Fig. 5. On rare occasions a phenomenon termed `plume ventilation' was seen, which caused significant temporal and horizontal variations in the sediment oxygen distributions. Here the fish (arrow) made a wriggling body movement, which channelled oxygenated water down along the body creating a `pocket' of oxygenated sediment around the fish. The plume, which typically lasted 20–30 min (the time before anoxic conditions were re-established around the fish), penetrated into the interstice as the oxygenated water was replenished by oxygen-depleted water leaving the gills. The oxygen was gradually consumed by microbes and chemical oxidation processes during sediment percolation.