The Amazon River basin and the waters in the Atlantic Ocean into which the Amazon flows are home to the world’s most diverse ecosystems. This region embodies a rich history of scientific discovery.
During the 1980s, one scientific team discovered that vast amounts of waterborne carbon seemed to simply disappear in transit between the upper and central reaches of the Amazon River and the sea. These researchers, part of the Carbon in the Amazon River Experiment (CAMREX) project, made early observations of organic matter and suspended sediments flowing through the upper and central reaches of the river. By 2002, researchers discovered that most of the carbon escaped the river as carbon dioxide (CO2), a phenomenon now recognized as being globally ubiquitous across inland waters at all latitudes.
What drives these large evasive gas fluxes? How do these processes evolve as the river meets the sea?
From 2010 to 2014, an international team of scientists led by Patricia Yager (University of Georgia) set out to decode the linkage between microbial and biogeochemical processes occurring along the lower reaches of the Amazon River and its plume, a broad swath just offshore where river water mingles with ocean water. The effort was called the River Ocean Continuum of the Amazon (ROCA) project.
During the ROCA project, Yager led a series of cruises through the tropical North Atlantic Ocean and into the river plume. Meanwhile, a team of ROCA collaborators led by Jeff Richey (University of Washington) simultaneously probed the lower reaches of the Amazon River, from the Óbidos downstream gauging station to the mouth of the river, where tides completely reverse the flow of the river. The force of the river is so strong that although tides reverse its flow, water can remain fresh a great distance offshore from the river mouth.
ROCA represented the first systematic effort to connect processes occurring in the lower reaches of the river to those occurring in the ocean plume. Previous understanding of the land-sea connection of the Amazon contained a data gap of some 1,000 kilometers between Óbidos and the river mouth, and there were no temporally overlapping studies in the river and plume.
Our team’s most recent project, which began in 2014, aims at further understanding biogeochemical dynamics in the lower river. This project is dubbed the Trocas Líquidas de Carbono do Ecossistema do Baixo Rio Amazonas: Da Terra para o Oceano e Atmosfera (Net Ecosystem Exchange of the Lower Amazon River: From Land to the Ocean and Atmosphere, or TROCAS).
Taking the Data
The goal of TROCAS is to develop a holistic understanding of how carbon speciation (e.g., carbon dioxide, carbonate minerals, organic matter) evolves as it travels from the landscape, through river networks to the sea, and, in the case of CO2, back to the atmosphere. The research framework is based on the concept of net ecosystem exchange, which tracks the evolution of the partial pressure of dissolved CO2 (pCO2) on a mass balance basis through defined boundaries of the river system.
We have recently completed the sixth TROCAS expedition (Figure 1). The first four TROCAS cruises involved performing measurements while the vessel was underway and doing cross-channel sampling transects along the entire study domain. We started near the river mouth at the city of Macapá and navigated upstream to Óbidos. Then we followed a water mass downstream (an approach called a Lagrangian mode) while also sampling the major clear-water tributaries—the Xingu and Tapajós rivers—each of which discharges a volume of water on the same order of magnitude as the Mississippi River.
During these expeditions, we performed a suite of experiments to measure how quickly different types of organic matter from terrestrial and aquatic plants were converted to CO2. We also investigated processes governing the production, emission, and oxidation of methane in the river, the influence of river hydrodynamics on in situ microbial respiration rates, and the optical signature of organic matter in the river that can be seen from space.
The physical flow of water through this complex and highly dynamic reach of the river is central to all of these questions. We measured river velocity and discharge in situ along the tidally influenced study domain and further compiled these data into a model capable of evaluating biogeochemical transformations (Base System for Environmental Hydrodynamics, SisBaHiA).
Carbon Inputs and Outputs
Data from the ROCA project allowed us to estimate that water took roughly 3–5 days to travel from Óbidos to the mouth. We considered this length of time to be significant relative to the 1–2 weeks it can take for vascular plants to turn over organic matter on the basis of initial incubation experiments.
After adding into the hydrodynamic model the actual river flow across the entire domain, along with bathymetric measurements, we now estimate that complex tidal dynamics extends the water transit time closer to 8–9 days (M. L. Barros et al., unpublished data, 2017). By comparison, more sophisticated incubation experiments showed that it took anywhere from hours to a day for organic matter derived from leachates of different plants to degrade, with organic matter leached from grasses and aquatic plants decomposing several times faster than that from harder wood tissues [Ward et al., 2016].
Continuous measurements made during discharge surveys and throughout the field campaign revealed an intriguing correlation between the river’s flow speed and the concentration of CO2 dissolved in the water. This observation motivated us to develop a shipboard system designed to measure microbial respiration rates under various degrees of mixing. Results from these experiments showed a direct link between microbial respiration and physical mixing rates across the lower Amazon River. Respiration rates measured with this system were an order of magnitude higher than those in past experiments that did not account for river flow and could almost entirely account for measured rates of CO2 outgassing [Ward et al., 2017].
From these insights, we have developed the perspective that although land-derived organic matter is rapidly and continuously degraded to CO2 in the river, constant input from the surrounding land and floodplains maintains high levels of reactive organic matter in the river until these sources are cut off in the inner sectors of the Atlantic Ocean plume.
In fact, measurements in the plume made during the ROCA project revealed observable levels of reactive land-derived organic matter that were degradable during both dark and light incubation experiments. These reactive molecules quickly disappeared as the water became saltier near the ocean, leaving behind relatively stable molecules that persisted throughout the plume.
We conducted experiments with and without light to mimic conditions at various locations in the Amazon River and its plume. The river remains dark below the water’s surface because of its high suspended sediment load, so microbial respiration is the primary pathway for organic matter decomposition upriver. However, as sediments settle in the plume, light can also begin to break down these molecules while also promoting primary production (plants’ conversion of inorganic carbon compounds into organic compounds). The stable molecules that persist throughout the plume might feed the pool of ~5,000-year-old dissolved organic carbon in the deep ocean [Medeiros et al., 2015].
Where the River Meets the Sea
The full suite of ROCA expeditions and the initial TROCAS expeditions laid the groundwork for interpreting chemical and biological signatures across the river-to-ocean continuum. However, we still had to answer one large question before we could accurately constrain fluxes to the ocean and atmosphere: How do tides influence the distribution and transformation of geochemicals near the river mouth?
Although our initial efforts were highly ambitious, they did not truly connect the river to the sea. The lack was due in part to the logistical difficulties involved in large oceanographic vessels taking samples close to shore and small river boats sampling far offshore. For example, an additional 150 kilometers remain between our river end point, Macapá, and the actual river mouth, and waters remain completely fresh more than 60 kilometers offshore from the mouth.
As such, we spent our final two TROCAS expeditions (November 2016, low water, and April 2017, late rising water) exploring as close to the river mouth as logistically possible in our current research vessel, the Mirage, and performing daily time series measurements in fixed locations throughout entire tidal cycles. Measurements made during the last two trips revealed that CO2 and methane concentrations can vary by order(s) of magnitude in small, but not insignificant, side channels, and these tidal effects are seen even in the main stem of the river (the main channel of the river, into which the tributaries flow).
On the most recent voyage, we traveled just beyond the final end point of the geographical river mouth (where water remained entirely fresh throughout the tidal cycle at surface and depth). We are still processing our geochemical measurements, but one striking observation emerged in real time. High levels of pCO2 persisted all the way to the river mouth, and gas fluxes measured with floating chambers here were similar to rates measured even as far upstream as Óbidos.
When scaled up across the lower river domain, these fluxes are significant not only on a basin scale but also globally. The most recent CO2 outgassing estimates by Sawakuchi et al.  suggest that including the lower reaches of the Amazon River in an updated basin-scale budget increases global outgassing estimates by as much as 40% because of the massive surface area that the lower river encompasses as it widens and channelizes.
These estimates still do not include the extension of freshwater into the ocean, 60 kilometers offshore, where surface area is order(s) of magnitude greater than for the river itself. Likewise, CO2 budgets for the plume in the Atlantic Ocean still do not include the inner reaches of the plume and nearshore waters, which likely maintain high levels of CO2 because of continued breakdown of any remaining reactive organic matter from the river.
Working Together to Find Answers
From our long-term involvement in Amazon research, we recognize that fully constraining the cycling of material through Earth systems requires close collaboration across disciplines and cultures. None of the important discoveries made in the Amazon throughout history would have been possible without the partnership of diverse groups of researchers and, of course, faith from funding agencies.
Our current TROCAS project represents a healthy collaboration among Brazilian and U.S.-based funding agencies, universities, national laboratories, and researchers that enabled an ambitious field and analytical effort. Through our efforts, we hope to inspire future generations to continue probing the connection between the land, ocean, and atmosphere to develop a holistic understanding of how Earth functions and responds to change.
Some of this work was presented at the American Geophysical Union’s 2017 Fall Meeting during the session “Progress in Biogeochemical Research of the World’s Large Rivers II” in a talk titled “The influence of tides on biogeochemical dynamics at the mouth of the Amazon River” (Abstract B54D-02).
CAMREX was supported by the National Science Foundation, NASA, and the government of Brazil. ROCA was supported by the Gordon and Betty Moore Foundation Marine Microbiology Initiative. TROCAS is funded by the São Paulo Research Foundation and the National Science Foundation.