Non-Timber Forest Products and Community Well-Being: The Impact of a Landscape Restoration Programme in Maradi Region, Niger

Abstract

The utilisation of non-timber forest products (NTFPs) represents a key strategy for coping with food insecurity in rural areas worldwide, often resulting in their gradual depletion and extinction. One of the objectives of landscape restoration programmes is to restore depleted and conserve scarce NTFPs, as well as to preserve their various ecosystem services. However, the impact of these programmes on the well-being of local communities through their access to NTFPs remains understudied. The study focuses on the arid areas of the southern and central parts of the Maradi region in Niger in the Sahel, analysing how the landscape restoration programme contributes to improved access to NTFPs and their subsequent commercialisation to increase household incomes. The probit model with propensity score matching (PSM) reveals, on data from 379 households, that socioeconomic factors such as age, male gender, education level, and number of children, as well as access to NTFPs through donations, increase the chance of being selected as the programme beneficiary. On the other hand, ownership of goats, practising free collection, or purchasing NTFPs were typical for non-beneficiaries. The PSM analysis indicates that beneficiaries sell 11.81% more NTFPs on the market (p < 0.05). Furthermore, beneficiaries earn an average of 7297.40 CFA francs from forest products, compared to 3281.37 CFA francs for non-beneficiaries (p < 0.01). To enhance the impact of the programme, policymakers should prioritise outreach to underrepresented households and incorporate livestock management strategies. Strengthening local markets, storage facilities, and processing infrastructure can boost economic returns. Additionally, combining the conservation of NTFP-producing species with sustainable income activities can support both environmental and community resilience.

1. Introduction

Non-timber forest products (NTFPs) are forest products other than timber, such as fruits, nuts, resins, medicinal plants, and fibres, which play a vital role in meeting household needs for food security, income generation, health, and cultural heritage, particularly in rural regions worldwide [1]. In the Sahel, where climate change intensifies land degradation and undermines agricultural stability, NTFPs have become essential for sustaining household resilience [2,3]. Such situation is typical for Niger, a country largely reliant on rain-fed agriculture, rural populations face compounded challenges such as erratic rainfall, rapid population growth, and persistent insecurity [4]. These conditions have deepened food insecurity, prompting communities to diversify their livelihoods [5]. NTFPs offer a crucial buffer, particularly during periods of scarcity, by providing supplementary food and income [6,7]. For instance, a recent study estimates that NTFPs contribute up to 46% of household income in Sahelian communities, with women and poorer households deriving an even higher share [8]. The significance of this contribution to local economies is evident, regardless the forest products are consumed directly or sold in raw or processed form [2,9,10].

However, the ecological sustainability of NTFPs harvesting is increasingly threatened by rising global demand, demographic pressures, and unsustainable harvesting practices [11,12]. In the Sahel, these pressures are compounded by recurrent droughts, weak governance structures, and limited economic alternatives [13]. Overdependence on key species such as Vitellaria paradoxa, Parkia biglobosa, and Adansonia digitata has further contributed to land degradation and biodiversity loss [14].

Various landscape restoration initiatives have been implemented globally, integrating reforestation, agroforestry, and community-based governance approaches [15,16]. International frameworks such as the Bonn Challenge and the UN Decade on Ecosystem Restoration are driving efforts toward large-scale land rehabilitation. In the Sahel region, particularly in Niger, integrated landscape restoration (ILR) has emerged as a holistic strategy to balance ecological sustainability with local development needs. Through interventions like assisted natural regeneration, the distribution of native species seedlings, and community-led management, ILR seeks to restore degraded ecosystems [17,18]. These practices not only rehabilitate ecological functions but also support the regeneration of important non-timber forest products (NTFP) species such as Balanites aegyptiaca, Sclerocarya birrea, and Ziziphus mauritiana, contributing to improved soil fertility, enhanced biodiversity, and increased resource availability [19].

In Niger, the ILR led by the World Food Programme (WFP) in the south-central Maradi region illustrates this approach. Indeed, since 2018, the WFP has been conducting land restoration initiatives with the aim of rehabilitating degraded land and strengthening food security [20]. The strategy is to incorporate agroforestry practices to improve soil fertility, increase agricultural productivity, and increase the availability of alternative trees that generate NTFPs [21,22]. A key aspect of these initiatives is the empowerment of local communities through a multifaceted support system, including the provision of Moringa species seedlings, training in nursery management, incentives to engage in sustainable land management practices, and access to market information [23]. This link between ecological restoration and food security highlights the strategic role of NTFPs [24,25].

Several studies have shown that restoration can enhance carbon sequestration, biodiversity conservation, and the resilience of farming systems if vegetation cover types are chosen wisely and appropriate land management is ensured [26,27,28,29]. Given these benefits, it is expected that households involved in integrated landscape restoration will experience greater improvements in ecological and livelihood outcomes compared to those without such interventions [30,31]. Effective community involvement remains a key success factor, although often hampered by various structural obstacles [29]. Despite the boom in landscape restoration initiatives in the Sahel, research on their impact on the well-being of local communities remains limited, particularly in the Sahel region, one of the world’s most vulnerable [32]. This gap hampers understanding of the social and economic implications of territorial resilience approaches, which are crucial to reconciling ecological sustainability and human development. NTFPs, although important in socio-economic terms, often remain marginalised in restoration policies, with little supervision of informal exploitation.

Against this background, the landscape restoration revives the fauna and flora that enhance the availability of NTFP to the communities. The present study aims to assess the impact of the ILR on community welfare in south-central Maradi. It provides answers to (i) what are the determinants that make the household be selected? (ii) What is the impact of the integrated landscape restoration on the availability of NTFP to the benefited households? This information provides empirical data useful for the development of policies integrating both ecological sustainability and social equity.

2. Theoretical Framework

The Collective Action Theory (CAT) offers a valuable framework for comprehending participation in land restoration programmes (see Figure 1). This theory posits that individuals or groups engage in collaborative endeavours to achieve shared objectives, particularly when the advantages of collective action exceed the disadvantages [33]. The phenomenon of collective action is influenced by factors such as resource dependence, group incentives, and institutional arrangements that facilitate cooperation [34]. In the context of land restoration, communities engage in collective efforts when they perceive long-term economic and socio-ecological benefits, such as improved soil fertility, increased agricultural productivity, and access to forest resources [35].

The welfare needs of local communities have been identified as a key driver of participation in land restoration programmes, particularly in resource-dependent regions such as the Sahel, where land degradation poses a threat to livelihoods. Individuals in these regions are more likely to be selected in restoration activities to secure long-term access to forest products, grazing lands, and improved agricultural conditions [34]. CAT posits that when individuals perceive tangible benefits, such as increased income from NTFPs or enhanced food security, they are more motivated to cooperate in communal restoration initiatives [33]. Moreover, external incentives, including financial support, training, and improved governance structures, further encourage participation by reducing transaction costs and increasing perceived gains [35].

Participation in land restoration programmes is also influenced by socioeconomic heterogeneity, household assets, and modes of access to NTFPs. Variations in social status and resource ownership influence individuals’ eligibility to be selected and engage in restoration efforts. Households with more financial resources and secure land tenure may participate actively, while those with fewer assets may face opportunity costs [36]. Furthermore, the type of NTFPs access, whether through free collection, purchase, or controlled distribution, has been found to modify participation by determining how much individuals rely on forest resources and whether they see restoration as a means of securing future benefits [37].

Socioeconomic characteristics have been shown to influence participation patterns, with gender, age, marital status, ethnicity, place of origin, lifestyle, and education level having a significant impact on individuals’ eligibility to participate [35,38]. For instance, older individuals with strong community ties may be more likely to engage in restoration efforts, while younger or migrant populations may have weaker incentives. Household assets such as livestock, including cattle, sheep, goats, and poultry, also play a role, as livestock owners may prioritise grazing lands over land restoration efforts [39,40]. Finally, the type of NTFPs access impacts engagement; households that freely collect forest products may have weaker incentives to restore degraded areas, whereas those who purchase NTFPs may support restoration efforts to ensure future availability [41,42].

3. Methodology

3.1. Study Site Characteristics

Niger is a landlocked country in West Africa, situated between latitudes 11° N and 23° N and longitudes 0° E and 16° E. It borders Algeria and Libya to the north, Chad to the east, Nigeria and Benin to the south, and Burkina Faso and Mali to the west [43]. With an estimated surface area of 1,267,000 km², it is the largest country in West Africa in terms of landmass. The landscape is predominantly arid and semi-arid, with over 80% of the territory covered by the Sahara Desert [44]. The Niger River, the country’s main source of water, flows through the southwestern part, supporting local agriculture and livelihoods [45]. The total population is over twenty-five million in 2023 and is expected to continue rising, as Niger has one of the world’s fastest-growing populations, mainly in rural areas [43,46]. In general, Niger’s forests are very scarce, covering less than 2% of its surface, and face severe soil degradation due to desertification, deforestation, and unsustainable agricultural practices [46]. Climate change, population pressure, and overgrazing further threaten food security and livelihoods. In response to these challenges, conservation efforts, including the Great Green Wall initiative, have been implemented to restore degraded land and combat desertification [47].

The Maradi region, particularly its central-southern area, has been identified as a priority intervention zone by the WFP due to its heightened vulnerability to food and nutritional insecurity [48]. This vulnerability is driven by a combination of climatic, economic, and demographic pressures that render local communities heavily reliant on humanitarian assistance and resilience-building initiatives [49]. While these challenges are acute in Maradi, they reflect broader structural issues affecting the entire Sahelian zone of Niger. The Nigerien Sahel is marked by extreme climatic variability, with frequent droughts and floods disrupting agropastoral systems and deepening the socio-economic fragility of rural households [50]. In response, the WFP has implemented an integrated household resilience programme in the region. Inclusion selection criteria included proximity to other agricultural development initiatives, such as those supported by the International Fund for Agricultural Development (IFAD), and the presence of intermediate markets. These conditions facilitate linkages between agricultural production zones in the south and pastoral areas in the north. Studying the Sahelian zone offers critical insights into how environmental and socio-economic factors shape natural resource management, food security, and the patterns of economic exchange across Niger’s diverse agroclimatic regions.

This study was carried out in two agroclimatic zones following a North–South rainfall gradient in the Maradi region, which is located in the south-central part of Niger, at around 13.5° north latitude and 7.1° east longitude [44], namely the North-Sahelian zone and the South-Sahelian zone (Figure 2). The South-Sahelian zone is characterised by a semi-arid climate, with annual rainfall varying between 500 and 700 mm [51]. The vegetation is less dense than in the northern Sahelian zone, with a wooded and shrubby savanna with predominantly sandy-silty soils, with good water retention capacity [52]. The North-Sahelian zone is the driest part of the Maradi region, marking a transition from the Sahara. It has a semi-arid to arid climate, with annual rainfall of between 300 and 400 mm [53]. The vegetation is mainly composed of shrub and tree steppe, with relatively sparse ground cover, and the soils are predominantly sandy to loamy, with low water retention capacity [52].

However, despite serious environmental problems [47], the Maradi region continues to be the main agricultural centre of Niger, often referred to as the “granary of Niger” due to its fertile soils. Local forest resources, notably baobab, acacia, and neem, are being rapidly depleted by deforestation and inadequate land management, although efforts are underway to promote reforestation and sustainable land tenure practices. The presence of small-scale industries, such as leather goods, handicrafts, and food processing, underlines the importance of the Maradi region as a commercial hub within Niger [46].

3.2. Sampling Procedure

A preliminary survey was conducted to identify areas rich in NTFPs through field visits. These visits were supplemented by consultations with development agents from the environmental department, representatives from non-governmental organisations (NGOs) partnering with the WFP, and village chiefs. These exchanges facilitated engagement with a range of stakeholders, including agricultural district chiefs, communal environmental department heads, and local authorities, to obtain data on the number of households needed.

The sample size (n) was determined using Daniel’s [54] formula:

$$n=t^{2}N/t^2+{\left(2e\right)}^2(N-1)$$

(1)

where n = sample size, t = confidence level, N = total population, e = accepted margin of error (set at 5%). Applying this formula, the final study sample consisted of 379 households.

A stratified sampling approach was employed at two levels: firstly, regarding the selection of villages, and secondly, regarding the selection of respondents. Villages were chosen based on their proximity to specific land use units, including pastures, restored sites, rock outcrops, continuously cultivated land, and fields with shrubs and trees. The selection process was guided by a land use map that was created from Sentinel-2A satellite imagery. The selection process was further based on the intensity of NTFPs exploitation, the sociolinguistic composition of the population to include all (major) local ethnicities, such as Peulh, Haoussa, and Touareg, prior exposure to restoration and development activities under WFP initiatives, and site accessibility as assessed through field observations and pre-surveys. The selection of respondents was conducted in accordance with a set of predetermined inclusion criteria. Participants were selected from the beneficiaries of the Resilience Programme. A subset of respondents was drawn from participants of the WFP resilience programme, who met at least one of the following criteria: physical disability (e.g., loss of an able-bodied arm) or living in extreme poverty. The remaining respondents were selected randomly, with the additional requirement of meeting specific socio-economic vulnerability criteria. The remaining participants were selected randomly but based on similar socio-economic vulnerability criteria. This methodological approach ensured a representative sample while accounting for socio-economic and ecological variations within the study area.

3.3. Data Collection

Data collection for this study was carried out in September 2023 and July 2024. A structured questionnaire integrated with the KoBoCollect system (version 1.29.3) was used to facilitate data collection using tablets. Through these questionnaires, respondents (179 beneficiaries and 200 non-beneficiaries) provided detailed inventories of the NTFP species they used. The interview technique proves to be the most appropriate in descriptive ethnobotany and favours the collection of data [9]. In addition, they reported on the quantities harvested, the proportions intended for personal consumption and for sale, their awareness of food crises occurring between two cropping seasons, the factors influencing these crises, the strategies adopted to mitigate their impact, and their socio-economic characteristics.

3.4. Data Analysis

To examine differences between beneficiaries and non-beneficiaries, a two-sample z-test for proportions (and means where appropriate) was first applied to compare their socioeconomic characteristics and access to NTFPs. This initial comparison helped identify significant baseline differences between the two groups. Subsequently, Propensity Score Matching (PSM) was employed to control for selection bias and to estimate the average treatment effect of program participation. PSM operates by matching beneficiary households (treatment group) with comparable non-beneficiary households (control group) based on observable characteristics. This methodological approach enables the estimation of mean differences in key welfare indicators, including the quantity of NTFPs taken to market, and income generated from NTFPs between the two groups. The study addresses the counterfactual question: “What would be the impact of program participation on household welfare if participant households had not taken part in the program?” The Average Treatment Effect on the Treated (ATT) is utilised to quantify this impact, as it measures the observed outcomes by comparing them to counterfactual estimates of what these outcomes would have been in the absence of participation [55]. The ATT is a method that captures the causal effect of programme participation on household welfare (market orientation of NTPFs collection and household dependency on NTPFs commercial collection), thereby enabling a rigorous assessment of its impact on livelihoods and economic well-being.

$$ATT=\mathrm{ᴛ}{\displaystyle \frac{PSM}{ATT}}=E\{E[Y^1|M=1,p(X)]-E[Y^0|\mathrm{M}=0,\mathsf{p}(\mathrm{X})] |\mathrm{M}=1\}$$

(2)

where $ATT$ measures the household that participated in the programme as the treatment on the outcome variables assessing household welfare (

Table 1. Description of the variables used in the propensity score matching (PSM) model (n = 379).

Variable		Description	Mean	Standard Deviation	Minimum	Maximum
Outcome variable
	Market orientation of collected NTFPs	NTFPs sold (%)	27.52	44.88	15.66	61.11
	Household dependency on NTFPs	Total value of NTFPs collected (CFA)	5195.98	9915.34	799.44	76,500.00
Treatment variable
	Programme beneficiary	Beneficiary = 1, Otherwise = 0	0.44	0.49	0	1
Explanatory variables
Household head characteristics
	Gender	Male = 1, Female = 0	0.67	0.46	0	1
	Age	Years	43.0	15.5	18	99
	Marital status	Married = 1, Otherwise = 0	0.90	0.29	0	1
	Origin	Indigenous = 1, Migrant = 0	0.91	0.27	0	1
	Ethnicity	Hausa = 1, Others = 0	0.90	0.29	0	1
	Livelihood style	Sedentary = 1, Nomad = 0	0.66	0.47	0	1
	Qur’anic education	Qur’anic education = 1, Others = 0	0.33	0.47	0	1
	Education	Primary = 1, Otherwise formal education = 0	0.18	0.38	0	1
Household characteristics
	Household size	Number of people living in the household	10.71	5.16	1	30
	Children	Number of children < 15 years living in the household	3.11	2.09	0	8
Livestock ownership
	Cattle	Number	0.46	1.37	0	20
	Goat	Number	1.97	2.75	0	20
	Sheep	Number	0.79	2.10	0	28
	Poultry	Number	2.30	5.04	0	30
NTFPs access type
	Free (no restriction)	Yes = 1	0.54	0.49	0	1
	Donation	Yes = 1	0.07	0.27	0	1
	Purchase	Yes = 1	0.17	0.37	0	1
	Legacy	Yes = 1	0.25	0.43	0	1

). $Y^1$ and $Y^0$ represent the possible outcomes in the treatment and control conditions, respectively, $M=1$ is a binary treatment indicator, where represents individuals who received the treatment, i.e., beneficiary, and $M=0$ represents a non-beneficiary household, $p\left(X\right)$ is the probability of a household taking part in the programme given the covariates $X$ and $E[Y^1|M=1,p(X)]$ is the expected outcome for a treated household (welfare), given the propensity score. $E[Y^0|M=0,p(X)]$ is the expected outcome for untreated farming households (non-beneficiary households), given the propensity score. This equation formally defines the Average Treatment Effect on the Treated (ATT) within the Propensity Score Matching (PSM) framework. It estimates the difference between the expected outcome for the treated group (M = 1) and the expected counterfactual outcome, representing what would have occurred had they not participated. The counterfactual is derived from a matched control group with comparable observable characteristics, ensuring that the analysis accounts for selection bias. By matching beneficiary and non-beneficiary farming households with similar propensity scores (Figure 3 and

Table A1. Post-matching covariance balance.

Variable	Sample	Mean		%Bias	%Reduce |Bias|	t-Test (t)	p > |t|
		Treated	Control
Household head characteristics
Gender	Untreated	0.67	0.67	−0.3		−0.04	0.972
	Matched	0.67	0.61	12.4	−4278.30	1.46	0.145
Age	Untreated	45.10	41.55	23.2		2.91	0.004
	Matched	45.10	42.04	20.0	13.80	2.39	0.017
Origin	Untreated	0.90	0.93	−10.2		−1.28	0.200
	Matched	0.90	0.89	3.7	63.30	0.41	0.684
Ethnicity	Untreated	0.93	0.88	16.7		2.08	0.038
	Matched	0.93	0.91	5.4	67.50	0.71	0.477
Livelihood style	Untreated	1.27	1.37	−20.1		−2.52	0.012
	Matched	1.27	1.23	10.2	49.30	1.31	0.189
Qur’anic education	Untreated	0.36	0.22	31.7		4.00	0.000
	Matched	0.36	0.37	−2.6	91.70	−0.29	0.770
Education	Untreated	0.19	0.10	25.7		3.26	0.001
	Matched	0.19	0.19	0.20	99.10	0.02	0.981
Household characteristics
Household size	Untreated	10.71	9.85	17.0		2.14	0.033
	Matched	10.71	10.05	13.3	21.80	1.56	0.119
Children	Untreated	3.11	2.76	16.8		1.00	0.316
	Matched	3.11	3.19	−3.9	76.60	0.19	0.849
Livestock characteristics
Cattle	Untreated	0.47	0.32	13.7		1.77	0.077
	Matched	0.47	0.41	5.5	60.10	0.58	0.564
Goat	Untreated	1.97	2.01	−0.9		−0.11	0.316
	Matched	1.97	1.94	1.1	−21.60	0.13	0.849
Sheep	Untreated	0.79	2.00	−0.9		−0.11	0.911
	Matched	0.79	1.94	1.1	−21.60	0.13	0.898
Poultry	Untreated	2.30	1.95	7.9		1.00	0.316
	Matched	2.30	2.23	1.7	78.70	0.19	0.849
NTFPs access type
Free (no restriction)	Untreated	0.54	0.76	−47.9		−6.05	0.000
	Matched	0.54	0.54	−0.1	99.70	−0.02	0.987
Donation	Untreated	0.07	0.03	21.1		2.69	0.007
	Matched	0.07	0.06	5.7	72.90	0.60	0.548
Purchase	Untreated	0.17	0.12	13,1		1.65	0.100
	Matched	0.17	0.17	4.4	66.70	0.50	0.617
Legacy	Untreated	0.25	0.16	22.0		2.78	0.006
	Matched	0.25	0.24	2.3	89.50	0.26	0.796

), this approach enhances the validity of causal inferences regarding the overall impact of the programme. To test the robustness of the propensity score matching (PSM) results, the study employed multiple matching algorithms, nearest neighbour (1), kernel, radius (0.5), and calliper (0.05), ensuring that the estimated treatment effects are not overly dependent on any single method.

4. Results

4.1. Characteristics of Beneficiary and Non-Beneficiary Households

Table 1 presents a descriptive summary of the variables incorporated in the PSM model, which was employed in this study to assess the impact of participation in a Land restoration programme on the quantity of NTFPs taken to market and the total financial value of NTFPs collected by the households.

The above table presents the outcome, treatment, and independent variables. In total, 27% of the NTFPs are taken to the market, with a 44% standard deviation, which indicates high variation in commercialization among the respondents. The mean NTFP value collected is 5196 CFA (XOF (West African CFA franc), with a standard deviation of 9915 CFA, indicating substantial variability among households. The maximum recorded income is 76,500 CFA, suggesting that certain households benefit considerably from NTFP collection. A mean value of 0.44 of the ILR participation implies that 44% of the households surveyed participated, while 56% did not.

The sociodemographic characteristics of the sample indicate that two-thirds of household heads (67%) are male, with an average age of 43 years, suggesting that most are within their economically active years. A significant proportion (90%) of household heads are married, which may reflect a higher level of household responsibility. The Hausa ethnic group dominates the sample, constituting 90% of respondents. Most of the participants are sedentary, meaning they are staying in one place without moving from one place to another, with 33% having received Qur’anic education, 18% having completed primary education. The average household size is approximately 11 members, with an average of three children, though household compositions vary widely. Regarding household assets, livestock ownership exhibits substantial variation. Poultry is the most owned livestock, followed by goats, while cattle are the least owned.

Furthermore, the findings provide insight into how households access NTFPs. The most prevalent access method is free access, utilised by more than half (54%) of households, followed by legacy access (25%) and purchase (17%), while donations (7%) represent the least common form of access.

Table 2. Comparison of the beneficiaries and non-beneficiaries of the land restoration programme.

Variable		Beneficiaries (n = 179)		Non-Beneficiaries (n = 200)		z-Score
		Mean	S.D.	Mean	S.D.
Outcome variable
	Market orientation of collected NTFPs	30.69	48.87	24.92	41.22	−1.620
	Household dependency on collected NTFPs	7256.10	12,836.42	3515.21	6172.57	−4.833 ***
Explanatory variables
Household head characteristics
	Gender	0.67	0.47	0.67	0.46	0.185
	Age	45.10	14.74	41.28	14.84	−3.130 ***
	Marital status	0.93	025	0.88	0.32	2.740 **
	Origin	0.90	0.29	0.92	0.25	1.206
	Ethnicity	0.91	0.28	0.90	0.29	1.389
	Livelihood style	0.77	0.44	0.62	0.48	2.657 **
	Qur’anic education	0.28	0.45	0.37	0.48	−1.642
	Education	0.19	0.39	0.22	0.41	−1.633
Household characteristics
	Household size	10.71	5.16	9.74	4.92	1.720 *
	Children	3.11	2.09	2.72	2.03	1.701 *
Livestock characteristics
	Cattle	0.47	1.37	0.31	0.74	−1.846 *
	Goat	1.97	2.75	1.98	3.14	0.040
	Sheep	0.79	2.10	0.30	1.01	−3.825 ***
	Poultry	2.30	5.04	1.93	3.67	−1.034
NTFPs access type
	Free (no restriction)	0.54	0.49	0.76	0.42	6.026 ***
	Donation	0.07	0.27	0.03	0.17	−2.748 **
	Purchase	0.17	0.37	0.12	0.33	−1.627
	Legacy	0.25	0.43	0.16	0.37	−2.597 **

Note: S.D. refers to Standard Deviation. Significant at * 10%, ** 5%, and *** 1%, respectively.

presents a comparative analysis of beneficiaries and non-beneficiaries of the land restoration programme across a range of sociodemographic, economic, and resource access variables. The findings offer insights into the characteristics of households that engage in the programme and the potential impact of participation on NTFP income. Households participating in the ILR reported a significantly higher mean of NTFP commercialization of (30.69%) and value collected (7256 CFA) compared to non-beneficiaries (24.92% and 3515 CFA), respectively. This substantial income disparity suggests that programme participation enhances NTFPs availability, potentially due to improved access, better resource management practices, and increased forest productivity facilitated by restoration activities. The gender distribution of both groups was found to be identical, with 67% of household heads being male, indicating that gender does not appear to be a significant determinant of participation. However, beneficiaries were found to be slightly older (45.10 years) than non-beneficiaries (41.28 years), suggesting that older individuals may be more likely to engage in restoration efforts, possibly due to a stronger reliance on forest resources for their livelihoods.

Furthermore, a higher proportion of beneficiaries (93%) are married in comparison to non-beneficiaries (88%), suggesting that family responsibilities may be a driving factor in participation, as households with more dependents may seek additional income opportunities through NTFP collection. Education levels also differ between the two groups. The results indicate that Qur’anic education is more prevalent among non-beneficiaries (37%) than beneficiaries (28%), while beneficiaries exhibit slightly higher representation in formal education (19% vs. 22%). This pattern suggests that households with higher levels of formal education may have greater awareness of, or willingness to participate in, land restoration programmes, possibly due to an improved understanding of their long-term economic and environmental benefits.

Household size is another distinguishing factor, with beneficiaries having larger household sizes (10.71 members) compared to non-beneficiaries (9.74 members) and slightly more children (3.11 vs. 2.72). This trend indicates that larger households may be more inclined to participate in restoration programmes, likely due to greater income needs and dependence on forest resources for both household consumption and market sales. With respect to livestock ownership, both groups possess livestock; however, beneficiaries generally possess more poultry (2.30 vs. 1.93), sheep (0.79 vs. 0.30), and cattle (0.47 vs. 0.31).

The various ways in which individuals obtain non-timber forest products are referred to as NTFP access types. Free access signifies that the resources are available without restriction, typically on communal lands. Donation describes receiving NTFPs in-kind, as a gift, transfer from relatives, etc. Purchase requires either buying the products or paying for harvesting rights. Legacy access is based on inherited or customary rights passed down through families or communities. Together, these categories reflect the mix of informal, traditional, and market-based mechanisms that govern NTFP use. A considerably higher proportion of non-beneficiaries (76%) access NTFPs without restriction compared to beneficiaries (54%), suggesting that unimpeded resource access may diminish the motivation to engage in restoration programmes. Conversely, beneficiaries exhibited a greater reliance on legacy access (25% vs. 16%) and purchase-based access (17% vs. 12%), suggesting that households that invest in NTFPs through inheritance or direct purchase are more inclined to participate in sustainable forest management practices. Furthermore, the higher reliance on donations among beneficiaries (7%, and 3%, respectively) suggests that some program participants may be more dependent on external support for NTFP access.

4.2. Utilisation of Non-Timber Forest Products, and Household Behaviour During Food Shortage

Specific parts of trees harvested by beneficiaries were categorised, including branches, leaves, bark, and whole trees (Figure 4). Overall, fruits were the most collected parts, mainly among non-beneficiaries. On the other hand, beneficiaries were more involved in other parts, such as leaves, bark, roots, and flowers.

Figure 5 shows that harvested tree parts are utilised for a variety of purposes, including food, fodder, construction, and medicine. Food was the major reason for using local trees. Generally, beneficiaries used local trees more for food, medicine, and construction materials, while non-beneficiaries were more active in harvesting fodder.

Households experiencing food insecurity employ a variety of coping mechanisms to mitigate the adverse effects of food shortages (Figure 6). A common strategy is the utilisation of NTFPs, underscoring the pivotal role that forest resources play in sustaining livelihoods during periods of food scarcity. In addition to reliance on NTFPs, households frequently resort to temporary migration, seeking alternative income sources in different locations to support their families. Another main strategy employed is the sale of livestock, which serves as a crucial financial buffer, enabling households to purchase essential food supplies. Furthermore, many households reduce the number of daily meals, a strategy that, while providing short-term relief, can have long-term nutritional and health consequences, particularly for vulnerable groups such as children and the elderly.

A distribution of household expenditures among different categories provides insights into the economic priorities of NTFP users (Figure 7). The figure highlights key spending areas such as household consumption, education, healthcare, and other financial commitments. A high proportion of income directed toward household needs, such as food, medicine, and clothing, suggests the role of NTFPs in sustaining livelihoods and enhancing food security. Conversely, less considerable investment allocated to education may indicate less effort to improve long-term economic stability and resilience.

4.3. Drivers of Participation in the Land Restoration Programme

Our findings revealed drivers significantly affecting the participation in the land restoration programme (

Table 3. Factors affecting participation in land restoration programme (Probit Regression) (n = 379).

Variable	Coefficient	Std. Error	z-Value	p-Value
Household head characteristics
Gender	−0.133	0.150	−0.89	0.376
Age	0.011	0.004	2.47	0.013
Marital status	0.436	0.211	2.07	0.039
Origin	−0.196	0.208	−0.95	0.334
Ethnicity	0.069	0.200	0.35	0.728
Livelihood style	−0.148	0.130	−1.14	0.256
Qur’anic education	0.447	0.141	3.16	0.002
Education	0.642	0.181	3.53	0.000
Household characteristics
Household size	0.006	0.014	0.44	0.660
Children	0.060	0.034	1.76	0.078
Livestock characteristics
Cattle	0.045	0.080	0.57	0.572
Goat	−0.059	0.024	−2.45	0.014
Sheep	0.212	0.053	3.94	0.000
Poultry	0.008	0.016	0.54	0.586
NTFPs access type
Free	−0.684	0.159	−4.29	0.000
Donation	0.679	0.301	2.25	0.024
Purchase	−0.427	0.238	−1.79	0.073
Legacy	0.024	0.213	0.11	0.909
Constant	−0.605	0.464	−1.30	0.192
LR chi2 (20)	119.35
Prob > chi2	0.000

Note: Std. Error. Refer to Standard Error.

). Sociodemographic characteristics of household head, such as age (β = 0.011, p = 0.013), being married (β = 0.436, p = 0.039), Qur’anic education (β = 0.447, p = 0.002), and primary education (β = 0.642, p = 0.000) have a positive effect. Similarly, household asset ownership, particularly sheep ownership (β = 0.212, p = 0.000), has been found to be positively associated with being selected, while goat ownership (β = −0.059, p = 0.014) has been observed to have a negative effect.

Households that accessed NTFPs for free were found to be less likely to be selected in the programmes (β = −0.684, p = 0.000). Conversely, those who received NTFPs through donations were more likely to be engaged (β = 0.679, p = 0.024). Households acquiring NTFPs through purchase were less inclined to be selected in the land restoration programmes (β = −0.427, p = 0.073).

4.4. Impact of the Land Restoration on Household Welfare

Table 4. Comparing the impact of land restoration programme on beneficiary and non-beneficiary households (n = 379).

Sample	Treated	Control	Difference	Std. Error	t-stat
Market orientation of NTFPs collected (%)
Nearest neighbour
Unmatched	30.69	24.92	5.76	3.55	1.62
ATT	30.99	19.18	11.81	5.29	2.23 **
Radius
Unmatched	30.69	25.13	5.56	3.58	1.55
ATT	30.69	25.12	5.57	3.12	1.78 *
Kernel
Unmatched	30.69	25.12	5.57	3.58	1.55
ATT	30.69	23.02	7.67	7.67	1.75 *
Calliper
Unmatched	30.69	25.12	5.57	3.58	154
ATT	30.69	23.73	6.91	6.05	1.85 *
Household dependency on NTFPs collected (total value in FCFA)
Unmatched	7256.10	3546.16	3709.16	780.41	4.75 ***
ATT	7297.40	3281.37	4016.02	1047.09	3.84 ***
Radius
Unmatched	7256.10	3546.16	3709.93	780.41	4.75 ***
ATT	7256.10	3546.16	3709.93	777.68	4.77 ***
Kernel
Unmatched	7256.10	3546.16	3709.93	780.41	4.75 ***
ATT	7256.10	3302.32	3953.77	913.16	4.33 ***
Calliper
Unmatched	7256.10	3546.16	3.709.93	780.41	4.75 ***
ATT	7181.36	3186.25	3995.11	1035.11	3.86 ***

Note: Std. Error. refers to Standard Error. Significant at * 10%, ** 5%, and *** 1%, respectively.

illustrates the impact of land restoration on key welfare indicators of beneficiaries and non-beneficiaries by comparing treated and control groups using both unmatched and Average Treatment on the Treated (ATT) estimations. The analysis focuses on two primary variables: the quantity (percentage) of NTFP taken to market, and the monetary value of NTFP collected. The unmatched results indicate that 30.69% of the collected products are marketed by the beneficiaries, compared to 24.92–25.13% for the control group, showing a positive but statistically insignificant difference of 5.56–5.76% (t = 1.62). The ATT estimation, however, reveals a more substantial and statistically significant increase of 5.57–11.81% (t = 2.23, significant at the 5% level).

The unmatched analysis indicates that the value of NTFP collected by the beneficiary group is worth 7256 CFA, in contrast to 3546 CFA for the non-beneficiary group. These results show a significant disparity of 3709 CFA (t = 4.75, significant at the 1% level). The ATT results further corroborate this trend, with an even higher NTFP value difference of 4016 CFA (t = 3.84, significant at the 1% level). This demonstrates a strong positive effect of land restoration on households.

5. Discussion

The objective of this study was to estimate the effect of social programmes, particularly land restoration initiatives, on household welfare and behaviour. Our results suggest important social, cultural, and economic implications. Land restoration programmes not only improve household income through NTFP commercialization but also influence community reliance on forests for food security, nutrition, and animal husbandry. At a cultural level, participation patterns show how restoration programmes reinforce collective responsibilities within households, especially among older and married individuals who embody community trust and resource stewardship. Economically, restoration strengthens household resilience by providing a stable income from NTFPs and diversifying coping strategies in periods of food scarcity. These findings resonate with the broader understanding of social programmes as mechanisms that go beyond environmental restoration to foster livelihoods, food security, and social cohesion.

5.1. How Does Participation in Land Restoration Programmes Affect the Utilisation of Non-Timber Forest Products?

Our study documented local households’ behaviour towards the collection and utilisation of NTFPs in the study site as well as the differences between beneficiaries and non-beneficiaries in terms of tree parts and modes of use. These results confirm those reported by [6,9,22,31,56], highlighting the diversity of uses and the influence of awareness-raising actions on collection practices. Furthermore, the distribution of collected parts provides insights into the sustainability of forest resource use. A higher proportion of fruit and leaves being collected suggests that communities are engaging in selective harvesting, which is generally considered more sustainable and allows for natural regeneration [57]. Conversely, a preponderance of whole trees being harvested may be indicative of unsustainable practices, which, in turn, could contribute to deforestation and long-term ecological degradation [58]. The composition of tree parts collected also has the potential to reflect resource availability, local regulations, or traditional practices regarding forest use. It is therefore vital to understand these dynamics to inform policies on sustainable forest management and conservation strategies.

The preponderance of tree parts employed for food signifies a pronounced reliance on forest resources for household purposes and highlights the imperative for alternative food sources to mitigate pressure on forests. This underscores the role of landscape restoration to boost agriculture as a means of livelihood [59]. The animal feed use highlights the multifunctional role of forests in community well-being, underscoring their significance not only as a source of food but also as a critical component for their animals. These insights underscore the necessity for integrated forest management policies that balance resource utilisation with conservation efforts. From a theoretical perspective, collective action theory helps to explain these behavioural patterns. Households act collectively when short-term private benefits (e.g., food security from tree cutting) are balanced with long-term community benefits (e.g., sustained NTFPs through selective harvesting). Participation in restoration programmes thus reflects both self-interest and the willingness to cooperate for shared resources.

5.2. What Drives the Participation of Households in Land Restoration Programmes?

Our findings indicate that older and/or married household heads are more likely to be engaged in land restoration activities. This phenomenon may be attributed to a heightened sense of responsibility towards community resources or an increased economic reliance on forest products, and the priority given to the most vulnerable as one of the conditions to be selected as a beneficiary. Consistent with the findings, a comparable effect of age and marital status of household head on land restoration programmes in central Ethiopia and China has been documented, where being older and married increased the chance of becoming a beneficiary [60,61]. Furthermore, educational background also positively and significantly influenced the participation, as education improves awareness about and engagement in programmes focusing on restoration and sustainable use of natural resources. These findings underline the experiences that farmers with a decent level of education are more likely to participate in land restoration programmes [61].

Household assets ownership, particularly livestock, is also very often associated with participation in restoration programmes. Nevertheless, our study interestingly identified sheep ownership as having positive implications towards participation, while goat ownership reduced the probability of participation. This can be explained that households with sheep may have a vested interest in sustainable land use, whereas goat owners may engage in practices less compatible with restoration (e.g., overgrazing). Similar studies reported how the type of farm animal kept by households affects their participation in the land restoration programme in arid zones of Ethiopia and Jordan, attributing the fact that owners of livestock that graze more (e.g., goats) are less likely to participate in the land restoration programme [42,62].

Finally, our study also highlighted the role of access to forest resources as one of the significant factors influencing participation in land restoration programmes. Households involved in NTPFs collection and use through donations tend to participate, while those with free access do not. This suggests that households with unrestricted access may not perceive a need to invest in restoration efforts. Similarly, households purchasing NTFPs were less inclined toward the land restoration programs. This confirms the theory that free access to NTFPs and tendencies to join or support forestry restoration and conservation partnerships and practices are rather weak [42]. The participation drivers identified here reinforce collective action theory: when resource use is unregulated or free, incentives for cooperation and restoration weaken; conversely, when resource scarcity or conditional access is present, communities demonstrate stronger cooperative behaviour.

5.3. How Does the Participation in Land Restoration Programmes Affect Household Welfare?

Our study reveals the impact of land restoration on key welfare indicators of beneficiaries and non-beneficiaries with respect to the commercialisation of collected NTPFs. It is generally observed that beneficiaries demonstrate a higher level of sales activity in relation to the forest products they collect, in comparison with non-beneficiaries, similarly to the recently published review [63]. In this instance, our study aims to address this research gap concerning the socioeconomic impact of land restoration programmes, particularly the increased market orientation of forest resources. The findings of the study therefore suggest a positive association between land restoration initiatives and an increase in the commercialisation of NTFPs. In line with these observations, participants involved in conservation and restoration programmes increase the market orientation of NTFPs harvesting when compared to non-participating households that use forest products predominantly for subsistence purposes. This practice has already been documented in other regions in the savannah zone of Africa, such as Burkina Faso and Ghana [64,65].

Furthermore, the trend in growing market-orientation has a rising tendency. Currently, beneficiaries show an average double income from selling NTPFs compared to non-beneficiaries. Nevertheless, our analysis shows that this difference will persist and even intensify in the near future. Our study therefore demonstrates a strong positive effect of land restoration programmes on participating households, which should be perceived as a nudge for increasing participation rate. The findings suggest that land restoration significantly enhances beneficiaries’ ability to collect forest products, increases their market participation, and boosts their household income. The results underscore the efficacy of land restoration programmes in enhancing rural livelihoods through the provision of sustainable forestry-derived economic opportunities, thereby aligning with the reported significant impact of land restoration programmes on beneficiary incomes in Ethiopia [42,66].

The high collection of fruits by non-beneficiaries reflects low food security, while that of leaves by beneficiaries is linked to World Food Programme (WFP) support on market gardening sites. These results are consistent with those of [67,68], who showed that development programmes promote the use of leaves by enhancing awareness of their nutritional value and supporting their integration into household diets, particularly through nutrition gardens. The selective collection of NTFPs by communities appears to be linked to the WFP’s multiple development initiatives, including training and awareness, promoting rational resource management. The composition of tree parts collected reflects that the diversity of NTFPs, combined with adapted regulatory mechanisms, contributes to better availability and sustainability of these resources [69]. Also, the preponderance of certain collected parts reflects a strong dependence of local populations on NTFPs for their resilience, as shown in the study by [70] around the Niassa reserve, Mozambique, where more than 90% of households depend on these resources for food, care, and income. The animal feed use highlights the multifunctional role of forests in community well-being, underscoring their significance not only as a source of food but also as a critical component for their animals. As Ref. [68] showed, the use of NTFPs in animal feed clearly illustrates the multifunctional role of forests in Sahelian agropastoral systems. The utilisation of NTFPs, underscores the pivotal role that forest resources play in sustaining livelihoods during periods of food scarcity. These results are consistent with those of [71], who highlighted that local communities in semi-arid areas of Tanzania rely heavily on NTFPs to improve their food security, particularly due to their vulnerability to climate-sensitive rain-fed agriculture. The development of these coping strategies, such as temporary migration, livestock sales, and meal reduction, has been widely documented and their central role highlighted in the decision-making process of the local household towards strengthening resilience to food insecurity in the Sahel [72,73]. The share of income from the sale of NTFPs allocated to food, health, and clothing confirms their role in livelihoods [8,68]. Conversely, low spending on education may reflect a priority given to immediate needs at the expense of long-term resilience. The participation of older and married people in restoration activities could be linked to a greater sense of responsibility for community resources or increased economic dependence on forest products, as well as the priority given to the most vulnerable, as one of the conditions for beneficiary selection. Consistent with findings reported by studies [60,61] that documented a comparable effect of the age and marital status of the household head on land restoration programmes in central Ethiopia and China, where being older and married increased the chances of becoming a beneficiary, basic education improves awareness of or engagement in these programmes. This result is consistent with a study documenting that farmers with early to moderate education levels are more likely to participate in land restoration programmes [61]. The positive effect of sheep ownership can be attributed to the fact that households that own sheep are expected to provide for this category of animals, as they are always attached to the house. The deleterious effect of goat ownership can be attributed to the fact that they are stray animals and therefore pose a threat to the resource. In accordance with the findings of other studies, the present study has demonstrated that the type of livestock kept by households has a significant impact on their engagement in land restoration programmes in arid regions of Jordan and Ethiopia. For instance, research conducted in Jordan and Ethiopia [42,63] has indicated that owners of livestock that graze extensively, such as goats, are less inclined to participate in these programmes. Households with free access to NTFPs are often already engaged in programmes, facilitating their access. In contrast, those who rely on donations are less involved and remain marginalised. These findings are in line with [74], who highlights the role of social capital in accessing NTFPs in Cameroon. A similar relationship between free access to NTFPs and forest restoration and conservation partners has been reported in Indonesia [61]. The more substantial and significant increase in beneficiaries’ income suggests a potential association between land restoration initiatives and an increase in the commercialisation of NTFPs. Consistent with these observations, studies have reported that savannah ecosystem conservation programme participants commercialise more NTFPs compared to non-beneficiaries in Burkina Faso and Ghana, respectively [64,65]. The significant disparity between beneficiaries’ and non-beneficiaries’ incomes demonstrates a positive effect of land restoration on households. The results suggest that land restoration significantly improves beneficiaries’ ability to collect forest products, increases their market participation, and increases their incomes. The results highlight the effectiveness of land restoration programmes in improving rural livelihoods through the provision of sustainable economic opportunities derived from forestry, which is consistent with the reported significant impact of land restoration programmes on beneficiaries’ incomes in Ethiopia [8,42,61].

5.4. Scope, Limitations, and Future Research

This study is facing certain limitations, namely with respect to data collection. The single-site focus might restrict generalisability. Reliance on self-reported household data may introduce recall or social desirability bias. In addition, our analysis did not fully consider intra-community power dynamics, gender roles, or institutional factors. Future research should use larger, cross-regional, and longitudinal datasets, complemented by mixed methods to validate findings. Comparative studies across restoration programmes, with greater attention to governance, local institutions, and gender dimensions, would strengthen understanding of collective action in restoration.

6. Conclusions and Policy Recommendations

This study examined household participation in Niger’s Integrated Land Restoration (ILR) programme and its implications for community welfare through access to non-timber forest products (NTFPs). By linking socioeconomic determinants, such as age, marital status, education, and livestock ownership, with households’ modes of accessing NTFPs, the analysis advances understanding of how restoration programs can achieve both ecological and social goals in Sahelian contexts.

Our results highlight three novel insights. First, underrepresented groups, particularly youth and unmarried individuals, face barriers to participation, underscoring the importance of social status in shaping engagement. Second, education emerges as a strong predictor of involvement, especially Qur’anic and primary education, pointing to the value of integrating restoration awareness into both formal and religious schooling. Third, households’ access to NTFPs significantly mediates participation, with free access reducing incentives to join, while donation-based access increases it. These findings emphasise that participation is not only a function of individual attributes but also of how households interact with local resource systems.

These results carry direct policy implications. To broaden participation, the ILR program should adopt youth- and women-focused outreach strategies, using schools, Qur’anic institutions, and community centres as platforms for awareness and training. Family- and community-based incentives should be tested to encourage responsibility among unmarried or younger members. The positive role of education suggests the need to embed restoration principles into local curricula and religious instruction, thereby fostering long-term environmental stewardship. Livestock dynamics also call for targeted interventions. Fodder support and improved grazing schemes could incentivize sheep owners, while goat-owning households require measures that address grazing conflicts and economic barriers. With respect to NTFPs, community-based management systems and regulated access policies are critical for aligning household incentives with sustainable use. In addition, conditional resource support, such as NTFP donations or alternative livelihood opportunities tied to program participation, can motivate households that might otherwise opt out.