عنوان مقاله [English]
Background and purpose: the importance of litter thickness in the regulation of microclimate and the physical, chemical and biological characteristics of the soil, the regulation of soil input and output exchanges and evolution and continuity of forest stands and the study of its nature in natural endemic hardwood and planted forests and different environmental conditions is an essential issue. In this study, different techniques of interpolation were analyzed and compared to estimate the spatial variability of litter thickness in the Arabdagh area located in Northeast Golestan province.
Materials and Methods: Litter thickness data were collected in 422 sample plots with an area of 400m2 using a systematic cluster network (400×600m, each cluster includes 5 plots with a distance of 100 meters) in different hardwood, softwood tree stands and shrub&herb lands including Pinus brutia, Cupressus sempervirens var. horizontalis, Pinus pinea, Pinus sylvestries and broad-leaved species such as Zelkova carpinifolia, Carpinus betulus, Acer velutinum, Parrotia persica and mixed broadleave. Then, after an initial analysis of the data in statistical software, their spatial and descriptive database was prepared in GIS environment . In order to produce litter thickness thematic map, the efficiency of different interpolation methods of EBK, OK, RBF, LPT, LDW and Co-kriging were compared. Cross-validation was performed to evaluate the accuracy of different interpolation techniques by the coefficient of determination (R2), mean relative error (MRE), mean error (MBE), mean absolute error (MAE) and root mean square error (RMSE).
Results: The results of this study presented that the highest thickness of the litter was in the natural broad-leaved stands and the lowest thickness in the Cupressus arizonica type in the region. Also, in the broad-leaved group, the pure type of Zelkova carpinifolia had the most changes (1.09 cm). While, in the needle-leaved group, the Pinus brutia type showed the most variation in the thickness of the litter. The results of this study showed that the Co-kriging interpolation method with the exponential (0.783), spherical (0.789) and Gaussian (0.791) models using the auxiliary data of Basel area per hectare and with least-squares error and the highest coefficient of determination had high capability Compared to kriging (0.8 to 0.817) and deterministic models (0.875 to 1.05) in spatial distribution interpolation of litter thickness in Arabdagh region.
Conclusion: Due to the effect of litter thickness on some stand and habitat factors such as regeneration , soil quality and Permeability and The intensity of some natural disturbances in the area such as wild fire, it is recommended to take an effective step in managing plan and forestry projects by evaluating it.
Due to the effect of litter thickness on some factors of habitat and habitat such as regeneration, soil quality and permeability and the severity of some common natural disturbances in the region such as fire and the results of this study, the Co-kriging interpolation method with the exponential and Gaussian models compared to other methods, due to the accuracy of the results can be more effective in determining the litter thickness of forest stands as well as managing similar stands and afforestation plans.