504
J.L. Probst et al.
mercury contents in the sediment decrease by about 50%, going from 20 to 80100°C drying temperature. For mercury analyses, the sediments are chemically
digested with nitric and sulphuric acids in the presence of KMnO 4. Mercury is
analyzed by cold vapour flameless atomic absorption spectrometry (CVFAAS) on
a Perkin Elmer 430 with a preconcentration of gold powder if necessary (Krempp
1996). To avoid water vapour condensation on windows, the detection cell is
heated at a temperature higher than the dewpoint. The absolute detection limit is
5 ng of mercury in the initial analyzed solution.
Mercury contents were analyzed in the different phases (aqueous phase,
suspended matter and bottom sediments) and in the different granulometric
fractions of the sediments. The bottom sediments were riddled to separate the
different sand fractions (fine 50-200 ~lm and coarse >200 ~lm). The clay and silt
fractions «50 ~lm) were separated by syphoning the sediments in suspension
according to Stokes law [clays «2 ~lm), fine silts (2-20 ~lm) and coarse silts (2050 ~lm)l. This granulometric separation exercise was made only for the bottom
sediments collected in February 1996.
3
Spatial and Seasonal Variations in Mercury Content
in the III River Basin
Mercury concentrations measured at the different stations upstream to downstream are low in the solutions, as already shown for other rivers (Beim and
Grosheva 1996; Grosheva and Hayashi 1996; Ikingura and Mutakyahwa 1996).
These concentrations are sometimes even below the detection limit, even in
industrial wastewater. All aqueous concentrations of mercury remain under the
standard limit allowed for drinking water (1 ~lg r', FAO/OMS 1972; EC 1980).
Nevertheless, the concentrations present a generally decreasing trend upstream to downstream in the basin, which can be attributed to river water
dilution by the different tributary inputs and to sorption processes onto colloids
(Fig. 3a).
It is interesting to note that the mercury concentrations in the aqueous phase
are higher during the low-water period of July 1996 than during the high flows of
February 1996. This difference between low and high-water periods must be
related to the difference in solute mercury concentration in industrial wastewater
(169 ng 1-' in July and under the detection limit in February) and to the
methylation processes of bottom sediment mercury. Contrary to the solutions,
the suspended matter (Fig. 3b) and the bottom sediment (Fig. 3c) show high
concentrations of mercury, as already shown for other rivers (Dauvalter et al.
1996; Dauvalter and Rognerud 1996; Ikingura and Mutakyahwa 1996). This is due
to sorption processes onto sediments because of the high total dissolved solid
concentrations in the river water (see next section) and of the high affinity of
mercury for solid phase (Chapman et al. 1982; Dinh 1982; Behra 1987; Tiffreau
et al. 1995; Tavares et al. 1996). However, as with the solutions, mercury
J.L. Probst et al.
mercury contents in the sediment decrease by about 50%, going from 20 to 80100°C drying temperature. For mercury analyses, the sediments are chemically
digested with nitric and sulphuric acids in the presence of KMnO 4. Mercury is
analyzed by cold vapour flameless atomic absorption spectrometry (CVFAAS) on
a Perkin Elmer 430 with a preconcentration of gold powder if necessary (Krempp
1996). To avoid water vapour condensation on windows, the detection cell is
heated at a temperature higher than the dewpoint. The absolute detection limit is
5 ng of mercury in the initial analyzed solution.
Mercury contents were analyzed in the different phases (aqueous phase,
suspended matter and bottom sediments) and in the different granulometric
fractions of the sediments. The bottom sediments were riddled to separate the
different sand fractions (fine 50-200 ~lm and coarse >200 ~lm). The clay and silt
fractions «50 ~lm) were separated by syphoning the sediments in suspension
according to Stokes law [clays «2 ~lm), fine silts (2-20 ~lm) and coarse silts (2050 ~lm)l. This granulometric separation exercise was made only for the bottom
sediments collected in February 1996.
3
Spatial and Seasonal Variations in Mercury Content
in the III River Basin
Mercury concentrations measured at the different stations upstream to downstream are low in the solutions, as already shown for other rivers (Beim and
Grosheva 1996; Grosheva and Hayashi 1996; Ikingura and Mutakyahwa 1996).
These concentrations are sometimes even below the detection limit, even in
industrial wastewater. All aqueous concentrations of mercury remain under the
standard limit allowed for drinking water (1 ~lg r', FAO/OMS 1972; EC 1980).
Nevertheless, the concentrations present a generally decreasing trend upstream to downstream in the basin, which can be attributed to river water
dilution by the different tributary inputs and to sorption processes onto colloids
(Fig. 3a).
It is interesting to note that the mercury concentrations in the aqueous phase
are higher during the low-water period of July 1996 than during the high flows of
February 1996. This difference between low and high-water periods must be
related to the difference in solute mercury concentration in industrial wastewater
(169 ng 1-' in July and under the detection limit in February) and to the
methylation processes of bottom sediment mercury. Contrary to the solutions,
the suspended matter (Fig. 3b) and the bottom sediment (Fig. 3c) show high
concentrations of mercury, as already shown for other rivers (Dauvalter et al.
1996; Dauvalter and Rognerud 1996; Ikingura and Mutakyahwa 1996). This is due
to sorption processes onto sediments because of the high total dissolved solid
concentrations in the river water (see next section) and of the high affinity of
mercury for solid phase (Chapman et al. 1982; Dinh 1982; Behra 1987; Tiffreau
et al. 1995; Tavares et al. 1996). However, as with the solutions, mercury