Abstract

Non-timber forest products (NTFPs) play a crucial role in food security, household income, and resilience in Sahelian systems. This study aimed to analyze the determinants of NTFP availability and degradation in two agroclimatic zones of the Niger Republic. Data were collected from 379 households using the KoboCollect method and have been analyzed with Stata 16 software through two models, including a linear regression for the availability and a logistic regression for the degradation. The linear regression is globally significant, and results show that the household size and the number of species collected are the main factors driving NTFPs availability. Species analysis reveals that Maerua crassifolia predominates in the northern area, while Diospyros mespiliformis is more important in the southern one. Other species such as Lannea microcarpa, Sclerocarya birrea, and Faidherbia albida are common across both areas, reflecting their central role in food, forage, and medicinal uses. The logistic model is also significant, with an estimated extinction probability of 71.6%. Households in the northern area are facing a lower species disappearance risk at about 13%, while each additional species harvested increases this risk by 2.4%. Further, education reduces the extinction risk by 11%, highlighting its role in raising awareness of sustainable practices. These findings indicate that both household dynamics and ecological conditions influence NTFPs’ sustainability highly, underscoring the need for management strategies such as Assisted Natural Regeneration to ensure biodiversity conservation and secure rural livelihoods.

Keywords

NTFPs, Availability, Degradation, Determinant, Niger

Share and Cite:

1. Introduction

Non-timber forest products (NTFPs) are an essential resource for rural populations in sub-Saharan Africa, providing food, traditional medicines, and vital income (Faye et al., 2023; Shackleton & Pandey, 2014; Shackleton et al., 2011). Beyond domestic needs, they contribute to ecosystem regulation and community resilience in the face of environmental disturbances (UICN & AGEREF-CL, 2024; Moussa et al., 2023). However, their availability is declining due to overexploitation, population growth, and climate change (FAO, 2019; Fandohan et al., 2020). In Niger, particularly in the Sahelian areas, NTFPs are a strategic component of livelihoods, where agriculture and livestock farming are highly exposed to climate hazards (Boubacar et al., 2022). Despite their vital role, few studies have simultaneously explored the factors determining their availability and those responsible for their degradation using an integrated quantitative approach. Understanding these dynamics is crucial in guiding sustainable management strategies that are tailored to the community’s needs and the preservation of ecosystems. This study adopts a dual econometric approach, including a linear regression in order to identify the variables that influence the NTFPs availability and a logit model to analyze their degradation determinants. Therefore, it aims to provide sound scientific information to support the conservation and sustainable use of these vital natural resources in the Sahelian areas. Specifically, it is to characterize the main factors that determine the availability of NTFPs and to identify the variables explaining their degradation. Hypotheses are that availability depends on socio-economic and environmental factors, while degradation is influenced by the intensity of exploitation, demographic pressure, and climate change.

2. Materials and Methods

2.1. Presentation of the Study Area

The study was conducted in two agroclimatic areas following a north-south rainfall gradient in the Maradi region, located in the south-central part of Niger Republic, at approximately 13.5˚ of North latitude and 7.1˚ of East longitude (Haweÿ et al., 2020), namely the northern and the southern Sahelian areas, as illustrated in , which shows the location of the selected villages within the study area. The southern Sahelian area is characterized by a semi-arid climate, with annual rainfall ranging from 500 to 700 mm (Katkoré et al., 2021). In the southern Sahelian area, vegetation is denser with a tree and shrub layer and predominantly sandy-loamy soils with good water retention capacity (WAEN, 2020). The northern Sahelian area is the driest part of the Maradi region, marking a transition to the Sahara. The climate is semi-arid to arid type, with an annual rainfall ranging from 300 to 400 mm (Soumana, 2021). The northern Sahelian area vegetation is made mainly of shrub and tree steppes, with relatively sparse vegetation cover. Soils are mainly sandy to loamy, with low water retention capacity (WAEN, 2020). The selected villages are presented according to their agroclimatic location, as shown in .

Figure 1. Selected villages’ location in the study area.

Table 1. Villages of the study area.

These villages were chosen based on their climatic differences, their proximity to the defined land use units, and their accessibility.

2.2. Sampling and Data Collection

A pre-survey was done by contacting local stakeholders in order to get household data by terroir. A sample was then determined using the following formula (Daniel, 2011):

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

where: n = sample size; t = confidence level; N = parent population; the estimated proportion of farmers using NTFPs as resources; with a margin of error of 5%.

The sample is then of 379 households heads old of at least 18 years. A two-stage stratified sampling was used to ensure adequate household representation. Villages were stratified by ecological proximity (land-use types and agroecological conditions) and sociolinguistic criteria to capture key differences in resource availability and harvesting practices. This approach allows the study to account for relevant population heterogeneity and enables meaningful comparisons between similar villages.Data collected from households heads included socio-economic characteristics, the fields size, and others information on NTFPs, such as the number of species collected, quantities harvested, collection location and distance, perceptions of climate change, pest attacks and excessive logging practices, as well as the degradation or disappearance level of NTFPs, measured by the binary variable “disappeared.” All information was collected via a questionnaire integrated into the KoBoCollect application (version 1.29.3) and checked to eliminate inconsistencies and missing values before statistical analysis.

2.3. Data Analysis

Household survey data collected with KoboCollect were processed and analyzed in Stata 16.0 software. Descriptive statistics were first computed, including species citation frequencies to identify the most important woody resources across the study area. To assess the determinants of NTFP availability, a linear regression model was estimated, linking household socio-economic characteristics, farm characteristics, and NTFP-related practices to the total annual quantity of NTFPs harvested per household.

Each variable included in the model was selected based on its theoretical potential to influence NTFP availability and harvesting. Household socio-economic characteristics reflect harvesting capacity, farm characteristics capture access to resources, and NTFP-related practices (number of species collected, collection distance, pest attacks, excessive logging, and climate perceptions) measure harvesting intensity and environmental and contextual factors.

The variable “perception of climate change” was included because farmers’ perceptions influence adaptive behaviors, such as selective harvesting or tree conservation, even in the absence of objective meteorological data (Dembélé et al., 2024).

The dependent variable being continuous, the ordinary least squares (OLS) method was chosen, as it provides unbiased and efficient estimators under classical assumptions (Greene, 2018). Statistical assumptions were tested: residual normality was assessed, and due to observed deviations, the dependent variable was log-transformed; homoscedasticity was examined using the Breusch-Pagan test and corrected via robust standard errors with the White test; model specification was validated (_hatsq p-value > 0.005); finally, no multicollinearity issues were detected (VIF < 10).

Such models are commonly used to explain household-level variation in NTFP collection (Babulo et al., 2009; Heubach et al., 2011). The model was specified as follows:

$\begin{array}{c}\text{Qthousehold}/\text{year}=\beta 0+\beta 1\text{zone}+\beta 2\text{Gender}+\beta 3\text{Education}+\beta 4\text{Ethnicity}\\ +\beta 5\text{HouseholdSize}+\beta 6\text{FieldArea}/\text{Hectares}\\ +\beta 7\text{NumberSpecies}+\beta 8\text{CollectionLocation}\\ +\beta 9\text{Distance}+\beta 10\text{Climate Change}\\ +\beta 11\text{PestAttack}+\int .\end{array}$

where $\text{Qthousehold}/\text{year}$ represents the total quantity of NTFPs collected per household per year and ϵ the error term.

In the degradation analysis, a binary logit model was applied to estimate the probability of a species being reported as extinct (1 = extinct, 0 = available), using the same explanatory variables. Marginal effects were calculated to ease interpretation. Similar Logit applications have been used to assess forest resource degradation and household harvesting pressures in West African and Sahelian contexts (Pouliot et al., 2012; Shackleton et al., 2001; Vedeld et al., 2004). These approaches provide robust insights into both the availability and the risk of disappearance of NTFPs in the study area. The Logit equation model was as follows:

$\begin{array}{c}\text{Logit}\left(P\left(\text{Missing}=1\right)\right)=\alpha 0+\alpha 1\text{Zone}+\alpha 2\text{Gender}+\alpha 3\text{Education}\\ +\alpha 4\text{HouseholdSize}+\alpha 5\text{Area}+\alpha 6\text{NumberSpecies}\\ +\alpha 7\text{CollectionLocatio}n+\alpha 8\text{Distance}+\alpha 9\text{ClimateChange}\\ +\alpha 10\text{PestAttack}+\alpha 11\text{Overharvesting}\\ +\alpha 12\text{FoodInsecurity}.\end{array}$

Indeed, although the pseudo R² of the logit model is relatively low (0.0592), which is common in the analysis of complex phenomena, the prediction test shows that the model correctly classified 69.68% of the observations. This confirms that the model retains an acceptable explanatory power and provides relevant insights into NTFP degradation.

2.4. Conceptual Diagram

As shown in , the conceptual diagram illustrates the relationships between household characteristics (sex, education level, household size, field size), the availability and degradation of non-timber forest products (NTFPs), and the main influencing factors (quantity of NTFPs collected per household per year, species disappearance, number of species exploited, collection distance, agro-climatic area, climate change, pest attacks, overlogging, and food insecurity.

Figure 2. Conceptual diagram.

3. Results and Discussion

3.1. Results

3.1.1. Socio-Demographic and Economic Characteristics of Households

lists the socio-economic characteristics of the households surveyed in the northern and southern Sahelian areas of the Maradi region. The majority of household heads are male gender (67.39%). The average age of household heads is approximately 41.8 years, and the mean household size is 9.91 persons, indicating large households with a substantial family labor force.

Table 2. Socio-demographic and economic characteristics of households.

Regarding education level, most of the household heads have received formal schooling (84.01%). The dominant ethnic group is Hausa (90.95%), followed by Fulani, Tuareg, and Kanuri. The cultivated area size per household ranges from 2.05 ha in the southern Sahel to 3.35 ha in the north, with an average size of 2.81 ha.

3.1.2. Species Providing NTFPs and Their Importance

A total of 42 woody species were identified, as presented in , where ten species are highlighted for their relatively high citation frequency. For transparency and comparability with other studies, the full list of all species and their importance according to agro-climatic areas is available in (Appendix 1).

In the North Sahelian area, Maerua crassifolia (15.05%), Lannea microcarpa (11.72%), and Piliostigma reticulatum (8.37%) emerged as the most prominent species, followed by Guiera senegalensis and Sclerocarya birrea (7.53% each). In contrast, the South Sahelian area was dominated by L. microcarpa (12.20%), Diospyros mespiliformis (11.72%), and S. birrea (8.35%), alongside Faidherbia albida and Hyphaene thebaica (7.70% each). These findings suggest that, despite the overall diversity of species recorded, a limited group of species accounts for most of the reported uses and cultural importance, illustrating their pivotal role in household subsistence strategies and local resilience.

Table 3. Woody species providing NTFPs and their importance according to the agro-climatic area.

3.1.3. Factors Influencing Availability

As shown in , the linear regression analysis conducted on 379 households shows that the model is overall significant (p < 0.001) with a coefficient of determination R² = 0.50. Two factors appear to be decisive: the number of species collected (β = 0.117; p = 0.009), which is positively and significantly associated with the total quantity collected, and household size (β = −0.017; p = 0.015), whose slightly negative effect suggests that internal resource distribution or activity diversification may limit the quantity collected per member. Other variables (gender, education, ethnicity, land area, collection distance, climate perceptions, parasitic attacks, and collection site) have no significant effects (p > 0.05), indicating a relative homogeneity in collection behaviors.

Table 4. Factors affecting NTFP availability.

3.1.4. Factors Influencing Degradation (Logit Model)

As shown in , the Logit model applied to 379 observations is overall significant (Wald χ² (11) = 25.11; p = 0.0088) with a pseudo R² of 0.0592, indicating that the included variables explain a significant portion of the NTFPs disappearance probability. Among the variables tested, ecological area shows a significant and negative effect (β = −0.63; p = 0.025), indicating that households located in certain areas are less likely to report NTFPs disappearance.

Table 5. Factors affecting NTFPs degradation.

3.2. Marginal Effects

As shown in , the Logit model revealed a high average probability of NTFP extinction, estimated at about 71.6%. Marginal effects showed that some variables influence this risk significantly. Belonging to the northern Sahelian area reduces the likelihood of species disappearance by 13% (marginal effect = −0.130; p = 0.027) compared to the southern one. Similarly, higher education levels are associated with an 11% lower probability of extinction (marginal effect = −0.111; p = 0.066). In contrast, the number of species harvested exerts a positive effect (marginal effect = 0.024; p = 0.091), indicating that each additional species collected increases the probability of extinction at about 2.4%. This result highlights the risk of overexploitation when households diversify harvesting practices excessively, thereby intensifying pressure on natural resources. Other factors, including the gender of the household head, household size, and the distance to the collection sites, did not show statistically significant effects. Likewise, insecurity was not found to significantly influence species extinction.

Table 6. Marginal effects.

Note: (*) dy/dx is for the discrete change of the dummy variable from 0 to 1.

3.3. Discussion

Results show that NTFPs exploitation is closely linked to household characteristics, with most respondents being married farmers and an average household size of around ten persons, representing both a source of labor and an increasing food demand (Faye et al., 2020; FAO, 2011; Shackleton et al., 2019). Beyond these socio-economic attributes, tree species play a central role in household livelihoods. The results confirm that M. crassifolia is particularly important in the northern Sahel, where it serves as both a source of food and fodder, in line with the findings of Nazirou et al. (2024). Similarly, D. mespiliformis appears more significant in the southern area, where it contributes to household income generation and provides multiple uses in food and traditional medicine, corroborating the observations of Ali et al. (2021). Other species such as L. microcarpa, S. birrea, and F. albida are present in both zones, highlighting their multifunctional value, while Z. mauritiana and B. aegyptiaca are also distinguished by their contribution to trading and income generation (Shackleton et al., 2015; Mahamane et al., 2021).

The linear regression result indicates that both woody species diversity and household labor determine the availability of Non-Timber Forest Products highly, supporting the evidence that in Sahel and West Africa, plant species richness enhances the access to NTFPs, improves dietary quality, and provides households with food and income opportunities (Osei et al., 2019). The lack of a significant effect of socio-economic variables supports (Ndiaye et al., 2024), indicating that collection remains a community-based activity, largely influenced by household composition and species diversity rather than individual characteristics or land size, as also noted by Mahamadou et al. (2020) and Soumana et al. (2023) in Niger and the Sahel. Furthermore, local governance and community-based management practices play a crucial role in the availability and sustainability of NTFPs, as highlighted by Moussa et al. (2023) in their study on the role of NTFPs in climate action in the West African Sahel.

Logit model results reveal a high probability of species extinction (71.6%), with significant marginal effects including that households in the northern area is facing lower risk of about 13% compared to those in the south, education reduces extinction probability by 11%, while harvesting a greater number of species increases this risk fot about 2.4% (Goussanou et al., 2019). These findings suggest that simultaneous exploitation of multiple species may exacerbate the risk of overharvesting, consistent with observations that increasing reliance on NTFPs for income can exceed the regeneration capacity of the harvested species (Twine et al., 2025; Ondo-Azi et al., 2025). Similar concerns have been raised in Central Africa, where unsustainable harvesting of Baillonella toxisperma wrecked both ecological resilience and local livelihoods (Reij & Garrity, 2016). Overall, these results highlight a transition from socio-economic factors to ecological pressures, showing that NTFPs’ sustainability in Sahelian systems is mainly conditioned by species availability, household dynamics, and local management practices. This underlines the importance of promoting biodiversity and implementing community-based strategies, such as Assisted Natural Regeneration, to sustain natural resources and strengthen resilience (IPCC, 2022; Reij et al., 2024).

4. Conclusion

This study highlights the importance of NTFPs in the livelihoods of Sahelian households and reveals the threats to their sustainability. Results confirm that the availability of NTFPs is primarily determined by demographic structure and the species diversity exploited, while their degradation depends on ecological conditions and the level of usage pressures. They also underscore the importance of integrating biodiversity and Assisted Natural Regeneration practices into management policies, strengthening community natural resource governance, diversifying income sources to reduce dependence on NTFPs, and implementing participatory monitoring of their dynamics. Such an approach would contribute to reconciling food security, household resilience, and ecological sustainability in Sahelian regions.

Funding

The authors gratefully acknowledge the financial support of the World Food Programme (WFP) for this study, through the memorandum of understanding signed with the Dan Dicko Dankoulodo University of Maradi (UDDM).

Appendix

Table A1. Complete list of 42 species and their importance according to the agro-climatic area.

Conflicts of Interest

The authors declare no conflicts of interest regarding the publication of this paper.

References