‘Field error’ related to marine sediment contamination studies

Investigators studying changes of heavy metals in marine sediments generally assumed that every future sampling at the same station would be very close to the initial sampling location. Thus, any observed variations can be assumed to be due to either seasonal changes or the changes in heavy metal input owing to nature or man's activity. However, it is obvious that it is impossible to carry out the subsequent samplings at the exact same initial point. Therefore, as a minimum, some statistical criteria should be established before generally assuming that a sample obtained at one point would be statistically similar to a sample collected, say, 10–50 m away in any direction. An assessment of the spatial distribution of cadmium, copper, iron, lead, zinc, total organic and carbonate contents at two stations in Haifa Bay (Israel), one adjacent to and the other far from a known input source of metals, are presented here. Inherent errors reflecting the local natural distribution of any particular studied parameter have been estimated with a ‘field error’ introduced into the calculation. A model of a sampling station based upon a boundary called the ‘statistically limiting radius’, r stn, is proposed.