Centre de Suivi Ecologique

Seagrass beds are critical coastal ecosystems that play a fundamental role in carbon sequestration, sediment stabilization, and marine biodiversity contributing directly to UN Sustainable Development Goal 14 (SDG 14), particularly target 14.5 which aims to conserve at least 10% of coastal and marine areas by 2030 (UN, 2015). Along the West African coast, and particularly in Senegal, their spatial distribution remains poorly documented. To address this gap, this study, based on 2022 satellite data and conducted under the SEDAD (Ecosystem solutions Ecosystem for sustainable adaptation) project framework. It proposes an integrated approach combining remote sensing, environmental modeling, and participatory knowledge. Satellite data (Sentinel-2 and Landsat 9) were processed in Google Earth Engine to extract spectral indices used as proxies of environmental gradients, including the Normalized Difference Turbidity Index (NDTI), the Normalized Difference Chlorophyll Index (NDCI), and the Normalized Difference Salinity Index (NDSI), as well as sea surface temperature derived from Landsat thermal data. These indices were integrated into a binary logistic regression model to predict the probability of seagrass presence. Additionally, field and participatory surveys were used to guide data interpretation. Spectral indices such as NDVI, EVI, MSAVI2, and NDWI were computed to enhance underwater vegetation detection. The logistic regression model achieved a strong predictive performance, with 81.1% overall accuracy and 76.0% recall. Model robustness was confirmed through a 5-fold cross-validation, ensuring the stability of predictive performance. Among the predictors, NDTI and surface temperature emerged as the most influential variables (38% and 34% relative importance, respectively), while NDCI and NDSI played a secondary role. These findings underscore the value of combining satellite-derived spectral indices with environmental spectral indices to map seagrass meadows in shallow coastal waters. The resulting probability map provides an operational basis for monitoring and managing priority habitats in Senegal’s Marine Protected Areas, and represents a replicable framework for coastal ecosystem assessment across West Africa.

Aim: Lung cancer is the top leading cause death globally, and early and precise detection is required to achieve better survival rates. This research tends to create an optimized Vision Transformer (ViT) model that can be used for real-time lung cancer detection with better generalization and efficiency. Leveraging deep learning and transformer-based models, the proposed model endeavors to realize high diagnostic accuracy along with computational efficiency for real-world usage. Materials and Methods: This research consists of two groups. Group 1: Recently, several start-ups explore the adoption of Vision Transformers (ViTs) to overcome some of the limitations CNNs exhibited. Group 2: Hybrid approaches that leverage the advantages of both ViTs and convolutional architectures, as well as advanced attention mechanisms, represent a strong opportunity to enhance lung cancer detection performance. Results: The Hybrid Attention Vision Transformer (HViT) model classification shows a 96.7% accuracy of 96.7%, , Recall of 96.63%, Precision of 94.8%, F1-score of 96.1% and AUC of 0.97. The model exhibits a substantial inference time reduction, processing images at 0.32 seconds per scan, and is most applicable for hence well-suite for real-time scenario. The model also generalizes well across datasets, with a 5.6% performance gain over conventional CNN-based models. Conclusion: The optimized Vision Transformer model proposed here effectively improves lung cancer detection with enhanced generalization and efficiency. The real-time processing capabilities with improved accuracy of this work make it a potential tool for early diagnosis, which can assist healthcare professionals in making timely decisions and enhancing patient outcomes.

The availability of current land cover and land use (LCLU) information for monitoring the status of land resources has considerable value in ensuring sustainable land use planning and development. Similarly, the need to provide updated information on the extent of LCLU change in West Africa has become apparent, given the increasing demand for land resources driven by rapid population growth. Over the past decade, multiple projects have been undertaken to produce regional and national land cover maps. However, using different classification systems and legends has made updating and sharing land cover information challenging. This has resulted in the inefficient use of human and financial resources. The development of the Land Cover Meta Language (LCML) based on International Organization for Standardization (ISO) standards offers an opportunity to create a standardized classification system. This system would enable easier integration of regional and national data, efficient management of information, and better resource utilization in West Africa. This article emphasizes the process and the need for multistakeholder collaboration in developing a standardized land cover classification system for West Africa, which is currently nonexistent. It presents the survey data collected to evaluate historical, current, and future land cover mapping projects in the region and provides relevant use cases as examples for operationalizing a standardized land cover classification legend for West Africa.

Le climat et la fauche influencent la production fourragère au Sahel. Cependant, l’effet combiné de ces paramètres reste à évaluer. Cette étude vise à évaluer l’effet conjoint de la pluviométrie et des pratiques de fauche sur la quantité et la qualité des fourrages. Des échantillons de phytomasse ont été prélevés au cours et à la fin de la saison des pluies 2021 par coupe intégrale sur 68 parcelles : 20 parcelles avec différents régimes hydriques et 48 parcelles avec différentes quantités d’eau combinées à différentes hauteurs et périodes de fauche. Les valeurs de la phytomasse sèche étaient respectivement comprise entre 2 932,2 ± 1 672,1 et 6 383,6 ± 2 962,6 kg/ha, 2 397,7 ± 6 263,4 et 15 059,2 ± 9 782,9 kg/ha pour les traitements de régime hydrique et du cumul de phytomasse fauchée. Le taux de cellulose brute (en % de Matière sèche) était compris entre 5,5 ± 0,9 % et 6,4 ± 1,2 %. La matière azotée digestible a varié entre 21,8 ± 67,96 et 67,2 ± 15,8 %. Quelle que soit la quantité d’eau reçue, des quantités et qualités équivalentes de fourrages ont été produites par les parcelles non fauchées et celles fauchées au début du cycle de développement des espèces. Pour une fauche à 0 ou 5 cm au-dessus du sol, la quantité de phytomasse était la même. Les parcelles fauchées au stade de fructification des espèces ont produit des fourrages composés de jeunes plantes en quantités plus faibles et de bonne qualité. La quantité et la qualité de fourrage herbacé en fin de saison des pluies ont surtout été influencées par le stade phénologique. Les programmes de gestion des ressources fourragères devront privilégier des périodes et hauteurs de coupe permettant une couverture optimale du sol afin de réduire les risques d’érosion.

Human-induced land cover land use (LCLU) changes such as agricultural extensification and forest degradation and loss have extensive negative impacts including biodiversity loss, land degradation, and a disruption to ecological services. In Senegal, where people are heavily reliant on dryland agricultural production, climate change and land degradation pose particularly significant threats especially as rapid population growth continues to fuel frequent LCLU change. Considering these challenges, approaches that facilitate increased insight into the spatial and temporal dynamics of land use are needed to implement sustainable land management practices and mitigation strategies. However, difficulties associated with Senegal’s highly variable phenology, sparse woody cover and small, irregular fields necessitate the use of Very High Resolution (VHR; < 3 m spatial resolution) data and modern techniques for modeling land use at sufficient scales.We take advantage of VHR data’s spatial resolution and Sentinel-1’s high temporal resolution by implementing an object-based data fusion strategy to model land use. By generating high resolution vector objects from single-date WorldView imagery and using the corresponding Synthetic Aperture Radar (SAR) time series to train a One-Dimensional Convolutional Neural Network (1D CNN), we can effectively leverage deep learning techniques to extract land use signals from multi-resolution and multi-temporal data in a near-autonomous manner.

This study examines the relationship between climate change and public health in Haiti, a country already facing socioeconomic challenges. The well-being of Haiti’s vulnerable population is expected to be further affected by climate change, leading to an increase in vector-borne, water-borne, and heat-related diseases. As one of the most vulnerable countries to climate change effects, Haiti is currently experiencing an increase in vector-borne diseases such as malaria, dengue, and chikungunya, as well as water-borne diseases and emerging zoonotic outbreaks. This study aims to improve planning, decision-making, and responses to public health challenges by utilizing health data, climatic information, and impact models. The methodology involves the creation of a comprehensive climate and health database to uncover detailed spatial-temporal relationships on a national scale. By evaluating disease indicators from historical periods (1950-2014) and future projections (2015-2100) using the Shared Socio-Economic Pathways (SSPs) from the multi-model ensemble mean of the CMIP6 models, target diseases, including malaria, meningitis, dengue, and heat-sensitive chronic diseases are assessed. Our results highlight a decrease in rainfall and a strong increase in temperatures, especially within western Haiti under the extreme SSP585 scenario. The ability of the impact models to simulate the seasonality and spatial distribution of malaria incidence, dengue and heatwaves was performed. The analysis of risks related to climate-sensitive diseases’ climatic parameters shows that Haiti’s west and central regions are mostly exposed to vector-borne and water-borne diseases. Models predict a decrease in malaria cases due to climate change with hot temperatures and a decline in rainfall, while dengue transmission patterns may undergo changes. These findings will inform the implementation of context-specific early-warning systems and adaptation strategies for climate-sensitive diseases while acknowledging the challenges of integrating climate-altered data into health policies.

Abstract Rainfall variability, low soil organic matter content, and costly inorganic fertilizers are the major agricultural constraints in Sub‐Saharan Africa. Integrated use of compost and manure is essential for sustaining soil fertility and increasing crop productivity. This study was conducted to evaluate the combined effects of compost and animal manure on mungbean growth and yield. The 12 treatments consisted of control, recommended dose of nitrogen, phosphorus, and potassium (NPK), 5 ton ha −1 of compost, 10 ton ha −1 of poultry, 10 ton ha −1 of cattle, and 10 ton ha −1 of sheep manure, and six combinations of organic amendments with 50% of their applied rate alone. These treatments were laid out in a randomized complete block design with six replications. Application of cattle manure at 10 ton ha −1 significantly increased mungbean seed yield by 66% and 84% compared to the recommended rate of NPK and control treatments, respectively. Similar observations were made on stem diameter, total pod weight, and number of seeds per pod. Plants amended with compost had the highest number of ramifications and number of pods than NPK fertilized plants, which recorded (9±) ramifications and (27±) pods per plant. On average, integrated use of 5 ton ha −1 of poultry manure + 5 ton ha −1 of cattle manure had the highest soil plant analysis development values, though not significantly different from NPK fertilizers. These results suggest that application of organic amendments could be an alternative to costly and inaccessible inorganic fertilizers for improving mungbean productivity under low‐input agriculture systems.

The mangrove ecosystem abounds in fish and wood resources exploited by local populations for cooking and house construction. This has resulted in a strong loss of mangrove surfaces whose causes are explained by the combined action of natural and anthropic factors. To minimize the regression of the mangrove in Casamance, the Oceanium Association has organized since 2006 reforestation campaigns of mangrove trees. To date, the effectiveness of carbon sequestration by mangrove plantations has never been evaluated in southern Senegal. The present study aims to estimate the produced phytomass (biomass) and the aerial carbon stock of mangrove plantations to promote mangrove restoration actions. Three study sites were selected, namely the communes of Oukout, Mlomp and Di&eacute;mb&eacute;ring, in the Ziguinchor region (Senegal). They polarize 24 plantations distributed in 8 villages of the study area. The characterization of the mangrove plantations was performed from May to July 2020 using 2.5 m x 2.5 m plots, arranged alternately on each transect set up by plantation. The overall amount of carbon sequestered in the plantations in 2020 was estimated at 0.155 tC/ha for stems, 0.389 tC/ha for leaves and 0.501 tC/ha for roots. These results correlated with plant height allowed the development of a regression model to assess the total carbon stock in mangrove plantations. This model explained 81% of the total carbon stock in 7&ndash;10-year-old plantations. Results from this study suggest that mangroves can be dynamic and promising areas for climate change mitigation. &nbsp; Key words: Mangrove, restoration, biomass, carbon sequestration, plantation.&nbsp;

Le bas delta du fleuve Sénégal est une zone d’intérêt écologique au regard de la pluralité d’aires protégées qui sont localisées. Ce milieu, qui se caractérise par la diversité de ses ressources naturelles, développe une très grande vulnérabilité imputable à des causes aussi bien naturelles qu’anthropiques. Depuis quelques décennies, cette fragilité du milieu se répercute sur la mangrove qui assume difficilement ses fonctions socioéconomiques et écologiques. Dès lors, il convient d’analyser la dynamique spatiotemporelle de cet écosystème afin de comprendre les facteurs qui interviennent dans sa dégradation. Dans cette étude, l’objectif a été atteint au moyen de la télédétection à travers le traitement d’images satellitaires Landsat 1980 et 2000, Sentinel 2021 combiné à des travaux de terrains. Cela a permis la réalisation de cartes d’occupation du sol et de cartes de changements. Les résultats démontrent une tendance globale à la hausse de la mangrove avec des espaces couverts qui sont passés de 1 105,82 ha en 1980 à 1 482,32 ha en 2021 malgré une perte de superficies de 166,92 ha sur cette période.