Abstract

Spatial variation in the distribution of copper, lead and zinc in sediments in Botany Bay, Australia, is described. Sampling was done according to a nested, hierarchical design, and variation was detected at spatial scales from 10 m to several kilometres.

In studies where the design of sampling is inadequate, the existence of variation at small spatial scales will confound comparisons at larger scales. Adequate replication at different scales, necessary to avoid confounding, has rarely been done in published studies. Implications of 'patchiness' for environmental sampling and monitoring, and the means of overcoming associated problems, are discussed.