Depth-specific variation of particulate organic carbon composition in the Amazon River

Sarah Rosengard, Marine Chemistry & Geochemistry
Advisor: Valier Galy, Marine Chemistry & Geochemistry



The Amazon River Basin, a global biodiversity hotspot, is one of the greatest reservoirs for reactive organic carbon on the planet. Encompassing an area of six million km2, it includes eight major tributaries that drain through several land cover types: flooded rainforest, drier savannah, high-elevation Andes, and more recently, cropland and pasture. Relatedly, the river network supplies ~6.3 x 1012 m3 water (Molinier et al. 1996) and ~14 x 1012 g particulate organic carbon (POC) (Richey et al. 1990) annually to the coastal Atlantic, comprising 20% of the global freshwater flux and ca. 8% of global annual riverine organic carbon export to the oceans. Accurately tracking the fate of this organic carbon transported from source to sea is significant for understanding the current state of the global carbon budget.
Several processes govern the fate of POC in river discharge: it can be respired in transit (Richey et al. 1990) or in the coastal Atlantic Ocean (Aller et al. 1996), buried for the shorter-term on the river bed (Guyot et al. 1996), or buried for longer time scales on the continental shelf (Kwon et al. 2009). The balance between these processes would reveal whether the Amazon River system is a net sink or source of atmospheric CO2 (Aufdenkampe et al. 2007).
Attempts to quantify each side of the balance have relied on accurately measuring carbon fluxes and describing the POC composition in suspended sediments. The majority of measurements were taken at Óbidos, the most downstream main-stem gaging station located 900 km upstream of the mouth. The assumption that quantities measured at Óbidos represent the entire river basin is complicated by several physical factors, one of which is variation in POC concentration and composition with depth. For large rivers like the Amazon, hydrodynamic sorting causes higher grain-size sediments to settle faster and thus concentrate at deeper depth (Bouchez et al. 2010; Rouse 1950). Bouchez et al.(2011) observed that at Óbidos, where the river is >50 m deep and several km wide, these deeper, coarser sediments had a different elemental composition than shallower, finer sediments. This proposal seeks funding to characterize the composition of POC in similar sediment profiles at Óbidos, and address two broader objectives of my thesis, comparing the residence time of POC across depths, and evaluating the accuracy of using such measurements to estimate the entire basin’s contribution to the coastal Atlantic Ocean.
Last year, my advisors and I visited Óbidos in April, at the rising water stage, and July, the falling water stage. During both trips, we collected 10-12 10-L water samples at specific depths and distances across the river cross-section, one large-volume (>100 L) sample at the surface, and sediments at the river bed. >0.2 μm diameter particles from the water samples were filtered and frozen within 1-2 days after collection. Sampling was coupled to the Sontek M9 Acoustic Doppler Current Profiler (ADCP), which provided velocity profiles from left bank to right bank, surface to bed.
The suspended sediments were freeze-dried before analysis at WHOI. The three levels of organic matter analysis employed are: (1) bulk (% organic C, % organic N, δ13C, δ15N, grain size, 14C), (2) compound-specific (lipid abundance, δ13C, 14C in the large-volume samples only), and (3) thermal stability derived from ramped oxidation of organic matter to CO2 (δ13C, 14C) (Rosenheim et al. 2008). These data allude to the source of the organic matter, its degradation state, and relatedly, its residence time in the catchment. Like prior studies (Mayorga et al. 2005), these interpretations will improve understanding of the river’s contribution to the coastal Atlantic Ocean. Unlike prior studies, these data will also provide a more accurate representation of Óbidos in the carbon budget of the Amazon system.
Preliminary Results
ADCP-derived discharge increased from 150,000 – 255,000 m3/s between April and July. We observed that compound-specific δ13C, bulk C/N and δ13C shifted with season, pointing to physical changes in the source of POC from the different tributaries (Moreira-Turcq et al. 2003). As expected (Bouchez et al. 2010; 2011), suspended sediment concentration increased with depth in our field samples, as well. % organic C decreased with depth, most likely due to dilution by more inorganic material in the deep, coarse grains. C/N ratios and δ13C also varied by depth, suggesting sorting of POC by source or degradational extent. By contrast, thermograms (CO2 vs. temperature) of the 10-L samples from ramped oxidation (100-700°C at 5°C/min.) do not vary with depth or time. If one uses thermal stability as a proxy for organic carbon lability (Rosenheim and Galy 2012), the relatively invariant thermograms suggests that the degradation history of POC does not change with depth, despite the bulk data showing variable source material.
Future Research Plans
To complete my graduate thesis, we plan to combine the bulk, compound-specific and thermal stability data with 14C content to better constrain the source of the different pools of POC we have characterized thus far. Quantifying 14C will also improve our understanding of their residence times in the Amazon basin. I plan to collaborate with the NOSAMS facility to learn how to prepare these samples before submitting them to the AMS. Both my advisors’ grant and the Ocean Ventures Fund supported the cost of field work and all preliminary measurements presented here. I would like to request financial support for the radiocarbon measurements to complete the dataset, and better connect Óbidos to the river basin and Atlantic Ocean.
Aller et al., Continental Shelf Research 16(5-6), 735-786 (1996).
Aufdenkampe, A.K., et al, Organic Geochemistry 38, 337-364 (2007).
Bouchez, J., et al., Geochemistry, Geophysics, Geosystems 12(3), (2011).
Bouchez, J., et al., Hydrol. Processes 25, 778-794 (2010).
Guyot, J.L., et al, IAHS Publication, 55-64 (1996).
Kwon, E.Y., et al., Nature Geoscience 2, 630-635 (2009).
Mayorga, E., et al., Nature 436(7050), 538-541 (2005).
Molinier, M., et al., IAHS Publ. 238, 209-222 (1996).
Moreira-Turcq, P., et al., Hydrol. Processes 17, 1329-1344 (2003).
Richey, J.E., et al., Limnology and Oceanography 35(2), 352-371(1990).
Rosenheim, B.E. and V. Galy, Geophysical Research Letters 39(19), (2012).
Rosenheim, B.E., et al., Geochemistry, Geophysics, Geosystems 9(4), (2008).
Rouse, H., Engineering Hydraulics, Wiley, New York, (1950).