Predicting bulk density using pedotransfer functions for soils in Sistan plain

Soil bulk density (BD) is important because of its direct effect on soil properties such as porosity, soil moisture availability, and hydraulic conductivity and its indirect effects on root growth and crop yield. Environmental processes and agronomic practices induce soil bulk density to vary greatly in both space and time. On the other hand, measuring it on a large scale requires a lot of time and is not economical. As a result, indirect methods are used to measure the bulk density when performing large-scale field activities. Pedotransfer Functions (PTFs) have been broadly implemented as indirect cost-effective and time-saving methods in predicting soil bulk density. The purpose of this study is to evaluate the existing Pedotransfer functions in order to determine the bulk density for different soils of Sistan region as well as calibration and provide new Pedotransfer functions for the study area.

After reviewing different reference, 64 different Pedotransfer functions (PTFs) published in different sources were selected to estimate the bulk density. These Pedotransfer functions were selected in such a way that 1)in a wide range of time scale (from 1957 up to date), 2) from wide regional, 3) from various soil land uses 4) from all types of regression techniques and 5) only using common and easily measured predictors such as sand, silt, clay and organic carbon. The soil samples collected in this study was 224 data, which was obtained from 112 points at two depths of 0-15 and 15-30. Three indicators of absolute mean error (ME), root mean square error (RMSE) and standard deviation of the predicted error (SDPE) were used to evaluate.

Among the existing Pedotransfer functions, Benites et al. (2007) with ME value equal to -0.0008, RMSE value equal to 0.1038 and SDPE equal to 0.1033 had the best results. Based on the RMSE value of Yang et al. (2007) with a value of 0.1038 with a rank of 1 and based on SDPE function with a value between 0.0976 Leonaviciute (2000) had the best results. For the study area, 5 presented relationships including linear relationship between BD and OC, linear relationship between OC and BD squares, exponential relationship between BD and OC, linear relationship between BD and OC logarithm and polynomial relationship between OC and BD were presented.

Based on the results it can be concluded that soil organic carbon (OC) is the most important factor in predicting soil bulk density and using soil organic carbon alone, soil bulk density can be predicted with relative accuracy. It can also be concluded that the 5 relationships developed in this study can be used to obtain the apparent density in the study area.