Estimating and constructing confidence intervals for spatial patterns between random and regular

Abstract

A new spatial pattern index is proposed that spans the continuum of patterns from random to regular. The proposed index is the parameter of the distribution of distances from random points to the closest individual. Knowing the distribution makes it possible to derive estimators, perform hypothesis tests, and construct confidence intervals anywhere on the continuum. Although analytical solutions to the estimators and critical values cannot be derived, numerical procedures are discussed to estimate the solutions. A simulated example is given that tests the hypothesis that spatial patterns in loblolly pine ( Pinus taeda L.) forests converge to a common pattern over time, across all initial patterns.