Relationship between Physico-Chemical Soil Properties and Drying of Coniferous Trees in the South of Tehran and Alborz by Multiple Linear Regression Models

In recent years, as a result of environmental stress, a large number of pine trees in parks and forestry in Tehran and Alborz provinces have dried up, that has conditions provided for contamination by secondary pests such as pine bark beetles (Orthotomicus erosus). In one hundred pine growing locations, three trees were randomly selected in each location, and in these locations, the average percentage of trees canopy dryness (three trees), the percentage of soil area under the trees canopy and the number of irrigations during in growing season (spring and summer) were measured; and thus the soils were sampled from the root activity zone at two depths of 0-30 and 30-60 cm. The percentage of coarse particles (> 2 mm) was measured in the soil samples after being dried in the vicinity of air, pounded and passed through a 2 mm sieve. In each soil sample, soil texture (percentage of sand, silt and clay particles), soil pH, electrical conductivity (ECe), equivalent calcium carbonate (CCE), organic matter and available concentration of sodium, potassium and phosphorus elements were measured. At a depth of 30-60 cm (subsurface soil), a soil sample was taken with a cylinder in order to measure the bulk density and soil porosity. Regression analysis was carried out using three methods Enter, stepwise by backward and forward with 15 series of variables as input variables to estimate or predict the percentage of pine drying (dryness of tree canopy). In total, the forward stepwise regression model, with R2=0.757 and RMSE=15.02% error, the most suitable model is the most appropriate model for assessing the percentage of tree drying, with two variables, the percentage of subsurface sand of the soil (30-60 cm) and coarse material (> 2 mm) of soil surface depth (0-30 cm), a more suitable model was recognized.