Pseudoreplication in experimental designs for the manipulation of seed germination treatments

In the published reports concerning the design of laboratory experiments examining signals that might trigger seed germination, there is currently a misunderstanding of what might constitute correct replication of the experimental treatment. This is particularly true for the study of dry heat, smoke and charcoal, where either an individual seed or a batch of seeds in a Petri dish or tray is being treated as the unit of replication of the experimental treatment, irrespective of whether or not those seeds were all subjected to the experimental manipulation simultaneously. Under these circumstances, the application of the treatment is unreplicated, while samples nested within that single application have been replicated, and the Petri dishes/trays are functioning solely as independent replicates of the variability in germination response within the seed batch used and variation within the pretreatment and post-treatment environments. Thus any observed difference in germination may be due to the germination treatment but, potentially, it could also be due to any chance event affecting the treated sample. There are a number of alternative experimental designs that avoid this problem. The essential point with these designs is that the application of the experimental manipulation to each replicate should be separated in space by the use of separate experimental equipment and/or in time by the repeated use of the same experimental apparatus.