Projected land-use change emissions surpass climate change-induced carbon sinks in Sub-Saharan African biomes

Abstract

Land-use and land-cover change (LULCC) and climate change (CC) effects on carbon dynamics in Sub-Saharan Africa (SSA) have the potential to transition the region’s carbon balance from sink to source. Future shared socioeconomic pathways (SSPs) for SSA show high vulnerability to CC, high urbanisation, and cropland and pasture demands expected to rise in the coming decades. Yet the realisations of these pathways may be biome-specific as these are expected to behave differently when it comes to carbon dynamics and also be affected by different demands. Here, we examine the impacts of LULCC and CC on Net Ecosystem Exchange (NEE), Net Primary Productivity (NPP), and carbon turnover time of six SSA biomes, namely tropical rainforest, montane forest, moist savanna, dry savanna, temperate grassland and semi-desert, between 2015–2100 for the five SSPs. We performed four simulations where we (i) kept LULCC and CC constant, (ii) varied LULCC and kept CC constant, (iii) kept LULCC constant and varied CC, and (iv) varied both LULCC and CC and calculated their effect on NEE, NPP and carbon turnover time. We find that LULCC effects supersede those of CC, resulting in a net carbon source despite the sink effect of CC. LULCC alone drives major carbon losses across all scenarios, with moist and dry savannas resulting in the strongest carbon sources with cumulative NEE values of 74.93 and 44.85 PgC by 2100 under SSP4. In contrast, CC alone led to consistent carbon sinks across all scenarios in the semi-desert and tropical rainforest, with the strongest sinks observed under SSP5 (−8.36 PgC and − 3.77 PgC, respectively). Moist savanna also acted as a sink under SSP5 (−4.82 PgC) and SSP3 (−1.73 PgC). Combined effects largely reflect LULCC trends, confirming its dominant role in shaping future carbon dynamics. We further observe that intensifying LULCC may decrease biome carbon turnover time by 50 %, especially in tropical rainforests under SSP3 and SSP4. These results suggest that the CC-induced carbon sinks will unlikely dampen the increasing emissions from LULCC in SSA that may not be offset by 2100. Ultimately, our study demonstrates that sustaining SSA as a carbon sink will require urgent, biome-specific land management policies that integrate socioeconomic realities with the region’s evolving climate and development pathways.