Human Directive (MSDF) perspective. Moreover, as reported by numerous

Human impacts can change marine ecosystem both
directly and indirectly causing, in the first case, overexploitation and loss
of habitat, while, in the latter ones, changes in interactions in the food web
and in the structure of environment (Goulletquer et al. 2014). Coastal areas
are particularly subject to human pressure, since in this zone are located most
of the anthropogenic activities, as coastal infrastructures (i.e. ports, defences against erosion)
and offshore installations (i.e. oil
and gas platforms, wind farms). Human pressure can modify the natural status of
of physical, chemical, and biological components that characterize marine
ecosystem (Reiss et al. 2014). Regular monitoring of sediment quality is an
important pursuant to assessing the possible influence of anthropogenic
pressure on ecosystem quality (Romeo et al. 2015) and give information to
support management in order to reach a Good Environmental Status (GES) in the Marine
Strategy Framework Directive (MSDF) perspective.

Moreover, as reported by
numerous studies (Cozar et al. 2014; Nuelle et al. 2014), the analysis of
marine sediments is important in the evaluation of the emerging pollutants
such as microplastics, that tend to accumulate in the sea-bottom. At present, microplastics
represent a major global concern affecting all world oceans, defined in the
MSFD as D10 into different categories including the plastic debris. This
material, in the environment is subject to a combination of physical,
biological and chemical processes that reduce its structural integrity (Cole et
al. 2011) producing high densities of smaller debris as microplastics (1-5 mm).
As reported in the Guidance on Monitoring of Marin Litter in European Saes
(2013) the monitoring of litter in seafloor cannot consider all coastal areas
because of limited resources, for this reason also opportunistic approach (i.e. data from other research activity
in the harbour) could be used to improve the existing monitoring plans. Recent
studies testify that microplastic can became a threat of biodiversity, becoming
a vector for the introduction of non-native marine species to new habitats on
floating (Barnes 2002; Derraik 2002; Winston 1982). In addition, because of
their size, microplastics are considered bioavailable to organisms throughout
the food web (Thompson et al. 2004). When ingested, plastics release chemicals compound
(nonylphenols, polybrominated diphenyl ethers, phthalates and bisphenol A)
together with adsorbed hydrophobic pollutants (i.e. PCBs, TBT, DBT, MBT). Even if, the use of these pollutants was
bounded, their extensive use in the past and their low water solubility make
them persistent and able to accumulate both in sediments and in biota (Harris
and Wiberg 2002) and are measured at significant
levels in marine ecosystems and marine food webs (D’Alessandro et al. 2016).

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now

The sea-floor integrity (D6, MSFD) reflects
characteristics of the sea bottom. These characteristics delineate the
structure and functioning of marine ecosystems, especially for species and
communities living on the sea floor (benthic ecosystems). Human activities may affect
this structure by damaging of sensible species and supporting opportunistic,
non-indigenous or scavenging species that have profit from disturbance of the
bottom and availability of dead organisms. Macroinvertebrates, due to their skill
to modify their community patterns in relation to natural and anthropogenic stress(Warwick
1988) , are considered great biondicators of marine ecosystem (Warwick 1993;
Romeo et al. 2015). Actually, a lot of benthic indices based on structure of macrofaunal
communities were created to assess the ecological quality status (EcoQ) to
support data for MSFD,, , e.g. AMBI (Borja
et al. 2000), M-AMBI (Muxika et al. 2007), BENTIX (Simboura and Zenetos 2002), BOPA
(Dauvin and Ruellet 2007). These indices, based on the subdivision of species
in different ecological groups, return a value of environmental
quality/disturbance status.

The aim of this paper
is to carry out a quality assessment of a high polluted harbour of the Ionian sub-region
and Central Mediterranean Sea through a multidisciplinary approach that
integrate biotic and abiotic parameters, in order to give data to improve the
monitoring plan of the MSFD regarding shallow waters.


Material and methods

Study area

The Augusta site is
located in the MSFD Ionian sub-region of central Mediterranean Sea, in a
harbour area with a high marine traffic activity. This area hosted a variety of
different chemical and petrochemical refining plants, a commercial harbour and
a basis of the Italian Navy and NATO activities (Sprovieri et al. 2007). The harbour
is closed to the South and East by artificial dams. Two main inlets connect
harbour with open sea: the south-east and the east inlet. The basin is
characterised by three different circulation systems: eastern inlet, dominated
by a tidal current with a northward flowing, south eastern inlet, characterized
by flowing parallel to the coast; the northern portion of the basin, instead,
is characterized by shallow seabed and scarcely affected by active currents (Sprovieri et al. 2007;
Romano et al. 2013). Three small rivers
flow in the area, Mulinello in the North and Marcellino and Cantera in the central
part of the bay (Fig. 1). Due to the dangerous contamination of air, seawater,
and marine biota documented in this area, Augusta coastal area has been
included by the Italian Government in the national remediation plan (G.U.R.I.,
L. 426/1998) and evaluated by the World Health Organization as providing a high
environmental risk.


Sampling activities and laboratory analyses

Samples were carried
out from hard and soft bottoms during the summer of 2013 (Table1, Figure1). Soft
bottom samples were collected by means of a Van Veen grab (0.1 m2)
along four transects perpendicular to coastline at three different depths (5,
10 and 20 m). For each sampling site four replicates were carried out, three of
which were used for biological analysis, and one for the environmental
characterization following the methods described in D’Alessandro et al. (2016)
and Romeo et al. (2015). Sediment characterization was carried out according to
Buchanan and Kain (1971) and the percentage of pebble, gravel, sand, silt and
clay was determined according to the ternary Wentworth scale (Wentworth, 1922).

Plastics debris were
classified according to Guidance on Monitoring of Marine Litter in European
Seas (JRC EU, 2013) adapted. Four size classes were identified: microplastics
(1-5 mm), macroplastics (5-10 mm), megaplastics (10-20 mm) and plastics (>
20 mm) (Barnes 2002; Claessens et al. 2011). Microplastics were extracted
according to Alomar et al. 2016 with some modifications. For each sample, 1 Kg
of sediment was dried at 50° C for 48 h and then sieved for 15 min by means stainless
steel sieves with mesh diameter of: 20; 10; 5; 0.5; 0.1 mm. For each fraction,
plastics were extracted by density separation method and then, sediments were
observed under Stereomicroscope (Zeiss Discovery.V8) with optical enhancement
with a maximum magnification of 80x. Accurate precautions have been used to
prevent contamination during all phases of study according to Woodall et al.
2015. Hard bottom samples were carried out by SCUBA diving, scraping a surface
of 400 cm2 from two pillars within refinery. Three samples were
collected at three different depths (0.5; 6.0 and 15.0 m), for a total amount
of 9 samples per pillar.

Chemical analysis
were conducted taking into account contaminants included in the MSFD monitoring
plane and other contaminants of interest taking into account, according
literature, different typologies of anthropogenic impact of the study area (Falandysz
et al. 2006; Fang et al. 2003; Reli? et al. 2016) . Trace elements (Cd, Hg, Pb,
As, Cr) and other elements (Cu, Ni, and Zn) concentrations were determined
through Inductively Coupled Plasma-Mass Spectrometry (ICP-MS; mod. Agilent
Technologies), according to the US-EPA 6020A method. Quantifications of PCBs
(congeners 28, 52, 77, 81, 101, 118, 126, 128, 138, 153, 156, 169, 180) were
conducted following US-EPA 8082A/2007 standard method. PAHs (acenaphthene,
acenaphthylene, anthracene, dibenzo(a,h)anthracene, benzo(b)-fluoranthene,
benzo(a)pyrene, benzo(ghi)perylene, benzo(k)fluoranthene, chrysene,
dibenzo(a,h)anthracene, fluoranthene, indeno(1,2,3,)pyrene, naphthalene,
phenanthrene, pyrene, perylene, acenaphthene) were determined according to
US-EPA Method 8270D. The chemical analyses of butyltins (BT) in the surface
sediments were conducted according to a modified method from Binato et al.
(1998) and Morabito et al. (1995). Tributyltin (TBT), dibutyltin (DBT),
monobutyltin (MBT) and total butyltins (?BT) were determined as described in Romeo
et al. (2015) and D’Alessandro et al. (2016). All results of chemical analysis
were calculated on dry weight (d.w.), trace elements were expressed in µg/Kg,
persistent organic pollutants in mg/Kg, while the butyltins concentrations were
expressed as ng Sn g-1.

In order to characterise the
benthic communities, the main biodiversity indices were calculated: number of
species (S), Shannon’s index (H’), and Pielou’s evenness (J) (Magurran 1991).
Benthic indices (i.e. AMBI, M-AMBI,
Bentix and BOPA) were calculated to evaluate the environmental status on soft
bottom fauna on abundance of each species. AMBI and M-AMBI were calculated by means of AMBI index software
(version 4.0, available at; BOPA index and its relative
environmental quality were calculated using the revisited formula proposed by
Dauvin and Ruellet (2007).


Statistical analysis

Data exploratory test
for univariate analysis was assessed evaluating collinearity between variables
in order to avoid type II error rates (Zuur et al. 2010). Thereafter, in order
to highlight the potential correlation between the biotic indices,
environmental variables, trace elements and organic contaminants, a Pearson
correlation matrix was produced (Feld et al. 2016). Subsequently, the
Generalized Variance Inflation Factors (GVIF) was calculated, retaining those
variables with negligible collinearity (a GVIF of 2 or less; Zuur et al. 2010).

In order to detect
the relationships between potentially explanatory environmental variables and
the diversity and biotic indices, linear regression models were applied. Prior
to analysis, the normality of the distribution of each diversity and biotic
indices was tested before applying parametric statistical tests (Shapiro and Wilk,
1965). Afterwards, Akaike’s Information Criterion (AIC; Akaike 1974) was used
to select the best model among all classes of competing models after applying
backward and forward stepwise variable elimination (Czado et al. 2007; Famoye
and Rothe, 2001).

Finally, the
statistical assumptions of independence, normality and homogeneity of variances
were tested using the gvlma function in the gvlma package in R (Pena and Slate,
2014). All statistical univariate analyses were performed using the software program
R, version 3.4.2, R packages: car, gvlma, PerformanceAnalytics (R Core Team,

Multivariate analyses
were performed using the PRIMER6 and PERMANOVA+ software packages (Anderson et
al. 2008; Clarke and Warwick 2001). A non-parametric multivariate
analysis of variance (PERMANOVA) was performed to evaluate potential variations
in the abundance of the four different plastic sizes in relation to the factor depth.
Data were transformed to square root and analysed
on the basis of Bray-Curtis similarity (4,999 permutations). When significant
differences (p < 0.05) among factor levels were detected, pairwise comparisons were computed. The potential differences among the composition of the macrobenthic community, in depth x transect, was assessed through a two-way crossed PERMANOVA analysis. Prior to analysis, abundance data were square root transformed and analysed on the basis of Gower distance. When significant differences (p<0.05) among factor levels were detected, pairwise comparisons were computed. SIMPER analysis was performed to assess the contribution of the different taxa to the average dissimilarity between groups. Spatial pattern of macrobenthic community composition was assessed using a Bray-Curtis similarity index (Bray and Curtis 1957) on the square root transformed abundance matrix. Then, for the normalized abiotic data, a similarity profile based on permutation was tested using the SIMPROF routine to group stations with similar (i.e. branch with p>0.05)
contaminant data. Groupings generated using the SIMPROF procedure were
validated with a dissimilarity matrix based on Euclidean distance that was used
in not agglomerative hierarchical clustering (routine CLUSTER). A Principal
Coordinates (PCO) analysis was also performed to describe the contaminant data
that most accounted for variation among the group identified by the SIMPROF



As regard the soft bottom sediment analysis, Augusta
harbour resulted mainly characterized by fine fraction with the highest values
in northern side (TR4_5, 95.17% of silt), while the highest percentages of sand
were recorded along TR1 (TR1_5 and TR1_10 with 93.74% and 41.07 % of sand,
respectively). In TR2_5 the major percentage of clay (12.21%) was recorded.
Pebble and gravel represent less abundant fractions showing percentages of
4.59% and 4.55% in TR1_5 (Table 1).

Regarding the plastic
analysis, sediments of Augusta harbour revealed the presence of a total of 38
fragments kg-1 dry sediment of all the dimensional categories.
Microplastics was the most abundant group (n = 20 particles kg-1 dry
sediment) followed by plastics (n = 8 particles kg-1 dry sediment),
macro and megaplastics, both with 5 particles kg-1 dry sediment.
Microplastics were found in 9 up to 13 stations, megaplastics and plastics in 4
stations while macroplastics in 3. Considering all the dimensional ranges, the
highest number of plastic debris was found in TR4_5 (12 particles kg-1
dry sediment), while in TR1_5 and TR2_5 sediments, presence of debris was not
detected. Microplastics showed highest abundance (n = 6 particles kg-1
dry sediment) in TR1_10, macroplastics and megaplastics in TR4_5 (both with 3
particles kg-1 dry sediment), megaplastics in TR1_20 (2 particles kg-1
dry sediment) and plastics in TR4_5 (4 particles kg-1 dry sediment)
(Table 1).

Regarding the hard bottom communities, a total of 4191
specimens were found in Augusta harbour. Mollusca represented the most abundant
group (76.54% of total abundance). Three non-indigenous species were recorded:
the bivalve Brachidontes pharaonis (57.27%
of total abundance) and the polychaetes Pseudonereis
anomala (0.84% of total abundance) and Branchiomma
bairdi (0.21% of total abundance). B.
pharaonis dominated the hard bottom community, followed by the native crustacean
Elasmopus rapax (6.54 % of total
abundance). The highest value of species richness was recorded in GAU2_0 (S =
24), the number of individuals showed the highest value in GAU1_0 (N = 1232), J
was highest in GAU2_6 (J = 0.92), while H’ was highest in GAU1_3 and GAU2_3 (H’
= 2.26) (Table 2).

The faunistic soft
bottom analysis highlighted the presence of a total of 2125 specimens of which
64.33% belonging to Polychaeta, 29.22% Mollusca, 1.88% Crustacea and 4.56% to
other minor groups (Nematoda, Nemertea, Echinodermata, Sipuncula and Chordata).
The most abundant species was the polychaete Aricidea (Aricidea) pseudoarticulata (Hobson, 1972), representing 24.99% of total benthic assemblage.
Among Polychaeta, three non-indigenous species Kirkegaardia dorsobranchialis (Kirkegaard, 1959), Notomastus aberans (Day, 1957) and Pista unibranchia (Day, 1963) were also
found. Among Crustacea, the Decapoda Alpheus
glaber (0,33 % of total abundance) resulted the most represented species
while the most abundant Mollusca was the bivalve Corbula gibba (11.58 % of total abundance) (Annex 1). The average
values of the diversity indices for each sampling station are reported in Table
2. The species richness showed the highest value (S = 18.7) in TR3_5 with a
negative peak in TR1_10 (S = 7.3); in the latter station the lowest value of H’
(1.6) was recorded. The highest values of H’ (2.4) was observed in TR2_5, while
lowest values of J (0.7) were recorded in TR1_20, TR2_20, TR3_5 and TR3_20
(Table 2). BOPA index showed highest values in TR2_20, TR3_5 and TR3_20
(0.1761), Bentix index in TR4_20 (3.82), AMBI in TR3_10 (3.202) and M-AMBI in
TR2_5 (0.91924). BOPA and M-AMBI indices showed a generalized High/Good
ecological status, AMBI assigned levels of disturbance that ranged between
undisturbed to slightly disturbed, while Bentix provided a classification that
ranged between high and moderate (Table 2).

Concerning the chemical
analysis, anthracene was the most abundant compound (158 µg/Kg in TR1_10),
followed by: benzo(b)fluoranthene
(107 µg/Kg in TR1_10), fluoranthene (56 µg/Kg in TR3_20), benzo(a)pyrene (91 µg/Kg in TR1_10), naphtalene
(86 µg/Kg in TR1_10), benzo(k)fluoranthene
(34 µg/Kg in TR1_10) and benzo(g,h,i)perylene
(<10 µg/Kg in all stations). The abundances of indenopyrene and PCBtot were < 5 µg/kg in all the sampled stations. Among trace elements, Zn showed the highest abundance values in the whole study area, reaching a value of 50.6 mg/Kg in TR1_10, followed by: Crtot (38.4 mg/Kg in TR3_5), Cu (18.8 mg/Kg in TR2_10), Pb (24.3 mg/Kg in TR1_10), Ni (15.4 mg/Kg in TR3_5), As (8.0 mg/Kg in TR1_10); Hg (9.49 mg/Kg in TR3_20) and Cd that showed abundance <0.3 mg/kg in all analysed stations. TBT, DBT and MBT showed the highest abundance in TR3_20 with values of 522 and 22 ng Sn g-1 respectively (Table 3).   Statistical analysis Pearson correlation coefficients, and their P-values, of the entire dataset were reported in Figure 2. This analysis highlighted that the grain-size composition is highly correlated with the AMBI index, in particular sand (0.7) and silt (-0.6). Regarding the plastic analysis, macro and plastic dimensional categories are highly related with J diversity index, with values of -0.6 and -0.8 respectively. Moreover, the microplastic resulted negatively correlated with the H' and with the M-AMBI (-0.7 and -0.8). TBT, DBT and MBT resulted only intra-correlated. Concerning the trace elements, Cu resulted negatively correlated with S, H' and M-AMBI, with values of -0.6, -0.6 and -0.7 respectively. M-AMBI resulted negatively related also with Zn (-0.6) and Pb (-0.7). The best linear models chosen after the model selection analysis were reported in Table 4. In detail, S resulted negatively related with the presence of microplastic showing a decreasing with the depth factor; J resulted positively related with sand, clay and the depth factor, while was negatively correlated with TBT. As regard the M-AMBI biotic index, resulted significant negatively correlated only with microplastic abundance. PERMANOVA analysis revealed no significant differences between the abundance of the four different plastic size categories by factor depth. Results of the two-way multivariate analysis conducted on the species abundance revealed significant differences in macrobenthic community composition between investigated depths (F1,24 = 2.8904; p<0.01) and between transects (F1,24 = 1.7146; p<0.01). Pairwise comparison showed significant differences among all the levels of the factor depth, except for 5 m vs 10 m, and among all the levels of factor transect, except for TR3 vs TR4. SIMPER tests showed the highest average dissimilarity between the macrobenthic composition at 5 and 20 m (? = 75.31), ascribable to the bivalve C. gibba, more abundant at 5 m, and the polychaetes Aricidea (Aricidea) pseudoarticulata, more abundant at 20 m depth. Moreover, the highest average abundances of the polychaetes Pseudoleiocapitella fauveli, at 10 m was responsible of the dissimilarities among this and the other investigated depths. As regards transect, SIMPER tests showed the highest average dissimilarity between the macrobenthic composition at TR1 and TR2 (? = 78.92), ascribable to the higher average abundances of the two polychaetes A. pseudoarticulata and P. fauveli recorded in TR2. While, C. gibba is the species that most contributed to the dissimilarities between TR3 and the other investigated transects (TR1 and TR2). Concerning the relationships between macrobenthic abundances vs. contaminant concentrations, plastic debris and grain-size composition, the similarity test based on SIMPROF analysis of environmental data revealed three distinct groups (A, B, C) (p<0.05) confirmed via hierarchical cluster analysis (Figure 3a). Their characterization is showed in Table 4. The PCO1 axis (Figure 3b) accounted for 33.6% of the total observed variation and clearly separated the three groups. PCO2 axis gathered 17.4% of the total variation. The superimposed vectors showed two sets of contaminants, one associated with the group A and one related with group C; no contaminants appeared to be associated with group B. Group A included the sampling station TR1_10 characterized by silt sediments with the highest concentrations of PAHs (anthracene) and trace elements, and the highest abundance of microplastics and macroplastic (Table 5). This group comprised 14 macrobenthic species, one of which, Kirkegaardia dorsobranchialis, non-indigenous. Group B comprised TR1_5 and TR2_5 characterized by the highest concentrations of the organotin compound DBT and by the presence of 59 macrobenthic species, three of which non-indigenous (Figure 3; Table 5). Group C was composed by the remaining sampling stations (7), that showed the highest concentrations of the organotin compounds TBT and MBT and the highest total number of macrobenthic species (88), three of which non-indigenous (Table 5). 


I'm Dianna!

Would you like to get a custom essay? How about receiving a customized one?

Check it out