Spatial and temporal variation of carbon stocks in a lowland tropical forest in West Africa

Understanding the nature and cause of spatial and temporal variation in forest carbon is critical for implementation of climate mitigation strategies such as REDD. Such knowledge is lacking and hard to acquire in resource poor regions such as West Africa’s Upper Guinea where benefits of such schemes for forest conservation could have great impact.We undertook a systematic and representative survey of an entire Upper Guinea forest – Gola Forest in southeast Sierra Leone - by measuring over 600 plots (0.125 ha) in order to quantify the level of spatial variation in C that might exist within a discrete forest type and relate this to historic and contemporary impacts on the forest. We modelled current C stocks and compared these with values calculated from historic surveys. Mean C content in above ground biomass was c. 160 Mg ha−1. The southern part of the forest which was subject to heavier logging in the 1980s had a lower C content (121–144 Mg ha−1) compared to the less disturbed central areas (186 Mg ha−1). Volumes of extracted timber and distance to settlements around the forest explained 42% of the variation in C content. Elevation, slope and other metrics of human impact such as distance to roads did not explain significant additional variation. A survey from the 1950s recorded much higher carbon content than currently found in the south of the forest. This accords with evidence that commercial logging in this area was destructively high. However, old surveys from the late 1960s/early 1970s in less disturbed areas recorded lower carbon content than present. Most of the past surveys were in areas that had not yet been commercially logged so the accumulation in biomass in the last 40 years implies recovery from a much older disturbance event, or a change in environmental conditions promoting growth.As a typical Upper Guinean forest, our results from Gola demonstrate the long term impacts of disturbance events on carbon stocks and how these can vary greatly at small scales, highlighting the need for representative and regionally relevant empirical data to inform REDD type initiatives. Nonetheless these forests remain important for carbon sequestration and storage in the region despite this history and the parts of the forest with no recorded logging activity that increased in carbon in recent decades with levels now akin to undisturbed plots reported from elsewhere in the region demonstrate the sequestration potential these forests provide once adequately protected.

► Carbon was assessed for an entire Upper Guinean forest in a representative way.
► Above ground carbon density varied spatially from 80 to 230 t per ha.
► Spatial variation reflected disturbance history especially logging and settlements.
► Areas never logged nonetheless gained carbon in the last four decades.
► Logged and farmed Upper Guinean forests remain important carbon sinks.