Several benthic species can be used interchangeably in integrated sediment quality assessment

Ecotoxicology and Environmental Safety 92 (2013) 281–288

Contents lists available at SciVerse ScienceDirect

Ecotoxicology and Environmental Safety journal homepage: www.elsevier.com/locate/ecoenv

Several benthic species can be used interchangeably in integrated sediment quality assessment A. Rodrı´guez-Romero a,n, A. Khosrovyan b, T.A. Del Valls c, R. Obispo d, F. Serrano e, M. Conradi f, I. Riba g a

´n Costera, Instituto de Ciencias Marinas de Andalucı´a (CSIC), avda. Repu ´ blica Saharaui s/n. Campus Rı´o San Pedro, Departamento de Ecologı´a y Gestio ´diz, Spain 11510 Puerto Real, Ca b ´diz, Polı´gono Rı´o San Pedro s/n, Puerto Real 11510, Ca ´diz, Spain UNESCO UNITWIN/WiCop, Facultad de Ciencias del Mar y Ambientales, Universidad de Ca c ´diz, Polı´gono Rı´o San Pedro s/n, 11510 Puerto Real, Ca ´diz, Spain UNITWIN/UNESCO/WiCoP. Physical Chemistry Department, University of Ca d ´pez 81, 28026 Madrid. Spain CEDEX Centro de Estudios de Puertos y Costas, Antonio Lo e ´ Area de Tecnologı´as del Medio Ambiente. Departamento de Ingenierı´a Civil, Edificio Polite´cnico., Universidad de Granada. Campus Fuentenueva s/n. 18071 Granada. Spain f Departamento de Fisiologı´a y Zoologı´a, Facultad de Biologı´a, Universidad de Sevilla. Reina Mercedes 6, 41012 Sevilla. Spain g ´diz. Polı´gono Rı´o San Pedro s/n, Puerto Real 11510, Ca ´diz, Spain Departamento de Quı´mica Fı´sica, Facultad de Ciencias del Mar y Ambientales, Universidad de Ca

a r t i c l e i n f o

abstract

Article history: Received 17 November 2012 Received in revised form 19 February 2013 Accepted 21 February 2013 Available online 24 March 2013

The selection of the best management option for contaminated sediments requires the biological assessment of sediment quality using bioindicator organisms. There have been comparisons of the performance of different test species when exposed to naturally occurring sediments. However, more research is needed to determine their suitability to be used interchangeably. The sensitivity of two amphipod species (Ampelisca brevicornis and Corophium volutator) to sediments collected from four different commercial ports in Spain was tested. For comparison the lugworm, Arenicola marina, which is typically used for bioaccumulation testing, was also tested. Chemical analyses of the sediments were also conducted. All species responded consistently to the chemical exposure tests, although the amphipods, as expected, were more sensitive than the lugworm. It was found that C. volutator showed higher vulnerability than A.brevicornis. It was concluded that the three species can be used interchangeably in the battery of tests for integrated sediment quality assessment. & 2013 Elsevier Inc. All rights reserved.

Keywords: Sediment Toxicity Amphipod Lethality Weight-of-evidence Sensitivity

1. Introduction Dredging is routinely carried out in harbors and navigational channels. However, oxygenation and mixing during the operations may form new chemical bonds causing an influx of contaminants (Casado-Martinez et al., 2007a). This necessitates a pre-dredging study of sediment contamination status to assess the possible its toxic impact. A common tool for testing sediment quality is toxicity bioassays conducted in natural and controlled conditions. Toxicological results together with chemical and physical characterization of the sediments form the basis of a weight-of evidence approach, acknowledged as a powerful tool for assessing sediment quality (e.g., Casado-Martinez et al., 2007b, 2008; Chapman et al., 2002;

n

Corresponding author. Fax: 34 956834701. E-mail addresses: [email protected] (A. Rodrı´guez-Romero), [email protected] (A. Khosrovyan), [email protected] (T.A. Del Valls), [email protected] (R. Obispo), [email protected] (F. Serrano), [email protected] (M. Conradi), [email protected] (I. Riba). 0147-6513/$ - see front matter & 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ecoenv.2013.02.015

Chapman and Anderson, 2005; Crane, 2003; DelValls and Conradi, 2000). The amphipods Corophium volutator and Ampelisca brevicornis have been used for sediment quality characterization in various Spanish ports (Casado-Martınez et al., 2006; Chapman et al., 2002; DelValls et al., 1998; Riba et al., 2003). Generally, amphipod species have been widely used in sediment quality analysis because of their abundance, high sensitivity to environmental contamination and high tolerance to many other environmental factors (Long et al., 2001). According to Casado-Martinez et al., (2007a), C.volutator and A.brevicornis provide similar toxicity responses in highly contaminated sediments but different responses in less contaminated ones. More evidence is required to determine that these two species can be used interchangeably for sediment quality assessments. The lugworm Arenicola marina is also used for sediment quality assessment because of its bioaccumulation capability (Casado-Martinez et al., 2007a, 2008; Morales-Caselles et al., 2008; Ramos-Gomez et al., 2011). The aim of this paper is to compare the sensitivity of C.volutator, A.brevicornis and A.marina to toxic exposure in acute toxicity bioassays to determine the extent they can be used interchangeably in the battery of tests for integrated sediment quality analysis,

282

A. Rodrı´guez-Romero et al. / Ecotoxicology and Environmental Safety 92 (2013) 281–288

including dredged environments. The test battery includes a number of bioassays designed to characterize sediment toxicity and is an intrinsic part of the weight-of-evidence approach.

2. Materials and methods

where percentage of fine particles (henceforth fines) exceeded 80%—muddy. For the rest of the cases, when sand was a prevailing substratum ( 4 60%) it was classified as more sandy; if mud prevailed (460%)—more muddy; if there was approximately equal percentage of sand and mud, then the sediment was sandy– muddy. 2.3. Toxicity tests

2.1. Sediment sampling Sediment samples were from four commercial ports along the Spanish coast: Huelva (H1, H2, H3), Santander (S1, S2, S3), Barcelona (B1, B2, B3) and Cadiz (CA1, CA2, CA3, CA4) (Fig. 1). The site CA2 is located in a fishing dock. CA3 and CA4 have a high accumulation of contaminants; B1-B3 and S2 are subject to intense shipyard and traffic activities. Huelva is affected by historic mining activities. At each site, sediments (three samples per site) were collected by 0.025 m2 Van Veen grab from approximately the top 20 cm and transported in a cool container to the lab, where they were homogenized with a Teflons spoon and sieved through a 2 mm mesh to eliminate debris. The sediments were then subsampled for physical characterization and chemical quantification, following Spanish recommendations for dredged materials (CEDEX, 2008). The sediment samples were stored at 4 1C in the dark and hermetically closed for at most two weeks. Field related work complied with quality assurance recommendations per ASTM (1991a,b); laboratory tests were conducted with appropriate quality control (blanks, controls, ambient conditions, etc.) as outlined by Morales-Caselles et al. (2008) and Riba et al. (2003). 2.2. Sediment analysis The total concentrations of select metals (Hg, Cd, Pb, Cu, Zn, As, Ni and Cr), 7 polychlorinated biphenyls (PCB: 28, 52, 101, 118, 138, 153, 180), polycyclic aromatic hydrocarbons (PAH: anthracene, benz(a)anthracene, benzo(ghi)perylene, benzo(a)pyrene, chrysene, fluoranthene, indene (1,2,3-cd) pyrene, pyrene and phenanthrene), organoclorine pesticides (POC: HCH, aldrin, DDT, dieldrin, endosulfan, ensosulfansulphate, endrin, endrin-aldehyde, heptachlor, heptachlor epoxide, hexachloro-1,3butadiene, hexachlorobenzene, lindeno) and tributyltin (TBT) were measured to determine the chemical content of the sediments. Concentrations of Cd, Pb, Cu, Zn, Ni and Cr were determined by microwave acid digestion in Teflon vessels and quantified using atomic absorption spectrometry. Cold vapor technique and hydride generation were used for Hg and As, respectively. Both metals were determined by atomic absorption spectrometry. Results were expressed as mg kg-1 dry weight. PCBs and POCs and 9 PAHs were extracted with cyclohexane and dichloromethane. PCBs and POCs were quantified by gas chromatography with electron capture detection (by US EPA method 8080) and 9 PAHs recommended by OSPAR were detected by HPLC with fluorescence detection (US EPA method 8310). TBTs were extracted with hexane and tropolone 0.05% and analyzed by selective ion monitoring gas chromatography-mass spectrometry (SIM GC/MS) Organic compound concentrations were expressed as mg kg  1dry weight. The accuracy of all analytical procedures was verified using the reference materials MESS-1 NRC and CRM 277 BCR for metals, and NRC-CNRC HS-1 for PCBs and PAHs, with a percentage of recovery higher than 90%. Detection limits ranged between 0.001 and 0.008 mg-kg—1 and 10–20 mg-kg—1 dry weight of sediment for metals and PAHs, respectively, and were 0.5 mg kg  1 dry weight of sediment for PCBs and 2 mg kg  1 dry weight of sediment for TBTs. Organic carbon content (TOC) in sediments was determined according to (MAPA, 1998) and expressed as percentage. For granulometry analysis, recommendations by Thain and Bifield (2001) were used: coarse particle size exceeded 2 mm, sand size ranged from 0.063–2 mm and fine particles had sizeo 0.063 mm. Sediments with the percentage of sand4 than 80% were considered sandy; those

Three independent 10 day static acute sediment toxicity tests were conducted using two crustacean amphipods (A.brevicornis and C.volutator) and a polychaete (A.marina). The percentage of mortality after exposure was selected as the toxicity endpoint. 2.3.1. Amphipod test Bioassays using the crustacean amphipods were conducted according to the procedures described in (Casado-Martinez et al., 2006; Morales-Caselles et al., 2008; Riba et al., 2003). Reference species of A.brevicornis and C.volutator were collected from clean areas of the coasts of Cadiz (Casado-Martinez et al., 2007b) and Galicia (Morales-Caselles et al., 2007), respectively, by sieving the sediment through a 0.5 mm mesh. The amphipods were transported to the laboratory in containers with seawater. In the laboratory, they were placed in 11 L aquariums with clean seawater and sieved sediment from the same locations and were acclimated for 7 days. During acclimation, aeration was provided, natural photoperiod was selected and no food was supplied. Prior to the toxicity tests, approximately 200 g of sieved sediment from the different study sites were placed in 5 replicates in 2 l glass beakers and with about 800 ml of overlying clean seawater. When the sediments settled down, aeration was provided. After 12 h, 20 amphipods were introduced into each test beaker. Natural photoperiod was selected and no food was provided during the experiment. Temperature (207 1 1C), pH (7.9–8.2), salinity (34–35%) and dissolved oxygen (Z 90%) in the seawater were controlled daily. Concentrations of total ammonia in interstitial water were measured by indophenol blue adsorption method using a Technicon TRAACS 800 autoanalyzer and monitored at the beginning (day 2) and at the end (day 8) of the bioassays. They were below 16 mg/l in all samples. 2.3.2. Polychaete test A.marina were collected from a clean area situated on the Cantabrian coast of north Spain (Casado-Martinez et al., 2008) by hand-digging and immediately transported to the laboratory in cool boxes containing clean seawater (the time between the collection and the arrival was about 24 h). In the laboratory, the lugworms were placed in aquariums with 5 cm of sieved sediment from the survey area and acclimated for 7 days. Prior to the test, 5 cm layers of sieved sediments were placed in 20 l aquariums (three replicates) and a clean seawater was added. When the sediment settled aeration was provided. 12 hours later, 10 A.marina were introduced into each replicate. During the acclimation and experimental periods, the natural photoperiod was selected, no food was provided and water was renewed on day 5. Bioassay was performed following to the protocol used by Thain and Bifield (2001). The temperature (177 1 1C), pH (7.8–8.4), salinity (34–35%) and dissolved oxygen (Z 85%) in seawater were controlled daily. 2.4. Sediment categorization The degree of sediment contamination was expressed as categories derived from the Spanish dredged material management framework (CEDEX, 2008) (Fig. 2). The framework uses a weight-of-evidence methodology by linking

Port of Santander

Spain

Port of Barcelona Mediterranean Sea

Port of Huelva

Port of Cádiz

Fig. 1. Map showing locations of the selected ports.

A. Rodrı´guez-Romero et al. / Ecotoxicology and Environmental Safety 92 (2013) 281–288

283

Grain size distribution ≤ 10 % fines and total organic carbon ≤ 1% No

Yes

BSPT ≤ 5 UT

No

Chemical Analysis

TOXICITY TESTS

Yes Ci ≤ AL1

Yes

Tox (+)

1º Liquid Phase Category A

Tox (-)

Category B Tox (+)

No

Category C Yes Yes

Ci ≤ IAL No

BSPT ≤ 5 UT

Tox (-)

Tox (-)

No

Tox Solid Phase Test (+)

Ye s Ci ≤ AL2

Yes No

Ci ≤ AL3

No

Fig. 2. Dredged material characterization scheme based on weight-of-evidence approach. BSPT-basic solid phase test Microtoxs expressed in toxicity units (UT), Ci–concentration of ith sediment chemical.

sediment chemical content to the results of toxicity bioassays. First, sediment chemical content was compared with different guidelines called action limits (AL)—AL1, AL2, Intermediate AL (IAL), AL3—adopted by the framework (CEDEX, 2008) (Table 1). Second, biological responses to the toxicity tests were considered to categorize sediment contamination level: A—low-contaminated sediment suitable for open-water disposal, B—moderately-polluted and C—highly-polluted ones. The disposal of C is not possible without further cleaning. In comparing the individual contribution of test organisms to the sediment categorization of the framework (CEDEX, 2008), only responses in whole-sediment bioassays were considered. The responses in Microtoxs BSPT (not shown and discussed in this work) were also taken into consideration (Fig. 2). They exceeded the threshold of 5 UT (toxicity units), used in the framework, in all sites except for CA1, B1, B2 and S1.

2.5. Statistical analysis The determination of toxicity of sediment samples was based on the statistically significant differences between sample and reference mean endpoints. Two different methods were applied and compared: 1) non-parametric Mann– Whitney U test (as normality of the data was not met) and 2) the previous method with an additional testing of the arithmetic difference between mortality rates of the reference and the sample means (if the difference is more than 20%, then the sediment is toxic). The second approach has been previously reported by MoralesCaselles et al. (2007). Non-parametric Spearman correlation analysis between all variable pairs was selected because of the small sample size and insensitivity to normality assumption. Multivariate (factor) analysis (with principal component as extraction method and Varimax as rotation method) was used to find associations among variables. The results of the associations show the bioavailability of the contaminants and their spatial relevance. Each extracted component represented combinations of highly correlated variables, whereas correlations among the components were weak. Correlation of each variable within the component was determined by component loadings. Positive component scores (linear combinations of the value of the variables with their corresponding loadings) were interpreted to determine spatial relevance, i.e., impact of the factor in each sample. The parameters used in the factor analysis were normalized first, to the reference (CA1) and then, maximum values, before being tested for the normality requirement. When necessary, square root or logarithmic transformations were applied. The data on organic contaminants were first normalized to TOC and then log transformed.

Component loadings (showing the weight of a variable in the given component) with absolute value of Z0.4 were used for interpretation, approximating Comreys’ cut-off of 0.55 (Morales-Caselles et al., 2007). The percentage of variance in the original data explained by each factor was also shown. Statistical analysis was performed using SPSS 15.

3. Results 3.1. Sediment characterization The physical and chemical properties of the sediments are shown in Table 2. Levels of organic contaminants (PAHs, PCBs and TBT) varied broadly between the stations and were higher in sediments collected in sampling sites where the ship traffic activities were intense (CA2, CA4, B3 and S2). They were not detectable in CA1 and S1. The highest concentrations for most of the metals were found in the Huelva sampling sites, especially in H3. Sediment categories derived by the framework (CEDEX, 2008) per species and per sample were the same. However, intra-port variability can be noticed in the ports of Barcelona (B]) and Santander (S]), suggesting impact from local point sources (Fig. 3). Station CA1 from Cadiz, where no chemical concentrations exceeded AL1 limits (category A), was used as a reference site. Spearman correlation analysis showed several relationships between metals and sediment physical properties (Table 3). Two methods of sediment toxicity determination (based on non-parametric Mann–Whitney U test) produced identical estimates for all species. Both methods estimated all samples to be toxic for the Corophium spp. For the Arenicola and Ampelisca spp., sample B2 was estimated to be non-toxic (Table 4).

A. Rodrı´guez-Romero et al. / Ecotoxicology and Environmental Safety 92 (2013) 281–288

284

Table 1 Sediment quality guidelines called action limits (AL)–AL1, AL2, Intermediate AL (IAL), AL3, adopted by Spanish dredged material management framework of 2008. Parameter

(total concentrations measured in sediment fine fraction o 2 mm)

Unit (dry weight)

Hg Cd Pb Cu Zn As Ni Cr P PCB P7 9PAH TBT

(mg/kg) (mg/kg) (mg/kg) (mg/kg) (mg/kg) (mg/kg) (mg/kg) (mg/kg) (mg/kg) (mg/kg) (mg/kg)

AL1

IAL

AL2

AL3

0.18 0.6 47 34 150 30 39 81 22 940 100

0.71 2.4 188 136 410 70 78 324 88 3760 200

2.84 12 752 680 1640 280 234 1296 440 18800 1000

14.2 60 3760 3400 8200 1400 1170 6480 2200 94000 5000

P 9–9 PAHs recommended by OSPAR (anthracene, benz(a)anthracene, benzo(ghi)perylene, benzo(a)pyrene, chrysene, fluoranthene, indene (1,2,3–cd) pyrene, pyrene and phenanthrene). P 7–PCBs congeners (28, 52,101, 118, 138, 153, 180).

Table 2 Physical–chemical characterization of the sediment (ports of Cadiz CA#, Huelva H#, Barcelona B# and Santander S#). Data are expressed as mg kg  1 dry weight, except for PCB, POC and TBT, which are expressed in mg kg  1 dry weight, and oxidizable carbon content (TOC) and fines expressed in % of dry weight. Marked areas show chemicals whose concentrations exceeded Action Level 1 limit, used in the Spanish dredged material management framework of 2008. Site

Fines

TOC

TBT

POC

PAHsa

PCBsb

Hg

Cd

Pb

Cu

Zn

As

Ni

Cr

CA1 CA2 CA3 CA4 H1 H2 H3 B1 B2 B3 S1 S2 S3

2.46 68.31 66.82 34.77 54.59 73.34 67.45 47.85 19.25 70.59 0.34 88.93 63.39

0.13 2.13 2.58 1.23 0.59 1.63 1.69 0.92 0.41 3.82 0.12 2.07 3.85

n.dc 742 37 718 o 12 16 41 24 62 1166 n.d c 791 24

0.01 23.1 8.9 14 0.01 0.01 9.6 34 12.6 167.8 0.01 861.9 0.01

n.dc 1.12 0.84 1.23 o 0.04 0.1 0.1 o 0.23 0.36 2.01 o 0.01 2.16 2.21

n.dc 243.3 6.6 21 5.4 5.5 14.1 28.2 34.7 361.4 n.dc 320.7 30.7

o 0.1 0.97 0.25 0.19 1.24 2.66 7.05 0.48 0.6 4.25 o 0.1 0.94 0.46

0.062 0.338 0.156 0.134 0.704 1.93 7.83 0.382 0.328 1.627 0.055 1.214 5.713

5.008 105.1 31.4 28.53 198.1 412.1 991 50.1 50.05 355 11.79 114.7 107.8

2.226 189.5 42.29 45.08 549.7 1636 1849 44.15 36.6 489.1 1.19 538.9 44.77

12 294.4 113.8 87.24 725.2 1584 2086 110.7 90.68 763.4 31.15 959.1 2869

2.524 15.08 13.12 5.043 126.6 294.5 533.6 12.07 9.54 27.46 8.66 42.15 28.73

8.711 22.78 44.23 12.88 21 28.49 28.49 28.96 14.16 27.21 2.899 33.11 43.99

4.458 60.67 34.85 105.7 42.28 66.93 77.64 37.69 23.15 96.74 5.635 90.64 52.52

a P 9–9 PAHs recommended by OSPAR (anthracene, benz(a)anthracene, benzo(ghi)perylene, benzo(a)pyrene, chrysene, fluoranthene, indene (1,2,3–cd) pyrene, pyrene and phenanthrene). P b 7–PCBs congeners (28, 52,101, 118, 138, 153, 180). c n.d. values below detection limits.

C B

B

B

C

C

C

B

B

Ampelisca

S3

Arenicola

S2

S1

A

B3

A

B2

H3

H2

H1

CA4

CA3

CA2

CA1

A

B1

Corophium A

Fig. 3. Schematic representation of sediment categories (A–C), according to the Spanish dredged material management framework of 2008. The categories are derived separately for each of the toxicological responses: by Ampelisca brevicornis, Corophium volutator, Arenicola marina. Category A (low contaminated) allows for open-water disposal of untreated dredged sediment, category B is moderately-polluted sediment, category C (highly-polluted) prohibits disposal of untreated sediment. Sediments are represented on X-axis.

3.2. Characterization of toxicity responses The Spearman correlation coefficients showed significant positive correlation between the mortality rates of all the studied species and the percentage of fines. The endpoints obtained from the amphipods were also correlated with TOC (Table 5). The toxicity endpoints of all the species tested demonstrated strong positive correlation with all metals (except for nickel in the case of A.brevicornis) (Table 6). For all the species studied, toxicity

endpoints did not show any statistically significant correlation with organic micro-pollutants. 3.3. Factor analysis Multivariate factor analysis with three factors explained 87% of the original variance and identified the following relationships between sediment conditions and biological effects on the organisms exposed to these conditions (Tables 7 and 8):

A. Rodrı´guez-Romero et al. / Ecotoxicology and Environmental Safety 92 (2013) 281–288

285

Table 3 Spearman correlation coefficients between metals and sediment physical properties: percentage of sediment fines and oxidizable organic carbon (TOC).

TOC Fines a b

Hg

Cd

Pb

Cu

Zn

As

Ni

Cr

0.40 0.73b

0.60a 0.71b

0.49 0.77b

0.42 0.77b

0.71b 0.78b

0.48 0.76b

0.76b 0.65a

0.57a 0.68a

Significant level at 0.05. Significant level at 0.01.

Table 4 Mortality rates of A.brevicornis, C.volutator and A.marina, categories of sediment samples, types of sediment substrate and sediment toxicity estimations by two methods: 1) statistically significant differences between sample and reference (CA1) means based on non-parametric Mann–Whitney U test; 2) the previous one with additional testing of arithmetic difference between the reference and the sample means of 20% (toxic, if 420%). Sediment

A.brevicornis

C.volutator

Site Category Type

Mortality, %

Sediment toxicity Mann–Whitney U/Mann–Whitney U and diff

Mortality, %

Sediment toxicity Mann– Whitney U/Mann–Whitney U and diff

Mortality, %

Sediment toxicity Mann– Whitney U/Mann–Whitney U and diff

CA1 B2 S1 CA4

3.33 16.67 26.67 30

– No/No Yes/Yes Yes/Yes

8 30 30 38

– Yes/Yes Yes/Yes Yes/Yes

3.33 0 0 26.67

– No/No No/No Yes/Yes

30

Yes/Yes

46

Yes/Yes

30

Yes/Yes

30 33.33

Yes/Yes Yes/Yes

60 60

Yes/Yes Yes/Yes

43.33 33.33

Yes/Yes Yes/Yes

43.33

Yes/Yes

48

Yes/Yes

26.67

Yes/Yes

53.33

Yes/Yes

64

Yes/Yes

43.33

Yes/Yes

56.67

Yes/Yes

70

Yes/Yes

46.67

Yes/Yes

70

Yes/Yes

70

Yes/Yes

40

Yes/Yes

83.33

Yes/Yes

88

Yes/Yes

60

Yes/Yes

93.33

Yes/Yes

96

Yes/Yes

73.33

Yes/Yes

A A A B

CA3 B S2 B1

B A

CA2 B H1

B

B3

C

S3

C

H2

C

H3

C

sandy sandy sandy more sandy more muddy muddy sandy– muddy more muddy sandy– muddy more muddy more muddy more muddy more muddy

A.marina

Table 5 Spearman correlation coefficients between mortality rates of A.brevicornis, C.volutator and A.marina, sediment grain size and oxidizable organic carbon (TOC). Species

Fines

Sand

TOC

A.brevicornis C.volutator A.marina

0.64a 0.72b 0.77b

 0.66a  0.73b  0.75b

0.59a 0.58a 0.52

a b

Significant level at 0.05. Significant level at 0.01.

The first factor (F#1) associated sediment fines, TOC, and all metal contaminants with the responses of all three organisms. The positive scores of F#1 were obtained for samples H1-H3, B1, B3, S2 and S3. F#1 explained 50% of variance in the original data. The second factor (F#2) linked Cu, Cr with all organic contaminants. The positive factor scores were shown in CA2, CA4, H3, B1-B3 and S2 (19% of the variance). The third factor (17% of the variance) revealed associations between sediment fine grains, Zn, Ni and PAH-based contaminants. This association was shown to be important for the samples CA2CA4, B3, S2 and S3.

4. Discussion The performances of three benthic species in acute bioassays with similar conditions were compared to analyze their relative contributions to sediment quality assessment. This study extends earlier works on toxicity responses of the same species, which however stood alone, without comparison of their relative sensitivities, for a few exceptions reported by e.g., Casado-Martinez et al. (2007b), Riba et al. (2003). The present attempt enables interchangeability and flexibility in using various species in the battery of tests under WOE context.

A. Rodrı´guez-Romero et al. / Ecotoxicology and Environmental Safety 92 (2013) 281–288

286

Table 6 Spearman correlation coefficients between mortality rates of A.brevicornis, C.volutator, A.marina and metal concentrations in sediments. Species

Hg

Cd

Pb

Cu

Zn

As

Ni

Cr

A.brevicornis C.volutator A.marina

0.76b 0.81b 0.8b

0.88b 0.95b 0.88b

0.89b 0.94b 0.92b

0.81b 0.84b 0.89b

0.87b 0.92b 0.84b

0.85b 0.91b 0.87b

0.49 0.58a 0.57a

0.56a 0.6a 0.63a

a b

Significant level at 0.05. Significant level at 0.01.

Table 7 Rotated component matrix with component loadings. Parameters

Component 1

FINES SAND TOC Hg Cd Pb Cu Zn As Ni Cr TBT PAH_TOTAL PAH_SUM PCB_COGENERS POC AMPELISCA_MORTALITY ARENICOLA_ MORTALITY COROPHIUM_ MORTALITY

2

0.72  0.7 0.41 0.87 0.88 0.93 0.86 0.84 0.95 0.52 0.49

3 0.48  0.52 0.83

0.43 0.41 0.67 0.63 0.89 0.46 0.45 0.82 0.88

0.74 0.72

0.94 0.97 0.95

PAH_SUM–9 PAHs recommended by OSPAR (anthracene, benz(a)anthracene, benzo(ghi)perylene, benzo(a)pyrene, chrysene, fluoranthene, indene (1,2,3–cd) pyrene, pyrene and phenanthrene). PCB_COGENERS—(28, 52,101, 118, 138, 153, 180).

Table 8 Positive component scores per sample. Sample

Component 1

CA1 CA2 CA3 CA4 H1 H2 H3 B1 B2 B3 S1 S2 S3

þ þ þ þ þ þ þ

Carpentier et al. (2002) identified two factors of sediment contamination: first, the organic content of the sediments is strongly related to the proportion of fines in sediment; second, fine particles tend to adsorb more organic matter and metal pollutants than coarse sediments. Indeed, high metal concentrations were measured in the samples with high organic content and high proportion of fines (Table 2). For example, metal load in fine and carbon-enriched sediments of H1-H3, S3 and B3 (Hg, Pb, Cu, Zn) often exceeded

2

3

þ þ

þ þ þ

þ þ þ þ

þ

þ

þ þ

different thresholds—Action Levels (Tables 1 and 2). The mortality rates of all tested species were strongly and positively correlated with the percentage of fines in the sediments (Table 5). In addition, the rates were very high in fine-based samples H1-H3, B3 and S3 (A.marina440%, A.brevicornis453% and C.volutator464%) and both methods of sediment toxicity determination estimated them as statistically significantly different from the reference (i.e., toxic) (Table 4).

A. Rodrı´guez-Romero et al. / Ecotoxicology and Environmental Safety 92 (2013) 281–288

The grouping of the factor F#1 assumes that metal toxicants could be cause for the mortality of A.brevicornis, C.volutator and A.marina in sediments H1-H3, B1, B3, S2 and S3 (Tables 7 and 8). High concentrations, bioavailable forms, different exposure pathways could equally increase species vulnerability. Bat et al. (1998) demonstrated reduced survival and burrowing activity of C.volutator with increasing concentrations of Cd, Cu and Zn in sediments. Mercury was very toxic for C.volutator as it tended to rapidly accumulate in their bodies at lower concentrations of the medium, according to Meadows and Erdem (1982). Adverse impact of dissolved metals on the studied organisms cannot be excluded as well e.g., C.volutator’s uptake of metals from contaminated water was found to be higher than from the contaminated sediment (Bat et al., 1998). The factor F#2 (Tables 7 and 8) defined contamination by metals (Cu and Cr) and organic chemicals that was not linked to mortality in sediments CA2, CA4, H3, B1-B3 and S2. In all those sediments except for B1 and B2, carbon enrichment and fine nature of sediments can be noticed. Likewise, all sites except for B1 had high measured concentration of at least one of the indicated (metal or organic) contaminants (often exceeding AL2 threshold) (Tables 1 and 2). However, F#2 can be interpreted as not polluted with regard to organic chemicals, which, if linked to F#1, confirms highly toxic impact of metal contaminants, at least, in the samples indicated by both factors (H3, B3 and S2). The Spearman correlation analysis also did not reveal statistically significant relationship between mortality and organic chemicals. Morales-Caselles et al. (2008) reported a similar low mortality of A.brevicornis and A.marina exposed to sediments that were highly contaminated by PAHs, while PAH accumulation occurred in the lugworm. Ramos-Gomez et al. (2011) showed enhanced enzymatic activity in A.marina affected by PAHs pollution. Hence, deleterious effect of organic contaminants on the organisms (all were burrowers and deposit feeders) over a longer time span cannot be excluded. The interpretation of the factor F#3 strengthened no-pollution effect by polycyclic aromatic hydrocarbons on any of the species. While the multivariate analysis associated the mortality rates of all species with metal concentrations in the sediments (F#1and F#2), an attempt was made to analyze relative sensitivity of these species to metal exposure. In highly-polluted sediments H2, H3 and S3, the lethal responses of both amphipods were almost identical and much higher than that of the lugworm. This pattern changed with low- and moderately-contaminated sediments (categories A and B). In low contaminated samples CA1, B2, S1 and B1 (category A), the responses of A.brevicornis were almost two-fold lower, compared to that of C.volutator, while the mortality of A.marina was either zero or identical to that of A.brevicornis. The same pattern was observed in moderately contaminated samples (category B). Generally, all species showed similar high mortality in highly contaminated sediment, but in low and moderately-contaminated medium the variability of the responses increased, though a trend of dependence on toxicant load in sediments remained unchanged. C. volutator demonstrated higher mortality in all tests, compared to A.brevicornis and A.marina. This is in agreement with the observations by Riba et al. (2003) and Casado-Martinez et al. (2007b). The concurrent uptake of metals by Corophium spp, from the overlying water (Bat et al., 1998) could enhance exposure of this species to toxicants. Besides, C.volutator is less tolerant to sandy sediments or habitat modification (Casado-Martınez et al., 2006, 2007b) and absence of sediments is stressful to them by itself (Bat et al., 1998). Despite the differences, the mortality rates of both amphipods were more representative of the metal load than that of the lugworm. For example, in completely different samples B1 (category A) and S3 (category C), the mortality rates of the lugworm were not very different (33% vs 40%, Table 4). Previous reports

287

(Ramos-Gomez et al., 2011) also confirm this: a higher body burden in A.marina from the control sediment than from the places where toxicity thresholds set by OSPAR and NOAA were exceeded. This pattern might be directly associated with the ability of A.marina to accumulate harmful substances (Morales-Caselles et al., 2008): uptake control mechanism regulates body burden of the lugworm depending on contaminant load in the medium. The consistency of the responses of the amphipods is likely related to their more pronounced vulnerability to toxic exposure, compared to the lugworm. The differences between all species could be explained by sensitivity to specific toxicants, uptake routes, tolerance to unfavorable sediment conditions, physiological or behavioral peculiarities. A sediment contaminant matrix is complex and even if toxic generally, it does not necessarily mean that the toxicity is caused by each component or that it is equally toxic for various organisms (Bat, 2005). Herewith, tube builders (e.g., A.marina) are more isolated from sediment contaminants and tubes reduce interstitial water contact with them (J.A. de-laOssa-Carretero et al., 2012). Nevertheless, Table 4 clearly shows distinct responses depending on sediment categories with exception of a few cases with possibly marginal impact on the organisms, e.g., mortality rates of A.brevicornis and C.volutator were almost similar in B1 (category A) and S2 (category B). Previous authors favored suitability of A.marina in acute toxicity bioassays (Bat, 1998, 2005; Thain and Bifield, 2001). The findings of the present study support this conclusion based on generally consistent responses obtained from all three species. Likewise, all three species, employed separately in sediment characterization by CEDEX (2008), provided the same categorization for all studied sediments, which differed considerably from each other by physical and chemical properties. Moreover, Spearman correlation analysis revealed strong positive correlations between mortality rates of the species (p o0.01): between the amphipods (0.97), between the Arenicola and Corophium spp. (0.94) and between the Ampelisca and Arenicola spp. (0.86). Such consistency provides flexible tools for sediment toxicity assessment, in particular for dredging purposes, assuming the predredging analysis of possible toxic effects, according to CEDEX (2008), and further evaluation of feasibility of different management options in this context.

5. Conclusions Contamination-dependent responses in the three species were demonstrated for metals but not for organic compounds. Mortality rates of C.volutator in low to moderately contaminated samples were higher than that of A.brevicornis and A.marina. C.volutator is more vulnerable to the physical or chemical condition of sediment. Taking into account the predictable and consistent responses by all species to sediment contamination, they are recommended for interchangeable employment in the battery of tests used for integrated sediment quality assessment, including for dredging purposes. The amphipods demonstrated repeatability of the results affirming their significance in acute toxicity testing. The work addressed the important issue of determining the eco-toxicity of naturally occurring sediments (in integrity of its constituents and complex interactions among them) eliminating the bias intrinsic to lab-designed sediment conditions.

Acknowledgments This work was supported by the Spanish Ministry of Science and Innovation through the grants CTM 2011-2843-CO2-02 and

288

A. Rodrı´guez-Romero et al. / Ecotoxicology and Environmental Safety 92 (2013) 281–288

CTM2012-36476-C02-01and by the grant RNM-3924 of the Junta de Andalucı´a. The authors thank port authorities of Ca´diz, Santander, Huelva and Barcelona as well as Ca´tedra UNESCO/ UNITWIN/WiCop for their assistance in data acquisition. References American Society for Testing and Materials (ASTM), 1991a. Standard guide for conducting 10 days static sediment toxicity test with marine and estuarine amphipods, Annual Book of ASTM Standards E 1367-90. American Society for Testing and Materials, Philadelphia, PA, pp. 310–390.. American Society for Testing and Materials (ASTM) , 1991b. Standard Guide for Collection, Storage, Characterization And Manipulation of Sediments for Toxicological Testing. Publ. E1391-90. Philadelphia, USA. Bat, L., 1998. Influence of sediment on heavy metal uptake by the polychaete Arenicola marina. Turk. J. Zool. 22, 341–350. Bat, L., 2005. A review of sediment toxicity bioassays using the amphipods and polychaetes. Turk. J. Fish. Aquat. Sci. 5, 119–139. Bat, L., Raffaelli, D., Marr, I.L., 1998. The accumulation of copper, zinc and cadmium by the amphipod Corophium volutator. J. Exp. Mar. Biol. Ecol. 223, 167–184. Carpentier, S., Moilleron, R., Beltra´n, C., Herve´, D., The´venot, D., 2002. Quality of dredged material in the River Seine basin (France). I physico-chemical properties. Sci. Total Environ. 295, 101–113. Casado-Martınez, M.C., Beiras, R., Belzunce, M.J., Gonzarlez-Castromil, M.A., Marırn-Guirao, L., Postma, J.F., Riba, I., DelValls, T.A., 2006. Inter-laboratory assessment of marine bioassays to evaluate environmental quality of coastal sediments in Spain: IV. Whole sediment toxicity test using crustacean amphipods. Cienc. Mar. 32, 149–157. Casado-Martınez, M.C., Branco, V., Vale, C., Ferreira, A.M., DelValls, T.A., 2008. Is Arenicola marina a suitable test organism to evaluate the bioaccumulation potential of Hg, PAHs and PCBs from dredged sediments? Chemosphere. 70, 1756–1765. Casado-Martinez, M.C., Ferna´ndez, N., Forja, J.M., DelValls, T.A., 2007a. Liquid versus solid phase bioassays for dredged material toxicity assessment. Environ. Int. 33, 456–462. Casado-Martinez, M.C., Forja, J.M., DelValls, T.A., 2007b. Direct comparison of amphipod sensitivities to dredged sediments from Spanish ports. Chemosphere 68, 677–685. CEDEX,2008. (Centro de Estudios de Puertos y Costas). Gestio´n ambiental del material de dragado, V seminario de ingenierı´a y operaciones portuarias. Concepcio´n, Chile.

Chapman, P.M., Anderson., J, 2005. A decision-making framework for sediment contamination. Integrated Environ. Assess. Manage. 1, 163–173. Chapman, P.M., Ho, K.T., Munns, W.R., Solomon, K., Weistein, M.P., 2002. Issues in sediment toxicity and ecological risk assessment. Mar. Pollut. Bull. 44, 271–278. Crane, M., 2003. Proposed development of sediment quality guidelines under the European Water Framework Directive: a critique. Toxicol. Lett 142, 195–206. DelValls, T.A., Conradi, M., 2000. Advances in marine ecotoxicology: laboratory tests versus field assessments data on sediment quality studies. Cienc. Mar. 26, 39–64. DelValls, T.A., Forja, J.M., Gormez-Parra, A., 1998. The use of multivariate analysis in linking sediment contamination and toxicity data for establish sediment quality guidelines: an example in the Gulf of Cadiz (Spain). Cien. Mar. 24, 127–154. de-la-Ossa-Carretero, J.A., Del-Pilar-Ruso, Y., Gime´nez-Casalduero, F., Sa´nchezLizaso, J.L., Dauvin, J.-C., 2012. Sensitivity of amphipods to sewage pollution. Estuarine Coastal Shelf Sci. 96, 129–138. Long, E.R., Hong, C.B., Severn, C.G., 2001. Relationships between acute sediment toxicity in laboratory tests and abundance and diversity of benthic infauna in marine sediments: a review. Environ. Toxicol. Chem. 20, 46–60. Ministerio de Agricultura, Pesca y Alimentacio´n (MAPA), 1998. Me´todos Oficiales de Ana´lisis en la Unio´n Europea. Madrid. Meadows, P.S., Erdem, C., 1982. The effect of mercury in Corophium volutator viability and uptake. Mar. Environ. Res. 6, 227–233. Morales-Caselles, C., Kalman, J., Micaelo, C., Ferreira, A.M., Vale, C., Riba, I., DelValls, T.A., 2008. Sediment contamination, bioavailability and toxicity of sediments affected by an acute oil spill: four years after the sinking of the tanker Prestige (2002). Chemosphere 71, 1207–1213. Morales-Caselles, C., Kalman, J., Riba, I., DelValls, T.A., 2007. Comparing sediment quality in Spanish littoral areas affected by acute (Prestige, 2002) and chronic (Bay of Algeciras) oil spills. Environ. Pollut. 146, 233–240. Ramos-Gomez, J., Martins, M., Raimundo, J., Vale, C., Martı´n-Dı´az, M.L., DelValls, T.A., 2011. Validation of Arenicola marina in field toxicity bioassays using benthic cages: Biomarkers as tools for assessing sediment quality. Mar. Pollut. Bull. 7, 1538–1549. Riba, I., DelValls, T.A., Forja, J.M., Gormez-Parra, A., 2003. Comparative toxicity of contaminated sediment from a mining spill using two amphipods species: Corophium volutator (Pallas, 1776) and Ampelisca brevicornis (A. Costa, 1853). Bull. Environ. Contam. Toxicol. 71, 1061–1068. Thain, J., Bifield, S., 2001. Biological effects of contaminants: sediment bioassay using the polychaete Arenicola marina. ICES Tech. Mar. Environ. Sci., 29.