Regarding silica, also diffuse but originating from the natural rock weathering
process, its concentrations oscillated around 3.8 mg Si L
À1 over the last 50 years.
Distinct periods of Si depletion are observed corresponding to uptake by diatoms
during blooms. In agreement with the observation of a strong reduction of blooms
after 2007, these silica depletions have been much less pronounced in recent years
than they were 20 years ago [24] (Fig. 5).
Suspended matter dynamics (not shown) did not change significantly over the
period, except for a decrease during low flow, from 20 mg L
À1 until the mid-2000s
to less than 10 mg L
À1 in recent years. This change might be due to the decrease in
phytoplankton biomass and the implementation of grass strips for stream protection
from erosion and pollution, rendered compulsory in 2005. River navigation is the
main contributor for maintaining particles in suspension during low flow through a
release of energy into the system [44]. The evolution of fluvial traffic since the 2000s
must therefore be taken into account to balance the effect of grass strip
implementation.
3.3 Modelling Nutrients and Phytoplankton Biomass
Long-term simulations with Riverstrahler match the observations (Fig. 6 for
1984–1990 [50] and 2007–2012 [29]). Once their formulations have been defined
and parameters determined based on experimental field and lab studies, no further
adjustment is carried out, and the comparison between simulations and observations
constitutes a validation test for the model.
Temporally refined simulations performed with the ProSe model across the Paris
urban area over the 2007–2012 period also validate the formalisms of the RIVE
model [27, 41, 42], even at the second order with a geostatistical analysis [31, 43,
44]. Discrepancies between observations and simulations raise new scientific questions, either on an incomplete representation of processes in the model or on the
accuracy of the forcing data (diffuse and point sources) or even on the quality of the
observed data [43].
4 River Metabolisms: Autotrophy, Heterotrophy and CO 2
Saturation
The metabolism of any ecosystem is well characterised by the intensity of their two
basic functions of autotrophy and heterotrophy, i.e. photosynthesis and respiration,
which directly interact with the carbon and oxygen cycles. In river systems,
according to the RCC [7], a distinct longitudinal pattern is predicted with ecosystem
metabolism shifting from dominant heterotrophy in small rivers receiving most of
their energy from allochthonous organic matter from the watershed to autotrophy in
Ecological Functioning of the Seine River: From Long-Term Modelling. . .
201
process, its concentrations oscillated around 3.8 mg Si L
À1 over the last 50 years.
Distinct periods of Si depletion are observed corresponding to uptake by diatoms
during blooms. In agreement with the observation of a strong reduction of blooms
after 2007, these silica depletions have been much less pronounced in recent years
than they were 20 years ago [24] (Fig. 5).
Suspended matter dynamics (not shown) did not change significantly over the
period, except for a decrease during low flow, from 20 mg L
À1 until the mid-2000s
to less than 10 mg L
À1 in recent years. This change might be due to the decrease in
phytoplankton biomass and the implementation of grass strips for stream protection
from erosion and pollution, rendered compulsory in 2005. River navigation is the
main contributor for maintaining particles in suspension during low flow through a
release of energy into the system [44]. The evolution of fluvial traffic since the 2000s
must therefore be taken into account to balance the effect of grass strip
implementation.
3.3 Modelling Nutrients and Phytoplankton Biomass
Long-term simulations with Riverstrahler match the observations (Fig. 6 for
1984–1990 [50] and 2007–2012 [29]). Once their formulations have been defined
and parameters determined based on experimental field and lab studies, no further
adjustment is carried out, and the comparison between simulations and observations
constitutes a validation test for the model.
Temporally refined simulations performed with the ProSe model across the Paris
urban area over the 2007–2012 period also validate the formalisms of the RIVE
model [27, 41, 42], even at the second order with a geostatistical analysis [31, 43,
44]. Discrepancies between observations and simulations raise new scientific questions, either on an incomplete representation of processes in the model or on the
accuracy of the forcing data (diffuse and point sources) or even on the quality of the
observed data [43].
4 River Metabolisms: Autotrophy, Heterotrophy and CO 2
Saturation
The metabolism of any ecosystem is well characterised by the intensity of their two
basic functions of autotrophy and heterotrophy, i.e. photosynthesis and respiration,
which directly interact with the carbon and oxygen cycles. In river systems,
according to the RCC [7], a distinct longitudinal pattern is predicted with ecosystem
metabolism shifting from dominant heterotrophy in small rivers receiving most of
their energy from allochthonous organic matter from the watershed to autotrophy in
Ecological Functioning of the Seine River: From Long-Term Modelling. . .
201