The Yellow Meadow Ant Lasius flavus: spatial analysis of its nest distribution
50 sampled on Yardley Hill in order to establish that these were working colonies. All proved to
have active nests at the time of sampling.
Investigation methods
A 20 x 20 metre quadrat was laid out using tapes over an area of each slope and the co-ordinates for
each anthill were recorded and plotted on graph paper. The number of anthills per sample quadrat
area (5m x5m) were also counted.
Students decided on the size of sample quadrats for each area to be set at 5m x 5m as this allowed
for individual anthills to fall within a quadrat square rather than straddling a number of smaller
quadrats (e.g. 1m x 1m), this being ascertained after some 'trial and error'.
Analysis of results
Statistical Method 1: Analysis of Variance
When in theory no interaction exists between individual organisms or with their environment then
we can say that a random distribution should result. Normally, however, interactions do occur and
some form of distribution pattern occurs.
Using data gathered in the field, the mean density of anthills per quadrat area (5m x 5m) and the
variance of the population sample are calculated (see Table 1).
Given that the Variance (V) =
X (x - xmean)2

n
Then comparing the mean (xmean) against the variance (v), a measure of pattern can be
discerned. The greater the ratio between x and v, the greater the degree of aggregation.
If xmean > v (indicates a regular distribution pattern)
If xmean < v (indicates an aggregated distribution pattern)
If xmean = v (indicates a random distribution pattern)
Results from the study based upon sample quadrats of 5m x 5m indicated a regular distribution
pattern (4.625>1.34, i.e. the mean is greater than variance) for Yardley Hill (south-facing), whilst
Daisey Plain (north-facing) demonstrated a pattern very close to a random distribution (1.40>1.31,
i.e. the variance is only marginally greater than the mean).
It is interesting to note that even when quadrat area sizes were increased fourfold (1 Om x 1 Om) and
summed horizontally, vertically or 'quartered', that a regular distribution pattern was still achieved
for Yardley Hill. However when this method was repeated for Daisey Plain, increasing the sample
area resulted in a movement away from randomness to one of greater regularity (see Table 2).
Similarly, reducing the sample quadrat to 2m x 2m caused a movement towards aggregation for
Yardley Hill (0.74