Consistent with participatory rural appraisal methods, we used the following data collection methods: informal interviews with key informants, field observation, participatory timeline development, focus group interviews with local farmers, and follow-up interviews in two selected communities of Narsingdi district. Field research and data collection took place in two sub-districts of Narsingdi – Shibpur and Belabo from November 15, 2011 to January 30, 2012. Both of these sub-districts straddle two agro-ecological zones–the terrace ecosystem zone of Madhupur Tract and the floodplain zone. About 37% of the total area of Shibpur and 35% of the total area of Belabo lie in Madhupur Tract (SRDI, 1991; SRDI, 2003). The Madhupur Tract part of the two sub-districts was covered by the agroforestry improvement project. From this area, two villages, ‘Kumertek’ from Shibpur and ‘Chitam’ from Belabo were selected for in-depth analysis. These two villages were selected because of relative accessibility for in-depth field research, evidence of long-term cropland agroforestry practices, and clear indications of socio-cultural activity related to local agroforestry practices.
At the beginning of the fieldwork, key informant interviews were conducted with six local people, three in Shibpur and three in Belabo, including a primary school teacher and social worker, officials of the agricultural section in Union Parishad (local-level administrative unit) and locally well-known and experienced farmers. These key informants provided firsthand information about the communities, experiences with agroforestry practices, and trends in agroforestry development. A timeline was created with six elderly farmers to gain a historical perspective on cropland agroforestry systems. This exercise involved the respondents recalling past events chronologically and capturing the major changes and transitions that took place over time. Factors that induced those changes were also revealed. The timeline was constructed in the two communities separately through a meeting with a small group of three elderly farmers in each area. They were asked to describe how and why agroforestry systems have changed over time in the region.
To obtain information on socio-cultural dimensions that are linked with current agroforestry systems, focus group interviews were conducted with local farmers in the two communities separately. The focus group was conducted with eight farmers in Shibpur and seven in Belabo based on a set of questions generated through the literature. At the same time there was flexibility for additional questions to be generated from the discussion. Since the main project was targeted at cropland agroforestry improvement, data collection in this research was focused on cropland agroforestry. In all cases, the participants were selected purposefully on the basis of their experiences, expertise and knowledge of cropland agroforestry so that they could act as ‘information rich cases’ (Baxter and Eyles, 1997). Finally, after the completion of the focus groups, follow-up interviews were conducted in-person with three participants from the focus groups and key informants. These interviews generated detailed insights on the issues that emerged as important and highlighted during the focus group and key informant interviews. For instance, additional investigation became necessary on issues related to the prevailing dominant agroforestry system – ‘jackfruit-Burmese grape’. Additionally, secondary data was collected to further understand the context of the study areas.
6. The study area Agroecologically, a distinct characteristic of villages in the Madhupur Tract is a terrace which generally stands one to ten meters above the adjoining floodplain (Rashid, 2006). The terrace areas consist of well-drained upland areas locally called “chala” and flood-prone valley areas locally called “baid” (SRDI, 1991; SRDI, 2003). Differences in vegetation can be seen between the uplands and the valleys. Fruit trees such as mango, jackfruit, litchi, jujube, olive, pineapple, banana, Burmese grape are common in the uplands. Under the big fruit trees (such as jackfruit, mango, litchi), various types of spices (such as turmeric, ginger), vegetables (such as cucumber, sponge gourd, chilli, papaya, eggplant/brinjal, bitter gourd) and shorter fruit species (such as lemon, pineapple, Burmese grape) are grown. In general, the jackfruit-based system is the dominant system in the terrace ecosystems of Madhupur Tract. Rice is the main crop in the valleys.
Shibpur is located about 56 kilometres north-east of Dhaka, which is the nearest city and capital of Bangladesh, and Belabo is located about 79 kilometres north-east of Dhaka. Both sub-districts are situated in the proximity of Dhaka-Sylhet national highway which acts as the main conduit from these regions to the other parts of the country. Within the sub-districts there are well developed paved regional road networks. The study villages have internal local road networks linked with the regional ones, where some are paved and some are unpaved. Also, there are rivers around the sub-districts that play a significant role in communication from these two regions to other places. Figure 3 depicts the map of the study district, Narsingdi, delineating its sub-districts and other important features such as the national highway with the thickest red line. In the inset, the position of the district is shown on the map of Bangladesh.
The study communities are very small and rural. Kumartek, the village from Shibpur, covers an area of 200 acres with 479 people living in 106 households. Chitam, the village from Belabo, constitutes an area of 219 acres with 766 people living in 167 households (BBS, 2007a). However, according to local informants, the populations of both communities have significantly increased within the last ten years. The literacy rates of Kumartek and Chitam are about 36% and 22% respectively among the population older than 7 years (BBS, 2007a). In both villages most households depend on agricultural activities for most of their income, although some are also engaged in non-farm activities such as business, service, non-agricultural labor (e.g. rickshaw/local van pulling, construction worker) and others. In terms of ownership of agricultural landholdings, about 80% of the population of both villages own agricultural lands in small and fragmented plots (.05 acre to 2.49 acres) (BBS, 2007b). There are smaller numbers of relatively well-off people who own substantial amount of land (above 2.50 acres), and landless people who may have only a very small portion of homestead area but no farm-holdings (BBS, 2007b). Most landless people work as sharecroppers or day laborers.