This study was conducted in four villages from two unions in Asasuni upazila of Satkhira district. The proximity to sea and existence of high saline areas having 16+ dSm-1 EC value (SRDI, 2010) in some places encourage the authors to select the district. The Chapra and Mohessor Kathi villages at Budhata union represents low saline area and the Sreeula and Moheskur villages at Sreeula union represents high saline area.
The first phase of this study was collected data on livelihood strategy of the people living in the study area. The income, expenditure, employment, and land-use pattern were considered as four parameters to represent livelihood strategy of the respondents and a well-designed interview schedule was administered and collected information of the above noted and some other socio-economic characteristics of the farmers. A total of 150 respondents (farmers) were interviewed by taking 75 randomly selected respondents from each of Budhata and Sreeula unions.
The second phase of this study was collected soil samples from 150 plots taking the biggest plot of each of the surveyed 150 farmers in the first phase for collecting socio-economic data. The collected soil samples were analyzed in the laboratory for getting EC and pH values. The third phase of this study was analyzed the collected socio-economic data and laboratory test results. Statistical techniques such as frequency distribution, correlation, t-test, and regression analysis are used to analyze the data. STATA and MINITAB are used for this purpose.
This study used frequency distribution to represented data on land ownership, freshwater source, distance from freshwater source, age, income, expenditure, land-use pattern, intensity of harvest, etc. These indicators were considered to represent livelihood strategy and socio-economic status of the farmers in this study. A t-test is used to check the statistical significance of mean difference between two groups for salinity and income. For salinity, the groups are high and low saline areas and the corresponding null hypothesis was: HO1: No difference in salinity level between low and high saline areas. Similarly, for income, the corresponding null hypothesis was: HO2: No difference in income level between low and high saline areas. In a correlation analysis, this study uses age, education, land holding, income, expenditure, salinity, pH, land-use pattern, water sources and occupation/employment. A pair-wise correlation gives the correlation coefficients among the pairs of the cited indicators. This study uses a multiple regression analysis to check the influence of salinity on livelihood strategy of the farmers. Four dependent variables for four multiple regression models are income, expenditure, employment and land-use pattern.
The income and expenditure are continuous variables which measure the yearly income and expenditure of the respondents in a unit of thousand Taka. For dichotomous type of dependent variables, such as employment and land-use patter, both linear probability model and logit model are estimated to check robustness. The corresponding explanatory variables were age, education, land holdings, income, expenditure, distance from fresh water source, salinity level, pH, land-use for shrimp, water sourcing from tube well and occupation as a day labor. For running the first and second regressions with income and expenditure as dependent variables respectively, the variables ‘income’ and ‘expenditure’ were dropped in both cases from the list of explanatory variables. Similarly, for running regressions with employment and land-use pattern as dependent variables, the corresponding variables were dropped respectively from the list of explanatory variables.