In the Botambi area, south of Bangui in the Central African Republic, deforestation linked to shifting cultivation is leading to the disappearance of the dense, semi-deciduous rainforest, which is being replaced by fallow land overrun by exotic shrubs and grassy savannahs, forcing farmers to clear new forests. The Assisted Natural Regeneration (ANR) technique was trialled in a heavily degraded forest landscape (the village of Boteke) and a less degraded one (Salanga). On the ANR plots, farmers selected and retained useful young trees in their fields whilst weeding their crops between 2019 and 2021, after which the plots were left fallow. The impact of ANR on the floristic diversity of post-crop fallow land was studied by analysing the kinetics of diversity parameters (at the time of the first ANR treatment, and at 7, 13, and 24 months after ANR) and the impact on the above-ground biomass of tree vegetation via mean square diameter and Lorey height at 5, 12, 29 and 53 months after Slash-and-burn. In each village, 11 ANR plots and 10 control plots without ANR, covering an area of 0.231 ha, were cleared and cultivated in early 2019 and monitored. At each inventory, all trees on the plots were identified by their botanical name, height and diameter at breast height. With regard to biodiversity, the relative frequency of species and the diversity index were calculated, and the vegetation in each plot was then classified based on the frequency of the two dominant species. Finally, all plots belonging to the same facies were grouped together to produce 9 'plot facies'. For all plots, diversity is higher at Salanga (Shannon index 2.41 and evenness 0.86) than at Boteke (Shannon index 1.90 and evenness 0.78), and the ANR allows the vegetation to evolve very quickly towards a certain 'stability'. In Boteke, above-ground biomass at 12 months was close to 0, but by 53 months it remained at 0.3 t/ha in the control plots, whereas it reached 0.7t/ha in the ANR treatment. At Salanga, above-ground biomass at 53 months was 0.9 t/ha in the control plots, whereas it reached 3.2 t/ha in the ANR treatment. ANR therefore promotes an accelerated and significant accumulation of above-ground biomass in fallow plots compared to the control, particularly in areas with low levels of degradation.
In Cameroon's Sudano-Sahelian zone, rising fuelwood demand is jeopardising Faidherbia albida agroforestry parklands. Faced with these alarming pressures and ongoing climate challenges, evaluating sustainable management practices and the contribution of F. albida agroforestry parklands to climate change mitigation has become a pressing priority. In these agroecosystems, total and rotational crown pruning are common practices, traditionally aimed at fuelwood and livestock feed collection. However, its impact on the trade-off between leaf removal and pod production remains poorly documented, particularly regarding long-term fuelwood yield potential. This research aims to develop tools for estimating the above-ground wood dry mass of F. albida components in the context of their use as renewable fuelwood and their carbon storage potential. Allometric equations for the different tree components were developed using data from the most recent pruning event (2023) of F. albida trees that had undergone three complete crown-pruning cycles over the past two decades. The predictors used to estimate fuelwood and above-ground wood dry mass were diameter at breast height (D, in cm), crown area (Ca, in m2), and total tree height (H, in m). The best-performing equations for each tree components relied on structurally meaningful combinations of these variables: (i) D2Ca, capturing the combined influence of stem size and crown development for estimating fuelwood obtained after pruning (DWpr, R2adj = 0.84) and the total above-ground wood dry mass of F. albida (TDWb, R2adj = 0.85), corresponding to the sum of DWpr and the post-pruning wood dry mass remaining after pruning (DWUpr); and (ii) D, reflecting the effect of stem diameter for predicting DWUpr, (R2adj = 0.76). The best crown area-only models for predicting DWpr and TDWb showed strong explanatory power, with adjusted R2 values of 0.82 and 0.64, respectively. The integration of structural characteristics visible from space, such as the tree crown area, provides promising prospects for large-scale biomass mapping using satellite images and remote sensing methods.
The study conducted in the Foret des Hraktas, located in eastern Algeria, aimed to model the growth in dominant height of the Aleppo pine (Pinus halepensis Mill.). The data were obtained from stem analyses of dominant trees sampled in 63 temporary measurement plots. Ten dynamic growth equations were fitted using the algebraic difference equations approach, including seven ADA (Algebraic Differences Approach) models and three GADA (Generalised Algebraic Differences Approach) models. Model evaluation was based on several criteria: statistical measures (RMSE, adjusted R2, bias), numerical and graphical analyses of residuals, and assessment of biological realism. With an accuracy of 99%, the GADA model (M8), derived from the Lundqvist-Korf base function, was selected for its superior statistical performance and biological realism. At the reference age of 35 years, three fertility classes were identified, corresponding to dominant heights of 5, 7, and 9 m. The distribution of the studied stands across these three productivity classes was 33% in the 5 m class, 43% in the 7 m class, and 24% in the 9 m class.
Forest ecosystems worldwide are increasingly affected by human activities, climate change, and environmental pressures, highlighting the need for reliable forest health assessments to support sustainable management. This study evaluates forest health in the forests of Khenchela, in Algeria’s Eastern Aurès region, using a methodological approach that integrates Geographic Information Systems (GIS) with the Fuzzy Analytic Hierarchy Process (Fuzzy-AHP). We selected eight environmental factors as indicators: vegetation vigour (NDVI, SAVI, NDMI), climate (temperature, precipitation), topography (altitude), and anthropogenic influence (proxied by distance to residential areas and rivers). We processed and normalised the spatial data layers using geospatial tools and determined their relative importance through Fuzzy-AHP. We then applied the weighted overlay technique to generate the final forest health map, which we validated using GPS-guided field surveys and high-resolution remote sensing data. The resulting map classified the area into five health categories: very healthy (6.94%), healthy (11.80%), moderately healthy (21.07%), low health (36.42%), and unhealthy (23.77%). Validation statistics demonstrated strong agreement between model outputs and reference data, with an overall accuracy of 85.58%. The study provides practical insights for forest managers, policymakers, and local communities, supporting responsible management and long-term conservation of mountain ecosystems. .
Since 2013, the mangrove in the village territory of Thiobon (Casamance estuary) has been regularly monitored, revealing a general trend toward regeneration. However, recent dynamics remain unclear. Is this regeneration continuing at the same pace? Are we witnessing stabilisation, localised regeneration or structural densification of the mangrove cover? This study seeks to answer these questions. This analysis combines Landsat 8 and 9 imageries at 15 m resolution, supervised classification, and expert post-processing. The results reveal a structural densification of the mangrove, with an increase in dense mangrove (+343.5 ha) and a decrease in semi-dense mangrove (-92.2 ha), highlighting an overall positive trajectory and providing useful information to support management and conservation decisions. Methodologically, the classification confusion matrices and comparisons of land-use class areas before and after expert post-processing indicate that the processing chain improves mapping reliability. Finally, the study primarily provides insight into recent dynamics of the Thiobon mangrove, while the methodological approach supports the characterisation of this trajectory.
