2. Materials and methods Case studies of two representative villages in Khulna District were undertaken to assess current cropping practices and their economic viability. The villages selected represent a wide range of typical climate, soil and cropping conditions prevailing in south-west Bangladesh. Five climate scenarios were considered together with three salinization scenarios. The impacts of climate change and salinity were disaggregated in the modelling
2.1. Modelling framework The methodological framework used for the biophysical modelling reported in this paper is illustrated in Fig. 2. The Agricultural Production Systems sIMulator (APSIM) modelling framework (see details in Section 2.7) was used for simulating both climatic and salinity impacts on wet-season (WS), dry-season (DS), and early-wet-season (EWS) rice crops, as APSIM has recently been enhanced to model the rice crop salinity response (Radanielson et al., 2018), beyond its inherent ability to simulate climate and environmental (crop management, soil moisture, nutrient) impacts. However, for maize, sunflower, and wheat, APSIM was employed to simulate only the climate change impacts as the current versions of APSIM-Wheat, APSIM-Maize, and APSIM-Sunflower do not respond to the level of soil salinity. Instead, the salinity impact was assessed by developing salinity response functions from studies in the literature and applying those, post-simulation, to the crop yields simulated in the absence of salinity. The future performance of some existing major DS crops, namely, watermelon and pumpkin, were extrapolated from data available in the literature and expert knowledge as they are not currently modelled within APSIM. Likewise, the future performance of the DS brackish-water shrimp activity was extrapolated from available evidence rather than explicitly modelled.
2.2. Village and household case studies: Two villages in Dacope Upazila in Khulna District with contrasting farming systems were selected for study. Between them, the two villages encompass the main types of farming that are currently practised in the south-western coastal zone, hence they provided a basis for assessing potential changes in farming systems with climate change and salinization. In the first village – Shaheberabad – arable land is intensively used for WS rice and DS non-rice cropping because of low levels of salinity, the availability of freshwater irrigation for DS cropping, and better access to extension services and markets. The second village – Uttar Kaminibasia – practises WS rice/fish farming and brackish-water shrimp farming in the DS due to high levels of salinity with no freshwater for irrigation. This village had better access to tidal water throughout the year but poorer access to markets and extension services. Farm-level data were collected through three case studies of farmers in each of three different farm-size classes, that is, small (< 0.6 ha), medium (0.6–2.6 ha), and large (> 2.6 ha), in each of the study villages (i.e., 18 case studies in total). The case studies involved face-to-face interviews with each farmer using a semi-structured questionnaire. Farmers were asked for details of their current cropping activities, including detailed management practices, decision points, factor inputs and costs, labour use and costs, and crop yields and prices (typical, high, and low values). Details on the methods and results of the economic research are provided in Paper 2.
2.3. Cropping systems Six cropping systems currently practised in the two case-study villages were assessed. In addition, a number of projects have trialled maize, sunflower, and wheat as potential DS crops in south-west Bangladesh and obtained promising results. For example, maximum wheat yields from the saline and non-saline areas of the south-western coastal zone (e.g., Khulna and Satkhira) are reported to range from 2.7 to 3.9 t/ha, indicating the yield potential of wheat in south-western Bangladesh. Similarly, highest sunflower yields range between 2.3 and 3.7 t/ha and maize yields range between 6.9 and 8.8 t/ha in Khulna and Patuakhali. Moreover, the key informant farmers of the case study villages recognized these as potential crops. Thus, maize, sunflower, and wheat systems were included in the simulation scenarios as potential future crops in the face of decreased access to freshwater irrigation, elevating soil and water salinity, and increased climatic risks. The key features of the potential crops are that they are significantly less irrigation-intensive than DS rice, require less fertilizer and pesticide than the major existing DS crops (for example, watermelon), and maize and sunflower are relatively tolerant to drought, pests and diseases, and salinity. On the other hand, farmers of Shaheberabad had stopped cultivating DS rice 5–8 years previously, mainly due to the availability of profitable substitute crops requiring less irrigation. Thus, ten cropping systems were examined – six existing, one former, and three potential cropping systems. These were each assessed under farmers' current management practices in each village.
2.4. Climate scenarios There are many projected climate scenarios used in estimating the impacts of climate change globally. Two emissions scenarios, namely, A2 and B1 from the IPCC's Special Report on Emissions Scenarios (SRES), were used to assess associated climate change impacts on crop production (IPCC, 2001). IPCC ranked B1 as the least harsh climate scenario and A2 as the second harshest climate scenario out of six scenarios (IPCC, 2001). The harsher climate scenario A2 was treated as the “pessimistic” scenario for this study while B1 was used to provide an “optimistic” climate scenario, using specific projections for Bangladesh (Karmalkar et al., 2012). The modelling was conducted for a historical climate period (1984–2013) and two future periods – 2030 and 2060 – giving five climate scenarios in all – historical, 2030 pessimistic (A2), 2030 optimistic (B1), 2060 pessimistic (A2), and 2060 optimistic (B1) as per IPCC (2007). At the time this study was conducted, the new IPCC scenarios had not yet been defined. The SRES scenarios used in this study relate to the RCP4.5 and RCP8.5 scenarios in AR5 (IPCC, 2013). Daily climate data for Khulna station (latitude 22°47′ N and longitude 89°32′ E) from 1984 to 2013 were incorporated into APSIM for simulating historical crop performance, including minimum and maximum temperature, precipitation, relative humidity, and sunshine hours (converted to solar radiation in MJ m−2 d−1 ). crops was assessed across historical (2004–2014) and two future (2030 and 2060) salinity scenarios. In addition, a “no salinity” scenario was used to separate out the offsetting effects of climate change. Both the “with-” and “without-salinity” scenarios were developed for the DS and EWS crops as soil and water salinity substantially affect the crops grown in those seasons. However, only the ‘without-salinity’ scenario was developed for WS rice. This is because the crop is essentially free from adverse salinity impacts as soil and water salinity fall to very low levels in the WS after the commencement of monsoon precipitation. The Bangladesh Soil Resources Development Institute (SRDI) has recorded reliable monthly mean soil salinity data for Batiaghata Subdistrict for 2004–2014. Due to the unavailability of recorded soil salinity data for Dacope Sub-district, the Batiaghata sub-district data were entered in the APSIM input files for modelling current impacts of salinity on DS rice and EWS rice. The Batiaghata data were used with some confidence because the soil salinity level of this location was consistent with one year's data collected from the case-study villages in Dacope Upazila (Kabir et al., 2016). The impact of salinity on other DS crops such as maize, sunflower, and wheat was assessed through a custom-developed salinity response function, applied to APSIM outputs.