As an example consider hedgerow intercropping in which an uneven slope results in non-parallel hedges (Fig. 2). Suitable strata might be rows next to a hedge, rows one row away from the hedge and remaining rows (Fig. 3).
Within each stratum a sample is taken. The sample consists of n sections of rows, each l metres long. Choice of l and n is discussed below. The n sections may be systematically or randomly chosen. If systematic sampling is used, care must be taken that biases are not introduced by, for example, only sampling rows on the up-hill side of hedges.