Modeling of soil exchangeable sodium percentage using easily obtained indices and artificial intelligence-based models

Model. Earth Syst. Environ. (2016)2:130 DOI 10.1007/s40808-016-0185-8 ORIGINAL ARTICLE Modeling of soil exchangeable sodium percentage using easily obtained indices and artificial intelligence-based models Ali Keshavarzi1 • Ali Bagherzadeh3 • El-Sayed Ewis Omran2 • Munawar Iqbal4,5 Received: 26 June 2016 / Accepted: 4 July 2016 Ó Springer International Publishing Switzerland 2016 Abstract Salinization and alkalization of land resources are the major obstacles to their optimal usage in many arid and semi-arid regions of the world, including Iran, since potential evapotranspiration is more noteworthy than precipitation in these areas. The amount of water that enters the soil is low and this results in salt accumulation in soils, which makes the soil infertile. Moreover, existence of salts, for example, sodium, in soils causes dispersion of soil particles and soil degradation, and intensifies soil erosion too. Monitoring exchangeable sodium percentage (ESP) variability in soils is both time-consuming and costly. However, in order to estimate the amounts of amendments and land management, it is necessary to know ESP variation and values in sodic or saline and sodic soils. Thus, introducing a method, which utilizes easily obtained indices to estimate ESP indirectly is more optimized and economical. Input and output data, i.e., ECe (dS m-1), clay (%), pH and ESP (%) were collected and measured from 100 soil samples in light of a stratified random sampling from Mashhad Plain, Khorasan-e-Razavi & Ali Keshavarzi 1 Laboratory of Remote Sensing and GIS, Department of Soil Science, University of Tehran, P.O.Box: 4111, Karaj 31587-77871, Iran 2 Soil and Water Department, Faculty of Agriculture, Suez Canal University, 41522 Ismailia, Egypt 3 Department of Agriculture, Islamic Azad University, Mashhad Branch, Emamyeh Boulevard, P.O.Box: 91735-413, Mashhad, Iran 4 Department of Chemistry, The University of Lahore, Raiwind Road, Lahore, Pakistan 5 Department of Chemistry, Qurtuba University of Science and Information Technology, Peshawar, KPK 25100, Pakistan Province, Northeast Iran. This study aims to propose some models to estimate ESP by easily obtained properties of soil. In this regard, the efficiency of artificial intelligence-based (AI) models (i.e., Artificial Neural Network, ANN, and Adaptive Neuro-Fuzzy Inference System, ANFIS) was investigated and compared. Accuracy results showed that owing to highest R2 and the lowest mean square error (MSE), ANFIS model predictions were superior to the MLP model for indirect estimation of soil exchangeable sodium percentage. Keywords Artificial intelligence  Prediction  Exchangeable sodium percentage  Mashhad plain  Iran Introduction Precision agriculture practices in arid and semi-arid areas like Iran require periodic information on soil salinity and alkalinity, which are the most essential issues threatening sustainable agricultural management (Keshavarzi and Sarmadian 2012; Kilic and Kilic 2007; Omran 2008). The amount of agricultural lands having salinity and alkalinity problems increase continuously as identified with climate, topography, groundwater level and quality of irrigation water (Postel 1989; Ayers and Westcot 1989; Kilic and Kilic 2007). The most widely recognized reasons of salinity and alkalinity are low precipitation, high evapotranspiration, and low quality of irrigation water. Saline soils contain soluble salts in adequate amounts to interfere the growth of most crop plants, yet they do not contain enough exchangeable sodium to adjust soil characteristics (Kilic and Kilic 2007). However, alkali soils incorporate exchangeable sodium in a sufficient quantity to interfere with the growth of most crops (Bohn et al. 1985). 123 130 Page 2 of 10 Model. Earth Syst. Environ. (2016)2:130 Soluble salts influence the productivity of soils in two principal ways: changing the osmotic potential of soil solution and increasing the content of exchangeable sodium, which produces in many soils an unfavorable physical condition (Pozdnyakova and Zhang 1999). There is a close relationship between soil properties and salinity and alkalinity (Kilic and Kilic 2007), which is related to soil texture, water content and bulk density (Pozdnyakova and Zhang 1999). Salinity stress poses three challenges, including water shortage (drought stress), ionic toxicity, and nutrient imbalances to crops (Sarani et al. 2015). The presence of abundant ions in the root zone causes the absorption and effectiveness of nutrients to decrease greatly, and on the other hand, increases absorption of any unnecessary elements (Pessarakli 1991). Saline soils have increased significantly in Iran and throughout the world. Approximately 44.5 M ha of arable lands is influenced by dry land salinity in Iran (Banaei et al. 2005; Sarani et al. 2015). Moreover, the application of irrigation water of low quality may result in increasing soil salinity. Management of irrigated arid and semiarid soils and land application of industrial and food processing wastes often requires frequent monitoring of soil salinity and sodicity variation. Such monitoring practices are feasible and economical only where simple, straightforward, and rapid methods are accessible. Despite the increasing prevalence of salinity worldwide, the estimation of exchangeable cation concentrations in saline soil remains problematic. In this situation, it is desirable to determine relationships among soil salinity indices. Soil electrical conductivity (EC), as a suitable index, which is reliable, cheap, and can be measured fast, has long been utilized by several researchers (Auerswald et al. 2001; Seilsepour and Rashidi 2008a; Adhikari et al. 2011). On the contrary, monitoring the changes of soil sodicity is costly in addition tedious. For saline or saline-sodic soils that are undergoing the amendment procedure, or when applying high sodium adsorption ratio (SAR) irrigation water or wastes, it is necessary to monitor the status of soil exchangeable sodium percentage (ESP) or SAR frequently. This, alongside pH and EC monitoring, is advisable for selecting and adjusting water and waste, estimating the amount of amendments and management practices (Robbins 1993). The SAR and ESP are two acknowledged indices for evaluating the degree of soil sodicity. The soil ESP is obtained by Eq. (1): ESP ¼ Naexchangeable ðme=100g soilÞ  100 CECðme=100g soilÞ ð1Þ As shown in Eq. (1), determining the cation exchange capacity (CEC) is necessary for estimating soil ESP. The CEC measurement in the laboratory is very costly and time-consuming, and contains errors (Rashidi and 123 Seilsepour 2008; Seilsepour and Rashidi 2008b, 2008c). In order to conquer the aforementioned issues, presenting a method that could utilize another parameter to calculate ESP in an indirect manner is more optimized and economical as well. Statistical methods have been widely used to model and predict the soil ESP from easily obt (...truncated)