4.2.1(b)(i)-(ii) - Sampling biodiversity in habitats

Why Sampling Matters

A habitat is a place where organisms live, such as a meadow, woodland edge, rocky shore or stream bank. Biodiversity in a habitat depends on the range of organisms present and how they are distributed, but a habitat is usually too large, uneven and changeable for every organism to be counted.

Sampling solves that problem by studying a smaller part of the habitat and using it as evidence for the wider area. A good sample is not just a small sample; it is a sample collected in a way that represents the habitat fairly.

Sampling matters because organisms are rarely spread evenly. Some plants grow only in damp hollows, some invertebrates stay near shrubs, and some organisms are active only at particular times. If only the easiest patch is sampled, the result may say more about the student's choice of patch than about the habitat.

Good sampling therefore aims to:

This lesson focuses on collecting samples. Calculating species richness, species evenness and Simpson's Index comes later; those calculations depend on the sampling being sound in the first place.

Random Quadrat Sampling

Random sampling means selecting sampling positions by chance. In fieldwork, this is often done by laying a grid over the survey area and using a random number generator or random number table to choose coordinates. The key idea is that the student does not choose positions by eye.

Quadrats are square frames used to sample organisms in a known area. They are most suitable for plants, lichens, algae on a surface, and slow-moving or sessile organisms. A quadrat can be used to record the number of individuals, the presence or absence of a species, frequency across quadrats, or percentage cover when individual organisms are hard to separate.

[DIAGRAM: sampling_strategy_equipment_map: Lesson 080: Sampling biodiversity in habitats - diagram 01; asset_slug: 080_m04_2_1_sampling_biodiversity_in_habitats__diagram_01; recommended_method: drawn_biology; description: A clean line diagram showing a habitat survey area with random quadrat coordinates, a systematic belt transect through an environmental gradient, stratified regions, and simple line icons for a quadrat, sweep net, pitfall trap and pooter.]

A clear random quadrat method has a sequence:

1. Mark out the habitat area with measuring tapes or a grid.
2. Generate random coordinates.
3. Place the same size quadrat at each coordinate using a consistent rule, such as placing the lower left corner on the coordinate.
4. Identify and record all target organisms in each quadrat.
5. Repeat with enough quadrats to make the sample more representative.
6. Calculate a summary such as mean number per quadrat or mean density.

Random does not mean careless. The quadrat size should suit the organism, the same counting rule should be used each time, and uncertain identifications should be handled consistently, for example by using an identification key.

For plants crossing a frame edge, use a pre-agreed rule, such as count individuals rooted inside and include only two designated boundary edges. For overlapping grasses, percentage cover may be more repeatable than pretending each shoot is a separate plant. Frequency is $100\times\text{quadrats containing a species}/\text{quadrats sampled}$; it measures occurrence, not how many individuals occupy each quadrat. If a species occurs in 6 of 20 quadrats, frequency is 30%.

A mean density of $15.2\,\mathrm{m^{-2}}$ across a representative $500\,\mathrm{m^2}$ habitat gives an estimated total of $7600=7.60\times10^3$ plants. This extrapolation assumes the sample density represents the unsampled area.

If a grid contains 100 equally eligible quadrat positions, a single random draw gives each position probability 1/100. Random selection reduces deliberate choice bias; it does not guarantee that a small sample will contain every rare species.

Non-random Sampling Designs

Non-random sampling means the positions are not chosen entirely by chance. That sounds less fair, but it can be useful when the habitat itself has a pattern that random sampling might miss.

Stratified sampling is often strongest when each stratum is sampled randomly. For example, if 40% of the habitat is short grass and 60% is shrub edge, a student could take 40% of their samples randomly from short grass and 60% randomly from shrub edge. This helps avoid missing organisms that occur only in one part of the habitat.

Systematic sampling is different. A line transect is placed across the habitat and samples are taken at fixed distances. If quadrats are placed along the transect, this becomes a belt transect. This is useful when the question is about change across space, such as how plant abundance changes from open grassland into woodland.