Globally significant greenhouse-gas emissions from African inland waters

Abstract

Carbon dioxide emissions to the atmosphere from inland waters-streams, rivers, lakes and reservoirs-are nearly equivalent to ocean and land sinks globally. Inland waters can be an important source of methane and nitrous oxide emissions as well, but emissions are poorly quantified, especially in Africa. Here we report dissolved carbon dioxide, methane and nitrous oxide concentrations from 12 rivers in sub-Saharan Africa, including seasonally resolved sampling at 39 sites, acquired between 2006 and 2014. Fluxes were calculated from published gas transfer velocities, and upscaled to the area of all sub-Saharan African rivers using available spatial data sets. Carbon dioxide-equivalent emissions from river channels alone were about 0.4 Pg carbon per year, equivalent to two-thirds of the overall net carbon land sink previously reported for Africa. Including emissions from wetlands of the Congo river increases the total carbon dioxide-equivalent greenhouse-gas emissions to about 0.9 Pg carbon per year, equivalent to about one quarter of the global ocean and terrestrial combined carbon sink. Riverine carbon dioxide and methane emissions increase with wetland extent and upland biomass. We therefore suggest that future changes in wetland and upland cover could strongly a ect greenhouse-gas emissions from African inland waters.C limate predictions necessitate a full and robust account of natural and anthropogenic greenhouse-gas (GHG) fluxes, especially for CO 2 (refs 1-3), CH 4 (ref. 4) and N 2 O (ref. 5), which together accounted for 94% of the anthropogenic global radiative forcing by well-mixed GHGs in 2011 relative to 1750 (ref. 6). Inland waters (streams, rivers, lakes and reservoirs) are increasingly recognized as important sources of GHGs to the atmosphere, with global CO 2 and CH 4 emissions estimated at 2.1 PgC yr −1 (ref.3) and 0.7 PgC yr −1 (CO 2 -equivalents; CO 2 e) (ref. 4) (1 Pg = 10 15 g), respectively. Considering that the oceanic and land carbon (C) sinks correspond to ∼1.5 and ∼2.0 PgC yr −1 (ref. 7), respectively, the GHG flux from inland waters is significant in the global C budget.In a recent global compilation of inland CO 2 data 3 , <20 data points (out of 6,708, that is, <0.3%) represented African inland waters (with the exception of South Africa, which has been densely sampled), even though they account for ∼12% of both global freshwater discharge 8 and riverine surface area 3 , and include some of the largest rivers and lakes in the world. Equally for the global CH 4 database, there is a strong under-representation of tropical inland waters, whereby a recent synthesis 4 resorted to extrapolating CH 4 fluxes from temperate rivers.The prevailing large uncertainty involved in GHG flux estimates for inland waters, essentially due to the paucity of available data, is coupled to a poor understanding of underlying processes, both of which preclude gauging of future fluxes in response to human pressures. In particular, there is a need to further understand the link between inland water GHG fluxes and ...

Effects of agricultural land use on fluvial carbon dioxide, methane and nitrous oxide concentrations in a large European river, the Meuse (Belgium)

“…Finally, in order to further characterize human-impacted river networks compared to relatively pristine ones, and also to compare temperate and tropical rivers we compared CH 4 and N 2 O in the Meuse with our own data-sets obtained in the Congo and Zambezi rivers (Borges et al, 2015a;Teodoru et al, 2015) (Fig. 12).…”

“…Inland waters are important players in the global budgets of longlived green-house gases (GHGs), acting as vigorous sources to the atmosphere of carbon dioxide (CO 2 ) (Raymond et al, 2013;Lauerwald et al, 2015;Borges et al, 2015a), methane (CH 4 ) (Bastviken et al, 2011;Borges et al, 2015a;Stanley et al, 2016), and nitrous oxide (N 2 O) (Seitzinger and Kroeze, 1998;Hu et al, 2016). The largest fraction of global CO 2 and CH 4 emissions from riverine networks occurs at tropical and sub-tropical latitudes (Bloom et al, 2010;Raymond et al, 2013;Lauerwald et al, 2015;Borges et al, 2015b) that are in general more pristine than their temperate counter-parts.…”

“…In pristine river networks, CO 2 and CH 4 emissions are driven by instream production related to the degradation of terrestrial organic matter (Cole and Caraco, 2001;Richey et al, 2002), as well as lateral inputs from groundwater and/or wetlands (Abril et al, 2014;Borges et al, 2015aBorges et al, , 2015b. Pristine rivers are usually nitrogen poor and seem to be low sources or even sinks of N 2 O, related to sediment denitrification that removes N 2 O from the water column (Richey et al, 1988;Baulch et al, 2011;Borges et al, 2015a).…”

“…Pristine rivers are usually nitrogen poor and seem to be low sources or even sinks of N 2 O, related to sediment denitrification that removes N 2 O from the water column (Richey et al, 1988;Baulch et al, 2011;Borges et al, 2015a).…”

“…Enhanced nutrient inputs will fuel primary production leading to low CO 2 and high CH 4 concentrations, the latter related to enhanced organic matter delivery to sediments . Other human impacts that affect carbon and nitrogen cycling in river networks that can potentially influence cycling of GHGs are river bank stabilization and floodplain drainage that disrupt the river-wetland connectivity that is important for CO 2 and CH 4 dynamics in rivers (Abril et al, 2014;Teodoru et al, 2015;Borges et al, 2015aBorges et al, , 2015bSieczko et al, 2016). The introduction of invasive animal species such as the zebra mussel (Dreissena polymorpha) in US rivers and lakes (Caraco et al, 1997;Evans et al, 2011) (Hussner, 2012), some with high production and biomass (Hussner, 2009); invasive floating macrophytes such as the water hyacinth (Eichhornia crassipes) have been documented to increase CO 2 and CH 4 levels in tropical rivers (Koné et al, 2009(Koné et al, , 2010, but this remains undocumented in temperate rivers.…”