LAUSR

ABSTRACT Sustainable management of plant diversity in a forest requires adequate information that is often derived from samples. With limited resources, the plot design has to be efficient. Most studies focused on evaluating effects of plot designs on sample estimates of species richness at landscape scale. This study aims at filling a knowledge gap by investigating how plot size and shape affect sample estimates of species composition at a local scale. Two census datasets with distinct forest ecosystems were used. Fifteen combinations of plot size and shape were simulated. Species compositional similarity between two combinations of plot design at a sample point was estimated by Jaccard and Sørensen indices, and their abundance-adjusted counterparts. Similarity in species composition decreased with increasing difference between two plot sizes. Plots with varying shapes were found to have different local species composition but could be similar in the number of observed species, which has not been explicitly reported elsewhere. For less species-rich forests, a 0.025–0.05 ha circular plot is recommended. For species-rich forests, a 0.05–0.1 ha rectangular plot with aspect ratio of at least 1:20 is recommended. The results should be of interest to stakeholders managing small area forests for conservation of plant diversity.