Greenhouse gas emissions from natural ecosystems and agricultural lands in sub-Saharan Africa: synthesis of available data and suggestions for further research

Abstract. This paper summarizes currently available data on greenhouse gas (GHG) emissions from African natural ecosystems and agricultural lands. The available data are used to synthesize current understanding of the drivers of change in GHG emissions, outline the knowledge gaps, and suggest future directions and strategies for GHG emission research. GHG emission data were collected from 75 studies conducted in 22 countries (n =  244) in sub-Saharan Africa (SSA). Carbon dioxide (CO₂) emissions were by far the largest contributor to GHG emissions and global warming potential (GWP) in SSA natural terrestrial systems. CO₂ emissions ranged from 3.3 to 57.0 Mg CO₂ ha⁻¹ yr⁻¹, methane (CH₄) emissions ranged from −4.8 to 3.5 kg ha⁻¹ yr⁻¹ (−0.16 to 0.12 Mg CO₂ equivalent (eq.) ha⁻¹ yr⁻¹), and nitrous oxide (N₂O) emissions ranged from −0.1 to 13.7 kg ha⁻¹ yr⁻¹ (−0.03 to 4.1 Mg CO₂ eq. ha⁻¹ yr⁻¹). Soil physical and chemical properties, rewetting, vegetation type, forest management, and land-use changes were all found to be important factors affecting soil GHG emissions from natural terrestrial systems. In aquatic systems, CO₂ was the largest contributor to total GHG emissions, ranging from 5.7 to 232.0 Mg CO₂ ha⁻¹ yr⁻¹, followed by −26.3 to 2741.9 kg CH₄ ha⁻¹ yr⁻¹ (−0.89 to 93.2 Mg CO₂ eq. ha⁻¹ yr⁻¹) and 0.2 to 3.5 kg N₂O ha⁻¹ yr⁻¹ (0.06 to 1.0 Mg CO₂ eq. ha⁻¹ yr⁻¹). Rates of all GHG emissions from aquatic systems were affected by type, location, hydrological characteristics, and water quality. In croplands, soil GHG emissions were also dominated by CO₂, ranging from 1.7 to 141.2 Mg CO₂ ha⁻¹ yr⁻¹, with −1.3 to 66.7 kg CH₄ ha⁻¹ yr⁻¹ (−0.04 to 2.3 Mg CO₂ eq. ha⁻¹ yr⁻¹) and 0.05 to 112.0 kg N₂O ha⁻¹ yr⁻¹ (0.015 to 33.4 Mg CO₂ eq. ha⁻¹ yr⁻¹). N₂O emission factors (EFs) ranged from 0.01 to 4.1 %. Incorporation of crop residues or manure with inorganic fertilizers invariably resulted in significant changes in GHG emissions, but results were inconsistent as the magnitude and direction of changes were differed by gas. Soil GHG emissions from vegetable gardens ranged from 73.3 to 132.0 Mg CO₂ ha⁻¹ yr⁻¹ and 53.4 to 177.6 kg N₂O ha⁻¹ yr⁻¹ (15.9 to 52.9 Mg CO₂ eq. ha⁻¹ yr⁻¹) and N₂O EFs ranged from 3 to 4 %. Soil CO₂ and N₂O emissions from agroforestry were 38.6 Mg CO₂ ha⁻¹ yr⁻¹ and 0.2 to 26.7 kg N₂O ha⁻¹ yr⁻¹ (0.06 to 8.0 Mg CO₂ eq. ha⁻¹ yr⁻¹), respectively. Improving fallow with nitrogen (N)-fixing trees led to increased CO₂ and N₂O emissions compared to conventional croplands. The type and quality of plant residue in the fallow is an important control on how CO₂ and N₂O emissions are affected. Throughout agricultural lands, N₂O emissions slowly increased with N inputs below 150 kg N ha⁻¹ yr⁻¹ and increased exponentially with N application rates up to 300 kg N ha⁻¹ yr⁻¹. The lowest yield-scaled N₂O emissions were reported with N application rates ranging between 100 and 150 kg N ha⁻¹. Overall, total CO₂ eq. emissions from SSA natural ecosystems and agricultural lands were 56.9 ± 12.7  ×  10⁹ Mg CO₂ eq. yr⁻¹ with natural ecosystems and agricultural lands contributing 76.3 and 23.7 %, respectively. Additional GHG emission measurements are urgently required to reduce uncertainty on annual GHG emissions from the different land uses and identify major control factors and mitigation options for low-emission development. A common strategy for addressing this data gap may include identifying priorities for data acquisition, utilizing appropriate technologies, and involving international networks and collaboration.