Location of forest plantations in Mozambique: Gains and losses in water, firewood and land availability

Highlights

• Industrial forest plantations affect access to water, fuel and food.

• Local communities have high preference for avoiding any loss of water availability.

• Acceptance of additional plantations relates negatively to current positive impacts.

• Differing preferences for plantations stem from village-specific conditions.

• Information on trade-offs serves the planning process of plantations.

Abstract

Industrial tree plantations in developing countries result in land-use changes affecting local livelihood possibilities and the availability of ecosystem services. Given the benefits that plantations also produce, the key question lies in making plantations compatible with other land uses. Our paper investigates trade-offs related to ecosystem services and land-use options by applying a discrete choice experiment in Niassa, Mozambique. The study addresses the experiences and preferences of local people concerning the impacts of plantations on the availability of selected ecosystem services (water and firewood), and the characteristics of plantations (distance from farm plot and type of land converted to plantation). The data collected with household interviews in the neighbourhood of four plantation sites were analysed with the random parameters logit model. Access to water and greater distance from farm plots were perceived as more important factors than easier access to firewood or the type of land converted to a plantation. Preventing declines in water availability is by far the most important factor for local people, and is valued higher than the improvement of water availability, in accordance with theoretical expectations. Both observed negative impacts and experienced positive social impacts of plantations induced preference heterogeneity for plantations to be located at greater distances from farmland, water availability, and tolerance for additional plantations in the area.

Introduction

Forest plantation area has expanded along with an increase in global demand for wood fibre. In 2015, the Food and Agriculture Organization (FAO, 2016) estimated a total of 291 million hectares of planted forest worldwide, which is about 7 per cent of the global forest cover. Between 2010 and 2015, the planted area was growing at a rate of 3.3 million hectares per year, the size of increase being, thus, about half of the annual natural forest loss (FAO, 2016). Many industrial plantations are situated in Latin America and Asia, but interest towards Africa has significantly increased due to land availability, low labour costs and government incentives (Barua et al., 2014).

Industrial forest plantations require large land areas. While providing employment and income, their establishment may affect the availability of ecosystem services, decrease local communities’ access to land, constrict agricultural activities and decrease the willingness to invest in agriculture. In developing countries, the welfare of local communities is often highly dependent on natural assets, and changes in the availability of ecosystem services may therefore have severe impacts. Conflicts have occurred between local people and investors or authorities due to these impacts (Gerber, 2011; German et al., 2011). Negative impacts of investments on land-dependent communities are not restricted to large-scale industrial forest plantations, but also apply to other industries such as mining and agriculture (for these, see e.g. Thondhlana, 2015; Mitchell, 2016; Schoneveld, 2016; Yengoh et al., 2016).

Industrial forest plantations impact land use and ecosystem services in various ways. If established on agricultural land or on land important for collecting various forest products, they directly impact peoples’ livelihoods and use of ecosystem services (Gerber, 2011; Nahuelhual et al., 2012; Nube et al., 2012; Obidzinski et al., 2012). Even though plantations provide certain ecosystem services for local people, the quality and quantity in the provision of ecosystem services are likely to change (De Groot and van der Meer, 2010; Heilmayr, 2014). Firewood and non-wood forest products (NWFPs) are often unavailable from monoculture plantations, and even if available, access may be denied. If access or trespassing is denied, the existence of plantations may increase the distance that local households must walk to access natural forests or agricultural plots. Bleyer et al. (2016) reported plantations to decrease the availability of land, NWFPs and firewood. The study on socio-economic impacts of plantations was conducted in Niassa, Mozambique, and is part of the same research as this paper, thus providing the background information for focusing on the trade-offs between potential environmental impacts of plantations on local people.

When plantations are established, impacts on local people depend on e.g. decisions concerning plantation locations, compensation measures and the involvement of local people in the decision-making process. Zoomers and Otsuki (2017) argue that the livelihood research framework should be integrated into plantation planning and policy intervention to make the investments more relevant to local development. To avoid social conflicts, plantation establishment should be matched with other land-use needs. Conflict risk can generally be decreased by examining local peoples’ preferences regarding plantation location and various ecosystem services and by considering their views during the planning process. As an example, life in rural areas of Mozambique, one of the poorest countries in the world according to United Nations (UN) statistics, is highly dependent on subsistence farming and ecosystem services. The number of private large-scale agricultural and forest investments has significantly increased in the country during recent years (Nhantumbo et al., 2013), and this number is likely to further increase in upcoming years. The welfare impacts on local people are, so far, poorly examined.

This study investigates the preferences of the local people in Mozambique for water, firewood and land, applying the discrete choice experiment (DCE) that provides a tool for assessing beforehand the importance of various impacts of land-use changes. The results can be utilized in plantation planning processes to minimize conflicts between local communities and investors, to avoid negative livelihood impacts, and to design suitable social responsibility projects. DCE has also been used to examine the impacts of forest-related land-use changes in the sub-Saharan context, but examples are scarce and scattered. For example, Kassahun and Bredahl Jacobsen (2015) applied the DCE in Ethiopia in the context of watershed management. Their study identified incentives that motivated land users to participate in the management of private and communal lands in the Ethiopian highlands, but did not examine attributes of plantation forestry or specific ecosystem services. Balana et al. (2011) analysed landholder preferences for reward-based land management contracts in Kenya. They concluded that the size of the land area to be committed, length of the contract period, and granting or prohibiting rights to harvest environmental products from the committed land are key attributes to focus on when designing land management contracts in the area. Gelo and Koch (2012) investigated the impact of community plantation attributes on the welfare of participants of a potential community forestry programme in Ethiopia. Their study sites were selected based on high deforestation levels and increased demand for woody biomass. They found significant preference heterogeneity for plantation attributes, especially for Eucalyptus plantations and area enclosures. Dissanayake et al. (2015) examined preferences of rural Ethiopian communities for attributes of the REDD + contract, participation in which would limit the use of existing forests in the community. For local people, the level and division rules regarding contract payments between households and communities and possible restrictions on using grazing land were important, while restrictions on firewood collection did not significantly impact their choices. Preferences for contract characteristics were subject to regional variation.

The relative importance of plantation characteristics and the availability of ecosystem services can stem from differences in social groups within a community. When planning for socially sustainable projects, identifying groups vulnerable to changes is especially important. According to Hope (2006), improved water accessibility (water source) benefitted especially female members of the community. Kibugi et al. (2016) identified disparities in the user rights and landownership between genders, women often being more dependent on land for subsistence, but not always included in negotiations concerning land ownership and compensation. Tyynelä et al. (2002) and Vihervaara et al. (2012) also found differences in preferences for or impacts of plantations between gender or wealth groups. According to Bleyer et al. (2016), the poorest and single-mother households benefitted less from the existence of plantations, and households without alternative income sources to farmland were more likely to report negative impacts of plantations.

Our study makes three contributions to the literature. First, it provides better understanding of trade-offs associated to large-scale industrial forest plantations in sub-Saharan Africa and the relative importance of ecosystem services to local people. This information is also useful for planning sustainable new plantations. Second, as plantations may differ in terms of their impacts on ecosystem services and, along this line, on local people and their perceptions and experiences, the empirical data are analysed with econometric models that allow investigating preference heterogeneity. Third, our study contributes to growing literature on the application of DCE in developing countries, and reports experiences on empirical application in sub-Saharan Africa, with special focus on the case where no monetary attribute is present in the design.

The rest of our paper is structured as follows. Section 2 presents the context of our study, i.e. the trade-offs between plantations and ecosystem services, and the study area. Section 3 describes the theoretical framework, econometric models and empirical data. Section 4 presents the results, Section 5 discusses and Section 6 concludes.