Assessing the Optimum Cluster Sampling Plan for Estimating the Quantitative Characteristics of Zagros Forests (Case study: watershed Olad Ghobad forests)

Abstract

Protection and conservation of forests require access to information and awareness of structural features such as number per hectare and canopy cover of trees. This study aimed to select the optimal cluster sampling design in estimating the quantitative characteristics of the forest according to the three measures including accuracy, precision, and time required for the sampling. For this purpose, 30 hectares of watershed Olad Ghobad forests in the west of Lorestan province were selected. The Cartesian coordinates of all the trees in them were determined. In Matlab2014 software environment, the position of all trees was simulated. Then different cluster sampling designs (triangular, square, star 1, linear, L-shaped, star 2) were defined in the research by taking into account the number of samples (30, 60, and 90) and three distances (10, 20, and 30 meters). These processes repeated 10,000 times. The results showed that there was no significant difference between the estimated mean of different designs of cluster sampling methods of the study area. Based on the number of samples and the distance studied, it was found that sampling error for linear design has the highest accuracy compared to other designs. The results of estimating the E%2× T index for the characteristic number of trees per hectare and canopy cover showed the lowest values for the square (770.99) and linear (2347) designs, respectively. The results of accuracy estimation for both mentioned characteristics of Star 2 Design showed the lowest values (18.83) and (-5.54), respectively. The results of accuracy estimation for both mentioned characteristics of Star 2 Design showed the lowest values (18.83) and (-5.54), respectively.

Keywords

• canopy cover
• cost
• e%2× t index
• forest inventory
• number per hectare
• precision