the result will be lower precision of the estimate -- how much lower depends
on the variation within each stratum. Strata may be defined by:
(i) Using experience of similar systems in a similar environment.
(ii) Using a small pilot study. Detailed measurements (e.g., row-by-row
yields) could be taken in a few units from which strata are defined for
use in the rest of the experiment.
(iii) Using of some other variable that is easy to measure and is correlated
with crop yield. An example for maize is plant height. A quick visual
assessment of plant height will give a good indication of the number of
rows over which the interface is affecting plant growth.