SCOPE 42 - Biogeochemistry of Major World Rivers

8.1 INTRODUCTION

Europe is the second smallest continent. It measures (without Iceland and Spitzbergen, but including other small marginal islands) 10330.5 X 10³ km²; this is equivalent to 6.94% of the total continental surface. Morphologically, Europe is highly structured and features the longest coastline of all continents relative to its area. At its southern margin this diversity arises from the ongoing collision of the African with the European plate which caused the uplift of the Alpine mountain ranges. At the northern margin, the glaciers of Quaternary ice ages cut deep scars into the sea level fringes of Fennoscandia and of the northern British Islands.

As a consequence of the geologically young morphology, Europe is occupied by a multitude of small basins drained by relatively short rivers (Figure 8.1; Table 8.1). The two largest rivers, the Wolga and the Danube (3694 and 2850 km long), do not rank among the top dozen world rivers in terms of length, basin area or discharge. Only six rivers (Wolga, Danube, Northern Dvina, Pechora, Neva and Rhine) qualify for the top 40 world rivers by discharge (e.g. Kempe 1982; Lerman 1981) which comprise some 48% of the total river discharge. Cumulatively the European rivers discharge some 2800 km³/year, i.e. 7.4% of the world total discharge of 37700 km³/year (Baumgartner and Reichel 1975), slightly more than one would expect from the relative size of Europe's area.

Europe, nevertheless, plays an important part in the global cycles of matter. Due to its temperate, humid climate and its high percentage of

Figure 8.1 Map of major European river systems

Table 8.1 European rivers ordered counterclockwise and their basic parameters, according to various sources (tributaries indented)

Table 8.2 Absolute and relative total dissolved solid (TDS). bicarbonate and dissolved inorganic carbon (DIC) transport from continents (after Kempe 1979)

limestones in surface rocks, it has the highest chemical weathering rate of all continents. Table 8.2 (Kempe 1979) shows that the average total dissolved ion (TDI) concentration as calculated from Livingstone's (1963) compilation of European rivers amounts to 182 ppm, much higher than on any other continent. Of all dissolved solids reaching the ocean 12.6% derive from Europe, i.e. double the amount one would expect from the relative area. With this portion Europe surpasses Africa and s. America in spite of their much higher water discharges. In the case of bicarbonate, the single most important ion, Europe delivers slightly less than S. America, but still more than Africa to the world ocean. The same is naturally true for the discharge of dissolved inorganic carbon (DIC) which is calculated from the HCO₃^- load (Table 8.2).

If one wants to estimate, however, the total flux of biogeochemically important compounds such as dissolved, particulate or total organic carbon (DOC, POC, TOC), dissolved nitrate (NO₃^-), nitrite (NO₂^-) and ammonia (NH₄⁺), organically bound nitrogen (N_org), dissolved ionic or total phosphate (PO₄^3-, T-PO₄) or particulate phosphate (PP), then the Livingstone data base does not suffice. Also, fluxes alone do not provide information about sources, transformations and sinks for these compounds within the river basin, nor do they allow the evaluation of time trends in concentrations which are valuable indications for changing levels of pollution in rivers and lakes. Because of the high population density in some of the European river basins (Table 8.3) and because of intensive agriculture and the highly developed industry in Europe, the input of these compounds to river systems has most likely changed dramatically since the last century and is probably still changing.

Table 8.3 Size of population in large European river basins (compiled from Helmer 1989; Pet tine et al. 1985)

 8.2 THE DATA BASE

Discharge of rivers has been monitored since the last century because the data were important for planning channels for shipping, reservoirs, power stations and irrigation schemes. Regular water quality monitoring, however, started only after the Second World War. In fact, even the longest hydrochemical records, those of the Elbe and of the Rhine, cover only some 30 years. The Elbe record of the Hamburgian Water Works covers the years 1954-81 and is still largely unpublished (Kempe 1982). Water quality monitoring of the Rhine started in 1963 (with some of the stations extending further into the past) under the auspices of the International Commission for Protection of the River Rhine against Pollution (Intern. Comm., since 1972; Intern. Comm., since 1976; Deutsch. Komm. since 1976). Results of monitoring of Spanish rivers are published by the Minister of Public Works since 1974 (Ministerio de Obras Publicas, since 1974) and French rivers are monitored for various parameters by the Minister of the Environment and the data are published since 1975 (Ministere de l'Environnement et al., since 1975). The Danube was sampled daily at Vienna between 1978 and 1981 and analysed for a large variety of constituents, also micropollutants, yielding a data set largely unpublished (Reuschel and Forster 1982).

Today, almost all European countries operate national or regional discharge and water quality monitoring networks. Publication and scientific evaluation of these data are, however, limited. European participants of the SCOPE/UNEP Project, 'Transport of Carbon and Minerals in Major World Rivers', were therefore asked to evaluate these existing records and to study specific regions or biogeochemical problems, rather than to set up new sampling programs.

Lugo (1983) and Cauwet and Martin (1982) evaluated the Spanish and French records, respectively, for total organic carbon transport. Kempe (1982) reviewed the data for the four largest French rivers, and of the Rhine, Weser and Elbe for long-term trends, long-term average transports and biogeochemical interactions. Pet tine et al.(1983,1985, 1987) gave accounts of Italian rivers and Reuschel and Forster (1982) reviewed some results derived from the Danube record. Pocklington and Pempkowiak (1983) and Pempkowiak (1985) calculated the organic carbon transport of the Vistula, and Romankevich and Artemyev (1985) did the same for the Russian rivers. Skoulikidis (1989) sampled within the SCOPE/UNEP River Project ten Greek rivers and discussed their chemistry in a doctoral thesis.

Headwater basins of various characteristics were studied in the Federal Republic of Czecho-Slovakia, Germany and Yugoslavia by Moldan (1987) (Elbe River), Hartmann (1983) (organic output of a Harz mountain bog) and Kempe and Emeis (1985) (carbonate chemistry and formation of travertine at Plitvice).

Much work has been devoted also to trace the fate of organic matter in estuaries. The Elbe, Weser and Ems Estuaries were studied by a cruise of the R/V Valdivia (Degens et al.1982). Several authors (Cadee and Laane 1983; Cadee 1987; Eisma et al.1983; Laane and Ittekkot 1983, 1985) studied the Ems Estuary and Eisma et al.(1985) compared it to the Gironde Estuary. Eisma et al.(1983) and Lindeboom and Merks (1983) described results obtained in parts of the Rhine Estuary and Cauwet and Meybeck (1987) studied the Loire and Gironde Estuaries, while Pet tine et al.(1987) investigated the situation of the Tiber Estuary. Recently, Artemyev and Romankevich (1988) studied organic carbon transport through the Northern Dvina Estuary.

Another point of gravity in the SCOPE/UNEP River Project is formed by the many studies dealing with the chemical characterization of the organic matter in rivers and estuaries. Seifert (1982, 1985) compared the composition with regard to carbohydrates in several European rivers and at various times of the year in the Elbe Estuary. Particulate carbohydrates were analysed in the Elbe Estuary by Lohse and Michaelis (1983) and Lohse (1983). Pempkowiak (1985) studied the fractionation into labile and stable organic matter in the Vistula Estuary and Mycke (1982, 1985) analysed Elbe River water samples for dissolved phenolic compounds.

Parallel to the SCOPE/UNEP Project, the UNEP Global Environmental Monitoring System (GEMS) was launched in 1977. Under the auspices of the World Health Organization (WHO) the Global Freshwater Quality Monitoring Project collects basic hydrochemical and health-related data from 43 lake, 61 groundwater and 240 river stations, 86 of which are in Europe (GEMS 1983). Meybeck (1987) described the project and gave first results. A more detailed account is given in 'Global Freshwater Quality¾A First Assessment' (GEMS 1989). Additional data compilations are also available from the OECD (1982, 1985).

8.3 TOTAL EXPORT

Rivers gain and lose water and dissolved or particulate matter along their course. They also experience substantial annual and interannual variations in their discharge, sediment load and concentration of the various chemical compounds transported. In large river systems, which derive their water from regions different in climate, high and low water stages and therefore the mobilization of material may occur in different seasons. The Rhine is such an example (Figure 8.2). The upper Rhine, which receives melt water from the Alps has the highest water discharge in June/July, while the lowland rivers in Germany have their highest runoff in February/March when the snow melts at the end of the winter. At Cologne, the hydrograph of the Rhine still shows a double peak, while the average discharge curve at Lobith (the Dutch/German border) hides the alpine signal under a smooth shoulder of decreasing discharge.

Figure 8.2 Long-term average monthly discharge of the Rhine at three stations: Basel, Cologne and Lobith (German/Dutch border) (Eisma et al. 1982)

Tributaries may play a more important role for the transport of water or a certain compound than the main stream itself. This is the case for the Danube, where the Inn contributes more water than the Danube and where more than 50% of the water and suspended matter is derived from the tributaries downstream of Budapest, i.e. the Drava, Sava and Tisza (Figure 8.3).

Total transports can therefore only be defined for a certain station. In most cases even the station closest to the mouth excludes some of the coastal tributaries and it does not give any information of how much matter really passes the estuary. In fact, certain estuaries may import more marine matter than they export terrigenous matter. Calculating river transports even for a specific station is, furthermore, not a straightforward task. Water discharge is mostly derived from daily readings of a gauge. Thus runoff is the best known mass transport in rivers. Other physical and chemical parameters are, however, mostly monitored in relatively large intervals. In the case of the Rhine stations, many important parameters are measured twelve or eight times per year only. The most simple way to obtain an estimate of the average transport (F_x) is to use the arithmetic mean (M_x) of the parameter (X) and to multiply it with the arithmetic mean of the discharge measured (M_Q) during the sampling days:

where the arithmetic mean is defined as:

(8.2)

(n = number of measurements, Xi individual measurement of parameter).

If no simultaneous discharge measurement is available, the otherwise available annual average discharge is also often used. This method of calculation is common and has been used with the GEMS data in Table 8.4 because only annual or long-term arithmetic mean concentrations have been published. If the samples are not equally spaced in time, the method becomes even more unreliable. Also, the widely spaced measurements may then miss a major flooding event, which perhaps could mobilize 50% of the total annual load. It would therefore be very important to study a daily chemical record and compare its calculated transport with values obtained from more widely spaced sampling events.

Figure 8.3 Longitudinal profiles of the Danube for discharge (right) and suspended matter transport (left) (data after  Lászlóffy 1967) 

Table 8.4. European rivers, ordered counterclockwise, and their transports of carbon and nutrients according to various sources (tributaries indented) (updated after Kempe et al. 1985)

Using arithmetic means gives an equal weight to each of the measurements. The concentration measured at low discharge has the same importance as the concentration measured at peak discharge. This introduces a serious bias in favour of low discharge concentrations in the transportation calculation. It is therefore much better to use discharge (Q) weighted means for the calculation of total transport:

In order to take into account irregular sampling intervals, the weighting should also be introduced for time ( expressed as numbers of days in a year, D_i) . An annual time-weighted concentration can be calculated from a set of measurements beginning not with the first day in the year and ending not with the last day in the year by:

	+	(8.4)

The combination of Equations (8.3) and (8.4) allows calculation of time and discharge weighted annual means and approximates the mass transport much better than transport calculated solely by the multiplication of concentration and discharge averages. Transport figures given by Kempe (1982) were derived in this manner (Table 8.4).

 Another method to calculate total mass transport can be used if the parameter in question shows a significant correlation with discharge. In such a case, the total discharge curve can be used to calculate fluxes simply by converting the individual gauge readings with the regression equation to concentrations. In a study of the Severn Estuary, Mantoura and Woodward (1983) describe such a case where the DOC concentration varies positively and highly significantly with discharge:

Since higher discharges occur in winter, they find that more than 70% of DOC transport occurs in winter. In summer (June, July) only two outliers were observed at low discharge so that they do not introduce significant errors in the mass transport estimate. High chlorophyll-a concentrations measured simultaneously indicated that these outliers mark plankton blooms in the river or in upstream lakes. The positive correlation with discharge suggests that most of the DOC in the Severn is leached from soil and is composed of rather stable humic matter. This conclusion is corroborated by the fact that the DOC passes conservatively through the Severn Estuary (Mantoura and Woodward 1983).

Mantoura and Woodward further suggest that because of the positive correlation of DOC with discharge, possibly the global total could be higher than published, e.g. by Meybeck (1982), because high water stages were not generally sampled in previous sampling programs. This conclusion is, however, more than premature because many rivers do not show any correlation with discharge and others show even strong negative correlations. Table 8.5 gives some of the relations found by Kempe (1982) for the large industrial rivers. Even the data of Hartmann (1983), who measured the DOC export from a mountain bog, does not yield a positive significant correlation between DOC and discharge (r = -0.44, n = 11). Bogs are probably a major source for upstream waters in Central and Northern Europe. The failure to find a significant correlation between their discharge and DOC output indicates at least that one has to be careful in generalizing findings from one drainage basin to another.

Table 8.4 lists the results of mass transport calculations for major European rivers. The values are very inhomogeneous, i.e. even for the same river they

Table 8.5 Correlation of organic carbon related parameters with discharge for large industrial rivers

apply often to different stations and to different years or sets of years. Also, data are mostly several years old before they are published. In consequence, no values for the most recent years can be given.

Even with these data at hand, it is not straightforward to calculate an estimate of the total continental discharge of organic carbon and nutrients. In Table 8.6 the cumulated loads plus the discharges which they represent are listed. If one assumes that the areas of Europe not represented have similar discharges, then one can calculate continental total discharges. This assumption, however, results in a total continental discharge of 2344 km³/year only, compared to the estimate of 2800 km³/year given by Baumgartner and Reichel (1975). Thus, two estimates for total loads are given in Table 8.6, corresponding to these two total discharge estimates. The load figures derived this way are only rough estimates at best. In the case of TOC the estimate is based on a sample representing 65.4% of total continental runoff, but for nutrients the data base for the estimate is much smaller. The totals calculated are (in 10⁶ t/year): 18.9 (22.5); 2.9 (3.4), 1.0 (1.2),0.31 (0.37) and 0.38 (0.46) for TOC, NO₃-N, NH₄-N, PO₄-P and P_tot (numbers in brackets refer to the Baumgartner and Reichel 1975 discharge).

A different approach to the problem is to average all the concentration values of all stations and multiply the average by the total continental runoff. This can, for example, be made for the published average of NO₃-N of all European river stations (Chapman 1989). This average amounts to 4.5 mg N/l resulting in an estimate of 13 X 10⁶ t N/year. A different value is obtained if one assumes that the distribution of concentrations in the GEMS data set is representative for Europe. If one integrates the area under the distribution function (tabulated in Chapman 1989, Table 8.6) then the average European concentration becomes 6.0 mg NO₃-N/l and 0.85 mg NH₄-N/l. The total fluxes amount to 16.8 and 2.4 X 10⁶ t N/year, respectively. These figures are much higher than the results in Table 8.6. This may in part be due to an insecurity in the GEMS data base because at some stations it is not clear if NO₃ or if (as was demanded) NO₃-N was reported.

Table 8.6 Estimates of total European river transports 

8.4 THE BIOGEOCHEMISTRY OF AN 'INDUSTRIAL RIVER': THE RHINE

Western, Central and Eastern Europe belong to countries with intensive agricultural and industrial activity. Rivers carry, intentionally or unintentionally, the wastes of these activities. The Rhine, the largest river crossing Central Europe, may therefore serve as an example of how civilization has altered the biogeochemistry of rivers. As the Rhine drainage system contains also large lakes, one can study simultaneously the deterioration of lacustrine and fluvial systems.

8.4.1 THE pCO₂ IN THE PLUVIAL AND LACUSTRINE REGIME

 The CO₂^-pressure (PCO₂) of a water is a suitable indicator of the dominance of either respiration or photosynthesis in an aquatic ecosystem. It can be calculated from standard hydrochemical parameters, i.e. pH, temperature, alkalinity (HCO₃^- + 2CO₃^2-) and the concentrations of the other main ions (Ca²⁺, Mg²⁺, Na⁺, K⁺, SO₄^2-, Cl^-) (for example; Kempe 1982). If the values calculated surpass 340 ppmv (ppmv = parts per million partial pressure), i.e. the atmospheric partial pressure, then the water suffers from a surplus of CO₂ and it becomes a source of CO₂ to the atmosphere. If the values fall below 340 ppmv, then the water is a sink for atmospheric CO₂. In the first case internal sources generate CO₂ by respiration of organic matter and in the second case CO₂ is consumed by ongoing photosynthesis in the water body.

In Figure 8.4 the downstream development of the pCO₂ in the Rhine River is summarized as apparent from the long-term time and discharge weighted averages for 15 monitoring stations (Kempe 1982). The lowest pCO₂ is found at station Stein (642 ppmv, average 1974-77) where the river leaves Lake Constance. Downstream the pCO₂ increases steadily until it reaches a long-term average of 6300 ppmv (1963-78), a value about 20 times the pressure of the atmospheric pCO₂. Below Braubach, the pCO₂ decreases in the long- term average curve (solid line). If one inspects the last year of the evaluated record, i.e. 1979 (dashed line in Figure 8.4), then the pCO₂ is found to monitoring stations. Times for which means apply are indicated. River flow is from right (outflow of Lake Constance) to left (Rhine Delta in the Netherlands) (Kempe 1982) increase all the way from Lake Constance to the Dutch-German border at station Bimmen/Lobith.

Figure 8.4 Longitudinal CO₂ pressure profile of the Rhine River as evident from long-term pCO₂ averages of 13 hydrochemical

Figure 8.5 Scheme of pCO₂ reaction to pollution with organics and nutrients in the lacustrine and fluvial environment as deduced from Rhine data (Kempe 1982)

Figure 8.5 summarizes the reaction of the biogeochemical system of the Rhine as apparent from the longitudinal profile to the input of wastes, i.e. organic matter and nutrients. In Lake Constance, dissolved nutrients stimulate photosynthesis in summer, depressing the pCO₂ of the lake surface water. During the plankton bloom, these values can be lower than 340 ppmv (Kempe 1982). In winter, pCO₂ values are significantly higher than in summer, indicating the predominance of respiration. This is characteristic for the hypolimnic water layer which is mixed with the epilimnic water at the end of October. The Rhine leaves Lake Constance with this lacustrine fingerprint, relatively overall low pCO₂-values and with characteristically lower summer than winter values. In the fluvial regime of the Rhine, this fingerprint is quickly lost. The overall pCO₂ rises rapidly and the mode of seasonal variation changes. Already at station Kembs one finds higher pCO₂ values in  summer than in winter. At Braubach., respiration dominates throughout the year. Lowest values are found in April and highest in October (Figure 8.6). This seasonal variation is possibly governed by the chemical composition of organics available for bacterial respiration. In April, the less polluted alpine melt waters start to increase in discharge, while in October low waters possibly lead to a concentration of the constant input from anthropogenic sources (sewage, industrial wastes) and hence to very high rates of respiration per volume of river water.

Figure 8.6 Comparison of pCO₂ seasonal variation for two Rhine stations. Station Stein monitors the outflow of Lake Constance and represents the lacustrine environment, while station Braubach represents the fluvial environment (Kempe 1988)

8.4.2 INTERDEPENDENCIES AMONG THE BIOGEOCHEMICAL PARAMETERS

The Weser, Rhine, Mississippi and Elbe have the highest pCO₂ annual averages of 22 large rivers investigated (Kempe 1982, 1984, 1988). This suggests that riverine high CO₂-values may be indicative of 'pollution'. The other possible source of a high pCO₂ could be tributary groundwater. Groundwaters also carry very high CO₂-pressures because they equilibrate with the high pCO₂ in the soil zone during infiltration. However, measurements in karst areas (Kempe 1982; Kempe and Emeis 1985) have shown that within a very short distance below springs the high pCO₂ is lost. Furthermore, the example of the Rhine shows that the pCO₂ increases downstream, which could not come about by recharges with small amounts of groundwater mixed into a constantly degassing running stream. Therefore internal respiration must be the reason for the high pCO₂ in rivers. Respiration is accompanied by an O₂ consumption leading to an at least equivalent deficit in dissolved oxygen. In Figure 8.7 the temperature dependence is plotted for the oxygen concentration at saturation and for the theoretical pCO₂, calculated under the assumption that all the oxygen is consumed by respiration to produce CO₂ and no exchange with air occurs. Even though less oxygen is available at higher temperatures and therefore less CO₂ can be produced, the pCO₂ rises with temperature due to the strong positive temperature dependence of the Henry Law constant. In the case of station Braubach, which has a 16-year average temperature of 12.8 °C and an average pCO₂ of 6300 ppmv, hardly any oxygen should be left in the river. However, the long-term mean amounts to 5.6 mg O₂/1. Similarly at station Lobith, which has an average temperature of 13.2 °C, an average pCO₂ of 5000 ppmv was found compared to an average O₂ concentration of 5.7 mg/l. Thus, more CO₂ is found in the river than explained by the apparent O₂ deficit. If one compares the actual pCO₂ (calculated from pH, etc.) with the pCO₂ which can be explained by the deficit of oxygen (Figure 8.8) for all individual measurements, one can see that another oxygen source must be available to fuel respiration in the river. Theoretically, both the actual and the oxygen deficit associated pCO₂ should have a regression with the slope of one. The observed slope is, however, much smaller and it is strongly determined by the high actual pCO₂-values.

Figure 8.7 The temperature dependence of (falling from left to right) the oxygen concentration at saturation compared to the temperature dependence of the pCO₂ (falling from right to left) arising from the total reduction of the oxygen (Kempe 1982)

This second oxygen source for respiration is most probably dissolved nitrate. Correlation analysis (Figure 8.9) shows that the nitrate concentration is highly positively correlated with the oxygen concentration and negatively correlated with pCO₂. If nitrate would act as a nutrient, i.e. fueling photosynthesis, the nitrate concentration should be positively correlated with the pCO₂ but negatively with the O₂-concentration. Thus, the riverine fingerprint of an industrial river includes also the inversion of the interdependence of pCO₂, pCO₂ and NO₃-concentration.

Figure 8.8 Relation between the actual pCO₂ calculated from monitoring data and the pCO₂ which is equivalent to the observed oxygen deficit for all measurements 1963- 78 at stations Braubach and Lobith (Kempe 1982). Regressions: Lobith: y = 0.194 * X + 2230, r = 0.351, n = 128; Braubach: y = 0.181 * X + 2097; r = 0.504, n = 128

The correlation matrices reveal some other interesting relations among biogeochemical parameters of an industrial river. TOC, BOD (biological oxygen demand) and COD (chemical oxygen demand) correlate highly significantly and positively with each other and with pCO₂ but negatively with O₂. This suggests that the bulk of the organic carbon in the river is rather labile and much of it can be re mineralized bacterially within the short residence time in the stream. PO₄^3- and NO₃^- correlate highly significantly and positively with each other, but only PO₄^3- correlates highly significantly and positively with discharge. Phosphate is therefore diluted with higher discharge but not nitrate. This may be related to the higher soil retention for phosphate than for nitrate, the latter being more easily leached by drainage water. Nitrate is negatively and highly significantly correlated with temperature. This correlation is also coherent with the hypothesis according to which nitrate may serve as an oxygen donor for microbial remineralization which would be most active in summer at elevated temperatures. The pH in the river is dependent on the pCO₂ of the water and is used to calculate pCO₂. It is therefore statistically not an independent variable. Consequently, and because pH drops under rising pCO₂, pH appears to be highly significantly correlated with most parameters which correlate with pCO₂ but with reversed sign.

From these correlations we can draw a picture of the biogeochemistry of an industrial river. With the increase of labile organics (measured, for example, as TOC) and nutrients, the BOD, COD and pCO₂ in the river rises because of the increased rate of respiration. Respiration consumes at first most of the oxygen and then part of the nitrate if temperatures are high enough for bacterial denitrification to occur. The phosphate concentrations are not affected, they just reflect the amount of input in relation to the background discharge.

Because of the evolution of free CO₂ in the river, the pH drops and the river becomes more acidic. This is in contrast to the lacustrine realm, where enhanced photosynthesis leads to a depletion of CO₂ and hence to an increase in pH. These shifts have a profound impact on the carbonate mineral equilibria. In general the waters of the Rhine have a hard water composition, i.e. high concentrations of Ca²⁺, Mg²⁺ and HCO₃^- because of the widespread occurrence of limestones in the tributary area. Therefore the waters in the tributaries are mostly saturated with respect to calcite and dolomite. In Lake Constance the increase in pH and hence the increase in CO₃^2- leads to an increase in the ion activity product of [Ca²⁺]*[CO₃^2-] and to an increase in

Figure 8.9 Correlation matrices for hydrochemical parameters of Rhine stations Lobith and Braubach, data 1963-78 (Kempe 1982)

 the supersaturations of carbonate minerals and causes the seasonal deposition of lacustrine chalk. In contrast to this is a strong undersaturation established almost year-round in the fluvial regime. In the 16 years of record, calcite saturation was reached at Braubach only four times, and at Lobith only seven times. At these stations dolomite saturation was never reached.

8.4.3 LONG-TERM TRENDS

The establishment of the distinctively different biogeochemical fingerprints for the fluvial and the lacustrine environment has taken place within the last century. Unfortunately, no complete record of these alterations is available. Zobrist and Stumm (1981) used, among other information, a hydrochemical analysis of the Rhine water taken at Arnheim 1854 in order to estimate the composition of the preindustrial Rhine in percent of its present composition (Lobith averages 1974-77) (Figure 8.10). The load of NaCl alone has increased 13 times.

The load of nitrogen- and phosphorus-bearing compounds has increased similarly (Figure 8.11, Van Bennekom and Salomons 1981). The annual load of PO₄-P at Lobith has increased from 8000 t/year in 1963 to 25000 t/year in 1978 with a high correlation with time (r = 0.80, n = 16 years, signif. > 99.9% ). Interestingly, the load of NO₃-N has not increased with time during the same period at Lobith (r = 0.09, n = 16 years)  and an average of 198000 t NO₃-N/year (I 53000 t) was transported. Annual transport for NO₃-N is, however, depending on discharge (r = 0.67, n = 16 years, signif. > 99% ) Kempe 1982. This result contrasts somewhat with the correlation of the NO₃ concentration with time which is highly significant and positive (see Figure 8.9). This paradox is possibly explained by some years of above average discharges (1965- 70) at the beginning of the investigated interval which yielded high transports at relatively low concentrations.

Figure 8.10 Postulated composition of the 'pristine' River Rhine in percent of the present composition (mean, station Lobith 1974-77; Zobrist and Stumm 1981)

Figure 8.11 Concentrations of important nutrients in the Rhine and their long-term concentration changes (after Van Bennekom and Salomons, 1981). Data before 1900 (after Clarke, 1924), Data between 1928 and 1945 represent annual averages measured at Rhenen, the Netherlands, calculated after Biemond's reports 'De watervoorziening van Amsterdam', 1940 I, p- 210 and 1948, p. 110. Data after 1955 annual averages measured at Lobith (Intern. Comm. for Protection of the Rhine against Pollution 1972, 1976)

Figure 8.12 Regression analysis of pCO₂-curves for stations Braubach and Lobith 1963-78 showing statistically significant trends of  increasing pressure until 1971 and decreasing pressure since 1972. Note that pCO₂ is plotted as its negative decadic logarithm, i.e. increasing ppCO₂-values notify therefore decreasing pCO₂ (Kempe 1982) 

Figure 8.13 Long-term (1963-78) records of biological oxygen demand (BOD₅) and chemical oxygen demand (COD) for the Rhine station Lobith and Braubach (Kempe 1982)

The long-term pCO₂-records reveal (Figure 8.12) that the Rhine had a highly significant increase in pCO₂ until 1972 and a statistically as significant decrease since then. These breaks in the pCO₂ records are evident in both the stations Braubach and Lobith and in the three Dutch delta stations. This congruence in five independently monitored stations is ample proof that pCO₂ calculations can in fact reproduce historic organic carbon records to a certain extent. For comparison the BOD and COD records for stations Lobith and Braubach are plotted in Figure 8.13. Again, one can see a break in the curve at around day 3200 (i.e. c. 1972) which again marks the year since when the building of sewage treatment plants in the drainage basin lead to a reduction in the load of organics in the Rhine. Similarly, the load of some heavy metals, e.g. cadmium, has decreased in the last few years (IOE 1984). This reversal in the slope of the long-term load curves has not, however, been realized for nutrients, the removal of which by present sewage treatment plants is inefficient. By amount, the Rhine discharges about half the nutrients of the Amazon with just 1/125 of its volume. The long-term average ratios of the biogeochemical elements at station Lobith are:

348 HCO_3--C : 31 CO₂-C : 79 TOC-C : 27 NO₃-N: 1 PO₄-P.

Compared to the Redfield ratios of marine plankton of 120 C : 15 N : 1 P this means that enough NO₃-N and PO₄-P is exported to bind half of the inorganic carbon load of the Rhine as organic carbon in the North Sea. The input of nutrients by rivers can thus lead to a marked increase in biological productivity in coastal seas. The northern Adriatic Sea is an example for an eutrophied sea: 50% of its nutrient balance is already supplied by rivers (Degobbis et al.1986).

8.5 DIVERSITY IN THE TRANSPORT OF ORGANIC MATTER: THE FRENCH RIVERS

Four rivers drain 69% of the French territory (Figure 8.1). They differ widely in their discharge characteristics, their seasonal variability and their element transports.

The Seine is the smallest and shortest river of the four, but also the one being the most polluted because it drains the industrial region of Paris. The Loire on the other hand is the longest (1010 km) of the four rivers and drains the largest basin (121 X 10³ km²). It is highly eutrophic and is therefore an interesting case if one wants to study input and transformation of organic matter and nutrients in fluvial systems. The Garonne (length: 575 km; drainage area: 56 X 10³ km²; discharge: 19 km³/year) is special because it forms, together with the Dordogne (10 km³/year), the largest French estuary: the Gironde, 75 km long and up to 2 km wide. Because of the long residence time of water in this estuary, the transformation of terrestrial organic matter at the fresh water-sea water boundary may be studied in detail there. The Rhône, 812 km long, has the largest discharge (60 km³/year) of the four rivers. Because of the damming of the Nile the Rhône became the most important single fresh water source to the Mediterranean Sea. Mixing in the Rhône Estuary is limited because of the Mediterranean microtidal conditions causing the Rhône water to flow out rapidly on top of the sea water. The estuaries of the Rhône, Loire and the Garonne have been investigated in the last few years.

The Rhône originates in the Alps and flows in a narrow valley. Therefore floods may occur any time of the year. In general (Figure 8.14) the summer is the low water season and discharge is rather stable (at c. 1000 m³/s) from July to December. But from December to June floods occur with an average rate of five per year. They have peak discharges of 5000 m³/s and last a few days only. During floods concentrations of suspended and dissolved matter vary considerably, thus floods must be taken into account when calculating total exports. Average concentrations of TSS range from 5 to 10 mg/1 but can increase to 100-300 mg/1 during flood peaks. Discharge data of the last years indicate a total discharge of only 55 km³/year, somewhat lower than previous

Figure 8.14 The annual variation of discharge of the Rhône River (Aug. 1986-Sep. 1987) (data by Compagnie Nationale du Rhône, Lyon). Note the flood events during the cold seasons and the relatively stable low water discharge during summer

estimates (Cauwet and Martin 1982). The organic content of suspended matter decreases with increasing river discharge. This causes POC concentrations to stay in a narrow range, i.e. between 1.8 and 2.5 mg/l. Compared to some other European rivers this concentration range is rather low. DOC also has a rather stable concentration and total DOC export is estimated to range between 95 and 132 X 10³ t/year. DOC discharge about doubles POC export resulting in a TOC export of about 150 X 10³ t/year. These values are somewhat lower than previously estimated by Cauwet and Martin (1982).

Because the Loire originates in the Massif Central where no summer snow melts occur, its discharge is low in the warm season (200m³/s). Its average discharge amounts to 850 m³/s (27 km³/year). Floods with up to 5000 m³/s can occur at any time of the year, but are most frequent during the rainy seasons, i.e. in spring and autumn. Between 1982 and 1984 the lower section and the estuary of the Loire were investigated during 12 cruises. In November 1983 low water conditions (220 m³/s) and in May 1983 flood conditions (3500 m³/s) were encountered. In the river, TSS concentrations varied between 20 and 84 mg/l. They increase with increasing water discharge (Figure 8.15a; Cauwet and Meybeck 1987). The POC content of TSS is negatively correlated with discharge and TSS concentration (Figure 8.15b): highest values (15%) were found in June (low water) and lowest in May (3%) (high water). Soils (humic substances and clay organic complexes with low nitrogen concentrations) and fluvial epiphytes, phytoplankton and bacteria ( characterized by high phaeopigment and nitrogen concentrations) are the two main sources of POC in the river (Cauwet 1985; Meybeck et al.1988; Relexans et al.1988; Saliot et al.1984). The POC to chlorophyll and the C to N ratios are lowest in summer when photosynthesis is highest in the river but high in winter and under flood conditions (February, May) when soil erosion dominates as a POC source (Figure 8.16).

The measured DOC concentrations range from 2.5 to 6.0 mg/l. Higher values occur during floods and low values during low waters (Figure 8.17). This suggests that DOC is leached from soils during rain storms and that it must be refractory in nature mainly and it supports the observation of Mantoura and Woodward (1983) that DOC behaves rather conservatively during estuarine mixing. In the upper estuarine zone (under tidal influence but above a significant salinity increase), however, exchange processes between DOC and POC also play a role according to our observations in the Loire (Cauwet 1985; Cauwet and Meybeck 1987).

The Garonne originates in the Pyrenees, therefore its floods (up to 4000 m³/ s) occur during the rainy seasons, i.e. in late autumn and winter and during the snow melt in spring. Low waters (down to 100 m³/s) occur from late spring to early autumn. TSS concentrations rise from 10 mg/l during low flow to 500 mg/1 during floods. Turbidity (T) and discharge (Q) correlate highly significantly with each other (log T = 1.121og Q -3.73; n = 228 r = 0.80; Lin 1988). In total 1 X 10⁶t of TSS enter the Gironde from the Garonne and 0.3 X 10⁶ t from the Dordogne annually. Similar to the Loire, the POC content of TSS decreases with increasing discharge and turbidity (Lin 1988). Its main source is soil humus during floods and autochthonous matter during low water even though primary productivity is lower in the Garonne than in the Loire (Relexans and Etcheber, 1982). Garonne and Dordogne deliver a total of 3.5 and 1.3 X 10³t of POC to the Gironde annually: 70% of this is associated with floods (Q > 900 m³/s) while low waters (<300 m³/s) account for only 5% of this. The degradability of the low water POC is five times higher than that of the high water POC (Figure 8.18) (Lin 1988). Similarly, DOC levels increase from 2.5 to 4.5 mg/l with mounting water discharge. The total DOC input to the Gironde amounts to 85 X 10³ t/year, 75% of this is transported with flood waters and is rather refractory in nature (Etcheber 1986; Lin 1988).

Figure 8.15 Variation of (a) total suspended matter concentration and (b) percentage of organic carbon in the suspended matter versus discharge in the Loire River (Cauwet 1985)

Figure 8.16 Seasonal variation of the POC/phaeopigment ratio in the Loire River.. Note low values during summer which are indicative of autochthonous material and high values during winter and floods which are indicative of soil humus mobilization (Cauwet 1985)

Figure 8.17 Variation of DOC concentration with discharge in the Loire River (Cauwet 1985)

Figure 8.18 Degradation experiments with particulate organic carbon collected from the Garonne River during various seasons (Lin 1988)

8.6 CONCLUSIONS

Europe occupies 6.94% of the total land surface of Earth and contributes 7.4% of total continental runoff. Even though these figures appear small, Europe, nevertheless, plays an important role for global biogeochemical cycles due to its high chemical weathering rate, its high population density, its intensive agriculture and its highly developed industry.

Organic matter discharged, often without any prior treatment, into the rivers tends to increase pCO₂ and, consequently, to decrease pH downstream from source to estuary. This pCO₂ increase, however, appears too high if compared to the O₂ deficit, suggesting the existence of a second oxygen source for respiration, such as dissolved nitrate. The predominance of respiration processes inverts the interrelations of O₂ and NO₃-N concentrations with pCO₂ characteristic for photosynthetic processes.

On the basis of the available data reported in Table 8.4, total European river loads can be estimated (Table 8.6). For TOC an estimate of 18.9 or 22.5 X 10⁶ t/year is derived, according to which figure of total continental discharge one accepts. Such an amount would correspond to about 7% of the global organic carbon discharge from rivers (Kempe 1988) and would be in accordance with the share of Europe in total world discharge of water.

Total inorganic nitrogen discharge (NO₃-N + NH₄-N) amounts to 4-6 X 10⁶ t/year using the estimate from the cumulated measured loads (Table 8.6). Total phosphorus discharge is in the order of 0.5 X 10⁶ t/year. Nutrient discharge figures are, however, still rather vague and change much, if different calculation procedures are employed. With regard to the distribution among the inorganic nitrogen species it is-according to the data in Table 8.4¾suggested that in most European rivers 2-20% is contributed by NH₄-N. This percentage is higher in only a few listed rivers. Data are too sparse to allow estimates on the percentages of dissolved and particulate forms of nitrogen and phosphorus as yet.

From a water quality point of view, organic carbon discharge to rivers is most important since they influence the oxygen budget and therefore the health of the river fauna. Nutrients are more of a problem once the river discharged to the coastal sea where eutrophication is a rising problem.

As was shown in the case of the French rivers the majority of DOC and POC enters the estuaries during flood conditions and is highly refractory in composition. In consequence, the larger part of the DOC and POC passes the estuarine zone conservatively. Degradable organic matter is introduced only as POC during low water levels in summer when autochthonous production in the rivers runs high. This part is degraded in the macrotidal estuaries of the Garonne and Loire because of their long residence times for particulate matter but is passed to sea in the microtidal environment of the Rhône estuary.

REFERENCES

Artemyev, V. E. and Romankevich, E. A. (1988) Seasonal variations in the transport of organic matter in North Dvina Estuary. In: Degens, E. T., Kempe, S. and Naidu, A. S. (Eds) Transport of Carbon and Minerals in Major World Rivers, Lakes and Estuaries, Pt.5. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 66, pp. 177-84.

Baidin, S. S. and Kosarev, A. N. (Eds) (1986) Caspian Sea: Hydrology and Hydrochemistry, Nauka, Moscow.

Baumgartner, A. and Reichel, E. (1975) The World Water Balance, R. Oldenbourg Verlag, Miinchen, Wien, 179pp.

Benedek, P. (1989) River case study: Danube. In: Meybeck, M., Chapman, D. and Helmer, R. (Eds) Global Freshwater Quality, A First Assessment, Ch. 14, pp. 219-31, GEMS, WHO, UNEP, Blackwell, Oxford.

Brockhaus (1952) Der Große Brockhaus, 12 Vol. 16th edn., E. Brockhaus, Wiesbaden.

Cadee, G. C. (1987) Organic carbon in the Ems River and Estuary: a comparison of summer and winter conditions. In: Degens, E. T., Kempe, S. and Gall Weibin (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt. 4. Mitt. Geol.- Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 64, pp. 359-74.

Cadée, G. C. and Laane, R. W. P. M. (1983) Behaviour of POC, DOC and fluorescence in the freshwater tidal compartment of the River Ems. In: Degens, E. T., Kempe, S. and Soliman, H. (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt.2. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 55, pp. 331-42.

Carlson, H. (Ed.) (1986) Quality Status of the North Sea, Deut. Hydrogr. Inst., Hamburg.

Cauwet, G. (1985) Dynamique de la matière organique dans les milieux marin et polyhalin. Son rôle dans les processus géochimiques aux interfaces. These de Doctorat d'Etat, Université de Perpignan, 178pp.

Cauwet, G. and Martin, J. M. (1982) Organic carbon transported by French rivers. In: Degens, E. T. (Ed.) Transport of Carbon and Minerals in Major World Rivers, Pt. 1. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 52, pp. 475-81.

Cauwet, G. and Meybeck, M. (1987) Seasonal fluctuations of carbon levels in a temperate river: the Loire (France). In: Degens, E. T., Kempe, S. and Gall Weiben (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt. 4. Mitt. Geol.- Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 64, pp. 349-57.

Chapman, D. (1989) Nitrates in freshwater. In: Meybeck, M., Chapman, D. and Helmer, R. (Eds) Global Freshwater Quality, A First Assessment, Ch. 9, pp. 121-37, GEMS, WHO, UNEP, Blackwell, Oxford.

Clarke, F. W. (1924) Data of Geochemistry, US Geol. Surv. Bull. 770, 841pp.

Czaya, E. (1981) Strome der Erde. Aulis Verlag Deubner & Co KG, Köln, 247pp.

Degens, E. T., Paluska, A. and Eriksson, E. (1976) Rates of soil erosion. In: Svensson, B. H. and Soderlund, R. (Eds) Nitrogen, Phosphorus and Sulphur- Global Cycles. SCOPE Rep. 7, Ecol. Bull., 22, 185-91, Stockholm.

Degens, E. T. et al. (1982) VALDIVIA Cruise October 81 (Elbe, Weser, Ems, German Bight), collection of papers. In: Degens, E. T. (Ed.) Transport of Carbon and Minerals in Major World Rivers, Pt. 1. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 52, pp. 585-749.

Deutsche Kommission zur Reinhaltung des Rheins (since 1976) Zahlentafeln der physikalisch-chemischen Untersuchungen, Koblenz.

Eisma, D., Cadee, G. C. and Laane, R. W. P. M. (1982) Supply of suspended matter and particulate and dissolved organic carbon from the Rhine to the coastal North Sea. In: Degens E. T. (Ed.) Transport of Carbon and Minerals in Major World Rivers, Pt.1. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 52, pp. 483-505.

Eisma, D., Boon, J., Groenewegen, R., Ittekkot, V., Kalf, J. and Mook, W. G. (1983) Observations on macro-aggregates, particle size and organic composition of suspended matter in the Ems Estuary. In: Degens E. T., Kempe, S. and Soliman, H. (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt.2. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 55, pp. 295-314.

Eisma, D., Bernard, P., Boon, J. J., Van Grieken, R., Kalf, J. and Mook, W. G. (1985) Loss of particulate organic matter in estuaries as exemplified by the Ems and Gironde Estuaries. In; Degens, E. T., Kempe, S. and Herrera, R. (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt.3. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 58, pp. 397-412.

Erkushova, S. A. and Ermakova, L. F. (1986) Inorganic forms of N, P and Si in waters of the mouth area of River Don in 1965-1975. In: Hydrology and Hydrochemistry of South Seas, Leningrad, Gidrometeoizdat, Proc. of GOIN 176, 58-71.

Etcheber, H. (1986) Biogéochimie de la matière organique en milieu estuarien: comportement, bilan, propriétés. Cas de la Gironde. Mem. I. G. B. A. 19, 379pp.

GEMS (1983) GEMS/WATER Data Summary, WHO Collaborating Centre for Inland Waters, Burlington, Ontario.

GEMS (1989) Meybeck, M., Chapman, D. and Helmer, R. (Eds) Global Freshwater Quality, A First Assessment, GEMS, WHO, UNEP, Blackwell, Oxford, 306pp.

Hartmann, R. (1983) Carbon and dissolved matter output from a Central Harz Sphagnum Bog. In: Degens, E. T., Kempe, S. and Soliman, H. (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt. 2. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 55, pp. 353-63.

Helmer, R. (1989) Water quality in Perspective. In: Meybeck, M., Chapman, D. and Helmer, R. (Eds) Globa[ Freshwater Quality, A First Assessment, Ch. 2, pp. 9-28, GEMS, WHO, UNEP, Blackwell, Oxford.

International Commission for Protection of the River Rhine against Pollution (since 1972) Jährliche Tätigkeitsberichte, Koblenz.

International Commission for Protection of the River Rhine against Pollution (since 1976) Zahlentafeln der physikalisch-chemischen Untersuchungen des Rheins sowie der Mosel/Koblenz, Koblenz.

IOE (1984) Input of Contaminants to the North Sea from the United Kingdom. Final Report for the Department of the Environment, Institute of Offshore Engineering, Edinburgh, 203pp.

Keller, R. (1962) Gewässer und Wasserhaushalt des Festlandes, Teubner, Leipzig,  520pp.

Kempe, S. (1979) Carbon in the freshwater cycle. In: Bolin, B., Degens, E. T., Kempe, S. and Ketner, P. (Eds) The Global Carbon Cycle. SCOPE Rep. 13, Wiley, Chichester, pp. 317-42.

Kempe, S. (1982) Long-term records of CO₂ pressure fluctuations in fresh waters. In: Degens, E. T. (Ed.) Transport of Carbon and Minerals in Major World Rivers, Pt. 1. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 52, pp. 91-332.

Kempe, S. (1984) Sinks of the anthropogenically enhanced carbon cycle in surface fresh waters. J. Geophys. Res. 89/D3, 4657-76.

Kempe, S. (1988) Freshwater carbon and the weathering cycle. In Lerman, A. and Meybeck, M. (Eds) Physical and Chemical Weathering in Geochemical Cycles NATO ASI Series C251, pp. 197-223, Kluwer Academic Publishers, Dordrecht, Boston, London..

 Kempe, S. and Emeis, K. (1985) Carbonate chemistry and the formation of Plitvice Lakes. In: Degens, E. T., Kempe, S. and Herrera, R. (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt.3. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd.58, pp. 351-83.

 Kempe, S., Mycke, B. and Seeger, M. (1981) Flußfrachten und Erosionsraten in Mitteleuropa 1966-1973. Wasser und Boden 1981(3), 126-31.

 Kempe, S. et al. (1985) Compilation of carbon and nutrient discharge from major world rivers. In: Degens, E. T., Kempe, S. and Herrera, R. (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt. 3. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 58, pp. 29-32.

 Laane, R. W. P. M. and Ittekkot, V. (1983) Behaviour of dissolved organic waste in a part of the Ems-Dollart Estuary: the Dollart. In: Degens E. T., Kempe, S. and Soliman, H. (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt.2. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 55, pp. 343-52.

Laane, R. W. P. M. and Ittekkot, V. (1985) Dissolved and particulate carbohydrates in the Ems-Dollart Estuary. In: Degens, E. T., Kempe, S. and Herrera, R. (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt.3. Mitt. Geol.- Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 58, pp. 385-95.

Lászlóffy, w. (1967) Die Hydrographie der Donau. In Liepoldt, E. (Ed.) Limnologie der Donau, Schweizerbartscher Verlag, Stuttgart.

Lerman, A. (1981) Controls on river composition and the mass balance of river systems. In: Martin, J. M., Burton, J. D. and Eisma, D. (Eds) River Inputs to Ocean Systems. Proc. Workshop Rome, 26-30 March 1979, UNEP, IOC, SCOR, 1-4.

Lin, R. (1988) Etude du potentiel de dégradation de la matière organique particulaire au passage eau douce-eau salée: cas de I'estuaire de la Gironde. Thèse de Doctorat, Université de Bordeaux I, no.218, 209pp.

Lindeboom, H. J. and Merks, A. G. A. (1983) Annual changes in nutrient, DOC and POC concentrations and their relationship with chemical and biological processes in a closed estuary. In: Degens, E. T., Kempe, S. and Soliman, H. (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt. 2. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 55, pp. 315-29.

Livingstone, D. A. (1963) Chemical composition of rivers and lakes. In: Fleischer, M. (Ed.) Data of Geochemistry, 6th edn, US Geol. Surv. Prof. Pap. 440G, 1-61.

Lohse, J. (1983) Kohlenhydrate in Schwebstoffen der Elbe. Diplomarbeit, Fachbereich Geowiss., Univ. Hamburg, 86pp. (unpublished).

Lohse, J.and Michaelis, W. (1983) Carbohydrates in particulate matter of the Elbe estuary. In: Degens, E. T., Kempe, S.and Soliman, H. (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt.2. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 55, pp. 371-84.

Lugo, A. E. (1983) Organic carbon export by riverine waters of Spain. In: Degens, E. T., Kempe, S. and Soliman, H. (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt.2. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 55, pp. 267-79.

Maltseva, A. V. (1980) Mean perennial discharge of organic substances from the territory of the USSR and its temporal variation. Gidrokhimicheskive Materiali, Leningrad. Gidrometeoizdat, 68, 14-21.

Mantoura, R. F. C. and Woodward, E. M. S. (1983) Conservative behaviour of  riverine dissolved organic carbon in the Severn Estuary: chemical and geochemical implications. Geochim. Cosmochim. Acta 47, 1293-309.

Meybeck, M. (1982) Carbon, nitrogen and phosphorus transport by world rivers. Amer. J. Sci. 282, 401-50.

Meybeck, M. (1987) The water quality of world rivers through the GEMS Program. In: Degens, E. T., Kempe, S. and Gall Weibin (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt. 4. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 64, pp. 1-17.

Meybeck, M., Cauwet, G., Dessery, S., Somville, M., Gouleau, D. and Billen, G. (1988) Nutrients (organic C, P, N, Si) in the eutrophic River Loire (France) and its estuary. Est. Coast. Shelf Sci. 27,595-624.

Ministère de I'Environnement et du Cadre de Vie, Ministère de la Santé et de la Sécurité Sociale and Ministère de I' Agriculture (since 1975) Rivières et Canaux, Annuaire de la Qualite des Eaux, Vol. I-VI, Paris.

Ministerio de Obras Publicas (since 1974) Antáisis de Calidad de Aguas, Dirección General de Obras Hidraulicas, Madrid.

Moldan, B. (1987) Sources of seven chemical elements in the water runoff of Bohemia. In: Degens, E. T., Kempe, S. and Gall Weibin (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt. 4. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 64, pp. 375-85.

Mycke, B. (1982) Preliminary results on free dissolved phenolic compounds in natural waters. In: Degens, E. T. (Ed.) Transport of Carbon and Minerals in Major World Rivers, Pt. 1. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 52, pp. 571-4.

Mycke, B. (1985) Chemofossilien aus dem selektiven Abbau organischer Geopolymere. Ph.D. Thesis, Fachbereich Geowiss., Univ. Hamburg, 145pp.

OECD (1982) Eutrophication of Waters, Monitoring, Assessment and Control, Organisation for Economic Cooperation and Development, Paris, 154pp.

OECD (1985) OECD Data Compendium 1985. Organisation for Economic Cooperation and Development, Paris.

Pempkowiak, J. (1985) The input of biochemically labile and resistant organic matter to the Baltic Sea from the Vistula River. In: Degens, E. T., Kempe, S. and Herrera, R. (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt.3. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 58, pp. 345-50.

Pet tine, M., La Noce, T., Pagnotta, R. and Puddu, A. (1983) The Tiber River: A picture of its water quality. In: Degens, E. T., Kempe, S. and Soliman, H. (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt.2. Mitt. Geol.- Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 55, pp. 281-293.

Pet tine, A., La Noce, T., Pagnotta, R. and Puddu, A. (1985) Organic and trophic load of major Italian rivers. In: Degens, E. T., Kempe, S. and Herrera, R. (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt.3. Mitt. Geol.- Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 58, pp. 417-29.

Pet tine, A., Puddu, A., La Noce, T. and Pagnotta, R. (1987) Behavior of organic carbon and nutrients in estuarine waters of the Tiber River. In: Degens, E. T., Kempe, S. and Gan Weibin (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt.4. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE-UNEP Sonderbd. 64, pp. 407-16.

Pocklington, R. and Pempkowiak, J. (1983) Contribution of humic substances by the Vistula River to the Baltic Sea. In: Degens, E. T., Kempe, S. and Soliman, H. (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt. 2. Mitt. Geol.- Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 55, pp. 365-70.

Relexans, J. C. and Etcheber, H. (1982) Cycles saisonniers de la matière organique particulaire à la limite amont de l'estuaire de la Gironde. C. R. Acad. Sci. 294, sér. II, 861-4.

Relexans, J. C., Meybeck, M., Billen, G., Brugeaille, M., Etcheber, H. and Somville, M. (1988) Algal and microbial processes involved in particulate organic matter dynamics in the Loire Estuary. Estuary Coast. Shelf Sci. 27, 625-44.

Reuschel, A. G. and Forster, H. P. (1982) Records of TOC and other hydrochemical parameters of the Danube at Vienna, 1978-1981. In: Degens, E. T. (Ed.) Transport of Carbon and Minerals in Major World Rivers, Pt. 1. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 52, pp. 507-22.

Romankevich, E. A. and Artemyev, V. E. (1985) Input of organic carbon into seas and oceans bordering the territory of the Soviet Union. In: Degens, E. T., Kempe, S. and Herrera, R. (Eds) Transport of Carbon and Minerals in Major World Rivers, Pt. 3. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 58, pp. 459-69.

Saliot, A. and Groupe de Chimie Organique du Greco-ICO (1984) Biogéochimie de la matière organique en milieu estuarien: stratégies d'échantillonnage et de recherches élaborées en Loire (France). Oceanol. Acta 7,191-207.

Schreier, M. (1981) Ökosystem Ostsee, eine Analyse des geoökologischen Natur- raumpotentials. Diplomarbeit, Fachbereich Biol., Univ. Bochum, 115pp. (unpublished).

Seifert, R. (1982) European rivers: first results. In: Degens, E. T. (Ed.) Transport of Carbon and Minerals in Major World Rivers, Pt. 1. Mitt. Geol.-Paläont. Inst. Univ. Hamburg, SCOPE/UNEP Sonderbd. 52, pp. 523-8.

Seifert, R. (1985) Organische Substanzen in europäischen Flüssen und Ästuarien. Diplomarbeit, Fachbereich Geowiss., Univ. Hamburg, 143pp. (unpublished).

Skoulikidis, N. T. (1989) Geobiochemie der größten Flüssen Griechenlands, Ph.D. Thesis, Fachbereich Geowiss., Univ. Hamburg, 313pp. (unpublished).

Tarasov, M. N. and Beschotnova, E. I. (1987) Hydrochemistry of the Lowe Wolga and the Regulation of Discharge (1935-1980). Gidrokhimicheskie Materiali, Leningrad, Gidrometeoizdat, 101, 120pp.

UNESCO (1971) Discharge of selected rivers of the world. A Contribution to the Intern. Hydrol. Dec., Vol. I-II, UNESCO, Paris.

Van Bennekom, A. J. and Salomons, W. (1981) Pathways of nutrients and organic matter from land to oceans through rivers. In: Martin, J. M., Burton, J. D. and Eisma, D. (Eds) River Inputs to Ocean Systems. Proc. Workshop Rome, 26-30 March, 1979, UNEP, IOC, SCOR, 33-51.

Zobrist, J. and Stumm, W. (1981) Chemical dynamics of the Rhine catchment area in Switzerland, extrapolation to the 'pristine' Rhine River input to the ocean. In: Martin, J. M., Burton, J. D. and Eisma, D. (Eds) River Inputs to Ocean Systems. Proc. Workshop Rome, 26-30 March, 1979, UNEP, IOC, SCOR, 52-63.