Biology › Practicals › Required practical 12: a belt transect and species distribution

Required practical 12: a belt transect and species distribution

Where a habitat changes along a line, from under trees into the open or up a shore, the species change with it. A belt transect places a quadrat at regular intervals along that line, records how much of a species is in each, and measures an abiotic factor at the same spot. Spearman's rank then says whether the two rise and fall together more than chance would give.

The capability this develops

This guide is one investigation under capability P12: Field sampling, distribution and behavioural response, which sets out the techniques the investigation belongs to, the calculations it carries and the errors that cost marks in all of them. The notes behind the biology are in Ecology, populations and environmental change.

Where this sits on the specifications

This library maps a qualification to its own specification one numbered heading at a time, against a document whose sha256 is recorded, and publishes the result as a route map. These are the headings whose practical work this investigation serves, on four of the five qualifications mapped so far. Each links to the route map the row belongs to.

Qualification	Heading
OCR H020/H420	6.3.1 Ecosystems
Cambridge International 9700	18.2 Biodiversity
Pearson Edexcel Biology A (Salters-Nuffield) 8BN0/9BN0	5.2 Biotic and abiotic control of numbers and distribution
Pearson Edexcel Biology B 8BI0/9BI0	10.1 iv The nature of ecosystems

Some boards number their practical activities, and those lists were read here from the same specifications as the headings above. This investigation is numbered on them as:

The other qualifications above are matched by specification heading only, because none of them numbers an activity that is this investigation. Use your awarding organisation’s current specification and its practical handbook as the authority on which activities your course requires and how they are assessed.

The principle

Random or systematic. For a site as a whole, or two sites compared, sample at random: lay two tapes at right angles as axes, generate coordinates with a random number generator, and place a quadrat at each pair. Along a gradient, such as from a wood into a field or up a shore, sample systematically. A line transect records the species touching a tape at regular intervals; a belt transect puts a quadrat at each interval, so it records abundance as well. Run at least three transects from random points along a baseline and take the mean at each distance.

Abundance in a quadrat is recorded as a count of individuals, where they can be told apart, or as percentage cover, the share of the quadrat's area the species covers, where they cannot. Dog's mercury and grasses spread in patches, so cover is the measure here.

The question a transect answers is about a relationship between two quantities measured at the same points. Spearman's rank tests whether they are correlated, without assuming the relationship is a straight line. A correlation, however strong, does not show that one causes the other.

Belt transect: A line across a habitat with a quadrat placed at regular intervals along it, recording the abundance of each species.
Percentage cover: The percentage of a quadrat's area covered by one species, seen from above.
Abiotic factor: A non-living condition of a habitat, such as light intensity, temperature, soil moisture or pH.

Apparatus

Two long tapes, one for the baseline and one for the transect
A 0.5 m × 0.5 m quadrat, gridded into 100 squares
A light meter, and a soil thermometer or temperature probe
Soil sample bags, a trowel, a balance and access to a drying oven, for soil moisture
A random number generator, for choosing where transects start
A record sheet, a clipboard and a pencil
Gloves for soil and leaf litter

Safety

Risk assessment. Work in groups within the agreed area, never alone, and check the weather and, on a shore, the tide times. Wear gloves for soil and leaf litter and wash your hands before eating. Drying ovens are hot. Trample as little as possible, put back turned stones as they lay, and return animals where they were found. Your school's risk assessment comes first; where it differs from this page, follow it.

Variables

Independent variable

Light intensity, measured at each quadrat with a light meter. It is not set by the investigator, so this is a correlation study rather than an experiment.

Dependent variable

The percentage cover of dog's mercury in each quadrat.

Control variables

The quadrat. One size, 0.5 m × 0.5 m, at every station.
The interval. A quadrat every 5 m along the transect.
How cover is judged. The gridded quadrat and one recorder, counting the squares more than half covered.
When light is read. All readings within a short time of each other, at the same height, sensor upwards.
Where light is read. At the quadrat, not once for the site.

Method

Laying out the transect

Run a baseline along the edge of the wood, and choose three starting points on it with a random number generator.
From each, lay the transect tape at right angles to the baseline, from under the canopy out into the clearing, for 35 m.
Place the quadrat on the same side of the tape at 0 m and at every 5 m, eight stations on each transect.

At each quadrat

Choose a size that holds several individuals but can still be recorded accurately, often 0.5 m × 0.5 m for short turf, and keep it throughout.
Record the percentage cover of dog's mercury as the number of the hundred squares more than half covered.
Take each reading at the quadrat, not once for the site.
Take all readings of a factor at the same time of day, since light and temperature change through the day.
Light: a light meter at a fixed height, sensor upwards, averaging three readings because cloud changes it.
Take the mean cover and the mean light at each distance across the three transects.

Random quadrats accumulating species until the count levels off — One quadrat recorded four species. Six quadrats had still missed three of the thirteen present. The flattening of the curve is the criterion for stopping, and it cannot be judged from a single sample.

Estimating a population, and comparing methods

For daisies on a playing field, the estimate is mean per quadrat ÷ quadrat area × field area, from random quadrats and again from quadrats placed by eye.
Quadrats placed by eye fall where daisies are easy to see, so that estimate is usually too high.
The dependent variable can be a feature of the organism, such as ivy leaf area in shade and in sun.
Choose plants at random and take the leaf from the same position on each, so leaf age is controlled.

Sample data

These readings are idealised. They are the eight stations the lesson works, each standing for the mean of three transects, and real readings scatter. Nothing here is a record of fieldwork.

Distance / m	Light / klx	Cover of dog's mercury / %	Rank of light	Rank of cover	d	d²
0	1.2	64	1	7	−6	36
5	2.0	70	2	8	−6	36
10	3.1	52	3	6	−3	9
15	4.5	40	4	5	−1	1
20	6.8	31	5	4	1	1
25	9.0	18	6	2	4	16
30	12.5	22	7	3	4	16
35	15.0	6	8	1	7	49

Null hypothesis: there is no significant correlation between light intensity and the cover of dog's mercury.

Both quantities are ranked from the lowest, 1, to the highest, 8, and d is the rank of light minus the rank of cover. The d² column adds to 164.

The statistic. r_s = 1 − (6 × 164) ÷ (8 × 63) = 1 − 1.95 = −0.95.

The critical value for n = 8 at p = 0.05 is 0.738, and 0.95, ignoring the sign, exceeds it, so the null hypothesis is rejected: there is a significant negative correlation between light intensity and cover.

Reading the graph

Plot cover on the vertical axis against light on the horizontal axis as a scatter graph, one point per station, and do not join the points. A second graph, a kite diagram or a bar chart of cover against distance along the transect, shows where along the line the change happens.

The pattern. Cover is highest under the canopy, 64% and 70% where light is 1.2 and 2.0 klx, and falls to 6% in the open at 15.0 klx. The one point out of order is at 25 m and 30 m, where cover is 18% and then 22%, which is why r_s is −0.95 and not −1.

What the test shows. A correlation this strong would arise by chance less than one time in twenty. It does not show that light controls where dog's mercury grows.

Moisture, temperature and competition from grasses also change from wood to clearing, so this does not show that light is the cause.

Testing the cause. Measure the other factors at the same quadrats, and look for a factor that changes along the transect without the others. Or move the question into an experiment: grow the plant at several shade levels with water and soil kept the same.

Evaluation

Each of these is a real source of error in this investigation, and the middle column is the part an answer earns marks for: what it does to the result, and in which direction.

Source of error	What it does to the result	What to do about it
Cover judged by eye	Recorders disagree about how much of a quadrat a plant covers, in either direction, so the cover values scatter between people.	A gridded quadrat and one recorder reduce that
Light read once, on one day	Light changes with cloud and the time of day, so a single reading may not represent the conditions the plants grow in.	Abiotic readings are a snapshot of one day, which a data logger left on site improves.
One transect	A transect run where the canopy happens to be thin, or where the plant happens to be absent, gives a pattern that belongs to that line only.	Run at least three transects from random points on the baseline and take the mean at each distance.
Quadrats too few, or too far apart	The point where cover changes fastest may fall between two stations, so the graph shows a smooth change where there is a sharp edge.	Plot the running mean to check enough quadrats were taken, and add stations at shorter intervals where cover changes most.
Trampling the transect	Plants flattened by the people recording them look like lower cover, by most at the stations visited most.	Walk beside the tape, not along it, and record each station once.

Questions

Each answer opens where you tap it.

Question 1 (2 marks)

Explain why a belt transect is used here rather than quadrats placed at random coordinates.

Check your answer

the investigation is about how cover changes along a gradient, from wood to clearing;

random quadrats might miss parts of the gradient, while a transect samples systematically at every distance along it.

Question 2 (3 marks)

Use the d² values in the table to show that Spearman's rank correlation coefficient is −0.95.

Check your answer

Σd² = 36 + 36 + 9 + 1 + 1 + 16 + 16 + 49 = 164;

r_s = 1 − (6 × 164) ÷ (8 × (8² − 1));

= 1 − 984 ÷ 504 = 1 − 1.95 = −0.95.

Question 3 (2 marks)

Explain why light is measured at each quadrat and not once for the whole site.

Check your answer

light varies along the transect, from under the canopy into the clearing;

the reading must belong to the quadrat whose cover it is compared with.

Question 4 (3 marks)

A student concludes that low light causes dog's mercury to grow. Evaluate this conclusion.

Check your answer

there is a significant negative correlation, so cover is higher where light is lower;

but a correlation does not show cause;

other factors such as soil moisture, temperature or competition from grasses also change along the transect and may be responsible.

Question 5 (2 marks)

A population of daisies was estimated from quadrats placed by eye and from quadrats at random coordinates. Explain why the first estimate is likely to be higher.

Check your answer

quadrats placed by eye tend to be put where daisies are easy to see, which is where they are most dense;

so the mean per quadrat is too high, and so is the estimate scaled up to the whole field.

Coverage maps are available for AQA 7401/7402, OCR Biology A H020/H420, Cambridge International 9700, Pearson Edexcel Biology A (Salters-Nuffield) 8BN0/9BN0 and Pearson Edexcel Biology B 8BI0/9BI0. Use the relevant route map to check what InkBiology covers, and your awarding organisation’s current specification as the final authority on examinable content.