You are on page 1of 24

Title: Dissolved oxygen and its response to eutrophication in a tropical black water river

Authors: Tim Rixena, Antje Bauma, Harni Sepryanib, Thomas Pohlmannc, Christine
Joseb, Joko Samiajib

Leibnitz-Zentrum für Marine Tropenökologie, Fahrenheitstr. 6, 28359 Bremen, Germany


University of Riau, Jl. Simpang Panam Km 12.5, Pekanbaru, Riau, Indonesia
Zentrum für Meeres- und Klimaforschung, Institut für Meereskunde, Universität

Hamburg, Bundesstr. 53, 20146 Hamburg, Germany

Corresponding author:
Tim Rixen
Zentrum für Marine Tropenökologie
Fahrenheitstr. 6,
28359 Bremen, Germany
Phone: +49 421 2380055
Fax: +49 421 2380033

The Siak is a typical, nutrient-poor, well-mixed, black water river in central Sumatra,
Indonesia, which owes its brown color to dissolved organic matter (DOM) leached from
surrounding, heavily disturbed peat soils. We carried out a 36-hour experiment in the
province capital Pakanbaru during the rainy season in November 2008. The data obtained

Rixen et al.

Dissolved oxygen and its response to eutrophication.


suggests in conjunction with model results that the decomposition of DOM was the main
factor influencing the oxygen concentration in the Siak which varied between ~100 and
140 µmol l-1. Additional DOM input caused by waste water discharges appeared to
reduce the oxygen concentrations by ~20 µmol l-1 during the peak time of family life in
the early morning and in the early evening. Associated enhanced nutrient inputs appear
to reduce the impact of the anthropogenic DOM by favoring the photosynthetic
production of oxygen in the morning. A reduction of 20 µmol l-1, which although perhaps
not of great significance in Pekanbaru, has strong implications for waste water
management in the fast developing areas downstream Pekanbaru where oxygen
concentrations rarely exceed 20 µmol l-1.

Keywords: Blackwater river, dissolved oxygen, dissolved organic matter, waste water,
Sumatra, Indonesia

1. Introduction
The exponential worldwide population growth and the associated rising food production
results in increased demand for nitrogen, which is a limiting nutrient for plants in many
terrestrial and marine systems (e.g.McElroy, 1983; Matson et al., 1999). In order to make
nitrogen available to plants, dinitrogen gas, which constitutes 78% of our atmosphere, has
to be converted into fixed or reactive nitrogen. According to recent estimates 39% (72.8
TgN yr-1) of the total anthropogenic fixed nitrogen (187 TgN yr-1 in 2005) is carried into
the ocean (Seitzinger et al., 2005; Duce et al., 2008; Galloway et al., 2008). The majority
of anthropogenic nitrogen input (54 TgN yr-1) is deposited as eolian dust in the open

Rixen et al.. often starts to respond to decreasing availability of oxygen when oxygen levels drop below 80 µmol l-1 and mass mortality can occur when the oxygen concentration falls below 22. Howitt et al. Besides eutrophication there are also natural processes such as black water events that lead to anoxic and hypoxic conditions in rivers and estuaries (Hamilton et al.8 TgN yr-1 is carried as dissolved inorganic nitrogen into the coastal ocean. Dissolved oxygen and its response to eutrophication. Subramanian. 2008). If the oxygen concentration falls below 5 µmol l-1 the microbial world also responds and denitrifying bacteria.1 µmol l-1 much lower in the Siak than in other non-black water rivers revealing a mean nitrate concentration of 142 µmol l-1 in South Asia (Baum.4 µmol l-1(Diaz and Rosenberg. which uses nitrate instead of dissolved oxygen for the oxidation of organic matter.Rixen et al. Indonesia (Baum et al. for example. 2001). Dead zones are. 2007. Black water events are flood events during which an enhanced leaching of DOM from leaf litter colors the water dark brown. 1997. 2008 and references therein).. we measured oxygen and DOC concentrations along the Siak and established a 36-hour time-series . 2008. 2008).8 – 65. of course.. 2007). not dead but reveal oxygen levels which are too low to sustain plant and animal life. 2008). at 5. Benthic fauna. become active (Codispoti et al. the subsequent decay thereof reduces the oxygen concentration in the water. The leaching of carbon from organic-rich peat soils and its subsequent decay is a further process that leads to oxygen poor conditions in the black water rivers draining peat soils such as the Siak river in eastern Sumatra. Due to the nutrientpoor peat soils dissolved inorganic nitrogen concentrations were. In order to study the response of an oxygen-poor black water river to eutrophication.. 3 ocean and a smaller share of 18.. where eutrophication is considered to be the main reason for the spreading of “dead zones” (Diaz and Rosenberg.

Tapung Kanan and S. We also developed a box-diffusion model to investigate the oxygen dynamics and interpret the results derived from the 36hour-time-series station. On inter-annual time scales the precipitation rates are influenced by the climate anomaly El Niño/Southern Oscillation (ENSO. Baum. Riau. The Siak passes through the eastern Sumatran lowlands and discharges into the Malacca Strait at river-km 370 km. the main tributary of the Siak. The Siak is one of the main peat-draining rivers in central Sumatra and originates at the confluence of its two headstreams. DOM decomposition and the resulting oxygen consumption were considered to be a main factor influencing the oxygen .. Study Area and Methods Central Sumatra experiences high rainfall and a weakly pronounced seasonality with a dry season (May – September) and a rainy season (October – April) due to the meridional variation of the inter-tropical convergence zone (Fig. et al. 1). Rixen. 2008). 2). S. 2007. originate in the peat swamps. Ropelewski and Halpert. 1987) which in its positive mode was referred to as La Niña in 2008. 1).Rixen et al. For the 36-hour experiment we selected a site in Pekanbaru which is the largest city along the river and the capital of the central Sumatran province. et al. Tapung Kanan and the Mandau. Due to high DOM input from these two peat-draining lowland rivers the DOC concentrations in the Siak increased from approximately 500 to 1300 and from 1300 to 1900 µmol l-1 around the Kanan/Kiri and Mandau junctions (Fig. 2. Tapung Kiri (Fig. The S. Dissolved oxygen and its response to eutrophication. 3a.. 4 station at the Siak during an expedition in November 2008 (Fig.

Oxygen concentration was determined using Winkler titration and we also measured the ammonia concentration shortly after taking the samples according to methods described . 5. 4) which continuously pumped water through a plexi-glas box which was equipped with electrodes to measure oxygen concentration and water temperature (WTW Tetra Con 325_3) every minute. A time series station was set up for 36-hours in Pekanbaru at a ferry boat jetty named “Pelita Pantei” which is hardly in use anymore. 2. 3b).849 inhabitants by 2031. the second largest city along the Siak. Tab.. Dissolved oxygen and its response to eutrophication. 5 concentrations which were inversely related to DOC concentrations along the Siak and dropped to 12 µmol l-1 at the onset of the estuary (Fig. We also took hourly water-samples to determine oxygen. 1). ammonia and DOC concentrations (Fig.Rixen et al. deforestation and conversion into palm oil and rubber estates as well as shrub lands (Laumonier. 3). In the last 4 years we have noticed a growing number of highways and two new bridges which were built downstream Pekanbaru near Siaksriindrapura. 1997. 1 Field Methods In November 2008.777 (in 2006) to 1. an expedition to the Siak was carried out in the course of which water samples were taken to determine DOC and dissolved oxygen concentrations using a Niskin bottle at a water-depth of one meter along the river (Fig. 1. 2006). Today the peat lands are heavily disturbed by drainage. Hooijer et al.284. The population of Pekanbaru has increased dramatically since the 1960s and local authorities estimate a further increase from 671. A submergible pump was deployed at the jetty (Fig. Due to these developments the Siak is slowly losing its role as a major transport path way in the region.

sealed. The oxygen concentrations measured using the Winkler method were used to calibrate the oxygen concentrations measured by the probe.Rixen et al. 5).. (2008) was developed. Current speed and direction were measured from the jetty by a current meter equipped with a conductivity. The water level was determined by a tide gauge fixed to the jetty. Due to heavy rainfall measurements were interrupted at 8 pm November 1th and resumed at 7 am November 2nd . Workhorse ADCP WH-s-1200 plus) equipped with a bottom tracking devices (WHBTK RIV) once due to logistical problems. Dissolved oxygen and its response to eutrophication. The samples were subsequently acidified (20% phosphoric acid) to a pH value of ~2. Unfortunately. 4). (Fig. The DOC concentrations obtained were used as an indicator for the amount of DOM contained in the system. we were only able to use an “Acoustic Doppler Current Profiler” (RD Inst.2 Model Setup In order to get an estimate on the importance of the individual processes which could affect the oxygen concentraions in the river. 6 in Grasshoff (1999). 2. a small box-diffusion model (Eq. The mean current velocities profile obtained across the river shows that current speeds were highest in the centre of the river at a water-depth of between zero and ~ five m and decreased towards the river banks where the Aanderaa current meter was deployed at a water-depth of 1 m (Fig. DOC samples were filtered through 0. In this model the water . temperature and depth sensor (Aanderaa 3231). 1) similar to the one presented by Rixen et al. and stored in darkness at ~4°C until they were analyzed after the expedition.45 µm filters into pre-combusted 20 ml FIOLAX ampoules. DOC was measured pursuant to a high temperature catalytic oxidation method by using a Dohrman DC-190 analyzer.

∂O ∂ ⎛ ∂O ⎞ = ⎜ Av ⎟+S ∂t ∂z ⎝ ∂z ⎠ 2 2 V Oxygen +C Oxygen + Pp (1) ‘AV’ is the diffusion coefficient for which a value of 370 cm2 s-1 was selected. The photosynthetic production of organic carbon and its decomposition were translated into oxygen production and consumption by applying a Redfield (C/O) ratio of 106:138 (Redfield et al.1 cm hr-1. which according to our previous model studies varied between 22. this suggests that diffusion affects the entire water column after approximately ~20 minutes (0.Rixen et al.3 hrs).0 m. 7 column of the river was divided into 50 cm thick layers (∆z) and a short time step of 3 s was applied. (1984) and ‘k’ is the piston velocity. The AV chosen implies a rapid mixing which corresponds with the well-mixed water body seen in the salinity and temperature profiles presented by Rixen et al.. SOxygen is the oxygen source term in the surface layer derived from the oxygen flux through the air-water interface (F Oxygen) by means of: S Oxygen =F Oxygen / ∆z. Given a diffusion velocity of u diff = 2 ⋅ AV / t and a mean water-depth of 9. 1963). Dissolved oxygen and its response to eutrophication.9 to 28. . (2008). The photosynthetic production of organic matter is indicated by Pp whereas values > zero indicate a net photosynthetic production of organic matter and values < zero represent respiration of the photosynthesized organic matter. The oxygen flux (FOxygen) is driven by the oxygen partial pressure (pO2) difference between the river and the atmosphere and was calculated according to Fick’s law: FOxygen=k * α (pO2-Atmosphere – pO2-River) (2) ‘α’ is the temperature and salinity-dependent solubility coefficient of oxygen (α=[O2]/ pO2) which was calculated according to Benson & Krause Jr.

016 hr-1.1 Oxygen and DOC Concentrations along the River The data measured during the expedition in November 2008 revealed almost the same trend as those measured during our previous expeditions with an increase and decrease of the DOC and oxygen concentrations along the river (Fig. et al. 3.73) (3) DOC(t) and DOCt0 were the DOC concentrations at a certain time (t) and at the beginning of the experiment (t0). 2008).8 mol of dissolved oxygen is used to oxidize 1 mol of DOM (DOM + 0. Contrary to the results obtained during our earlier expeditions the DOC and oxygen concentrations were slightly lower and higher respectively in 2008.8: C Oxygen = 0. The first derivative of equation (3) describes the DOM decomposition rate (DOMdecomposition) which was converted into oxygen consumption (Coxygen) by multiplying it by 0. Results and Discussion 3.Rixen et al.27) * exp(-0.. 3).8 indicates that only 0. This might have been caused by the high .8 * ∂ DOC (t ) ∂t (4) The factor 0. Dissolved oxygen and its response to eutrophication.. 8 A DOC decomposition experiment carried out during our previous study suggested that 27% of the DOC in the Siak was fast degradable with a half-life of 43 hrs (Rixen.8 O2 -> CO2) because oxygen organically bound within the dissolved organic matter could also be used to oxidize DOC in addition to oxygen dissolved in the river water (Rixen. The decay constant (λ) was -0. The exponential decreasing DOC concentration observed in the decomposition experiment was described as follows: DOC(t) = (DOC t0 * 0.016 * t) + (DOC t0 * 0. 2008 and references therein). et al.

the correlation between the DOC and the oxygen concentrations indicated as during our earlier expeditions that increasing DOC inputs lowered oxygen concentrations by increasing the oxygen consumption (see Eq. 3. the oxygen minimum between 7 pm and 9 pm was not linked to an increase in the DOC concentration indicating that processes other than the decomposition of riverine DOM influenced the oxygen concentration at Pelita Pantai (Fig. 4). suggesting that varying DOM inputs could have been responsible for the changing oxygen concentrations along the river. However.2 Oxygen and DOC at Pelita Pantai The DOC and oxygen concentrations measured during the 36-hour experiment at Pelita Pantai correspond with the general trend observed along the Siak river (Fig. 5b). The maximum as well as the minimum in the morning were associated with decreasing and increasing DOC concentrations.Rixen et al. 3. 3. respectively. impeded our 36-hour time-series experiment and diluted the river and the ground water feeding into the Siak. 9 precipitation rates which. oxygen levels also showed significant variations at Pelita Pantai (Fig. 3). Although the amplitude of diurnal variation was much lower than the spatial variations along the river. Since DOC concentration increased along the river despite its decomposition. as indicated above. it is assumed that DOC inputs exceed the DOC decomposition along the river.3 Waste water discharges . However. 5a). Dissolved oxygen and its response to eutrophication. At around 7 am and between 7 pm and 9 pm oxygen concentrations dropped by ~20 µmol l-1 and revealed their maximum at around 10 am.

During previous expeditions in March (rainy season) and September 2004 (end of the dry season) NH4 concentrations were determined in some of the urban waste water channels draining the city of Pekanbaru (Tab. 10 In the early morning and in the evening the oxygen minima were associated with enhanced NH4 concentrations which at ~ 8 µmol l-1 were more than twice as high as during the remaining day and night (Fig. The product of the resulting waste water discharge and the mean NH4 concentration in the waste water channels suggests a NH4 waste water discharge of 84. Assuming a NH4 concentration of ~ 4 µmol l-1 as measured at around noon at Pelita Pantai (Fig. 3).. This NH4 discharge hardly exceeds the waster water NH4 discharge which accordingly could double NH4 concentrations in the Siak (Tab. this water discharge would suggest a NH4 discharge of 92. 2003). 3). At the end of the dry season the impact of the waste water discharges on the NH4 concentrations in the Siak could have been even much higher because of the lower water . et al. Due to this common daily routine the times at around 7 am and 7 pm are generally peak times of water demand and waster water discharge in residential areas (Walski et al.3 m3 s-1 (=23 x 109 l day-1.255 mol day-1 during the rainy season in March 2004. (Baum. 5a).. 2).621 mol day-1.Rixen et al. During these two phases of the day we noticed an enhanced activity of people who prepared themselves for the day and for the night after returning home from work.777 inhabitants with the water consumption per capita of 95 l capita-1 day-1 as provided by local authorities for 2006 (Tab. 5a). 2007). In order to estimate the anthropogenic NH4 inputs into the Siak the waste water discharge was calculated by multiplying the population of 671. At this time the water discharge of the Siak measured at the tidal unaffected catchment area upstream Pekanbaru was 268. Dissolved oxygen and its response to eutrophication.

2004) could of course also affect the NH4 concentrations as such industries are located a few km upstream Pelita Pantai at the Siak. However. in general. et al. In order to estimate what processes are relevant for explaining the oxygen concentrations we used the box diffusion model described earlier (Eq. This implies that anthropogenic DOM inputs could. 3) but surprisingly the early evening NH4-peak was not associated with a pronounced increase in the DOC concentrations.. Dissolved oxygen and its response to eutrophication. et al.Rixen et al. suggested that the photosynthesis of organic matter was low and the associated oxygen production an insignificant oxygen source in the Siak (Baum. in principle. 11 discharge in the Siak (Tab. Industrial effluents from the rubber and paper processing industries which are often enriched in NH4 (Agamuthu. Korhonen et al. . 1-3). oxygen inputs caused by oxygen fluxes across the airwater interface and.. 2). 3). 2002. Rixen. contribute to the increasing DOC concentrations as observed along the river (Fig. A repeat of the same exercise with the DOC instead of the NH4 concentrations suggested that anthropogenic DOM inputs could have increased the DOC concentrations from 750 to 786 ( ~ 5%) and from 750 to > 5000 (660%) µmol l-1 during the rainy and the dry season respectively (Tab. The low light penetration depth of 20 – 40 cm and the relatively low chlorophyll concentrations. also by the photosynthesis of organic matter. Choorit et al. the timing strongly indicates that the early morning and evening NH4-peaks are mainly caused by urban waste water discharges. 1999.. 3.. 2007.4 Oxygen Dynamics Oxygen concentrations are the product of oxygen consumption mainly caused by the remineralization of organic matter.

a temperature of 29. This view was so far supported by the observed diurnal variation of the oxygen concentrations at Pelita Pantai which did not reveal the typical increasing oxygen concentrations during the day due to photosynthesis and the subsequent respiration of organic matter decreasing the oxygen concentrations during the night (e. oxygen concentration were calculated by using a mean water-depth of 9 m. 1927). in order to estimate the time it takes to re-establish a steady state after enhanced DOM inputs increased the DOC concentration. However. In order to estimate how long it takes to establish a steady state at which oxygen inputs across the air-water interface equal oxygen consumption caused by the DOM decomposition. Since it was assumed that DOM inputs compensate DOM decomposition as discussed above. 6b). DOC(t0) was first increased and later decreased during the model run. We therefore assumed that there was no photosynthetic production of oxygen at Pelita Pantai. five model experiments will be discussed that were carried out in order to check what assumptions are required to reproduce the measured diurnal oxygen variations at Pelita Pantai The mean water-depth and temperature determined were considered in all the experiments and the DOC concentrations measured were used as DOC(t0) to calculate the oxygen consumption (Tab.g. The mean DOC concentration of 858 µmol l-1 was considered as DOC(t0) and used to calculate the oxygen consumption (Fig. Tab. In the following.. 1). Dissolved oxygen and its response to eutrophication. 1). . DOC(t0) and the resulting oxygen consumption remained constant during the model run. 6.. 12 2008). Butcher et al. The results showed that it takes up to ~ 5 days (~ 144 hrs) until a steady states is established after changing the DOM input and the resulting oxygen consumption (Fig.7 °C and ‘k’ value of 25 cm hr-1.Rixen et al.

By selecting a ‘k’ value of 26 cm hr-1 the oxygen concentrations calculated varied within the range of those measured and partly even revealed the same trend (Fig. In a second step we used the same parameters but considered an equilibration time of only 6 hours which almost equals a tidal ebb or flood phase respectively. 5). Dissolved oxygen and its response to eutrophication. Raymond and Cole. modeled and measured oxygen concentrations deviated from one other. the modeled oxygen concentrations corresponded fairly well with the measured concentrations apart from the morning peak at 10 am and the oxygen minimum during the early evening (Fig. especially during the day when the modeled oxygen concentrations were much higher than the measured ones. In general it is assumed that increasing current velocity and a decreasing water depth enhance the oxygen inputs across the air-water interface. At Pelita Pantai counteracting effects caused by tides which de-/increased water-depth and in-/decreased current speeds could have been responsible for the absence of a strong tidal signal in the oxygen concentration (Fig. the difference between the model and the measured oxygen concentrations could be further reduced if one assumed that the impact of current speeds on ‘k’ exceeded . However. Nevertheless. By additionally reducing the ‘k’ value by 2 cm hr-1.Rixen et al. 7a). It therefore appears that in addition to DOM decomposition and residence time. 7b). Current velocity and water-depth are other factors which could strongly influence oxygen concentrations in the water through their influence on the bottom friction and subsequently on the piston velocity ‘k’ (see Eq. other processes controlled the oxygen concentrations at Pelita Pantai. 2. 2001). 13 The aim of the first experiment was to observe whether the oxygen concentrations at Pelita Pantai represent a steady state concentration.

Assuming that the resulting net primary production rate of organic matter reached values of up to 0. Tab. On the other hand the net photosynthetic production might be of greater ecological relevance because it occurred at the right time by reducing the impact of waste water DOM on the oxygen concentrations in the morning (Fig. 1). when darkness prevents primary production. 8.8 µmol C l-1 hr-1. The best agreement between calculated and measured oxygen concentrations were obtained by increasing the DOC concentrations in the early morning and in the evening by up to 30% (Fig. Furthermore. of course. one could explain the elevated oxygen concentration at around 10 am and the respiration of photosynthesized organic matter (= primary production rates < 0) could help to explain the decreasing oxygen concentrations after 10 am.Rixen et al. 1) because of additional inputs of fast degradable DOM. 14 those of the water-depth and enhanced current speeds increased the ‘k’ values (Fig. Tab. However. waste water discharges were also believed to accelerate the photosynthesis of organic matter in the morning by enhancing the inorganic nutrient concentrations in the Siak. we believe that a proportion of the photosynthesized organic matter was respired during the early evening and thereby contributed to the oxygen drop at this time which. 1) and therefore has only a minor effect on the oxygen consumption during the day and night. oxygen minima prevailed for a couple of hours and the oxygen drop of ~20 . 5). The assumed primary production rates are 3 to >10 times lower than the oxygen consumption rates caused by the DOM decomposition (Tab.7c. Accordingly the oxygen minimum disappeared shortly after sunrise in the morning. could also be explained by considering an enhanced input of waste water DOM. Dissolved oxygen and its response to eutrophication. In the evening.

15 µmol l-1 was very pronounced on November 1st and 2nd (Fig. a decrease in the oxygen concentrations caused by waste water discharge was not of significant relevance at Pekanbaru during the observation period. This could change.. Associated enhanced nutrient concentrations appear to weaken the impact of the DOM on the oxygen levels in the morning by favoring the photosynthetic production of oxygen. Contrary to other systems. et al. 2008) that the decomposition of riverine DOM is the main process that controls the oxygen concentrations along the river as well as its diurnal variation at Pekanbaru. e. Conclusion Our results indicate in line with our previous study (Rixen. As the oxygen concentration remained above all the thresholds discussed above. Nevertheless. 5)..Rixen et al.. the Siak seem to owe the positive effect of anthropogenic inorganic nutrient input to its well mixed water body and the low light levels which reduce the photosynthesis and the oxidation of organic matter at night. During the period of investigation waste water discharge seem to reduce the oxygen concentration by approximately 20 µmol l-1 because of additional DOM inputs during the peak-time of family life in the early morning and evening. during the dry season when anthropogenic DOM discharges could have a much more significant effect on the DOC concentrations in the river . these results call for sophisticated waste water management strategies especially in the fast developing regions downstream Pekanbaru where DOM concentrations were much higher and oxygen concentrations rarely exceeded 20 µmol l-1. 3a). especially downstream Pekanbaru where oxygen concentrations rarely exceeded 20 µmol l-1 (Fig. .g. 4. Dissolved oxygen and its response to eutrophication.

957. E. C. Indonesia. Dissolved inorganic nutrients in the peat-draining river Siak.B. Baum. Pentelow.. especially the rector of the University of Riau for his support. We are also grateful to P. References Agamuthu. FB 2 Biology/Chemistry. J. Rixen. Samiaji. Indonesia. Jose. 16 Acknowledgments We would like to thank all scientists and students from the University of Pekanbaru who helped us during our field work. Coastal and Shelf Science 73. 1999.. Liebezeit. Relevance of peat draining rivers in central Sumatra for the riverine input of dissolved organic carbon into the ocean.. 1984. Wöstmann. Smith for providing the generic mapping tools (GMT). 1927. Schönig. Butcher.. Baum. p. B.. 2009. Baum. Woodley.F.. T. P. Estuarine. University of Bremen. R.. F. 620-632. Central Sumatra. R. Bremen... A. J. 98..H.T. G. A... Benson. Krause Jr. Biochemistry Journal 21. Samiaji. Specific biogas production and role of packing medium in the treatment of rubber thread manufacturing industry wastewater.W. Diurnal variations of the gaseous contents of river waters.ZMT). Wessels and W. Dissolved oxygen and its response to eutrophication. For helpful discussion we would like to thank the anonymous reviewer. Rixen. 2007. Tropical blackwater biogeochemistry: The Siak river in central Sumatra... 563-570.K. A. 945 . The concentration and isotopic fractionation of oxygen dissolved in freshwater and seawater in equilibrium with the atmosphere. . 151-155..Rixen et al. submitted to Aquatic Sciences. 2008. Limnology and Oceanography 29.W. This study was carried out within the frame of the bilateral Indonesian-German research program SPICE II (Science for the Protection of Indonesian Coastal Marine Ecosystems) sponsored by the German Federal Ministry of Education and Research (Grant No.A... T.. Bioprocess and Biosystems Engineering 21. 03F0392C . our colleagues from the ZMT and Rena Kerr. J. the Indonesian Ministry of Marine Affairs and Fisheries (DKP) and the Ministry for Research and Technology (RISTEK). D.

0° gridcells). S. Thongpradistha. Italy. The Vegetation and Physiography of Sumatra. 889-892. Williams.J.P. K.L. Questions. 2008. M. Galloway. 2002. A.J. Assessment of CO2 emissions from drained peat lands in SE Asia.A. Dissolved oxygen and its response to eutrophication. Spreading Dead Zones and Consequences for Marine Ecosystems.. A. Pedersen. Sorensen. F. A...Rixen et al. FAO.. Baldwin... Seitzinger. Estuaries and Coasts 24. Gas exchange in rivers and estuaries: Choosing a gas transfer velocity... 1997. J. 2008. Voss... 2006. H.S. monthly precipitation (monitoring) product (1. Seitzinger.R. J.. Martinelli... Townsend. Freney.. Kauppinen.. J. An anoxic event and other biogeochemical effects of the Pantanal wetland on the Paraguay River.P.. H. Schlitzer. Jickells. 926-929... Ehrhardt. Offenbach am Main. Ganeshram.. T.A. Melack. R.. Delft...R.A.. Widerkiewicz. WL Delft Hydraulics.. N. L. M.. Grasshoff..H. J. Kremling... 2004. The globalization of N deposition: ecosystem consequences in tropical environments. Modeling black water: Predicting water quality during flooding of lowland river forests.. Methods of seawater analysis. B.. J. Weinheim. Moore.. Page. L. Yoshinari. Kogevinas. S.K. Kluwer Academic Publishers.CO2... I.. Sippel.N. Heederik... M. Matson. Baker.. J. 36. P. Nature 302. Vitousek. Thanakoset. p. R... M. P. K. O.. and Potential Solutions. Sutton. Howitt... K. Liukkonen. 1:5000000. Rees. J. 2001.. W.. Cole. R. Bekunda. Transformation of the Nitrogen Cycle: Recent Trends. D. Christensen. 1997. Capone. Diaz..S... The oceanic fixed nitrogen and nitrous oxide budgets: Moving targets as we enter the anthropocene? Scientia Marina 65..E. T.. DWD. Geider. LaRoche... Sunyer. W. Oschlies. D. Duce. p.H. Townsend. Burdorf. 3rd ed. International Archives of Occupational and Environmental Health 77. Marine biological controls on atmospheric CO2 and climate.. J. J.J. Westberg.. Identification and cultivation of photosynthetic bacteria in wastewater from a concentrated latex processing factory. A. S... J.R.M. Altieri.. J.. K. Noparatnaraporn.. Deutscher Wetterdienst. S. M. K.... 229-242. 451-460.. Uematsu.A.N. P. Brandes. Middelburg.F.W. T.G. Rosenberg. A. Astrakianakis. P. Galloway. Codispoti. A.. Prospero.M. Y. H. 312-317. Nickovic.1058. Saalo. Arrigo... P. Soil map of the world. . Raymond. Liu. L.L. K.. 2001. K. S. Teschke.. P. R.. 1999.M. Ecological Modeling 203. Ward. Global Precipitation Climatology Centre. Impacts of Atmospheric Anthropogenic Nitrogen on the Open Ocean... R. 85-105.. Liss. Occupational exposure to chemical agents in the paper industry.. S.M. Ahrens. 2008. Korhonen. R. S. C. 893-897. 328-329. 1999.. K..A... 1055 . 17 Choorit. Osvoll. Szadkowska-Stanczyk. A. M. Devol. M.A. 2006. Silvius. Rix. W. J. Z. J.R. G.S. J.B. J. Sasaki..R. McElroy..... Biotechnology Letters 24. D. Langlois. Ulloa. M. 257-272. Paerl.A. Boffetta. D. M..J. B. McDowell. Kuypers. Biogeochemistry 46. Science 320. 2003. 1983. Cornell.. Laumonier. Cai. Dentener. Erisman.. Science 320. Rome. FAO / UNESCO.. R..A. Calheiros. S.. S. A. T.M. Wösten. Wiley-VCH.. Zamora. 67-83. L.A.W. G. Hamilton.W... Delft Hydraulic report Q3943. Germany. 2007.. Science 321. Naqvi. Limnology and Oceanography 42. T... P. Peat . Hooijer.

Koelle. T. V. Samiaji.. 2008. 2003. 2005.1418. doi:10. J.W. A. Beusen. 26-77. 18 Redfield. CT (USA). Seitzinger. F.M. Dumont. Halpert. B. Ketchum.Rixen et al. Dissolved oxygen and its response to eutrophication. Rixen. Chase. Sources and dilivery of carbon. Grayman. M.A. 1987. Nitrogen transport by rivers of south Asia. M. W. D. Current Science 94. Beckwith.. Wiley & Sons. in: Hitt..A.. T..C. Waterbury. W.F.. Global and regional scale precipitation patterns associated with El Nino/Southern Oscillation. pp... 1413 . Advanced Water Distribution Modeling and Management.. Biogeochemistry 90.V... The Influence of organisms on the composition of sea-water.. A.. a tropical black water river in central Sumatra on the verge of anoxia. C.F. Bouwman..H... S.1029/2005GB002606.. nitrogen. Walski. Monthly Weather Review 115..). Baum. New York. 16061626.S.. and phosphorus to the coastal zone: An overview of global nutrient export from watersheds (NEWS) models and their implication. . N (Ed. Harrison. The Siak. D.... E. A. E. Haestad Press. Global Biogeochemical Cycles 19. 1963. Ropelewski. A. Jose. 129-140. Balzer. Subramanian. S. C. Pohlmann.. Savic. Haestad.. The sea. J. 2008.A. T. Richards.H.

Tapung Kiri in the highlands. The locations of the main cities (Pekanbaru. Samples collected during our former expeditions are indicated by open circles and sampling sites during the expedition in November 2008 are represented by black circles. The Kanan/Kiri and the Mandau junctions are at river-km 155 and 245. 3: DOC (a) and oxygen concentrations (b) measured at a water-depth of 1 m versus river-km. Fig. Dissolved oxygen and its response to eutrophication.Rixen et al. Fig. Perawang. The river-km zero represents the origin of the S. The Southern Oscillation Index (SOI) was obtained from http://www. 1: Study area: The Siak with its headstreams S. Tapung Kanan are shown by the large black . Tapung Kanan.ncep. 19 Figure Captions Fig. Tapung Kiri and S. At river-km 320 increasing salinity indicates the beginning of the estuary (salinity data are not shown) and at river-km 370 the Siak discharges into the Malacca Strait. The mean oxygen and DOC concentrations in the Mandau and S. and Siaksriindrapura) are indicated by triangles. Peat soil distribution (marked in grey) is obtained from the FAO (2003). and its tributary Mandau.cpc. 2: Precipitation rates obtained from DWD (2006) and averaged for the area 1°S – 1°N and 100-102°E are indicated by the grey bars. The black bold line shows the precipitation rates smoothed with a three-month moving and also smoothed with a three-month moving average.noaa. The dark grey bars show the months during which the expeditions were carried out.

20 squares. black circle) as well as water temperature (b). 4: River profile obtained at Pelita Pantai in Pekanbaru after the 36-hour experiment at November 8th. (a. 3. 5: Oxygen (grey line). Dissolved oxygen and its response to eutrophication.c) are the results derived from the experiments 2. Data measured during former expedition are indicated by open circles and the 2008 data by black circles. 4. Modeled oxygen concentrations versus water-depth (b). The black squares indicate the position of our equipment at the beginning of the experiments (further mid stream) and after resetting the experiments on November 2nd (closer to pier). The light gray bar indicates the high water periods and the dark grey bars the night times. Fig.Rixen et al. Fig. NH4 (a) and DOC concentrations (b. Fig. The used piston velocities are given as numbers and shown in c by black circles. water level and current speed (c) versus time. The bold black bar indicate the time of the thunder storm. 7: Comparison of modeled oxygen concentrations (bold line) and measured oxygen concentrations (grey line) as shown also in figure 5 a versus time. . b. 18:00. 6: Modeled oxygen concentrations versus time (a) indicating the time it takes until a steady state is established after a change in the DOC concentrations and therewith oxygen consumption rate. Fig.

21 Fig. . In order to calculate the oxygen concentrations shown in figure 8a the DOC concentrations marked by the grey circle were used instead of the measured ones. The DOC concentrations indicated by the black circle were measured and also shown in figure 5 c. Dissolved oxygen and its response to eutrophication. 8: Modeled (black line) and measured oxygen concentrations (grey line) (a) as well as primary production rates used in the model (b) and DOC concentrations used to calculate the oxygen consumption rates (c) versus time.Rixen et al.

0 24.6 29.0 0.0 9.2 -0.6 2. 1: Hourly averaged water-depth (WD).3 4.1 7.6 746.9 30.0 8.0 0.0 30.7 4.0 0.4 29.0 0.6 29.6 29.0 0.3 21.0 773.8 2.0 6.0 9.1 4.5 2.0 3.9 6.0 9.8 29.0 0.7 29.4 -0.5 775.8 2.0 0.3 27.9 1027.0 0.0 0. Time 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 7:00 8:00 9:00 10:00 11:00 12:00 13:00 14:00 15:00 16:00 17:00 18:00 19:00 20:00 21:00 22:00 23:00 0:00 1:00 2:00 3:00 4:00 5:00 6:00 7:00 Mean WD [m] CS Temp.0 827.0 2.5 888.8 6.2 -0.1 24.0 0.9 25.8 0.0 0.0 0.6 -0.5 9.2 0.0 0.6 2. Due to missing samples the DOC and NH4 concentrations indicated by the italic number were interpolated.0 10.4 -0.5 4. current speed (CS.0 2.4 4.2 -0.8 25. 5a).0 0. 8 a.Rixen et al.2 29.0 0.9 3.3 2.5 28.1 10.4 896.2 3.2 4.2 0.9 25..4 29.2 745.0 0.7 28.6 3.2 4.7 3.5 19.0 3.0 9.3 24.1 6.6 7.5 9.5 9.8 2. 8c).0 0.0 K [cm hr-1] OC Pp [µmol l-1 hr-1] 26.5 18.9 28.8 25.4 3. 5c).0 4.2 18.6 2.5 9. watertemperature (Temp.0 0.7 29.6 27.0 3.7 29.0 3.0 783.5 8.0 0.0 4.0 9.0 0.1 15.4 -0.6 29.3 870.2 -0.6 944. 7c).2 865.7 875.0 857.0 9.0 24.8 9.5 24.1 6.0 9.0 9.2 912.8 4.6 802.0 3.8 29.6 29.7 745.0 0.6 2.4 754.7 5.0 4.2 3.3 25.0 9.6 2.3 3.2 9.2 -0.3 28.8 0.5 8.7 5.8 29.0 3.0 27.2 -0.7 6.1 24.6 29.7 29.1 3.4 29.8 2.0 0.2 6.6 5. 5b) as well as the determined DOC (Fig.8 4.0 0.0 0.8 0.3 27.5 8.0 24.8 8.5 26.9 1017. 5b) and NH4 concentrations (Fig.4 9.0 0.0 0.7 24.4 821.6 DOC [%] 30.5 9.3 818.9 20.0 0.8 -0.5 9.0 6.0 0.7 18. 22 Tables Tab. Measured data Date Nov.9 2.8 0. Dissolved oxygen and its response to eutrophication.5 29.7 4.0 30.0 21.0 29. the resulting oxygen consumption rates (C oxygen) and the primary production rate (Pp) used to reproduce the measured oxygen concentrations at Pelita Pantai as shown by the bold line in Fig.2 7.9 24.0 30.2 3.9 1029.4 6.8 29.0 900.5 8. Negative Pp values indicate the respirations of photosynthesized organic matter.0 23.5 29.0 12. Fig.7 20.1 28.8 3.3 863. Given are also the deviations from the DOC concentrations in % (Fig.9 3.6 4.5 9.8 .0 0.0 3.5 10.0 9.7 3. Fig.2 21. [cm s-1] [°C] Model assumptions NH4 [µmol l-1] DOC [µmol l-1] 9.8 5.8 933. the piston velocity ‘k’ (Fig.4 3.1 821.8 9.5 29.0 0.

Juanda Sampling Il. Riau DOC NOx NH4 [µmol l-1] 16. 23 Tab.356 233.465 367 213 190 137 1.776 296.034 10.236 135 127 234 159 1. Dissolved oxygen and its response to eutrophication.005 13.563 [µmol l-1] [µmol l-1] 13. 2009).242 664 1.2: Waste water channels draining the city of Pekanbaru as well as DOC.511 856 1.. NOx (NO3 + NO2) and NH4 concentrations measured in March and September 2004 whereas the NOx and NH4 concentrations were obtained from (Baum et al. Waste water channel Il.292 263 247 1.911 286.Rixen et al.356 14.948 330. Riau 2 month March September March September March Il.320 595 . Karag March Mean September March September Il.

465 18.777 ----- ----- Capita 95 ----- ----- l capita-1 day-1 63.768 92.876 786 18.320.281.568.097 mol day-1 µmol l-1 mol day-1 mol day-1 µmol l-1 mol day-1 . Within the row entitled ‘combined’ water. Waste water Population water consumption/capita March 2004 Water discharge NH4 concentration NH4 discharge DOC concentration DOC discharge September 2004 Water discharge NH4 concentration NH4 discharge DOC concentration DOC discharge River Combined Units 671.366.400 176.621 750 17.Rixen et al. (2007) in the tidal unaffected area upstream Pekanbaru.1 as well as water discharges measured by Baum et al.815 l day-1 µmol l-1 1.254.000 3.770 20.138.676. The NH4 and DOC concentrations were obtained by dividing the NH4 and DOC discharges by the water discharge and indicate the impact of waste water discharge on the respective concentrations.155.168 63. 3: Calculation of waste water discharges into the Siak based on NH4 and DOC concentrations given in Tab.217 5.815 l day-1 595 4 14 µmol l-1 37.219.958.818.818.815 23. Dissolved oxygen and its response to eutrophication.815 3.240 52.944 286.857 13.320 4 8 84.273 750 2.000 23. DOC and NH4 discharges from the waste water and river were added.911 887. 24 Tab. The population and the information on water consumption per capita which were used to calculate the waste water discharge were obtained from local authorities for the year of 2006.