Forest management is considered as a strategy of mitigation and adaptation to climate change because of its ability to contribute to improve local livelihood and reduce carbon emission from forest. This study investigated the impacts of community-forest management on changes in land use and land cover, vegetation composition and structure as well as carbon stocks in Missirah Forest located in south-eastern Senegal. Land use and land cover change was mapped using Landsat images of 1990, 2003, and 2014 combined with ground truth data.

The direction, rate of change and transition among land use and land cover types were determined. Forest inventory was conducted by sampling randomly from a number of permanent sampling plots. Data was collected from 94 circular plots of 1256m2 in elevated lands and 57 rectangular plots of 400m2 in riparian forest. Biomass data was collected using destructive sampling and carbon stocks estimated by means of a model. Socio-economic data on drivers of deforestation and forest degradation was collected through a structured survey among 136 selected households in five villages using multi-stage sampling. The results of the land use and land cover revealed six classes: riparian forest, tree savanna, shrub savanna, degraded shrub savanna, croplands and settlements. Vegetation types decreased in all periods with the exception of shrub savanna that experienced an increase of 1.46 % between 1990 and 2003. The entire forest cover showed a decrease of 9.08 % between 1990 and 2003 and 13.63 % from 2003 to 2014. Croplands experienced a continual increase with a rate of more than 100% from 1990 to 2003. The transition to less wooded vegetation (31.58 %) was higher than transition to more wooded vegetation (13.91 %). Analysis of Variance (ANOVA) showed a significant difference between species richness in 2002 and 2013 (p-value = 0.0003) which were 50 and 42 respectively. Prescribed species for charcoal production experienced the highest decline in their Importance Value Index (IVI). The mixed model ANOVA applied on structural parameters revealed that parameters showed a significant decline with the exception of stem density suggesting the forest was not recovering from harvests. The average carbon density of the forest was estimated at 34.10 Mg C ha-1. It varied from 71.87 Mg C ha-1in riparian forest to 12.73 Mg C ha-1 in tree savanna and 11.42 Mg C ha-1in shrub savanna. Most of the carbon stock (67 %) was found in five out of the fifty four species. The decreases in forest cover from1990 to 2014 resulted in a loss of 24.43 % of total carbon stocks. Local perceptions indicate a general decline in vegetation quality. Age group and location of communities significantly affected the rating of the level of degradation. Species cited as declining in numbers were those of high economic value and the perceptions were consistent with results of the inventory.

The perceived drivers of vegetation degradation were charcoal production, bush fire, seasonal migration of cattle and illegal logging. Main economic activities and location of communities significantly affected the ranking of the perceived drivers. This study revealed that the conditions under which forest are managed currently do not constitute a sustainable response to deforestation and degradation induced by charcoal production.