Experimental designs for mixed-species plantations

摘要

Although an appropriate design is available for mixtures of two species-replacement series of multiple densities-studies concerning competition among thre" or more tree species have not been as etegant. A fundamental problem in mixtures of more thantwo species is that there are many treatment combinations. As plots must be large if individual tree competition indices are used, studies of more than two species require more land and seedlings than is practical for a single study. A solution to this problem is to use a systematic or clinal design where species composition varies gradually across a large plot. For three species, the plot is triangular with each species dominating a vertex of the plot. With large numbers of species, no satisfactory design is available.