SCOPE 42 - Biogeochemistry of Major World Rivers

1.1 INTRODUCTION

The hydrological cycle between oceans, atmosphere and continental fresh waters comes to a close in the estuaries. Estuaries are recipients of rivers, they are gigantic mixing vessels where waters with very different characteristics mix and where large physico-chemical gradients are created, larger than anywhere else in the fresh water or in the sea water realm. Therefore many biogeochemical characteristics created naturally or anthropogenically in rivers upstream develop their full environmental impact in estuaries. Vice versa, the study of estuaries teaches much about the geochemistry of particular river basins, their water discharge, their sediment load and their nutrient concentrations as well as about the physical structure of mixing waters. Since estuaries are geographically large features they are also accessible to remote sensing techniques facilitating their study greatly. In this chapter we present results obtained by remote sensing methods for a number of estuarine mixing problems and for a number of important rivers, results that illustrate the general role of rivers for the structure and chemistry of coastal seas.

River runoff forms a low salinity zone in the upper layer of an estuary and tends to move seaward, while the sea water, which has been entrained into the surface water, is replaced by deeper sea water with a movement towards the river. Lateral and vertical mixing causes sea water to be added to the effluent which is carried in an offshore direction. Vertical mixing in an estuary is rapid and directly involves river water of negligible salt content and sea water of high salinity. In moving seaward through the estuary, the upper layer is continuously increasing in volume and salt content. Deeper water rises to replace that which is carried seaward in the surface layers, whether entrained into the overlying surface layers or frictionally driven by the stress exerted by the overlying seaward-flowing surface layers. This process describes river-induced upwelling. In addition, wind-induced upwelling may drastically influence the hydrographic conditions in the estuary or the characteristics of the effluents.

The different effects of river effluents on the limiting nutrients have been studied by Park et al. (1972) who estimated the nutrient ratio from monthly nutrient concentrations for the Columbia River. During May to September the nitrate/phosphate ratio is generally less than the normal assimilation ratio of 16: 1 and may reach a minimum of 1 in June and July; nitrate then becomes the limiting nutrient in the effluent. Conversely, during the rest of the year, the nitrate/phosphate ratio is greater than 16: 1 and may reach a maximum of 67.

This fact is of importance because it may show that the productivity in an estuary may be controlled by a nutrient which has to be supplied by vertical mixing from sea water below the effluent. In other words, eutrophication in an estuary or the near coastal area may not necessarily occur when the effluent leaves the estuary; rather, very specific hydrographical conditions have to prevail in order to initiate a eutrophic effect.

Also the effect of growth-limiting metals and chelators on the growth of phytoplankton has to be considered. As shown by Barber et al. (1971), newly upwelled water has to be conditioned through natural organic compounds that are necessary to support growth rate. This suggests that phytoplankton synthesizes and releases a quantity of conditioning agents, and as the agent accumulates in the water, the specific rate of increase of the population continuously accelerates in an autocatalytic fashion until the inherent growth rate of the population is realized or until other rate-regulating properties of the system, such as light and nutrient supplies, exert their control.

These scenarios show the complicated nature of the processes involved to initiate eutrophication in a river effluent. As wind mixing, upwelling, as well as the discharge of effluent, are rather fast changing processes, it is difficult to identify the location of eutrophication and to monitor their dynamics with conventional methods. Therefore, an attempt has been made to use information from satellite altitudes in order to identify the distribution patterns of sediments in estuaries, to estimate the range of the major river discharge, and to investigate the eutrophic effect of river discharge and its interaction with other water masses.

1.2 REMARKS ON THE SALTWEDGE

Based on conventional parameters it is obvious that eutrophication in the saltwedge is a result of remineralization processes within the wedge in connection with upwelling processes. Therefore the recycling of nutrients through micro-organisms is probably the most important factor stimulating enhanced phytoplankton growth. The generally accepted form of recycling of nitrogen through bacterial decay of marine organisms is based in its final step on the oxidation of ammonia to nitrite and consequently nitrate. However, it has been shown that all three species (NH₄⁺, NO₂^- and NO₃^-) can be used by phytoplankton.

Some phytoplankton species, e.g. dinoflagellates, on the other hand are able to use dissolved organic nitrogen compounds for their nitrogen requirements. This in fact shows that nitrogen can be taken up in different forms which may not be monitored with the standard procedures of measuring nutrient levels and uptake kinetics.

Figure 1.1 Salinity, density and temperature along 1 °N in the Amazon estuary during May/June

The stratification of the river plume for the Amazon is given in a section along l °N during May/June, where the river outflow is clearly visible in the salinity and density distribution (Figure 1.1). In the offshore region slight upwelling is indicated by salinities higher than 36 per thousand, by increased density (above 236s_t) and by temperatures below 27 °C.

The origin of this water can be traced to a depth of about 10-150 meters as can be shown by data presented in Figure 1.2. Typical for all seasons is the distinct salinity maximum which is mainly a result of water from higher latitudes exposed to heating at the surface. The salinity maximum in this data set is more pronounced during November than May. Important for the mechanisms and interaction of this water with the saltwedge is the low concentration of nutrients, especially nitrate being almost zero. Therefore the transport of this water does not necessarily contribute to the nutrient concentration but rather acts as a medium to bring water from the saltwedge to levels with high illumination. There mineralization and recycling of nutrients in the saltwedge can be described with salinity and data on the apparent oxygen utilization (AOU) in Figure 1.3. With respect to oxygen the surface layer still is unsaturated on two stations with low salinities while the further downstream station with higher salinities has AOU values of zero for the upper layer. This indicates that primary production as well as wind-induced turbulence compensate for biological oxygen uptake.

Figure 1.2 Distribution of salinity, total phosphate and nitrate in an offshore station east of the Amazon plume

As Edmond et al. (1981) pointed out, at the start of mixing between the river and ocean water, suspended matter from the Amazon settles rapidly and the resulting water is sufficiently transparent for increasing photosynthesis to commence at around 7 per thousand salinity. As a result, high diatom productivity has been observed at a salinity between 7 and 15 per thousand, which depletes dissolved nutrients. Such a diatom bloom has been observed by Edmond et al. (1981), in May/June 1976, which removed nitrate and phosphate from the surface layer while the underlying saltwedge was found to be enriched by re mineralization of planktonic residues. Also the regeneration of carbon, phosphorus, silica and nitrogen has been studied by the same authors indicating that 58% of the nitrogen was unaccounted for, having been dissolved to species other than nitrate and nitrite. Therefore nitrogen in other forms, as well as different uptake kinetics and decay mechanisms for nitrogen in the metabolic processes may account for the importance of the saltwedge.

Figure 1.3 Apparent oxygen utilization and salinity for three hydrographic stations in the Amazon plume

As upwelling and intense vertical mixing in the river effluent can easily be induced by topographical conditions and wind factors, the strong halo-stratification is broken down and nutrient-enriched saltwedge waters and suspended material are introduced to the surface. Therefore, as has been shown with data from different locations in the vicinity of the Amazon effluent, no simple relationship exists between the conservative parameters temperature and salinity and the non-conservative nutrients and oxygen. This is especially true for nitrite which occurs in high concentration at low salinities as well as at high salinities.

In the effluent of the Amazon a very thin layer of brackish water overlays ocean water with S > 35 per thousand. This stratification can easily be broken into separate lenses with fresh water from the surface or high saline water from below. Satellite data indicated separated patches with different reflection values which show the mixing between the river effluent and the ocean water. Two processes can be recognized: the first is the injection of ocean water from below to the surface; the second process is the separation of plumes from the major river effluent into the ocean. Both processes may have the same or similar impact on the biogeochemistry of the ocean/river system although the intrusion of high saline water to the surface may be regarded as an upwelling process.

Biological productivity, nutrient regeneration and physical action on a river effluent make it difficult to describe in generalized form the behavior of nutrients and their uptake and remineralization. Therefore all available stations for the area of the Amazon effluent in the upper 25 m have been grouped in the 5 per thousand increments and are presented in Figure 1.4. Silica is removed from the surface layer continuously. Nitrate concentrations show a very similar tendency up to a salinity of about 20 per thousand. In comparison, the average nutrient concentrations at zero salinity (Edmond et al. 1981) are: 128 µmol SiO₂/l, 0.52 µmol PO₄^-3/l, 8.5 µmol NO₃^-1/l and 0.15 µmol NO₂^-1/l. The high nitrite concentrations at the first mixing between river and ocean water may be interpreted as a consequence of microbiological aerobic activities and not by denitrification because all stations show oxygen levels above those where denitrification starts. Nutrient regeneration is mainly expected at the saltwedge where especially phosphate data show a higher relative enrichment from remineralization processes of biogenous debris from diatom blooms (Edmond et al. 1981) or deposited and trapped particulate material from the Amazon. The resulting higher concentration of nitrate, as well as of phosphate, is visible at higher salinities. For both nutrients intermediate maxima occur at salinities of about 30 per thousand. These salinities are found in locations with strong vertical haline gradients; therefore any turbulence or vertical motion to the euphotic layer increases productivity. Edmond et al. (1981) reported that at higher salinities the concentration of particulate carbon and nitrogen increases, which indicates the separation of large particles with low carbon concentration but also shows the increasing primary productivity toward higher salinities. Edmond et al. (1981) observed at salinities less than 10 per thousand and in the saltwedge, that the particulate matter is largely composed of Amazon detritus. 'Beyond the region of high productivity out to oceanic salinities, the surface samples are predominantly biogenic, many being bright green and composed of a tremendous diversity of diatoms.'

Figure 1.4 Nutrient concentration in the Amazon plume based on average data over 5 per thousand increments in the upper 25-meter layer. Numbers in the curves give the number of samples used

Possible ways of enhancing the primary productivity in the Amazon effluent have been discussed earlier, and upwelling may play a major part in eutrophication. Upwelling may include under the circumstances in the Amazon effluent only slight uplifting of the halocline into higher light intensity levels. Therefore wind conditions may initiate locally limited upwelling and contribute to patchiness in the distribution of biomass. Hulburt and Corwin (1969) showed the distribution of eight abundant neritic species. Their distribution pattern shows not only that highest concentrations are found in the vicinity of the estuary but also that patchiness is difficult to explain by data such as salinity alone.

Huang et al. (1983), by examining the output fluxes for the Changjiang of dissolved silicon, nitrogen and phosphorus, showed that SiO₂-Si, NO₃-N and salinity, because of their linear correlation, could, for the time of investigations, be calculated knowing that the concentrations of dissolved SiO₂ and NO₃-N in the Changjiang mouth were 122.2 µg-at/l and 58.17 µg-at/l, respectively. In consideration of the Changjiang River runoff at 925 km³/year the output fluxes were estimated to be: 316 x 10⁴ t SiO₂-Si/year (or: 677 x 10⁴ t SiO₂/year); 75 x 10⁴ t NO₃-N/year.

With respect to a survey on silicon, phosphorus and nitrogen, Huang et al. (1983) concluded that none of these nutrients is a limiting factor in primary production for the Changjiang. For nitrogen, this was based on the observations that inorganic nitrogen exists mainly in NO₃-N form, which makes up 95-97% of the total dissolved inorganic nitrogen. The highest NO₃-N concentration reaches 77.4 µg-at/l, three to four times higher than that of 18 years ago, due to the rapid increase of chemical fertilizer consumption and discharge of industrial sewage and waste water. This is also evidenced by an unusual atomic ratio between nitrate and phosphate which ranged for NO₃-N : PO₄-P from 88 to 150. This high N : P ratio is characteristic for the three largest Chinese rivers when compared to other large rivers, which have N : P ratios close to 15 (Kempe 1984).

Basin-like estuaries such as the Rio de la Plata (see Figure 1.5) for instance have the exceptional feature of river discharge being injected into the basin and mixing migrates through a semi-enclosed region toward the ocean. During this migration, all processes such as sedimentation, remineralization and plankton growth are progressing through the system. Although there is no real steady state to be expected, the zone of eutrophication should show up as a region almost perpendicular to the axis of the basin but in an offshore direction.

Turbidity maxima as a function of the tide have been observed in the Elbe River by Koske et al. (1966). The maximum transparency of the water has been found at low tide; total particle load and particulate carbon concentration showed minimum values at the same time. Maximum values in particulate matter have been observed shortly after the maximum current velocity was reached. This can be explained by the redistribution of suspended matter which is deposited during the lowest current velocity. Similar processes may appear in other estuaries which are under the influence of tidal currents. Therefore, in such systems local maxima of suspended matter can be postulated in the vicinity of the estuary, while further out, eutrophication, as a result of mixing between the effluent and coastal water, can be expected. Although only a few satellite images have been available for the Rio de la Plata region, it has been shown that the different expected turbidity maxima can be resolved to a high degree (Szekielda et al. 1983).

Figure 1.5 The Rio de la Plata observed in the visible

1.3 PATCHES IN RIVER PLUMES

The smallest patch included in the following analysis is based on continuous salinity recordings with silver/silver chloride electrodes in the effluent of the Rhône River (Szekielda and Kupferman 1973). Repetitive crossings of the outflowing Rhône water and an anchor station in the effluent showed that patches in the range from 0.1-0.6 km have a resident time of 0.004 days, while patches at a scale of several kilometers may exist for at least in the range of some hours (Figure 1.6). Observations from the GOES (Geostationary Operational Environmental Satellite) on the ocean/effluent boundary of the Amazon indicated that the lifespan of patches between 30 and 50 km is in the range of 0.5 to 1 day. The next range of changes in the Amazon River effluent in the neighborhood of 100 km is based on time sequences also observed with GOES (shown in Figure 1.7). Basically, the intrusion of salt water lenses into the effluent, the separation of fresh water lenses into the ocean, and the oscillation of the ocean effluent boundaries have been observed to have a residence time of about 5 days for a patch size of about 90 km.

Figure 1.6 Continuous salinity recordings in the Rhône River effluent (modified after Szekielda and Kupferman 1973). (a) Repeated salinity recordings. Upper trace represents the salinity distribution at a wind speed of approximately 10 m/s, lower trace at calm conditions-recordings were taken six hours apart. (b) Salinity variations at an anchor station six miles offshore from the Rhône Estuary

The satellite observations for the Changjiang in connection with regular oceanographic observations showed that during the period of investigations March through November, the main current pattern is very similar. As a consequence of river discharge, tidal action, erosion, upwelling and the coastal hydrography make it difficult to identify the area of eutrophication, although patterns recognized in the satellite data show increased reflectance in the offshore patches connected with upwelling and coastal runoff. Patches seen in the offshore region may not only be derived from the Changjiang. Studies have suggested that the offshore patches are being generated from the flat area off Jianggang in connection with river discharge, upwelling processes and development of algae blooms (Figure 1.8).

Figure 1.7 Intrusion of a salt water patch into the river effluent of the Amazon. The patch outlined in the shaded area appeared in GOES data on 2 December 1981 and was not detected on 7 December 1981. (b) Patches along the ocean effluent boundary of the Amazon region as detected in GOES data. (c) Separation of a fresh water patch from the Amazon effluent into the neighbouring ocean. The patch was not observed after 18 September 1981

For the summertime Limeburner et al. (1983) found that fresh water and sediments from the Changjiang flow initially to the south along the shore on the Chinese inner shelf. The river discharge eventually flows northeastward after mixing with Taiwan Current water. Some low salinity, low sediment river discharge flows directly northeastward from the river mouth as an effluent as shallow as six meters.

Figure 1.8 Details on the eddy structures of suspended matter on 9 March 1982 in the Changjiang plume off Jianggang

Le (1983) pointed out that the path of the Changjiang effluent, on average, is toward the southeast during April, May and September while the path during summertime is in the northeast direction with a sharp turning point within the near-shore section (Figure 1.9). This is a generalized conclusion and it has to be kept in mind that deviations from such an observation may occur frequently. Le (1983) presented data from which he concluded that apparently three turning patterns described above are associated with different factors, such as the amount of Changjiang discharge, the portioning of discharge through the estuary branches, the position and strength of the coastal current and/or the Taiwan Warm Current as well as wind stress. The most important factor is the discharge and its distribution into the delta arms. Field data show that the higher the runoff, the stronger will be the tendency of the Changjiang diluted water to turn left, and vice versa.

In view of the fact that freshened coastal water is transported in an offshore direction, it is questionable whether the data displayed in Figure 1.9 actually show the pathway of the Changjiang diluted water. Based on the satellite data, it seems rather plausible that the fresh water in the offshore region north of the mouth is generated in conjunction with discharge from other tributaries.

Figure 1.9 Paths of the Changjiang diluted water (CDW) as inferred from salinity measurements. The averaged paths in April, May and September are towards the southeast, but the paths in summer turn to the northeast and the locations of their turning point are within the near-shore section. Paths of the CDW observed in different months have various patterns and this is the case even in summer (after Le 1983)

Such a conclusion also is supported from the bathymetry of the investigated area, which shows that the different water masses identified in satellite images are closely connected with bathymetry as well as with the circulation system.

Under the assumption that the separated patches in the offshore regions are affiliated with tidal action over the coastal plain and that each tide would generate one patch to be dispersed offshore, one can at least estimate the transport characteristics. Other typical patterns observed are shown in Figures 1.10 and 1.11. One can recognize that the distribution pattern in particulate matter in the coastal area is not identical with the thermal pattern. This, in fact, proves that even in that region most of the signatures are due to varying concentrations of suspended matter. With respect to the patches observed in the offshore region on 16 Apri1 1981, it is remarkable that the size of the patches and their distance from each other is close to what has been observed on 9 March 1982. Two injection sites for the patches can be recognized and it is evident that the patches are being partly generated over the flat area (Figure 1.12). This in fact shows an additional offshore current component but still confirms the data from March. A detailed comparison between the reflectance values and the thermal data showed also that there is no simple relationship between the two parameters. This indeed indicates the non-conservative behavior of the pattern which, inter alia, includes sedimentation and eutrophication processes while the patch is being carried into the region.

A temperature section for the same month but different year (Figure 1.13) shows that the thermocline is at about 10 meters. Therefore, it may be speculated that low wind stress or turbulence may only be necessary to bring the cold water to the surface. Also, tidal action in connection with wind stress will modify the path of the Changjiang diluted water.

 1.4 LIFESPAN OF PATCHES

 The mixing of fresh and salty water causes not only a vertical salinity gradient, i.e. stratification, but also lateral inhomogeneities in the form of patches.

During the last several years, our studies with satellites on estuaries have shown that a rather complex pattern exists in the distribution of particulate matter. This patchiness is mainly the impact of gravitational and meteorological forces and also, to a high degree, is a consequence of diffusion and advection processes. As these processes are of high importance to the development and production mechanisms of organic matter, especially in a complex environment such as a river effluent, the mechanics of mixing between fresh and salt water determines the rate of primary production in the river effluent. The saltwedge in the effluent may be regarded as the major compartment for the recycling of nutrients and growth-stimulating agents and, in consequence, is the dominant factor for the development of the food chain in the effluent.

Figure 1.10 Comparison between the pattern in the visible (a) and the thermal patches (b) on 16 Apri1 1981, Changjiang coastal area

Figure 1.11 Comparison between visible (a) and infrared (b) data in the upwelling water of the Changjiang coastal area

Figure 1.12 Comparison between infrared recordings (a) and data in the visible (b) over the coastal area of the Changjiang

Figure 1.13 Temperature section off the Changjiang in August 1981 (after Lime-burner and Beardsley 1982)

As mixing is not a steady-state process, gradients are developed in the effluent. Patches generated in this manner range in scale from the molecular level to patches as large as 300 kilometers in diameter. From pure physical impact alone, it is evident that, if one considers a conservative parameter, the patchiness is determined by the factors involved in the mixing processes. In another dimension, namely the growth of organisms or the decomposition of organic material, the situation becomes more complex. It has been shown that the size of a generated patch is one of the factors that control the development of the food chain or the breakdown of an existing biological system. This means that a diffusing phytoplankton patch in the effluent is affected by eddies larger than the patch size as a whole, while smaller eddies disperse the phytoplankton. Shear and/or spatial variations in the velocity field of eddies with similar size as the plankton patch also deform the patch by advecting one part of the patch relative to another.

In an estuary and within the effluent of a river, the relationship between plankton and the dilution process is increasingly complicated by variable light conditions, sedimentation of organic material and the recycling of nutrients from the saltwedge. In a river effluent, highly stratified in a very thin layer of about 10 meters with respect to physical, chemical and biological parameters, even small-scale processes such as wind stress and fluctuations in current shear, will influence not only the distribution patterns of organic matter but also change the conditions for recycling of nutrients and primary production. Consequently, the size and lifespan of a physically identified patch is of importance to the biological conditions within the patch.

With regard to patchiness in oceanic parameters, one has to keep in mind that with different patch sizes, different processes are also involved in their dispersal and dilution with adjacent water masses. As was outlined by Okubo (1978), the division between one part of the motion which is assigned to the advective processes or to the diffusive processes, depends not only upon the spectrum but also on the scale of the phenomena which we observe in the ocean. This means that the continuous spectrum containing the scale range from microturbulence, or even down to molecular diffusion to the ocean wide circulation, has to be linked to the pure mixing effect and also to the time-space-dependent chemical and biochemical processes in the ocean.

1.5 DISCUSSION AND CONCLUSIONS

In discussing patchiness in the distribution of particulate matter in the oceans one has to consider biological, physical and chemical conditions simultaneously, which makes it difficult to interpret single measurements of biological parameters properly. So far no instantaneous measurements exist and, at best, measurements of chlorophyll, temperature and salinity can be considered quasi-synoptic over distances of 10-100 kilometers in a few hours (Steele 1978). Compared to the time scales of changes in relation to the lifespan of organisms, species composition and food chain relations, this may be considered sufficient to resolve the size in patchiness. However, advection and diffusion are on very different time scales but both may influence the physiological behavior and growth rate of primary producers. In response to changes in wind stress, turbulence and currents may also act on a water mass and can modify the distribution patterns of particulate matter in the horizontal and vertical direction. Also the strong density gradient in the effluent plays an important role in the recycling, remineralization and uptake kinetics of nutrients, especially those released by particles which may have been trapped in the pycnocline. A few remarks are necessary to explain the important role of particulate matter carried through the effluent to the marine environment. Larger particles, mainly containing silicate, settle very fast, while the smaller particles and organic debris will remain longer in suspension.

Vertical motions affect the residence time of particles at a given depth in the water column either positively or negatively depending on the direction of movement. Since it is difficult to obtain in situ observations on the influence of vertical movements on the residence time of particles in a given space, a few theoretical aspects will be considered. The settling velocity of a particle in the ocean depends upon the radius of the particle, the density of the particle and the density of kinematic viscosity of sea water whereby the density of sea water is a function of temperature and salinity. Assuming that a particle is spherical and has a constant viscosity and mass, Stoke's Law can be applied in estimating its settling velocity ( v),

where g is the gravitational acceleration constant (980 cm/s²), (p the density of the particle in g/cm³, (w the density of sea water in g/cm³, r the radius of the particle in cm and v the viscosity of sea water in cm²/s. Based on data collected from a reference station at 15° 14' W, 27° 30' N, the settling velocity as a function of the diameter of particles and the density of sea water was calculated. It shows that even at low vertical velocity to the surface, particles with a diameter of < 5 µm will rise to the surface. In considering the sinking velocity of particles, the change in weight and modification in the shape of  particles have been taken into account; however, estimates (in which a weight loss of 0.3% was assumed) show that over 30 days and relative to dissolution in the upper layers, there were no significant changes in the sinking velocity. Thus it may be concluded that particle size and vertical velocity are the major determinants for the position of particles at a given time. Residence time of particles at a given depth may be altered, particularly with smaller particles, by the vertical motion within an upwelling system as was demonstrated in a study which estimated the position of <5 µm particles as a function of time and vertical velocities. The results showed that in the upwelling zone all particles <5 µm were trapped within the euphotic zone with an increased residence time.

Steele (1978) pointed out that the minimum scale at which phytoplankton growth can maintain patch-structure in face of horizontal turbulent dispersion is of the order of a few kilometers. In view of the complex nature of the dynamical behavior of an effluent, it is therefore questionable that the real distribution pattern of particulate matter can appropriately be resolved by ship measurements alone. In light of these considerations, analysed satellite data provide an important input for further guidance on how to monitor patchiness on a large scale.

The summary of the data on the lifespan of observed patches is presented in Figure 1.14. In comparison with the time scale of biological growth, one recognizes that patches only in the range of kilometers and upward may be of statistical significance, although one has to keep in mind that if steady-state conditions appear, development of a plankton bloom may occur in much smaller patches.

These findings should be compared with experiments measuring the dispersion of dyes (Okubo 1978) based on 'point source' injections, by deriving the empirical relation between the standard deviation (km) and time t (days):

The deviation between the patches as observed in the present study can be explained by the fact that the dispersion of dye injections measures only a limited range of possible mixing processes, while the satellite data include processes over all the possible ranges and, henceforth, represent a more realistic observation than the extrapolation of dye experiments into large-scale mixing processes, as has been suggested by Steele (1978) in order to explain patches in an ocean system.

Although it has been shown that the color gradients undergo rapid changes and build separate eddies or patches, a few general conclusions can be derived for the investigations, which can be confirmed with data reported earlier (Szekielda 1982). Based on the reflectance measurements from satellites, it is evident that sedimentation of particles occurs very fast in the vicinity of the estuary at a rather low salinity. Milliman et al. (1975) pointed out that more than 95% of the terrigenous sediment present in Amazon surface waters settles out within the river mouth before the salinity reaches 3 per thousand. As a consequence, higher light intensities and mineralization of organic detritus, largely from coastal vegetations, cause relatively high nutrient concentrations (Cadée 1975). In conclusion, the area of eutrophication is a vital region in the ecosystem of coastal areas and may play a more important role than one thought in the past.

Figure 1.14 Relationship between patch size (in kilometers) and lifespan (in days) as based on satellite observations and at the lower range based on continuous salinity measurements in the Rhône River

From the data presented it can be concluded that a certain patch size is required in order to develop the eutrophication in the effluent of a river. This, however, is a function of the stability of a water mass, i.e. mixing processes initiated by wind stress, for example, may locally vary the potential for plankton development and, on the opposite side, a steady state may develop maximum plankton concentrations. Within the effluents, it may be speculated that maximum concentrations of plankton should be observed at the outer part of the plume surrounding the estuary in a 'circular fashion'. This is under the assumption that a continuous outflow of the river water is present under normal or average conditions. Under extreme weather conditions, for instance, total lack of wind stress may create a patch which may not be in the lifespan-patch size relationship, as shown in Figure 1.14. Therefore, the data in this figure may represent an average condition with the understanding that anomalous patches may be created which do not fit the established correlation. Nevertheless, the results shown give an insight into the important correlation between the biological expectation within a generated patch and the required patch size. For instance, the food chain may rarely develop in patch sizes of hundreds of meters with a low lifespan, although the mixing processes will generate an environment more suitable for biological development at a higher level of the lifespan-patch size relationship. The concept of patches at this point disregards the vertical stratification of the marine environment. However, the gradients dealing with the observed relationship are extreme and it is speculated that the vertical stability of a patch also is related to its size and survival time.

The importance of an equilibrium between the physical conditions, such as wind stress and the nutrient pool, for example, would still have to be considered separately.

REFERENCES

Barber, R. T., Dugdale, R. C., MacIsaac, J. H. and Smith, R. L. (1971) Variations in phytoplankton growth associated with the source and conditioning of upwelling water. Inv. Pesq. 35(1), 171-93.

Cadée, G. C. (1975) Primary production of the Guayana Coast. Neth. J. Sea Res. 9(1) , 128-43.

Edmond, J. M., Boyle, E. A., Grant, B., and Stallard, R. F. (1981) The chemical mass balance in the Amazon plume I: The nutrients. Deep-Sea Res. 28A(11), 1339- 74.

Huang, S., Yang, J., Ji, W. and Yang, X. (1983) Silicon, nitrogen and phosphorus in the Changjiang River mouth water. In: Sedimentation on the Continental Shelf, with Special Reference to the East China Sea, China Ocean Press, pp. 241-50.

Hulburt, E. M. and Corwin, N. A. (1969) Influence of the Amazon River outflow on the ecology of the western tropical Atlantic. J. Mar. Res. 27, 55-72.

Kempe, S. (1984) Sinks of the anthropogenically enhanced carbon cycle in the fresh water system. J. Geophys. Res. 89 D/3, 4657-76.

Koske, P. H., Krumm, H., Rheinheimer, G. and Szekielda, K.-H. (1966) Untersuchungen über die Einwirkung der Tide auf den Salzgehalt, Schwebstoffgehalt, Sedimentation und Bakteriengehalt in der Unterelbe. Kiel. Meeresforsch. 22, 47- 63.

Le, K. (1983) Some results from the study of the path of Changjiang diluted water. In: Sedimentation on the Continental Shelf, with Special Reference to the East China Sea, China Ocean Press, pp. 375-81.

Limeburner, R. and Beardsley, R. C. (1982) Hydrographic station data obtained in the East China Sea. Technical Report WHOI-82-10, 159pp.

Limeburner, R. , Beardsley, R. C. and Zhao, J. (1983) Water masses and circulation in the East China Sea. In: Sedimentation on the Continental Shelf, with Special Reference to the East China Sea, China Ocean Press, pp. 285-94.

Milliman, J. D. C., Summerhayes, C. P. and Baret to, A. T. (1975) Oceanography and suspended matter off the Amazon River. J. Sed. Petrol. 45, 189-206.

Okubo, A. (1978). Horizontal dispersion and critical scales for phytoplankton patches. In: Steele, J. H. (Ed.) Spatial Pattern in Plankton Communities. NATO Conf. Ser. IV, Mar. Sci. pp. 21-41.

Park, P. K., Osterberg, C. L. and Forster, W. O. (1972) Chemical budget of the Columbia River. In: Printer, A. T. and Alverson, D. L. (Eds) The Columbia River Estuary and Adjacent Ocean Waters, Univ. of Washington Press, pp. 123-34.

Steele, J. H. (1978) Some comments on plankton patches. In: Steele, J. H. (Ed.) Spatial Patterns in Plankton Communities. NATO Conf. Ser. IV, Mar. Sci. pp. 1-20.

Szekielda, K.-H. (1982) Investigations with satellites on eutrophication of coastal regions. 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. 13-37.

Szekielda, K.-H. and Kupferman, S. (1973) Heterogeneities in salinity in a river plume. Estuar. Coast. Mar. Sci. 1, 419-24.

Szekielda, K.-H., McGinnis, D. F. and Gird, R. (1983) Investigations with satellites on eutrophication of coastal regions. 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. 55-84.