Sampling with a quadrat smaller than 30 cm, e.g. 15 cm (Fig. 1.5c), provides a
more accurate picture of the spatial variability of these organisms. Some of these
units will land near or around clusters, which will thus give very large numbers of
the animals. Others will land in spaces among clusters, which will thus give very
small numbers of the animals. The average of the replicates should be the same as
when large quadrats are used (when scaled to the same area), as long as each set of
quadrats representatively samples the population. The variance will, however, be
much larger because it will more accurately represent the true pattern of dispersion,
thus identifying the scale at which the ecological processes causing these patterns
are operating.