CHAPTER 22 . (hemometric Applications to Seawater Analysis
419
lived for 16 weeks before the analysis. Another consideration to be made is that the major
difference between the two sampling locations is that one is polluted by urban fall and
ship discharges and the other one is pristine. It is reasonable to believe that the difference in the sea conditions at the two locations is reflected by the difference in composition of the natural components in mussels. Further studies of similarities and differences among the components naturally present in samples from different unpolluted
locations are desirable, to make the method suitable for monitoring coastal areas.
Mussels were also studied for heavy metal pollution samples collected in 12 stations
in the eastern coast of the Bothnian Sea, adjacent to the town of Pori in Finland
(Lindstrom et al. 1988). In this area the two main pollution sources are the local titanium oxide industry and the River Kokemaenjoki, responsible for both industrial and
agricultural discharges. Aims of the study were: (i) to investigate how far metals reach
from the Vuorikemia factory effluent, (ii) to distinguish the waters coming from the
factory from those coming from the river and (iii) to study biological effects of pollution to mussels during a short (2-3 months) and a long (years) period. In both soft
parts and shells, a selective fractionating procedure for the determination of AI, As,
Cd, Co, Cr, Cu, Fe, Hg, Mn, Pb, Sb, Ti, V and Zn and the treatment of the chemical data
with SIMCA method gave the best correlation to the pollution sources with data from
the calcite shells. The soft tissue data of the caged mussels correlated well with the
pollution load contributed by the factory. Ongoing pollution is best measured by soft
tissue method while long term or past pollution is better characterized by data from
the shells.
A major recent advance in environmental science has been the use of molecular
markers for the identification of organic inputs in aquatic systems. Generally speaking, molecular markers are compounds whose structure can be related to specific
sources; for example long-chain alkylbenzenes and trialkylamines can be related to
contributions from anionic or cationic detergents, respectively. The usefulness of PCA
and FA in environmental studies when using molecular markers for sample description was evaluated in a study involving the determination of aliphatic and chlorinated
hydrocarbon, fatty acids, alcohols, chlorophylls and some detergents in water and in
particulate (Grimalt and Olive 1993). Seven sampling sites were selected in the delta
of the Ebre River, in the Mediterranean Sea. Samples of surface sea water and of about
3 m depth were collected in Fangar and Alfacs Bays and in the Ebre River bed, as well
as in the irrigation and in the drainage channels. The sampling was carried out every
2 months covering a period of 18 months. 40 samples were collected in total. The use
of GC and GC-MS techniques for hydrocarbon, alcohol and fatty acids fractions allowed the identification and determination of 122 compounds and two standard industrial mixtures of PCBs, while by UV spectrophotometry chlorophyll pigments and
phaeophorbide were determined. PCA analysis of the data shows that PC1 explains
34% of the variance and corresponds to a mixture of compounds originating from
various sources. This indicates that there is a concurrence of biogenic and anthropogenic inputs, which is not surprising because the river is the major geographical source
which determines the distribution in the delta. PC2 explains 14% of the variance and
shows a rather complex distribution. As concerns the score plot, PC1 scores afford a
good differentiation between the bay water and those of the river and channels, while
PC2 scores separate the irrigation channel waters from those of the river and of the
discharge channel. FA was performed in an iterative way, by replacing in each cycle
419
lived for 16 weeks before the analysis. Another consideration to be made is that the major
difference between the two sampling locations is that one is polluted by urban fall and
ship discharges and the other one is pristine. It is reasonable to believe that the difference in the sea conditions at the two locations is reflected by the difference in composition of the natural components in mussels. Further studies of similarities and differences among the components naturally present in samples from different unpolluted
locations are desirable, to make the method suitable for monitoring coastal areas.
Mussels were also studied for heavy metal pollution samples collected in 12 stations
in the eastern coast of the Bothnian Sea, adjacent to the town of Pori in Finland
(Lindstrom et al. 1988). In this area the two main pollution sources are the local titanium oxide industry and the River Kokemaenjoki, responsible for both industrial and
agricultural discharges. Aims of the study were: (i) to investigate how far metals reach
from the Vuorikemia factory effluent, (ii) to distinguish the waters coming from the
factory from those coming from the river and (iii) to study biological effects of pollution to mussels during a short (2-3 months) and a long (years) period. In both soft
parts and shells, a selective fractionating procedure for the determination of AI, As,
Cd, Co, Cr, Cu, Fe, Hg, Mn, Pb, Sb, Ti, V and Zn and the treatment of the chemical data
with SIMCA method gave the best correlation to the pollution sources with data from
the calcite shells. The soft tissue data of the caged mussels correlated well with the
pollution load contributed by the factory. Ongoing pollution is best measured by soft
tissue method while long term or past pollution is better characterized by data from
the shells.
A major recent advance in environmental science has been the use of molecular
markers for the identification of organic inputs in aquatic systems. Generally speaking, molecular markers are compounds whose structure can be related to specific
sources; for example long-chain alkylbenzenes and trialkylamines can be related to
contributions from anionic or cationic detergents, respectively. The usefulness of PCA
and FA in environmental studies when using molecular markers for sample description was evaluated in a study involving the determination of aliphatic and chlorinated
hydrocarbon, fatty acids, alcohols, chlorophylls and some detergents in water and in
particulate (Grimalt and Olive 1993). Seven sampling sites were selected in the delta
of the Ebre River, in the Mediterranean Sea. Samples of surface sea water and of about
3 m depth were collected in Fangar and Alfacs Bays and in the Ebre River bed, as well
as in the irrigation and in the drainage channels. The sampling was carried out every
2 months covering a period of 18 months. 40 samples were collected in total. The use
of GC and GC-MS techniques for hydrocarbon, alcohol and fatty acids fractions allowed the identification and determination of 122 compounds and two standard industrial mixtures of PCBs, while by UV spectrophotometry chlorophyll pigments and
phaeophorbide were determined. PCA analysis of the data shows that PC1 explains
34% of the variance and corresponds to a mixture of compounds originating from
various sources. This indicates that there is a concurrence of biogenic and anthropogenic inputs, which is not surprising because the river is the major geographical source
which determines the distribution in the delta. PC2 explains 14% of the variance and
shows a rather complex distribution. As concerns the score plot, PC1 scores afford a
good differentiation between the bay water and those of the river and channels, while
PC2 scores separate the irrigation channel waters from those of the river and of the
discharge channel. FA was performed in an iterative way, by replacing in each cycle