In this research, an indicator-based methodology was used to assess energy efficiency, and descriptive data obtained from various types of primary and secondary sources were used to support the findings. Bagda (shrimp)-based agricultural systems (S), Bagda-rice-based agricultural systems (SR), rice-based agricultural systems (R), Galda (prawn)-rice vegetable-based integrated agricultural systems (I) and traditional practices-based agricultural systems (T) were selected for examination based on discussion with local partners and backed up by a review of the literature that describes agricultural practices followed in the area. The data for this study were collected in 2011 and pertain to the cropping season of 2010.
2.1. Description of the agricultural systems All of the study sites examined are located between 22.3500°N and 90.6525°E. The agricultural systems consisting of S, SR, and I are located in Shyamanagar Upazila, Kalijang Upazila and Dumuria Upazila, respectively. Each of these Upazilas (local administrative units) is located in the Ganges tidal floodplain of the southwest coastal belt. In addition, R is situated in Kalaroa Upazila, further north in the floodplain while T is located in Bhola sadar Upazila in the more recently formed Meghna estuarine floodplain east of the other sites. Agriculture in all of the sites is affected by the tropical monsoon climate called Koppen Am. Rice, the staple food of the local people, is cultivated in each location. Rice and other crops occupies the entire agricultural area in R and T, while in S, SR and I, one-third to a-half of the total agricultural land is dedicated to shrimp/prawn cultivation.
2.2. Data collection Primary data were collected by deploying household questionnaire surveys with responses obtained via face-to-face interviews with 212 sampled farmers in the study areas. After field investigation, observation and discussion with the local experts, it was found that the farmers were practicing homogenous agriculture systems in the study sites. A pilot survey was carried out to develop a final version of the questionnaire. Key criteria for selection of the study sites included the dependency of the population for livelihood on local agriculture, positive attitude of the community, community cohesiveness, time-tested and knowledgeable farmers and eagerness to take part in focus group discussions (FGD), as well as support from local NGOs and local government administration. To save time and money, purposeful random sampling (Cohen and Crabtree, 2006) was carried out to select 40, 60, 59, 22 and 26 representative households from S, SR, R, I and T, respectively. The households within each site were selected through stratified random sampling and represented landless labourers and farmers, ranging from those with marginal holdings to those who cultivate more than 2 ha (BBS, 2010). Various documents produced by the Government of Bangladesh and NGOs were also consulted as sources of secondary data. In addition to the individual interviews and secondary data collection, 5 focus group discussions and 20 key informant interviews were conducted to support and validate the data collected. Information related to agricultural products like rice, shrimp, vegetables and other crops as well as inputs like seeds, fertilisers, pesticides, human labour, bullock power, machinery and fuel was collected to calculate energy and economic efficiency.