Fig. 1.13 (a) Sampling to handle uncertainty about the scale of an impact due to an outfall. It has
been proposed that any impact would only extend over a small distance (stippled), so samples are
taken from that area and from controls on either side, outside the area expected to be affected. The
area influenced may, however, be much larger (shaded), so other control samples are taken at much
greater distances from the outfall. (b) Sampling to test the hypothesis that assemblages in sediments
are influenced by the presence of a rocky reef (reefs are shaded). For two replicate reefs, n = 5 cores
are taken next to the reef (0 m) and at 10, 20 and 40 m from it. The reef will cause samples at 0–10 to
differ more than 10–20 or 0–20 to differ more than 20–40 (i.e. more than similar distances where there
are no reefs; for more details, see the text)