Students from Axe Valley Community College carried out a stratified random
sampling exercise on strips 1 - 4 on the Heathland Restoration Project site at Offwell in March. The purpose was to investigate
whether there was any difference in the % cover of Bluebells in spring-cut strips, as opposed to autumn-cut
strips. The exercise also tested whether there was any difference between the % cover of
Bluebells in the top of the strips (which are more influenced by neighbouring woodland)
and the bottom of the strips, which are more open.
The survey was carried out early in the spring,
before the spring-cut strips were cut. At the time, the Bluebells were developing and were
not yet flowering. This meant that care had to be taken not to misidentify sedges and
woodrushes (which are also common in all the strips) as Bluebells. These can look quite
similar to Bluebells in the vegetative stages.
||Groups of students were
assigned to sample the different strips for Bluebells. They surveyed ten random 1m2 quadrats in
section 1. Ten random quadrats were also sampled in the top (western) half of each of
sections 2 - 4, as well as in the bottom half of each of these sections. (Strip 1 is much
smaller than the other strips and so was not divided into top and bottom halves. Diagram) A total of 70 quadrats
were therefore sampled.
Students sampling the bottom of strip 4 (left)
|The random quadrat locations were chosen using a random number table to select x and y
co-ordinates to walk. A line across the centre of each of the strips acted as the x axis.
For example, if the first random number selected was
15, then a student would walk 15 paces along the centre line, beginning at the boundary of
the relevant strip. If the next random number was 6, the student would then turn 900 and walk six paces up. At this point the quadrat would be laid
down and sampled. When students were sampling the bottom half of the strips they would
walk the required number of paces down 900
from the centre line, rather than up.
determining the percentage
cover of Bluebells in each quadrat. In each group, the same person estimated the %
cover each time to reduce the effect of observer bias.
Now try your skills at processing and analysing the
raw data they collected, by carrying out the exercise devised for the Axe Valley students
by Dr Wendy Graham.