Marine Pollution Bulletin 123 (2017) 313–323

Contents lists available at ScienceDirect

Marine Pollution Bulletin

journal homepage: www.elsevier.com/locate/marpolbul

Ecological quality status of the Adriatic coastal waters evaluated by the MARK organotin pollution biomonitoring

⁎ A. Erdeleza, M. Furdek Turkb, ,A.Štambukc,I.Župand, M. Pehardaa a Institute of Oceanography and Fisheries, Šetalište I. Meštrovića 63, 21000 Split, Croatia b Department for Marine and Environmental Research, Ruđer Bošković Institute, Bijenička cesta 54, 10000 Zagreb, Croatia c Department of Biology, Faculty of Science, University of Zagreb, Roosveltov trg 6, 10000 Zagreb, Croatia d Department of Ecology, Agronomy and Aquaculture, University of Zadar, Trg kneza Višeslava 9, 23000 Zadar, Croatia

ARTICLE INFO ABSTRACT

Keywords: The aim of this study was to evaluate the post-legislation change in tributyltin (TBT) pollution at Croatian TBT pollution Adriatic coast. Gastropod trunculus and sediments were collected, nearly 10 years after TBT based Imposex antifouling paints were banned, at 12 locations along the coast where a previous study was conducted in 2005. The study showed a decline of TBT levels over the investigated period, although all gastropods populations were Adriatic Sea highly affected by imposex meaning that prohibition did not result in the recovery of populations. The further Ecological status assessment aim was to propose the Ecological Quality Ratio (EQR) boundaries for potential use of H. trunculus as a principal Mediterranean bioindicator in the assessment of the ecological status of the Mediterranean regarding TBT pollution, under the Water Framework Directive (WFD). According to the proposed EQR classes, the WFD target for achieving the Good ecological status of the marine environment by 2015 was not reached.

1. Introduction et al., 1995). All these traits make this a reliable bioindicator of TBT pollution. During the last few decades, imposex in H. trunculus has Organotins, namely tributyltin (TBT), were introduced into the been widely investigated in relation to TBT pollution in many Medi- marine environment mainly through the application in antifouling terranean countries, e.g. Malta (Axiak et al., 1995, 2003), Italy (Terlizzi coatings. TBT may cause various biological effects on different non- et al., 1998, 1999, 2004; Chiavarini et al., 2003; Pelizzato et al., 2004; target organisms, but the main toxic effect caused by this compound in Garaventa et al., 2006, 2007), Israel (Rilov et al., 2000), marine ecosystem is occurrence of imposex in prosobranch gastropods (Vasconcelos et al., 2006a), Croatia (Prime et al., 2006; Garaventa (Smith, 1981; Bryan et al., 1987; Graceli et al., 2013). Imposex is de- et al., 2006, 2007; Stagličić et al., 2008) and Tunisia (Lahbib et al., fined as superimposition of male sex characteristics in females (Smith, 2007, 2009, 2010). The majority of these studies investigated spatial 1971). Previous studies demonstrated that some gastropod species de- distribution of imposex occurrence and some have related it to TBT velop imposex even at TBT concentrations in seawater as low as pollution. − 1ngSnl 1 (Gibbs et al., 1988), and that imposex intensity increases as Because the organotins exhibited negative effects on non-target environmental TBT concentration increases and as TBT bioaccumulates organisms, primarily gastropods and bivalves, their use in antifouling (Bryan et al., 1987; Oehlmann et al., 1998a). At higher TBT con- coatings is nowadays banned in many countries worldwide, including centrations advanced stages of imposex occur, severely affecting fe- all of (Sonak et al., 2009). All antifouling coatings containing males' reproductive capabilities, population recruitment and structure organotins are banned by the International Convention on the Control (Gibbs et al., 1988). of Harmful Anti-fouling Systems on Ships (AFS Convention) since the Up to date > 200 species have been proposed to monitor TBT pol- year 2008. Croatia ratified AFS Convention in 2006, and in the same lution worldwide, including Hexaplex trunculus (Horiguchi, 2017). year TBT was included in the List of hazardous substances that are Banded - H. trunculus is a common Mediterranean gastropod. prohibited for use (Official gazette 17/2006), however ships containing This species inhabits littoral, is locally abundant, has limited mobility, organotins coatings have not been banned from entering Croatian ports does not have planktonic larval stage and develops imposex at TBT until 2008. In EU, the prohibition of organotins on ships entered into − concentrations even lower than 1 ng Sn g 1 dry weight (d.w.) (Axiak force through Directive 2002/62/EC and Regulation EC/782/2003.In

⁎ Corresponding author. E-mail address: [email protected] (M. Furdek Turk). http://dx.doi.org/10.1016/j.marpolbul.2017.08.039 Received 18 January 2017; Received in revised form 24 July 2017; Accepted 16 August 2017 Available online 25 August 2017 0025-326X/ © 2017 Elsevier Ltd. All rights reserved. A. Erdelez et al. Marine Pollution Bulletin 123 (2017) 313–323

Fig. 1. Location of sites sampled for Hexaplex trunculus and sediments in 2015 (M1 = Marina, M2 = Milna, M3 = Biograd, H1 =Baška Voda, H2 = Rogač,H3 = Vis, B1 = Zaton,

B2 = Lastovo, B3 = Crvena Luka, P1 = Brsečine, P2 = Drage, P3 = Mali Ston).

Annex VIII of the Water Framework Directive (WFD) 2000/60/EC or- (2004) and Garaventa et al. (2006, 2007) investigated imposex indices ganotins are declared as one of the main pollutants, and TBT is enlisted in H. trunculus collected in Venice area (Italy) in the time period as one of the priority substances in the field of water policy (2455/ 2002–2003, the former including also coastal area around western Is- 2001/EC) that should be monitored in order to evaluate the chemical tria peninsula that is part of Croatia. Furthermore, Stagličić et al. status of a water body. The presence of TBT in the environment is (2008) analysed banded dye-murex collected in 2004 and 2006 at regulated by the prescribed maximum allowed concentration in sea- Kaštela bay (Croatia), while Carić et al. (2016) presented data on im- − water. Considering that this concentration is extremely low, 0.2 ng L 1 posex in H. trunculus collected from several sites in Dubrovnik area in − (0.08 ng Sn L 1) for all of Europe, chemical analysis of TBT at such low 2006. In all of these studies the investigated populations were seriously levels are often difficult to perform even with the most sensitive ana- affected by imposex (VDSI ranged from 2 to 5). The first observation of lytical methods. For this reason, the highly specific biological response imposex in H. trunculus in the central Croatian Adriatic was conducted to TBT, i.e. imposex in gastropods, is often used as a valuable tool in in 2005 (Prime et al., 2006) and represents the baseline data for the estimating the impact of TBT pollution in the marine environment. evaluation of temporal trend in imposex occurrence conducted in this Therefore, imposex became a mandatory element under the OSPAR study. It demonstrated high levels of imposex at 12 investigated loca- environmental monitoring programme (OSPAR, 2004). Furthermore, in tions categorized by boating activity, although three of them were ca- the WFD monitoring programme it was proposed as one of the biolo- tegorized as “pristine” areas. gical quality elements that could be used for evaluation and classifi- WFD set the year 2015 as the target deadline to achieve Good cation of ecological status of coastal waters (WFD-UKTAG, 2014). For ecological and chemical status of the Mediterranean marine environ- this to be accomplished, classes of ecological status based on imposex ment regarding TBT (Laranjeiro et al., 2015). The question is whether levels should be defined as proposed by Laranjeiro et al. (2015). this has been accomplished? Based on the above presented review of Several recent studies reported the widespread appearance of im- literature data it can be stated that new data are necessary to evaluate posex in the Mediterranean gastropod populations, such as in Sardinia the environmental impact of the TBT ban so this question could be and Tunisia (Anastasiou et al., 2015; Boulajfene et al., 2015; Abidli answered. et al., 2013), however data on temporal trends of imposex status in the The aims of this study were 1) to provide the first insight on tissue Mediterranean related to the TBT ban is limited (the only data is given burden of butyltins (TBT and its degradation products, dibutyltin (DBT) by Lahbib et al. (2009) for the Tunisian coast). Also, no recent data on and monobutyltin (MBT)) in populations of H. trunculus in central temporal trends of imposex occurrence in the Adriatic, the northern- Croatian Adriatic, and to assess the relationship between butyltin most arm of the , are available. However, measure- concentrations in the tissue and sediment with the imposex level; 2) to ments of TBT concentrations in seawater and mussels from Croatian evaluate the effect of enforced law restrictions on the use of organotins coast for the time period 2009–2010 showed that eastern Adriatic was in antifouling paints over a 10-year period by analysis of changes in polluted with organotin compounds even after the ban (Furdek et al., imposex occurrence in banded dye-murex populations at 12 location in 2012). the central Croatian Adriatic; 3) to propose Ecological Quality Ratio The first study of imposex occurrence in the Adriatic Sea was con- (EQR) boundaries for the potential use of imposex in H. trunculus as a ducted by Terlizzi et al. (1998) in Brindisi area, Italy. Pelizzato et al. biological quality element in the assessment of the ecological status of

314 A. Erdelez et al. Marine Pollution Bulletin 123 (2017) 313–323 coastal waters regarding TBT pollution in the frame of WFD directive, per each group). The samples were lyophilized and stored in the dark at and to use proposed EQR boundaries to evaluate if the Good Ecological −20 °C until organotin analysis. Surface sediment samples were also Status regarding TBT is achieved along the Eastern Adriatic Coast. lyophilized and stored in the dark at −20 °C until organotin analysis. The extraction of butyltins (BuTs) from the tissue was performed by 2. Materials and methods the method described in details by Furdek et al., 2012. Briefly, BuTs − were extracted from the homogenized tissue (1–1.5 g) in 0.1 mol l 1 2.1. Study area and sampling HCl in methanol (10 ml) by ultrasonic stirring for 30 min. The sus- pension was centrifuged at 4200 rpm for 10 min and supernatant (1 ml) Samples of H. trunculus and sediments were collected over a 6 week was used for further analysis. Then, simultaneous derivatization with period (May–June 2015) from 12 locations along the Croatian Adriatic NaBEt4 solution (1 ml 1% (w/v)) and extraction into hexane (1 ml) coast (Fig. 1). The sampling locations were adopted from the study were carried out in a sodium acetate-acetic acid buffer (20 ml, − conducted in year 2005 by Prime et al. (2006). Locations were cate- pH = 4.8, c = 0.4 mol dm 3) by mechanical shaking at 400 rpm for gorized by the expected intensity of the boating activity: (i) nautical 30 min. The extraction of BuTs from the sediments (fraction < 2 μm) marinas (M), (ii) village harbours (H), (iii) sheltered bays that were was performed by the method previously described by Milivojevič seasonally used for boating activities (B), and (iv) reference sites very Nemanič et al. (2009) and Furdek (2015). Briefly, BuTs were extracted rarely used for boating activities (P). For comparison of imposex be- from the sediment (1–2 g) by acetic acid (20 ml) and ultrasonic stirring tween years 2005 and 2015 data from Prime et al. (2006) as well as for 30 min. The simultaneous derivatization and extraction step was unpublished data from that study (Peharda, unpublished data) were performed in the same manner previously described for the biological used. samples. The detection of BuTs (TBT, DBT and MBT) in tissue and se- From each location 60 adult individuals (> 40 mm) were collected diments was carried out on a gas chromatograph (GC, Varian CP3800) by snorkelling and SCUBA diving at 1–10 m depths. As the lifespan of with a pulsed flame photometric detector (PFPD, Varian). The quality H. trunculus is around 7–8 years (Axiak et al., 2003), and at 3 years of control was performed by the analysis of standard reference material age this species already reaches 60 mm of shell length (Vasconcelos certified for BuTs in mussels (CE 477, ERM, Geel, Belgium) and for BuTs et al., 2006b), individuals larger than 40 mm were considered suitable in marine sediments (PACS 2, Ottawa, Canada). Tripropyltin (TPrT) for assessing the contamination status of the marine environment. The was used as internal standard while quantification of BuTs was per- samples were transported to the laboratory, maintained in plastic tanks formed by applying standard addition calibration method. The results with flowing seawater and examined for imposex level within 2 days, as obtained for BuTs were in agreement with the certified values, con- in Prime et al. (2006). The surface sediments (0–2 cm) were collected firming the accuracy of the applied analytical method for the de- by SCUBA diving using plastic cups, and were stored at −20 °C im- termination of butyltin compounds in the investigated samples. The − mediately after the sampling. limits of detection (DL) were 0.9, 1.0 and 1.9 ng Sn g 1 d.w. for TBT, − DBT and MBT in tissue, and 1.5, 2.1 and 6.1 ng Sn g 1 d.w. for TBT, 2.2. Imposex analyses DBT and MBT in sediments, respectively.

The shell length (apex to aperture) was measured to the nearest 2.4. Statistical analyses 0.1 mm using Vernier callipers, the shell was cracked in a vice, and soft body was gently removed. The were not narcotized. After the Statistical analysis was performed using Statistica 8.0. The homo- operculum was removed, the longitudinal cut was performed on the geneity of variances was tested by Levene's test. Difference between hypobranchial gland to observe sexual features of each individual. years 2005 and 2015 for RPLI and VDSI were tested by Mann-Whitney Males were identified by the presence of penis and vas deference, and U test. Kruskal-Wallis test was used for: 1) comparison of imposex in- the absence of the vaginal opening and/or capsule gland. Females were dices between and within different boating categories, and 2) com- identified by the presence of the vaginal opening and capsule gland. parison of butyltins concentrations within categories of sites according The length of straightened penis from its tip to its junction with body to boating activities. Spearman rank correlation analyses was pre- wall was measured for males and imposexed females with digital formed between imposex parameters and butyltins tissue burden. For Vernier callipers to the nearest 0.1 mm. Three imposex indices were samples with butyltins levels below the detection limit (DL), half of DL determined for each location: (i) the relative penis length index (RPLI, value was used. Stroben et al., 1992)defined as (mean female penis length) / (mean male penis length) × 100, (ii) the cubed relative penis size index (RPSI, 3. Results Gibbs et al., 1987)defined as (mean female penis length)3 / (mean male penis length)3 × 100, and (iii) the vas deferens sequence index (VDSI) 3.1. Spatial and temporal variation of imposex intensity which represents the mean score of vas deferens stages of all in- dividuals, following the general scheme for the determination of vas Hexaplex trunculus females exhibiting imposex were observed at all deferens stage (VDS) proposed by Gibbs et al. (1987) and Stroben et al. sampling sites in the year 2015 (Table 1). Even at three sites that were (1992), later modified by Axiak et al. (1995) and Terlizzi et al. (1999). categorized as reference sites, > 90% of collected females had certain Due to the fact that imposex stages are ordinal variables, some authors level of imposex. The lowest VDSI, RPLI and RPSI values were recorded calculate VDSI as a median score (Ho and Leung, 2014). However, since at reference site P3, the only site within marine protected area. At this VDSI is defined as a mean score of different VDS in females and this site VDSI ranged from 0 to 3. Population from sites M1,M2 and M3 definition is used in most of the existing studies (Garaventa et al., 2007; (nautical marinas) exhibited the highest RPSI (6.7–20.4%) and VDSI Lahbib et al., 2008; OSPAR, 2004; WFD-UKTAG, 2014), we followed (> 4.7) values, showing the highest intensity of imposex. Very high the same principle in order to be able to compare data of this study with levels of imposex were also observed at village harbours, especially at data from literature and that given by OSPAR and WFD. location H3. Populations from seasonally used sheltered bays also showed moderate to high imposex level, though to a lesser degree than 2.3. Chemical analyses in marinas and harbours. Sterile females (those with VDS stage 5) were

determined at nine sites. Sterility was absent in village harbour H2, After measurements of imposex stages, the visceral coil and the rest sheltered bay B3 and reference site P3. of the soft body of randomly selected individuals from each location Comparison of imposex indices in populations from different types were pooled into three male and three female samples (5 individuals of sites showed that for RPLI (Fig. 2a) and VDSI (Fig. 2b) statistically

315 A. Erdelez et al. Marine Pollution Bulletin 123 (2017) 313–323

Table 1 Summary of data for Hexaplex trunculus collected in 2015 along the Croatian Adriatic coast (shell height, percentage of females affected by imposex, percentage of sterile females, penis length ± standard deviation, relative penis length index (RPLI), relative penis size index (RPSI), vas deferens sequence index (VDSI) ± standard deviation, butyltin degradation index (BDI) ± standard deviation).

Category Location Sex No. of Shell length Females Sterile Penis length RPLI (%) RPSI (%) VDSI BDI ind. (mm) affected by females (mm) imposex (%) (%) Visceral coil Rest of the soft tissue

Nautical marinas M1 M 20 63.4 ± 3.5 10.6 ± 2.2 11.4 ± 2.4 9.6 ± 3.1 F 20 63.5 ± 4.1 100 70 5.5 ± 1.7 51.6 13.8 4.7 ± 0.4 4.8 ± 2.4 3.6 ± 0.9

M2 M 20 78.3 ± 2.8 14.5 ± 4.1 9.2 ± 3.4 6.5 ± 3.7 F 20 81.0 ± 5.6 100 80 5.9 ± 2.4 40.5 6.7 4.9 ± 0.3 8.0 ± 4.7 4.0 ± 0.6

M3 M 20 62.0 ± 4.3 14.8 ± 2.0 6.0 ± 3.8 4.5 ± 2.6 F 20 63.9 ± 2.9 100 60 8.7 ± 2.3 58.6 20.1 4.8 ± 0.4 4.6 ± 2.6 2.0 ± 0.5

Village harbours H1 M 20 62.3 ± 6.8 12.2 ± 2.8 15.0 ± 2.3 5.1 ± 3.9 F 20 67.1 ± 4.8 100 35 4.9 ± 1.7 39.8 6.3 4.6 ± 0.4 / 8.8 ± 0.9

H2 M 20 68.9 ± 5.6 13.3 ± 2.6 7.7 ± 1.5 6.5 ± 3.7 F 20 69.7 ± 5.3 100 0 6.4 ± 1.5 48.1 11.1 4.1 ± 0.3 8.0 ± 4.7 4.0 ± 0.6

H3 M 20 69.2 ± 6.0 13.6 ± 3.4 / 7.6 ± 1.5 F 20 63.5 ± 4.3 100 20 4.4 ± 1.9 32.6 3.4 4.2 ± 0.6 21.2 ± 20.7 17.1 ± 16.1

Sheltered bays B1 M 15 56.5 ± 6.3 12.2 ± 2.7 / / seasonally F 20 64.5 ± 3.9 100 5 4.1 ± 1.2 33.5 3.8 4.0 ± 0.6 4.9 ± 1.8 8.2 ± 4.5

used B2 M 20 74.5 ± 6.8 13.4 ± 2.4 / 6.5 ± 4.4 F 20 72.0 ± 7.0 100 10 4.1 ± 2.0 31.0 3.0 3.7 ± 1.0 / 2.4 ± 2.4

B3 M 20 56.0 ± 6.2 10.8 ± 2.3 12.0 ± 8.7 39.8 ± 23.6 F 20 67.6 ± 3.7 100 0 3.5 ± 1.1 32.3 3.4 3.9 ± 0.3 / 12.2 ± 10.6

Reference sites P1 M 20 61.3 ± 4.9 11.7 ± 4.9 15.2 ± 7.5 2.3 ± 1.6 very rarely F 20 64.9 ± 2.3 100 5 4.3 ± 1.9 36.9 5.0 3.9 ± 0.7 / 7.5 ± 1.7

used P2 M 20 57.3 ± 7.6 13.1 ± 1.9 / 4.3 ± 1.6 F 20 64.2 ± 4.9 100 5 4.4 ± 1.6 33.6 3.8 4.1 ± 0.3 4.6 ± 2.7 7.4 ± 4.2

P3 M 20 59.3 ± 5.3 14.3 ± 2.3 / / F 20 65.7 ± 5.0 90 0 0.9 ± 0.3 6.6 0.0 2.2 ± 1.2 / /

significant difference can be observed between nautical marinas, village harbour H3 (U = 123, p = 0.04) (Fig. 3b). Females with ventrally split harbours and sites with seasonal or rare boating activity (RPLI: capsule gland (caused by the growth of vas deferens tissue) with its H = 58.05, p < 0.001; VDSI: H = 107.42, p < 0.001), whereas va- lumen open into the mantle cavity are at VDS stage 5 (Terlizzi et al., lues of latter two categories did not statistically differ from each other. 1999), and this females are considered as sterile females (Fig. 4). Imposex indices in 2015 (RPLI and VDSI) between sampling sites Comparison of appearance of sterile females with VDS stage 5 showed within the same boating category were significantly different, except that in 2015 their percentage was much higher in all nautical marinas for sheltered bays and VDSI for nautical marinas. Population from re- (M1,M2 and M3), as well as in harbour H1 (Fig. 3c). At the sheltered bay ference site P3, the only marine protected area site, had significantly B2 and reference sites P1 and P2, sterile females were detected in 2015 lower RPLI values than at any other site (p < 0.007). even though they were not present in 2005. The comparison of imposex indices in H. trunculus collected at 12 locations along the Croatian Adriatic coast in years 2005 and 2015 is presented in Fig. 3. Significantly higher values for RPLI were noted in 3.2. Distributions of butyltins in tissues of Hexaplex trunculus and in the year 2005 at most sampling sites, except at village harbour H1 surface sediments (U = 165, p = 0.34) and reference site P1 (U = 143, p = 0.12) (Fig. 3a). The comparison of VDSI in years 2005 and 2015 (Fig. 3b) Concentrations of butyltins (TBT, DBT, MBT) in tissues of H. trun- showed slightly higher VDSI in 2015 at several locations (M1,M3,H1, culus collected at 12 sites along the Adriatic coast were determined P1 and P2) but only at two sites, nautical marina M3 and reference site separately in the visceral coil (digestive gland and gonads) and in the P2, statistically significant increase in VDSI values is observed rest of the soft body, both in females and males, and the results are (U = 118, p = 0.03, and U = 92, p < 0.001, respectively). Significant presented in the Fig. 5. The total BuT concentration reduction in VDSI values from 2005 to 2015 was observed at village (∑BuT = TBT + DBT + MBT) in visceral coil tissue of females in − nautical marinas (73.3 ± 9.3 ng Sn g 1 d.w.) was 2 times higher than

Fig. 2. Comparison of Hexaplex trunculus imposex indices between different sites categorized by boating activity (M- nautical marinas; H-village harbours; B-sheltered bays; P- reference sites) in the year 2015 according to: (a) relative penis length index (RPLI), (b) vas deferens sequence index (VDSI). Letter marks (a, b, c) denote statistically significant differences between categories (p < 0.05).

316 A. Erdelez et al. Marine Pollution Bulletin 123 (2017) 313–323

Fig. 3. Comparison of Hexaplex trunculus imposex indices between different sites cate- gorized by boating activity (M-nautical marinas; H-village harbours; B-sheltered bays; P- Fig. 5. Concentrations of butyltins (monobutyltin MBT, dibutyltin DBT, tributyltin TBT) reference sites) in the year 2005 and 2015 according to: (a) relative penis length index in diverse female and male tissue types (visceral coil tissue and rest of the body tissue) of (RPLI), (b) vas deferens sequence index (VDSI) + SD, and (c) percentage of sterile fe- Hexaplex trunculus collected at 12 sites of different boating activity (M-nautical marinas; males. Asterisks denote statistically significant differences between years (p < 0.05). H-village harbours; B-sheltered bays; P-reference sites) in the central Croatian Adriatic coast in the year 2015. Mean values are presented with respective standard deviation: mean + SD. Letter marks (a, b, c, d) denote statistically significant differences in ∑BuT between categories (p < 0.05).

− in village harbours (42.5 ± 11.2 ng Sn g 1 d.w.), 8 times higher than − in seasonally used sheltered bays (8.9 ± 2.0 ng Sn g 1 d.w.) and 18 − times higher than in reference sites (4.1 ± 1.1 ng Sn g 1 d.w.). In the rest of the soft body of females ∑BuT in nautical marinas − (95.7 ± 4.6 ng Sn g 1 d.w.) was 2 times higher than in village har- − bours (47.4 ± 7.4 ng Sn g 1 d.w.), 10 times higher than in sheltered − bays (9.5 ± 2.3 ng Sn g 1 d.w.) and 5 times higher than in reference − sites (19.2 ± 5.1 ng Sn g 1 d.w.). The distribution of BuT concentra- tions in males showed similar trends as in females but the differences between categories were less pronounced. The highest concentrations

of ∑BuT were determined in specimens from nautical marina M3 con- cerning all tissue types, although in the males' rest of the body tissue at

this location the concentration was almost the same as at location M2. At several sites with seasonal or rare boating activity ∑BuT was very − low (< 6.0 ng Sn g 1 d.w.) or even below detection limits, mainly in

females' tissue (P1,P2,P3,B2). The measured TBT concentrations were − generally low (< 15 ng Sn g 1 d.w.), except in nautical marinas where Fig. 4. Hexaplex trunculus female in VDS stage 5 indicating split capsule gland (white − 1 error) and vas deferens (black error). Scale bar 5 mm. concentration usually ranged from 5 to 60 ng Sn g d.w., while at sheltered bays and reference sites the values were always near or even

317 A. Erdelez et al. Marine Pollution Bulletin 123 (2017) 313–323 below the detection limit. In most cases TBT had the lowest portion in presented in Fig. 7 showed sharp increase of imposex indices (VDSI) at ∑BuT. Considering all tissue types, MBT was often the predominant very low levels of TBT, reaching highest VDS stages (VDSI > 4) al- − butyltin specie since its proportion ranged from 27 to 87%. In order to ready at TBT concentrations lower than 5 ng Sn g 1. In particular, VDS determine the prevalence of TBT over its degradation products, and stage 5 corresponds to mean TBT and ∑BuT concentrations of 4.3 and − thus the occurrence of recent TBT input into the investigated environ- 41.6 ng Sn g 1 d.w. in the visceral coil tissue, respectively. ments, butyl degradation index (BDI) was calculated (BDI = Statistically significant correlations were established between im- (MBT + DBT) / TBT, as proposed by Diez et al. (2002) for sediments posex indices (RPLI, RPSI, VDSI) and TBT and ∑BuT concentrations in and Sousa et al. (2009) for molluscs) for the population from each lo- both types of tissues (Table 2). The correlations in the visceral coil were cation. BDI values higher than 1 indicate that no recent TBT inputs slightly stronger between imposex indices and ∑BuT concentrations, occur, while BDI values < 1 show that TBT prevail over its degrada- while in the rest of the soft body stronger correlations with TBT were tion products and therefore suggest its recent inputs into the environ- observed. However, imposex indices correlated stronger with ∑BuT ment. The average BDI in our study ranged from 4.6 ± 2.6 (M3)to concentrations in the visceral coil tissue than in the rest of the body 21.2 ± 20.7 (H3) for the visceral coil tissue, while BDI values for rest tissue. Significant correlations were also observed between various of the soft tissue were generally slightly lower ranging from 2.0 ± 0.5 imposex indices (VDSI vs RPLI, VDSI vs RPSI, RPLI vs RPSI; r = 0.67;

(M3) to 39.8 ± 23.6 (B3). r = 0.59; r = 0.93; respectively, p < 0.001). Kruskal-Wallis test confirmed that ∑BuT and TBT were significantly higher in populations from nautical marinas than in those from shel- 4. Discussion tered bays and reference sites considering all tissue types of both fe- males and males, p < 0.05 and p < 0.01, respectively. However, 4.1. Evaluation of TBT pollution and imposex status of H. trunculus at the there was no significant differences nor between nautical marinas and Croatian Adriatic coast village harbours nor between sites with seasonal and rare boating ac- tivity. It is worth mentioning that concentrations of butyltins in dif- In order to evaluate current level of pollution of the Adriatic Coast ferent tissue types of males and females were not significantly different with organotin compounds, degree of imposex was investigated in within the population from the same boating category. The only ex- gastropod H. trunculus collected at 12 locations along the coast. To ception was TBT concentration in females' rest of the body tissue be- elucidate the effects of marine traffic on the ecological status of the tween sites H1 and H2. population, the sampling locations were divided into four groups Spatial variations in butyltins concentrations in sediments collected (nautical marinas, village harbours, sheltered bays and reference sites). at 12 locations in the central Adriatic coast in the year 2015 are shown As most of the studied populations demonstrated high degree of im- in Fig. 6. The results indicated distinct differences among sites of dif- posex, and none was completely free of it, it can be concluded that ferent intensity of boating activities. However, due to the lack of re- incidence of imposex caused by TBT pollution is still widely present plicates at each location those differences were not statistically sig- along the Croatian coast and was strongly related to the intensity of nificant. The ∑BuT concentrations in marinas and harbours were boating activities - it decreased from marinas to the sites with occa- − comparable (104.2 in marinas vs 135.5 ng Sn g 1 in harbours), while sional and rare boating activities (Fig. 2). These results indicate that at sheltered bays and reference sites were below the detection limit. The TBT has been introduced into the marine environment after the ban of highest ∑BuT concentration was found in the village harbour H2 TBT based antifouling paints in 2006. The study on organotins in sea- − (299.1 ng Sn g 1), and the lowest in sediment sample from the nautical water and mussels Mytilus galloprovincialis conducted at Croatian coast −1 marina M2 (81.9 ng Sn g ). The concentrations of TBT followed the during 2009–2010 also demonstrated inputs of TBT after its prohibition same distribution pattern as BuT. Calculated BDI values were 1.9, 2.2 (Furdek et al., 2012). and 2.1 for sites M1,M2 and H2, respectively, while they were < 1 for Notable TBT concentrations are still observed in various marine sites M3 (0.9) and H3 (0.7). environments even years after the total ban of its use in antifouling paints (e.g. Kim et al., 2014; Pougnet et al., 2014), often related to resuspension of contaminated sediments and/or possible new inputs 3.3. Relation between imposex indices and butyltins concentration in (e.g. Langston et al., 2015; Suzdalev et al., 2015). However, the reg- Hexaplex trunculus ulatory restrictions resulted in recovery of many gastropod populations along the European Atlantic coast, including species Nucella lapillus, Comparison of TBT and ∑BuT tissue concentrations and VDS stages Tritia reticulata, nitidus and Stramonita haemastoma, which were all previously substantially affected by imposex (e.g. Evans et al., 1996; Birchenough et al., 2002; Huet et al., 2004; Bray et al., 2011; Cuevas et al., 2014; Langston et al., 2015; Laranjeiro et al., 2015; Nicolaus and Barry, 2015), while Guðmundsdóttir et al. (2011) pre- sented data on the increase of imposex levels near few small harbours in Iceland after intervening regulations. One of the aims of this study was to evaluate the post-legislation change in TBT pollution level and the effect that TBT ban had on H. trunculus populations. With this in mind, this study was conducted 10 years after organotins in antifouling coatings were banned in Croatia at the same sampling sites where Prime et al. (2006) conducted their research in 2005. Statistical comparisons of imposex incidence in years 2005 and 2015 (Fig. 3a, b) lead to dissimilar conclusions regarding temporal trend in imposex occurrence in Croatia. RPLI in 2015 was significantly lower for most sampling sites than in 2005, while VDSI values were similar, except at locations M3 and P2 where they were Fig. 6. Concentrations of butyltins (monobutyltin MBT, dibutyltin DBT and tributyltin significantly higher in 2015, and location H3 where was lower. TBT) in sediments collected at 12 sites of different boating activity (M-nautical marinas; Therefore, when comparing imposex levels in years 2005 and 2015 H-village harbours; B-sheltered bays; P-reference sites) in central Croatian Adriatic coast according to RPLI values a declining trend can be observed at all sites, in 2015. while VDSI index suggested that imposex level mainly remained the

318 A. Erdelez et al. Marine Pollution Bulletin 123 (2017) 313–323

Fig. 7. Relation between tributyltin (TBT) (a) and total butytin (∑BuT) concentration and (b) VDSI in the visceral coil (vc) and in the rest of the soft tissue (rt) in Hexaplex trunculus.

Table 2 reduction in TBT pollution was not sufficient to result in a recovery of Spearman rank order correlations between imposex indices and butyltins concentrations H. trunculus population. Although imposex is considered as an irrever- ∑ ff (TBT, BuT) in visceral coil tissue and rest of the soft body tissue. Di erent p-levels are sible phenomenon, when discussing a declining trend of RPLI values it applied. is worth to mention the hypothesis of the limited reversibility of im- VDSI vs RPLI vs RPSI vs VDSI vs RPLI vs RPSI vs posex, in particular FPLI, as suggested by Bryan et al. (1993) and Sousa ∑BuT ∑BuT ∑BuT TBT TBT TBT et al. (2005, 2007). However, this should be further investigated and

⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ verify whether this applies to VDSI also. Visceral coil 0.65 0.70 0.59 0.58 0.66 0.59 tissue In order to elucidate the level of TBT pollution of the marine en- ⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ Rest of the 0.48 0.52 0.48 0.58 0.64 0.60 vironment in the central Adriatic, in addition to imposex indices, the soft BuTs concentrations in the tissue of H. trunculus (visceral coil and the tissue rest of the soft tissue) and surface sediments were determined. The

⁎⁎⁎ ∑BuT, regardless of sex and tissue type, ranged from 0 to p < 0.001. −1 ⁎⁎ p < 0.01. 348.3 ng Sn g d.w., with the highest concentrations found in mar- inas, somewhat less in ports, and the lowest concentrations (often near same over a 10-year period. It is therefore questionable which imposex or below the detection limits) in the sheltered bays and reference sites ff parameter is more reliable for assessing the contamination and ecolo- (Fig. 5). Even though the di erence in BuTs concentrations between fi gical status of the marine environment regarding TBT pollution, espe- those categories was not statistically signi cant, this study nevertheless cially taking into account that both of them correlate well with TBT demonstrated that BuTs body burden in H. trunculus is related to the concentration in H. trunculus (Table 2). Stroben et al. (1992) suggested degree of boating activity, the same as found for imposex indices. It that VDSI is the most reliable indicator for imposex in T. reticulata be- should be emphasized that imposex represents a dose-dependent but cause of the seasonal changes in the penis length, while Pelizzato et al. irreversible response to the total TBT accumulated throughout the life (2004) preferred female penis size (RPSI) when comparing con- of an individual (Laranjeiro et al., 2015; Axiak et al., 2003), and taminated sites because of the larger range of variations in RPSI than in therefore the degree of imposex determined in this study could be VDSI. Whereas both of the studies were conducted in the same time of partly due to the past exposure to TBT. Since the depuration time of the year (May–July; fresh and not narcotized animals), penis length TBT in molluscs is usually described by the half-life of one to several indicators are considered comparable, in particular because RPSI and months (Gomez-Ariza et al., 1999; Sousa et al., 2009), the determined RPLI were below 100%. However, according to OSPAR, WFD and many BuTs concentrations in the tissue of H. trunculus undeniably indicated other studies, the key imposex parameter to be monitored for evalua- that the investigated populations were exposed to TBT pollution over tion of TBT pollution is VDSI because of its biological implication as it the last few months. However, despite the possible time discrepancy reveals the reproductive capability of population (OSPAR, 2004; WFD- between imposex development and TBT accumulation, rather weak but fi UKTAG, 2014; Oehlmann et al., 1998b; Laranjeiro et al., 2015). still statistically signi cant correlations were found between imposex The conclusion based on determined VDSI values, that no declining indices (VDSI, RPLI and RPSI) and TBT concentration (Table 2). This is trend in imposex levels is observed, is further supported by the per- in agreement with observation of many studies that the intensity of centage of sterile females (those with VDS stage 5) which was higher at imposex in gastropods is related to the levels of accumulated TBT most locations in 2015 (7 out of 12; Fig. 3c). Having in mind that VDS (Axiak et al., 1995; Pellizatto et al., 2004; Ho and Leung, 2014). An provides information on reproductive capability of a population and even stronger correlation was observed between imposex indices and ∑ risk of its extinction, we can undoubtedly conclude that the prohibition BuT concentrations, particularly in the visceral coil. This is due to the ∑ fl of TBT based antifouling paints in Croatia did not result in the recovery fact that BuT concentrations more reliably re ect the total uptake of of H. trunculus population. However, it should be mentioned a model the TBT. Tributyltin is bioaccumulated through feeding and initially experiment on the imposex development in H. trunculus exposed to accumulates in the visceral coil where it undergoes dealkylation to DBT − different levels of TBT concentration (5 and 50 ng l 1) described by and MBT through the P-450 dependent mixed-function oxidase system. Abidli et al. (2009). This study demonstrated that female penis length Later, all three are transported to other body organs (Axiak et al., 1995; was significantly higher in individuals exposed to higher TBT con- Pellizzato et al., 2004). The results of our study are in agreement with fi centration after 6 months of exposure, while VDSI was the same for this observations as the statistically signi cant correlations were found both exposure levels. According to this research, penis length in H. between all TBT, DBT and MBT in the visceral coil (Spearman rank trunculus better indicates the level of TBT concentrations in the en- correlation; TBT vs DBT: r = 0.76; DBT vs MBT: r = 0.85; TBT vs MBT: vironment since the highest VDS stage develops even at lower con- r = 0.70; p < 0.05), as well as between TBT, DBT and MBT con- centration. Therefore, the difference between imposex indices observed centrations in the visceral coil (vc) and in the rest of the soft tissue (rt) in our study could be interpreted as follows; RPLI decline from 2005 to (Spearman rank correlation; TBTvc vs TBTrt: r = 0.79; DBTvc vs DBTrt: 2015 suggests that the level of TBT pollution in the marine environment r = 0.69; MBTvc vs MBTrt: r = 0.52; p < 0.05). Many published pa- ff decreased along this period, while VDSI demonstrates that apparent pers reported di erent distribution patterns of BuTs in the H. trunculus

319 A. Erdelez et al. Marine Pollution Bulletin 123 (2017) 313–323 tissue. Some authors stated significantly higher concentrations in the concentrations in sediments at sites categorized as nautical marinas and − visceral coil tissue (Axiak et al., 1995, 2003; Pellizzato et al., 2004; harbours were 302.3 and 296.2 ng Sn g 1, and in those categorized as − Garaventa et al., 2006), while others found higher load of organotins in small ports and reference sites were 63.1 and 53.1 ng Sn g 1, respec- the rest of the body tissue (Garaventa et al., 2007). In our study sig- tively. Due to lack of replicate samples in our study, and the fact that nificant difference in the accumulation of BuTs in the visceral coil and these two studies were not conducted at the same sites, the reliable in the rest of the body tissue was not observed. Distribution of BuTs in comparison of the results is not possible, but nevertheless a slight de- the investigated populations showed the highest proportion of MBT and crease in ∑BuT concentrations in marinas and harbours, as well as in the lowest of TBT (MBT > DBT > TBT) thus showing that TBT is ports and reference sites, could be observed. However, taking into ac- faster metabolized to DBT and MBT and eliminated from the organism, count that average sedimentation rate in the coastal area is around than is being accumulated from the environment. This, together with 0.5–1 cm per year, it is difficult to determine temporal changes in the BDI values calculated for H. trunculus tissue which were higher than surface sediments (0–5 cm) over the period shorter than 5 years. 1 at all locations (Table 1), strongly suggests that there was no recent After all, on the basis of the presented results, both BuTs con- inputs of TBT into the investigated environments, at least not in the last centrations in H. trunculus and sediments, we can conclude that TBT several months. However, it should be mentioned that high BDI values pollution rather decreased during the last 10 years in Croatia, con- in gatropods could be also explained by raised availability of MBT and firming the effectiveness of the ban. However, the decreased TBT input DBT as a result of TBT degradation in surface sediments (Ruiz et al., into the marine environment did not result in the decrease of imposex 2008). This is supported by the statistically significant correlations that level in H. trunculus and the recovery of this population because the was found in our study between MBT and DBT concentrations in sedi- concentrations of TBT in the environment are still higher than those ments and H. trunculus tissue (Spearman rank correlation; DBTsed vs that cause advanced stages of imposex in the local gastropod popula- DBTvc: r = 0.62; p < 0,001; DBTsed vs DBTrt: r = 0.62; p < 0.05; tions. MBTsed vs MBTvc: r = 0.47; p < 0,05). Correlation between TBT in Among all gastropod species investigated so far, H. trunculus is sediment and tissue was not found. considered as the most sensitive with respect to its biological response The observation of BuTs tissue concentrations in H. trunculus over to TBT (Axiak et al., 1995; Abidli et al., 2009). This is further supported investigated time period was not possible because our study re- in our study as the highest VDS stage was reached at TBT and ∑BuT − presented the first data on BuTs tissue burden in populations of H. concentrations below 5 and 50 ng Sn g 1 d.w., respectively. These va- trunculus for the central part of the Croatian Adriatic. When comparing lues are significantly lower than those reported by Pelizzato et al. the mean BuTs concentrations in our study to those of previous study on (2004) and Garaventa et al. (2007) who reported that VDS stage 5 − H. trunculus in the Northern Adriatic for the period 2002–2003 (∑BuT corresponds to approximately 90 and 235 ng Sn g 1, respectively. This − ranged between 192 and 4233 ng Sn g 1 d.w., Garaventa et al., 2007), corroborates the conclusion made by Garaventa et al. (2007), that it can be noticed that concentrations are lower in our study conducted correlation analyses of VDSI and butyltins tissue concentration could be in 2015. They are also lower than TBT concentrations in H. trunculus unreliable at high concentrations because investigated populations from other areas in the Mediterranean in the period 1992–2003 (Axiak usually carried body burdens of butyltins which are above the satura- et al., 1995; Pelizzato et al., 2004). tion limits. Distribution of BuTs concentrations in sediments also showed ex- pected differences between different categories of sites regarding the 4.2. Evaluation of ecological status classes for Water Framework Directive intensity of boating activity. The ∑BuT concentration in marinas and based on imposex levels in Hexaplex trunculus harbours were rather similar (average concentrations are 104.2 and − 135.5 ng Sn g 1 d.w., respectively), while in sediments from sites with According to the EU Water Framework Directive (WFD, Anex V, rare marine traffic (bays and reference sites) BuTs were below the de- 2000/60/EU), one of the biological quality elements for the classifi- tection limits. Tributyltin bioaccumulated in H. trunculus at these sites cation of ecological status of surface waters is composition and abun- − was also low (ranged from 0.9 to 3.3 ng Sn g 1 d.w.), confirming low dance of benthic invertebrate fauna. Although not specifically referred BuTs levels in these environments. It is important to underline that to in the WFD, the imposex assessment could be used as a metric within sample from harbour H1 was gravel-grained sediment which could ex- the benthic invertebrate biological quality element since it represents a plain low level of BuTs determined at this site since this type of sedi- specificeffect of a particular contaminant on a sensitive indicator ment has very low ability to adsorb TBT (Furdek et al., 2016). Many species (WFD-UKTAG, 2014; Laranjeiro et al., 2015; Ruiz et al., 2017). published papers demonstrated that some sediments, as a consequence To perform evaluation of the ecological status according to WFD, each of high affinity of TBT for the adsorption onto particulate matter and biological quality element should be evaluated by comparing the very slow TBT degradation in sediments (half lives ranged from several measured conditions (the observed value) against that described for years to decades), represent a long-term storage of TBT (Omae, 2003). reference conditions (minimally disturbed), and this is reported as an Due to resuspension of contaminated surface sediments, its desorption Ecological Quality Ratio (EQR). The evaluation of the ecological status back into the water column may occur, which results in a new threats to of the environment is performed by determining the EQR values which biota (e.g. Ruiz et al., 2008; Kim et al., 2014; Pougnet et al., 2014; define five ecological status classes (High, Good, Moderate, Poor and Langston et al., 2015; Suzdalev et al., 2015). Although H. trunculus lives Bad), in accordance to the effects of anthropogenic disturbance in the in the direct contact with sediments, the significant correlations be- biological community. The detailed description of each ecological tween TBT concentration in sediments and tissue or imposex indices status class, based on the information from OSPAR (2004) and WFD were not confirmed in our study. A reason for absence of such corre- (Annex V, 2000/60/EC), is given by Laranjeiro et al. (2015). When lations could be the influence of the sediment characteristics on the TBT imposex is used as a biological quality element, the EQR values should adsorption. Some of the sediment samples (from locations M2,H1,B1, be calculated by converting the determined VDSI values into EQR, ac- B2,B3,P1,P2) were mainly sandy and gravel sediments which usually cording to Eq. (1) (WFD-UKTAG, 2014; Laranjeiro et al., 2015): have low amount of organic matter and thus low capacity to adsorb (M− O) TBT. On the contrary, sediments containing more fine particles have EQR = M (1) higher affinity for adsorption of organic matter and TBT, resulting in much slower TBT degradation in these sediments which consequently (M is maximum VDSI that a population may attain, and O is the represent a long-term source of pollution (Furdek et al., 2016). The observed VDSI). recent study conducted in the period 2009–2013 along the Croatian Imposex in species N. lapillus is already proposed as a tool for WDF Adriatic coast (Furdek, 2015) demonstrated that mean ∑BuT ecological status assessment of coastal waters in the UK (WFD-UKTAG,

320 A. Erdelez et al. Marine Pollution Bulletin 123 (2017) 313–323

2014), while Laranjeiro et al. (2015) proposed ecological status classes environmental data relating to the of imposex levels in H. trunculus for N. lapillus, T. reticulata and Littorina littorea. Because various gas- (VDSI < 1), the EQR value and the corresponding VDSI are difficult to tropod species have different sensitivity to TBT and a different path of establish for this category. For this category Laranjeiro et al. (2015) imposex development (Axiak et al., 1995), the proposed ecological adopted the VDSI value (VDSI < 0.3) for N. lapillus and T. reticulata set status classes for those species, and the corresponding EQR values are by OSPAR (2004). not applicable for H. trunculus. Furthermore, none of these species is The Good status class is defined by the absence of the risk for po- widely distributed in the Mediterranean and therefore cannot be used in pulation extinction, which, when using imposex as s parameter for WFD monitoring programme in this area. Due to its ubiquity in the ecological status assessment, requires at least absence of sterile females. Mediterranean, and a very sensitive response to TBT contamination, H. The data available for H. trunculus in the Mediterranean, presented in trunculus could be the principal bioindicator for the purpose of WFD Fig. 8, indicate that sterile females appear in populations with monitoring programme in the Mediterranean. For this to be realized, VDSI > 3.0 (confidence interval p = 0.95). However, some sterile fe- the appropriate EQR values for the ecological status classes should be males were also observed in a few populations with 1.8 < VDSI < defined. Therefore, one of the aims of this work was to propose such 3.0. Therefore, the Good status class, which indicates lack of sterile values on the base of the available data for imposex indices (VDSI and females in the population, should be defined with the interval percentage of sterile females (%S)) in H. trunculus adopted from the VDSI < 1.8. The Moderate status implies low risk of population ex- literature (this study; Axiak et al., 1995; Terlizzi et al., 1998; Rilov tinction, meaning that sterile females may appear, while Poor status et al., 2000; Prime et al., 2006; Garaventa et al., 2007; Stagličić et al., corresponds to a major deviation in population abundance caused by 2008; Lahbib et al., 2011), following the approach described in the high female sterility. According to Laranjeiro et al. (2015), the VDSI work of Laranjeiro et al. (2015). Considering that the principal aim of boundary between the Moderate and the Poor class is related to dis- WFD is to preserve diversity and abundance of each population, the tinction between low and high risk of population extinction and occurrence of female sterility can be one of the factors in the evaluation therefore corresponds to the incidence of > 50% of sterile females. The of boundaries of the ecological status classes regarding imposex. It is relationship between VDSI and percentage of sterile female presented in also very important to notice that endocrine modulators can have ex- Fig. 8 demonstrated that 50% of sterile females corresponds to ap- tensive consequences for the population fitness beyond the merely in- proximately VDSI 4.5 so this should be the proposed value for the ducing sterility in subset of individuals. TBT has been previously shown boundary between Moderate and Poor class. The Bad status refers to to affect traits related to Darwinian fitness, such as growth, mortality population extinction and the absence of specimen individuals due to and fecundity of H. trunculus already at environmentally relevant con- extremely high TBT pollution. centrations (Abidli et al., 2009; Lahbib et al., 2012). Thus further stu- The VDSI values that were proposed as the classes boundaries were dies should aim to reveal the extact impact on the population fitness at converted into EQR boundaries for ecological status classes using Eq. TBT concentrations lower than those causing female sterility. (1), and are presented in Table 3. EQR boundaries for N. lapillus and T. The relationship between the percentage of sterile females and the reticulata (Laranjeiro et al., 2015), the gastropods most often used for corresponding VDSI determined in 105 different populations from 9 imposex monitoring at the European Atlantic coast, are shown for studies is presented in Fig. 8 (this study-12 populations; Axiak et al. comparison. However, the proposed boundaries represent only the in- (1995) – 9 populations; Terlizzi et al. (1998) – 15 populations; Rilov itial step in establishing the EQR classes for imposex in H. trunculus and et al. (2000) – 16 populations; Prime et al. (2006) – 12 populations; should be further validated and confirmed on the basis of a much larger Garaventa et al. (2007) – 8 populations; Stagličić et al. (2008) – 7 po- set of data. Further studies of imposex appearance at very low TBT pulations; Lahbib et al. (2008) – 20 populations; Lahbib et al. (2011) – 6 levels when VDSI is ≤1 are particularly needed. populations). According to the proposed EQR classes for H. trunculus presented High ecological status class corresponds to undisturbed environ- above, the ecological status of the marine environment in the Central mental conditions and is therefore achieved if no (or minor) anthro- Adriatic for the year 2015 was evaluated. The results indicated that pogenic alteration occurs. This status equals to extremely low TBT none of the surveyed sites reached the classification of Good status (not concentrations in water, usually close to zero and often below detection even the protected marine areas). All investigated nautical marinas (M1, limits of the most sensitive analytical methods. Due to the lack of M2,M3) and harbour H1 are evaluated as the environments of Poor ecological status, while other village harbours, sheltered bays and re- ference sites are categorized as Moderate class environments. Therefore, the data presented in this study showed that WFD target for achieving the Good ecological status by the year 2015 is not reached. If the same classification of ecological status is applied using the data from 2005, it can be concluded that the legislative restriction of TBT based antifouling paints did not result in the recovery of the in- vestigated environments during the last 10 years.

5. Conclusion

The prohibition on use of TBT in antifouling coatings resulted in a decline of BuT levels in the central Croatian Adriatic. However, all in- vestigated populations of Hexaplex trunculus are still highly affected by imposex, meaning that the prohibition of TBT based antifouling paints did not result in the recovery of this population. The level of imposex in the investigated populations correlates with TBT and total butyltins body burden. Ecological Quality Ratio (EQR) boundaries for the po- tential use of imposex in H. trunculus as a tool for the ecological status Fig. 8. The relationship between the percentage of sterile females (%S) and VDSI de- assessment of coastal waters regarding TBT pollution, in the frame of termined in 105 different populations of Hexaplex trunculus from the Mediterranean: WFD, were proposed. According to proposed EQR boundaries, WFD polynomial fit equation VDSI = 3.0258 + 0.0501 ∗ x − 0.0003 ∗ x2 with confidence in- target for achieving the Good ecological status by the year 2015 is not terval (conf.int.) p = 0.95. reached at Eastern Adriatic Coast.

321 A. Erdelez et al. Marine Pollution Bulletin 123 (2017) 313–323

Table 3 Proposed EQR values defined for use of imposex as biological quality element for the ecological status assessment in the frame of WFD.

Ecological status class Nucella lapillus (Laranjeiro et al., 2015) Tritia reticulata (Laranjeiro et al., 2015) Hexaplex trunculus (proposed in this study)

Proposed EQR classes Corresponding VDSI Proposed EQR classes Corresponding VDSI Proposed EQR classes Corresponding VDSI

High 1.00–0.95 0.0–0.3 1.00–0.93 0.0–0.28 na na Good 0.95–0.50 0.3–3.0 0.93–0.80 0.28–0.8 0.64 < 1.8 Moderate 0.50–0.25 3.0–4.5 0.80–0.40 0.8–2.4 0.64–0.10 1.8–4.5 Poor 0.25–0.00 4.5–6.0 0.40–0.00 2.4–4.0 0.10–0.00 4.5–5.0 Bad Population absence Population absence na – not applicable in this study.

Acknowledgement degradation study using species-specific stable isotopic tracers. J. Hazard. Mater. 307, 263–273. Garaventa, F., Pellizzato, F., Faimali, M., Terlizzi, A., Medakovic, D., Geraci, S., Pavoni, This work has been supported by the Croatian Science Foundation B., 2006. Imposex in Hexaplex trunculus at some sites on the North Mediterranean under the project number 7555. Coast as a base-line for future evaluation of the effectiveness of the total ban on organotin based antifouling paints. Hydrobiologia 555, 281–287. Garaventa, F., Centanni, E., Pellizzato, F., Faimali, M., Terlizzi, A., Pavoni, B., 2007. References Imposex and accumulation of organotin compounds in populations of Hexaplex trunculus (, ) from the Lagoon of Venice (Italy) and Istrian coast (Croatia). Mar. Pollut. Bull. 54, 615–622. Abidli, S., Lahbib, Y., El Menif, N.T., 2009. Effects of TBT on the imposex development, Gibbs, P.E., Bryan, G.W., Pascoe, P.L., Burt, G.R., 1987. The use of the dog-whelk, Nucella reproduction and mortality in Hexaplex trunculus (Gastropoda: Muricidae). J. Mar. lapillus, as an indicator of trybutyltin (TBT) contamination. J. Mar. Biol. Assoc. UK 67, Biol. Assoc. U. K. 89, 139–146. 507–523. Abidli, S., Lahbib, Y., Gonzalez, P.R., Alonso, J.I.G., El Menif, N.T., 2013. Imposex and Gibbs, P.E., Pascoe, P.L., Burt, G.R., 1988. Sex change in the female dog-whelk, Nucella butyltin burden in (, Gastropoda) and sediment from the lapillus, induced by tributyltin from antifouling paints. J. Mar. Biol. Assoc. UK 68, Tunisian coast. Hydrobiologia 714, 13–24. 715–731. Anastasiou, T., Chatzinikolaou, E., Mandalakis, M., Arvanitidis, C., 2015. Imposex status Gomez-Ariza, J.L., Morales, E., Giraldez, I., 1999. Uptake and elimination of tributyltin in and organotin concentrations in Hexaplex trunculus (Linnaeus, 1758) from the port of clams, Venerupis decussata. Mar. Environ. Res. 47, 399–413. Cagliari (Sardinia). In: 11th Panhellenic Symposium on Oceanography and Fisheries, Graceli, J.B., Cavati Sena, G., Iguatemy Lopes, P.F., Carvalho Zamprogno, G., da Costa, Mytilene, Lesvos Island, Greece. M.B., Locateli Godoi, A.F., Moscardi dos Santos, D., Rodrigues de Marchi, M.R., dos Axiak, V., Vella, A.J., Micallef, D., Chircop, P., 1995. Imposex in Hexaplex trunculus Santos, Fernandez, 2013. Organotins: a review of their reproductive toxicity, bio- (Gastropoda: Muricidae): the first results from biomonitoring of tributyltin con- chemistry, and environmental fate. Reprod. Toxicol. 36, 40–52. tamination in the Mediterranean. Mar. Biol. 121, 685–691. Guðmundsdóttir, L.Ó., Ho, K.K.Y., Lam, J.C.W., Svavarsson, J., Leung, K.M.Y., 2011. Axiak, V., Miccalef, D., Muscat, J., Vella, A., Mintoff, B., 2003. Imposex as a biomoni- Long-term temporal trends (1992–2008) of imposex status associated with organotin toring tool for marine pollution by tributyltin: some further observations. Environ. contamination in the dogwhelk Nucella lapillus along the Icelandic coast. Mar. Pollut. Int. 28, 743–749. Bull. 63, 500–507. Birchenough, A.C., Evans, S.M., Moss, C., Welch, R., 2002. Re-colonization and recovery Ho, K.K.Y., Leung, K.M.Y., 2014. Spatio-temporal comparisons of imposex status and of populations of dogwhelks Nucella lapillus (L.) on shores formerly subject to severe tissue organotin concentration in the whelk Reishia clavigera collected along the TBT contamination. Mar. Pollut. Bull. 44, 652–659. coasts of Dapeng Bay and Daya Bay, Shenzhen, China. Mar. Pollut. Bull. 85, 254–260. Boulajfene, W., Boukhicha, J., Deidun, A., Berto, D., Romeo, T., Ben Hassine, O.K., Tlig- Horiguchi, T., 2017. Contamination by organotins and its population-level effects in- Zouari, S., 2015. Biomonitoring of the environmental contamination by organotins in volved by imposex in prosobranch gastropods. In: Horiguchi, T. (Ed.), Biological the Gulf of Tunis: occurrence of imposex in Stramonita haemastoma (Linnaeus, 1767). Effects by Organotins. Springer Japan, Tokyo, pp. 73–99. Mar. Freshw. Res. 66 (9), 778–785. http://dx.doi.org/10.1071/MF13301. Huet, M., Paulet, Y.M., Clavier, J., 2004. Imposex in Nucella lapillus: a ten year survey in Bray, S., McVean, E.C., Nelson, A., Herbert, R.J.H., Hawkins, S.J., Hudson, M.D., 2011. NW Brittany. Mar. Ecol. Prog. Ser. 270, 153–161. The regional recovery of Nucella lapillus populations from marine pollution, fa- Kim, N.A., Shim, W.J., Yim, U.H., Hong, S.H., Ha, S.Y., Han, G.M., Shin, K.H., 2014. cilitated by man-made structures. J. Mar. Biol. Assoc. UK 92, 1585–1594. http://dx. Assessment of TBT and organic booster biocide contamination in seawater from doi.org/10.1017/S0025315411001317. coastal areas of South Korea. Mar. Pollut. Bull. 78, 201–208. Bryan, G.W., Gibbs, P.E., Burt, G.R., Hummerstone, L.G., 1987. The effect of tributyltin Lahbib, Y., Abidli, S., Le Pennec, M., Flower, R., El Menif, N.T., 2007. Morphological (TBT) accumulation on adult dogwhelks, Nucella lapillus: long-term field and la- expression and different stages of imposex in Hexaplex trunculus (: boratory experiments. J. Mar. Biol. Assoc. UK 67, 525–544. Muricidae) from Tunisian coasts. Cah. Biol. Mar. 48, 315–326. Bryan, G.W., Burt, G.R., Gibbs, P.E., Pascoe, P.L., 1993. Nassarius reticulatus (: Lahbib, Y., Abidli, S., El Menif, N.T., 2008. Imposex level and penis malformation in Gastropoda) as an indicator of tributyltin pollution before and after TBT restrictions. Hexaplex trunculuc from the Tunisian coast. Am. Malacol. Bull. 24 (1), 79–89. J. Mar. Biol. Assoc. UK 73, 913–929. Lahbib, Y., Abidli, S., Chiffoleau, J.F., Averty, B., El Menif, N.T., 2009. First record of Carić, H., Klobučar, G., Štambuk, A., 2016. Ecotoxicological risk assessment of antifouling butyltin burden and imposex status in Hexaplex trunculus (L.) along the Tunisian emissions in a cruise ship port. J. Clean. Prod. 121, 159–168. coast. J. Environ. Monit. 11, 1253–1258. Chiavarini, S., Massanisso, P., Nicolai, P., Nobili, C., Morabito, R., 2003. Butyltins con- Lahbib, Y., Abidli, S., Chiffoleau, J.F., Averty, B., El Menif, N.T., 2010. Imposex and centration levels and imposex occurence in snails from the Sicilian coasts (Italy). butyltin concentrations in snails from the lagoon of Bizerta (Northern Tunisia). Mar. Chemosphere 50, 311–319. Biol. Res. 6, 600–607. Commission Directive 2000/62/EC, 2000. EU Water Framework Directive. Off. J. Eur. Lahbib, Y., Abidli, S., Rodriguez Gonzalez, P., Alonso, J.I.G., El Menif, N.T., 2011. Communities 327. Monitoring of organotin pollution in Bizerta Channel (Northern Tunisia): temporal Cuevas, N., Ignacio, J., Alonso, G., Laretta, J., Germán Rodríguez, J., Sariego, C., Zorita, trend from 2002 to 2010. Bull. Environ. Contam. Toxicol. 86, 531–534. I., 2014. Monitoring the effectiveness of the European tributyltin regulation oh the Lahbib, Y., Abidli, S., El Menif, N.T., 2012. TBT pollution in Tunisian coastal lagoons as Basque coast (northern ) by assessing imposex in two gastropod species indicated by imposex in Hexaplex trunculus (Gastropoda: Muricidae). Transit. Water. (Nassarius reticulatus and Nassarius nitidus). Chem. Ecol. 30, 636–642. Bull. 6, 17–24. Decision 2455/2001/EC of the European Parliament and of the council establishing the Langston, W.J., Pope, N.D., Davey, M., Langston, K.M., O'Hara, S.C.M., Gibbs, P.E., list of priority substances in the field of water policy. Off. J. Eur. Communities 331. Pascoe, P.L., 2015. Recovery from TBT pollution in English Channel environments: a Diez, S., Abalos, M., Bayona, J.M., 2002. Organotin contamination in sediments from the problem solved? Mar. Pollut. Bull. 95, 551–564. Western Mediterranean enclosures following 10 years of TBT regulation. Water Res. Laranjeiro, F., Sánchez-Marín, P., Galante-Oliveira, S., Barroso, C., 2015. Tributyltin 36, 905–918. pollution biomonitoring under the Water Framework Directive: proposal of a multi- Evans, S.M., Evans, P.M., Leksono, T., 1996. Widespread recovery of dogwhelks, Nucella species tool to assess the ecological quality status of EU water bodies. Ecol. Indic. 57, lapillus (L.), from tributyltin contamination in the North sea and Clyde sea. Mar. 525–535. Pollut. Bull. 32, 263–269. Milivojevič Nemanič, T., Milačič, R., Ščančar, J., 2009. A survey of organotin compounds Furdek, M., 2015. Distribution and Behaviour of Organotin Compounds in the Coastal in the northern Adriatic Sea. Water Air Soil Pollut. 196, 211–224. Marine Environment of the Eastern Adriatic. (Doctoral thesis) University of Zagreb, Nicolaus, E.E.M., Barry, J., 2015. Imposex in the dogwhelk (Nucella lapillus): 22 years Faculty of Science. monitoring around England and Wales. Environ. Monit. Assess. 187, 736. Furdek, M., Vahčić, M., Ščančar, J., Milačić, R., Kniewald, G., Mikac, N., 2012. Organotin Oehlmann, J., Bauer, B., Minchin, D., Schulte-Oehlmann, U., Fioroni, P., Markert, B., compounds in seawater and Mytillus galloproviciialis mussels along the Croatian 1998a. Imposex in Nucella lapillus and intersex in Littorina littorea: interspecific Adriatic Coast. Mar. Pollut. Bull. 64, 189–199. comparison of two TBT-induced effects and their geographical uniformity. Furdek, M., Mikac, N., Bueno, M., Tessier, E., Cavalheiro, J., Monperrus, M., 2016. Hydrobiologia 378, 199–213. Organotin persistence in contaminated marine sediments and porewaters: in situ Oehlmann, J., Stroben, E., Schulte-Oehlmann, U., Bauer, B., 1998b. Imposex development

322 A. Erdelez et al. Marine Pollution Bulletin 123 (2017) 313–323

in response to TBT pollution in Hinia incrassata (Ström, 1768) (Prosobranchia, Nassarius reticulatus (L.) along the Portuguese coast. Appl. Organomet. Chem. 19, Stenoglossa). Aquat. Toxicol. 43, 239–260. 315–323. Omae, I., 2003. Organotin antifouling paints and their alternatives. Appl. Organomet. Sousa, A., Matsudaira, C., Takahashi, S., Tanabe, S., Barroso, C., 2007. Integrative as- Chem. 17, 81–105. sessment of organotin contamination in a southern European estuarine system (Ria de OSPAR, 2004. OSPAR/ICES Workshop on the evaluation and update of background re- Aveiro, NW Portugal): tracking temporal trends in order to evaluate the effectiveness ference concentrations (B/RCs) and ecotoxicological assessment criteria (EACs) and of the EU ban. Mar. Pollut. Bull. 54 (10), 1645–1653. how these assessment tools should be used in assessing contaminants in water, se- Sousa, A., Laranjeiro, F., Takahashi, S., Tanabe, S., Barroso, C.M., 2009. Imposex and diment and biota. In: Final Report. OSPAR commission. organotin prevalence in a European post-legislative scenario: temporal trends from Pelizzato, F., Centanni, E., Marin, M.G., Moschino, V., Pavoni, B., 2004. Concentrations of 2003 to 2008. Chemosphere 77 (4), 566–573. organotin compounds and imposex in the gastropod Hexaplex trunculus from the Stagličić, N., Prime, M., Zoko, M., Erak, Ž., Brajčić, D., Blažević, D., Madirazza, K., Jelić, Lagoon of Venice. Sci. Total Environ. 332, 89–100. K., Peharda, M., 2008. Imposex incidence in Hexaplex trunculus from Kaštela Bay, Pougnet, F., Schäfer, J., Dutruch, L., Garnier, C., Tessier, E., Dang, D.H., Lanceleur, L., Adriatic Sea. Acta Adriat. 49, 159–164. Mullot, J.U., Lenoble, V., Blanc, G., 2014. Sources and historical record of tin and Stroben, E., Oehlmann, J., Fiorini, P., 1992. The morphological expression of imposex in butyltin species in a Mediterranean bay (Toulon Bay, France). Environ. Sci. Pollut. Hinia reticulata (Gastropoda: Buccinidae): a potential biological indicator of tribytiltin Res. 21, 6640–6651. pollution. Mar. Biol. 113, 625–636. Prime, M., Peharda, M., Jelić, K., Mladineo, I., Richardson, C.A., 2006. The occurrence of Suzdalev, S., Gulbinskas, S., Blažauskas, N., 2015. Distribution of tributyltin in surface imposex in Hexaplex trunculus from the Croatian Adriatic. Mar. Pollut. Bull. 52, sediments from transitional marine-lagoon system of the south-eastern Baltic Sea, 810–812. Lithuania. Environ. Sci. Pollut. Res. 22, 2634–2642. http://dx.doi.org/10.1007/ Regulation (EC) No 782/2003, 2003. of the European Parliament and of the Council of 14 s11356-014-3521-4. April 2003 on the prohibition of organotin compounds on ships. Off. J. L. 115, Terlizzi, A., Geraci, S., Minganti, V., 1998. Tributyltin (TBT) pollution in the coastal 0001–0011 09/05/. waters of Italy as indicated by imposex in Hexaplex trunculus (Gastropoda, Rilov, G., Avital, A., Evans, S.M., Benayahu, Y., 2000. Unregulated use of TBT-based Muricidae). Mar. Pollut. Bull. 36, 749–752. antifouling paints in Israel (eastern Mediterranean): high contamination and imposex Terlizzi, A., Geraci, S., Gibbs, P.E., 1999. Tributyltin (TBT) induced imposex in the levels in two species of marine gastropods. Mar. Ecol. Prog. Ser. 192, 229–238. neogastropod Hexaplex trunculus in Italian coastal waters: morphological aspects and Ruiz, J.M., Barreiro, R., Couceiro, L., Quintela, M., 2008. Decreased TBT pollution and ecological implications. Ital. J. Zool. 66, 141–146. changing bioaccumulation pattern in gastropods imply butyltin desorption from se- Terlizzi, A., Delos, A.L., Garaventa, F., Faimali, M., Geraci, S., 2004. Limited effectivness diments. Chemosphere 73, 1253–1257. of marine protected areas: imposex in Hexaplex trunculus (Gatropoda, Muricidae) Ruiz, J.M., Carro, B., Albaina, N., Couceiro, L., Míguez, A., Quintela, M., Barreiro, R., populations from Italian marine reserves. Mar. Pollut. Bull. 48, 188–192. 2017. Bi-species imposex monitoring in Galicia (NW Spain) shows contrasting Vasconcelos, P., Gaspar, M.B., Castro, M., 2006a. Imposex in Hexaplex (Trunculariopsis) achievement of the OSPAR Ecological Quality Objective for TBT. Mar. Pollut. Bull. trunculus (Gastropoda: Muricidae) from the Ria Formosa lagoon (Algarve coast – 114 (2), 715–723. southern Portugal). Mar. Pollut. Bull. 52, 337–341. Smith, B.S., 1971. Sexuality in the American mud snail, Nassarius obsoletus (Say). Proc. Vasconcelos, P., Gaspar, M.B., Pereira, A.M., Castro, M., 2006b. Growth rate estimation of Malacol. Soc. London 39, 377–378. Hexaplex (Trunculariopsis) trunculus (Gastropoda: Muricidae) based on mark/re- Smith, B.S., 1981. Tributyltin compounds induce male characteristics on female mud capture experiments in the Ria Formosa lagoon (Algarve coast, southern Portugal). J. snails Nassarius obsoletus = Ilyanassa obsolete. J. Appl. Toxicol. 1, 141–144. Shellfish Res. 25, 249–256. Sonak, S., Pangam, P., Giriyan, A., Hawaldar, K., 2009. Implications of the ban on or- WFD-UKTAG, 2014. UKTAG coastal water assessment method – bentic invertebrate ganotins for protection of global coastal and marine ecology. J. Environ. Manag. 90 fauna. In: Assessment of Imposex in Nucella lapillus. Water Framework Directive – (1), S96–S108. United Kingdom Technical Advisory Group. Sousa, A., Mendo, S., Barroso, C.M., 2005. Imposex and organotin contamination in

323