Efficiency of the Transects Branch-Sampling Line Method to Estimate the Canopy and Number of Trees per Hectare in Zagros Forests (Case Study: Karazan Forest of Ilam)

One of the basic parameters in forest management is the detailed knowledge of the trees’ growing stock for planning, which is operational by forest inventory. Excessive human interference and dependence on the forests have changed their natural structure. Although the circumstances of wood harvesting are not provided in these forests, their role is essential in soil and water conservation, the livelihood of local people, and environment improvement. Sampling methods must be accurate, inexpensive, and easily implementable in the wild. This study presents a new sampling method, called a branching transect, in the Iranian Zagros forests and similar forests. Features of the new method include proper accuracy, easy implementation in nature, simplicity of statistical calculations, and low cost. The number of trees per hectare (density) in a forest stand represents the existing situation, and its monitoring is very important to know the changes in forest stands. The use of distance sampling methods, especially in open forests, has grown in recent years due to the reduction of inventory costs and optimal accuracy. The research mainly aims to evaluate the efficiency of a proposed distance method (a branched sampling line) with easy implementation and good accuracy in a 100 × 100 m network and the GIS environment.

The studied area is located in the southwestern part of the Manshet and Qalarang protected area in Ilam County. First, 100% inventory was applied in a 90-hectare area of the studied forest, and then 10 and 20-m sampling lines with branches of up to a maximum of 20 m (90 sampling points) were used for this study. In this method, a transect is used that includes some sub-transects. The length of the main transect, sub-transect, the number of trees measured in each sub-transect, and the number of sub-branches in this method are changeable based on the homogeneity, heterogeneity, density, and structure of the forest and are determined according to the forest and expert opinions. In general, less numbers of sub-branches and trees can be considered when the forest is denser or more homogeneous. It is better to consider more sub-branches in dense forests, such as Zagros forests so that a wider area can be covered under inventory and an acceptable number of trees can be examined in transects. Due to the density and heterogeneity of the studied forest area, 20-meter transects with 2, 4, and 6 sub-branches were used in this study. To reach the study goal of the comparison among distance sampling methods, the results of these samplings should be compared to a real amount to have the final judgment on their benefits and weaknesses. Accordingly, the results were compared to 100% inventory in the study area. In 100% inventory, geographical features and the measured quantities for each tree were entered into the GIS environment. This research is the first to use the sampling method of a sample line with 2, 4, and 6 m lateral branches. To evaluate the methods, they compared with 100% inventory based on the accuracy rate and t-test.

The normality of data was tested through the chi-square test, and the unpaired Student`s t-test was used to compare the average numbers and canopy of trees in a hectare, resulting from real amount sampling (100% inventory). Based on the whole amount of trees in the area (12079( and the whole area of the canopy of trees (62.32 m2), the real average amount of trees per hectare and the real average value of the canopy area per hectare were calculated as 141 and 3704, respectively 12). The 6-branched two-tree sample line with 141 trees per hectare and the four-branched one-tree sample line method with a canopy of 3815 m2 per hectare showed the best results compared to the real mean. The analysis showed that most of the used sampling methods  (the branched sampling line with six different patterns 6b2t- 6b1t- 2b1t- 2b2t- 4b2t-4b1t) yielded acceptable results for the evaluation of open forests. However, the branched sampling line method in most networks showed a lower statistical error and was closer to the real value in terms of the mean number of trees per hectare.

Totally, the branched sampling line has an accuracy percentage within 10% to estimate the number of trees per hectare in most networks. In terms of the mean number of trees per hectare, this method was not significantly different from the actual value with a probability of 95% and was more accurate than the other distance methods tested in this research. Considering the low percentage of sampling errors, acceptable accuracy, and easy implementation in nature, the branched sampling line was evaluated as a more suitable method than the other approaches. Due to the novelty of the presented sampling method, there was no case for comparison, in other words, transect branching has not been applied in other areas that can be compared with the results of this study. Finally, the results of this review for the parameters of the number and area of canopy of trees per hectare and its comparison with the actual value and flexibility of this method (changeability of the number of branches, number of trees in each branch, and the length of the main and sub-transects) demonstrate that it is a proper sampling method in dense and even normal forests.