C4.1.1 Populations as interacting groups of organisms of the same species living in an area C4.1.2 Estimation of population size by random sampling
Who counts as one population?
- A population is a group of organisms of one species living in one area, close enough to interact and, normally, to breed with each other.
- Reproductive isolation separates one population from another: frogs of the same species round a pond across a motorway could breed with the first group if they met, but the road keeps them apart.
- Counting every individual is almost never possible (there are too many, they hide, or counting would damage the habitat), so ecologists take a sample and scale it up to an estimate.
- The sample must be random: every individual, or every spot, must have the same chance of being picked.
- Sampling error is the difference between the estimate and the true size. It shrinks as the sample grows, but never quite vanishes.
The detail
- Frogs round one pond are a population.
- Ecologists want to know how big a population is, and whether it is growing or shrinking, to manage a fishery, protect a rare orchid or track a pest.
- Even a random sample gives an estimate that differs a little from the true size.
How we know
- Nature of science: random sampling trades certainty for possibility. A sample can never give the true size, only an estimate with an error.
- A random sample at least makes that error unbiased: as likely to be too high as too low.
- Scientists report the error with the estimate, so others can judge how far to trust it.
Watch out
- If you choose where to sample, you drift towards the spots that look interesting, and the estimate is biased.
A population is a group of organisms of one species living in one area, close enough to interact and, normally, to breed with each other. Frogs round one pond are a population. Frogs of the same species round a pond on the far side of a motorway are a different population: they could breed with the first group if they ever met, but the road keeps them reproductively isolated.
Ecologists want to know how big a population is, and whether it is growing or shrinking, to manage a fishery, protect a rare orchid or track a pest. Counting every individual is almost never possible: there are too many, they hide, or counting would damage the habitat. So they take a sample and scale it up to an estimate.
The sample must be random: every individual, or every spot, must have the same chance of being picked. If you choose where to sample, you drift towards the spots that look interesting, and the estimate is biased. Even a random sample gives an estimate that differs a little from the true size; that difference is the sampling error. It shrinks as the sample grows, but never quite vanishes.
- Population: one species, one area, interbreeding.
- Estimate from a random sample; counting all is rarely possible.
- Sampling error: estimate minus true size. Bigger samples, smaller error.
A student estimates the daisies on a field by placing quadrats wherever daisies look thickest. What will the estimate be?
Choosing the thickest patches biases the sample upwards: the mean per quadrat is higher than the true mean, so the estimate for the whole field is too high. Random coordinates remove the choice.