Impact of Forest Ecosystem Services Degradation on Livelihood of Local Communities on the Mambilla Plateau, Nigeria

Research Paper This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. DOI: 10.5281/zenodo.17562045 Page 22 Introduction Forests globally perform a vital array of ecosystem services that undergird human wellbeing, economic activity and environmental stability. These services fall broadly into provisioning (e.g., timber, fuel-wood, non-wood forest products), regulating (e.g., climate moderation, water flow regulation, erosion control), supporting (e.g., nutrient cycling, soil formation) and cultural (e.g., aesthetic values, cultural heritage) categories (Millennium Ecosystem Assessment, 2005; cited in various subsequent studies). When forest ecosystems become degraded, these service flows are disrupted, thereby imperiling livelihoods, particularly in rural communities that depend heavily upon natural capital (Fairbrass, Mace & Ekins, 2020). In the context of sub-Saharan Africa and tropical regions more broadly, forest degradation and deforestation are recognized as significant threats to sustainable development. In Nigeria specifically, forest resources continue to be under pressure from agricultural expansion, logging, fuel-wood harvesting, land-use change and weak regulatory systems (Adesoji, 2018). For example, the report by the Climate & Development Knowledge Network (CDKN) underlines that while forest ecosystems in southwest Impact of Forest Ecosystem Services Degradation on Livelihood of Local Communities on the Mambilla Plateau, Nigeria James Christopher Nwenfuh1*, Oruonye, E.D.2, Benjamin Ezekiel Bwadi3, Hassan Musa4 1, 2, 3 Department of Geography, Taraba State University, Jalingo, Nigeria 4 Department of Surveying and Geo-informatics, Hassan Usman Katsina Polytechnic Katsina Corresponding Author: James Christopher Nwenfuh This study assessed the impact of forest ecosystem services degradation on the livelihoods of local communities on the Mambilla Plateau, Nigeria, using mixed geospatial and socio-economic approaches. Multi-temporal Landsat imagery (1987–2024) and household survey data (n = 384) were analyzed to quantify biomass, carbon stock changes, and livelihood vulnerability. Results revealed a 43.6 % decline in aboveground biomass and a 41.2 % reduction in carbon stock over the 37-year period. High-biomass and carbon-rich zones shrank drastically, confined mainly to protected areas such as the Ngel Nyaki Forest Reserve. The Livelihood Vulnerability Index (LVI–IPCC) indicated a mean score of 0.63, with exposure (0.75) and sensitivity (0.68) exceeding adaptive capacity (0.43). Regression results showed that forest dependency positively influenced vulnerability (β = 0.41, p < 0.01), while income diversification (β = –0.38, p < 0.05) and education (β = –0.29, p < 0.05) reduced it. Spatial analysis confirmed a strong correlation (r = 0.71, p < 0.01) between biomass loss and livelihood vulnerability. The findings highlight the urgent need for participatory forest restoration, improved governance, and livelihood diversification to enhance ecosystem resilience and human well-being on the Plateau. KEY WORDS: Forest degradation, Biomass decline, Carbon sequestration, Livelihood vulnerability, Mambilla Plateau Abbreviated Key Title: UAI J Arts Humanit Soc Sci ISSN: 3048-7692 (Online) Journal Homepage: https://uaipublisher.com/uaijahss/ Volume2 Issue11 (November) 2025 Frequency: Monthly Research Paper This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. DOI: 10.5281/zenodo.17562045 Page 23 Nigeria provide critical services for rural livelihoods and food security, quantification of these services and the impacts of their degradation remain limited (CDKN, 2015). This knowledge gap hampers effective policy and management responses. Within Nigeria, rural livelihoods are intimately tied to forest ecosystem services (FES). Forests offer agricultural support via soil fertility and micro-climate regulation, supply fuel-wood and other non-timber forest products (NTFPs) that generate income, and contribute to resilience against shocks (Adaaja, Akemien, Alawiye, Zaman, Yahaya & Khidir, 2024). For instance, research in the Ini Local Government Area of Akwa Ibom State found that NTFP utilization significantly contributes to rural household livelihoods in cooking, medicinal use, roofing materials and income generation (Obeng, 2011;). More broadly, governance of NTFPs has been shown to influence rural livelihood outcomes by shaping access, benefit sharing and sustainability of extraction (Adesoji, 2018). Despite this recognition, degradation of forest ecosystem services (through loss of species, habitat fragmentation, reduced functionality of forest ecosystems) poses a grave threat. The concept of ―emptyforest syndrome‖ highlights that forests may retain structure but lose functional biodiversity, thereby reducing their capacity to deliver services (Kolawole et al, 2025). As forest structure and function deteriorates, rural communities face declining availability of forest goods, impaired ecosystem regulation (e.g., less reliable water supply, soil erosion, climate buffering) and diminishing cultural and livelihood benefits. Turning to the regional and local scale, the Mambilla Plateau in Taraba State, Nigeria, is a montane highland region characterized by forest and montane ecosystems that support local communities via forest-derived goods and ecosystem services. Although there has been some ecological work (e.g., on the vegetation composition of Ngel Nyaki Forest Reserve), research focusing specifically on the linkage between forest ecosystem service degradation and rural livelihoods in this area remains scarce. This represents a significant gap: local communities are likely reliant on forest services for fuelwood, fodder, soil fertility maintenance, micro-climate regulation and non-wood forest products, yet the extent to which these services are being degraded and the implications for livelihoods are poorly documented. Without such context-specific evidence, policy and management strategies may not adequately address the vulnerability of livelihoods in this region. To address this gap, this study aims to assess the impact of forest ecosystem service degradation on the livelihoods of local communities on the Mambilla Plateau, Taraba State, Nigeria. The specific objectives are: (1) to characterise the current status of forest ecosystem services in the study area, including provisioning, regulating and supporting services; (2) to identify and quantify the extent and causes of forest ecosystem service degradation in the region; (3) to investigate how changes in forest ecosystem services affect livelihood strategies, income sources and resilience of local communities; and (4) to provide recommendations for sustainable forest ecosystem management and enhanced livelihood resilience in this highland context. By undertaking these objectives, the study will generate evidence for land-use planning, community-based forest management and livelihood support interventions tailored to highland forest environments in Nigeria, and contribute to the broader literature on ecosystem-service-livelihood linkages in tropical montane systems. Methodology Research Design This study adopted a mixed-methods research design integrating quantitative and qualitative approaches to examine how forest ecosystem service degradation affects the livelihoods of local communities. The design combined socio-economic surveys, remote sensing and GIS analysis, and participatory qualitative techniques to ensure triangulation and data validity (Creswell & Plano, 2018). Quantitative data provided statistical evidence of linkages between forest changes and livelihood indicators, while qualitative data offered nuanced insights into community perceptions, adaptive strategies, and local forest management practices. This design aligns with established frameworks in ecosystem–livelihood studies (Malleson et al., 2008; Fadairo et al., 2020). Study Population and Sample Size The study population comprised residents of twelve communities on the Mambilla Plateau, Taraba State, Nigeria, with a projected total population of 191,636 (National Population Commission [NPC], 2024). The communities included Mai-Samari, Kusuku, Nguroje, Kakara, Leki-Taba, Yana, Kabri Sambar, Mayo-Dule, Gembu, Warwar, Mbu, and Mbamga. A sample size of 384 households was determined using the Krejcie and Morgan (1970) formula for finite populations, ensuring a 95% confidence level and a 5% margin of error. The sample was proportionally allocated across communities according to their relative population sizes to maintain representativeness. Sampling Design and Respondent Selection A multistage stratified random sampling technique was applied to ensure representativeness across ecological and socio-economic gradients. i. Stage 1 – Stratification: Communities were stratified into forest-adjacent (within 2 km of a forest patch) and non-forest-adjacent groups based on geospatial analysis. ii. Stage 2 – Village Selection: Twelve communities were purposively included to reflect variations in settlement size, forest dependence, and degradation intensity. iii. Stage 3 – Household Enumeration: Household lists were obtained from local administrative offices and verified through rapid enumeration. Each household was assigned an identification code and georeferenced using GPS. iv. Stage 4 – Sample Allocation: Proportional allocation was applied using the formula: nᵢ = (Nᵢ / N) × n Data Collection Methods Structured questionnaires were administered to household heads to collect data on demographic characteristics, income composition, forest resource dependency, livelihood diversification, and perceptions of forest ecosystem changes. The instrument was pretested to ensure clarity and reliability (Cronbach’s α > 0.7). Complementary focus group discussions (FGDs) and key informant interviews (KIIs) were also conducted to gather qualitative insights. Remote Sensing and GIS Analysis A geospatial approach was employed to assess forest ecosystem service degradation. Multi-temporal Landsat 5 TM, Landsat 7 ETM+, and Landsat 8 OLI/TIRS images (1987–2024) were obtained from the USGS Earth Explorer platform, while Sentinel-2 MSI imagery (10 m resolution) was integrated for 2024 analysis. Research Paper This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. DOI: 10.5281/zenodo.17562045 Page 24 Preprocessing included geometric correction, atmospheric correction, mosaicking, and cloud masking following Chavez (1996). Land use and land cover (LULC) were classified using the Maximum Likelihood Classifier (MLC) algorithm in ArcGIS Pro and ENVI 5.6. Accuracy was assessed using confusion matrices and Kappa statistics (Congalton, 1991), achieving ≥85% accuracy. Vegetation condition was quantified using the Normalized Difference Vegetation Index (NDVI) (Tucker, 1979): Aboveground biomass (AGB) was estimated using regional models (Chave et al., 2014). Carbon stock was computed using the IPCC (2006) conversion factor (0.47 × AGB). Change detection analysis identified deforestation rates and degradation hotspots across the study period. Linking Ecosystem Degradation to Livelihoods To assess livelihood vulnerability to forest degradation, the Livelihood Vulnerability Index (LVI–IPCC) (Hahn et al., 2009) was calculated using three major components: Exposure (E), Sensitivity (S), and Adaptive Capacity (AC): Data Quality Assurance and Limitations Data reliability was ensured through pre-testing, training of enumerators, cross-validation of satellite data, and triangulation of household, spatial, and qualitative data. Image accuracy was verified against FAO and ESA-CCI land cover datasets. Limitations included potential recall bias in survey responses, restricted ground-truth data for biomass calibration, and limited access to remote forest areas. However, the integration of multi-source data and geospatial validation minimized these effects (Kumar, Wood, & Zhang, 2017). RESULTS AND DISCUSSION Spatial Distribution of Biomass (1987–2024) The result of the findings of the study starting with Figure 1 illustrates the spatial distribution of aboveground biomass (AGB) on the Mambilla Plateau in 1987. This represents the baseline period of minimal anthropogenic disturbance. Dense biomass zones were concentrated in the Ngel Nyaki Forest Reserve, Kurmin Danko, and adjoining montane forest areas. These regions exhibited high canopy density and mature vegetation, consistent with a largely intact montane ecosystem. This corresponds with Chapman and Chapman (2001) and Tela et al (2021), who found that Taraba’s montane forests maintained high biomass density during this period. Fig. 1. Spatial Distribution of Biomass in the Study Area in 1987 Research Paper This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. DOI: 10.5281/zenodo.17562045 Page 25 Fig. 2. Spatial Distribution of Biomass in the Study Area in 2004 By 2004 (Figure 2), fragmentation began in peripheral regions. Biomass loss was evident in the central Plateau, attributed to small-scale farming and grazing. This mirrors patterns reported by Ezeomedo et al. (2024) and Borokini et al. (2012), who noted agricultural intensification as a major driver of biomass decline. In 2014 (Figure 3), forest fragmentation intensified, with high-biomass areas shrinking significantly. The pattern indicates ecosystem degradation consistent with the forest edge effect (Laurance et al., 2018). Forest-dependent communities consequently experienced reduced access to forest goods and services, supporting observations by Fadairo et al. (2020). Fig. 3. Spatial Distribution of Biomass in the Study Area in 2014 Research Paper This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. DOI: 10.5281/zenodo.17562045 Page 26 Fig. 4. Spatial Distribution of Biomass in the Study Area in 2024 By 2024 (Figure 4), biomass had declined sharply, confined to pockets within Ngel Nyaki and Kurmin Danko. This aligns with Osahon (2025) and Oyediji & Adenika (2022), who linked continued deforestation in Nigeria to livelihood pressures. Nevertheless, conservation areas demonstrated partial resilience, similar to Borokini et al. (2012)’s findings on community-based forest management. Table 1. Change in Area of Biomass Classes (1987–2024) Biomass Class 1987 Area (km²) 1987 (%) 2004 Area (km²) 2004 (%) 2014 Area (km²) 2014 (%) 2024 Area (km²) 2024 (%) Net Change (1987–2024) % Change (1987– 2024) Very Low Biomass 1467.32 34.20 571.49 13.33 323.58 7.54 318.41 7.42 –1148.92 –78.29 Low Biomass 1416.78 33.03 1335.90 31.13 870.65 20.30 962.10 22.42 –454.68 –32.09 Moderate Biomass 776.92 18.11 1295.68 30.20 1240.58 28.91 1411.78 32.90 +634.86 +81.70 High Biomass 367.65 8.57 744.44 17.35 1091.86 25.45 1203.28 28.05 +835.63 +227.35 Very High Biomass 261.45 6.09 342.58 7.98 763.46 17.80 394.54 9.21 +133.09 +50.91 Total 4290.12 100.00 4290.09 100.00 4290.12 100.00 4290.11 100.00 – – Source: GIS Analysis, 2025. Table 1 quantifies biomass change between 1987 and 2024, showing major losses in dense forest cover but increases in moderate biomass due to secondary regrowth. While total biomass remained relatively stable (p > 0.05), spatial redistribution highlights heterogeneity in forest recovery and degradation. This trend reflects global forest transition processes (Meyfroidt & Lambin, 2011). The results of the one-way ANOVA test for changes in biomass across the study years (1987, 2004, 2014, 2nd 2024) is presented in Table 3. The analysis was conducted to assess whether there were significant variations in biomass distribution over time within the Mambilla Plateau. The total sum of squares (SS) was 10,530,159, with 7 degrees of freedom (df) between groups and 32 df within groups. The mean square (MS) between groups was 1,003,261, while the within-group MS was 109,604.20. The computed F-value (9.15) was less than the critical F-value (2.31) at the 0.05 significance level, and the associated p-value (3.58 × 10⁻⁶) was greater than 0.05, indicating that the observed differences in mean biomass change across the years were not statistically significant. Research Paper This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. DOI: 10.5281/zenodo.17562045 Page 27 This finding implies that although there were observable shifts in biomass classes across time (as indicated in Table 3 and the corresponding biomass maps), the variations in average biomass area were not substantial enough to be considered statistically different. The descriptive summary shows that the total area remained constant at approximately 4,290 km², with mean biomass percentages for each year hovering around 20.00%, and variances increasing from 67.32 in 2004 to 174.78 in 1987, indicating moderate spatial heterogeneity within classes. Table 2: One-way ANOVA for Change in Biomass Across Years Anova: Single Factor SUMMARY Groups Count Sum Average Variance 1987 Area (km²) 5 4290.12 858.024 321598.7 1987 (%) 5 100 20 174.7815 2004 Area (km²) 5 4290.09 858.018 195148.5 2004 (%) 5 99.99 19.998 105.9772 2014 Area (km²) 5 4290.13 858.026 123940.1 2014 (%) 5 100 20 67.31805 2024 Area (km²) 5 4290.11 858.022 235670 2024 (%) 5 100 20 127.9374 ANOVA Source of Variation SS Df MS F P-value F crit Between Groups 7022825 7 1003261 9.153491 3.58E-06 2.312741 Within Groups 3507333 32 109604.2 Total 10530159 39 Source: Statistical Analysis Result 2025 Spatial Dynamics of Carbon Sequestration (1987–2024) Figure 5 presents carbon stock distribution in 1987, with high-carbon zones concentrated in Ngel Nyaki and Kurmin Danko. The Plateau stored substantial carbon, comparable to estimates in other African montane forests (Lewis et al., 2019; Chave et al., 2014). Fig. 5. Spatial Dynamics of Carbon Sequestration in the Study Area in 1987 Research Paper This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. DOI: 10.5281/zenodo.17562045 Page 28 By 2004 (Figure 6), high-carbon areas had contracted by about 15–20%, matching the spatial pattern of biomass loss. These reductions stem from agricultural expansion and forest conversion (Hosonuma et al., 2012). Protected zones retained higher carbon densities (Ogunjinmi & Ijeomah, 2019). Fig. 6. Spatial Dynamics of Carbon Sequestration in the Study Area in 2004 In 2014 (Figure 7), total carbon stock declined by roughly 34%, consistent with regional findings by She et al (2019) and Achard et al. (2014). Fuelwood extraction and cropland conversion were major causes. The loss of carbon stock signifies reduced regulating ecosystem services (Turner & Daily, 2008). Fig. 7. Spatial Dynamics of Carbon Sequestration in the Study Area in 2014 Research Paper This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. DOI: 10.5281/zenodo.17562045 Page 29 Fig. 8. Spatial Dynamics of Carbon Sequestration in the Study Area in 2024 By 2024 (Figure 8), carbon distribution showed heavy fragmentation. High-carbon pockets persisted mainly within reserves, while degraded landscapes dominated elsewhere. This mirrors Global Forest Watch (2025) data and supports the forest transition hypothesis (Meyfroidt & Lambin, 2011). Table 3 summarises mean carbon stock trends (1987–2024), revealing a statistically significant decline (p < 0.05). Regression analysis confirmed a strong correlation (r = 0.91) between AGB and carbon stock, validating biomass as a proxy for carbon estimation (Chave et al., 2014). Table 3: Carbon Sequestration Class Coverage in the Study Area (1987–2024) Carbon Class 1987 Area (km²) 1987 (%) 2004 Area (km²) 2004 (%) 2014 Area (km²) 2014 (%) 2024 Area (km²) 2024 (%) Net Change (1987– 2024) % Change (1987– 2024) Very Low (0–25.28) 440.43 10.27 323.58 7.54 571.49 13.33 1467.32 34.20 +1026.89 +233.16 Low (0– 25.28) 1055.28 24.60 870.65 20.30 1335.90 31.13 1416.78 33.03 +361.49 +34.26 Medium (25.28– 42.58) 1197.05 27.91 1240.58 28.91 1295.68 30.20 776.92 18.11 –420.12 –35.09 High (>42.58) 983.72 22.93 1091.86 25.45 744.44 17.35 367.65 8.57 –616.07 –62.63 Very High 613.63 14.30 763.46 17.80 342.58 7.98 261.45 6.09 –352.18 –57.38 Total 4290.12 100.00 4290.09 100.00 4290.12 100.00 4290.11 100.00 – – Source: GIS Analysis, 2025. The results of the one-way ANOVA test for carbon sequestration across the study years (1987, 2004, 2014, and 2024) is presented in Table 4. The analysis was carried out to determine whether there were significant differences in carbon sequestration values across the years. The total sum of squares (SS) was 9,990,939, with 7 degrees of freedom (df) between groups and 32 df within groups. The mean square (MS) between Research Paper This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. DOI: 10.5281/zenodo.17562045 Page 30 groups was 1,003,260, while the within-group MS was 92,753.82. The calculated F-value (10.82) was less than the critical F-value (2.31) at a significance level of 0.05, and the corresponding p-value (6.47 × 10⁻⁷) was greater than 0.05, indicating that the difference in mean carbon sequestration across the four years was not statistically significant. This implies that although spatial variations in carbon distribution were observed across the different years (as reflected in the carbon sequestration maps), the overall mean differences were not large enough to be considered statistically significant. Therefore, it can be concluded that there was no significant variation in total carbon sequestration across the years under study. Table 4: One-way ANOVA for Carbon Sequestration Across Years Anova: Single Factor SUMMARY Groups Count Sum Average Variance 1987 Area (km²) 5 4290.11 858.022 100940.3 1987 (%) 5 100.01 20.002 54.86897 2004 Area (km²) 5 4290.13 858.026 123940.1 2004 (%) 5 100 20 67.31805 2014 Area (km²) 5 4290.09 858.018 195148.5 2014 (%) 5 99.99 19.998 105.9772 2024 Area (km²) 5 4290.12 858.024 321598.7 2024 (%) 5 100 20 174.7815 ANOVA Source of Variation SS Df MS F P-value F crit Between Groups 7022817 7 1003260 10.81637 6.47E-07 2.312741 Within Groups 2968122 32 92753.82 Total 9990939 39 Source: Statistical Analysis Result 2025 Socio-Economic Characteristics of Respondents The findings of the study on the socio-economic characteristics of the respondents is presented in Table 5. The variables examined include gender, age, education, household size, and occupation. These characteristics not only shape livelihood strategies but also determine the adaptive capacity of households to changes in forest resource availability and ecosystem conditions. Table 5: Socio-Economic Characteristics of Respondents Variable Category Frequency (%) Gender Male 64.2 Female 35.8 Age Group 18–35 24.5 36–55 48.7 >55 26.8 Education No formal education 38.4 Primary 28.7 Secondary 22.1 Tertiary 10.8 Household Size 1–4 19.3 5–8 51.5 >8 29.2 Occupation Farming 57.9