High diversity of the components which constitutes stand is very important for increament and enrichment of biodiversity which is one of the considerable criteria of sustainable forest management. There are numerous indexes that are developed to quantify the stand structure diversity. However, sampling methods and sample size of these indexes are yet unclear. Thus, determination of proper sampling desing for stand diversity indexes is the aim of this study. Horizontal (Dominance Index, Species Mining Index and Uniform Angle Index) and vertical indexes (Gini Coefficient and L-moment statistics) are calculated in the 100 by 100 m sized sample plots in the 25 stands which are located in the borders of Sinop-Ayancık Forest District. Then stand diversity indexes are estimated in the 1 hectare sized each 25 stand by using different combinations of Angle-Count Sampling, Single Tree Sampling, Fixed Area Plot Sampling and Six-Tree Sampling methods. Then estimated values hes been compared with reference (actual) values. Three different pattern such as Simple Ramdom, Systematic and Trackside of Single Tree Sampling are also evaluated. Effects of different sample plot size on estimation of diversity indexes has been determined by taking 100, 200, 300, 400, 500, 600, 700 and 800 m2 sample plots on Fixed Area Plot Sampling method. This study shows that sampling methods and sample size has a significant effect on estimation of stand diversity indexes. Diameter based vertical diversity indexes like Gini Coefficient and L-Moment statistics; neighbourhood based indexes like Dominance and Uniform Angle indexes are found to be more sensitive. However, another neighbourhood based index, Species Mining index is found to be less effected by the sample plot size. It is undertood that a single 800 m2 sample plot is enough for sampling Gini Coefficient and L-Moment statistics. Furthermore, 2 sample plot for L-Moment statistics and 3 or 4 sample plot for Gini Coefficient will be enough for sampling the stand with Six-Tree sampling method. Optimal sample size is determined as 30 trees for sampling Gini Coefficient and L-Moment statistics with Single Tree Sampling method. A 400 m2 of sample plot for the Dominance index and a 800 m2 of sample plot for the Uniform Angle index may be enough for sampling with Fixed Area Plot Sampling. It can be said a 500 m2 of sample plot for the Species Mining index is enough for sampling. Optimal number of sampling is estimated as 6 for Dominance index; 4 for Uniform Angle index and 5 for Species Mining index with Six-Tree Sampling method. Optimal sample size is estimated as 30 trees for all horizontal diversity indexes (Dominance, Species Mining and Uniform Angle indexes) with Single Tree Sampling. The suggested sample size and sample methods may be used for making forest management plans if the stands have similar structure with the stands in sthis study. Keywords: Optimal sampling design, Sampling Plot, Stand-Based Forest Inventory.
Tez (Yüksek Lisans) - Süleyman Demirel Üniversitesi, Fen Bilimleri Enstitüsü, Orman Mühendisliği Anabilim Dalı, 2013.
Kaynakça var.
High diversity of the components which constitutes stand is very important for increament and enrichment of biodiversity which is one of the considerable criteria of sustainable forest management. There are numerous indexes that are developed to quantify the stand structure diversity. However, sampling methods and sample size of these indexes are yet unclear. Thus, determination of proper sampling desing for stand diversity indexes is the aim of this study. Horizontal (Dominance Index, Species Mining Index and Uniform Angle Index) and vertical indexes (Gini Coefficient and L-moment statistics) are calculated in the 100 by 100 m sized sample plots in the 25 stands which are located in the borders of Sinop-Ayancık Forest District. Then stand diversity indexes are estimated in the 1 hectare sized each 25 stand by using different combinations of Angle-Count Sampling, Single Tree Sampling, Fixed Area Plot Sampling and Six-Tree Sampling methods. Then estimated values hes been compared with reference (actual) values. Three different pattern such as Simple Ramdom, Systematic and Trackside of Single Tree Sampling are also evaluated. Effects of different sample plot size on estimation of diversity indexes has been determined by taking 100, 200, 300, 400, 500, 600, 700 and 800 m2 sample plots on Fixed Area Plot Sampling method. This study shows that sampling methods and sample size has a significant effect on estimation of stand diversity indexes. Diameter based vertical diversity indexes like Gini Coefficient and L-Moment statistics; neighbourhood based indexes like Dominance and Uniform Angle indexes are found to be more sensitive. However, another neighbourhood based index, Species Mining index is found to be less effected by the sample plot size. It is undertood that a single 800 m2 sample plot is enough for sampling Gini Coefficient and L-Moment statistics. Furthermore, 2 sample plot for L-Moment statistics and 3 or 4 sample plot for Gini Coefficient will be enough for sampling the stand with Six-Tree sampling method. Optimal sample size is determined as 30 trees for sampling Gini Coefficient and L-Moment statistics with Single Tree Sampling method. A 400 m2 of sample plot for the Dominance index and a 800 m2 of sample plot for the Uniform Angle index may be enough for sampling with Fixed Area Plot Sampling. It can be said a 500 m2 of sample plot for the Species Mining index is enough for sampling. Optimal number of sampling is estimated as 6 for Dominance index; 4 for Uniform Angle index and 5 for Species Mining index with Six-Tree Sampling method. Optimal sample size is estimated as 30 trees for all horizontal diversity indexes (Dominance, Species Mining and Uniform Angle indexes) with Single Tree Sampling. The suggested sample size and sample methods may be used for making forest management plans if the stands have similar structure with the stands in sthis study. Keywords: Optimal sampling design, Sampling Plot, Stand-Based Forest Inventory.