2.1. Description of the study area and the study villages Coastal Bangladesh is comprised of 19 districts and divided into three sub-zones (e.g. south-eastern, south-central, and south-western). This study focussed on the south-western (south-west) coastal zone comprising Khulna, Bagerhat, and Satkhira district. The biophysical characteristics of these three districts. The land type of the region is dominated by medium-high land (60%) depending on flood depth, and it is suitable for cultivating at least two crops in a year (BBS, 2018b). Khulna and Satkhira have more than the double land area under triple cropping than Bagerhat, while Bagerhat has a large area for single cropping.
summarizes the soil properties of southwest coastal districts. Clays and clay loams are the dominant soil textures in the area, with low to medium levels of soil organic carbon Soil salinity ranges from very slight to extremely high. Nitrogen and phosphorus levels are low, although potassium levels (low to optimum) are more suitable for crop production in south-west coastal Bangladesh. The three study villages – Atlia, Dhanibunia, and Jhorjhoria – are located in the Dumuria Upazila (sub-district) of Khulna district. The key informant farmers from the villages were selected to collect data on (i) perceptions of environmental change and its impacts on ecology, farming practices and biophysical performance of crops, and (ii) drivers of landuse change and cropping systems intensification. The arable area in Atlia village is medium-high land and mostly free from flash floods. The diversified rice and non-rice crops are grown in the village because of the availability of freshwater for irrigation and a low salinity level (2.0-4.0 dS/m). Dhanibunia (mostly medium-high to medium low land) has moderate salinity (3.0-8.0 dS/m). Arable lands of the village are mostly used for freshwater prawn culture in the gher (modified rice fields) in wet season instead of Aman rice (wet season rice), year-round vegetables are cultivated on dike of the gher, and irrigated Boro rice is grown in the gher in the dry season using available freshwater in the canals. Low lying Jhorjhoria has high salinity (>12.0 dS/m) and scarce freshwater, with production systems commonly based on brackish water shrimp in the dry season using tidal saline water and Aman rice in the wet season.
2.2.1. Primary data collection Primary data on the perception of changes in temperature, rainfall and salinization and the impacts of these changes on cropping systems was collected through focus group discussion (FGD) with selected experienced farmers during November-December 2018 and December 2019. The farmers (e.g. progressive, knowledgeable and experienced) for the FGDs were purposively selected in consultation with the local elites, village leaders and local extension personnel of the Department of Agricultural Extension. Two FGD were conducted in each study village, and 8–12 farmers attended in each FGD (total n = 54). All the participating farmers were male and their ages ranged between 40–55 years old. On average farm size of the participants was from 0.51 ha to 1.5 hectares. The main themes of discussion in the FGD were changes in temperature, rainfall and salinization, the impacts of these changes on crops and aquaculture systems and adaptation drivers behind cropping systems intensification. We sought Human Research Ethics Approval, and the University of New England approved the application (Approval no. HE18-231 and HE19-216).
2.2.2. Secondary data collection Secondary data were based on daily climate records for the south-west region from the Bangladesh Meteorological Department (BMD) and the area and production data of crops and aquaculture from the Bangladesh Bureau of Statistics (BBS), Department of Agricultural Extension (DAE) and Department of Fisheries (DOF). The daily climate data such as maximum, minimum, and mean temperature and rainfall from 1972 to 2013 was collected from the Khulna meteorological station. Three meteorological stations are located in southwest coastal region of Bangladesh. Khulna station is located in the middle of the whole region and nearest (within 20 km) to the study locations. Therefore, climatic data from the Khulna weather station was considered representative of the present research. The area data of rice, non-rice crops, and shrimp and prawn in the south-west coastal zone from 1990 to 2018 was collected from the BBS, DAE and DOF. The soil salinity data were collected from the secondary sources in particular soil salinity reports of Soil Resource Development Institute (SRDI), Bangladesh. The SRDI categorizes soil salinity into four classes: nonsaline (<2.0 dS/m), slightly saline (2.0-4.0 dS/m), moderately saline (4.1-8.0 dS/m), strongly saline (8.1-12 dS/m) and very strongly saline (>12 dS/m) (SRDI, 2010).
2.3. Data analysis Secondary data processing and analyses were carried out using R version 4.0.3 (R Core Team, 2020). The daily data for temperature (maximum and minimum) and rainfall were aggregated into monthly and seasonal data according to the growing season of the crops. Pre-monsoon (mid-Mar to mid-Jun) data represent the early wet season, monsoon data (mid-June to mid-September) represent the wet season and post-monsoon (mid-Sep to mid-Dec), and winter season (mid-Dec to mid- Mar) data represent the dry season. Linear regression was used to assess temporal changes in the environmental variables; maximum temperature (Tmax), minimum temperature (Tmin), and rainfall. The changes were evaluated at monthly and seasonal intervals over the years for which data was available. The regression results are reported as P-values (of the slope coefficient), b (the slope coefficient) and R2 (coefficient of determination), with 95% confidence intervals shown to indicate variability. The findings from the FGDs in three villages and the KII with farmers, extension agents and researchers yielded varying perceptions of climatic change, its impacts and intensification driver. The validation workshops provided an opportunity to clarify the range of views on these issues. These data were analysed using thematic analysis based on expected issues informed by the literature review and emerging concepts from the textual analysis.