6.6 Combating eutrophication in the Seine river and the Seine Bight :
scenarios analysis6 Drainage basin use and nutrient supply by rivers to the
coastal zone.6.4 Origin and processes of nutrient transformation in the river continuum6.5 Modelling phytoplankton and nutrient in drainage networks

6.5 Modelling phytoplankton and nutrient in drainage networks

6.5.1 Seasonal and geographical variations of phytoplankton development and nutrients

By coupling the hydrological and ecological models, it is possible to calculate the mean phytoplankton growth rate, integrated over the water column, at any point in the hydrographic network. Phytoplankton only develops when its growth rate is higher than the rate of dilution by lateral inflows. Therefore, phytoplankton can only grow significantly in rivers from order 5 upward (Figure 8). As the depth increases with the order, the mean phytoplankton growth rate decreases in the higher orders under light limitation. The low growth rates upstream are the result of nutrient limitation.

josFig8

Figure: Comparison of the variations in the dilution rate and phytoplankton growth versus the stream order, for a spring situation (June).

 

As described above, the model is able to adequately simulate the seasonal variations of phytoplankton in the main sub-basins as well as in the main axis of the Seine River (Figure 9). As predicted by the comparison between the dilution rate and the phytoplankton growth rate, the phytoplankton biomass is only significant from order 5 upward and therefore increases from upstream to downstream. In accordance with the observations, the model predicts a spring peak dominated by diatoms and a summer peak dominated by Chlorophyceae. In the downstream part of the sub-basins, the model reproduces the sudden declines in spring bloom. In the summer, phytoplankton biomass again increases downstream of Paris and reaches its maximum in the estuary, downstream of Poses [].

josFig9a

Figure: Seasonal and geographical variations of phytoplankton development in the Seine (1991). a) outlet of the major sub-basins and mouth of the river; b) Marne from order 4 (Donjeux) to order 6 (mid 6: Reuil, outlet: Neuilly).

 

Simulations of the seasonal variations in phytoplankton biomass are also shown for 10 different years with various hydrological conditions at the outlet of the Marne sub-basin, taken as 9 sub-basins (Figure 10). The control of the spring bloom of phytoplankton by the flow rates is confirmed. As soon as the flow rates have decreased in late winter or early spring, the phytoplankton builds up a large biomass that remains high if the water flow stays low (in 1991, 1993, 1997). The early spring bloom can be disrupted by a late rainy event (in April 1998), whereas when high flow rates are maintained until spring, the bloom may be delayed and reduced in intensity (in 1995 and 1999).

josFig10

Figure: Simulated seasonal variations of phytoplankton development (dots) at the outlet of the Marne (Neuilly-sur-Marne) for two three year periods (1991-1993: dry; 1999-2001: wet) throughout a ten year period. Seasonal variations of water fluxes are shown for comparison.

 

The model was used to investigate in what conditions the different nutrients might become limiting for the phytoplankton and what would be the impact on its mean summer and spring levels (Table 4).

Simulated reduction in spring phytoplankton biomass (mg L-1 of chlorophyll a, mean value in April and May, expressed in percentage) as a function of change factor in total inputs (F = 1 for the reference situation): diffuse and point inputs of nitrogen (N) and phosphorus (P), diffuse inputs of dissolved silica (Si). Results for 1991 at the outlet of the Marne.

F

N P S
Chl a %
0.01 47.8 25.919.3
0.1 94.2 44.322.8
0.3 99.4 76.9 39.5
0.5 100.0 96.6 48.6
1 100.0 100.0 100.0
1.5 100.0101.9119.2
3 100.0103.3125.0

This simulation confirms that nitrogen is far from being limiting : only a very strong reduction in the fluxes (99%) would cause any effect on the spring averages of the phytoplankton biomass. A phosphorus reduction scenario is more realistic given its point sources, but it would still have to be considerable (90% reduction) to affect the spring bloom of phytoplankton. The phytoplankton is also very sensitive to dissolved silica variations since a 50% reduction in the input significantly decreases its biomass. Although such a reduction is not feasible, because of the diffuse origin of silica due to rock weathering, its retention in stagnant systems (large reservoirs) could nevertheless influence its concentration downstream and subsequently change the phytoplankton successions as observed in the Danube [237]. In the Seine basin the reservoir impact is not sufficient to markedly lower the silica on the whole basin.

At the sub-basin outlets where the phytoplankton decline occurs, phosphorus and nitrogen are not limiting. Regarding silica, river algal bloom dominated by diatoms may temporarily bring down its concentration to very low levels which would doubtless be limiting for the phytoplankton.

6.5.2 Autotrophy vs heterotrophy in eutrophic and/or polluted rivers: the Seine, the Loire, the Mosel and the Scheldt Rivers

Most of major Western European rivers have been extensively studied but rarely in an integrated manner, at the scale of their watershed basins. Because of the modular structure of the RIVERSTRAHLER model, it is possible to adapt the procedure to each case. The four rivers chosen here are situated in morpho-climatic domains similar to that of the Seine []. They can therefore be used to explore an increasing gradient of population density and of agricultural and industrial activity: their watershed basins are fairly small (Table 5). At their mouth the stream-orders are 8 for the Loire and the Seine, and 7 for the Mosel and the Scheldt. The hydrological regimes are all characterised by winter and early spring floods and summer low flow. The population density increases clearly from the Loire to the Scheldt according to a South-North gradient. As in the case of the Seine, human activity has greatly contributed to altering the hydrological regime of the rivers (navigation channels, wetlands draining, construction of reservoirs for flood protection). Among the rivers, the Loire is relatively untouched by engineering works, although it has two large dams in the upstream sector, there is no navigation.

Simulated reduction in spring phytoplankton biomass (mg L-1 of chlorophyll a, mean value in April and May, expressed in percentage) as a function of change factor in total inputs (F = 1 for the reference situation): diffuse and point inputs of nitrogen (N) and phosphorus (P), diffuse inputs of dissolved silica (Si). Results for 1991 at the outlet of the Marne.

river

watershed mean max. depth reservoirs % arable popul
systems area, km2 discharge order regul. land density
m3s-1 inh, Km2
Seine 64 500 420 8 yes 3 major 46 195
Loire 116 000 580 8 no small 38 63
Scheldt 6 200 80 7 yes none 39 426
Mosel 17 890 300 7 yes none 20 190

As for the Seine River, phytoplankton develops from upstream to downstream in the Loire, Mosel and Scheldt (Figure 11). At the outlet of the hydrographic network, the level of nutrients and the main pattern of their seasonal variations are found by the model (Figure 12).

josFig11

Figure: Simulations, by the RIVERSTRAHLER model, of phytoplankton biomass (Chl a) in the Seine (1994), the Loire (1989), the Mosel (1994), the Scheldt (1987) at the outlet of the their network.

 

josFig12

Figure: Simulation of seasonal variations of water fluxes (m3 s-1), nitrates (mg N l-1), phosphates (mg P l-1), silica (mg Si l-1) in the Seine, the Loire, the Mosel an d the Scheldt.

 

In classical ecology [], the trophic state of a system is determined by the equilibrium between the autotrophic metabolism that produces oxygen and the heterotrophic metabolism that consumes it. On this basis, one can therefore establish a diagnosis of the autotrophic-heterotrophic state of a system and its variations in time and space. Autotrophic-heterotrophic diagrams (P/R diagrams) show, on the y-axis, the intensity of the metabolic production of oxygen or carbon in the system (P, algal photosynthesis) and on the x-axis, that of the processes of oxygen consumption or of organic matter degradation (R, respiration of bacteria, zooplankton and the benthos as well as of phytoplankton). In this type of diagram, the systems at equilibrium have an oxygen concentration at saturation level and fall on the diagonal, with increasing distance from the origin as the biological activity rises. Predominantly autotrophic systems, net producers of oxygen, lie above the diagonal (P/R >1) while the systems dominated by heterotrophic processes, net consumers of oxygen, fall below the diagonal (P/R <1), (Figure 13). The P/R diagrams are therefore a means of summarising the trophic state of an aquatic system during its evolution in time and space. Although such diagrams can be drawn on the basis of experimental data [], the model, validated at the scale of the hydrographic network, makes it possible to calculate the values of P and R at any point in the network and in all seasons.

josFig13

Figure: Schematic representation of P vs. R diagrams. For closed system, without nutrient input primary production and respiration are equilibrated, P vs R point being situated on the diagonal. When nutrients are added to the system, primary production increases and oxygen is super saturated, P vs. R point being above the diagonal, the system is autotrophic. In a system enriched in organic matter, mineralization increase and oxygen is consumed, P vs. R point being below the diagonal, the system is heterotrophic.

 

According to the river-continuum concept RCC [] a system not subjected to human intervention has an essentially heterotrophic behaviour in the lower-order streams (1 to 4) even without organic wastes inputs from urban or industrial origins. Their functioning is dominated by allocthonous fluxes of organic matter from their watershed basins and their primary production is weak. The system becomes progressively autotrophic from upstream to downstream.

The comparative analysis by the model of the P/R diagrams at a spatial scale in the 4 large rivers, shows that they, like the Seine, are very eutrophic judging by their high phytoplankton production (maxima of over 2.5 g C m2d-1). The case of the Scheldt is a special one, with heavy organic contamination that causes heterotrophic conditions in any order of the drainage network (Figure 14).

As in the Seine, the primary production in the Mosel and Loire increases from upstream to downstream, beginning in orders 4 to 5. In contrast to the Seine and the Mosel, sudden declines of the phytoplankton production, leading to heterotrophy, occur only in the estuary of the Loire. In the Mosel, the system becomes heterotrophic already at Hauconcourt, the upstream part of the canalised principal axis, and remains so until the confluence with the Rhine (Detzem). Regarding the Seine, the system becomes heterotrophic downstream of Paris, but recovers after 150 km, at Poses, where the estuary begins, only to return to heterotrophy inside the estuary (Figure 14).

josFig14

Figure: Geographical variations of Production vs. Respiration (mg C m-2 day-1) ), in the Mosel (1993, from upstream to downstream at order 5: Mi: Millery; Ha: Hauconcourt; Si: Sierck; De: Detzem), and from upstream -order 2 to 4-, to downstream -order 6 to 8- in the Scheldt (1987), in the Seine (1991) Loire (1989). Spring situations (June).

 

The analysis of the P/R diagrams at a seasonal scale in the Oise and the Marne tributaries of the Seine shows that the basin outlets are often heterotrophic, except in April and May, whereas the zones further upstream are always autotrophic (Figure 15). The most autotrophic sectors are those where the phytoplankton is always in the growth phase with still moderate allochthonous inputs (point sources). These sectors are then permanently oxygenated but the intense algal activity may increase the pH which poses problems for drinking water treatments []. The degradation of insufficiently treated discharges, which increase from upstream to downstream, then overtakes the algal production at the outlet of the sub-basins; phytoplankton biomass may become an autochtonous input of highly labile organic matter that leads to heterotrophic conditions (Figure 15). From the beginning of the transit through the Paris urban area, heterotrophic conditions dominate all year round (Figure 15). Downstream of Paris, after the impact of the treatment plant at Achères, these conditions become even more pronounced. A restoration phase that re-establishes an autotrophic state occurs at Poses, at the mouth of the estuary.

josFig15

Figure: Seasonal P/R variations (1991) in the Marne, Oise basin, and in the main axis of the low Seine river. Model outputs at a) Paris, Conflans (upstream the Oise confluence and downstream the waste water effluents of the largest Waste Water treatment Plant of the Seine Basin, 6.5 106 inhabitant equivalent), and Poses the limit of the freshwater estuary: b) two Marne stations (La Ferté, 2/3 of order 6 and Neuilly, near the Seine-Marne confluence, order 6) and two Oise stations (Compiègne: Oise-Aisne confluence, order 6; Méry: Oise, upstream Seine confluence, order 7).

 

P/R diagrams analysis led to an interpretation of the upstream-downstream functioning of these large impacted and regulated rivers as a succession of heterotrophic and autotrophic sequences. The diagrams confirm the RCC theory but show clearly how the systems have been changed by human intervention and allow to quantify the effect of these interventions. It is clear that the strong human pressure on the Seine contributes to an increase in the alterations between heterotrophic and autotrophic conditions in the system; when the pressure is extreme, as in the case of the Scheldt, the sequence is limited to heterotrophy.


6.6 Combating eutrophication in the Seine river and the Seine Bight :
scenarios analysis6 Drainage basin use and nutrient supply by rivers to the
coastal zone.6.4 Origin and processes of nutrient transformation in the river continuum6.5 Modelling phytoplankton and nutrient in drainage networks