Estimating the spatial distribution in forest stands by counting small angles between nearest neighbours

摘要

Methods for describing the structure of a tree population have been used for a considerable time as an integral part of forest research. In this paper we investigate the properties of a simple measure to describe the spatial pattern of tree positions,based on the following idea: The three angles between the three nearest trees from a given reference point are each classified as being either "large" or "small", where small is defined as being smaller than some predetermined critical angle. The measure is simply the proportion of small angles out of three. Practical tests have shown that the method is easy to apply and is thus potentially cost-effective. We investigate the power of a test that is based on this measure to test the hypothesis of complete spatial randomness. We show that its performance is comparable to alternative methods that are more costly to implement. It is also shown that the expected value and standard error of the measure can be estimated unbiasedly.