Evaluation of distance methods for estimating population density in Populus euphratica Olivier natural stands (case study: Maroon riparian forests, Iran)

Abstract

The aim of this study was to determine the performance of distance methods in terms of accuracy, precision, bias, consumed time and sampling efficiency in the Maroon riparian forests, Iran. 40 estimators were used to evaluate the density of Populus euphratica Olivier trees in pure and mixed stands. Fifty quadrates (30 × 30 m) were established in each stand. To evaluate the accuracy, precision, bias, consumed time and efficiency of sampling techniques, relative root mean square error – RRMSE (%), coefficient of variation – CV (%), relative bias – RBIAS (%), t × RBIAS², t × E², where t is study time and E (%) is sampling error at a confidence level of 95%, and efficiency ratio between method j and k (Ef_jk) were used. A compound of three basic distance estimators sampling method and n-tree were the best in both stands according to all criteria for density estimation. Moreover, variable area transect by Parker (g = 3) and quadrat method were the best methods for density estimation only in pure stand, while the angle order-point-centred quarter method was superior in mixed stand. Regarding to the results, we recommend the use of compound of three basic distances (BDAV3) and basic distance-nearest neighbour (BDNN2) for density estimation of P. euphratica stands in riparian forests.