The study was conducted in the Satkhira district, a salinity-affected coastal area of the Bay of Bengal, situated in southwestern Bangladesh. The SW regions of Bangladesh (Khulna, Bagerhat and Satkhira districts) are operating eighty percent of the country’s shrimp farms. Satkhira has been identified as the most promising area for brackish water shrimp culture due to year-round moderate to high water salinity. Mostly, shrimp is cultivated in this area between February and November when the water of the surrounding rivers becomes saline. The dry season is from November to February; with its high water salinity and scarcity make it hardly suitable for shrimp cultivation. During summer monsoon, from July to October, some farmers grow rain-fed transplanted rice as the overall salinity becomes low. Satkhira district is divided into seven sub-districts. Among them, only Kaliganj and Shyamnagar sub-districts have been considered in this study, because there are a large number of shrimp farms are operating in the area due to the available saline water and the closeness to the river channels. Both sub-districts are located close to the world’s largest continuous mangrove forest. Here, an OSP is implemented by WAB trading international. The OSP has about 200 staff members whose education levels vary from secondary to doctoral degrees. Most of the staff members are local farmers. The OSP works according to an “internal control system” (ICS). The ICS includes internal quality management procedures, internal training, and internal inspections done by the staff as a means to prepare for the external, independent, third-party inspection by IMO. The internal trainers cannot be an inspector and vice versa.
Data was collected between October and December in 2009 during the late harvesting season. This study applied both quantitative and qualitative data. Research was done in collaboration (to help identifying the respondents) with WAB Trading International. WAB cooperates with 160 organic farmers’ groups (15–40 per group) and 3,379 individual organic farmers. From these 160 groups, 12 per sub district were selected through a stratified random sample (stratum = sub-region). In every group, farmers were again selected through stratified purposive random sampling based on the strata gher size (small, medium and large). A total of 144 organic shrimp farmers, 24 in each stratum from each sub-district, were sampled.
Primary data was collected during a face-to-face field survey using a pre-tested questionnaire. Pre-tests were done on 14–16 October with 6 non-sampled shrimp farmers. The questionnaire contained both pre-coded and open-ended questions. The questionnaire was developed in English and then translated into Bengali by the first author to ensure efficient communication with farmers during interviews. Each respondent was given a brief introduction about the nature and purpose of the study before the interview commenced. Then the questions were asked in sequence, with replies being recorded directly onto the questionnaire. For each interview, the time required was about 40 minutes.
As a means of triangulating the data derived from questionnaires, several topics relevant to the study such as farmer’s views and experiences in shrimp culture activities, where presented and discussed in focus groups. A total of 8 focus group discussions were conducted in both sub-districts. Each focus group session comprised 8–12 individuals and the duration of each discussion was approximately an hour. Focus group discussions were conducted only with organic shrimp farmers. The discussions were recorded with a digital voice recorder, and organized with the help of WAB staffs (identification of convenient venues and time). The focus group discussions were held inside collection centers of WAB and in farmer’s residences. During the discussion, the first author of this paper acted as moderator of the sessions. WAB staff was not present during the focus groups.
In addition, 10 transect walks were performed systematically with shrimp farmers by walking across the gher sites at the beginning of the study to build rapport. Transect walks allow researchers to speak and observe with farmers directly at the sites relevant to the research. Thus, they provided informal information on resource use patterns and helped to understand the farming practices and daily livelihood activities. The experiences from the transect walks were also useful for validating farmers’ answers from the questionnaires.
Questionnaire interview data were coded and entered into a database system using MS-Access (Microsoft 2003). The statistical package for social science (SPSS 15.0 for Windows) was used to produce descriptive statistics. Comparisons among farmer’s categories were made by ANOVA F-test and Spearman correlation. The ANOVA was followed by a Tukey Post-hoc comparison of means. Differences are reported as significant at a level of p≤0.05. In some cases, data was normalized using the log transformation.