 138 beetles, of which 31 are marked. Estimate the population size and state one assumption.
Step 1: write the formula and the valuesM = 124, N = 138, R = 31estimate = (124 × 138) ÷ 31Step 2: calculate= 17112 ÷ 31 = 552about 552 beetlesStep 3: an assumptionthe marked beetles mixed fully back into the population and the marks did not rub off
Assumptions and limitations
Capture–mark–release–recapture only works if all of these hold, and in the field they rarely all do:
Marked animals mix randomly back into the population before the second capture.
The mark does not change survival — it must not make the animal easier for predators to see.
The mark stays visible and does not wash or rub off.
The population does not change much between the two captures: no big birth, death or migration events.
How to use this in an answer. If marks rub off, R is too small, so the estimate comes out too large. Being able to say which way an error pushes the result is a high-level skill examiners reward.
💡 Exam tip
Show every step of a calculation. Method marks survive an arithmetic slip.
Round sensibly and give units: “about 552 beetles”, not “552.0”.
If asked to justify random sampling, the key words are avoids bias and representative.
If the site has a gradient, say systematic or transect — random would hide the pattern.
State assumptions as full sentences, not single words.
To improve reliability: more quadrats, larger samples, repeat on different days, use the same observer.
⚠ Common mix-up
Throwing the quadrat over your shoulder. That is not random sampling — use random coordinates.
Mixing up N and R in the Lincoln index. N is the whole second sample; R is only the marked part of it.
Using quadrats for animals. They only work for organisms that stay still.
Saying percentage cover gives population size. It gives abundance, which is not the same thing.
Forgetting to multiply by the area ratio. The mean per quadrat is not the answer on its own.
Claiming a bigger sample removes bias. It improves reliability; only random placement removes bias.
Up next: How Ecosystems Keep Functioning — from measuring populations to asking what keeps the whole system running year after year.
Want this explained one-to-one?
Book a free session with an experienced IB ESS tutor and get your trickiest topics made simple.
