Ecological Sampling and Data Analysis

A quadrat is a frame enclosing a known area. It is suitable for plants and slow-moving animals because they are unlikely to move into or out of the sample during counting. For a clearly identifiable plant such as a daisy, count the individuals inside the frame. When individuals are difficult to distinguish, as with grass or moss, estimate percentage cover: the percentage of the quadrat occupied by that species. A subdivided quadrat helps you judge the area covered. Percentage cover measures abundance but does not directly give a number of individual plants.

Estimating a population using random quadrats

The first part of AQA required practical activity 7 uses sampling to estimate the population size of a common species in a habitat. A suitable investigation is to count daisies in a grassed area.

You need a quadrat of known dimensions, tape measures or marked strings, a way to choose random coordinates, and a recording sheet on a clipboard. Quadrats measuring 25 cm × 25 cm or 50 cm × 50 cm are suitable examples. Identify the target species before starting so that everyone counts the same organism.

1. Measure the survey area. For example, place two 20 m tape measures at right angles to mark two sides of a 20 m × 20 m square.
2. Select a pair of random coordinates using a random number generator or numbered slips drawn from two bags. If using bags, replace the slips after each selection.
3. Use the coordinates to locate the sampling position. Place the same reference corner of the quadrat at each selected position, keeping the whole frame within the survey area.
4. Count and record the daisies inside the quadrat. Use a consistent rule for plants on the boundary.
5. Repeat at many randomly selected positions, recording every count, including zeros.

Random selection reduces bias: the person sampling does not deliberately choose patches with many daisies or places that are easiest to reach. Taking many samples helps capture differences between patches.

Scaling a sample up to the whole habitat

First calculate the mean number of individuals per quadrat:

$$\text{Mean per quadrat}=\frac{\text{total number counted}}{\text{number of quadrats}}$$

Then scale up according to how many quadrat-sized areas fit into the habitat:

$$\text{Estimated population}=\text{mean per quadrat}\times\frac{\text{total habitat area}}{\text{area of one quadrat}}$$

Equivalently, multiply the total number counted by the total habitat area divided by the total area sampled.

For example, suppose 50 daisies are counted across ten 25 cm × 25 cm quadrat samples in a 20 m × 20 m survey area. Each quadrat has an area of $0.25\times0.25=0.0625\text{ m}^2$, so the ten samples cover $0.625\text{ m}^2$. The habitat area is $400{\text{ m}}^{}$.

$$\text{Estimated population}=\frac{400}{0.625}\times50=32\,000\text{ daisies}$$

This estimate assumes that the sampled density is representative of the whole survey area. It does not mean that exactly 32,000 daisies have been counted.

Using a transect to investigate an environmental factor

Random sampling is useful for estimating abundance across an area. To investigate how distribution changes across a habitat, use a transect: a line along which samples are taken. A line transect can record species touching the line; placing quadrats at regular intervals also records abundance in a known area beside it.

The second part of required practical activity 7 investigates the effect of a factor on the distribution of the species. For daisies, light intensity can be investigated along a transect running from beneath a tree towards an open area.

Lay a tape measure in a straight line across this change in habitat. Place the same-sized quadrat at regular distances along it. At each position, record the distance along the transect, the number of daisies and the light intensity measured with a light meter. Keep the counting method consistent, and take light readings in the same way at each quadrat. Alternatively, percentage grass cover can be recorded instead of counting daisies.

Regular sampling reveals how abundance changes along the habitat rather than estimating an overall population. Measure the factor itself: distance from the tree is not a direct measurement of light intensity. Repeating the investigation along additional transects helps establish whether the pattern is consistent.

If daisies are more abundant where light intensity is higher, that suggests an association between light and their distribution. Light is needed for photosynthesis, so it can affect plant growth. However, other factors, such as soil moisture or competition, may also vary along the transect. A field association alone does not prove that light caused the pattern.

Summarising counts with mean, median and mode

The arithmetic mean uses every count: add the counts and divide by the number of samples. It is useful for estimating population size, even when it is not a whole number.

Consider these teaching data for buttercup counts in ten open-field quadrats:

$5,\ 5,\ 7,\ 1,\ 2,\ 5,\ 3,\ 4,\ 7,\ 6$

The total is 45, so the mean is $45\div10=4.5$ buttercups per quadrat. This is an average, not a claim that a quadrat contains half a buttercup.

For the median, arrange the counts in order:

$1,\ 2,\ 3,\ 4,\ 5,\ 5,\ 5,\ 6,\ 7,\ 7$

There are ten values, so the middle positions are fifth and sixth. Their mean is $(5+5)\div2=5$, giving a median of 5. With an odd number of values, there is one middle value.

The mode is 5 because it occurs three times, more often than any other value. The three measures describe the same data in different ways; they need not be equal.

Choosing and interpreting a graph

Choose a graph that matches the data. A bar chart can compare mean abundance between named habitats, such as woodland and open field. A scatter graph can show the relationship between abundance and a numerical environmental factor, such as light intensity or altitude.

Put the environmental factor on the horizontal axis and abundance on the vertical axis. Label both axes, including units where applicable. Choose evenly spaced scales that include the full range of values and spread the points across a useful amount of the graph.

Data for Plotting abundance against an environmental factor

This dataset has altitudes from 0 to 70 m and buttercup counts from 0 to 84 per quadrat. Suitable scales would run from 0 to 70 m in 10 m intervals horizontally and from 0 to 90 buttercups in intervals of 10 vertically. The point $(30,45)$ means that the quadrat at an altitude of 30 m contained 45 buttercups.

The points show a general decrease in abundance as altitude increases. If asked for a line of best fit, draw a line or curve following the overall trend rather than joining every point. Do not automatically join the points on a scatter graph. Describe the overall relationship while recognising variation in the individual results.

Judging the quality of a sampling investigation

A small number of quadrats may miss dense patches or bare areas. More randomly located samples generally give a more representative population estimate. Consistent species identification, quadrat area and counting rules make results comparable.

A transect examines only one strip of a habitat, so repeating it in different positions strengthens the evidence. Measuring the environmental factor at each quadrat is more informative than assuming it changes steadily with distance. Other changing factors remain a limitation when explaining the relationship.

Carry out fieldwork with minimal disturbance: avoid trampling the sampling area unnecessarily and do not uproot plants to count them.

• Abundance: how many individuals occur; distribution: where they occur.
• Random quadrats: estimate abundance across a habitat. Use random coordinates, count the target species and repeat many times.
• Transect: investigate change across a habitat. Sample at regular intervals and measure the environmental factor at each position.
• Count individuals where possible; use percentage cover for plants that are difficult to count separately.

Essential calculations

$$\text{Mean}=\frac{\text{sum of counts}}{\text{number of samples}}$$