The quality of chemical analysis and the accuracy of the data were
checked with a blank sample and sediment reference materials STSD-3
and MESS-4 (Canadian Certified Reference Materials Project) that were
included in each batch of 13 samples during the course of analysis. The
results indicate a good agreement between the certified and analytical
values for all metals analyzed in both two certified materials, except for
Fe and As. The recoveries of Fe and As were around 87% and 84%,
respectively, in the certified material STSD-3, but there were near 95%
for MESS-4 (Supplementary material).
It is important to note that the metal analyzes have been conducted
on the total fraction and not on a wet sieved fraction. The sieving allows
to reduce the variance arising from grain-size differences, but we are
interested by the ecological risk, and thus by the entire sediment and
the “real” metal concentrations.
4. Evaluation of the sediment contamination
Numerous methods have been put forward for quantifying the degree of metal enrichment in sediments associated to anthropogenic
inputs (Ridgway and Shimmield, 2002). Various authors (Salomons and
Forstner, 1984; Muller, 1969; Håkanson, 1980) have thus proposed to
evaluate the impact scales (or ranges) of the pollution by converting the
analytical results into broad descriptive bands of pollution ranging from
low to high intensity. In order to get a global overview of the ecological
risk, three of these methods, presented hereunder, will be discussed in
the following sections.
4.1. The chemical contamination index method
A crucial first step in evaluating the impact of sediment pollution
and the level of contamination of a given area is to select a reference
background or baseline sample of known metal composition. Two kind
of reference backgrounds are generally proposed: average crustal or
shale values (Turekian and Wedepohl, 1961; Taylor, 1964; Taylor and
McLennan, 1985; Wedepohl, 1995) or a comparable local sediment
unaffected by anthropogenic activity. The choice made for this paper is
discussed on Section 6.1.
4.1.1. Contamination factor (CF)
The Contamination Factor (CF) is the ratio obtained by dividing the
concentration of each metal in the sediment by its background values
(Håkanson, 1980):
=
CF
C
C Background
Site
(1)
Håkanson (1980) defined four classes of CF: low (CF < 1), moderate (1 < CF < 3); considerable ((3 < CF < 6) and very high contamination factor (CF > 6).
4.1.2. Pollution load index (PLI)
The pollution load index (PLI) gives an assessment of the overall
toxicity status of the sediment as a result of the contribution of several
metals (Tomlinson et al., 1980). It is defined as the n
th root of the
multiplications of the contaminations factors (CF metals):
=
×
×
× ………×
PLI (CF
CF
CF
CF )
1
2
3
n
1/n
(2)
where CF is the contamination factor of each metal (Eq. (1)). Values of
PLI above 1 imply that metal pollution exists, whereas there is no
pollution below one.
4.1.3. Geo-accumulation index (Igeo)
Muller (1969) proposed to estimate the enrichment of metals concentrations above background by calculating a geoaccumulation index:
Igeo. He calculated this index according to:
=
Igeo log (C /1.5. B )
2
n
n
(3)
where C n is the heavy metal concentration and B n is the geochemical
background value. Here again the choice of this background value has a
considerable influence on this index, but the factor of 1.5 is used to
minimize a possible variation in the background value due to lithogenic
effects. Muller (1969) then proposed a seven-level classification of Igeo
ranging from below zero to above 5, corresponding to a unpolluted
(Igeo < 0), moderately (1–2), strongly (3–4) and extremely polluted
(> 5).
4.1.4. Enrichment factor (EF)
A very common approach to estimate the anthropogenic impact on
sediments is to calculate a normalized enrichment factor (EF) for metals
above uncontaminated background levels (Salomons and Forstner,
1984; Dickinson et al., 1996; Hornung et al., 1989). Many researchers
have applied this factor in contamination assessments (Feng et al.,
2004; Reddy et al., 2004; Çevik et al., 2009; Bastami et al., 2012;
Hamdoun et al., 2015). The EF method presents the advantage to
normalize the heavy metals concentrations by two means. Firstly, by
the use of a reference metal which is supposed to be non affected by
anthropogenic inputs or chemical reaction during diagenesis (and if
possible which also allows to normalize relatively to the grain size);
secondly with the same ratio estimated from a background reference
sample.
The enrichment factors are calculated according Sutherland (2000):
=
EF (X/Y)
/(X/Y)
.
sediment
Background reference value
(4)
where X = concentration of the metal of interest and Y = Concentration of the reference metal.
Al or Fe are often used as reference metals (Windom et al., 1989,
Din, 1992; Ravichandran et al., 1995, Chen et al., 2007 and Huerta-Diaz
et al., 2008), but other metals have been proposed like Mn, Co or Cs
(Stewart, 1989; Matthai and Birch, 2001; Roussiez et al., 2012). Al was
not analyzed in our case and thus we used Fe as the reference tracer to
differentiate natural from anthropogenic components. Indeed, Keresten
and Smedes (2001) explained that Fe can be used as a proxy for Al and
clay fraction unless it is affected by post-depositional diagenetic processes. It should not be the case here since the samples were collected in
surface of the sediment. For the background reference values, we use a
specific local background described in section 6.1. Finally, we refer to
the five category of pollution index proposed by Andrews and
Sutherland (2004) to define the pollution assessment: EF <
2 = minimal
pollution;
2 < EF < 5 = moderate
pollution;
5 < EF < 20 = significant pollution; 20 < EF < 40 = high pollution; and EF > 40 = extreme pollution.
4.2. The ecological risk assessment method
Ecological risk assessment is a process that evaluates the likelihood
that adverse ecological effects may occur or are occurring as a result of
exposure to one or more stressors (U.S. EPA, 1992). It is suited to
transform scientific data into meaningful information about the risk of
human activities to the environment (U.S. EPA, 1998).
4.2.1. Potential ecological risk index (RI)
The potential ecological risk index (RI) proposed by Håkanson
(1980) is an approach supposed to represent the sensitivity of biological
communities to the overall toxic substances present in the sediments. RI
is calculated using:
∑
∑
=
=
RI
E
T · C
i
r
i
r
i
f
(5)
where: C f
i is the value of the concentration of a metal i divided by its
background value, and T r
i is the “toxic response factor” for the metal i,
which reflects its toxicity levels and the sensitivity of bio-organism to it.
The toxic response factors were respectively, 5, 30, 2, 5, 5, 10 and 1
for Pb, Cd, Cr, Cu, Ni, As and Zn (Håkanson, 1980) and 5, 2 and 1,
I. Ahmed et al.
Marine Pollution Bulletin 136 (2018) 322–333
325
checked with a blank sample and sediment reference materials STSD-3
and MESS-4 (Canadian Certified Reference Materials Project) that were
included in each batch of 13 samples during the course of analysis. The
results indicate a good agreement between the certified and analytical
values for all metals analyzed in both two certified materials, except for
Fe and As. The recoveries of Fe and As were around 87% and 84%,
respectively, in the certified material STSD-3, but there were near 95%
for MESS-4 (Supplementary material).
It is important to note that the metal analyzes have been conducted
on the total fraction and not on a wet sieved fraction. The sieving allows
to reduce the variance arising from grain-size differences, but we are
interested by the ecological risk, and thus by the entire sediment and
the “real” metal concentrations.
4. Evaluation of the sediment contamination
Numerous methods have been put forward for quantifying the degree of metal enrichment in sediments associated to anthropogenic
inputs (Ridgway and Shimmield, 2002). Various authors (Salomons and
Forstner, 1984; Muller, 1969; Håkanson, 1980) have thus proposed to
evaluate the impact scales (or ranges) of the pollution by converting the
analytical results into broad descriptive bands of pollution ranging from
low to high intensity. In order to get a global overview of the ecological
risk, three of these methods, presented hereunder, will be discussed in
the following sections.
4.1. The chemical contamination index method
A crucial first step in evaluating the impact of sediment pollution
and the level of contamination of a given area is to select a reference
background or baseline sample of known metal composition. Two kind
of reference backgrounds are generally proposed: average crustal or
shale values (Turekian and Wedepohl, 1961; Taylor, 1964; Taylor and
McLennan, 1985; Wedepohl, 1995) or a comparable local sediment
unaffected by anthropogenic activity. The choice made for this paper is
discussed on Section 6.1.
4.1.1. Contamination factor (CF)
The Contamination Factor (CF) is the ratio obtained by dividing the
concentration of each metal in the sediment by its background values
(Håkanson, 1980):
=
CF
C
C Background
Site
(1)
Håkanson (1980) defined four classes of CF: low (CF < 1), moderate (1 < CF < 3); considerable ((3 < CF < 6) and very high contamination factor (CF > 6).
4.1.2. Pollution load index (PLI)
The pollution load index (PLI) gives an assessment of the overall
toxicity status of the sediment as a result of the contribution of several
metals (Tomlinson et al., 1980). It is defined as the n
th root of the
multiplications of the contaminations factors (CF metals):
=
×
×
× ………×
PLI (CF
CF
CF
CF )
1
2
3
n
1/n
(2)
where CF is the contamination factor of each metal (Eq. (1)). Values of
PLI above 1 imply that metal pollution exists, whereas there is no
pollution below one.
4.1.3. Geo-accumulation index (Igeo)
Muller (1969) proposed to estimate the enrichment of metals concentrations above background by calculating a geoaccumulation index:
Igeo. He calculated this index according to:
=
Igeo log (C /1.5. B )
2
n
n
(3)
where C n is the heavy metal concentration and B n is the geochemical
background value. Here again the choice of this background value has a
considerable influence on this index, but the factor of 1.5 is used to
minimize a possible variation in the background value due to lithogenic
effects. Muller (1969) then proposed a seven-level classification of Igeo
ranging from below zero to above 5, corresponding to a unpolluted
(Igeo < 0), moderately (1–2), strongly (3–4) and extremely polluted
(> 5).
4.1.4. Enrichment factor (EF)
A very common approach to estimate the anthropogenic impact on
sediments is to calculate a normalized enrichment factor (EF) for metals
above uncontaminated background levels (Salomons and Forstner,
1984; Dickinson et al., 1996; Hornung et al., 1989). Many researchers
have applied this factor in contamination assessments (Feng et al.,
2004; Reddy et al., 2004; Çevik et al., 2009; Bastami et al., 2012;
Hamdoun et al., 2015). The EF method presents the advantage to
normalize the heavy metals concentrations by two means. Firstly, by
the use of a reference metal which is supposed to be non affected by
anthropogenic inputs or chemical reaction during diagenesis (and if
possible which also allows to normalize relatively to the grain size);
secondly with the same ratio estimated from a background reference
sample.
The enrichment factors are calculated according Sutherland (2000):
=
EF (X/Y)
/(X/Y)
.
sediment
Background reference value
(4)
where X = concentration of the metal of interest and Y = Concentration of the reference metal.
Al or Fe are often used as reference metals (Windom et al., 1989,
Din, 1992; Ravichandran et al., 1995, Chen et al., 2007 and Huerta-Diaz
et al., 2008), but other metals have been proposed like Mn, Co or Cs
(Stewart, 1989; Matthai and Birch, 2001; Roussiez et al., 2012). Al was
not analyzed in our case and thus we used Fe as the reference tracer to
differentiate natural from anthropogenic components. Indeed, Keresten
and Smedes (2001) explained that Fe can be used as a proxy for Al and
clay fraction unless it is affected by post-depositional diagenetic processes. It should not be the case here since the samples were collected in
surface of the sediment. For the background reference values, we use a
specific local background described in section 6.1. Finally, we refer to
the five category of pollution index proposed by Andrews and
Sutherland (2004) to define the pollution assessment: EF <
2 = minimal
pollution;
2 < EF < 5 = moderate
pollution;
5 < EF < 20 = significant pollution; 20 < EF < 40 = high pollution; and EF > 40 = extreme pollution.
4.2. The ecological risk assessment method
Ecological risk assessment is a process that evaluates the likelihood
that adverse ecological effects may occur or are occurring as a result of
exposure to one or more stressors (U.S. EPA, 1992). It is suited to
transform scientific data into meaningful information about the risk of
human activities to the environment (U.S. EPA, 1998).
4.2.1. Potential ecological risk index (RI)
The potential ecological risk index (RI) proposed by Håkanson
(1980) is an approach supposed to represent the sensitivity of biological
communities to the overall toxic substances present in the sediments. RI
is calculated using:
∑
∑
=
=
RI
E
T · C
i
r
i
r
i
f
(5)
where: C f
i is the value of the concentration of a metal i divided by its
background value, and T r
i is the “toxic response factor” for the metal i,
which reflects its toxicity levels and the sensitivity of bio-organism to it.
The toxic response factors were respectively, 5, 30, 2, 5, 5, 10 and 1
for Pb, Cd, Cr, Cu, Ni, As and Zn (Håkanson, 1980) and 5, 2 and 1,
I. Ahmed et al.
Marine Pollution Bulletin 136 (2018) 322–333
325