Study area
The study was conducted at Raipura Upazila (local government unit under a district in Bangladesh) of Narsingdi districtin Bangladesh. The total area of the Upazila is 312.77 km2, which lies between 23.9607° North and 90.8750° East ([4]). The soil formation of the Upazila is flood plain and grey piedmont. The Upazila has a uniform temperature, high humidity and heavy rainfall which occur from June to October. The average annual temperature is maximum 36 °C, minimum 12.7 °C with annual rainfall 2376 mm. Raipura Upazila, consists of 24 Union Parishads (local government unit under an Upazila in Bangladesh), 113 Mouzas (smaller units of Union Parishad mainly used for land demarcation) and 241 villages. The total population of the Upazila is 454 546 where male 231 449 and female 223 097. Among the total population, Muslim populations are 92%, Hindu 7%, and others 1%. The population density is 2 219 persons km-2 . Average literacy of the Upazila is 34%. The people of the study area are involved with various occupations. The main occupations include agriculture 41%, commerce 14%, agriculture laborers 13%, weaving 12%, service 5%, wage laborers 3%, fishing 2%, industry 2%, and others 9%. Total cultivable land 24 393 ha; single crop 31%, double crop 61% and treble crop 8%; cultivable land under irrigation 39% ([4]). The rivers flown over the Upazila are Meghna, Old Brahmaputra, Arial Khan and Kakan. Among them, Meghna is the dominant river flowing inside this Upazila, creating a unique culture and economy in this Upazila. As Raipura Upazila is the plain land, the overflow of Meghna especially at the rainy season, creates floods in this area. There is no government-owned forest lands in this Upazila. Only privately-owned homestead forests or village grooves/jungles are present here, but these forests are going to be converted from high-biodiverse to lower ones and from the native species to the exotic species (pers. comm.).
Methods of data collection
The study was conducted using the stratified random sampling technique over a period of four months, from November 2008 to February 2009. It was a time-intensive case study in which total time frame planned for the study was flexibly considered. On site, locating the samples was completed within 15 days starting from November 2008 and the primary data collection was finished by the end of February 2009. But in research planning, data collection time was selected purposely considering the easiness of data collection, season and availability of research crews. November through February in Bangladesh is usually non-rainy dry period. Data collection at that period was thought to be easier. However, local volunteers’ support substantiated to finishing the data collection within four months.
Sampling procedure
The sequence of sampling was from Upazila to Union, from union to village and then village to households. Out of 24 Unions of Raipura Upazila, 4 unions were selected randomly as the rural area. The selected unions were Chanderkandi, Mohespur, Paratali and Radhanagar. From each of the selected unions in the rural area, 2 villages were selected randomly. From each of the village, 10 households making 80 in total from the rural area were selected randomly. It was found in the study area that a village had 90-120 numbers of households. So, 10 households in a selected village (around 10% sampling) was considered enough. The list of the unions and its villages were collected from the Upazila administration. After selecting the villages, the households’ name and location was identified with the help of the respective Union Parishad office. The randomization was done with the help of the random number table. With the help of the volunteers involved by the ward and Union Parishad office, the final location of the households was determined and thus data collection was easier with the help of them.
Collection of primary data
Before collecting final data, a reconnaissance survey was carried out to have an overall idea of the study site. However, the final data collection was started from November 2008 and ended on February 2009. The major data included demography of the respondent, land ownership, household dwelling characteristics, types of biomass, biomass source, biomass fuel consumption per month and monthly household expenditure. All kinds of field data were collected by the direct interviewing of the households’ head and physical observation. Some of the facts (5%) delivered by the respondents were cross-checked with the help of the other neighboring people and key persons in the society. The facts mostly included biomass source, biomass fuel consumption per month and monthly household expenditure. In some cases, a little adjustment in the data was needed with a re-discussion with the respective households. Household’s head was the respondent of the study, where all the respondents were Muslim. Male respondents were highest, 96%, while the female were only 4% in the study area. The literacy rate of the respondents was 41%. Business was found as the highest, 42% in all the occupations of the respondents followed by farming, 32%, and so on. Most of the female respondents were housewife, 97%. However, the interviewers were composed of one male and one female at every team. But the local volunteers made our tasks easy to access the facts from the respondents of both male and female.
Data analysis
To find out the dominant biomass used for energy, frequency of use was confirmed for every end-uses of biomass energy carriers. The frequency of the uses was weighted by the amount used of each biomass type. Then the dominant biomass fuel was ranked with the percentage of each biomass type among the total weights. For analyzing the income and energy expenditure, the national currency measurement was used. The national currency of Bangladesh is taka (tk) which could be converted to US $ as US$ 1= tk 70. The conversion can be verified by the average currency exchange rate between taka and US dollars during the study period. The expenditure for each biomass type was calculated summing all the explicit costs of each biomass fuel used by all the households. Biomass fuel used was considered as dry-matter, as the fuels were physically observed and verified to be air-dried at oven-dry condition before burning. So, the moisture content of plant derived biomass fuel was assumed to be 20-25% ([11], [28]). The collected field data were compiled and analyzed with the help of calculator, Microsoft® Excel and statistical package SPSS 13.0. Linear regression analysis was carried out to show the effect of different factors to the energy expenditure. In addition to this, Spearman’s rho (ρ) correlation test was carried out to show the relationship between the variables.