Sediment type, carbonate and organic carbon
Grain size composition of recent sediments in Edremit Gulf display high sand contents as (mean)45% followed by silt and clay with a median value of 43% and 10%, respectively (Table 1). Their composition is mainly silty and sandy, with varying rates of coarse and fine, thus they include the entire range from mud to silt, sandy silt, silty sand and sand (Fig. 2a). Considering the spatial distribution of the sand-sized material, except for Ayvalık port and Gömeç coastal areas, the southern part of Edremit Gulf and the northwest coastal zone around Müsellim Passage where it is connected with the open Aegean Sea has a high concentration of over 50% (Fig. 2b). Silt contents (> 50%) prevailing in the zone between the southern shore of Edremit inner gulf and Assos basin, high values (> 75%) are observed in Assos basin and Altınoluk regions (Fig. 2c). While clay, the finest sedimentary unit with a maximum content of 35%, shows distribution in deeper areas of the Assos basin and southern offshore section of Küçükkuyu, it also has high values in coastal areas of Ayvalık, Gömeç and Ören (Fig. 2d).
Fine-grained sediments (silt-clay) show higher contents in deeper sections of Edremit Gulf with exceptions on south-eastern coastal areas between Ayvalık and Burhaniye. Usually, grain-size distribution trends were expressed by depth variations, with finer particles being found in the deeper sections (Assos basin and offshore Küçükkuyu) and coarser sediments in shallow areas (northern and southern entrance of Gulf). Coarse grained sediments are widespread on the shelf area, which is normally shallower than 50m, at the southern and northern entrances (around Alibey island and Musellim passage) where Edremit Gulf is connected with the open Aegean Sea. This configuration can be evaluated as a result of bottom water movements developed under the effect of cyclonic water circulation (Giamali et al. 2020), which is characteristics of the northern Aegean Sea.
Organic carbon(Corg) content show high variability, in between 0.37% and 3.44%, and mean value of 1.67% (Table 1); the highest organic carbon values (up to 3.0%) are detected both coastal areas of southern and northern exits and additionally the eastern part of the inner Gulf (Fig. 2e). While the higher contents of the Corg is a factor of limited circulation in Ayvalık Bay, other areas of local maxima possibly related to waste-water treatment plants (WWTP) located around the Edremit Gulf. Potential effects of these locations on Corg, carbonate and trace element contents (eg As, Cr, Cu, Ni, Pb, Zn) will be discussed in the following sections.
According to microscobic observations Edremit Gulf sediments contain autochtonous biogenic carbonates, such as foraminifera, bivalves, gastropod shells, and their fragments. Total carbonate content of Edremit Gulf sediments showed high variability, with a maximum level of 65%, minimum level of 7% and mean value of 27% (Table 1) that indicates that on average of more than one-fifth of the sediment samples consist of biogenic particles. Carbonate distribution (Fig. 2f) shows different trends in comparison to Corg, with maxima (> 50%) in southern entrance of the Edremit Gulf around Alibey island and Northern exit of the Gulf.
Contents and areal distribution of trace elements
Investigated heavy metal levels of 68 surface sediments of Edremit Gulf presented in Table 1. The overall results demonstrated that the contents of trace elements in recent sediments were in the range of Al: 1.8–8.17%, Fe: 1.24–4.86%, As: 16–63 mg/kg, Cr: 26–200 mg/kg, Cu: 6.4–51.5 mg/kg, Ni: 14.6–137 mg/kg, Pb: 16.8–63.7 mg/kg, Zn: 31–123 mg/kg. The comparison between metals concentrations in the current study and the most recently published data at marine environments in Mediterranean Sea are given in Table 2. In the limited number of studies in which results on As concentrations were given, the mean As values were lower than the values measured in this study (Martínez-Guijarro et al. 2019; Karageorgis et al. 2020, Radomirović et al. 2021). While the average values of Cr, Cu and Ni from other metal concentrations are lower than Christophoridis et al. (2019), Karageorgis et al. (2020), Radomirović et al. (2021) and higher from Martínez-Guijarro et al. (2019), Karditsa et al. (2014), Abbasi and Mirekhtiary (2020), Ben Amor et al. (2019), Balkıs et al. (2020). The average values of Pb and Zn are only higher than Koukounari et al. (2020), Martínez-Guijarro et al. (2019), Abbasi and Mirekhtiary (2020) and lower than the others listed in Table 2. Overall, the metal concentrations studied in this study showed significant fluctuations compared to some of the most recent studies on coastal areas mentioned in Table 2. When the mean concentrations of heavy metals in the study region of Edremit Gulf are examined, Cr has been the highest value with 86.2 mg/kg followed by Zn (72.3 mg/kg), Ni (55.2 mg/kg), Pb (35.9 mg/kg), As (29.2 mg/kg), Cu (22.7 mg/kg) and Al (6.15%), Fe (3.21%). Based on Fig. S1, which shows the spatial distribution of trace elements in the sediments from the Edremit Gulf, the distribution patterns of trace element concentrations of As, Cr, and Ni tend to decrease from the same coastal sources, while Cu, Pb, Zn and Al and Fe showed commonly deep basin concentrations. In the eastern half of the Gulf, both northern and southern side, the levels of metals were relatively high, showing that discharges of small rivers and streams were the possible pollution sources. Contents of some trace elements, as As in some areas have four times higher than their background levels. Highest metal (Cr, Zn, Ni) concentrations distributed mainly eastern half of the north side of the Gulf and central part of the southern shore around Gömeç. While high As contents are mainly focused on the coastal regions of southern side, in the north they spread only at the center of Küçükköy area and decrease relatively without showing any depocenter in the offshore part of the Gulf. The distribution model of the center with high As values are predominantly in accordance with the local wastewater treatment plant discharge locations.
Unlike other metals, Cu, Pb and Zn, which show approximately the same distribution patterns, whose common features are an increase in concentration from the coast to the open sea have high concentrations above the reference values in shallow and deep depocenters. All these element contents tend to decrease in the offshore direction where lower concentrations were observed than the accumulations on the eastern coastal areas. The Al and Fe values, which showed concentrations close to reference values, contain high values in natural depocenters shaped in accordance with the bottom morphology of the Gulf (Fig. S1). Since the coastal areas with open sea connections in the west of the Gulf are exposed to more sediment movement compared to the inner parts of the bay, they have mainly erosional/transport rather than depositional areas. The average concentrations of heavy metal contents of the inner coastal canters that depositional trends more active are in the order Cr > Zn > Ni > Pb > As > Cu. The only study done on the basis of limited sediment samples across the Edremit Gulf belongs to Meriç et al. (2012). According to this study, as a result of Co, Cu, Cr, Mn, Ni, Pb, Fe and Al elements, it was concluded that the metal enrichment in Edremit Gulf originated from the metal ores located in the northwest Aegean Region.
Relationship between element concentrations and sediment grain size
Organic matter and sediment grain size are the most important parameters of sediments that affect the concentration of metals (Giordano et al. 1999; Cai et al. 2011). Fine particles absorb soluble metals in seawater and transport them to the surface sediments of the seafloor. In general, the levels of heavy metals increase with decreasing grain size (Förstner and Wittmann, 1983) and the fine fraction (< 63 µm) is generally preferred in environmental research as it has the ability to bind elements through complex adsorption and/or absorption mechanisms (Tessier et al. 1984). Among all metals examined in this study, Al (r = 0.569, P < 0.001), Cr (r = 0.515, P < 0.001), Cu (r = 0.774, P < 0.001), Ni (r = 0.673, P < 0.001), Pb (r = 0.534, P < 0.001). Zn (r = 0.783, P < 0.001) showed an extremely positive correlation with fine material consisting of silt and clay (Table S4). In contrast, the content of As (r = 0.232, p < 0.10), which correlates poorly with the fine material, is also not associated with the coarse sized material. This indicates that grain size is not the predominant factor controlling the As distribution in the surficial sediments of the Gulf and probably a different source of As from all other metals. The TOC levels in the sediments varied from 0.37–3.44%. Heavy metals tend to be absorbed by biological matters (Ianni et al. 2010), so the organic matter is a good metal filter (Wasserman et al. 1998; Weiping et al. 2014) and plays an important role in metal reserves in sediments. Edremit Gulf sediment TOC and heavy metal data show no such correlation, except for Cu and Pb, which are positively correlated with low values (Table S4) and therefore predominantly indicate anthropogenic and/or lithogenic source other than biological source.
Trace element interrelations and source identification
In Pearson's correlation coefficients (Table S4), a positive and significant correlation was observed with the fine material, lithogenic Al and Fe and other elements except As, while it was negatively correlated with CaCO3, which was positively correlated with coarse grained material and represented biogenic carbonates (also see Figs. 2b and 2d). All trace elements except As are distinctly related to Al and Fe suggesting that their common origin is due to lithogenic or similar transport patterns. Various anthropogenic effects on the region were determined by evaluating the relationships between trace elements, organic carbon, carbonate and grain size. For this purpose, by grouping variables with common geochemical behaviour. Principal Component Analysis (PCA), in which the varimax rotation is applied with Kaiser Normalization, has been performed to the existing data to define the heavy metal sources in Edemit Gulf surface sediments. By extracting the eigen values and eigen vectors from the correlation matrix, the number of important factors and the percentage of variance described by each were calculated.
Table S5 demonstrates the results of varimax transformed factor values and Eigen values and communalities. The results showed that there are five Eigen values higher than 1 and these five factors cover 83% of the total variance. The first factor accounts for 51.5% of the total variance and concentrates strongly on Ni, Cr, Co (> 0.7) and moderately to low for Fe, V, Al, Mn and Cu which are loaded > 0.4. Factor 1, which is partly positively loaded lithogenic elements associated with fine-grained soil including silt (0.6) and clay (0.2) indicates an anthropogenic-lithogenic origin with high Cr and Ni values predominantly from mining waste. As can be seen from Fig. 3a, it points to the important input originating from the northern shores of the bay and affecting a large section of the Gulf. Factor 2 constitutes 12.9% of the total variance, with Sn, Zn, Pb, Cu scores above > 0.7 and secondarily with relatively high values of Silt and clay dominated by Al, Fe and V (> 0.4). Considering the relevant metals and the spatial distribution model given in Fig. 3b, together with the partial lithogenic source, it indicates an anthropogenic pollution associated with the Ayvalık cantered maritime traffic. Traffic could be considered a notable source of Cu, Ni, Pb, Sn and Zn pollution, particularly since there are urban areas of around the gulf. Moreover, runoff is an important source of heavy metals in the marine surrounding, while contents of the heavy metals from transporting emissions and ships are correlated with the density of traffic (Christophoridis et al. 2019). Factor 3 is loaded primarily by Ba, Al, and Nb > 0.7, as well as moderately by Rb, V, Pb and Fe (0.4), making up 9.5% of the total variance and representing the lithogenic factor with high Al value. Low (> 0.4) level loadings (Al, Fe,) within the first two factors indicate the partial lithological effect within the respective group. In total, these three factors account for 73.9 % of the total variance and are shown in a three-dimensional space (Fig. 3). As can be seen from Fig. 3 (top left), the first three factors were found to overlap locally in regions around Al and Fe. This combination consisting of the contributions of Al and Fe (Table S5), which are among the basic parameters of all three factors, indicates that the main factor in Edremit Gulf is predominantly lithogenic. The main element group of first three factors includes primary heavy metals (Ni, Cr, Co, Cu, Pb and Zn) and nutrients (organic carbon), which are normally thought as the anthropogenic pollutants mostly related to the discharge of agricultural, industrial wastes and untreated domestic sewage and also marine and land traffic activities (Qi et al. 2010; Lu et al. 2010). Anthropogenic sources of heavy metals are indicated in the world wide marine areas in numerous former studies (Abbas et al. 2016; Birch, 2017; Martínez-Guijarro et al. 2019; Karageorgis et al. 2020).
Factor 4 accounts for 5.3% of the total variance with high contribution of As and Sb and low contribution of Mn, Co, Ni, Fe, Cr, Cu, respectively. In the spatial distribution model given in Fig. 3d, high values are mainly concentrated in the locations where wastewater treatment systems and geothermal resources are located, and the relevant positive metal charges indicate the wastewater discharge -geothermal factor in this study. As with high loading (0.9), which is the highest representative of factor 4, was suggested to be associated with geothermal waters in Somay (2016). Factor 5, which accounts for the lowest 3.9 % of the total variance, is strongly correlated with Mo, S, organic carbon, almost all of which have high loading values (> 0.68). This factor has been defined as the "organic matter sulfurization" factor due to the following environmental characteristics (Fig. 3e). According to Qi et al. (2010) S has negative correlations with anthropogenic and lithogenic components and positive correlations with biogenic elements, showing a marine source instead of terrigenous origin. High OC, high Mo and low As conditions demonstrate possible circumstances of sulfurized organic material in sediments. High Mo levels point out that reactive iron was limiting factor, preventing As capture (Tribovillard, 2020).
Establishing natural background levels in the study area
Local geochemical background concerns the element content in the sediments without any anthropogenic enrichment, i.e., the element content is supplied by natural sources such as biogenic and terrigenous components. After preliminary trials using three different sediment cores it was found that there is no significant difference in the background levels between the three locations of the Edremit Gulf. To distinguish the pre-industrial deposits in the Gulf, available two 14C dated core data taken from northeastern part of Edremit Gulf were used from Yümün and Önce (2017). In the Edremit Gulf, the sediment accumulation rates were calculated at 0.57 cm y− 1 and 0.75 cm y− 1, thus element contents of deepest core levels (88-150cm) are used as background values. Table 1 displays the pre-industrial values of metals defined as background values for each core locations.
Risk assessment and contamination
The pollution indices are useful tools to assess the sediment contamination. In this study, several relaiable contamination and risk indicators utilized in previous investigations, as well as new indices were used to asses pollution degree and potential ecological risk (Bourliva et al. 2018; Christophoridis et al. 2019). The Enrichment Factor (EF), Geoaccumulation Index (Igeo), Contamination Factor (CF), Modified Pollution Index (MPI) and Toxic Risk Index (TRI) were used. Morover, SQGs along with the QTEL and QPEL were used to estimate the seabed surficial sediment contamination in sampling region.
Enrichment Factor (EF)
To analyze if the heavy metals in Edremit Gulf surficial sediments were affected by anthropogenic activities, enrichment factors (EFs) based on the earth’s crust were computed using Eq. 1 and the results are shown in Fig. 4. Calculated EF values exhibited ranges of 1.3–8.9 for As, 0.4–2.7 for Cr, 0.3–1.3 for Cu, 0.3–2.3 for Ni, 1.4–5.1 for Pb and 0.6–1.8 for Zn. The highest EFs were calculated by As followed by Pb, while average EFs indicated decreasing order of As > Pb > Cr > Ni > Zn > Cu with respective values as 3.1, 2.5, 1.3, 1.1, 1.0 and 0.7. The EF values of all determined heavy metals (except Cu and Zn) are higher than 1.5 that suggesting their anthropogenic source. Alternatively, the EF values of As, Pb > 5 and Cr, Ni > 2 indicated significant to moderate enrichment respectively. The basic distribution of EF values between stations (a) and within each heavy metal (b) were shown in Fig. 4. While moderate and significant enrichments are observed on EF values of As, Pb, Cr and Ni, all values for Cu and Zn indicate minimal enrichment. When the spatial distribution of EF-As values, which constitute the highest values among all metals studied, is examined, almost all of the southern coastal areas and the central part of the northern coast of around Küçükkuyu demonstrated higher EF values than those in the central part of the Gulf (Fig. 4c). The high EF-As values observed in the northern part where EF-Pb values are minimal, were evaluated as untreated urban sewage and geothermal origin. EF of Cr and Ni display generally minimal values except northern coast around Küçükköy and Edremit, which shows moderate enrichments related with mining activity (Fig. 4e, f).
Geoacumulation Index (Igeo)
The Igeo values of the heavy metals in Edremit Gulf’s surficial sediments were computed based on the crust of the earth using Eq. 2 and results are presented in Fig. 5. All analyzed samples displayed negative Igeo values of Cu and Zn and were defined as “non-polluted” regarding these two metals. Calculated Igeo values ranged from − 0.7 to 1.7 for As, -2.4 to 0.6 for Cr, -3.4 to -0.4 for Cu, -2.8 to 0.4 for Ni, -0.8 to 1.0 for Pb, -2.2 to -0.2 for Zn. The highest Igeo values were calculated by As followed by Pb, while average Igeo values indicated decreasing order of As > Pb > Cr > Ni > Zn > Cu with respective values as 0.5, 0.2, -0.8, -1.0, -1.1 and − 1.7, which was parallel to the trend traced in the enrichment factor values. The Igeo values of Cu and Zn and majority of Cr and Ni in the Edremit Gulf, which proposed that the sediments are not polluted by anthropogenic origins. The general distribution of Igeo values between stations (a) and within each heavy metal (b) is shown in Fig. 5. While moderately contamination is observed on Igeo values of only As, nearly all values for Pb indicate uncontamination to moderately contamination. While a few uncontaminated to moderately contaminated values are observed in Igeo values of Cr and Ni, a significant part of the remaining values consists of uncontaminated as in Cu and Zn.
When the spatial distribution of Igeo-As values is examined, which constitute the highest values among all metals studied, almost all of the southern coastal areas and the central part of the northern coast around Küçükkuyu exhibited higher Igeo values than those in the central part of the gulf (Fig. 5c) the same trend as EF values (Fig. 4c). It has been observed that the Igeo values of the Pb data throughout the area and the Cr and Ni values are uncontaminated to moderately contaminated only in a few stations, especially around Küçükkuyu, on the north coast. Except for the high values of Cr and Ni observed in the limited area around Küçükkuyu, all of the data are uncontaminated (Igeo < 0) throughout the bay as well as Cu and Zn (Fig. 5d,e,f). As shown in Fig. 5g, both EF and Igeo showed similar results for trace element contamination levels in the sediments of Edremit Gulf. As and Pb were dominating contamination, Cr and Ni pollution were also observed in limited coastal regions, proposing that main driving factor for the enrichment of As, Pb, Cr and Ni was anthropogenic origin. Nevertheless, Cu and Zn were firstly originated from natural sources like mentioned at recent work of Wang et al. (2020).
Pollution indices
The calculated contamination factor (Cf) (Eq. 3), pollution load index (PLI), degree of contamination (Cdeg) (Eq. 4) and modified degree of contamination (mCd) (Eq. 5) values for heavy metals are given in Table S6. Cf values of the trace elements were found in the order of As (1.62) > Pb (0.99) > Zn (0.86) > Cr (0.84) > Cu (0.82) > Ni (0.71). Between the calculated Cf values only displayed the moderate contamination for arsenic while the rest of the Cf values exhibited low contamination for Pb, Zn, Cr, Cu and Ni. Since the average pollution load index (PLI) was nearly 1 (0.99), which indicated baseline level of pollution for the entire area of investigation. The interpretation of PLI value is defined as polluted (PLI > 1); baseline level of pollution (PLI = 1) and not polluted (PLI < 1) (Liu et al. 2005; Chakravarty and Patgiri, 2009). The degree of contamination (Cdeg) and mCd values were calculated as 5.84 and 0.97 respectively. These values correspond Low degree of contamination for Cdeg and Nil to very low contamination for mCd.
Multi elemental pollution indices are beneficial to single element ones, as for instance EF, as they combine the additional impact of multiple pollutants, which are frequently existing in industrialized and urbanized surroundings, in sediment quality evaluations. Among other indices within the study, the Modified Pollution Index (MPI) presented by Brady et al. (2015) has also been utilized. Several researchers tried the MPI in varied sampling regions and deduced that it is more confident than other unidimensional and multidimensional indices. This idea has been lately reported by Karageorgis et al. (2020). It was computed by Eq. 6 and could be thought of as an improvement of the Håkanson's modified degree of contamination index (mCd) (Eq. 5) (Håkanson, 1980) in the sense that a suite of elements is combined to generate a single value, in the way that it uses the CFmax to develop a weighted average, hence considering the effect of contamination of one element, that otherwise could be decreased. Sediment samples with 10 < MPI, 5 < MPI < 10, 3 < MPI < 5, 2 < MPI < 3, 1 < MPI < 2 and MPI < 1 were categorized as extremely polluted, heavily, moderately-heavily, moderately, slightly and unpolluted, respectively (Brady et al. 2015). According to Fig. S2, indicating the spatial distribution pattern of MPI, the Edremit Gulf, is entirely polluted starting from slightly to heavily. In the southwestern corner of the area, northern side of the Alibey island heavily polluted sediments were observed close to old Pb mining area (MPI: 6.6). Similarly to the spatial distribution of the As and Pb EFs (Fig. 4c,d), the MPI is decreasing to moderately and slightly polluted ranges towards the east and north respectively. The degradation of sediment quality is coherent with the hydrological regime forcing the dispersion of treated wastewater of the outfall systems both southern and northern side of the Gulf.
Sediment quality parameters
The limit values determined in the sediment quality guidelines (SQG) and the results of the metal analysis performed in the samples were given in Table 1. As seen in Fig. 6 showing the entire data, our study showed that all measured Zn concentrations in sediment samples were below the corresponding TEL value indicating rare (9%) toxic effect that is not associated with adverse biological effects. Mean levels of As, Cr, Cu and Pb were between corresponding TELs and PELs which only the mean value of Ni exceeds PEL. As, Pb, Cr, Cu and Ni concentrations were in the range between the corresponding TELs and PELs by 90%, 78%, 75%, 65% and 35% respectively, which indicate occasionally toxic (22%) adverse biological effects. Ni, As and Cr were the limited metals with concentrations exceeding the PELs by 60%, 10% and 6% respectively and this indicates Ni has the highest adverse biological effect as frequently toxic (56%). Therefore, adverse biological effects may be formed more frequently due to the concentrations of Ni, As and Cr. According to the NOAA sediment quality criteria (ERL and ERM), the levels of Zn for all sediments were less than the corresponding ERL which indicates minimal toxic (25%) effect. The concentrations for Pb, Cu, Cr and Ni are lower than the corresponding ERLs by 94%, 90%, 50% and 12%, respectively. As, Cr, Ni, Cu and Pb contents are in the range of the corresponding ERLs and ERMs by 100%, 50%, 41%, 10% and 6% respectively indicating occasional adverse effects. Ni was the only metal with levels exceeding the ERM (by 47%) which indicates frequently toxic (75–100%) effects (Fig. 6).
Toxicity estimation and risk assessment
To classify the sediments toxicities several principles have been defined. Mearns et al. (1986), categorized the data of amphipod survival tests nontoxic (average survival:96–96.5%), marginally toxic (average survival:76.5–83%), clearly toxic (average survival < 76%), and highly toxic (average survival < 20%). Later Swartz et al. (1995) categorized the results of amphipod survival tests as nontoxic (< 13% mortality), uncertain (13–24% mortality), or toxic (> 24% mortality) and according to Hansen et al. (1996); sediments were categorized as toxic if mortality was higher than 24%, while sediments with ≤ 24% mortality were thought as nontoxic. Like Zhang et al. (2017) this research regarding the uncertainties in the toxicity estimation results of different organisms and end points, a wider toxicity ranges of 10–30% was characterized as “uncertain”, < 10% was defined as “nontoxic”, while > 30% was “toxic”.
If more than 75% of sediment samples in the marine environment are correctly estimated to be toxic or nontoxic, sediment quality guidelines (SQGs) are considered to supply a reliable basis for evaluating sediment quality (Long et al. 1998; MacDonald et al. 2000). Based on this, the concentrations of 6 trace elements (As, Cr, Cu, Ni, Pb and Zn) determined in the surface sediment samples of the Edremit Gulf were compared with the TEL and PEL values in the previous sections (Table 1, Fig. 6). When sediment quality guidelines (SQGs) are evaluated by considering the number of samples with QPEL> 1, QPEL≤ 1 ≤QTEL and QTEL <1 calculated on the basis of the given TEL and PEL, all QTEL and QPEL values vary between 4.9–24.4 and 1.2–6.7 respectively (Fig. S3, top). Based on these data, virtually all of the sampling locations are higher than 1 both for QTEL and QPEL and the whole area shows toxic (30%) characteristics as indicated in Fig. S3.
Figure S3 demonstrates the contribution of each trace element to the total toxicological risks, i.e the ratio of Qi,PEL/QPEL. In the 68 sediment samples, the Qi,PEL of As, Cr, Cu, Ni, Pb and Zn accounted for 8.45%-41.84%, 9.73%-25.57%, 2.76%-11.38%, 14.82%-52.86%, 4.48%-21.04 and 4.22%-15.17% of QPEL, respectively. The results proposed that Ni made a dominant contribution to the toxicities of trace element mixtures in the sediments, which was followed by As, Cr, Pb, Zn and Cu. These results showed that Ni and As were the major metal contaminants in the sediments of Edremit Gulf. In contrast, the risks caused by Cu and Zn were relatively low.
MacDonald (1994), proposed SQGs where the adverse effects of TEL is considered reliable if the minimal effect range is less than 10%, while if the adverse effects exceed 65% of the probable effect range the PEL is considered consistent. Thus, the TEL of trace metals in sediments would display only limited acute toxic impacts to exposed aquatic organisms. However, if the exposure period is long enough the possibility of chronic toxic impacts cannot be refused. Hence, a TRI integrating the TEL and PEL was used to evaluate toxic risks in terms of both chronic and acute toxic impacts in aquatic organisms exposed to sediment polluted by trace elements. Overall TRI results, displayed moderate toxic risk (10 < TRI ≤ 15) at 23 locations, while no toxic risk (TRI ≤ 5) and considerable toxic risk (15 < TRI ≤ 20) were observed at three (Stn; 9, 18, 55) and two (Stn; 24, 45) stations respectively (Fig. 7). The mean TRIi values of metals showed a decreasing order of TRIAs (2.91) > TRINi (2.62) > TRICr (1.23) > TRIPb (0.87) > TRICu (0.86) > TRIZn (0.45) with mean contributions of 33.5%, 27.8%, 13.3%, 10.5%, 9.8%, 5.2% respectively. Calculated TRI results indicate that As accounts for most of the overall sediment toxicity. Compared to the others, the most prominent difference of As and Ni, which provide the most important contribution to TRI, is mainly their relatively low TEL values and their high concentrations measured in surface sediments. This points out the potential toxicity of sediments in Edremit Gulf and stands out As and Ni as two heavy metals deserving more concern.
Sediment Transport Patterns
Using the GSTA model the sediment transport vectors were generated. The transport pattern that was developed is identified and displayed in Fig. 8, which shows the grain size trend and transport directions, denoted by arrows, in the Edremit Gulf. The vectors represented the net transport directions of surface sediments, and the vector length indicates the importance of grain size trend, rather than the scale of transport rate.
The circulation of the coastal marine reflects the directional flow features on the dispersion and regional tendency of the sediments. The primary driving forces for sediment transport in shallow nearshore are wind induced waves and associated movements. The conclusion of the ADCP measurements in Edremit Gulf showed that currents were primarily driven by winds and the associated fluctuation of sea surface layers. The ADCP data recorded for the shallow (< 50m) zone demonstrated that the surface currents are generally slightly faster than the bottom currents that form in the opposite direction (TUBITAK 115Y180, 2018). The surface waters that moved towards the east at the southern zone and in the opposite direction on the northern coastal areas displayed complex and small-scale cycles that occur in the southern part of the Gulf because of the coastal structure. Small scale cyclonic cycles in the comparatively shallow eastern part and an anti-cyclonic cycle in the coastal zone around the eastern section of the islands were observed. The general circulation of this region is not only wind-driven but also the general thermohaline circulation of the Aegean Sea, particularly during the summer months, contribute significantly to the deep-water circulation system of Edremit Gulf.
In this study, two distinct sediment transport modes consisting of strong basinward transport sediments consisting of mainly river-born and westerly directed weak transport of basin sediments were recognized in Edremit Gulf. As tracers for sediment movement elemental, carbonate, carbon content and contamination assessment data were evaluated. Within this system, utilizing tracing and hydrodynamic information of the pattern of sediment transport was also evaluated. Even though the main controls of sediment transport are definitely hydrodynamic, the depositional zones are strongly influenced by the topography where the geomorphological characteristics trapping the sediments. Spatial distribution of the trace elements, PCA components, degree of sediment pollution and several other indices are consistent with the determined residual grain-size trend pattern of Edremit Gulf transport vectors (Fig. 8).