The original methods for site selection in Mozambique predicted the identification of two sites with open access and with high and low HIV/AIDS prevalence. Because of the difficulty in establishing a working basis with communities without previous contacts with the Extension Service, this approach had to be modified initially in order to identify communities. Provincial and district authorities suggested that within the limited time available it would not be feasible to make the initial contacts with the communities to establish the basis for the study. Based on this information, alternative suggestions were discussed in the field, together with the mission coordinator, the consultant and the provincial Extension Service. The original idea of establishing a comparison between high and low HIV/AIDS prevalence within the open access areas was changed; such areas without any management would not offer accessibility for the present study. Community management areas, while under management but not accessible to resource use, did not differ much from open access areas and were therefore considered feasible for this category of study site. On the other hand, to increase the probability of finding affected or afflicted households, a decision was made to find a place where population groups at high risk of HIV/AIDS were concentrated. Based on this view, a forest concession that employs migrant labour would represent particular conditions for high occurrence of HIV/AIDS.
Study sites were selected based on forest management regime and HIV/AIDS prevalence criteria. Miombo woodlands in Mozambique occur north of the Limpopo. Because the southern coastal region of Gaza and Inhambane has been highly modified by agriculture, miombo woodlands in these regions lost most of their original characteristics. Dominant woodlands still maintaining the characteristics of the original miombo may be found north of the Save river, covering most of the forest areas in the central and northern provinces. Considering accessibility aspects, the Beira corridor region (Manica and Sofala) was selected to conduct the study. This region also is the one with the highest HIV/AIDS prevalence (see Section 3). However, within this region, the districts along the highway Beira-Machipanda (EN6) are those that present the highest prevalence index. Districts that are away from that highway have been estimated to have low prevalence indices.
Table 5. General characteristics of study site's forest management regime and HIV/AIDS prevalence
Forest management regime
(estimated district index)
Community forestry, access to forest resources is controlled by the natural resource committee
Forest concession, access to forest resources is controlled by the Concessionaire in coordination with the traditional leaders
Table 5 above presents the general characteristics of the locations selected for field observations. The HIV/AIDS statistics are those of the district, extrapolated from observations carried out in similar districts. In fact, none of the selected districts has a sentinel post for HIV/AIDS monitoring. The prevalence indices for these districts were estimated on the basis of similarity with districts located in areas of similar accessibility, social organization and the level of investment, as well as the concentration of migrant workers. The index by similarity with the suburban post of Chingussura (Beira) was then attributed to the district of Nhamatanda, while the index based on similarity with the post of Caia was attributed to the district of Cheringoma (see Table 4).
The social profile was conducted with the objective of identifying afflicted, affected and unaffected households. A meeting was held with community leaders to present the study objective. Although HIV/AIDS was not specifically mentioned, the need was stressed to identify households with widows, orphans, child-headed households and households with chronically ill or recently deceased adult. Orphans were defined based on CNCS (1999) as children, aged 15 or under, who had lost one or both parents. During this exercise, families were identified that are economically stable (wealthy) as well as those with few resources (poor). This served as the basis for selection of households for detailed interview (Section 4.6).
The objective of the woodland transect is to inventory the various types of forest and tree resources present within the village and surrounding areas, and describe their ecological, spatial and management attributes. The woodland transect was conducted after the first meeting with the community, in which the initial approach produced a preliminary list of woodland products and tree species. A male and a female community member were part of the team in both communities, and the Zangua community also included a child.
The vegetation type, tree species and woodland products were identified as the team went through. Species were identified by community members by their local name and the botanical name was obtained from the research team's experience and consultation on relevant literature (e.g. Koning, 1993; Palgrave, 1983; van Wyk, 1993; and van Wyk and van Wyk, 1997). Some species used for several purposes were not identified botanically either because they were not seen during the transect observation or there were not enough elements for full identification.
Access rules were not particularly checked during this exercise since it had been observed during the introductory community meeting that rules apply in general for specific products, such as charcoal in Mucombedzi, which is regulated by the forest resource management committee, and the need for authorization by the concessionaire at Zangua.
Mucombedzi does not have a village centre and households are scattered within the forest so that each household has its own piece of forest, agricultural and pasture land around the house. Distances were not measured in this case because there were no village centres; however, the distance to the selling point on the road was estimated.
The mapping exercise was conducted to produce the resource map in each community. The research team facilitated the preparation of the map showing the location of the community in relation to the main features (e.g. roads, rivers, sawmill) and the limits in relation to other neighbouring communities. The community residential area was then indicated on the map as the central point and location where they got the resources identified in the focus group. The major products, such as charcoal in Mucombedzi and palm wine in Zangua, were located in relation to the village. The community members drew the lines and, using flipcharts, indicated the key features and location of the woodland resources on the map.
A community meeting with all households was called in advance through the community leader. Because the period of the field work (November-December) is a period of agricultural activity, the meeting was set for a time in which most of the family members would be available, to avoid their delaying the activities or absence of family members. The need was stressed that all men, women, youth and children be called to the meeting so that most of the community groups would be represented. This exercise had the objective of producing (a) the seasonal calendar and activities ranking, for the purpose of identifying different types of production and income-generating activities and their seasonal distribution in the village; (b) contingencies and coping strategy matrix ranking, so as to identify the various sources of safety-nets in the community and their relative value under certain crises; (c) use of forest and tree products in order to identify the various benefits provided by forests and trees, rank the relative importance of these benefits and identify the most important species that provide them. Flipcharts were used to facilitate the interactions and participation of all participants.
A total of 12 households per community were selected for interview. The social profile information was used to identify the target households. The objective of this exercise was to obtain in-depth information on how households use woodland products, how they view the changes in resource use through time, and the coping strategy in different crises including illness and death. The questionnaire prepared for this purpose was used as the basis for the interviews, while direct observations also stimulated additional questions. All interviews were announced in advance and were conducted in the household. The respondent was selected in the household, depending on who was more able to talk about forest resource use. In some households there were more than one respondent at the same time, as the mother and daughter or wife and husband responded alternately. In this case, whoever agreed at the beginning of the interview was recorded as respondent.
Three traditional healers were interviewed in Mucombedzi and two in Zangua. They were interviewed using the questionnaire prepared for that purpose, with the objective of identifying the use of medicinal plants obtained from the forest, as well as the common illnesses, and measuring species-specific changes in the forest specifically in relation to HIV/AIDS. Discussion took place on the trade, rate of change (of medicinal plants and patients) through time, production, protection and access rights to the medicinal plants, and were conducted with a view to identifying the role of the traditional healer in HIV/AIDS treatment as well as the management of woodland resources. Because not all traditional healers would be willing to share their knowledge of medicinal plants, they were advised at the beginning of the interview that they could decline to respond to certain questions regarded as secret.
Key informants are those representing institutions and organizations either working on HIV/AIDS or natural resource management. Institutions at the national level included CNCS, the Ministry of Health and MADER. National institutions also had provincial representatives who dealt with each province on a more operational level. While the role of national institutions is to coordinate, NGOs and local institutions had a more operational role. Local organizations included the health centre, the district administration, NGOs and the forest concession.
The objective of key informant interviews was to understand the organization, programmes and institutions at both local and national levels currently addressing HIV/AIDS and livelihoods in the communities. No formal questionnaire was developed for this purpose. Discussions varied depending on the organization but common questions included the identification of HIV/AIDS action areas (prevention, treatment and care, mitigation) and their relation with other institutions, particularly CNCS and MADER programmes.
Strengths, weaknesses, opportunities and threats (SWOT) analyses were conducted after the household interviews and the major problems and coping strategies identified. This exercise was conducted in a single group with male and female members of the community. Time limitation did not allow the division of the group into males and females separately, as predicted in the original methodology. This exercise was conducted to present the issues related to woodland use and to discuss possible interventions to improve woodland benefits to households, as well as to reduce the problems associated with reliance on forest and tree products. Based on the information initially available from the focus group and the household interviews, the benefits from the forest and the coping strategies were presented to initiate the discussion. Using flipcharts, community members contributed to identifying the strengths, weaknesses and opportunities.
Forest inventory was not conducted as such, but the observations during the woodland transect exercise contributed to identifying the forest resource in general including tree species, tree size, distribution and woodland physiognomy.