CHAPTER 22 . (hemometric Applications to Seawater Analysis
409
lar composition for the particulate matter and the sediments, while the composition of
heavy metal in soft tissues reflects that of seawater body (Adami et al. 1998).
The composition of surface water (within 1 m depth) can result different for sampling sites even very close. This can happen in particular for lagoon water, as observed
for two sites in Canal Grande in Venice (Gennaro et al. 1993), due to the lower circulation of waters, to the coastal geographical morphology and to the structure of the
hydraulic net of channels differently affecting the cumulation processes and the circulation of clean water from the open sea. Also the effect of tide, of winds and of different sources of anthropogenic pollution playa role. Lagoon waters generally present
larger problems of pollution, with respect to the open sea, and their composition can
give useful information (Gennaro et al.1994; Gennaro et aL1995). At this purpose and
to study also the seasonal dependence, monthly samples were collected in 1993 in three
different stations characterized by different level of pollution (Marengo et aL1995).
13 variables (mainly inorganic anions) were determined. Unfortunately in 1993 the
weather behaved atypically as concerns both rains and temperature and the raw data
do not allow any direct correlation with the season, resulting predominant the differences
due to the geographical positions. Only a PCA treatment permitted to characterize the
samples as a function of seasonality and allowed to select the significant discriminating
factors. The trend which is a function of the season can be envisaged in the plot of PC2 vs.
PC3 and the discriminating variables resulted to be water pH, temperature, dissolved oxygen and phosphate and nitrate content. As a matter of fact, these species are the more affected by phytoplankton activity, that in turn depends on seasonality.
22.2.1.2
Seawater Vertical Distribution
As already mentioned, seawater composition is also a function of water depth. Studies
of the vertical distribution can increase the information concerning the geochemical
circumrotation of the elements. A study was devoted to find correlation between composition and depth in the open-sea region in the south-western part of the Black Sea,
at about 200 miles from the town of Va ran a (Simeonov et al. 1993). The chemical components and the physical parameters collected for waters at different depths were studied by PCA, FA and hierarchical cluster analysis. The study was performed for 9 sampling depths and 22 variables, including NH3, P, Zn, Mn, Cu, Fe, 02' pH, salinity, alkalinity, Eh (m V), temperature and total suspended matter. The total variance could be
explained by 5 PCs, being the 78.2% explained by the first two PCs. The plot of the loadings (PC1) reported in Fig. 22.1 permits the identification of two groups of data: the
first includes the data for the samples collected between 0 and 100 m and the second
one collects the data for the waters sampled between 125 and 200 m. The depth of 100 m
seems to correspond to a variation in the composition that can be explained by two
major effects:
i. the action of the wind able to mix the water mass of Black Sea up to this depth, and
ii. the effects of vertical density gradients.
These effects in turn cause accumulation of suspended matter at depths of about
100 m due to the retaining of suspensions by the more dense water layers. Additional
409
lar composition for the particulate matter and the sediments, while the composition of
heavy metal in soft tissues reflects that of seawater body (Adami et al. 1998).
The composition of surface water (within 1 m depth) can result different for sampling sites even very close. This can happen in particular for lagoon water, as observed
for two sites in Canal Grande in Venice (Gennaro et al. 1993), due to the lower circulation of waters, to the coastal geographical morphology and to the structure of the
hydraulic net of channels differently affecting the cumulation processes and the circulation of clean water from the open sea. Also the effect of tide, of winds and of different sources of anthropogenic pollution playa role. Lagoon waters generally present
larger problems of pollution, with respect to the open sea, and their composition can
give useful information (Gennaro et al.1994; Gennaro et aL1995). At this purpose and
to study also the seasonal dependence, monthly samples were collected in 1993 in three
different stations characterized by different level of pollution (Marengo et aL1995).
13 variables (mainly inorganic anions) were determined. Unfortunately in 1993 the
weather behaved atypically as concerns both rains and temperature and the raw data
do not allow any direct correlation with the season, resulting predominant the differences
due to the geographical positions. Only a PCA treatment permitted to characterize the
samples as a function of seasonality and allowed to select the significant discriminating
factors. The trend which is a function of the season can be envisaged in the plot of PC2 vs.
PC3 and the discriminating variables resulted to be water pH, temperature, dissolved oxygen and phosphate and nitrate content. As a matter of fact, these species are the more affected by phytoplankton activity, that in turn depends on seasonality.
22.2.1.2
Seawater Vertical Distribution
As already mentioned, seawater composition is also a function of water depth. Studies
of the vertical distribution can increase the information concerning the geochemical
circumrotation of the elements. A study was devoted to find correlation between composition and depth in the open-sea region in the south-western part of the Black Sea,
at about 200 miles from the town of Va ran a (Simeonov et al. 1993). The chemical components and the physical parameters collected for waters at different depths were studied by PCA, FA and hierarchical cluster analysis. The study was performed for 9 sampling depths and 22 variables, including NH3, P, Zn, Mn, Cu, Fe, 02' pH, salinity, alkalinity, Eh (m V), temperature and total suspended matter. The total variance could be
explained by 5 PCs, being the 78.2% explained by the first two PCs. The plot of the loadings (PC1) reported in Fig. 22.1 permits the identification of two groups of data: the
first includes the data for the samples collected between 0 and 100 m and the second
one collects the data for the waters sampled between 125 and 200 m. The depth of 100 m
seems to correspond to a variation in the composition that can be explained by two
major effects:
i. the action of the wind able to mix the water mass of Black Sea up to this depth, and
ii. the effects of vertical density gradients.
These effects in turn cause accumulation of suspended matter at depths of about
100 m due to the retaining of suspensions by the more dense water layers. Additional