Background and Objective
Given the decline in agricultural land over the past decades, determining land suitability is a vital step toward sustainable agriculture and environmental preservation. Alfalfa is one of the most important forage crops in the world, playing a vital role in providing livestock feed and supporting the development of the livestock industry. This crop has a special place in crop rotation due to its high production potential, excellent nutritional value, and positive role in improving soil structure. However, to sustainably increase yield and prevent soil quality decline, it is essential to pay attention to optimal land management. Considering the strategic importance of alfalfa cultivation, the necessity of land suitability assessment studies for sustainable exploitation of soil resources, and the research gap in the study area in the field of applying advanced decision-making models, this study aims to determine the suitability of land units under sprinkler irrigation, identify susceptible and limiting areas, and finally, compare the efficiency of the multi-criteria ELECTRE TRI method and the parametric square root method in a part of the lands of the Moghan Plain.
Material and method
The study area, with an area of 3930 ha, is a part of the Dasht-e-Moghan, which is located between 47°43' to 47°50' east longitude and 39°24' to 39°28' north latitude. The average annual temperature and rainfall of the region are 15.4 ℃ and 271.5 mm, respectively. The soil moisture and temperature regimes were aridic borders on xeric and thermic, respectively. For sampling, the study profiles were first gridded and surveyed at 500 m intervals. Finally, a total of 160 profiles were dug, and samples were taken from their different horizons. Finally, 46 soil map units were separated at the regional level. To achieve the suitability of land units, eight key attributes, including soil depth, calcium carbonate, gypsum, pH, electrical conductivity, exchangeable sodium percentage, slope, and climate, were selected using Principal Component Analysis (PCA) and Euclidean distance. Criterion weights were determined via a pairwise comparison matrix and the thresholds for the ELECTRE-TRI model were established based on the crop requirement table. Subsequently, the parametric square root method and the multi-criteria ELECTRE-TRI method were used for land suitability evaluation for alfalfa cultivation.
Results
The results of land suitability assessment showed that, based on the square root method, 11.61% of the land was classified in the S1 class, 63.5% in the S2 class, 24.36% in the S3 class, and 0.52% in the N1 class. On the other hand, the use of the ELECTRE TRI multi-criteria model for qualitative assessment of land suitability showed that 26.79% of the land was in the S1 class, 60.74% in the S2 class, 12.39% in the S3 class, and 0.08% in the N1 class. Slope, soil depth, salinity, and exchangeable sodium percentage are the most important characteristics limiting alfalfa growth in the region. The ELECTRE-TRI model results were more consistent with field reality based on crop yield in units under alfalfa cultivation. Hotelling's Trace and Wilks' Lambda multivariate tests confirmed the accuracy of class discrimination in the ELECTRE-TRI method. Furthermore, the LSD test revealed significant differences among soil properties across the land suitability classes differentiated by this approach. Ultimately, the ELECTRE-TRI method exhibited higher accuracy than the parametric square root method for land suitability evaluation due to its appropriate determination of transition boundaries, application of realistic thresholds, and use of fuzzy logic principles.
Conclusion
From a practical perspective, the achievements of output maps provide planners of the study area with an efficient decision-making tool for implementing smart and spot-based management of inputs, including local amendment of saline-sodic soils and monitoring of phosphorus dynamics, and by separating inappropriate lands, it prevents the waste of investment in the development of sprinkler irrigation systems in sensitive lands. In this regard, it is suggested to the operators and executive experts of the region to prioritize corrective measures such as adding manure fertilizers to improve the physical condition of the soil, optimal use of phosphate fertilizers, and the use of biological fertilizers in units facing limitations. Finally, it is suggested to use exploring the potential of other modern multi-criteria decision-making models, measuring the dynamics of suitability boundaries under climate change scenarios, and integrating this methodology with machine learning algorithms to predict production potential on a large scale are among the most important research needs for future studies. |