In the village soil of Thiobon in lower Casamance (southern Senegal), site being established as Marine Protected Area (MPA), the main activities of the population (rice growing, harvesting of fish products, salt production, …), happens in mangrove zone. These mobilize the population for several months in the year. This mangrove ecosystem, vulnerable and fragile, has undergone profound changes since the 1960s under the combined actions of rainfull variability and main. This study proposes to analyze the spatio-temporal dynamics of Thiobon mangrove. The method used is based on the processing and analysis of satellite data (Landsat imagery acquired in 1972, 1986, 2000 and 2017), rainfall data (1960 to 2017), and on the perception of populations of change in their terroir (field investigation). Map results indicate an overall decline on 52% in mangrove area in tanne during drought years (1970 to the late 1990s). During 2000–2017, characterized by the almost normal return of the rainfall and activities of reforestation done by the local population, about 69% of the mangrove areas lost between 1972 and 2000 regenerated. This dynamic is the result of external and internal forces exerted on this ecosystem and evolve since the early 1970s.

No abstract

No abstract

Changes in land use and land cover have attracted considerable scientific interest in recent years because of their marked influence on hydrological cycles. In the tropics, widespread vegetation degradation and changes in land use that occurred from the 1970s onwards have affected the hydrological dynamics of catchments. In this study, land cover changes in the Casamance catchment upstream of Kolda in recent years were analysed and their role in hydrological evolution was estimated. These changes had occurred in environmental conditions between 1987 and 2018. Remote sensing and geographic information system (GIS) techniques were used to monitor changes in the basin as well as changes in NDVI values. Landsat 5 and 8 satellite images were used respectively. The extraction of land cover in the basin was carried out through the application of unsupervised nesting classification processes. The results indicated that a regression of forest cover in the Casamance River Basin upstream of Kolda and the classification of the area into seven main land categories (forest, savannah, slash-and-burn, cultivated areas, habitat, plantations and water) was significant. Like rainfall, the flow rates, which declined during the 1980s, increased from the 2000s onwards, although the trend was not significant. This increase in runoff is the result of vegetation degradation and increased rainfall. The results highlight the importance of integrating land use information into assessments of water availability in a region where water is a strategic resource.

Cette étude vise à détecter et cartographier la dégradation et la régénération de la couverture végétale durant ces deux dernières décennies en Haute-Casamance, située dans le sud du Sénégal. Elle s’appuie sur l’analyse des tendances de séries temporelles d’images MODIS NDVI acquises sur la période 2000-2018, sur la cartographie des changements d’occupation et d’utilisation des sols (LULCC) faite à partir d’images Landsat (ETM+ 1999 et OLI 2018), et sur une analyse conjointe des tendances de NDVI et LULCC. L’étude a révélé d’importants changements en termes d’occupation et d’utilisation des sols, notamment la conversion de la forêt à la savane (13,5 %), et de la savane aux zones de culture et d’habitat (7,9 %). Des tendances négatives significatives (p<0,1) du NDVI (en lien avec la conversion forêt-savane et forêt-zones de culture et d’habitat) sont observées sur 2,1% de la zone étudiée et indiquent une dégradation de la végétation en particulier le long de la frontière avec la Gambie, sur l’axe Pata-Ndorna, dans le centre de la zone d’étude, et autour de Médina Gounass. Par ailleurs, des tendances positives significatives (p<0,1) sur 1,8 %  de la surface de la zone d’étude témoignent de la régénération dans le sud et sont liées à la conversion des zones de culture en savane et en forêt. L’analyse conjointe des tendances NDVI et LULCC indique une dégradation significative au sud de Kerewane et de Niaming, sur l’axe Ndorna-Pata, autour de Médina Gounass, et vers Saré Coly ; et une régénération significative à la frontière nord-est, au nord de Mampatim, au sud vers Bagadadji et Médina El Hadji, et entre Linkering-Paroumba. Cette analyse a également révélé que les deux séries d’images (NDVI et Landsat) sont cohérentes spatialement et que l’essentiel des tendances de productivité de la végétation est expliqué par les changements d’utilisation des sols et non par le climat.