One of the truisms of sampling design is that the design depends on the objectives. Too often objectives are not defined properly, with the result that the data collected cannot be used to answer the questions posed. A good example is that of a monitoring programme that aims to detect changes in an assemblage of benthic organisms caused by eutrophication but where the magnitude of change was not specified in the objectives, with the result that the monitoring programme was so loosely designed that insufficient samples were taken. A posteriori analyses of the results may show that the monitoring would take 10 years to detect a 10% change in the faunal composition. You may think that this is an unrealistic and hypothetical example, but our experience shows that far too often results such as this are the norm. We return to the types of monitoring in Chapter 11, but for now let us start with perhaps the simplest case: we wish to survey an area of coastal soft sediment simply to find out what is there (i.e. to map the habitats and prepare for a more detailed quantitative study of the benthic assemblages). Up to the last couple of decades, sampling subtidally below diveable depths was usually done blind. One had to resort to charts, perhaps prepared in the nineteenth century, which have depths and descriptions of sediments made from soundings done with handlines with candlewax in a hollowed-out part of the lead weight that touched sediment particles, enabling the sediment type to be crudely mapped. Since the 1980s huge technological advances have been made in mapping sediments. Two types of instrumentation have been developed: depth sounders of various types and remote-operated vehicles (ROVs). With sounders, accurate maps of the contours of the seabed can be produced and then indications of the hardness and roughness superimposed on the depth and good three-dimensional images produced with modern software. Sophisticated multibeam echosounders have been used to map the whole continental shelves of many countries. Now that the satellite-based differential global positioning system (DGPS) is generally available with an accuracy to a few metres, mapping of subtidal sediments has become much easier and more accurate.