Jonathan Deborde

Amazon River carbon dioxide outgassing fuelled by wetlands

River systems connect the terrestrial biosphere, the atmosphere and the ocean in the global carbon cycle. A recent estimate suggests that up to 3 petagrams of carbon per year could be emitted as carbon dioxide (CO2) from global inland waters, offsetting the carbon uptake by terrestrial ecosystems. It is generally assumed that inland waters emit carbon that has been previously fixed upstream by land plant photosynthesis, then transferred to soils, and subsequently transported downstream in run-off. But at the scale of entire drainage basins, the lateral carbon fluxes carried by small rivers upstream do not account for all of the CO2 emitted from inundated areas downstream. Three-quarters of the world’s flooded land consists of temporary wetlands, but the contribution of these productive ecosystems to the inland water carbon budget has been largely overlooked. Here we show that wetlands pump large amounts of atmospheric CO2 into river waters in the floodplains of the central Amazon. Flooded forests and floating vegetation export large amounts of carbon to river waters and the dissolved CO2 can be transported dozens to hundreds of kilometres downstream before being emitted. We estimate that Amazonian wetlands export half of their gross primary production to river waters as dissolved CO2 and organic carbon, compared with only a few per cent of gross primary production exported in upland (not flooded) ecosystems. Moreover, we suggest that wetland carbon export is potentially large enough to account for at least the 0.21 petagrams of carbon emitted per year as CO2 from the central Amazon River and its floodplains. Global carbon budgets should explicitly address temporary or vegetated flooded areas, because these ecosystems combine high aerial primary production with large, fast carbon export, potentially supporting a substantial fraction of CO2 evasion from inland waters.

Divergent biophysical controls of aquatic CO2 and CH4 in the World’s two largest rivers

Carbon emissions to the atmosphere from inland waters are globally significant and mainly occur at tropical latitudes. However, processes controlling the intensity of CO2 and CH4 emissions from tropical inland waters remain poorly understood. Here, we report a data-set of concurrent measurements of the partial pressure of CO2 (pCO2) and dissolved CH4 concentrations in the Amazon (n = 136) and the Congo (n = 280) Rivers. The pCO2 values in the Amazon mainstem were significantly higher than in the Congo, contrasting with CH4 concentrations that were higher in the Congo than in the Amazon. Large-scale patterns in pCO2 across different lowland tropical basins can be apprehended with a relatively simple statistical model related to the extent of wetlands within the basin, showing that, in addition to non-flooded vegetation, wetlands also contribute to CO2 in river channels. On the other hand, dynamics of dissolved CH4 in river channels are less straightforward to predict, and are related to the way hydrology modulates the connectivity between wetlands and river channels.

Seasonal Pattern of the Biogeochemical Properties of Mangrove Sediments Receiving Shrimp Farm Effluents (New Caledonia)

Coastal tropical shrimp farming may impact the adjacent ecosystems through the release of large quantities of effluents rich in nutrients. In New Caledonia, mangroves are considered as a natural biofilter to reduce impacts on the surrounding World Heritage listed lagoon. Our main objective was to understand the influence of effluent discharge on the biogeochemistry of mangrove sediments. A monitoring of the physico-chemical parameters of mangrove sediments was carried out during a whole year, including active and non active periods of the farm. The parameters studied were: i) benthic primary production (Chl-a concentrations), ii) physico-chemical parameters of sediments (redox potential, pH, salinity, TOC, TN, TS, δ13C and δ15N), iii) concentrations of dissolved nitrogen, iron and phosphorus. A mangrove developing in the same physiographic conditions, presenting the same zonation, and free of anthropogenic input was used as reference. The concentration of benthic Chl-a measured at sediment surface in the effluent receiving mangrove was twice to three times that measured in the control zone whatever the season. We thus suggest that nutrients inputs significantly increased the phytobenthic production in the effluent receiving mangrove during the whole year, even after the cessation of discharges and because of natural seasonal dynamic of phytobenthos. Although the flow of surface OM was increased, the OM content at depth was not higher than in the control mangrove. However, the contribution of mangrove detritus to the sedimentary organic pool was higher probably as a result of higher density and much greater individual size of the mangrove trees. Unlike the control mangrove sediment, the effluent receiving mangrove sediment was not stratified, redox potential values were high and presence of Fe3+ was detected down to 50 cm depth, probably as a result of a larger root system, allowing a better sediment oxygenation and accentuated OM decomposition processes, and thus limiting ecosystem saturation.