Africa had been thought to be a potentially large carbon sink — of great value in efforts to mitigate carbon dioxide emissions. But an analysis now reveals that it could be a net source of greenhouse gases that will increase global warming.
One of the biggest challenges in curbing climate change is to obtain robust estimates of greenhouse-gas emissions and sequestration. Such estimates are necessary for predicting the effects of land-use-related mitigation efforts on the release and uptake of the three main greenhouse gases — carbon dioxide, nitrous oxide and methane — produced by complex terrestrial processes. Writing in Biogeosciences, Valentini et al.1 have risen to this challenge by providing a full greenhouse-gas assessment for Africa. Their paper addresses the burning need for evidence-based knowledge2 to devise effective policies for reducing the impact of human activities on climate and the effects of climate change on societies, and to decide where investment is required to meet those aims.
Few reliable estimates of carbon flux are available for Africa, because information about its carbon budget is partial, fragmented and inconsistent. Furthermore, greenhouse-gas budgets generally vary in precision and clarity because of the different methods used to produce them. Existing data are therefore subject to many uncertainties associated with incomplete data sets, short observation periods and unknown fluxes from land-use change and forest fires. Other problems include the limited data available to calibrate the results of modelling studies, and the empirical assumptions made about factors such as deforestation rates, carbon density in major pools and losses of carbon from soil.
Valentini and colleagues used several approaches to solve these problems. They include: data collection from ground-level observations, such as studies of biomass, tree measurements and ecosystem inventories of gas fluxes (Fig. 1); satellite-based data gathering; modelling of carbon fluxes; and derivation of emissions data from secondary sources of information, such as records of agricultural activity. In this way, the authors show that Africa is central to global mitigation efforts because of its large forested areas, the diversity of its ecosystems and the volume of emissions3 it produces as a result of land use and land-use change.
Until now, scientific opinion has held that Africa could help to reduce emissions or sequester carbon if deforestation of large areas, such as the Congo Basin, could be avoided, or if tree and forest cover could be increased through sustainable practices, for example agroforestry and plantation management. The authors concur that Africa is a small carbon sink on an annual timescale. But, more surprisingly, they find that it may be a net source of radiative forcing — reradiation of heat back towards Earth's surface by greenhouse-gas molecules — when methane and nitrous oxide are included in the annual budgeting.
One of the merits of this study is that Valentini et al. disaggregate their data, determining the flux of greenhouse gases from different carbon pools (biomass and soil carbon) and land-cover classes (such as croplands, forests and shrub lands), and from the sources and sinks associated with various forms of land use. Furthermore, they establish the temporal evolution of the net budget for different forms of land use in the four sub-Saharan African regions. They find that the emission rate of carbon dioxide from the continent was nearly constant between 1980 and 2009, but that there were regional variations: western Africa had the largest increase of emissions and eastern Africa the lowest; only central Africa had decreased emissions during that period. These differences might reflect the regional frequency and magnitude of vegetation fires, which are used mostly to expand cropland and shift areas of cultivation. The use of firewood, or of fire in land management, leads to substantial carbon emissions because biomass is rapidly oxidized.
Are these conclusions merely implications associated with data uncertainty, or plausible outcomes of an integrated approach? There is previous evidence to suggest the latter, such as the finding4 that emissions produced by tropical deforestation are only partially offset by the carbon sequestration associated with forest regrowth.
Valentini and colleagues' results are crucial to the debate about greenhouse-gas emissions related to land use in Africa2, where most 'game changers' are occurring through shifts in either agricultural development models (family farming to agri-business; labour intensive to capital intensive; small scale to large scale) or production systems (farming and agriculture versus conservation)5. These structural changes in agricultural systems were not considered explicitly in the paper, but will need more attention if we are to work out precisely how each one affects greenhouse-gas emissions. If emissions tend to increase because of these changes, then strategies could be developed to address either the specific land-use processes that cause emissions, or, most importantly, the drivers of change. This approach is central to climate-smart agriculture, an emerging concept that minimizes greenhouse-gas emissions by encouraging sustainable farming and land-use practices.
This paper could also instigate a turning point in the design of programmes for counteracting climate change in Africa. Such programmes are currently based mainly on the assumption6,7 that agriculture, forestry and other land uses (AFOLU) collectively form the main terrestrial carbon sink. In Africa, however, emissions from land-use change are exceptionally high, and are much greater than those from burning fossil fuels. This makes Africa unique among the continents. However, devising new programmes on the basis of Valentini and co-workers' findings will be difficult, because accounting for the three major greenhouse gases in the AFOLU sector requires high investment and skills not always available in developing countries2,8.
Valentini et al. did not fully consider two main aspects of the African carbon cycle. The first is the key role of agroforestry, which could potentially sequester large amounts of carbon, reduce deforestation and provide many ecosystem services, such as wood fuel9. The second is the part played by peat lands and by vegetation in flood plains. Although such regions are small in terms of surface area, they produce substantial carbon emissions, especially when they undergo land-use changes. This is particularly true when pristine peat lands are converted10 to fields for agriculture or plantations. Further studies are needed to determine whether either of these aspects of the carbon cycle could change the outcome of the authors' findings.
Other future directions of research should include the creation of an inventory of databases of information from in situ studies, as part of a collaborative framework to improve the availability of low-cost greenhouse-gas data. Such frameworks, which are being promoted worldwide, help to harmonize data-collection methods and improve the consistency of the collected data. The advent of new generations of satellites also provides opportunities to substantially improve carbon budgeting for Africa, and possibly the budgeting of other greenhouse gases.
About this article