S. F. Nabavi; N. Yaghmaeian Mahabadi; S. M. Soltani
Abstract
Introduction: highly increases in population growth rate, in particular in developing countries, cause high pressure on agricultural resources. In north of Iran, the pressures are compounded by restricted rice paddy fields. Reliable and relevant land suitability evaluations are vital requirements for ...
Read More
Introduction: highly increases in population growth rate, in particular in developing countries, cause high pressure on agricultural resources. In north of Iran, the pressures are compounded by restricted rice paddy fields. Reliable and relevant land suitability evaluations are vital requirements for land use policy and decision making to support sustainable rural development. Therefore, it is necessary to employ and compare the classic model (FAO), new release technique (multi-criteria decision making strategy) and the capabilities of fuzzy systems to assess land suitability. In recent years, multi-criteria evaluations including Boolean overlay operators and weighted linear combination methods have been increasingly used. Also, using the Ordered Weighted Average (OWA) method can improve the above-mentioned techniques. The OWA method is able to calculate the degree of risk taking and risk aversion of individuals and apply them to the selection of the final option. Therefore, the purposes of the current study were to explore the most reliable method of land suitability evaluation for rice by using integrated fuzzy decision making and determine the optimum depth of soil for quantitative land suitability evaluation for rice production in Amol, Mazandaran province. Materials and Methods: Two-hundred soil samples from 50 observation points at four depths of 0 to 25, 25 to 50, 50 to 75 and 75 to 100 cm with a constant interval were selected. After crop harvesting and taking soil samples from four depths in 50 observation points, and from the genetic horizons of representative pedon excavated in the region, the parameters needed for land suitability evaluation of rice were measured. Then, land suitability classes were calculated using the parametric (square root), Fuzzy-AHP and Fuzzy-AHP-OWA methods and were compared in four depths from 0 to 25, 0 to 50, 0 to 75 and 0 to 100 cm. In Fuzzy-AHP method, Kandel membership functions were used to determine the membership degree and analytic hierarchy process (AHP) was used to determine the weight of each of the effective land properties in crop yield. In Fuzzy-AHP-OWA method, criteria weights were obtained from AHP method and ordered weights using linguistic fuzzy quantifiers. Results and Discussion: The results showed significant difference between the potential yield (5.5 t/ha) and the average of actual yield (3.9 t/ha) in the study area. With respect to the same and acceptable agricultural management of all plots, this difference might be due to soil limitations and subsequently a decrease in the numerical value of the soil index. Except for 0 to 25 cm soil depth, actual yield for the other soil depths showed a positive significant correlation with all calculated land indices by parametric, Fuzzy AHP and Fuzzy-AHP-OWA methods. The compatibility percentage between the representative pedon and observation points was remarkable for 0-50 and 0-100 cm depths in three studied methods. Considering time and cost consuming for land evaluation, this finding shows that 0 to 50 cm soil depth information might be a relevant alternative for the optimal depth to evaluate land suitability for rice in studied paddy fields. The results of the Fuzzy-AHP method showed that soil texture and organic carbon content were the most important soil properties for rice production. The results of land evaluation using Fuzzy-AHP-OWA method showed that with increasing the levels of risk (decreasing the value of α from 1000 to 0.0001), areas with a higher suitability degree occupy greater area. This can be explained by the fact that the strategy associated with the fuzzy quantifier all (α=1000) represents the worst-case scenario (the lowest criterion value is assigned to each location) and under the strategy associated with the fuzzy quantifier, at least one (α=0.0001), the land suitability pattern is composed of the best possible outcomes. The highest correlation coefficients (R2= 0.37) were obtained for Fuzzy-AHP-OWA (α=1) based on correlation between actual yields and calculated land indices for different depths of each parametric method, Fuzzy-AHP and Fuzzy-AHP-OWA. This is due to the high trade-off among the evaluation criteria in the half fuzzy quantifier (α=1). Conclusion: The proposed approach based on Fuzzy-AHP-OWA has great potential to model land use suitability evaluation problem. Half fuzzy quantifier is introduced as the best scenario using Fuzzy-AHP-OWA method for rice land suitability evaluation. This is due to the high trade-off among the evaluation criteria in this quantifier. Given the fact that the land suitability studies are often costly and time consuming, the land suitability evaluation by using 0-50 cm results might be a relevant alternative for the optimal soil depth required for land suitability evaluation in paddy fields.