To accommodate these augmentations later in the trial requires contingency plots to be set down at the start, for large inter-seasonal variances experienced in many tropical areas prohibit useful contrasts between agroforestry treatments established in different seasons