chromium, copper, and nickel

Do trace metals (chromium, copper, and nickel) influence toxicity of diesel fuel for free-living marine nematodes? Amor Hedfi, Fehmi Boufahja, Manel Ben Ali, Patricia Aïssa, Ezzeddine Mahmoudi & Hamouda Beyrem Environmental Science and Pollution Research ISSN 0944-1344 Environ Sci Pollut Res DOI 10.1007/s11356-012-1305-2

1 23

Your article is protected by copyright and all rights are held exclusively by SpringerVerlag Berlin Heidelberg. This e-offprint is for personal use only and shall not be selfarchived in electronic repositories. If you wish to self-archive your work, please use the accepted author’s version for posting to your own website or your institution’s repository. You may further deposit the accepted author’s version on a funder’s repository at a funder’s request, provided it is not made publicly available until 12 months after publication.

1 23

Author's personal copy Environ Sci Pollut Res DOI 10.1007/s11356-012-1305-2

RESEARCH ARTICLE

Do trace metals (chromium, copper, and nickel) influence toxicity of diesel fuel for free-living marine nematodes? Amor Hedfi & Fehmi Boufahja & Manel Ben Ali & Patricia Aïssa & Ezzeddine Mahmoudi & Hamouda Beyrem

Received: 28 July 2012 / Accepted: 6 November 2012 # Springer-Verlag Berlin Heidelberg 2012

Abstract The objective of this study was to test the hypotheses that (1) free-living marine nematodes respond in a differential way to diesel fuel if it is combined with three trace metals (chromium, copper, and nickel) used as smoke suppressants and that (2) the magnitude of toxicity of diesel fuel differs according to the level of trace metal mixture added. Nematodes from Sidi Salem beach (Tunisia) were subjected separately for 30 days to three doses of diesel fuel and three others of a trace metals mixture. Simultaneously, low-dose diesel was combined with three amounts of a trace metal mixture. Results from univariate and multivariate methods of data evaluation generally support our initial hypothesis that nematode assemblages exhibit various characteristic changes when exposed to different types of disturbances; the low dose of diesel fuel, discernibly non-toxic alone, became toxic when trace metals were added. For all types of treatments, biological disturbance caused severe specific changes in assemblage structure. For diesel fuel-treated microcosms, Marylynnia bellula and Chromaspirinia pontica were the best positive indicative species; their remarkable presence in given ecosystem may predict unsafe seafood. The powerful toxicity of the combination between diesel fuel and trace metals was expressed with only negative bioindicators, namely Trichotheristus mirabilis, Pomponema multipapillatum, Ditlevsenella murmanica, Desmodora longiseta, and Bathylaimus capacosus. Assemblages with high abundances of these species should be an index of healthy seafood. When nematodes were exposed to only trace metals, their response looks special with a distinction of a different list of indicative Responsible editor: Philippe Garrigues A. Hedfi : F. Boufahja (*) : M. Ben Ali : P. Aïssa : E. Mahmoudi : H. Beyrem Faculty of Sciences of Bizerte, Laboratory of Environment Biomonitoring, Coastal Ecology and Ecotoxicology Unit, University of Carthage, 7021, Zarzouna, Tunisia e-mail: [email protected]

species; the high presence of seven species (T. mirabilis, P. multipapillatum, Leptonemella aphanothecae, D. murmanica, Viscosia cobbi, Gammanema conicauda, and Viscosia glabra) could indicate a good quality of seafood and that of another species (Oncholaimellus mediterraneus) appeared an index of the opposite situation. Keywords Diesel fuel . Trace metals . Toxicity . Meiobenthic nematodes . Seafood quality

Introduction The invention and widespread of internal combustion engines at the beginning of the twentieth century caused a radical revolution in petroleum refining. Ever since then, immense progress has been made to make diesel fuel better known and produced throughout the world (Ajiwe et al. 2003). Several generic types of diesel fuel additives can have a significant effect on emissions. These include smoke suppressants which are principally colloidally dispersed or solubilized carboxylate forms of metals (arsenic, cadmium, chromium, lead, nickel, manganese, selenium, mercury, iron, and copper) (Rising et al. 2004). Added to diesel, these compound catalyst diesel particulate traps regeneration and inhibits soot formation during the combustion process and thus, greatly reduce visible smoke emissions. It can be seen clearly that petroleum industry makes easy the release of diesel fuel associated with metals into the environment and especially water plans through field drainage, groundwater discharge, or aerosols. Sites used by the metallurgical industry as well are frequently contaminated by both metals and mineral oil (Fijalkowska et al. 1998; Riis et al. 2002) and may be also be a source of contamination of aquatic ecosystems following the same routes, notably if they are located in the coastal domain.

Author's personal copy Environ Sci Pollut Res

Clearly, up to certain point, we are faced with two contracting tasks. Through its economic interest, there is no doubt that fossil diesel plays a vital role in the generation of energy to drive mechanics. Through its environmental relations, the immediate and post-impact of oil spillage and pollution has done much to destroy aquatic lives. Non-combustible metals present in a fuel are typically released into the environment during the combustion process. With low-grade fuels, trace metals can particularly be released in significant concentrations (Rising et al. 2004). The above-mentioned problems and other facts necessitated further search or rather awakened new interest for diesel fuel and trace metals interactions in terms of toxicity for humans. To estimate health risks related to such combination of two different classes of toxicants, a set of three trace metals frequently present in the commercial diesel fuel was selected for a detailed survey: chromium, copper, and nickel. Species-specific responses to pollutant exposure are the basis of most experimental studies in ecotoxicology. It is true that there is a large amount of literature on the ecological effects of oil and trace metals on macrofaunal seafood such as fishes, crabs, lobsters, shrimps, and prawns, but limited observations have been made on meiofauna (Beyrem et al. 2007) which, as we think, should be the first step toward assessing potential risks for any stressor. Meiofauna are small metazoan organisms (1 mm–40 μm according to Vitiello and Dinet 1979) which have a distinct biological and operational identity relative to any one of the marine components. These organisms are abundant and diverse even in habitats subjected to considerable physical and chemical stresses and comprise an important link in the so-called “small food web” (bacteria, protists, and meiofauna) and in the global marine food web. They would be expected to be highly susceptible to sediment-associated pollutants because they live and feed in the sediments. Any effects of disturbances on them are likely to be passed through all the system. The short generation times of the major meiofaunal taxa, the free-living nematodes on the order of weeks (Singh and Ingole 2011; Boufahja et al. 2012), result in a faster potential response time to stress and faster bioaccumulation in them which could transfer pollutants to higher trophic levels (Hamels et al. 2001) since they are preyed upon by macrobenthic predators such as polychaetes, crabs, and fishes (Flach et al. 2002; Cruz Rosa and Bemvenuti 2005). There have been rare studies assessing combined toxicity assay of oil and metals on macrofauna seafood, probably because of their big size needing big experimental containers. However, field results showed sufficiently that contaminant interactions can affect species’ ability to survive in natural environments. The combination of phenanthrene and

copper caused synergistic increases in oxygen consumption and ammonia production in the marine mussel Mytilus edulis (Moore et al. 1984). Lemaire-Gony et al. (1995) found synergistic immuno-toxicological affects (phagocytic index and phagocytic activity) caused by combined cadmium and benzo[a]pyrene exposure in the European sea bass Dientrarchus labrax. Klerks (1999) found that grass shrimp’s (Palaemonetes pugio) physiological acclimation to metal or polycyclic aromatic hydrocarbons (PAH) stress was reduced when pre-exposed to mixtures of metals and PAH. Moreover, Klerks and Moreau (2001) showed that the heritability of resistance to metal or PAH stress in Cyprinodon variegatus decreased as the number of components in the metal–PAH mixture increased. Free-living marine nematodes are well suited to experimental studies since small sediment samples are required without preliminary sieving in the field. These worms are inherently quick to reproduce, easily maintained, and more stable, both qualitatively and quantitatively, than macrofauna (Millward et al. 2004; Mahmoudi et al. 2005). It is thus more reasonable to perceive, and monitor changes in, community structure of meiobenthic nematodes than in that of macrobenthic seafood. The main goals of this work were to: (1) compare descriptors of nematodes exposed to diesel with those exposed to diesel combined with three trace metals (chromium, copper, and nickel) (H1: no differences in term of toxicity between the two type of treatments, H2: the toxicity of diesel fuel is influenced, negatively or positively, when mixture of trace metals is added) and (2) assess possible changes in the taxonomic composition of the nematofauna in order to determine indicative species of seafood quality.

Materials and methods Field site Subtidal sediments with their natural meiobenthic assemblages were collected during 25–27 January 2006 from Sidi Salem beach (37° 17.385′ N 09° 52.289′ E) in Bizerte Bay (Tunisia) (Fig. 1). This undisturbed dissipative beach is a simple flat, generally has fine sand; with waves (N–S and NW–SE directions) breaking far from the intertidal zone and dissipating their force gradually along wide surf zones (Ben Garali et al. 2008). The suitability of this site for this kind of bioassays was validated referring to two bases: (1) the concentrations of the study trace metals (chromium, copper, and nickel) were absolutely lower than the threshold effects level (TEL) of NOAA (1999) (Table 1) and (2) the nematode community was acceptably diverse (species number018–35 according to Beyrem and Aïssa 2000).

Author's personal copy Environ Sci Pollut Res 9 °5 5 ’

Si d i Sa l e m b e a c h

Mediterranean Sea

N Sa m p l i n g s i t e

B I Z ERT E BAY

T U N I SI A

Km 0

1

2

Fig. 1 Aerial photograph of Bizerte bay (Tunisia) showing the location of Sidi Salem beach and the position of the sampling site

Collection and manipulation of sediment During the sampling days, water depth was 1.2±0.3 m; salinity was 36.5±1.1 PSU; temperature was 15.9±2.4 °C; dissolved oxygen content was 12.12±1.73 mg/L, and the sediment had a median particle diameter of 0.37±0.12 mm, with 99.65±0.24 % coarse size fraction (>63 μm) and 0.83 ±0.2 % organic matter content. To collect the sediment, Plexiglass hand-cores (10 cm2, 3.6 cm internal diameter) were used to a depth of 10 cm. The fresh sediment was transported to the laboratory, stored in two plastic containers with aeration, and then gently homogenized with a plastic spoon. Finally, the sediment was transferred to the experimental units (i.e., microcosms). Microcosms Microcosms consisted of 570 ml glass bottles (Mahmoudi et al. 2005). The same amount of homogenized fresh sediment (200 g) was added to each microcosm; bottles were carefully filled with 1 l filtered seawater (1 μm) from the native site (i.e., Sidi Salem beach, Tunisia) and left for 1 week for acclimatization before application of treatments (Mahmoudi et al. 2005). During this period, microcosms were connected to aquarium air-stone diffusers, and each one was stoppered with a rubber bung with two holes for inflow

and outflow of air. Care was taken to avoid the variability of water temperature (average 14.4± 0.5 °C at noon), saturation of dissolved oxygen (average 15.30± 2.06 mg/L), and salinity (average 37.3±0.6 PSU) into the microcosms for the duration of the experiment. Metal and diesel fuel spiking of sediments The sediment was defaunated by three successive freezing and thwarting in order to kill all present meiofauna (Austen et al. 1994; Gyedu-Ababio and Baird 2006; Hermi et al. 2009), and then it was wet sieved to remove the larger particles (>63 μm). To resolve our problem, seven treatments in four replicates were used, comprising a control, three levels of diesel fuel, and three levels of trace metals (chromium, copper, and nickel) added to the low dose of diesel fuel, hypothetically less toxic than the others (Table 1). To produce the diesel fuel-spiked sediments, appropriate doses of diesel fuel were added to aliquots of 100 g (dry mass (dm)) in order to obtain final concentrations of 1, 10, and 20 mg/kgdm after being mixed with 200 g of fresh, meiofauna rich sediment (Table 1). Chlorides of the chosen trace metals (CrCl3, CuCl2, and NiCl2) were used as contaminants in varying concentrations. Chloride form of metals was used because it is the most ubiquitous form in seawater. Metal chloride solutions

0 1 10 20 0 500 800 1300 0 700

1414 2180 0 250 550 900 500 ppm Cr+700 ppm Cu+250 ppm Ni 800 ppm Cr+1,414 ppm Cu+550 ppm Ni 1300 ppm Cr+2180 ppm Cu+900 ppm Ni 1 mg/kg Di+500 ppm Cr+700 ppm Cu+250 ppm Ni 1 mg/kg Di+800 ppm Cr+1,414 ppm Cu+550 ppm Ni 1 mg/kg Di+1,300 ppm Cr+2,180 ppm Cu+900 ppm Ni

C Di(L) Di(M) Di(H) C Cr(L) Cr(M) Cr(H) C Cu(L)

Cu(M) Cu(H) C Ni(L) Ni(M) Ni(H) Mix(L) Mix(M) Mix(H) Di(L)+Mix(L) Di(L)+Mix(M) Di(L)+Mix(H)

1307.9±102.6 1,922.4±117.0 10.2±3.3 117.8±20.5 396.4±50.9 781.7±70.6 404.5 ppm Cr+621.4 ppm Cu+154.7 ppm Ni 668.3 ppm Cr+1,335.7 ppm Cu+411.5 ppm Ni 1204.9 ppm Cr+2,053.1 ppm Cu+804.5 ppm Ni 0.82 mg/kg Di+394.1 ppm Cr+611.9 ppm Cu+132.5 ppm Ni 0.80 mg/kg Di+660.2 ppm Cr+1,318.8 ppm Cu+402.7 ppm Ni 0.77 mg/kg Di+1,197.8 ppm Cr+2,044.6 ppm Cu+790.3 ppm Ni

0.02±0.00 0.72±0.10 6.24±1.05 13.38±4.00 11.4±1.7 383.7±30.5 662.3±16.9 1,188.5±48.2 12.1±3.3 594.8±63.2

Actual

15.9

18.7

52.3

TEL

47.75±10.78 26.50±6.40 239±52.5 123.50±30.45 106.75±24.11 50±9.25

239±52.5 156.25±21.48 89±15.33 64.50±12.07 239±52.5 142.75±18.57 76.25±9.03 52.25±10.68 239±52.5 84±8.41

Abundance

C uncontaminated control, Di diesel fuel, Cr chromium, Cu copper, Ni nickel, Mix mixture of chromium, copper, and nickel, L low contamination, M medium contamination, H high contamination, TEL threshold effects level of NOAA (1999)

Targeted

Treatment

Table 1 Targeted and actual concentration of trace metals (parts per million dry mass±SD) and diesel fuel (milligrams per kilogram dry mass±SD) in microcosm sediments at the end of experiment and abundance of nematodes (individuals±SD)



were made in distilled water, and quantities of 100 gdm of sediment were contaminated with appropriate doses of chromium or copper or nickel in order to obtain final concentrations reported in Table 1 after being mixed with 200 g of fresh sediment. In contrast, 100 g of the defaunated, uncontaminated sediment was added to the 200 g fresh natural sediment in every control microcosm. Each microcosm was then gently filled with filtered natural seawater (1 μm) (Austen and McEvoy 1997; Mahmoudi et al. 2005). Toxicants were mixed into the sediment with a food blender and the amended sediment was left to equilibrate for 1 week at 5 °C before microcosms were assembled (Hermi et al. 2009). Decrease of abundance within a given disturbed community is logically occurred before the reduction of its diversity. Therefore, it was essential before species identification was commenced to realize that targeted doses can cause already changes in terms of abundance. The choice of diesel fuel doses was based on their numerical negative impact on meiobenthic nematodes from Ghar El Melh lagoon (Tunisia) (Mahmoudi et al. 2005) (Table 1). The choice of metal doses followed three assumptions: (1) In anticipation of the current study, many lethal toxicity bioassays were performed using several concentrations of metals. The targeted metal doses (Table 1) were finally chosen on the basis of ratios calculated by dividing the nematode abundance for each couple of treatments. The retained ratio had to be higher than 1.5 to separate the controls from the low doses and higher than 1 to differentiate between two consecutive doses (Table 1). (2) Approximate concentrations of copper tested for meiofauna by Austen and McEvoy (1997) were used in this study. (3) Maximal metal concentrations reported in published studies were taken into account to decide on the high doses to be used herein (Moore and Ramamoorthy (1984)—1,337 ppm Cr dm; Somerfield et al. (1994)—2,532 ppm Cu dm; Nicolaidou and Nott (1998)— 889.4 ppm Ni dm). The setup was left for 30 days as nematodes have a short regeneration time (Carman et al. 1995; Millward et al. 2004), and an aliquot of sediment from each microcosm was analyzed to determine toxicant concentrations. Trace metal analyses were carried out, using the methods described in Yoshida et al. (2002), by means of a Varian Spectra AA20 atomic-absorption spectrometer with air/acetylene flame and autosampler. Other sediment samples were preserved in a freezer at −17 °C until total hydrocarbon concentrations were determined by infra-red spectrophotometry (Danovaro et al. 1995). Processing of samples The content of each microcosm was passed through a sieve of 1 mm mesh (to remove macrofauna and detritus) using a jet of filtered water and collected in a 5-L jar. Meiofauna

was then decanted and sieved over a 40 μm mesh sieve (Mahmoudi et al. 2005). The resuspension and decantation procedure was repeated several times to ensure that no organisms were left in the beaker (Wieser 1960; Vitiello and Dinet 1979). Samples were preserved in 4 % formalin and stained with Rose-Bengal (0.2 g/L). All sorted nematodes were enumerated under a ×50 stereo microscope using a Dollfus counting box, and 100 worms (when available) from each microcosms were randomly picked out for identification. Nematodes were mounted on microscope slides after the procedure by Seinhort (1959), which permit their examination to genus level using the pictorial keys by Platt and Warwick (1983; 1988) and Warwick et al. (1998) and to species level using the NeMys Database from the Marine Biology Section of the University of Gent (Boufahja et al. 2007). Equally, we have measured in millimeters the total length (L) and the maximum width (Wd) of nematodes by using Motic Educator software in order to estimate their biovolume (V) in nanoliters by using the equation of Warwick and Price (1979)—V 0530 × L × (Wd)2. Wet mass (micrograms wet mass) of each specimen was obtained by using a specific gravity of 1.13 μg/nL (Wieser 1960) and converted to dry mass (micrograms dry mass) assuming a dry/wet mass ratio of 0.25 (Juario 1975).

Data analysis Multi- and univariate techniques were used for data analysis using the software package PRIMER v6.0 (Clarke and Gorley 2006) and STATISTICA v5.0 from StatSoft. Statistical differences among treatments were investigated with one-way analysis of variance with six parameters: abundance (I), individual biomass (bi), number of nematode species (S), Margalef’s species richness (d), Shannon–Wiener diversity index (H’), and Pielou’s evenness (J’). The uptake data were first checked for fulfillment of parametric analyses, i.e., normality and homoscedasticity, using Kolmogorov–Smirnov test and Bartlett test, respectively. When needed, data were transformed and re-checked to know if transformation improved their suitability to apply parametric assumptions. In cases of significance, paired a posteriori comparisons were performed with the Tukey’s HSD test using 95 % confidence limits. The Bray–Curtis coefficient was used as measurement of similarity. Data of abundance of nematode species were transformed as square-root in order to down-weight the contribution of dominant species to the similarity matrix. A numerical ordination plot of all nematode assemblages using non-metric multidimensional scaling (MDS) was carried on the same matrix (Clarke and Warwick 2001).

Author's personal copy

I abundance, bi individual biomass, S number of species, H’ Shannon–Wiener index, d Margalef’s species richness, J’ Pielou’s evenness

6.55±0.43 5.94±0.45 (p00.447) 4.33±0.58 (p00.004) 2.52±0.26 (p