Soil conductance classification for crop performance assessment using electromagnetic induction and geospatial techniques in coastal region of Indian Sundarbans

Sustainable crop production in the coastal zone of the Indian Sundarbans presents a significant challenge due to seasonal salinity build-up, particularly during the post-monsoon season. The measurement of apparent electrical conductivity (ECa) of soil using electromagnetic induction (EM) technique has emerged as a popular method for rapidly assessing soil salinity, serving as a proxy for traditional cost and labour-intensive laboratory analysis methods. This study investigated two test sites, Bijoynagar (9.24 ha) and Sonagaon (9.69 ha), to assess the spatial pattern of ECa at four depths of exploration (DOE)- 30 cm, 50 cm, 80 cm, and 160 cm below the surface using the DualEM-1HS. The measured ECa data were interpolated using Ordinary Kriging (OK) to get contiguous grid layers of ECa at respective depths which are subsequently classified into six Soil Conductance Units (SCUs) using K-means clustering technique. The Bijoynagar site, where mung beans and a few other vegetables were cultivated in the rice fallow during the survey recorded lower ECa than the Sonagaon site. The major part of the Sonagaon site was laid fallow during the survey, with patches of summer rice grown sporadically. The coefficient of variation of ECa within each SCU was most pronounced at the surface level (up to 30 cm depth), which was often influenced by agricultural practices such as tillage and irrigation. Strong correlations were observed between the laboratory-measured electrical conductivity of soil-water suspensions (ECe) and the cluster numbers, as evidenced by a high Spearman Rank Correlation coefficient (0.905). The ECa of the crop rhizospheric layer (surface to 50 cm depth) exhibited the highest correlation coefficients for mung bean stover yield, biomass yield, and rice grain yield. Spearman rank correlations between cluster numbers and the yield of summer rice and mung beans grown in different SCUs were also significant at the 5 % level. The investigation of the electromagnetic induction approach for assessing apparent soil conductivity has been crucial in understanding local fluctuation of soil salinity and its impact on performance of the two major crops grown in the Sundarbans. The improved EM survey involving hand-held GPS and Android application ‘GeoTracker’ followed by geostatistical analysis, and clustering in open-source GIS platform customized in this study offers a robust package of practices that enabled quick, precise and effective classification of land into Soil Conductance Units for site specific crop management in the coastal saline region of Indian Sundarbans.