Acute and sub-chronic toxicity bioassays of Olive Mill Wastewater on the Eastern mosquitofish Gambusia holbrooki

Ecotoxicology and Environmental Safety 175 (2019) 48–57

Contents lists available at ScienceDirect

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

Acute and sub-chronic toxicity bioassays of Olive Mill Wastewater on the Eastern mosquitofish Gambusia holbrooki

T

⁎

Ioannis Lerisa, Eleni Kalogiannia, , Catherine Tsangarisb, Evangelia Smetia, Sofia Laschoua, Evangelia Anastasopouloua, Leonidas Vardakasa, Yiannis Kapakosa, Nikolaos Th. Skoulikidisa a b

Hellenic Centre for Marine Research, Institute of Marine Biological Resources and Inland Waters, Anavissos, 19013 Attica, Greece Hellenic Centre for Marine Research, Institute of Oceanography, Anavissos, 19013 Attica, Greece

A R T I C LE I N FO

A B S T R A C T

Keywords: Bioassays Biomarkers Acute toxicity Enzyme activities LC50

Olive oil production generates large volumes of wastewaters mostly in peri-Mediterranean countries with adverse impacts on the biota of the receiving aquatic systems. Few studies have however documented its toxicity on aquatic species, with an almost total lack of relative studies on fish. We assessed the acute and sub-chronic OMW toxicity, as well as the acute and sub-chronic behavioural, morphological and biochemical effects of OMW exposure on the mosquitofish Gambusia holbrooki. LC50 values of the acute bioassays ranged from 7.31% (24 h) to 6.38% (96 h). Behavioural symptoms of toxicity included hypoactivity and a shift away from the water surface, coupled with a range of morphological alterations, such as skin damage, excessive mucus secretion, hemorrhages, fin rot and exophhalmia, with indications also of gill swelling and anemia. Biochemical assays showed that OMW toxicity resulted in induction of catalase (CAT) and inhibition of acetylcholinesterase (AChE) activities. The implications of our results at the level of environmental policy for the sustainable management of the olive mill industry, i.e. the effective restriction of untreated OMW disposal of in adjacent waterways, as well as the implementation of new technologies that reduce their impact (detoxification and/or revalorization of its residues) are discussed.

1. Introduction Olive oil production is an important industry, mostly in the periMediterranean countries (95% of global production), which, despite its economic benefits, poses a threat to the aquatic environment due to the high volume of the wastewater produced (Nogueira et al., 2015; Paraskeva and Diamadopoulos, 2006). Olive mill wastewater (OMW) is often discharged untreated or partially treated in adjacent aquatic ecosystems with severe impacts on them and their biota (Karaouzas et al., 2011b; Karaouzas, 2018). These effects of pollutants may be more severe at intermittent Mediterranean aquatic systems due to the reduced dilution capacity of the receiving waterways (Kalogianni et al., 2017; Karaouzas et al., 2018; Smeti et al., 2019). OMW composition may vary according to the olive oil extraction method used, the olive variety, as well as soil and climatic conditions and cultivation methods (Justino et al., 2012; Kapellakis et al., 2006); its typically high phenolic content and high organic load (BOD and COD) render it, however, one of the most significant pollutants for aquatic ecosystems in the Mediterranean (Pavlidou et al., 2014). In addition to its organic load, OMW is rich in ammonium and inorganic phosphorous, trace metals, such as ⁎

copper, manganese and nickel, and has high acidic and suspended solid content (Paredes et al., 2005; Pavlidou et al., 2014). In Greece, the usual treatment and disposal practice involves neutralization of the OMW with lime, to reduce acidity, and disposal in evaporation ponds/ lagoons or directly to adjacent streams (Karaouzas et al., 2011b). Disposal of OMW, however, causes significant environmental pollution with detrimental effects not only on the quality of surface water, but also on the quality of ground water and soil (Karaouzas et al., 2011b; Mekki et al., 2008); the latter may include infiltration of groundwater by phenolic compounds (S’Habou et al., 2005) and stronger water repellency and sorptive capacity for agrochemicals in polluted soils (Peikert et al., 2014). Toxicity tests have been widely applied to assess the potential toxicity of effluents to aquatic organisms, as well as to facilitate ongoing monitoring of aquatic systems receiving such discharges (for a review see Mitchell et al., 2002; Pedrazzani et al., 2019). However, the available studies on the toxic effects of OMW to aquatic organisms have mostly focused on OMW lethal effects on algae or macroinvertebrates of freshwater and coastal marine ecosystems (Fiorentino et al., 2003; Karaouzas et al., 2011a; Paixão et al., 1999; Pavlidou et al., 2014). To

Corresponding author. E-mail address: [email protected] (E. Kalogianni).

https://doi.org/10.1016/j.ecoenv.2019.03.025 Received 4 December 2018; Received in revised form 5 March 2019; Accepted 6 March 2019 0147-6513/ © 2019 Elsevier Inc. All rights reserved.

Ecotoxicology and Environmental Safety 175 (2019) 48–57

I. Leris, et al.

dilution water was allowed to stand for 30 min before the introduction of the test fish. The OMW solutions in each container were constantly agitated using air-stones connected to an air-pump (Modell 20 A-10010, Hagen, Germany).

our knowledge, only one study, demonstrated lethality and various behavioural effects of OMW exposure on fish (Adakole, 2011). Thus, there is a great gap in our knowledge on the potentially lethal effects of OMW and on its potential adverse effects on behaviour, morphology and biochemistry of fish species. Similarly, there are no studies documenting the chronic toxicological effects of OMW discharge on the native fish fauna. The limited information available on OMW effects on native freshwater fauna and the high endemicity and often threatened status of the local fish fauna (Barbieri et al., 2015; Hermoso and Clavero, 2011), renders the information gained through such tests very valuable from a conservation perspective. This information could factor significantly in environmental policy decision making and managerial measures formulation to mitigate OMW discharge impacts on the local fish fauna. In this study, acute and sub-chronic toxicity tests were used to assess lethal and sub-lethal effects of OMW exposure of the Eastern mosquitofish Gambusia holbrooki Girard, 1859, to various dilutions of OMW. More specifically, the aim was to provide basic acute toxicity data (lethal concentration 50, LC50), to investigate shifts in behaviour (swimming activity, position in the water column, stress-related behaviours), to describe morphological alterations and to assess biochemical biomarkers of oxidative stress, biotransformation and neurotoxicity in fish subjected to OMW.

2.3. Wastewater sample characterization Physicochemical parameters of the undiluted OMW and its dilutions used, such as temperature, pH, conductivity, and dissolved oxygen (DO) were measured using a Portable multiparameter Aquaprobe AP-200 with a GPS Aquameter (AP 2000, Aquaread Ltd, UK) and Turbidity with a portable turbidimeter (2100Q, Hach, USA). Major ions (calcium, magnesium, sodium, potassium, bicarbonate, sulphate, chloride), silicate, nutrients (nitrate, nitrite, ammonium, phosphate, total phosphorus), total phenols, biochemical oxygen demand (BOD5), chemical oxygen demand (COD) and suspended matter were determined. Specifically, Ion Chromatography was applied using a Metrohm Ion Chromatographer for major ions and nitrates. Nitrites, ammonium, orthophosphate, total phosphorus and COD were determined photometrically using the spectrophotometer MERCK-FARO 300. The determination of ΝΗ4+ was carried out according to Berthelot's Reaction (EPA 350.1, APHA 4500-NH3 F, ISO 7150–1 and DIN 38406–5); the determination of ΝΟ2- was performed according to Griess Reaction (EPA 354.1, APHA 4500-NO2- B,DIN EN26 777 D10); PO43- was determined using the method of phosphomolybdenum blue (PMB) (DIN EN 1189 D11, ISO 6878/1, APHA 4500-P E and EPA 365.2 + 3) and COD was determined using potassium dichromate as oxidant (DIN ISO 15705, EPA 410.4, APHA 5220 D, and ASTM D1252-06 B). Alkalinity was determined by titration using an Automatic Titrator, Tim 900 (Radiometer) and the biochemical oxygen demand (BOD5) was determined with the WTW OxiTop System. Total phenol concentration was determined using the colorimetric method 4 aminoantipyrine (Government Gazette 438, 1986; Standard Methods 420.1, 1999). After extraction with chloroform, the sample was analyzed by a UV–Vis Perkin Elmer spectrophotometer, at a wavelength of 460 nm. Water quality and chemical parameters were also measured for water samples from the fish collection site, as well as for each treatment group at the start and end of the acute test (at 0 and 96 h respectively). Values obtained were then compared to the limits for the dispensing of olive oil mill waste waters in surface waters set out for the Regional Units of Athens and Thessaloniki (Government Gazette 582/Β, 1979 and 82/Β, 1994).

2. Materials and methods 2.1. Experimental subjects A total of 102 adult mosquitofish were used for the acute and subchronic tests. The alien mosquitofish was selected as test fish species because of its ready availability, ease of maintenance, non-threatened status and wide distribution in olive oil production areas, often in cohabitation with small-bodied native cyprinids in aquatic systems of Greece and other Mediterranean countries receiving OMW effluents (Barbieri et al., 2015). Fish of both sexes were collected in February 2018 from a semi-natural population inhabiting a large artificial pond in Athens, Greece using fish nets. The fish were transferred to the laboratory and housed in three 30 L plastic containers (33 × 48 × 28 cm) equipped with air-driven sponge filters for aeration and filtration. The fish were allowed to acclimatize in laboratory conditions for approximately two weeks prior to the experiment at temperature 20–22 °C, pH 7.0–7.2, dissolved oxygen 6–8 mg/l, and natural photoperiod. During acclimatization they were fed once daily ad libidum with commercial flake food (TetraMin, Tetra, Germany). The subjects were not fed for 48 h prior to the test. Aerators were used both during acclimatization and the experiment. Mortality during acclimatization did not exceed 4%. Only males and visibly non-gravid females were used in the experiment, to avoid possible effects related to females’ reproductive stage (e.g. due to different energetic demands). The fish that were not used in the experiment remained in HCMR facilities for breeding purposes.

2.4. Pilot study A pilot study was conducted, prior to our experiments, in order to determine the most appropriate concentrations to accurately calculate the LC50 of OMW. Based on previous trials and relevant literature (Adakole, 2011), we tested five OMW concentrations (5%, 7.5%, 10%, 15%, and 20%) using two fish per concentration. At 24 h, we recorded 100% mortality at 10%, 15% and 20% OMW concentrations, and 50% mortality at 7.5% OMW concentration. No mortalities were thereafter recorded until the end of the pilot study (96 h). We therefore selected 3%, 6%, 7%, 8% and 9% concentrations for the experiment.

2.2. Olive mill wastewater sample preparation We obtained untreated OMW from an olive mill in Kalamata (Southern Greece) in late December 2017 during its operation. The solution was aliquoted in ~500 ml portions in polypropylene water bottles, and the aliquots were frozen and stored at –20° prior to the experiment. On days 0 (start of the acute test), 5 (first phase of the subchronic test), and 10 (second phase of the sub-chronic test), the appropriate amount of OMW was defrosted for solution renewal. For the toxicity tests, we used 17 transparent cylindrical 5 L tanks (diameter: 17 cm, height: 26 cm) made from food-grade plastic. The containers were initially filled with dechlorinated tap water and OMW was added to achieve the desired final concentrations, after being brought to ambient temperature and shaken well. The mixture of toxicant and

2.5. Acute toxicity bioassay The OMW toxicity on G. holbrooki was determined during 96 h of exposure (4 days) according to the OECD guideline 203 (1992) and expressed as LC50 at 24 h, 48 h, 72 h and 96 h that represent the lethal wastewater concentration in which mortality reaches 50% of the initial number of the exposed fish. OMW samples were diluted with dechlorinated tap water to achieve nominal concentrations of 3%, 6%, 7%, 8% and 9% OMW. Dechlorinated tap water without any effluent (nominal concentration of 0% wastewater) was used as control. For the tests, three replicates of each OMW concentration were used (except for 49

Ecotoxicology and Environmental Safety 175 (2019) 48–57

I. Leris, et al.

the 9% concentration where two replicates were used), with 6 fish in each replicate, at the onset of the acute test (a total of 12–18 fish at each OMW concentration). Acute toxicity values were converted into toxic units following Persoone et al. (2003), to classify the toxicity level of the agro-industrial effluent.

2.8. Sub-chronic toxicity bioassay Following the acute test, the surviving subjects in the control, 3% and 6% treatments were further exposed to their respective OMW solutions for 10 days, to examine any possible physiological effects of prolonged sub-lethal exposure. All OMW solutions were renewed with freshly prepared ones at the beginning and the middle (day 5) of the sub-chronic treatment. Exposure tanks were aerated throughout the sub-chronic exposure, and the fish were fed prior to the solution changes. We continued taking behavioural records every 24 h during the subchronic exposure test, in case there was a shift in behaviour after prolonged exposure to OMW and we also examined the containers daily for mortalities, as described above. On the final day the subjects were euthanized by immersion in iced water, body mass (TW), total length (TL) and condition factor (K) were measured/calculated and fish were swiftly decapitated under a stereoscope (5–14 individuals per treatment). The head and decapitated body of each individual fish were frozen in liquid nitrogen and stored at −80 °C for biochemical analysis.

2.6. Behavioural observations and mortality count During the experiment, the following behavioural responses were examined: general swimming activity (levels: normal/hypo-/slightly hypo-/hyper-/slightly hyperactive), position in the water column (levels: surface, column, bottom and any combination of these), presence or absence of stress-related behaviours, such as elevated breathing rate (opercular movement faster than normal), darting (rapid unpredictable movements/dashing), jumping (attempts to leap out of the water), presence or absence of gulping events (gasping for air in the surface indicating a severe lack of oxygen) and presence or absence of loss of equilibrium (when fish could no longer hold their upright position and typically fell on their side, while still breathing and reacting to touch). The overall level of each behaviour scored for each replicate, was determined by the most severe symptom observed. For example, in a replicate with six subjects, if five of them exhibited normal swimming activity but one was hypoactive, the overall activity of the group was scored as hypoactive. All OMW solutions were dark coloured and highly turbid (turbidity range 279–805 NTU) which made behavioural observations difficult, as light could only penetrate 2–3 cm (from the top, sides or bottom of the container). Thus, most behavioural observations (e.g. regarding swimming activity or breathing rate) were only possible when fish swam in one of these visible zones, an effort however was made to account for all fish present during the observation. Behavioural observations were taken at 0.5, 1, 2, 4, 6, 24, 48, 72, 96 h from the start of the acute test. During each observation bout (1–2 min per replicate), we examined the test container from all sides, the surface and the bottom. Individuals observed within the visible 2–3 cm zone in the surface, sides or bottom of the container were given a position ‘surface’, ‘column’ and ‘bottom’ respectively. Fish not seen during the observation were considered to be swimming in the water column and were given a ‘column’ position score. All possible options were selected if fish were observed in all of these positions. To accurately report mortalities at major experimental time points (6, 24, 48, 72 and 96 h), a more thorough examination was required. At these time points, after observing and scoring the behaviour of fish in each container, the air-stone was removed, and the solution was poured into a second identical container through a net to hold the fish. The fish were gently transferred to a small container with a small quantity of the respective solution to look for mortalities. A fish was considered dead when there was lack of opercular movement and no reaction when prodded with a glass probe. Loss of equilibrium was typically noted during this examination. The dead fish were removed and kept in ice for post mortem morphological and histological examination, while the solution and the surviving fish were returned into their original container.

2.9. Biochemical biomarkers analyses Activities of the following enzymes were measured in the fish samples: catalase (CAT), an enzyme of the antioxidant defence against oxidative stress that can be induced by exposure to organic contaminants and metals (Lushchak, 2011), glutathione S-transferase (GST), an enzyme of phase II biotransformation pathway involved in the detoxification of xenobiotics and products of oxidative stress (Hayes and Pulford, 1995) and acetylcholinesterase (AChE), a key enzyme of the nervous system involved in neurotransmission that can be inhibited by exposure to various chemicals (Lionetto et al., 2013). The above enzyme activities were used as biomarkers of pollutant exposure/effects (Tsangaris et al., 2016). Biochemical biomarkers CAT and GST activities were measured in the decapitated body of the fish. Tissues were weighed, cut in small pieces and then homogenized using a Potter-Elvehjem homogenizer (Heidolph Electro GmbH, Kelheim, Germany) in 1:10 (w/v) 100 mM KH2PO4/K2HPO4, pH 7.4, 1 mM EDTA. Homogenates were centrifuged at 10,000 ×g for 30 min. All preparation procedures were carried out at 4 °C. CAT activity was measured through the loss of H2O2 that was measured colourimetrically with ferrous ions and thiocyanate on a microplate reader (BIOTEK Synergy HTX MultiMode Microplate Reader) (Cohen et al., 1996). CAT activity was determined by the difference in the absorbance at 490 nm per unit of time. CAT activity results are expressed in terms of the first-order reaction rate constant (k) and protein content as follows: U/mg proteins = k/mg proteins = [ln (A1/A2)/(t2 − t1)]/mg proteins where U represents units, ln is the natural log, and A1 and A2 are the observed mean absorbance at 490 nm at two time points, t1 = 1 min and t2 = 4 min. GST was measured by the method of Habig and Jakoby (1981) with 1chloro-2,4-dinitrobenzene (CDNB) as a conjugation substrate, adapted to microplate reading by McFarland et al. (1999). Activity was expressed as nanomoles of conjugate per minute per milligram of proteins. AChE activity was measured from the head of the fish. Tissues were weighed, cut in small pieces and then were homogenized using a PotterElvehjem homogenizer in 1:30 (w: v) 0.1 M Tris–HCl buffer containing 0.1% TRITON X 100, pH 7. Homogenates were centrifuged at 10,000 ×g for 20 min. AChE activity was assayed by the method of Ellman et al. (1961), adapted to microplate reading by Bocquené et al. (1993). Enzyme activity was expressed as nanomoles of acetylthiocholine hydrolyzed per min per mg of proteins. Total protein content in the tissue extracts was measured using bovine serum albumin (BSA) as a standard (Bradford, 1976).

2.7. Morphological alterations During the acute test, morphological examinations were conducted on dead fish (total fish examined 52) that were removed from the tanks and examined under a stereoscope for skin mucus secretion, fin damage and hemorrhages. These alterations were rated on a scale of 0 (absence) to 2 (severe symptoms). Fish were also examined externally for presence of abdominal swelling and exophthalmia. Due to the small size of the fish, only a limited number of specimens (3) were dissected for examination of internal organs. 50

Ecotoxicology and Environmental Safety 175 (2019) 48–57

I. Leris, et al.

2.10. Statistical analyses

Table 1 Water quality and chemical characteristics of the fish collection site, the OMW effluent and the regional unit limits of Olive Oil Waste Water Disposal Concentrations in Surface Waters in the Regional Units of Athens & Thessaloniki.

Treatment effects on fish survival (censored at 96 h) were analyzed using the Kaplan-Meier survival test in SPSS (v.23, IBM, USA). Overall and pairwise comparisons were conducted using the Mantel-Cox log rank test. LC50 at 24, 48, 72 and 96 hrs and the associated 95% confidence limits were calculated using probit analysis in SPSS. Acute toxicity values were converted into toxic units using the formula TU = (1/LC50)× 100 (Persoone et al., 2003). In an effort to determine the variables affecting fish mortality, physico-chemical variables measured at the beginning of the experiment were first log transformed and then Pearson correlation analysis was performed to define relationships between variables. When correlations were over 99%, a variable was removed. The selected variables were then used as explanatory variables in generalized linear mixedeffects models (family binomial), with experimental replicate as a random factor and the death of a fish as a binary response variable (0 if fish alive, 1 if fish dead). Fish mortality data used in the statistical analysis were those obtained at 24 h of the OMW addition and at 96 h of the OMW addition (end of acute test), to explore the relationship of chemical variables and short (24 h) and longer term (96 h) mortalities. Model selection was based on the Akaike Information Criterion (AIC), whereby the model with the lowest AIC is retained, aiming in minimizing information loss and separating noise from structural information (Burnham and Anderson, 1998). For the selection of the best model, it is not the absolute value of AIC that is important, but rather the difference between the considered models (ΔAICi = AICi – AICmin). Models with ΔAIC < 2 are considered as equally explaining variation of the response variable, therefore, within this range, selection was based on parsimony. Another criterion is the ‘weight of evidence’ (wi) of the selected model, i.e. the relative likelihood of the model, considered the data. The closest wi is to 1, the greater the likelihood that the selected model is the best (Burnham and Anderson, 1998). The relative appearance of each variable in models that could adequately explain variation in fish death was also used as an indicator of an importance of a variable. Statistical analysis was performed with the statistical program R v.3.3.3 (R Development Core Team, 2017). Package lme4 v.1.112 (Bates et al., 2015) was used for mixed effect models and package MuMIn v.1.15.6 (Bartón, 2016) was used for the model selection. Possible group differences in total length (TL), body weight (TW) condition factor (K) and biochemical biomarkers of fish after the subchronic test were analyzed with one-way ANOVAs, followed by Tukey’s multiple comparison tests where necessary, in SPSS. Normality and homogeneity of variance were assessed using Kolmogorov-Smirnov’s and Levene’s tests respectively. Data were log-transformed where necessary. The level of significance was set at p < 0.05.

Parameters

Site

OMW

Regional Unit Limitsa

Temperature (°C) pH DO (mg/l) Conductivity (μS/cm) TDS (mg/l) Turbidity (NTU) Phenols (μg/l) COD (mg/l) BOD (mg/l) TSS (mg/l) Ca (mg/l) Mg (mg/l) Na (mg/l) K (mg/l) HCO3 (meq/l) SO4 (mg/l) Cl (mg/l) N-NO3 (mg/l) N-NO2 (mg/l) N-NH4 (mg/l) P-PO4 (mg/l) Total P (mg/l)

13 7.91 10.05 461 300 5.70 5.5 8.5 5.9 6.39 64.14 5.99 7.08 2.69 3.91 9.24 8.68 0.036 0.001 0.006 0.001 0.006

5.53 0.0 6778 4324 9211 11,100 > 10,000 3049 146.22 310 100.3 55.5 2247 24.85 120.15 421 0.112 0.150 1.4 109.63 114.8

9

a

500 180 60 1000

11.3 0.91

10

Government Gazette 582/Β, 1979 and 82/Β, 1994.

3. Results 3.1. Water quality The physico-chemical characteristics of the OMW are shown in Table 1. Comparing the values determined in the pure effluent to the upper limits of OMW disposal concentrations in surface waters in the Regional Units of Athens & Thessaloniki, the following can be noted: Total phenols far exceeded the limit value (500 μg/l) since they were recorded more than 20 times higher in the waste (11,100 μg/l). The same also applies for both BOD and COD where extremely high values were recorded indicating extensive organic load. Total Phosphorus content (114.8 mg/l) was also far exceeding the upper limit of 10 mg/l. Nitrates, nitrites and TSS did not exceed limit values. Water quality and chemical parameters obtained before (0 h) and after (96 h) the acute test (with similar values expected during the subchronic test) at the six treatment groups revealed an increase along the OMW concentration gradient (i.e. effect increasing from 3% to 9%), while a gradual decrease was found in DO and in pH.

Fig. 1. Fish survival after exposure to an OMW concentration gradient during the acute test.

3.2. Acute toxicity 3.2.1. Fish mortality Fish survival varied significantly among treatments (Mantel-Cox log rank test, χ2 = 97.87, p < 0.001, Fig. 1). The survival distribution of the control group differed significantly from all OMW groups (Mantel-Cox log rank test, χ2 > 4.14, p < 0.05) with the exception of the 3% group (Mantel-Cox log rank test, χ2 = 0.94, p = 0.33). The distribution of the 3% group differed significantly from 7%, 8% and 9% groups (Mantel-Cox log rank test, χ2 > 25.28, 51

Ecotoxicology and Environmental Safety 175 (2019) 48–57

I. Leris, et al.

there were no findings of hyperactivity and only one observation of a slightly hyperactive subject (Table 3). OMW-exposed fish were more hypoactive compared to the control individuals, with increased hypoactivity observed in all OMW concentrations, especially in the first 24–48 h. Fish became more hypoactive when they were closer to death, but in surviving fish hypoactivity remained constant or even decreased (e.g. in 7% after 48 h). In the higher OMW concentrations, individuals also started exhibiting loss of equilibrium and elevated breathing rate as they approached death (at 7% and 8% at 24–48 h). No events of gulping, jumping or darting were observed.

p < 0.001) but did not differ from the 6% group (Mantel-Cox log rank test, χ2 = 2.02, p = 0.16). The 7% group differed significantly from all other OMW groups (Mantel-Cox log rank test, χ2 > 4.90, p < 0.05). The 8% and 9% groups did not differ between them (Mantel-Cox log rank test, χ2 = 1.89, p = 0.17) but differed significantly from all other groups (Mantel-Cox log rank test, χ2 > 4.90, p < 0.05). Median lethal concentration values (LC50 and 95% Confidence Limits) of OMW calculated at 24, 48, 72, and 96 h were 7.31 (95% CL [6.65, 7.84]), 6.70 (95% CL [6.25, 7.04]), 6.44 (95% CL [6.03, 6.72]) and 6.38 (95% CL [6.05, 6.68]) respectively, indicating the high toxicity of the OMW on the tested fish species. The respective Toxic Units (TU and 95% Confidence Limits) calculated at 24, 48, 72, and 96 h were 13.68 (95% CL [12.76, 15.04]), 14.93 (95% CL [14.20, 16.00]), 15.53 (95% CL [14.88, 16.58]) and 15.67 (95% CL [14.97, 16.53]). Based on the five-class hazard classification system used for wastewaters discharged into the aquatic environment (Persoone et al., 2003), samples tested were classified as highly toxic (Class IV – high acute toxicity, 10–100 TU). There was high correlation among many variables of OMW waste and those that were correlated over 99% were removed from further analysis (K+ concentration, BOD and Turbidity). The high correlation between variables resulted in many models that could adequately explain variation, using variables interchangeably (delta < 2 in at least the first 10 models, Table 2). However, the model with the best fit was the most parsimonious, with only one variable adequately explaining variation in fish deaths. The concentration of total phenols was the most important variable explaining fish deaths, irrespective of the time of their death (after 24 h or after 96 h of OMW exposure, Table 2). Its importance was reinforced by its highest relative appearance when considering all possible models that adequately represent mortality variation (0.59 at 24 h and 0.82 at 96 h), followed by NH4 (0.58 at 24 h and 0.60 at 96 h) and PO4 concentration (0.48 at 24 h and 0.46 at 96 h).

3.2.3. Morphological alterations The fish exposed to OMW had skin lesions and damaged fins (shredded appearance) accompanied by excessive mucus covering the entire body (Fig. 2A, B). Hemorrhages were detected at the base of the soft rays of the caudal, anal and dorsal fins, and near the area of the anus (Fig. 2A, B, C). Moreover, the subjects exhibited abdominal swelling, exophthalmia and anemia in the gills (increased lamellar thickness and decreased interlamellar area, Fig. 2D, E). Spot hemorrhages were observed in kidney tissue samples obtained from the three individuals examined (Fig. 2F). A summary of all observed morphological symptoms is presented in Table 4. 3.3. Sub-chronic test Only the control, 3% and 6% groups continued in the sub-chronic test. In this test, four fish died in the control group (23.5%), two in the 3% group (11.8%), and nine in the 6% group (64.3%). In terms of behavioural responses, the swimming activity of the control fish remained normal, similar to the 3% group which, however, showed some signs of hypoactivity. The 6% group had equal observations of normal activity and slight hypoactivity as the 3% group throughout the subchronic test. Regarding the position of fish in the test tank, both the control and 3% groups were distributed in the entire water column with most observations including all three possible locations (surface, column and bottom), in contrast to the 6% group which appeared more often closer to the bottom of the tank. Only three individuals from the 3% group and one from the 6% group exhibited loss of equilibrium during this test and there were no reports of gulping, jumping and darting in any of the treatments. The breathing rate could not be accurately recorded in the OMW exposed groups.

3.2.2. Behavioural responses Concerning the relative position of the subjects in the test tank along the vertical axis, OMW exposure resulted in a general shift away from the surface. Specifically, in the control treatment, fish were more evenly distributed in the container (inhabiting all three possible locations in the tank), in comparison to the OMW treatments where they were mostly observed in the column and at the bottom of the tank. Concerning fish activity, only ‘hypoactivity’ was assessed, since

Table 2 Global model with all considered variables. Replicate is considered as a random factor (1 | replicate). The first 10 models with the lowest AIC are presented. Models selected to best explain fish deaths are in bold. Death24 correspond to deaths after 24 h, Death96 correspond to total deaths after 96 h. Global model: Death ~ (1 | replicate) + pH + DO + Conductivity + Total phenols + TSS + Cl + NO2 + NH4 + PO4

Death24 ~ (1 | replicate) + Total phenols Death24 ~ (1 | replicate) + Total phenols + NH4 Death24 ~ (1 | replicate) + DO + Cl + NH4 + PO4 Death24 ~ (1 | replicate) + DO + Total phenols + NH4 Death24 ~ (1 | replicate) + PO4 Death24 ~ (1 | replicate) + Total phenols + NO2 Death24 ~ (1 | replicate) + DO + Total phenols Death24 ~ (1 | replicate) + DO + TSS + Cl + NH4 + PO4 Death24 ~ (1 | replicate) + DO + Cl + NH4 + PO4 + Total phenols Death24 ~ (1 | replicate) + Total phenols + NH4 Death96 ~ (1 | replicate) + Total phenols Death96 ~ (1 | replicate) + pH + DO + Total phenols + Cl + NH4 Death96 ~ (1 | replicate) + DO + Total phenols Death96 ~ (1 | replicate) + pH + Total phenols Death96 ~ (1 | replicate) + TSS + Cl + NH4 + PO4 Death96 ~ (1 | replicate) + Total phenols + NO2 Death96 ~ (1 | replicate) + Total phenols + TSS Death96 ~ (1 | replicate) + Conductivity + Total phenols Death96 ~ (1 | replicate) + Total phenols + Cl Death96 ~ (1 | replicate) + Total phenols + NO2 + NH4

52

df

AIC

delta

weight

3 4 6 5 3 4 4 7 7 5 3 7 4 4 6 4 4 4 4 5

81.1 81.2 81.6 82 82.2 82.5 82.6 82.6 82.7 82.8 38.2 38.5 39.7 39.7 39.8 40 40 40.1 40.1 40.1

0 0.11 0.49 0.88 1.16 1.47 1.52 1.55 1.59 1.7 0 0.38 1.52 1.56 1.62 1.8 1.84 1.92 1.94 1.95

0.161 0.153 0.126 0.103 0.09 0.077 0.075 0.074 0.072 0.069 0.194 0.161 0.091 0.089 0.086 0.079 0.077 0.074 0.074 0.073

Ecotoxicology and Environmental Safety 175 (2019) 48–57

I. Leris, et al.

Table 3 Behavioural symptoms during exposure to OMW concentration gradient. , surface column and bottom; , column and bottom; , column; , bottom position; 0, no-; I, moderate; II, severe symptoms. nr, no record (could not be scored due to low visibility); –, no surviving subjects for observation; †, only absence (0) or presence (I) reported; * , no surviving subjects at 24 h, thus only data from 0.5, 1, 2, 4 and 6 h observations were used. OMW (%)

Surviving subjects 24 h

48 h

Position

72 h

96 h

24 h

Hypoactivity 48 h

72 h

96 h

24 h

Loss of equilibrium†

48 h

72 h

96 h

24 h

48 h

72 h

Elevated breathing† 96 h

24 h

48 h

72 h

96 h

0

17

17

17

17

0

0

0

0

0

0

0

0

0

0

0

0

3

17

17

17

17

II

I

II

I

0

0

0

0

nr

nr

nr

nr

6

15

15

15

14

II

II

II

II

0

0

0

0

nr

nr

nr

nr

7

11

5

1

1

II

II

0

0

I

I

0

0

I

I

nr

nr

8

6

1

0

0

9

0

0

0

0

–

–

–

II

II

–

–

I

0

–

–

I

nr

–

–

–

–

II*

–

–

–

0*

–

–

–

I*

–

–

–

Fig. 2. A. Excessive mucus on the skin and fins and hemorrhages on the caudal area (8% OMW exposure at 72 h). B. Excessive mucus and hemorrhages on the anus area (8% OMW exposure at 6 h). C. Hemorrhages at the base of the dorsal fin (7% OMW exposure at 24 h). D. Exophthalmia and abdominal swelling (6% OMW exposure at 96 h). E. Gill filament (6% OMW exposure at 24 h) with the lamellae anemic, distended and covered in mucus. F. Spot hemorrhages in kidney tissue (8% OMW exposure at 24 h).

During the sub-chronic test, five individuals were also examined for morphological alterations (two from 0%, one from 3% and two from 6%). Control group fish showed no evidence of skin and fin damage, but both exhibited exophthalmia and abdominal swelling. The individuals from the 3% treatment showed some signs of skin and fin degeneration and exophthalmia but no abdominal swelling. One of the 6%-individuals showed only moderate skin and fin alterations, while the other 6%-individual exhibited exophthalmia and abdominal swelling and minimal skin and fin damage. All specimens showed moderate hemorrhages on the body. Treatment groups did not differ, in terms of TL (One-way ANOVA, F2,30 = 1.06, p = 0.36), TW (One-way ANOVA, F2,30 = 1.77, p = 0.19) and condition factor (One-way ANOVA, F2,30 = 2.9, p = 0.07), after the sub-chronic exposure (n = 31).

3.4. Biochemical responses CAT activities (Fig. 3A) varied significantly between our treatment groups (One-way ANOVA, F2,30 = 6.98, p < 0.01). In particular, the CAT activity of the control group differed significantly from the 3% group (Tukey HSD post hoc test, p < 0.01), while its difference from the 6% group was marginally non-significant (Tukey HSD post hoc test, p = 0.07). The effect of treatment on AChE activities (Fig. 3C) was also significant (One-way ANOVA, F2,30 = 3.51, p < 0.05) with the highest concentration (6%) exhibiting significantly lower activity than the control treatment (Tukey HSD post hoc test, p < 0.05). GST activities showed no significant differences among treatments (One-way ANOVA, F2,30 = 1.4, p > 0.05), although at the highest OMW concentration (6%) they were slightly lower than in control fish (Fig. 3B).

Table 4 Morphological changes (post-mortem examination) during exposure to OMW concentration gradient. 0, no-; I, moderate; II, severe symptoms. The number of subjects exhibiting the recorded symptom level is presented in superscript (total number examined also presented). nr, no record (no mortalities for examination); –, no remaining subjects; †, only absence (0) or presence (I) reported. OMW (%)

0 3 6 7 8 9

Examined subjects

Skin mucus

Fin damage

Hemorrhage

Abdominal swelling†

Exophthalmia†

24 h

48 h

72 h

96 h

24 h

48 h

72 h

96 h

24 h

48 h

72 h

96 h

24 h

48 h

72 h

96 h

24 h

48 h

72 h

96 h

24 h

48 h

72 h

96 h

0 1 3 7 11 12

0 0 0 7 5 –

0 0 0 4 1 –

0 0 1 0 – –

nr I1 I3 II4 II 9 I9

nr nr nr II 7 II 4 –

nr nr nr I4 II 1 –

nr nr I1 nr – –

nr 11 II 3 II 4 II 9 II 12

nr nr nr II 7 II 5 –

nr nr nr II 4 I1 –

nr nr I1 nr – –

nr II 1 II 3 II 4 II 11 II 12

nr nr nr II 6 II 5 –

nr nr nr Ι3 II1 –

nr nr Ι1 nr – –

nr Ι1 Ι3 Ι7 Ι 11 Ι 12

nr nr nr Ι6 Ι5 –

nr nr nr Ι2 Ι1 –

nr nr Ι1 nr – –

nr Ι1 Ι3 Ι7 Ι 11 Ι 12

nr nr nr Ι7 Ι5 –

nr nr nr Ι2 Ι1 –

nr nr Ι1 nr – –

53

Ecotoxicology and Environmental Safety 175 (2019) 48–57

I. Leris, et al.

phenols to fish mortality both at the onset and the end of the acute test. Total phenol concentration in our OMW sample was indeed 20 times higher than the maximum amount levels of phenols (500 μg/l) allowed in the wastewater by the Regional Unit Limits of Athens and Thessaloniki, Greece and 10 times higher than those (1000 μg/l) by the European Union (Urban Water Directive 91/271/EC). BOD5, COD and Total Phosphorus content were also extremely high in our sample, far exceeding regional values. Karaouzas et al. (2011b) in an in situ study of OMW effects to stream macroinvertebrates and ecological status showed extremely high values of phenols, BOD5, COD, chloride and TSS during OMW discharge episodes, that caused a significant decrease in dissolved oxygen concentrations. Acute toxicity test on the Eastern mosquitofish in the present study revealed LC50 values (6.38 at 96 h) comparable to those reported by Adakole (2011) for fingerlings of Clarias gariepinus (5.54). LC50 values for the mosquitofish at 24 h (7.31%) were, however, higher than those reported for OMW toxicity in macroinvertebrate bioassays (2.64–3.36 for Gammarus pulex and 3.62–3.88 for Hydropsyche peristerica (Karaouzas et al., 2011a). Similarly, Paixão et al. (1999) reported LC50 values at 48 h of 2.24 in a Daphnia bioassay, a value lower than the relevant LC50 found in our study (6.70 at 48 h). This discrepancy between the lethal concentrations of OMW reported for fish and invertebrates is normal, as invertebrates have been previously found to be more sensitive to pollutants. For example, exposure to coal mine wastewater of a range of invertebrate and vertebrate species was not acutely toxic to the two fish species tested (Hypseleotris compressa and Pseudomugil signifier) as opposed to cladocerans (Daphnia carinata) and planarians (Dugesia sp.) that exhibited significant mortalities after 48 and 96 h exposure, respectively (Lanctôt et al., 2016). Our study has shown that the Eastern mosquitofish appears to be suitable for acute toxicity testing of OMW. However, given the adaptability of the mosquitofish to a wide range of, often degraded, environmental conditions (Vargas and De Sostoa, 1996), the possibly lower sensitivity of native species occupying similar habitats, such as small cyprinids, would require further assessment. 4.2. Behavioural responses Toxic activity of OMW was also observed at the behavioural level. Exposure to OMW resulted in hypoactivity, in contrast to previous findings where hyperactivity (‘rapid and erratic swimming’) was reported after exposure to similar concentrations of the effluent (Adakole, 2011). Exposure to phenols is known to cause initial hyperactivity possibly due to temporary irritation by the chemical (Rice et al., 1997) followed by a stage of hypoactivity, mainly attributed to nervous paralysis and tissue damage (reviewed by Alabaster and Lloyd, 1980). Similarly, Saxena et al. (2005) reported an initial increase in surface movement followed by lethargic behaviour and bottom dwelling of the related species Gambusia affinis (Baird & Girard, 1853) exposed to detergents. Moreover, fish generally reduce their activity when ambient turbidity increases, primarily to reduce predation risk in the absence of visual cues (e.g. Leahy et al., 2011). Thus the high turbidities in all OMW solutions could provide an additional explanation for the observed decrease in activity levels in the non-control treatments. Hypoactivity was more severe in the higher concentrations and as the fish approached death, but in surviving subjects it remained at constant levels or it was reduced, which could potentially indicate a form of acclimation to the presence of the pollutant. Regarding positioning in the water column, OMW exposure resulted in a general tendency of the fish shifting away from the surface. This could be partially linked to the decreased swimming activity observed in the OMW treatment groups, as hypoactive fish would typically stay at the bottom of the container, often completely immobile as they approached death. Several subjects exhibited loss of equilibrium in both the acute and sub-chronic test. This behavioural symptom typically appears as the fish approach death, and – in combination with altered swimming – it is

Fig. 3. CAT (A), GST (B) and AChE (C) activities and lipid peroxidation at the end of sub-chronic exposure (mean ± SE). * , difference compared to control treatment (p < 0.05).

4. Discussion 4.1. Fish mortality This study has shown a dose-responsive increase in mortality of the tested fish to OMW exposure (LC50 values ranged from 7.31% to 6.38%), with our analysis indicating the predominant contribution of 54

Ecotoxicology and Environmental Safety 175 (2019) 48–57

I. Leris, et al.

hydroxyl radical (HO•), a highly reactive and toxic form of reactive oxygen species (ROS) involved in oxidative stress, which may cause lipid, protein, and DNA damage (Kehrer, 2000). Changes in CAT activities are indicative of increased ROS production that can be due to several environmental factors, including exposure to chemical contaminants acting as pro-oxidants and hypoxia (Limón-Pacheco and Gonsebatt, 2009). Higher CAT activities in OMW-exposed fish compared to control fish suggest induction of the cellular antioxidant defense against ROS. ROS were possibly produced by substances with prooxidant properties present in OMW, such as phenolic compounds (Eghbaliferiz and Iranshahi, 2016). Accordingly, phenolic compounds found in OMW were considered responsible for oxidative stress effects, i.e. enhanced lipid peroxidation, in OMW-exposed mussels Mytilus galloprovincialis (Danellakis et al., 2011). Hypoxic conditions can also enhance ROS production and induce antioxidant enzymes and/or oxidative stress in fish (Lushchak, 2011; Lushchak et al., 2005; Onukwufor et al., 2017). Furthermore, chemicals and hypoxia may exert interactive effects on antioxidant enzyme activities (Dasgupta et al., 2016; Onukwufor et al., 2017). Thus, potential hypoxia caused by reduced oxygen consumption as indicated by the observed morphological alterations (but not the behavioural responses) in the OMW-exposed fish, may have additionally affected CAT activities since hypoxic conditions can enhance ROS production and induce oxidative stress and/or antioxidant enzymes in fish (Lushchak, 2011; Lushchak et al., 2005; Onukwufor et al., 2017). In a study on the effects of OMW exposure on amphipods G. pulex and caddisfly larvae of H. peristerica (Karaouzas et al., 2011a), high concentrations of phenolic compounds in OMW were regarded as responsible for GST induction; however such effects were not evident in our study. GSTs are enzymes of the phase II biotransformation pathway that detoxify a variety of chemical compounds by conjugation of toxic electrophiles with glutathione, whereas certain GST isozymes catalyze the reduction of organic peroxides and thus act as antioxidants (Hayes and Pulford, 1995). GST can be induced by organic contaminants as part of the phase II biotransformation pathway whereas GST inhibition can be regarded as a non-specific response to chemicals (Regoli et al., 2003). Our results on GST activities imply that the phase II biotransformation pathway was not induced by organic chemical constituents of OMW. In contrast, a decreasing tendency (although not statistically significant) was noted in GST activities of the exposed fish which could indicate a non-specific response to chemicals present in OMW. In the current study, AChE activities of the exposed fish decreased along the OMW concentration gradient. AChE is an enzyme involved in nerve impulse transmission, and its inhibition is an established biomarker of neurotoxicity (Fulton and Key, 2001). AChE inhibition can result from exposure to various chemical contaminants (Lionetto et al., 2013), and has been shown in aquatic organisms exposed to OMW, i.e. M. galloprovincialis (Danellakis et al., 2011), G. pulex and larvae of H. peristerica (Karaouzas et al., 2011a). AChE inhibition that has been shown to impair several behavioural processes in fish and to modify swimming performance in G. holbrooki (Sismeiro-Vivas et al., 2007) could be related to the hypoactivity observed in the OMW exposed fish in the current study.

used to characterize a subject as moribund, an alternative endpoint in acute fish toxicity tests (Dang et al., 2017; Rufli, 2012; Toth, 2000). Moribund individuals are practically debilitated and are thus expected to be more susceptible to predation in the wild, even if their symptoms are reversible (Rufli, 2012). In the acute phase, only fish from 7% and 8% treatments exhibited loss of equilibrium, however it is possible that fish in the 9% treatment also presented it (just before death) but we could not record it, as all of them died between 6 and 24 h. Interestingly, after prolonged exposure to OMW in the sub-chronic test, fish from 3% and 6% also started exhibiting this terminal symptom. This is particularly important as it highlights the possibility that OMW exposure, even in lower sub-lethal concentrations, could severely impact survival in a natural population. Although the concentration of DO in the test solutions decreased linearly along the OMW gradient, there were no observations of jumping, air gulping or gasping in the surface. We did however observe elevated breathing rates in subjects from higher OMW concentrations (7–9%). Mosquitofish have been previously reported to have extremely low oxygen demands, a characteristic that can explain their effectiveness as global invaders (Stoffels et al., 2017). Gambusia affinis, a closely related species of mosquitofish, can survive in severely hypoxic environments (0.2–0.3 mg/l O2) if given access to the surface (Odum and Caldwell, 1955), thus our DO concentrations (> 7.37 mg/l) were possibly not low enough to trigger severe hypoxia-related behaviours, even at the higher OMW concentrations (8–9%). 4.3. Morphological responses Several morphological alterations, such as skin damage, excessive mucus secretion, hemorrhages, fin rot and exophhalmia were observed in the exposed fish of the current study. Excessive mucus production is a common response mechanism in fish, as it creates a barrier between the body and the pollutant and reduces the exposure to the toxic agent, thus reducing the irritation (Kasherwani et al., 2009). It has been reported elsewhere that fish exposed to phenols show excess mucous secretion from skin and gills (Saha et al., 1999). Even at low concentrations, phenols have been found to negatively affect fish gills, kidney, liver, and skin after long-term exposure (reviewed by Alabaster and Lloyd, 1980). More specifically, exposure of sunfish, Lepomis gibbosus, to low (sub-lethal) concentrations of OMW resulted in gill changes similar to the ones presented here, and these could be attributed to the effluent’s high phenolic content (Koca and Koca, 2016). Fish gills are very sensitive to changes and hence an important indicator of waterborne toxicants, since these may cause damage to gill tissues, thereby reducing the oxygen consumption and disrupting the osmoregulatory function of aquatic organisms. The severity of damage to the gills depends on the concentration of toxicants and the period of exposure (Oliveira Ribeiro et al., 1996). Moreover, increased TSS may also cause significant gill clumping and other structural alterations (Cumming and Herbert, 2016) which could potentially account for the elevated breathing rate observed in individuals from higher OMW concentrations. Gill clumping reduces surface area and thus the respiratory capacity of the fish, decreasing thus subsequently the metabolic rate of the fish (Kasherwani et al., 2009). Saxena et al. (2005) also reported a combination of excessive mucous secretion and gill hemorrhages to Western mosquitofish exposed to detergents. Finally, the open lesions observed in the current study could provide an entry point for various pathogens, further highlighting the adverse indirect effects of OMW exposure to fish survival.

4.5. Environmental policy implications Our results suggest that toxicity studies, such as the current one, are highly relevant to environmental policy decision making, as they indicate the importance of a rigorous implementation of the prohibition of untreated OMW disposal on aquatic ecosystems and of the necessity of on-site treatment using various separation techniques for OMW detoxification (McNamara et al., 2008; Arvaniti et al., 2011). The main challenge, however, remains the revalorization of OMW residues into high added-value products (Federici et al., 2009), e.g. for heat and power generation and for the production of biofuels (Negro et al., 2017) or to amend or irrigate agricultural soils (Roig et al., 2006; Doula et al.,

4.4. Biochemical responses Toxicity of OMW was observed also at the biochemical level, as higher CAT activities and lower AChE activities were observed in fish after sub-chronic exposure. CAT is an antioxidant enzyme that detoxifies hydrogen peroxide (H2O2), the main cellular precursor of the 55

Ecotoxicology and Environmental Safety 175 (2019) 48–57

I. Leris, et al.

2017). This is a challenge due to the seasonal and small scale nature of the enterprises involved in olive oil production in Mediterranean countries, as well as to the high costs often associated with such technologies. However, the effective disposal and/or revalorization of OMW waste arises as a priority, given the huge volumes of water utilized in the industry, the amounts of OMW produced and the deleterious effects of its disposal in Mediterranean waterways.

doi.org/10.1016/0165-0270(96)00011-8. Cumming, H., Herbert, N.A., 2016. Gill structural change in response to turbidity has no effect on the oxygen uptake of a juvenile sparid fish. Conserv. Physiol. 4, cow033. https://doi.org/10.1093/conphys/cow033. Danellakis, D., Ntaikou, I., Kornaros, M., Dailianis, S., 2011. Olive oil mill wastewater toxicity in the marine environment: alterations of stress indices in tissues of mussel Mytilus galloprovincialis. Aquat. Toxicol. 101, 358–366. https://doi.org/10.1016/j. aquatox.2010.11.015. Dang, Z.C., van der Ven, L.T.M., Kienhuis, A.S., 2017. Fish embryo toxicity test, threshold approach, and moribund as approaches to implement 3R principles to the acute fish toxicity test. Chemosphere 186, 677–685. https://doi.org/10.1016/j.chemosphere. 2017.08.047. Dasgupta, S., DiGiulio, R.T., Drollette, B.D., L. Plata, D., Brownawell, B.J., McElroy, A.E., 2016. Hypoxia depresses CYP1A induction and enhances DNA damage, but has minimal effects on antioxidant responses in sheepshead minnow (Cyprinodon variegatus) larvae exposed to dispersed crude oil. Aquat. Toxicol. 177, 250–260. https:// doi.org/10.1016/j.aquatox.2016.05.022. Doula, M.K., Moreno-Ortego, J.L., Tinivella, F., Inglezakis, V.J., Sarris, A., Komnitsas, K., 2017. Olive mill waste: recent advances for the sustainable development of olive oil industry. In Olive Mill Waste. Academic Press, pp. 29–56. https://doi.org/10.1016/ B978-0-12-805314-0.00002-9. Eghbaliferiz, S., Iranshahi, M., 2016. Prooxidant activity of polyphenols, flavonoids, Anthocyanins and carotenoids: updated review of mechanisms and Catalyzing metals. Phyther. Res. https://doi.org/10.1002/ptr.5643. Ellman, G.L., Courtney, K.D., Andres, V., Featherstone, R.M., 1961. A new and rapid colorimetric determination of acetylcholinesterase activity. Biochem. Pharmacol. 7, 88–95. https://doi.org/10.1016/0006-2952(61)90145-9. Federici, F., Fava, F., Kalogerakis, N., Mantzavinos, D., 2009. Valorisation of agrotion of agrooD. 2009-2952rffluents and waste: concept, opportunities and the case of olive mill wastewaters. J. Chem. Technol. Biotechnol. 84 (6), 895–900. https://doi.org/10. 1002/jctb.2165. Fiorentino, A., Gentili, A., Isidori, M., Monaco, P., Nardelli, A., Parrella, A., Temussi, F., 2003. Environmental effects caused by olive mill wastewaters: toxicity comparison of low-molecular-weight phenol components. J. Agric. Food Chem. 51, 1005–1009. https://doi.org/10.1021/jf020887d. Fulton, M.H., Key, P.B., 2001. Acetylcholinesterase inhibition in estuarine fish and invertebrates as an indicator of organophosphorus insecticide exposure and effects. Environ. Toxicol. Chem. https://doi.org/10.1002/etc.5620200104. Habig, W.H., Jakoby, W.B., 1981. Assays for differentiation of glutathione S-transferases. Methods Enzymol. 77, 398–405. https://doi.org/10.1016/S0076-6879(81)77053-8. Hayes, J.D., Pulford, D.J., 1995. The glutathione s-transferase supergene family: regulation of GST and the contribution of the lsoenzymes to cancer chemoprotection and drug resistance part II. Crit. Rev. Biochem. Mol. Biol. 30, 521–600. https://doi.org/ 10.3109/10409239509083492. Hermoso, V., Clavero, M., 2011. Threatening processes and conservation management of endemic freshwater fish in the Mediterranean basin: a review. Mar. Freshw. Res. 62, 244–254. https://doi.org/10.1071/MF09300. Justino, C.I.L., Pereira, R., Freitas, A.C., Rocha-Santos, T.A.P., Panteleitchouk, T.S.L., Duarte, A.C., 2012. Olive oil mill wastewaters before and after treatment: a critical review from the ecotoxicological point of view. Ecotoxicology 21, 615–629. https:// doi.org/10.1007/s10646-011-0806-y. Kalogianni, E., Vourka, A., Karaouzas, I., Vardakas, L., Laschou, S., Skoulikidis, N.T., 2017. Combined effects of water stress and pollution on macroinvertebrate and fish assemblages in a Mediterranean intermittent river. Sci. Total Environ. 603–604, 639–650. https://doi.org/10.1016/j.scitotenv.2017.06.078. Kapellakis, I.E., Tsagarakis, K.P., Avramaki, C., Angelakis, A.N., 2006. Olive mill wastewater management in river basins: a case study in Greece. Agric. Water Manag. 82, 354–370. https://doi.org/10.1002/ppul.23368. Karaouzas, I., 2018a. Agro-Industrial Wastewater Pollution in Greek River Ecosystems (The Rivers of Greece) In: Skoulikidis, N., Dimitriou, E., Karaouzas, I. (Eds.), The Handbook of Environmental Chemistry 59. Springer, Berlin, pp. 169–204 (The Rivers of Greece). Karaouzas, I., Cotou, E., Albanis, T.A., Kamarianos, A., Skoulikidis, N.T., Giannakou, U., 2011a. Bioassays and biochemical biomarkers for assessing olive mill and citrus processing wastewater toxicity. Environ. Toxicol. 26, 669–676. https://doi.org/10. 1002/tox.20606. Karaouzas, I., Skoulikidis, N.T., Giannakou, U., Albanis, T.A., 2011b. Spatial and temporal effects of olive mill wastewaters to stream macroinvertebrates and aquatic ecosystems status. Water Res. 45, 6334–6346. https://doi.org/10.1016/j.watres.2011.09. 014. Karaouzas, I., Smeti, E., Vourka, A., Vardakas, L., Mentzafou, A., Tornés, E., Sabater, S., Muñoz, I., Skoulikidis, N.T., Kalogianni, E., 2018b. Assessing the ecological effects of water stress and pollution in a temporary river - Implications for water management. Sci. Total Environ. 618, 1591–1604. https://doi.org/10.1016/j.scitotenv.2017.09. 323. Kasherwani, D., Lodhi, H.S., Tiwari, K.J., Shukla, S., Sharma, U.D., 2009. Cadmium toxicity to fresh water catfish Heteropneustes fossilis (Bloch). Asian J. Exp. Sci. 23, 149–156 (https://doi.org/0971-5444). Kehrer, J.P., 2000. The Haber-Weiss reaction and mechanisms of toxicity. Toxicology. https://doi.org/10.1016/S0300-483X(00)00231-6. Koca, S., Koca, Y.B., 2016. Genotoxic and histopathological effects of olive-Mill wastewater on Lepomis gibbosus. Water Air. Soil Pollut. 227. https://doi.org/10.1007/ s11270-016-2769-3. Lanctôt, C., Wilson, S.P., Fabbro, L., Leusch, F.D.L., Melvin, S.D., 2016. Comparative sensitivity of aquatic invertebrate and vertebrate species to wastewater from an operational coal mine in central Queensland, Australia. Ecotoxicol. Environ. Saf. 129,

5. Conclusions Olive mill wastewater discharge into surface waters has severe effects on the local biota at different levels of biological organization as it was shown with the tested fish species in our study. Our study has shown a range of toxicological impacts on the tested fish species that included acute and sub-chronic fish mortality, adverse behavioural and morphological effects and changes at the biochemical level. This study fills the existing knowledge gap on OMW basic toxicity on freshwater fish and our results have clear implications for OMW potential effects on the native fish fauna. Clearly more research is needed in the shortterm and cumulative impacts of OMW exposure, however, based on our results, behavioural, morphological and/or biochemical tests appear to have implications and can inform the sustainable management of the OMW receiving waterways and the formulation of measures to mitigate the OMW adverse effects on the local fauna. Acknowledgements This research was funded by the EU Seventh Framework Programme (FP7/2007-2013) Globaqua project (No. 603629). The authors wish to thank P. Kouraklis for field assistance, A. Latsiou for help with fish maintenance, N. Kapetanaki for assistance in chemical analyses, I. Karaouzas for critically reading the manuscript and B. Zimmerman for English proof-reading of the text. HCMR had secured all necessary permits for fish collection from the Greek Ministry of Environment, Energy and Climate Change (permit 9ZE24653Π-ΖΟ6). Conflict of interest The authors declare that there is no conflict of interest. References Adakole, J.A., 2011. Toxicological assessment using Clarias gariepinus and characterization of an edible oil mill wastewater. Brazilian. J. Aquat. Sci. Technol. 15, 63–67 (https://doi.org/http://dx.doi.org/10.14210). Alabaster, J.S., Lloyd, R., 1980. Monohydric phenols. In: Alabaster, J.S., Lloyd, R. (Eds.), Water Quality Criteria for Freshwater Fish. Butterworth, Boston, pp. 103–126. Arvaniti, E.C., Zaglis, D.P., Papadakis, V.G., Paraskeva, C.A., 2011. New techniques in olive mill wastewater treatment (OMW). In: Migliorini, Paola; Minotou, Charikleia; Lusic, Drazen; Hashem, Yousry and Martinis, Aristotelis (Eds.) Book of Abstracts of International Conference on Organic Agriculture and Agro-Eco Tourism in the Mediterranean. Barbieri, R., Zogaris, S., Kalogianni, E., Stoumboudi, M., Chatzinikolaou, Y., Giakoumi, S., Kapakos, Y., Kommatas, D., Koutsikos, N., Tachos, V., Vardakas, L., Economou, A.N., 2015. Freshwater fishes and lampreys of Greece: an annotated checklist. Monogr. Mar. Sci. 8, 130. Bartón, K., 2016. MuMIn: Multi-Model Inference. R package version 1.15.6. 〈https// CRAN.R-project.org/package=MuMIn〉. https://doi.org/10.1111/j.1420–9101. 2010.02210.x. Bates, D., Maechler, M., Bolder, B., Walker, S., 2015. Fitting linear mixed-effects models using lme4. J. Stat. Softw. J. Stat. Softw. 67, 1–48. https://doi.org/10.1097/COC. 0b013e31822d9cd6. Bocquené, G., Galgani, F., Burgeot, T., Le Dean, L., Truquet, P., 1993. Acetylcholinesterase levels in marine organisms along French coasts. Mar. Pollut. Bull. 26, 101–106. https://doi.org/10.1016/0025-326X(93)90099-6. Bradford, M.M., 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 772, 248–264. https://doi.org/10.1016/0003-2697(76)90527-3. Burnham, K.P., Anderson, D.R., 1998. Model Selection and Multimodel Inference: A Practical Information-Theoretic Approach. Springer-Verlag, New York. https://doi. org/10.2307/3802723. Cohen, G., Kim, M., Ogwu, V., 1996. A modified catalase assay suitable for a plate reader and for the analysis of brain cell cultures. J. Neurosci. Methods 67, 53–56. https://

56

Ecotoxicology and Environmental Safety 175 (2019) 48–57

I. Leris, et al.

Shoqeirc, J.H., Dag, A., 2014. Characterization of topsoils subjected to poorly controlled olive oil mill wastewater pollution in West Bank and Israel. Agric. Ecosyst. Environ. 199, 176–189. https://doi.org/10.1016/j.agee.2014.08.025. Persoone, G., Marsalek, B., Blinova, I., Törökne, A., Zarina, D., Manusadzianas, L., NaleczJawecki, G., Tofan, L., Stepanova, N., Tothova, L., Kolar, B., 2003. A practical and user-friendly toxicity classification system with microbiotests for natural waters and wastewaters. Environ. Toxicol. 18, 395–402. https://doi.org/10.1002/tox.10141. Pedrazzani, R., Bertanza, G., Brnardić, I., Cetecioglu, Z., Dries, J., Dvarionienė, J., Škrbić, B., 2019. Opinion paper about organic trace pollutants in wastewater: toxicity assessment in a European perspective. Sci. Total Environ. 651, 3202–3221. https://doi. org/10.1016/j.scitotenv.2018.10.027. R Development Core Team, 2017. R: A language and environment for statistical computing. Regoli, F., Winston, G.W., Gorbi, S., Frenzilli, G., Nigro, M., Corsi, I., Focardi, S., 2003. Integrating enzymatic responses to organic chemical exposure with total oxyradical absorbing capacity and DNA damage in the European eel Anguilla anguilla. Environ. Toxicol. Chem. 22, 2120–2129. https://doi.org/10.1897/02-378. Rice, P.P.J., Drewes, C.C.D., Klubertanz, T.M.T., Bradbury, S.P.S., Coats, J.R., 1997. Acute toxicity and behavioral effects of chlorpyrifos, permethrin, phenol, strychnine, and 2,4-dinitrophenol to 30-day-old Japanese medaka (Oryzias latipes). Environ. Toxicol. Chem. 16, 696–704. https://doi.org/10.1897/1551-5028(1997) 016<0696:ATABEO>2.3.CO;2. Roig, A., Cayuela, M.L., Sánchez-Monedero, M.A., 2006. An overview on olive mill wastes and their valorisation methods. Waste Manag. 26 (9), 960–969. https://doi.org/10. 1016/j.wasman.2005.07.024. Rufli, H., 2012. Introduction of moribund category to OECD fish acute test and its effect on suffering and LC50 values. Environ. Toxicol. Chem. 31, 1107–1112. https://doi. org/10.1002/etc.1779. S’Habou, R., Zairi, Ben Dhia, H., 2005. Characterisation and environmental impacts of olive oil wastewater disposal. Environ. Technol. 26, 35–45. https://doi.org/10.1080/ 09593332608618582. Saha, N.C., Bhunia, F., Kaviraj, A., 1999. Toxicity of phenol to fish and aquatic ecosystems. Bull. Environ. Contam. Toxicol. 63, 195–202. https://doi.org/10.1007/ s001289900966. Saxena, P., Sharma, S.S., Sharma, S.S., Suryavathi, V., Grover, R., Soni, P., Kumar, S., Sharma, K.P., 2005. Effect of an acute and chronic toxicity of four commercial detergents on the freshwater fish Gambusia affinis Baird & Gerard. J. Environ. Sci. Eng. 47, 119–124. Sismeiro-Vivas, J., Abrantes, N., Pereira, J.L., Castro, B.B., Gonçalves, F., 2007. Shortterm effects of Quirlan® (chlorfenvinphos) on the behavior and acetylcholinesterase activity of Gambusia holbrooki. Environ. Toxicol. 22, 194–202. https://doi.org/10. 1002/tox.20256. Smeti, E., von Schiller, D., Karaouzas, I., Laschou, S., Vardakas, L., Sabater, S., Tornés, E., Monllor-Alcaraz, L.S., Guillem-Argiles, N., Martinez, E., Barceló, D., López de Alda, M., Kalogianni, E., Elosegi, A., Skoulikidis, N., 2019. Multiple stressor effects on biodiversity and ecosystem functioning in a Mediterranean temporary river. Sci. Total Environ. 647, 1179–1187. https://doi.org/10.1016/j.scitotenv.2018.08.105. Stoffels, R.J., Weatherman, K.E., Allen-Ankins, S., 2017. Heat and hypoxia give a global invader, Gambusia holbrooki, the edge over a threatened endemic fish on Australian floodplains. Biol. Invasions 19, 2477–2489. https://doi.org/10.1007/s10530-0171457-6. Toth, L.A., 2000. Defining the moribund condition as an experimental endpoint for animal research. ILAR J. 41, 72–79. https://doi.org/10.1093/ilar.41.2.72. Tsangaris, C., Moschino, V., Strogyloudi, E., Coatu, V., Ramšak, A., Abu Alhaija, R., Carvalho, S., Felline, S., Kosyan, A., Lazarou, Y., Hatzianestis, I., Oros, A., Tiganus, D., 2016. Biochemical biomarker responses to pollution in selected sentinel organisms across the Eastern Mediterranean and the Black Sea. Environ. Sci. Pollut. Res. 23, 1789–1804. https://doi.org/10.1007/s11356-015-5410-x. Vargas, M.J., De Sostoa, A., 1996. Life history of Gambusia holbrooki (Pisces, Poeciliidae) in the Ebro delta (NE Iberian peninsula). Hydrobiologia 341, 215–224. https://doi. org/10.1007/BF00014686.

1–9. https://doi.org/10.1016/j.ecoenv.2016.03.003. Leahy, S.M., McCormick, M.I., Mitchell, M.D., Ferrari, M.C.O., 2011. To fear or to feed: the effects of turbidity on perception of risk by a marine fish. Biol. Lett. 7, 811–813. https://doi.org/10.1098/rsbl.2011.0645. Limón-Pacheco, J., Gonsebatt, M.E., 2009. The role of antioxidants and antioxidant-related enzymes in protective responses to environmentally induced oxidative stress. Mutat. Res. - Genet. Toxicol. Environ. Mutagen. https://doi.org/10.1016/j.mrgentox. 2008.09.015. Lionetto, M.G., Caricato, R., Calisi, A., Giordano, M.E., Schettino, T., 2013. Acetylcholinesterase as a biomarker in environmental and occupational medicine: new insights and future perspectives. Biomed. Res. Int. https://doi.org/10.1155/ 2013/321213. Lushchak, V.I., 2011. Environmentally induced oxidative stress in aquatic animals. Aquat. Toxicol. https://doi.org/10.1016/j.aquatox.2010.10.006. Lushchak, V.I., Bagnyukova, T.V., Lushchak, O.V., Storey, J.M., Storey, K.B., 2005. Hypoxia and recovery perturb free radical processes and antioxidant potential in common carp (Cyprinus carpio) tissues. Int. J. Biochem. Cell Biol. 37, 1319–1330. https://doi.org/10.1016/j.biocel.2005.01.006. McFarland, V.A., Inouye, L.S., Lutz, C.H., Jarvis, A.S., Clarke, J.U., McCant, D.D., 1999. Biomarkers of oxidative stress and genotoxicity in livers of field- collected brown bullhead, Ameiurus nebulosus. Arch. Environ. Contam. Toxicol. 37, 236–241. https://doi.org/10.1007/s002449900510. McNamara, C.J., Anastasiou, C.C., O’Flaherty, V., Mitchell, R., 2008. Bioremediation of olive mill wastewater. Int. Biodeterior. Biodegrad. 61 (2), 127–134. https://doi.org/ 10.1016/j.ibiod.2007.11.003. Mekki, A., Dhouib, A., Feki, F., Sayadi, S., 2008. Assessment of toxicity of the untreated and treated olive mill wastewaters and soil irrigated by using microbiotests. Ecotoxicol. Environ. Saf. 69, 488–495. https://doi.org/10.1016/j.ecoenv.2007.04. 008. Mitchell, E.J., Burgess, J.E., Stuetz, R.M., 2002. Developments in ecotoxicity testing. Rev. Environ. Sci. Bio. 1 (2), 169–198. https://doi.org/10.1023/A:1020842718996. Negro, M.J., Manzanares, P., Ruiz, E., Castro, E., Ballesteros, M., 2017. The biorefinery concept for the industrial valorization of residues from olive oil industry. In Olive Mill Waste. Academic Press, pp. 57–78. Nogueira, V., Lopes, I., Freitas, A.C., Rocha-Santos, T.A.P., Gonçalves, F., Duarte, A.C., Pereira, R., 2015. Biological treatment with fungi of olive mill wastewater pre-treated by photocatalytic oxidation with nanomaterials. Ecotoxicol. Environ. Saf. 115, 234–242. https://doi.org/10.1016/j.ecoenv.2015.02.028. Odum, H.T., Caldwell, D.K., 1955. Fish respiration in the natural oxygen gradient of an anaerobic spring in Florida. Copeia 1955, 104. https://doi.org/10.2307/1439312. Oliveira Ribeiro, C.A., Fanta, E., Turcatti, N.M., Cardoso, R.J., Carvalho, C.S., 1996. Lethal effects of inorganic mercury on cells and tissues of Trichomycterus brasiliensis (Pisces; Siluroidei). Biocell 20, 171–178. Onukwufor, J.O., Stevens, D., Kamunde, C., 2017. Combined effects of cadmium, temperature and hypoxia-reoxygenation on mitochondrial function in rainbow trout (Oncorhynchus mykiss). Aquat. Toxicol. 182, 129–141. https://doi.org/10.1016/j. aquatox.2016.11.015. Paixão, S.M., Mendonça, E., Picado, A., Anselmo, A.M., 1999. Acute toxicity evaluation of olive oil mill wastewaters: a comparative study of three aquatic organisms. Environ. Toxicol. 14, 263–269. https://doi.org/10.1002/(SICI)1522-7278(199905) 14:2<263::AID-TOX7>3.0.CO;2-D. Paraskeva, P., Diamadopoulos, E., 2006. Technologies for olive mill wastewater (OMW) treatment: a review. J. Chem. Technol. Biotechnol. https://doi.org/10.1002/jctb. 1553. Paredes, C., Cegarra, J., Bernal, M.P., Roig, A., 2005. Influence of olive mill wastewater in composting and impact of the compost on a Swiss chard crop and soil properties. : Environ. Int. 305–312. https://doi.org/10.1016/j.envint.2004.10.007. Pavlidou, A., Anastasopoulou, E., Dassenakis, M., Hatzianestis, I., Paraskevopoulou, V., Simboura, N., Rousselaki, E., Drakopoulou, P., 2014. Effects of olive oil wastes on river basins and an oligotrophic coastal marine ecosystem: a case study in Greece. Sci. Total Environ. 497–498, 38–49. https://doi.org/10.1016/j.scitotenv.2014.07.088. Peikert, B., Schaumann, G.E., Keren, Y., Bukhanovsky, N., Borisover, M., Garfha, M.A.,

57