Efficiency of edge effect correction methods for Ripley's K-function in spatial analysis of Christ's thorn jujube trees (Ziziphus spina-christi) in Zagros

Correction of edge effect in Ripley's K-function is important for obtaining unbiased results in spatial pattern analysis of trees. This research aimed at studying the capability of Ripley's K-function corrected by three major edge effect correction methods for square plots (guard, toroidal and weighted) to identify the spatial patterns of Christ's thorn jujube (Ziziphus spina-christi) in Fars Province. A true and two simulated 200 × 200 m2 plots with different spatial distributions of trees (aggregated, clustered, and random) were selected to investigate the analytical power of corrected Ripley's K-function. When computed with no correction, the results showed that Ripley's K-function was biased and the bias increased with increasing distance. The theoretical values departed the simulation envelopes in the true (aggregated) and two simulated (random and dispersed) plots of Christ's thorn jujube trees illustrating that edge effect correction was necessary. The guard method showed unbiased results for the random pattern (root mean squared error less than 20) while the root mean squared error of this method was high for the non-random patterns (more than 120 in clustered and about 80 in aggregated patterns). The toroidal and weighted methods were efficient in pattern analysis of dispersed and aggregated distributions of Christ's thorn jujube, respectively (with the least root mean squared error less than 20 and 35, respectively). In general, it was concluded that the analytical power of corrected Ripley's K-function varies based on the implemented correction method and the type of investigated spatial pattern.