with permanent plot designs, samples taken at the same sites over time are unlikely to be considered true replicates; the data must be analyzed as a repeated measures design.
further, if plots along transects are to be considered replicates, they need to be spaced randomly or distant enough so that adjacent plots are practically independent of one another.