Comparative assessment of soil fertility across varying elevations.

Land elevation exerts a significant influence on soil fertility through affecting macro and micro climatic conditions and geomorphological processes. To evaluate the soil fertility at different elevation classes, namely 1600-2000, 2000-2400, 2400-2800, and > 2800 m, 350 surface soil samples (0-30 cm) were collected from the agricultural lands of northwestern Iran. Soil properties, including soil texture, calcium carbonate (CaCO3), pH, electrical conductivity (EC), organic matter (OM), and soil macronutrients (TN, P, and K) and micronutrients (Fe, Mn, Zn, and Cu), were measured. Finally, the interpretation and classification of the soil samples were made using the nutritional value index (NIV). The comparison of the NIV index based on elevation changes showed that the Gomez method classifies the soil properties in an optimal order as evidenced by its tendency towards the center of the data. However, the Common method is more consistent with the observed trend. After classifying the NIV index using the Common method, it was determined that CaCO3 and soil salinity are not the limiting factor for soil fertility in different elevation classes. However, in all elevations, high pH, low OM at elevations > 2800 m, total nitrogen (TN), available phosphorous (AP), and micronutrients deficiencies, except Zn at the elevation of 1600-2000 m, are the main limiting factors for soil fertility of agricultural lands. The results provide further insight into the elevation-based land evaluation and may supports grower's decision on nutrient management and crop selection strategies.

Digital mapping of soil erodibility factor in northwestern Iran using machine learning models.

Understanding the spatial distribution of soil erodibility factor (K-factor) at the district scale is essential for managing water erosion risk. In this research, we performed to predict the low and high classes of K-factor in the northwest of Iran. Based on this, soil sampling was performed at 64 points using the grid sampling method with 1 km spacing. To calculate the K-factor, the distribution of particle size and organic carbon (OC) were determined. In addition, 21 terrain attributes were calculated by Digital Elevation Model (DEM) to add value to the soil data. Then, K-factor was modeled using Random Forest (RF) and Artificial Neural Network (ANN) models. In the next step, a non-linear Multiple Logistic Regression (NMLR) was used to obtain low and high classes of K-factor. The results showed that the performance of RF is superior to ANN with a high coefficient of determination [R2 = 0.85] and good accuracy [RMSE = 0.003 (Mg ha h/ha MJ mm)]. Therefore, the RF was employed for predicting the K-factor spatial distribution. Finally, using the NMLR model, the study area was divided into low and high classes of K-factor with good correlation [R2 Cox and Snell = 0.78, R2 Nagelkerke = 0.65]. The areas of these two classes were 60.4% for low class and 39.6% for the high class of K-factor. Based on these results, it was concluded that the resultant map of low and high classes of K-factor could be used by farmers and managers for managing soil water erosion risks in the study area.

Urbanization influences the distribution, enrichment, and ecological health risk of heavy metals in croplands.

The contamination of urban soils with heavy elements due to the rapid development of urbanization and urban services has become a major environmental and human health challenge. This study provides insight into the urbanization controls on combined pollution severity and health risk potential of heavy metals in corn-cultivated urban versus non-urban soils. A multifaceted assessment was conducted using enrichment factor (EF), ecological risk (ER), bioconcentration factor (BCF), transmission factor (TF), hazard index (HI), and carcinogenic risk (CR). The results indicate a significant increase in the concentration of all metals in urban farmlands. When compared to the non-urban soils, EF implies a significant increase of all metals in the urban soil, downgrading this index from minimal enrichment (EF < 2) in the control soils to moderate enrichment (2 ≤ EF < 5) in the urban soils. Likewise, the average ER value showed an increase in the urban soils than in the control soils in the order of Fluvisols (66.6%) > Regosols (66.1%) > Cambisols (59.8%) > Calcisols (47%). The BCF and TF values for different elements decreased in the order of Cd (0.41-0.92) > Cu (0.1-0.23) > Zn (0.1-0.18) > Ni (0.01-0.03) > Pb (0.005-0.011) and Zn (0.75-0.94) > Cu (0.72-0.85) > Pb (0.09-0.63) > Cd (0.17-0.22) > Ni (0.01-0.21), respectively, which indicates that certain metals were not mobilized to the extent that they had been accumulated in the plant roots. The total carcinogenic risk was ranged from 5.88E-05 to 1.17E-04 for children and from 1.17E-04 to 2.30E-04 for adults, which implies a greater associated health risk for children.

Selecting environmental factors to predict spatial distribution of soil organic carbon stocks, northwestern Iran.

Knowledge of environmental factors controlling soil organic carbon (SOC) stocks can help predict spatial distribution SOC stocks. So, this study was carried out to select the best environmental factors to model and estimate the spatial distribution of SOC stocks in northwestern Iran. Soil sampling was performed at 210 points by multiple conditioned Latin Hypercube method (cLHm) and SOC stocks were measured. Also, environmental factors, including terrain attributes, moisture index, and normalized difference vegetation index (NDVI), were calculated. SOC stocks were modeled using random forest (RF) and partial least squares regression (PLSR) models. Modeling SOC stocks by RF model showed that the efficient factors for estimating the SOC stocks were slope height (slph), terrain surface texture (texture), standardized height (standh), elevation, relative slope position (rsp), and normalized height (normalh). Also, the PLSR model selected standardized height (standh), relative slope position (rsp), slope, and channel network base level (chnl base) to model SOC stocks. In both RF and PLSR methods, the standh and rsp factors were suitable parameters for estimating the SOC stocks. Predicting the spatial distribution of SOC stocks using environmental factors showed that the R2 values for RF and PLSR models were 0.81 and 0.40, respectively. The result of this study showed that in areas with complex land features, terrain attributes can be good predictors for estimating SOC stocks. These predictors allow more accurate estimates of SOC stocks and contribute considerably to the effective application of land management strategies in arid and semiarid area.

Predicting spatial distribution of soil organic matter using regression approaches at the regional scale (Eastern Azerbaijan, Iran).

Soil organic matter (SOM) is one of the important factors in arid and semiarid areas, which describes the soil quality. Spatial estimation of SOM is important to understand the SOM storage and the emphasis of the SOM in the global carbon cycle and environmental issues. Mapping of SOM content can have significant uses in environmental modeling. In the current study, various methods have been evaluated for estimating the SOM content through soil samples and using auxiliary variables in the west of Eastern Azerbaijan province, Iran. In this study, support vector machine (SVM), multi-factor regression (MFR), and multi-factor weighted regression model (MWRM) approaches have been suggested for predicting and investigating the spatial distribution of SOM. In total, 155 surface soil samples (from the depth of 0 to 30 cm) were obtained. These soil samples were randomly divided into training data set (105 soil samples) and testing data set (50 samples). According to the results, SOM is affected by soil properties as well as environmental factors (normalized difference vegetation index (NDVI)). In this study, clay, silt/sand, NDVI, and soil moisture were used as auxiliary variables in the estimation of SOM. Three methods were compared to determine a suitable method for spatial estimation of SOM, and results showed that SVM has the lowest estimation error (RMSE = 0.100, MAE = 0.07, and MRE = 3.32) and highest regression coefficient (R2 = 0.719) during SOM estimation. The present results show the indirect effect of elevation by controlling auxiliary variables and confirm the importance of auxiliary variables in spatial distribution patterns of SOM.

Predicting the spatial distribution of soil mineral particles using OLI sensor in northwest of Iran.

Texture is one of the most important soil properties that knowledge of the spatial distribution is essential for land-use planning and other activities related to agriculture and environment protection. So, this study was performed to supply the soil texture spatial distribution using standardized spectral reflectance (ZPC1) index of Landsat 8 satellite images in the northwest of Iran. The soil sampling was performed using a random method in 145 points. Mineral soil particles including clay, silt, and sand were determined, and soil texture was calculated. In this study, Landsat 8 satellite images were used to interpolate the soil texture spatial distribution. In the first step, the principal component analysis (PCA) was obtained. Then, PCA1 was standardized using a z-score (ZPC1), and regression techniques were used to create proper relationships between ZPC1 and the primary soil particles. Then, spatial distribution of soil particles was used to create a spatially distributed map of the soil textural classes. The results showed that the standardization of the first component reduced the standard deviation of PCA1 from 23.6 to 10.8. The results of comparing ZPC1 with soil mineral components showed that with increasing the amounts of soil clay and sand, the ZPC1 value decreases and increases, respectively. The results showed that the ranges of the spatial distribution of clay and sand were similar to the laboratory-measured amounts. The results of texture class prediction using the soil texture triangle showed that the amount of similarity between the measured and predicted classes was 53.79%.