Editorial: Inference on biological populations

Designing an unreplicated field trial essentially involves firstly selecting the plots for the check varieties, and secondly arranging the check varieties among these plots. Selecting the check plots appears to be very similar to choosing sites for a monitoring network, or choosing sites in a region at which to take a sample. The problems appear to be even closer if spatial dependence is postulated, when another aim in choosing the sites is to allow efficient estimation of the dependence. In this paper, the designs of monitoring networks and spatial samples, and some related design problems, are considered to see if they have implications for the design of unreplicated field trials. Copyright © 2001 John Wiley & Sons, Ltd.