The point-centered quarter method (PCQM) is often used in phytosociological surveys of tropical forests. The method has a basic assumption that individual trees in the forest have a completely random spatial pattern. In this study, the effect of deviation from total spatial randomness on PCQM estimate of forest density was analysed through Monte Carlo simulation of hypothetical forests with regular and clustered spatial patterns and with different densities. The influence of sample size was also analysed, but showed no marked effect on estimation biases. The relative bias on the tree density estimation varied from +70.3% (regular lattice spatial pattern) to -75.7% (strongly clustered spatial pattern). Tree density did not affected estimation bias, except for the totally randomized spatial pattern and randomized regular lattice pattern. The point-centered quarter method overestimates (positive bias) tree density for regular patterns and underestimates this parameter (negative bias) for clustered patterns. Previous knowledge of tree spatial pattern in a forest is necessary for correct implementation and interpretation of results in this method.
density; bias; spatial pattern; point-centered quarter method; tropical forest; Monte Carlo simulation
