1. Introduction
Until the last decade, the concept of a vicious circle, fostered by the World Bank [
1,
2] and related international agencies [
3], dominated the debate on poverty–environment linkages. This concept first appeared in the Brundtland Report launched by the World Commission on Environment and Development (WCED) [
4]. According to this report, “Many parts of the world are caught in a vicious downward spiral: poor people are forced to overuse environmental resources to survive on a daily basis, and their impoverishment of their environment further impoverishes them, making their survival more uncertain and difficult” ([
4]; p. 27). Contrary to the assumptions of WCED, later studies [
5,
6,
7] show that poverty–environment linkages do not conform to a simple downward-spiraling two-way relationship. According to these findings, existing social networks and institutions that govern access to natural resources mediate the relationships between poverty and environmental degradation and improvement. In addition, the popular portrayal of the vicious circle has treated poverty in uniform and generic ways and thereby caused the broader context behind this poverty to be under-represented. Given that the natural resource base constitutes an important source of income in rural developing countries, especially for the poorer groups in society, a better understanding of the poverty–environment relationship can lead to better policies that support the natural resource assets of the poor.
Poverty itself is contextualized differently across academic disciplines and regions of the world [
8]. It is now generally accepted that human poverty has many dimensions, and it is not just poverty of income or not having things necessary for material well-being. Human poverty also means the deprivation that people suffer throughout their lives which differs among nations. Furthermore, differences occur within developing countries when rural areas are compared to their urban counterparts. The result has been the emergence of different ways of approaching poverty–environmental linkages. On one hand, the state of the environment is increasingly a focus of development practitioners when determining the magnitude of poverty. On the other hand, a major concern for environmentalists is the role of poverty in resource degradation [
9]. While the former focuses on the poor state of the environment as a factor in pushing rural households that are dependent on natural resources into poverty, the latter concerns itself with the poor as agents of environmental degradation [
10]. Can the potential reinforcing interactions between environmental degradation and poverty provide explanations as to why this relationship is far from being linear? Additional empirical research is needed to further specify the relationship.
In addition, the poverty–environment vicious circle hypothesis suggests that economic growth is needed to break the poverty-environment downward spiral. Economic growth policies are expected to reduce poverty while providing incentives for investment in the land [
10]. However, a simple generalization of the relationship between poverty and environmental degradation is misleading [
9]. This is because economic growth is not always inclusive of the poor and policy instruments often fail to address inequality. On the other hand, previous studies have limited environmental degradation to soil erosion and deforestation at the expense of attention being paid to other environmentally degrading activities such as overgrazing, use of pesticides, etc. [
6]. Understanding the environmental entitlements or resource rights of farmers provides valuable information for understanding land-use decision making [
11,
12].
In Burkina Faso, approximately 70 percent of the population is rural and depends on farming and livestock for their livelihoods [
13]. The rural population depends on the natural resource base and adopts different resource management strategies that can either improve or degrade the environment. The study reported here examines the role of different rural wealth groups in contributing to environmental degradation in four rural communities in southern Burkina Faso. Given that many farmers in the Sahel manage complex crop and livestock portfolios [
8], expanding the focus to include other environmentally degrading activities aside from soil erosion and deforestation is invaluable for further understanding of the poverty–environment nexus in the Burkina Faso context. Previous Burkina Faso studies identify the cutting and selling of fuel wood [
13,
14], cotton cultivation [
15], and the conversion of forests to croplands [
16] as the main drivers of environmental degradation. However, studies on the relationship between individual wealth status and environmentally degrading activities are lacking.
For the purposes of this study, environmental degradation is defined as the deterioration of the environment through depletion of resources such as air, water and soil, leading to the destruction of ecosystems and the extinction of wildlife [
17]. This study focuses on the depletion and degradation of land and forest resources as examples of environmental degradation. In addition to the above cited activities, overgrazing resulting from livestock stocking densities that exceed the available fodder supply is also considered to be a cause of degradation of land and forest resources [
18]. Agro-pastoral systems dominate in the Sahel, and livestock numbers have increased significantly during the last 30 years in southwestern Burkina Faso [
19]. Such increases inevitably lead to increases in rangeland and fodder demand that is likely to affect the management of natural resources and especially forests. Understanding the livelihood activities of farming households characterized by different economic means will provide insight into the natural resource management problems implicit in poverty–environment linkages. These problems are considered to be largely related to agriculture [
20] and may be responsible for the depletion of both individual and common-pool resources.
Southern Burkina Faso offers greater opportunities for rain-fed agriculture, fuel wood supply, forest and tree products, fodder supply, etc., compared to the country’s central and northern regions, which suffer from periodic drought. The favorable climate and soil conditions attract migrants and agribusiness investors in search of arable land. Increased demand for land has caused land scarcity [
21] and resulted in higher competition among different land uses [
22], which not only threatens environmental sustainability [
16] but also causes deforestation [
23,
24]. Consequently, previous studies in Burkina Faso [
16,
23,
24] focus on population–environment rather than poverty–environment interactions. Although these studies are important for assessing land degradation, they address only some of the potential causes. This study tests the poverty–environment relationship through the following questions: (i) When households are categorized based on poverty and wealth, which groups are more responsible for environmental degradation? (ii) Does poverty constrain the adoption of land management practices that are considered to improve the land? Understanding the relationship between poverty, wealth, and natural resource management activities leading to environmental degradation is important for prescribing policy measures to mitigate these problems.
2. Materials and Methods
2.1. Description of Study Area
This paper is based on field research conducted in four adjacent community forest villages: Cassou, Vrassan, Dao, and Kou, all in the Ziro province, southern Burkina Faso (
Figure 1). These villages were chosen under the framework of the Building Biocarbon and Rural Development (BIODEV) project in West Africa. This project was financed by Finland’s Ministry of Foreign Affairs as an initiative to achieve developmental benefits by building biological and natural carbon resources through improved agroforestry and forest management practices. This area lies within the South-Sudanian climate zone, with annual precipitation of 800–1000 mm. Rains consisting of short intense storms fall over a single wet season lasting for approximately four months from June to September [
25]. During the hot season, the average daily temperature stands at 30 °C, with peaks of 40 °C as a result of hot dry air that blows from the Sahara Desert. The area is characterized by low relief and homogenous soil types including silt-clay cambisols, sandy lixisols, and loamy ferric luvisols [
26].
The average population density in the Ziro province was estimated at 28 persons/km² in 2006 [
27], but this figure is increasing due to rural–rural in-migration [
16]. The population consists of three main ethnic groups: Gourounsi (indigenous), Mossi (originating from the central plateau) and Fulani (originating from the north of Burkina Faso).
The farming system is dominated by crops grown under a discontinuous cover of scattered trees that constitutes the so-called parklands. Parklands are considered to be agroforestry systems, but their biodiversity depends on the original vegetation cover, the number and type of trees and shrubs spared during conversion to farmland, the needs of farmers, etc. [
20]. Subsistence production includes the cultivation of cereals (such as sorghum, sesame, maize and millet) and tubers (yam and sweet potatoes) and animal husbandry. In addition to the above, a more complex and lucrative production system exists that involves the extraction of fuel wood and non-timber forest products, the cultivation of cash crops (cotton and fruit-tree plantations) and ranching [
20]. The natural flora is dominated by perennial grasses such as
Andropogon gayanus Kunth,
A. ascinodis C.B. Clarke, and
Schizachyrium sanguineum (Retz.) Alston [
28]. Tree species commonly found in the parklands include
Vitellaria paradoxa C. F. Gaertn,
Parkia biglobosa (Jacq.) R. Br. ex G. Don., and
Tamarindus indica L., amongst others. Forests in the study villages are under one of two management regimes: protected and classified forest. The classified forests, or national parks, (25 percent) are strictly protected from livestock and farming activities, while the protected forests (
chantiers d’aménagements forestiers—CAF and
forêts villageoises) are subject to field expansion and managed by local communities in collaboration with the government [
14].
2.2. Sampling and Data Collection
The first step of field data collection consisted of constructing a participatory poverty assessment (PPA) based on local indicators developed during a focus group discussion (FGD) in each of the four villages. Although approximately 70 percent of the population in Burkina Faso is rural and supported by the informal economy, national poverty assessment is based on income criteria [
21]. Use of the money metric criterion [
29,
30] in such rural communities as in Burkina Faso is misleading and is not the most adequate and applicable means for assessing poverty. Such a criterion is more applicable in urban areas where it is possible to assess income and expenditures using money as the unit of assessment. This is not the case in rural areas where banks are not available for money to be deposited, and which are dominated by an informal economy where wealth is stored in the form of assets. In rural areas, the informal economy sustains the livelihoods of households through natural resource and land-based economic activities such as farming, logging, trade, etc. This rural informal sector is highly complex and often rooted in traditional resources and land rights [
31]. Therefore, the livelihood approach for assessing a household’s wealth in relation to its asset holdings has been widely applied in rural developing countries [
32]. In this approach, the real assets of a household are the unit of measurement, not money.
A total of five participants, including two women and three men from each of the three ethnic groups, participated in each FGD. The aim of the FGDs was to develop a poverty profile for the study area through a participatory exercise and to identify indicators of environmental degradation. The participants of the FGDs had to satisfy two conditions. Participants had to have lived in the community in or before 2003, which coincides with the start of the period used by the study for assessing deforestation. This time period is considered to be sufficiently long enough for the participants to know the level of well-being of other households. Participants were selected to represent a cross-section of the community in characteristics such as gender, ethnicity, wealth status, and neighborhood. Based on these criteria, participants were selected from among all the ‘sub-chiefs’ and women leaders of farmer management groups. This is because in Cassou, for example, the Mossi and Fulani ethnic groups live in separate quarters from the indigenous ethnic group (Gourounsi). Each of the ethnic groups has a leader, or sub-chief, who knows all the families in the village within that group. The women leaders were additionally selected because all the sub-chiefs were men.
Following Narayan et al. [
33], the participants of each FGD were guided during the discussion to: (i) list local indicators used to assess wealth status in the community; and (ii) describe the specifications of each local indicator and its corresponding wealth category (see
Table 1 for indicators). The resulting indicators differed slightly across the villages; therefore, the 20 participants of the four FGDs participated in a final meeting in Cassou to agree upon a common list (see the PPA below for more details). Participants also identified the following as environmentally degrading activities or indications of environmental degradation: field expansion (leading to deforestation), cotton cultivation, fuel wood exploitation, overgrazing, soil fertility loss, and local perceptions of tenure insecurity. Local participants identified tenure security as a factor playing a role in the adoption land management practices such as assisted natural regeneration. Tenure insecurity is likely to act as a disincentive for land management practices with long time horizons [
25] compared to the use of practices such as composting, which are more immediately effective [
34].
These outputs from the FGDs were reinforced with a literature review to locate corroborating scholarship on the activities identified by the FGDs as environmentally degrading in the region. These activities were found to be consistent with those identified in previous studies in Africa and Latin America [
6,
15,
35,
36]. A detailed questionnaire was then designed to collect both quantitative and qualitative data related to the environmentally degrading activities and which also addressed a specific set of land management practices (use of fallows, planting pits, composting, stone bunds, and live hedges) as well as household and farm characteristics (to be used for wealth categorization).
With the assistance of local youth leaders, 200 households from the study villages were randomly selected from a list of all households considered to fall within the different wealth status groups. Furthermore, 10 farms from each proposed wealth group were randomly selected for an additional farm survey, for a total of 30 farms. During the farm surveys, field sizes were estimated alongside other specific features of the farming systems such as fallows, evidence of assisted natural regeneration, live hedges, planting pits, stone bunds, use of compost, etc. The purpose of the farm survey was to confirm the interview questionnaire data.
2.3. Analytical Methods: Categorizing Households Based on Local Indicators Derived from Participatory Poverty Assessment (PPA)
As described above, participatory research methods were applied in which community members defined the wealth criteria based on local indicators [
6]. Twelve wealth status indicators and their descriptions (
Table 1) were identified during the FGDs. As a part of this process, participants selected a schema of wealth groups, with households categorized based on these indicators. A different set of wealth groups was identified in each study village, and three wealth groups were adopted for a common list. In Cassou, participants identified the groups: rich, fairly rich, poor, and poorest; in Dao and Kou: rich, fairly rich, and poor; and in Vrassan: rich and poor. To correct for these differences in income groups across locations, Ravnborg et al. [
37] use a mean value, which was adopted in our study as follows: (4 + 3 + 3 + 2)/4 = 3. Thus, the three income groups of non-poor, fairly poor and poorest were numbered 1, 2 and 3, respectively.
The next stage in this method was to transform the numbers to scores representing poverty levels. The qualitative rankings were quantified using the following equation adapted from Ravnborg et al. [
37] as follows:
where S = Well-being score; A = Income group of the household based on local perception of well-being indicators (
Table 1); and P = the total number of wealth groups. The result was multiplied by 100 to avoid operating with decimals, resulting in:
Level 1—[(1 − 1)/(3 − 1)] × 100 = 0, where 0 implies non-poor household
Level 2—[(2 − 1)/(3 − 1)] × 100 = 50, where 50 implies fairly poor household
Level 3—[(3 − 1)/(3 − 1)] × 100 = 100, where 100 implies poorest household
During the interviews, data were collected from each household based on the local indicators agreed upon in the common list (
Table 1). In addition, data were collected on household resource management strategies self-reported by farmers, based on their perceptions, to capture activities considered environmentally degrading. Each household was assigned a corresponding score for all 12 indicators, which was later averaged by the first author to classify the household into its corresponding wealth group. The threshold values were then calculated to define the range of each wealth group based on the MEAN of all 200 household as follows: 0 to 50 (non-poor), 50 to 75 (fairly poor) and 75 to 100 for the poorest. Within these threshold values, it was possible to assign each individual mean to a category.
2.4. Study Variables Indicating Environmentally Degrading Activities
The collected activities-related data were used to create variables of two kinds: numeric measurement-based variables and categorical variables (often self-reported). The former variables include mean annual deforestation (2003–2013), cotton cultivation, cutting and selling of fuel wood, and overgrazing based on cattle numbers. Categorical data were gathered through farmers’ self-reported assessments on the following: overgrazing, soil fertility loss, and tenure security as an incentive for assisting natural regeneration. Overgrazing occurs in both categories because both numerical data and self-reported assessments of respondents were recorded. Several variables based on numeric measurement-based data were also converted to additional categorical variables for analyses (see below).
2.4.1. Deforestation
Through the interviews, data were collected on changes in farm area between 2003 and 2013. Farm size included areas under shifting cultivation with fallows. A recent study in Burkina Faso finds that farmers without fallows are more likely to expand their fields, thereby causing deforestation [
38]. The lack of strict monitoring of community forest areas exposes them to encroachment. As such, field expansion into protected forest areas has been identified as the dominant proximate driver of deforestation in Burkina Faso [
39,
40]. In this study, the cultivation of fallows was not considered to represent deforestation because they represent potential areas for reuse when the rotation cycle is completed or as need arises. Therefore, deforestation in the current study focuses on the expansion of fields into protected forest.
The difference in farm area was calculated as the change in farm area, excluding fallow, between 2013 and 2003. The result was divided by 10 to get the mean annual change value for the 10-year-period. A categorical variable was created in addition to the continuous variable. Households that did not experience a change in farm area were assigned the value 0, while those that cleared forests were assigned the value 1. Jones et al. [
41], in an assessment of deforestation driven by farming systems, apply a mean annual area of forest cleared using this methodology. For additional details, see Etongo et al. [
38]. Land is not for sale due to customary rules that prevail in these communities, thereby reducing the options for farm expansion.
2.4.2. Cotton Cultivation
Aside from the role of cotton as a driver of deforestation in Burkina Faso, where a threefold increase in cultivated area (ha) occurred between 1992 and 2007 [
15], the use of pesticides also constitutes a threat to the environment [
42]. An increase in the annual rate of pesticide consumption over the last two decades is attributed to the treatment of cotton fields [
43]. The term pesticide covers a wide range of compounds including insecticides, fungicides, herbicides, etc. A pilot study in Burkina Faso on agricultural pesticide poisoning indicates its effect on the environment, livestock and human health [
44]. This study employs both a continuous variable of cotton produced (in kg) and a categorical variable for presence or absence of cotton cultivation at the farm level.
2.4.3. Cutting and Selling of Fuel Wood
Fuel wood is a major source of household energy in Burkina Faso and is collected from fields, fallows, forests, and plantations. The traditional measuring unit is a cart-driven system called a
charet. A
charet full of fuel wood is estimated at three m
3. The estimated monthly fuel wood consumption per household in 2013 (for both subsistence use and sale) was recorded in
charets and converted to cubic meters. A recent study in Burkina Faso estimates the average rural household’s daily fuelwood consumption to be 0.04 m
3 [
45], which is roughly the same as estimated in Kenya [
46]. Based on this daily estimate, monthly and annual consumption should stand at 1.24 m
3 and 14.88 m
3, respectively. These estimates provide a guide on household fuel consumption. In addition, information was collected on quantity of fuel wood sales and places of collection. This study employs both a continuous variable of reported monthly fuel wood consumption and a derived categorical variable which assumes fuel wood sales when reported monthly consumption was greater than 1.24 m
3.
2.4.4. Overgrazing
Average cattle herd sizes were collected for the last five years (2009–2013) to assess overgrazing in rangelands and farmlands. Studies on total livestock units (TLUs) in relation to stocking density consider overgrazing to occur when demand for fodder exceeds supply [
18,
47]. Niemeijer and Mazzucato [
47] study total livestock units (TLUs) and arrive at a similar conclusion that overgrazing occurs when stocking density increases without an increase in fodder. Studies in the West African Sahel [
19] and eastern Mediterranean [
18] find that increases in livestock numbers are not consistent with fodder supply, thereby causing overgrazing. Furthermore, another study in Peru reports losses in range species as a consequence of high stocking rates, drought, and the absence of fallow areas [
7]. In this study, the number of cattle was used as a continuous variable indicating overgrazing. In addition, a categorical variable was created using the mean value of 9.05 calculated among the 200 respondents, with two categories formulated as follows: households with ≤9 cattle and households with ≥10 cattle. Thus, owning more than the mean number of cattle was defined as indicative of overgrazing. To reinforce the above analysis, farmers also self-reported practices of overgrazing on their own farms in terms of environmental degradation as either low, moderate, and high.
2.4.5. Farmers’ Assessment of Tenure Insecurity on FMNR
The role of indigenous tree species in remediating land degradation through farmer-managed natural regeneration (FMNR) is widely acknowledged in the Sahel [
48,
49]. FMNR fosters tree ownership and land tenure security for farmers. Also known as assisted natural regeneration, FMNR is the protection of indigenous tree species in the Sahel so that they can regenerate naturally to maturity [
49]. It provides environmental benefits such as restoration of tree cover, increased biodiversity, climate change adaptation and mitigation, etc. [
50]. In the interviews, farmers were asked if they participate in assisted natural regeneration. However, due to the sensitive nature of this line of questioning (as indigenous species are protected by law), this question was asked in the context of other questions to determine farmers’ self-reported assessments of the state of natural regeneration of indigenous trees on their farms, who owns the trees, and how tenure security influences this practice. For this study, farmers’ assessments of the role of tenure security for FMNR was used because these data were assessed to be the most accurate proxies for determining FMNR practice due to the sensitivity of the issue. Additionally, two studies in the Sahel indicated the importance of FMNR as a land management practice that improve tenure security and also tree cover on farms, thereby reducing environmental degradation [
48,
49].
2.4.6. Farmers’ Assessment of Soil Fertility Loss
A study in Rwanda finds that local perceptions of soil fertility align well with soil fertility measured using local indicators such as crop yield, soil softness, indicator plants, and soil color [
51]. Karltun et al. [
52] also find consensus between farmers’ indigenous knowledge and scientifically validated indicators of soil fertility. Using locally reported soil fertility thus is a viable alternative to direct or indirect measurement. Farmers’ perceptions of soil fertility loss were adopted for this study and assessed at three levels: high, moderate and low. Three levels of soil fertility loss were adopted in the current study because this approach was found in Ethiopia to be effective when using perception-based methods [
36].
2.5. Statistical Analysis
In order to compare the indicators of environmental degradation across the households in different wealth categories, we use descriptive statistics and non-parametric tests to analyze the data collected. Specifically, we applied Chi square to check the relationships between categorical variables. We use one-way ANOVA to compare the variation within the continuous variables across a single factor and determine whether significant differences existed among their means [
53].
To run the ANOVA, Levene’s test was used as a prior test for homogeneity of variance within the continuous variables. When a significant difference was found in the one-way ANOVA test at the 5 percent level, the Scheffe pairwise multiple comparison test was conducted for cases exhibiting homogenous variance [
54]. For variables that exhibited non-homogenous variance, a Welch test was performed to correct the violation; then, Dunnett’s C pairwise multiple comparison test was applied [
55]. The multiple comparison tests highlighted where mean differences varied between pairs of wealth categories for the selected indicators.
4. Limitations
This study faces several limitations. First, as part of the process of grouping households into different wealth status categories, one of the roles of the FGDs was to identify every household in each category. With a total of 900 households in the four villages, having detailed information for all households was challenging. To overcome this limitation, the twenty participants from the four villages agreed on a common list of local indicators to use in categorizing the households. The researchers then collected interview data from households based on these indicators and later classified households into different categories of wealth. Thus, the PPA approach could only be implemented in a limited way. However, the study maintained the key aspect of participant input into how poverty was defined.
In addition to limitations arising from the application of the PPA, the use of PPA itself introduced a key limitation into the study’s analytical methods and findings. Both the number of cattle and the amount of land held by a household were used as independent variables in the categorization of households into the wealth groups. The use of these variables in wealth categorization is a direct reflection of local definitions of household wealth and a function of the PPA process. However, these variables also play a key role in the analysis of household engagement in environmentally degrading activities. The study treated number of cattle as a dependent variable to assess overgrazing. Furthermore, several of the environmental degradation dependent variables logically correlate with farm size: annual mean deforestation, as measured in hectares, likely reflects farm size, as does cotton production in kilograms. We nonetheless consider our findings to have merit, precisely because they highlight the role of wealth, as defined locally, as a driving force in environmentally degrading activities at this research site.
Next, self-reported assessments by farmers can be biased, especially if the issue under investigation is considered sensitive. Such data can only be gathered indirectly because a direct question might be too sensitive and can affect the engagement of respondents. Because parkland tree species are protected by law, farmers are likely not to provide information if such species are fast degrading on their farmlands. Also, unequal and small sample sizes might not effectively reveal differences between groups. Finally, applying a unified list of indicators across multiple villages requires careful consideration of potential differences in the socio-cultural settings. In our case, it was easier because these villages are close to each other and share ethnicity and culture.
5. Conclusions
This case study of the poverty–environment nexus not only reveals that the relationship between wealth or poverty status and environmentally degrading activities is site specific in general, as well as activity specific; it also provides insights into resource management practices across different wealth groups at this site. The current study reveals that non-poor and fairly poor farmers engaged more often in environmentally degrading activities compared to the poorest farmers. Approximately 93 percent of non-poor and 22 percent of fairly-poor farmers experienced high rates of overgrazing, while none of the poorest farmers experienced high rates of overgrazing. This is because the non-poor and fairly-poor farmers possess the available resources to own large herds of cattle. This study in fact categorized households into wealth status groups in part based upon number of cattle owned. Equipped with capital, the non-poor and fairly-poor farmers engaged more effectively in the cutting and selling of fuel wood and cotton cultivation than did the poorest farmers. Based on farmers’ self-reported assessment of the role of their tenure security in relation to assisting natural regeneration of indigenous tree species, tenure insecurity constitutes a major constraint on FMNR. FMNR is known to reduce environmental degradation, and due to their perceived inability to participate in this practice, the non-poor contribute more to environmental degradation. Parklands in the Sahel have existed for centuries because of the regeneration of indigenous tree species that provide multiple livelihood and environmental benefits.
Furthermore, deforestation through field expansion proved to be significant for the fairly-poor and non-poor households. Cotton cultivation, which can motivate field expansion and lead to heavy pesticide use, is capital intensive. The lack of available resources among the poorest farmer limits their effective participation in this practice. Other activities considered environmentally degrading, such as the cutting and selling of fuel wood, are dominated by the fairly poor and non-poor farmers.
On the other hand, self-reported assessment of soil fertility loss was highest among the poorest farmers. In addition, the adoption rates of land management practices considered to improve the land (and help ameliorate soil fertility loss) were relatively low for the poorest households. Some of these land management practices, e.g. planting pits, use of compost and stone bunds, are both labor and capital intensive, which may explain the low rates of adoption among the poorest farmers.
Therefore, the results of our study indicate that the non-poor and fairly-poor farmers contribute toward environmentally degrading activities relatively more, while poverty constrains the adoption of sustainable land management practices for the poorest farmers. Further research is needed in the following areas: (i) a comparison of natural resource management strategies for in-migrant farmers versus indigenous farmers; (ii) the relationship among ethnicity, poverty, and environmental degradation; (iii) the role of tenure security, land quality, and land fragmentation on sustainable land management; and (iv) the factors influencing FMNR in Burkina Faso.