**3.1 Methodological approach for discharge measurement in large rivers**

Muste & Merwade (2010) describe recent advances in the instrumentation used for investigations of river hydrodynamics and morphology including acoustic methods and remote sensing. These methods are revolutionizing the understanding, description and modelling of flows in natural rivers.

Challenges and Solutions for Hydrodynamic and Water Quality in Rivers in the Amazon Basin 73

net discharge monitoring stations in the U.S. These stations provide a near real-time data flow from all stations on the Internet, generating data of water column depth, or the flow

However, a significant challenge for large rivers is the fact that the channel bed, sediment and/or sand banks move location over time. Thus, the discharge-curves need to be updated through regular measurements of the depth, breadth and speed of the river at the monitoring stations. Such difficulties have led to the replacement of conventional techniques with measurements of velocity and net discharge. But at the same time, a vast storage capacity for the data obtained is required, especially when the ADCP data are combined with topography. For example, the data storage requirements have increased from the order of hundreds of thousands to the order of millions of bathymetric and hydrodynamic

This obstacle requires massive investments in instruments with extraordinary data processing abilities in order to store, group, process and quickly distribute data in a myriad of different formats to fulfil information needs of users. On the other hand, the numerical models developed to accommodate the 3D information of hydrodynamics and bathymetry

Dinerhart & Burau (2005) used the ADCP in the Sacramento River (CA/USA) in diurnal tidal rivers for mapping velocity vectors and indicators of suspended sediment. They observed that in surface waters, the ADCP is particularly useful for quickly measuring the current discharge of large rivers with non-permanent flows, presenting several advantages

Another important parameter in the biogeochemical cycle of aquatic ecosystem is the longitudinal dispersion. The longitudinal dispersion coefficient (D) is an important parameter needed to describe solute transport along river currents (Shen et al., 2010). This parameter is usually estimated with tracers. For economic and logistical reasons, the use of the

The same authors argue that these shortcomings can be overcome with the use of ADCP simultaneously with tracers in the stretch of river, by examining the conditions under which both methods produce similar results. Thus, ADCP appears to be an excellent alternative / addition to the traditional tracer based method, provided that care is taken to avoid spurious data in the computation of weighted average distances used in the representation

Stevaux et al. (2009) studied the structure and dynamic of the flow in two large Brazilian rivers (the Ivaí and the Paraná) using eco-bathymetry and ADCP together with samples of suspended sediments. This occurred in two phases of the hydrological cycle (winter and summer). The methodology proved to be valid and easily transferable to other river systems of similar dimensions. For example, at the confluences of river estuaries with complex hydraulic interactions resulting from the integration of two or more different flows, constituting a "competition and interaction" environment. This is because continuous changes occur in flow velocity, discharge and structure, in addition to the changes in the physical and chemical properties of water quality and channel morphology. These dynamic systems are very important in river ecology, reflecting many features and limiting

According to Stevaux et al. (2009), from a hydrological perspective, the confluences can be considered as likely sites of turbulence and convergent or divergent movements, forming upward, downward or lateral vortices. These effects generate chaotic motion, generating

is only available with the use of intensive techniques like ADCP or MBEs.

such as visualisations of time based flows and sediment dynamics in tidal rivers.

stage discharge-curves at each location.

information points (Muste & Merwade, 2010).

latter is prohibitive in large rivers.

conditions of the environment.

of the average conditions of the river stretch in question.

Stone & Hotchkiss (2007) report that accurate field measurements of shallow river flows are needed for many applications including biological research and the development of numerical models. Unfortunately, data quantifying the velocity of river current are difficult to obtain due to the limitations of traditional measurement techniques. These authors comment that the mixing processes and transport of sediment are among the most important impacts on aquatic habitat. The velocity of large rivers is typically measured in either stationary or moving boats with reels or ADCP (Acoustic Doppler Current Profiler).

ADCPs are designed to measure the velocity of the current in a section of a watercourse, producing a velocity profile of the section based on the principle of sound waves from the Doppler effect. This effect is a result of the change in the frequency of the echo (wave) which varies with the motion of the emitting source or reflector. Using this technique, it is possible to measure more accurately net discharge i) in sites and ii) on occasions where the task of measuring flow is more difficult with traditional techniques. At the same time results from ADCP are comparable with results from techniques using traditional methods and can be used to evaluate the qualitative discharge of suspended sediments. In both cases, the technique can be applied in specific monitoring programs (Abdo et al., 1996).

The ADCP has some technical advantages over more traditional techniques (e.g. quantitative net discharge) in places where there are difficulties in applying traditional methods, such as large rivers, during the wet season, discontinuous river sections, and some authors recommend that its use should become more common in estuaries (Guennec & Strasser, 2009). The main advantages of using ADCP are: a greater quantity and quality of data, improved accuracy (5%); measurements are obtained in real time, with a high rate of reproductibility. The technique for measuring liquid discharge using ADCP technique is also faster is also faster than conventional methods and can be used in large and small water bodies. Furthermore, it requires less effort, does not need alignment, allows for the correction of detours in discrete river sections, and estimating the motion of sediment on the river bed. It also demonstrates a good correlation with the more conventional methods, permitting to obtain of section profiles, river width, flow velocity, the qualitative distribution of suspended sediments, measurement time, boat speed, water temperature and salinity (Guennec & Strasser, 2009).

According to Muste & Merwade (2010) recent advances in instrumentation for the analysis of river flows include the combination of acoustic methods with remote sensing to quantify variables and hydrodynamic and morphological parameters in natural bodies of water, and notably the degree of importance of these new technologies is more evident when applied to large rivers under tidal influence (Abdo et al. 1996; Martoni & Lessa, 1999).

These instruments can be quick and efficient in providing detailed multidimensional measures that contribute to the investigation of complex processes in rivers, especially hydrodynamics, sediment transport, availability of habitats and ecology of aquatic ecosystems.

Muste & Merwade (2010) describe how to quantify the hydrodynamic characteristics and morphology of complex channels. In addition to the ability to extract information that is available through conventional methods in the laboratory, the ADCP and MBES (Multibeam Echosounder) can provide additional information that is critical to the understanding and development of modelling processes in rivers, for example providing a 3D view of river hydrodynamics that was previously unavailable to hydrological studies of large rivers.

A major challenge for studies involving large rivers in the Amazon is the operation of flow meters. For example, the United States Geological Survey (USGS) operates more than 7,000

Stone & Hotchkiss (2007) report that accurate field measurements of shallow river flows are needed for many applications including biological research and the development of numerical models. Unfortunately, data quantifying the velocity of river current are difficult to obtain due to the limitations of traditional measurement techniques. These authors comment that the mixing processes and transport of sediment are among the most important impacts on aquatic habitat. The velocity of large rivers is typically measured in either stationary or moving boats with reels or ADCP (Acoustic Doppler Current Profiler). ADCPs are designed to measure the velocity of the current in a section of a watercourse, producing a velocity profile of the section based on the principle of sound waves from the Doppler effect. This effect is a result of the change in the frequency of the echo (wave) which varies with the motion of the emitting source or reflector. Using this technique, it is possible to measure more accurately net discharge i) in sites and ii) on occasions where the task of measuring flow is more difficult with traditional techniques. At the same time results from ADCP are comparable with results from techniques using traditional methods and can be used to evaluate the qualitative discharge of suspended sediments. In both cases, the

technique can be applied in specific monitoring programs (Abdo et al., 1996).

salinity (Guennec & Strasser, 2009).

ecosystems.

The ADCP has some technical advantages over more traditional techniques (e.g. quantitative net discharge) in places where there are difficulties in applying traditional methods, such as large rivers, during the wet season, discontinuous river sections, and some authors recommend that its use should become more common in estuaries (Guennec & Strasser, 2009). The main advantages of using ADCP are: a greater quantity and quality of data, improved accuracy (5%); measurements are obtained in real time, with a high rate of reproductibility. The technique for measuring liquid discharge using ADCP technique is also faster is also faster than conventional methods and can be used in large and small water bodies. Furthermore, it requires less effort, does not need alignment, allows for the correction of detours in discrete river sections, and estimating the motion of sediment on the river bed. It also demonstrates a good correlation with the more conventional methods, permitting to obtain of section profiles, river width, flow velocity, the qualitative distribution of suspended sediments, measurement time, boat speed, water temperature and

According to Muste & Merwade (2010) recent advances in instrumentation for the analysis of river flows include the combination of acoustic methods with remote sensing to quantify variables and hydrodynamic and morphological parameters in natural bodies of water, and notably the degree of importance of these new technologies is more evident when applied to

These instruments can be quick and efficient in providing detailed multidimensional measures that contribute to the investigation of complex processes in rivers, especially hydrodynamics, sediment transport, availability of habitats and ecology of aquatic

Muste & Merwade (2010) describe how to quantify the hydrodynamic characteristics and morphology of complex channels. In addition to the ability to extract information that is available through conventional methods in the laboratory, the ADCP and MBES (Multibeam Echosounder) can provide additional information that is critical to the understanding and development of modelling processes in rivers, for example providing a 3D view of river hydrodynamics that was previously unavailable to hydrological studies of large rivers. A major challenge for studies involving large rivers in the Amazon is the operation of flow meters. For example, the United States Geological Survey (USGS) operates more than 7,000

large rivers under tidal influence (Abdo et al. 1996; Martoni & Lessa, 1999).

net discharge monitoring stations in the U.S. These stations provide a near real-time data flow from all stations on the Internet, generating data of water column depth, or the flow stage discharge-curves at each location.

However, a significant challenge for large rivers is the fact that the channel bed, sediment and/or sand banks move location over time. Thus, the discharge-curves need to be updated through regular measurements of the depth, breadth and speed of the river at the monitoring stations. Such difficulties have led to the replacement of conventional techniques with measurements of velocity and net discharge. But at the same time, a vast storage capacity for the data obtained is required, especially when the ADCP data are combined with topography. For example, the data storage requirements have increased from the order of hundreds of thousands to the order of millions of bathymetric and hydrodynamic information points (Muste & Merwade, 2010).

This obstacle requires massive investments in instruments with extraordinary data processing abilities in order to store, group, process and quickly distribute data in a myriad of different formats to fulfil information needs of users. On the other hand, the numerical models developed to accommodate the 3D information of hydrodynamics and bathymetry is only available with the use of intensive techniques like ADCP or MBEs.

Dinerhart & Burau (2005) used the ADCP in the Sacramento River (CA/USA) in diurnal tidal rivers for mapping velocity vectors and indicators of suspended sediment. They observed that in surface waters, the ADCP is particularly useful for quickly measuring the current discharge of large rivers with non-permanent flows, presenting several advantages such as visualisations of time based flows and sediment dynamics in tidal rivers.

Another important parameter in the biogeochemical cycle of aquatic ecosystem is the longitudinal dispersion. The longitudinal dispersion coefficient (D) is an important parameter needed to describe solute transport along river currents (Shen et al., 2010). This parameter is usually estimated with tracers. For economic and logistical reasons, the use of the latter is prohibitive in large rivers.

The same authors argue that these shortcomings can be overcome with the use of ADCP simultaneously with tracers in the stretch of river, by examining the conditions under which both methods produce similar results. Thus, ADCP appears to be an excellent alternative / addition to the traditional tracer based method, provided that care is taken to avoid spurious data in the computation of weighted average distances used in the representation of the average conditions of the river stretch in question.

Stevaux et al. (2009) studied the structure and dynamic of the flow in two large Brazilian rivers (the Ivaí and the Paraná) using eco-bathymetry and ADCP together with samples of suspended sediments. This occurred in two phases of the hydrological cycle (winter and summer). The methodology proved to be valid and easily transferable to other river systems of similar dimensions. For example, at the confluences of river estuaries with complex hydraulic interactions resulting from the integration of two or more different flows, constituting a "competition and interaction" environment. This is because continuous changes occur in flow velocity, discharge and structure, in addition to the changes in the physical and chemical properties of water quality and channel morphology. These dynamic systems are very important in river ecology, reflecting many features and limiting conditions of the environment.

According to Stevaux et al. (2009), from a hydrological perspective, the confluences can be considered as likely sites of turbulence and convergent or divergent movements, forming upward, downward or lateral vortices. These effects generate chaotic motion, generating

Challenges and Solutions for Hydrodynamic and Water Quality in Rivers in the Amazon Basin 75

Fig. 2. Net discharge measured with ADCP (600Hz). a) North Channel profile, with Q = 1.3x105 m3/s (12/05/2010), b) South Channel, with Q = 1.2x105 m3/s (12/07/2010).

approximately 30 Tg C yr-1 into the atmosphere.

**4. Carbon and nutrient biogeochemistry at the Mouth of the Amazon River** 

Brito (2010) prepared a review based on some of the main studies from the ROCA project that indicated that the tropical North Atlantic Ocean can be considered a source of

But Subramaniam et al (2008) found a carbon sink of similar magnitude with biologically measures of approximately 28 Tg C yr-1 from the atmosphere to the ocean, resulting from nitrogen and phosphorus in the river, in addition to the fixation of N2 in the plume of the Amazon River. Thus, the Amazon plume reverses the normal oceanic conditions, causing carbon capture and sequestration of CO2, defined as the net remover of carbon from the

secondary currents of differing velocities and directions, including some that feedback to the flow main current.

For these authors these dynamics induce the main movement of the sediment formed in the river bed and consequently the main source of variation and alteration in the shape of the channel bed. In this case, highlights the main factors controlling mixing in channel confluences: morphological, such as the confluence angle and asymmetry of the channel bed; and hydraulics, such as momentum and contrast in the flow density. These results also confirmed that the identification and understanding of flow mixing at river confluences is very important in studies of pollution, nutrients, dispersal of dissolved oxygen, and other ecological variables (Rosman, 2007; Bastos, 2010; Cunha, 2008; Pinheiro, 2008). The rate of movement of the flow can be used to determine flow predominance in the confluence.

### **3.2 ADCP Measurements in North Channel**

According to ANA (2008) measurements occurred in the section of the channel, located between *Amapá* Coast and island of *Marajó*, North Channel of the Amazon River, with the total time to perform the measurement of the 12 hours and 40 minutes.

Due to the effect of tidal flow and the flow directions of the channel of the Amazon River at its mouth, to determine the actual flow of the river the flow is continuously measurement during a tidal cycle. Or more precisely, it is necessary to perform measurement during a wave variation of flow generated due to influence of the tide.

Considering the peculiarities of the flow measurement under the influence of the tide and large transect of the North Channel of the Amazon River at its mouth (about 11.9 km), the traditional methods of measuring flow in large rivers do not apply to flow measurement in this situation.

The measurements of flow at the mouth or the Amazon were only possible through the development of equipment for flow measurement by Doppler effect (ADCP) due to the drastic reduction of time required for measurement and reduction of risk associated. Calculation of the total duration of measurement: the total length is determinate by measurement the time difference between initial and final wave flow due to the tidal cycle, expressed in seconds.

Since the wave due to the tidal cycle can be represented by a periodic wave, the start and end times are obtained by the abscissas corresponding to the intersection points of the flow versus time curve with a straight parallel to the axis "x" vs Q (discharge), whose first derivative has the same sign, i.e. at points where the curve is increasing (positive derivative) or descending (negative derivative). The calculation of the actual flow in the section is done by determining the area under curve Q x t, which corresponds to the total volume of water that passed through the measurement section during the period of the wave flow, divided by the total time duration. Determination of the flow of the North Channel were performed on different days, the first is the simple sum of three part of effective flow rates measured. The second involves the propagation of waves flow measurements performed by the same reference time, the sum of curves "overlapping" and integrating.

The expected outcome of this type of information is the realisation of regular "in situ" hydrodynamic measurements, to understand the relationship between river hydrodynamics and biogeochemical factors along this key stretch of the Amazon River (Transect T2) near Macapá-AP. The idea is to integrate them to control upstream hydrodynamics and biogeochemistry, as well as to understand how the ecosystem may responds to anthropogenic climate change (Fig. 2).

secondary currents of differing velocities and directions, including some that feedback to the

For these authors these dynamics induce the main movement of the sediment formed in the river bed and consequently the main source of variation and alteration in the shape of the channel bed. In this case, highlights the main factors controlling mixing in channel confluences: morphological, such as the confluence angle and asymmetry of the channel bed; and hydraulics, such as momentum and contrast in the flow density. These results also confirmed that the identification and understanding of flow mixing at river confluences is very important in studies of pollution, nutrients, dispersal of dissolved oxygen, and other ecological variables (Rosman, 2007; Bastos, 2010; Cunha, 2008; Pinheiro, 2008). The rate of movement of the flow can be used to determine flow predominance in the confluence.

According to ANA (2008) measurements occurred in the section of the channel, located between *Amapá* Coast and island of *Marajó*, North Channel of the Amazon River, with the

Due to the effect of tidal flow and the flow directions of the channel of the Amazon River at its mouth, to determine the actual flow of the river the flow is continuously measurement during a tidal cycle. Or more precisely, it is necessary to perform measurement during a

Considering the peculiarities of the flow measurement under the influence of the tide and large transect of the North Channel of the Amazon River at its mouth (about 11.9 km), the traditional methods of measuring flow in large rivers do not apply to flow measurement in

The measurements of flow at the mouth or the Amazon were only possible through the development of equipment for flow measurement by Doppler effect (ADCP) due to the drastic reduction of time required for measurement and reduction of risk associated. Calculation of the total duration of measurement: the total length is determinate by measurement the time difference between initial and final wave flow due to the tidal cycle,

Since the wave due to the tidal cycle can be represented by a periodic wave, the start and end times are obtained by the abscissas corresponding to the intersection points of the flow versus time curve with a straight parallel to the axis "x" vs Q (discharge), whose first derivative has the same sign, i.e. at points where the curve is increasing (positive derivative) or descending (negative derivative). The calculation of the actual flow in the section is done by determining the area under curve Q x t, which corresponds to the total volume of water that passed through the measurement section during the period of the wave flow, divided by the total time duration. Determination of the flow of the North Channel were performed on different days, the first is the simple sum of three part of effective flow rates measured. The second involves the propagation of waves flow measurements performed by the same

The expected outcome of this type of information is the realisation of regular "in situ" hydrodynamic measurements, to understand the relationship between river hydrodynamics and biogeochemical factors along this key stretch of the Amazon River (Transect T2) near Macapá-AP. The idea is to integrate them to control upstream hydrodynamics and biogeochemistry, as well as to understand how the ecosystem may responds to

total time to perform the measurement of the 12 hours and 40 minutes.

wave variation of flow generated due to influence of the tide.

reference time, the sum of curves "overlapping" and integrating.

anthropogenic climate change (Fig. 2).

flow main current.

this situation.

expressed in seconds.

**3.2 ADCP Measurements in North Channel** 

Fig. 2. Net discharge measured with ADCP (600Hz). a) North Channel profile, with Q = 1.3x105 m3/s (12/05/2010), b) South Channel, with Q = 1.2x105 m3/s (12/07/2010).

#### **4. Carbon and nutrient biogeochemistry at the Mouth of the Amazon River**

Brito (2010) prepared a review based on some of the main studies from the ROCA project that indicated that the tropical North Atlantic Ocean can be considered a source of approximately 30 Tg C yr-1 into the atmosphere.

But Subramaniam et al (2008) found a carbon sink of similar magnitude with biologically measures of approximately 28 Tg C yr-1 from the atmosphere to the ocean, resulting from nitrogen and phosphorus in the river, in addition to the fixation of N2 in the plume of the Amazon River. Thus, the Amazon plume reverses the normal oceanic conditions, causing carbon capture and sequestration of CO2, defined as the net remover of carbon from the

Challenges and Solutions for Hydrodynamic and Water Quality in Rivers in the Amazon Basin 77

One of the main unifying conceptual frameworks in biological oceanography is the idea that the structure of the phytoplankton community profoundly affects the export and sequestration of organic material. That is, the biological carbon pump and chemical nutrient cycles (Michaels & Silver, 1988; Wassman, 1988; Peinert et al., 1989; Legendre & Le Fevre,

Cyanobacteria are recognised as a particularly important group of organisms in the carbon cycle, occurring in a wide variety of ecosystems, especially in aquatic environments. Cyanobacteria can survive in extreme conditions and are found in habitats with wide ranges of temperature, salinity and nitrogen availability (Falconer, 2005). The abundance of cyanobacteria varies seasonally, as a consequence of changes in water temperature and solar radiation as well as weather conditions and nutrient supply (Falconer, 2005). It is known that their distribution it is not homogeneous on the surface or in water column. The distribution of cyanobacteria assemblages may vary depending on gradients such as depth, salinity, temperature, space and seasonality. However, as well as a lack of studies in this area, it is difficult to analyze the dynamics and diversity of planktonic groups such as cyanobacteria, especially when the analytic scale is at the species level, where diversity is

In the North and the South Channel in Amapá, sampling of water quality was conducted with i) quarterly and ii) in the Channel North monthly frequency (Brito, 2010). Quarterly collections are used to obtain vertically and horizontally integrated samples for the calculation of dissolved and particulate loads in the water column with the use of 15 to 20 metre boats. The monthly samples are used for seasonal interpolation and are obtained from the surface and 60% of water depth. The measures routine collected are used to derive parameters relating to the ion and nutrient system, carbonates, organic material, water

The depth and surface samples are obtained by immersion pumps, where water is then pumped into a graduated cylinder of 2 litters, which is flooded for at least three times prior

Transect T2 (Table 1) defines the main flow of the river to the north of the island of Marajo. As sea water never passes the dividing line at the mouth of the river in front of the city of Macapá (Nikieme et al., 2007), this is considered a final a purely fluvial compounent of the river.

Left Bank - North Channel Amazon S0 03 32.2 W51 03 47.7

Middle River - North Channel Amazon S0 04 35.9 W51 01 46.7 Right Bank - North Channel Amazon S0 05 01.9 W51 00 21.9

Left Bank - South Channel Amazon S0 09 51.8 W50 37 48.9 Middle River - South Channel Amazon S0 10 43.0 W50 36 59.4

Right Bank - South Channel Amazon S0 11 59.8 W50 35 59.7

1995, Ducklow et al., 2001).

high and the river area to be sampled is enormous.

discharge and suspended sediment.

**4.1 Methods and preliminary biogeochemistry results of transect T2** 

to sampling and the overflow is maintained during the procedures sampling.

Point Local Coordinates

Table 1. Geographical coordinates of sampling sites

atmosphere to the ocean (Dilling, 2003; Battin et al, 2008; Ducklow et al., 2008; Legendre and Le Freve, 1995).

Subramaniam et al (2008) revealed the importance of symbiotic associations of diazotrophic diatoms (DDAs) in nitrogen fixation in the Amazon plume and showed that the chemical outputs associated with these organisms represent a regionally significant carbon sink. DDAs or other agents of N2 fixation have also been found in other tropical river systems, such as the Nile (Kemp et al., 1999), Congo (AN, 1971), the South of China Sea (Voss et al., 2006) and the Bay of Bengal (Unger et al., 2005), and it is speculated that these have global significance, as a previously neglected biological carbon pump.

These results suggest that techniques used to study inland waterways of the Amazon may be applied to other systems e.g. the Amazon plume. However, knowledge about the magnitude, spatial extent and final destination of this plume is limited. The importance of connections with the processes that occur upstream are also very poorly known. Independent measurements of net community production, diazotrophic production and flow of particles near the surface of the plume agree with the export of carbon (Subramaniam et al., 2008), but the ultimate fate of carbon and nitrogen and the sensibility of the plume front to global climate change are currently unknown.

The microbial community is a driving force behind the processing of material along the Amazon continuum, from land to sea. Cole (2007) suggested that the biosphere should be considered as a metabolically active network of sites that are interconnected by a fluvial network.

Despite indications that the organic carbon derived from soil is resistant to degradation on land, remaining stored for decades or centuries (Battin et al., 2008), once released into aquatic ecosystems, there is evidence that this carbon is dissolved rapidly in the rivers in a matter of days or weeks (Cole & Caraco, 2001).

High levels of CO2 and low O2 concentrations are often found in muddy rivers (Brito, 2010) which suggests that organic carbon derived from the soil represents a substantial carbon source for the heterotrophic network of the river ecosystem (Richey, 1990).

Current knowledge about the diversity and dynamics of bacterioplankton comes almost exclusively from studies of lakes (Crump et al., 1999). Various small rivers have also been sampled (Cottell et al., 2005; Crump & Hobbie, 2005), but from 25 of the world's major rivers, only four had their genetic sequences recorded in the bacterioplankton "Genbank" (National Center for Biotechnology and Information, U.S. National Library of Medicine): the Columbia River, USA (Crump et al., 1999), the Changjiang River, China (Sekigushi et al, 2002), the Mackenzie River, Canada (Galand et al., 2008) and the Paraná river, Brazil (Lemke, 2009).

The lack of information regarding bacterioplankton in large rivers limits understanding of global biogeochemical cycles and the ability to detect community responses to biotic and anthropogenic climate impacts in these critical ecosystems (Crump et al., 1999).

In less turbid areas of the continuum, the process of photosynthesis can reduce or even reverse the CO2 emission rate (Dilling, 2003). Likewise, when the river meets the sea, the loss of suspended sediment increases the penetration of light sufficiently to stimulate marine primary production (Smith & Demaster, 1995). Once light has been removed as a limiting factor nutrients released by river "metabolism" allow phytoplankton blooms, whose community structure is probably dependent on concentrations and ratios of (limiting) nutrients such as nitrogen, phosphorus and iron (Dilling, 2003; Subramaniam et al., 2008).

atmosphere to the ocean (Dilling, 2003; Battin et al, 2008; Ducklow et al., 2008; Legendre and

Subramaniam et al (2008) revealed the importance of symbiotic associations of diazotrophic diatoms (DDAs) in nitrogen fixation in the Amazon plume and showed that the chemical outputs associated with these organisms represent a regionally significant carbon sink. DDAs or other agents of N2 fixation have also been found in other tropical river systems, such as the Nile (Kemp et al., 1999), Congo (AN, 1971), the South of China Sea (Voss et al., 2006) and the Bay of Bengal (Unger et al., 2005), and it is speculated that these have global

These results suggest that techniques used to study inland waterways of the Amazon may be applied to other systems e.g. the Amazon plume. However, knowledge about the magnitude, spatial extent and final destination of this plume is limited. The importance of connections with the processes that occur upstream are also very poorly known. Independent measurements of net community production, diazotrophic production and flow of particles near the surface of the plume agree with the export of carbon (Subramaniam et al., 2008), but the ultimate fate of carbon and nitrogen and the sensibility

The microbial community is a driving force behind the processing of material along the Amazon continuum, from land to sea. Cole (2007) suggested that the biosphere should be considered as a metabolically active network of sites that are interconnected by a fluvial

Despite indications that the organic carbon derived from soil is resistant to degradation on land, remaining stored for decades or centuries (Battin et al., 2008), once released into aquatic ecosystems, there is evidence that this carbon is dissolved rapidly in the rivers in a

High levels of CO2 and low O2 concentrations are often found in muddy rivers (Brito, 2010) which suggests that organic carbon derived from the soil represents a substantial carbon

Current knowledge about the diversity and dynamics of bacterioplankton comes almost exclusively from studies of lakes (Crump et al., 1999). Various small rivers have also been sampled (Cottell et al., 2005; Crump & Hobbie, 2005), but from 25 of the world's major rivers, only four had their genetic sequences recorded in the bacterioplankton "Genbank" (National Center for Biotechnology and Information, U.S. National Library of Medicine): the Columbia River, USA (Crump et al., 1999), the Changjiang River, China (Sekigushi et al, 2002), the Mackenzie River, Canada (Galand et al., 2008) and the Paraná river, Brazil

The lack of information regarding bacterioplankton in large rivers limits understanding of global biogeochemical cycles and the ability to detect community responses to biotic and

In less turbid areas of the continuum, the process of photosynthesis can reduce or even reverse the CO2 emission rate (Dilling, 2003). Likewise, when the river meets the sea, the loss of suspended sediment increases the penetration of light sufficiently to stimulate marine primary production (Smith & Demaster, 1995). Once light has been removed as a limiting factor nutrients released by river "metabolism" allow phytoplankton blooms, whose community structure is probably dependent on concentrations and ratios of (limiting) nutrients such as nitrogen, phosphorus and iron (Dilling, 2003; Subramaniam et

anthropogenic climate impacts in these critical ecosystems (Crump et al., 1999).

significance, as a previously neglected biological carbon pump.

of the plume front to global climate change are currently unknown.

source for the heterotrophic network of the river ecosystem (Richey, 1990).

matter of days or weeks (Cole & Caraco, 2001).

Le Freve, 1995).

network.

(Lemke, 2009).

al., 2008).

One of the main unifying conceptual frameworks in biological oceanography is the idea that the structure of the phytoplankton community profoundly affects the export and sequestration of organic material. That is, the biological carbon pump and chemical nutrient cycles (Michaels & Silver, 1988; Wassman, 1988; Peinert et al., 1989; Legendre & Le Fevre, 1995, Ducklow et al., 2001).

Cyanobacteria are recognised as a particularly important group of organisms in the carbon cycle, occurring in a wide variety of ecosystems, especially in aquatic environments. Cyanobacteria can survive in extreme conditions and are found in habitats with wide ranges of temperature, salinity and nitrogen availability (Falconer, 2005). The abundance of cyanobacteria varies seasonally, as a consequence of changes in water temperature and solar radiation as well as weather conditions and nutrient supply (Falconer, 2005). It is known that their distribution it is not homogeneous on the surface or in water column. The distribution of cyanobacteria assemblages may vary depending on gradients such as depth, salinity, temperature, space and seasonality. However, as well as a lack of studies in this area, it is difficult to analyze the dynamics and diversity of planktonic groups such as cyanobacteria, especially when the analytic scale is at the species level, where diversity is high and the river area to be sampled is enormous.
