Introduction | Methodology | Results | Conclusions | Recommendations

South Kilimanjaro: Methodology

Sampling design

Fieldwork was carried out along transects around the HMFS in Moshi rural district. Two transects of the HMFS were selected on the basis of accessibility, levels of degradation and prevailing land uses. These were forest areas around Mweka and Lyasomboro villages in Kibosho east and Marangu east Wards, respectively. The former formed Ngomberi-Mweka transect while the later constituted Isari-Lyasomboro transect. Both Mweka and Lyasomboro villages border the HMFS. Location of the villages was geo-referenced by using Geographical Positioning System (GPS) GARMIN GPS 12XL Software 4.00. A household formed the sampling unit. This composed of a husband, wife, children and family dependants living in the family. A household also composed of single fathers and mothers, unmarried or widowed.

Sampling procedure and Sample size

A stratified simple random sampling procedure was employed to obtain a sample of fifty respondents from each village. The total sample size was one hundred respondents. The number of people in each village register was divided into strata based on sex and respective sub-villages. There were seven strata in Mweka village and four in Lyasomboro village. An almost equal number of females and males comprised the sample (n) (Table 2.2)

Table 2.2 Sampling procedure and sample size

Source: Field Survey (2001)

A table of 5-Digit random numbers generated in the computer software LIMDEP version 5.1, developed by Green (1988), was used in picking a random sample from the village registers. A sample size (n) in each stratum was picked on the basis of its proportion to a sampling frame (N). The criterion was that If N<10 one digit was picked from the random numbers’ table and two digits if N >10<100. The rule, however, was that a zero, any number greater than N or a repeated number in the table of random numbers was discarded. The random number picked in the table was corresponded to the name of the bearer of that number on the register for each stratum. The total number of the sample (n) constituted a gross proportionate number of selected individuals in each stratum (Table 2.2).

Types and sources of data

Both primary and secondary data were collected for the study. Primary data included land use cover patterns and changes, socio-economic characteristics, population, perceptions, status of the forest, and management of the HMFS (Table 2.3). Secondary data comprised of management history of the HMFS, socio-economic characteristics, population dynamics, and experiences on land use changes and their impacts. Primary data were derived from structured interviews using questionnaires, in-depth interviews of key informants, discussions, field observation and a mosaic of Landsat 7TM+ Satellite images for October 1999 and February 2000 (Plate2.1). Secondary data were obtained from literature survey, 1952 and 1982 land use/cover maps of the southern slopes of the mountain (Yanda and Shishira, 2001), reports, files and documents (Table 2.3). A literature survey was undertaken at the University of Dar es Salaam, District Administrative Secretary’s Office in Moshi and the Catchment Forestry office–Moshi.

Table 2.3. Types and data sources

Data collection techniques

Data collection was done through a methodological triangulation; this entailed the use of multiple methods to study a phenomenon or a single problem. Triangulation is a process of data collection, which involves looking at an object from more than one standpoint (Mwanje, 2001) The methods used included interpretation of satellite imagery, questionnaire interviews, field observations, in-depth interviews and discussions. As a strategy, these methods were put to play to gather information and allow for the crosschecking of facts.