Study area This study was conducted in six villages (Keshabpur, Baliadanga, Brahmakati, Ramchandrapur, Durgapur, and Madhyakul) of KeshabpurUpazila (a sort of sub-district) in southwestern Bangladesh from September 2009 to August 2010. Upazila is approximately located between 22° 25' and 23° latitude and 89° 25' and 89° 38' longitude. Upazila's area is 259 km2, and it comprises 9 union parishads, 142 mouzas and 143 villages. In 2009, the total population was approximately 200,229, and the majority of inhabitants were Muslims (80.1%) (BBS, 2009). Most households are predominately dependent on agriculture. In general, men are the primary breadwinners, and women spend most of their time as homemakers. Langurs have been residing in KeshabpurUpazila for over 100 years. One pair of Hanuman langurs was released into Ramchandrapur village by devotees. Since the initial release, the langurs have been living in the area and spread to the other villages of the Upazila (personal communication with an octogenarian Hindu person). Currently, a sizable population lives in the study villages, especially Ramchandrapur, Durgapur and Madhyakul, which are home to the majority of the langurs. Eight of eleven groups live in these three villages (Khatun et al., in press). Langurs come into conflict with humans because of their crop-raiding habits (Khatun et al., 2012a). To minimise crop damage, as well as for the conservation of the langurs, the governmental „Biodiversity Conservation and Nature Development Project. supplies food to the langurs in the study villages. We, therefore, divided the study villages into the following two areas based on how much additional food was distributed to the villages by the conservation project: (1) A high conservation status area (Keshabpur, Baliadanga, and Brahmakati), which is close to the Upazila headquarters. The langurs also received handouts from the villagers. (2) A low conservation status area (Ramchandrapur, Durgapur, and Madhyakul), which is approximately 12 km north of the Upazila headquarters. Visitors hardly visit the area to see the langurs.
Data collection We collected data on people's attitudes and perceptions towards the conservation of crop-raiding langurs using a semi-structured questionnaire (Box 1). In total, 410 people were randomly interviewed: 51% were males, and 49% were females. Additionally, 57.4% were Muslims, and 42.6% were Hindus. Data were collected by a main researcher (UHK, female PhD candidate) and three trained field assistants who were familiar with the local language. Among those interviewed were the household heads, their wives, and other adults (≥18 years) in the house who were willing to answer the questions because some adults felt too shy to speak in front of the researcher and the assistants. The questionnaire had a combination of closed- and open-ended questions regarding background information (age, gender, level of education, religion, and household size) and socio-economic status (occupation, landholding status, and cultivated farm size in acres) of the household. The open-ended questions asked for an assessment of the respondent's thoughts concerning the conservation of common langurs in the area (Box 1).
Statistical analyses Demographic variables were coded as follows: gender (male or female), age (young ≤ 40 years or old > 40 years), household size (small household ≤ 4 or large household > 4), and religion (Muslim or Hindu). Socio-economic variables included the following: access to mass media (yes or no), education (uneducated- no schooling/less than three years of basic education or educated- at least three years of basic education), landholding status (low < 0.1 acres, 0.1 < medium < 0.5, and high > 0.5 acres), and occupation (farmers, who were engaged in crop production and farming, and non-farmers, who were service holders, businessman, students, elected representatives of villages, day labourers, van pullers, etc.). All demographic and socio-economic variables were considered as independent variables in the analyses. Pearson Chi-square tests were used to evaluate differences in the answers of respondents. Stepwise regression analyses were applied to determine which independent variables controlled the locals? attitudes in relation to the various dependent variables. In the analyses, independent variables were coded as dummy variables. Only significant results are reported in the results. For all tests, p < 0.05 was considered to be statistically significant. The SPSS (version 16) statistical package was used to perform all analyses. Throughout the text, respondents are referred to as households, people, locals, or villagers. A household is defined as those who reside under the same roof, share income, and are typically close relatives. We considered “do not know” responses to be missing values in the analyses.