Development of sediment reference sample for toxicity testing using Microtox Solid Phase test and Metal Fractionation using single extractions
Chemical characterisation of pollutants using fractionation techniques and bioassays are useful monitoring tools for sediment quality assessment. However, a common criticism of sediment bioassays is the lack of an appropriate reference sediment sample which sample sediment toxicity can be comparatively assessed. In this study an approach of obtaining a reference sediment sample by cleaning the sediment samples with metals was tested. Metal fractionation was carried out by applying single extraction techniques modified from a sequential extraction scheme proposed by Tessier et al (1979). The total metal concentrations were characterised using nitric acid digestion. The sediment samples before and after the extractions were analysed using the Microtox Solid Phase Test (SPT). Comparison of total metal concentration with various sediment quality guidelines suggests that the sediments are polluted due to higher concentrations of Cu , Ni , Pb , Cd and Zn. The fractionation studies reveal that metals are contained mainly within Fe-Mn Oxide phase.The comparison of the results of the SPT with various sediment classification methods suggests that the sediments are moderately toxic to non toxic. However, the results of changes in the toxicity of sediment residues obtained after each extraction compared to unprocessed sediment toxicity results are not statistically significant. But the comparison of toxicity results of sediment residues obtained after HNO3 and NaOAc digestion with the toxicity value of replicate1 of unprocessed sediment suggests a marginal decrease in the toxicity of sediments while the comparison of toxicity values of MgCl2 , NH2OH.HCl, HNO3+H2O2 indicates an increase in the toxicity of sediment residued. The comparison of toxicity values of all sediment residues with that of replicate2 of unprocessed sediment indicates an increase in the toxicity of the sediments after extractions.
Get wisdom, get understanding. Wisdom is supreme; therefore get wisdom. Though it costs all you have, get understanding (Proverbs 4: 5,7). Then you will know the truth, and the truth will set you free (John 8: 32).
With the growing interest in the rules that govern the fate of pollutants in urban environments, the sediments of urban rivers pose a particularly challenging scientific problem as many persistent contaminants (e.g. metals, persistent organic pollutants (POPs)) tend to concentrate in river bed sediments. Therefore, the assessment of sediment quality is recognised as a critical step in understanding the risks associated with man made pollution in the riverine system (De Miguel et al , 2005). Depending upon the conditions in the river, pollutants bound to sediment may become bioavailable and impose toxicity on aquatic organisms. Chemical analysis alone is not adequate to explain effects of chemicals present in the sediment (Beg and Ali, 2008) as they do not prove that adverse effects are occurring (Luoma et al , 1995) , thus for optimal characterization and assessment of pollution , issues concerning both concentration and toxicity should be addressed (Mowat et al , 2001).
Therefore, because of the need to establish a cause -effect relationship between the concentration of pollutants and consequent environmental damage and to measure the possible synergistic effect of complex mixture of chemicals(Girotti et al , 2008), Microbial toxicity tests based on bacteria have been widely used in environmental toxicity screening due to the similarity of complex biochemical function in bacteria and higher organisms (Mowat et al , 2001) .Among the bioassays solid phase tests are useful and widely used as test organisms are exposed to whole sediments which include water soluble and non polar substances and thus offer a high relative realism for toxicity assessment of sediments. However, sediment toxicity tests require reference sediment exclusive of contaminant with similar physico – chemical characteristics as the test sediments (Guzzella , 1998).
The microtox test based on bacterial bioluminescence which uses V. Fischeri bacteria as test organism represent one of the most suitable test for sediment toxicity assessment as it can be used on extracts as well as directly to the sediment (solid phase test) ( Calace et al , 2005).
As it is now widely recognised that the total concentrations of Heavy Metals indicate the extent of contamination, but they provide little information about the forms in which Heavy Metals are present, or about their potential for mobility and bioavailability in the environment (Lake et al , 1987) , knowledge on metal speciation in the sedimentary environment may be of more importance for hazard assessment than the total metal concentrations( Farkas et al , 2007). For this reason, sequential extraction procedures are commonly applied because they provide information about the fractionation of metals in the different lattices of the sediments and other solid samples (Margui et al , 2004).
It is against this background that an investigation into establishing a reference sediment sample for solid phase bioassays was undertaken in relation to Microtox solid phase test utilising single extractions of metal fractions using -same conditions and procedures described in the sequential extraction procedure mentioned in Tessier et al 1979 .
Aims and Objectives:
The main aim of the study is to assess whether the approach of cleaning the sediment with metals using single extraction steps of sequential extraction is an appropriate alternative to develop a sediment reference sample or not.
In order to obtain a reference sample exclusive of metals, the following procedure was adopted:
Each extraction step described in the Tessier scheme was applied to separate aliquots of sediment samples using the same extraction conditions and chemicals described in the scheme (see section 3.8 for details). After the extraction step washed and dried residue sediment samples were analysed for toxicity using the Microtox solid phase test. A reduction in the toxicity could be expected as the metals were removed using chemicals. Microtox solid phase test was also conducted on unprocessed sediment so that a relative comparison between toxicity measurements could be made.
The objectives of the investigation are summarised as follows :
2.1 Urban River Sediments and Pollution:
Urban rivers have been linked with water quality issues since the nineteenth century when it was usual practice to discharge untreated domestic and industrial waste into water courses. Since then the situation has been improved due to e.g. the management curtailment of pollution at sewage treatment plants. However, because of high population densities in urban areas due to variety of sources of pollution the degradation of urban rivers is still important today (Goodwin et al , 2003).
When released into the river environment many anthropogenic chemicals bind or adsorb on to particulate matter and depending upon river morphology and hydrological conditions such particulate matter along with associated contaminants can settle out along the water course and become part of the bottom sediments(Vigano et al , 2003). Thus , sediments are considered as repositories for physical and biological debris and for many pollutants (Calmano et al , 1996).
Further more , under various physical , biological and chemical conditions (e.g. aqueous solubility ,pH, redox , affinity for sediment organic carbon , grain size of sediments , sediment mineral constituents and quantity of acid volatile sulfides) these contaminants may become bioavailable and result in a toxic impact on aquatic biota(Ingersoll et al , 1995).
Nowdays, escalating evidence of environmental degradation have been confirmed where water quality guidelines for contaminants are not surpassed but, still organisms in or near the sediments are adversely affected (Ingersoll et al , 1995).
Thus, with a vision to protecting aquatic biota, improving water quality and managing problems of resuspension and the land deposition of dredged materials, sediment quality assessment has been a crucial scientific and legislative issue in recent years. ( Calmano et al 1996 ; Nipper et al 1998).
2.2 Water Framework Directive (WFD) :
The European Union’s(EU) Water Framework Directive (WFD) which came in effect on 22 December 2000, is one of the most important pieces of environmental legislation and is likely to transform the way water quality is being monitored within all member states ( Allan et al , 2006).
The main objective of the Directive is to improve, protect and prevent further deterioration of water quality across Europe and it aims to achieve and ensure good quality status of all water bodies throughout Europe by 2015. Thus the necessity of addressing water quality issues associated with urban rivers has been increased within Member States(Goodwin et a, 2003). Under the WFD , three modes of monitoring strategies are specified and at each strategy level chemical monitoring , biological/ecological assessment , physico-chemical and hydro morphological tools have been covered to assess the water quality status of the body(Allan et al,2006).
In the WFD, EU commission places emphasis on establishing quality standards related to the concentrations of priority substances and substances which may cause harm in water , sediment or biota . (Crane , 2003).
2.3 Sediment and Pollutants Sources in Urban Rivers :
Urban river system is much more complex in its sediments and pollutant sources. Sediments may be released into urban rivers due to erosion of land surface through variety of physical and chemical processes, the rapid run off from impervious surfaces, routing through drainage network, retention tanks and winter gritting roads (Goodwin et al, 2003). These sediments may contain or associated with pollutants such as hydrocarbons , garden and animal wastes , fertilisers , pesticides , oils , detergents , deicing chemicals , street litter (Hall, 1984 ; Chapman, 1996) and trace and heavy metals (Collins et al, 2007). Moreover, Combined Sewer Overflow (CSO) events also augment the pollutant and sediment load due to its own contaminant load and the erosion and wash out of in-sewer sediments (Fierros et al , 2002). Due to the wide variety of sources and river dynamics there exist a wide spatial and temporal variation in the properties of sediments.
2.4 River Sediment Composition and dynamics :
River sediments are mainly composed of mineral particles originated from the parent rocks due to erosion process, particulate organic matter adsorbed on mineral particles or particle sized organic matter which originates from plant detritus and animal debris, adsorbed nutrients and toxic inorganic and organic pollutants (Chapman , 1996). However , with respect to their behaviour in nature , sediments can be classified in two distinctively different groups a) fine sediments with particles smaller than 50m (i.e silt and clay) and b) coarse sediments with size exceeding 50m ( i.e. sands and gravels) (Salomons et al , 1984).
The erosion, transportation and deposition of sediment is a function of river flow velocity, particle size, water content of the material (Chapman , 1996) , channel structure and degree of turbulence(Goodwin et al , 2003). Under certain hydraulic conditions sediments can be transported in suspension or by traction along the bottom which is often called ‘Bed Load’. The suspension mechanism initiates the movement of fine particles while the Bed Load causes the movement of coarse particles (Chapman , 1996). More over, within urban catchments rapid runoff and CSO events trigger river flow events with short peak times and high peak flows which step up transport of sediments and associated pollutants (Goodwin et al , 2003).
2.5 Sediment Quality Assessment:
Historically, the assessment of sediment quality has often been limited to chemical characterisation. It helps to classify what are the contaminants and what is their concentrations(McCauley et al , 2000) and it provides information about the condition of sediments and processes within them(Wolska et al , 2007). However, quantifying contaminant concentration alone can not provide enough information to evaluate adequately potential adverse effects, possible interaction among chemicals or the time dependent availability of these materials to aquatic organisms ( Ingersoll et al , 1995) because it is impractical to analyze all the compounds and their synergistic/antagonistic effects contributing to toxicity(Plaza et al , 2005). As the bioavailability of pollutants to aquatic biota and their effects on the biota is the key concern in sediment risk assessment , ecotoxicological testing (bioassays) of sediments which study the toxic effects of sediment contaminants on living organisms ( e.g. fish , plants , bacteria , algae) has been extensively used ( Rand et al , 1995).
Thus, to understand the fate of pollutants in sediments and their impacts on aquatic biota , a tiered biological and chemical assessment methods have been implemented (Calmano et al , 1996) . The sediment quality triad methodology, one of the most widely used tiered approach based on weight of evidence combines 1) Identification and quantification of contaminants (i.e. chemical analyses ) , 2) Measurement and quantification of Toxicity based on bioassays (toxicity tests) and 3) Evaluation of in situ biological effects(e.g. Benthic community structure) (Calmano et al , 1996 ; McCauley et al , 2000 ).
Principal advantages are that it can be used for any sediment type (Calmano et al ,1996) and as both biological and chemical components are used , environmental significance of contaminated sediments is addressed (McCauley et al , 2000). However the cause -effect relations are not always identified due to the synergistic/antagonistic effects of chemicals causing toxicity in sediments (Calmano et al , 1996 ; McCauley et al , 2000) . Furthermore, the assessment is very site specific and does not allow empirical calculations of chemical specific guidelines ( Mc Cauley , 2000).
2.6 Metals in Urban Sediments and Sources :
Metals are natural components of biosphere (Luoma , 1983) and they are introduced in to the aquatic environment through many lithogenic and anthropogenic sources(Zhou et al , 2008). Chemical leaching of bedrocks , water drainage basins and run off from banks are considered as the major lithogenic sources of metals (Zhou et al , 2008) while emissions from industrial processes ( e.g. mining , smelting , finishing , plating , paint an dye manufaturing) (Rand et al , 1995) and through urban sewage, house hold effluents, drainage water, business effluents , atmospheric deposition and traffic related emissions transported with storm water (Karvelas et al , 2003) are the major anthropogenic sources of metals in the aquatic environment. Upon discharge to the aquatic environment metals are partitioned between solid and liquid phase (Luoma , 1983) and eventually as a result of settling metals associated with solid phase accumulate in bottom sediments(Farkas et al , 2007).Thus , sediments are main sink of metals in aquatic environment. In brackets : minimum- maximum values ; in italic :arithmatic mean values
2.7 Toxic metals and their forms in sediments : Although some metals are essential micronutrients (e.g. Mn, Fe, Cu,Zn) , almost all metals are toxic to aquatic organisms and human health if exposure levels are sufficiently high (Luoma , 1983). Among the toxic metals cadmium, chromium, copper, lead, nickel, zinc, mercury and arsenic are of prime importance due to their association with anthropogenic inputs. Under different physical, biological or chemical conditions the toxicity of metals in sediments is a matter of bio availability (Jennett et al ,1980). Thus in order to estimate the bio availability of metals and their potential toxicity it is desirable not only to determine the total concentration but also the different chemical forms or ways of binding between metals and sediments(Albores et al , 2000).
In sediments depending upon various physical, chemical and biological conditions , metals partitioned into different chemical forms associated with a variety of organic and inorganic phases (Farkas et al , 2007).In river sediments metal can be bound to various compartments e.g. adsorbed onto clay surfaces or iron and manganese oxy hydroxides, present in lattice of secondary minerals such as carbonates, sulphates or oxides, occluded within amorphous material such as iron and manganese oxyhydroxides, complexed with organic matter or lattice of primary minerals such as silicates (Gismera et al , 2004). Due to natural and anthropogenic environmental changes these associations can be altered and metals can become more or less bio available or mobilized within different phases. These influential factors include pH, temperature , redox potential , organic matter decomposition , leaching and ion exchange processes and microbial activity(Filgueiras et al ,2002). Thus in relation to their mobility and bioavailability, in order of decreasing interest the major metal fractions are : 1) Exchangeable ,2) Bound to carbonates , 3) Bound to Fe-Mn Oxides , 4) Bound to organic matter and 5) Residual .
2.7.1 Exchangeable Metals :
In this fraction , weakly adsorbed metals retained on the solid surface by relatively weak electrostatic forces that can be released by ion exchange processes in the sediment are included(Filgueiras et al , 2002). These metals are considered the most available form of metals present in the sediments (Morrison , 1985).
2.7.2 Metals Bound to Carbonates :
Metals in this fractions are co-precipitated with carbonates which exist as cement and coating (Morrison , 1985) and this phase can be an important adsorbent for metals in the absence of organic matter and Fe-Mn oxides (Filgueiras et al , 2002).
2.7.3 Metals bound to Fe-Mn Oxides :
Metals in this fraction are associated with Iron and Manganese oxides which exist as nodules , concretion and cement between particles or simply as a coating on particles. Iron and Manganese oxides are considered as excellent scavengers of metals and are thermodynamically unstable under anoxic conditions (Tessier et al , 1979).
2.7.4 Metals bound to organic matter :
In this fractions metals associated with various forms of organic materials such as living organisms, plant and animal detritus or coatings on mineral particles are included. This fraction is considered to be less mobile due to their associations with humic substances of higher molecular weights(Filgueiras et al , 2002).
2.8 Sequential Extractions :
A sequential extraction procedure (SEP) also known as sequential extraction scheme (SES) can be used to determine above mentioned binding fractions of metals in the sediment. In this process, given sediment sample is subjected to a series of increasingly strong , phase specific reagents under controlled condition which extract our metals from the particular physic-chemical phase of interest(Bird et al , 2005).
Depending upon target fractions, a wide variety of chemical extractants can be used (see fig.2.1) and thus in the literature various sequential extraction schemes are available which differ in the use of extractant, target phase and the order of attack to separate particular form of metals. The majority of the schemes are variants of a scheme proposed by Tessier et al (1979). Many researchers have reported difficulties in comparing the results of SES due to their wide variation in the use of chemicals and target phase. Thus, in an effort to harmonize the different methodologies and to make the comparison of results easier , Community Bureau of Reference (BCR) proposed a three step extraction procedure along with a reference sediment material to certify the protocol (Mossop and Davidson , 2003).
2.9 Advantages and problems of sequential extractions :
The application of sequential extraction techniques , though time consuming provide valuable information about the origin , mode of occurrence, biological and physic-chemical availability , mobilisation and transport of metals within the sedimentary matrices(Tokalioglu et al , 2000).However, since their initial development, sequential extraction schemes have been criticized for the lack of selectivity of reagents, issues of re adsorption and redistribution of metals solubilised during extraction and changes in speciation due to sample pre- treatment and its general methodology ( Gleyzes et al , 2002).
In the sequential extraction scheme, the reagents are expected to attack only the target phase without solubilising the other phases. However, it has been found that the reagents are not selective and may affect other phases also. Thus the sequential extractions are termed as operationally defined fractionation techniques. This lack of selectivity may cause re-adsoprtion and re distribution of metals among the target phases. Moreover, incomplete dissolution of some phases and changes in pH may also lead toward re adsorption and redistribution problems (Gleyzes et al, 2002). Various researchers have reported the problem of re adsorption and redistribution for many sequential extractions for each phase.
Despite these limitations sequential extractions are widely accepted for metal fractionation in sediment samples to assess the mobility and bioavailability of metals .
2.10 Single Extractions :
To reduce lengthy procedures and thus making sequential extractions a part of routine analysis, various alternatives(e.g. microwave heating and ultrasonic shaking) to conventional extraction procedures have been employed (Albores et al , 2000). One of the alternatives to reduce the lengthy and laborious sequential process is to use single extractions. In single extractions the same reagents and operating conditions as the sequential extractions are applied to different sub- sample (Albores et al ,2000) and, except for first step , the metal concentrations in each individual step can be obtained by subtracting the results obtained in two successive steps(Filgueiras et al , 2002). Initially this technique was suggested by Tack et al (1996) in which first three steps mentioned in Tessier’s Scheme were extracted simultaneously while, for organic matter bound metals, it was suggested that the sample should be extracted first for reducing metals and should then be re treated with hydrogen peroxide step to remove organic matter and thus release metals bound to this phase.
2.11 Bioassays : A useful monitoring tool
Bioassays measure changes in physiology and behaviour of living organisms resulting from stress induced by biological or chemical toxic compounds which can cause disruption of e.g. metabolism. Thus, bioassays help to establish cause / effect relationship between the concentrations of pollutants and consequent environmental damage (Girrotti et al , 2008).
Historically fish and macro invertabrates bioassays are the first in the series of toxicity bioassays involving animals. As these bioassays were found useful in assessing the acute toxicity of chemicals and effluents and often predicted their effects of aquatic biota and habitat, they have been extensively used in the screening of chemicals and regulatory compliance monitoring (Blaise et al , 1998). However , these conventional bioassays require longer test duration along with additional time(e.g. acclimatisation) for preparations of the test (Ribo and Kaiser , 1987). Moreover toxicity was found a trophic level property and thus it was realized that protection of aquatic resources could not be ensured by conducting bioassays solely at macro organism level (Rand et al , 1995).
Therefore an urgent requirement of cost effective , multi trophic and faster bioassays was strongly felt which led to development of micro scale testing procedures involving bacteria , protozoa , micro algae and micro invertabrate (Blaise et al , 1998). Distinct advantages of microbial testing procedures include :1) ease of handling ,2 ) short testing time , 2) reproducibility of results (Mowat et al , 2001) and 4) cost effectiveness (Wadhia and Thompson , 2007).
2.12 Sediment Toxicity Tests :
As Van Beelen (2003) stated, toxicity is not a substance property only , but it is the combination of the substance , the organisms , the conditions and the exposure duration that can cause toxic effects. Thus on the basis of this basic principle sediment toxicity tests can be classified according to : 1) test end points , 2) test organisms and 3) routes of exposure (Nipper et al , 1998) .
According to test end points most sediment bioassays can be classified as acute (having a short period of exposure from hours to days) or chronic (having longer period of exposure from days or ,weeks to months) types ( Burton , 1991 ; Nipper et al 1998).With a view to identifying polluted areas , acute tests can be used as screening tools in the first tier of a risk assessment while chronic tests can be employed in later stages to evaluate the long term effects of contaminants on organisms (Nipper et al , 1998).
Depending upon the goals and stages of assessment a wide variety of organisms have been utilised within sediment bioassays. A complete list has been compiled by other authors (e.g. Nendza , 2002 ). The majority of tests have utilised bacteria, rotifers, amphipods, insects, polychaetes, crustaceans, bivalve, echinoid and fish (Nendza , 2002).
According to the routes of exposure or test phases sediment bioassays can be classified in four major groups : 1) Elutriate tests (Water extractable) , 2) Extractable (with solutes other than water) , 3) interstitial or pore water and 4) whole sediment or solid phase tests (Burton , 1991 ; Nipper et al ,1998).
Each type of test has its own strengths and weaknesses (see appendix 1) . Elutriates may represent only a part of multiple sources of contamination due to different levels of solubility of each contaminant in water. Moreover, water elutriation could underestimate the types and concentrations of bioavailable organic contaminants present as many organic contaminants are not water soluble (Ronnpagel et al , 1995). Solvent extract tests are useful in screening the sediments for the presence of toxic chemicals but these tests do not provide a realistic assessment of sediment toxicity to benthic biota as the extraction procedures can release the contaminants from the sediments which are otherwise not bioavailable(Nipper et al , 1998). As pore waters are considered as major route of exposure to many contaminants to some organisms, toxicity tests involving sediment pore waters have been widely used. However their sensitivity may be meaning less relative to other exposure routes due to manipulation and laboratory artefacts (Chapman et al , 2002a).
Whole sediment tests offer much more realism and ecological relevance compared to other test as the organisms are directly tested against the sediments(Burton , 1991). The solid phase tests appreciate the toxicity due to soluble /insoluble and organic/inorganic material without extraction (Calace et al , 2005) and as the test provide direct contact between the test organisms and sediment particles , it increases the probability for the measurement of responses to particle bound and marginally soluble toxicants(Qureshi et al , 1998) . However, they present a series of limitations due to sediment typology, loss of organisms which can lead to an overestimation of sediment toxicity due to sorption of bacteria on particles during the tests(Calace et al , 2005).
2.13 Sediment Toxicity Tests and Problem of Reference Sediment :
In conventional sediment risk assessments, the toxicity of test sediment is compared to that of reference sediments or to a reference condition as this would allow an assessment of whether the chemicals present in the sediment pose a hazard or not (Chapman et al, 2002b). Moreover , as test organisms are sensitive to the sediment properties (Van Beelen ,2003) it is required to distinguish the response of the test organism to the sediment properties along with the associated contaminants. Thus, a source of representative, uncontaminated and non toxic sediment is essential for the sediment toxicity assessment (Suedel and Rogers , 1994) .
A reference sediment may be defined as a sediment possessing similar characteristics (e.g. pH, redox potential, particle size distribution and percent organic carbon) as close as to the test sediment but without chemicals that might be a problem ( Burton Jr. et al , 1992 ; Chapman et al, 2002b) . This reference sediment can be used as an indicator of sediment conditions exclusive of the specific pollutant(s) of interest and provides site specific basis for comparing the results of test sediment with that of the non toxic sediment (Lamberson et al , 1992).
As no natural sediment is likely to be totally uncontaminated and have the same characteristics as the sediment being assessed, obtaining a reference problem for comparison is an inherent problem with sediment toxicity testing (Beg and Ali , 2008).
Ideally the reference sediment is collected from a nearby unpolluted area near to the site of concern. The potential advantages of field collected sediment as reference sediment are: a) sediment properties and characteristics are close to the test sediments and b) preparations are not time consuming (Suedel and Rogers , 1994). However , field collected sediments may contain pollutants other than the pollutants of concern which may show back ground toxicity to the test organism and thus lead to false positive results for toxicity comparisons. In case of highly urbanised catchments, it is particularly difficult to find a nearby clean area for reference sediment as there are chances that the whole catchment is heavily polluted (e.g. River Brent which passes from highly urbanised catchment ). Moreover , Walsh et al (1991) noted inconsistent compositions among samples collected at different time and locations which makes relative comparison more difficult.
As a contribution to addressing the issue of representative reference sediment sample for toxicity assessment, formulated reference sediment samples were developed (Suedel and Rogers , 1994).With the aim of matching the physical and chemical characteristics of natural sediments , formulated sediments are prepared using various combinations of sand , silt and clay sized particles , organic matter and calcium carbonate (Still et al , 2000). To optimise the artificial sediment’s representativeness and quality, artificial sediments are preconditioned for each constituents (Verrhiest et al 2002). Gonzalez (1996) has also tried to improve the ‘natural characteristics’ of artificial sediments through the addition of components such as Acid Volatile Sulfide (AVS) to artificial sediments.
There fore, artificial sediments can offer several advantages over field collected sediments which include a) absence of background contaminants, b) well characterised and reproducible composition and c) absence of indigenous biota (Burton , 1996). However, the principal limitation with artificial sediments is to match the organic carbon content qualitatively, key factor affecting the fate and kinetics of sediment bound materials and thus bioavailability (Suedel and Rogers , 1994).
A possible solution to the problem of reference sediment could be provided if naturally contaminated sediment can be ‘cleaned’ through the removal of pollutant(s) and then tested for toxicity testing. This technique could help retain the sediment physical and chemical characteristics of the natural sediment in the reference sediment with the exception of the contaminant(s) of interest.
In an experiment of involving the development of a non toxic reference sample, Kwan and Dutka(1996) washed natural field collected sediment with water until a negative response was obtained in the monitoring bioassay (Toxi-chromotest) . They found the sediment sample non toxic at the ratio of 1:5 (sediment: ToxiChromotest test reaction mixture). In another experiment Beg and Ali (2008) extracted organic contaminants from two different sediment samples using solvents of varying polarity in Soxhlet extraction. The extraction was done overnight using hexane followed by 2nd overnight extraction using dichloromethane which further followed 3rd overnight extraction using methanol. The toxicity of both sediment was samples analysed before and after the extraction . A drastic reduction in toxicity of PAHs rich sediment sample was observed while the toxicity of metal rich sediment which were extracted for PAHs reduced marginally after the extraction.
Thus, with a view to establishing a reference sediment for robust toxicity testing previous studies suggest that sediment can be washed for the contaminant of interest using chemical extraction processes and a non toxic reference sediment for the toxicity comparison can be obtained.
Therefore , in an experimental design , sediments contaminated with metals could be treated with chemicals and conditions applied in sequential extractions of metals which extract out particular form of metals from the sediments . As the bioavailability of metals is dependent on the metals forms , after extraction of these metals from the sediment a reduction in the toxicity of sediment could be expected and a reference sediment sample for robust sediment toxicity analysis by washing/cleaning to remove all bio available forms of metal from the sediment, established.
2.14 Bio Luminescence based Bacterial Bioassays :
As sediment micro organisms are crucial for the biodegradation of organic matter and the cycling of nutrients and while these microorganisms are susceptible to toxic pollutants(Van Beelen , 2003) , monitoring microbial responses has been recommended as an early warning indicator of ecosystem stress and a means of establishing toxicant criteria for terrestrial and aquatic eco systems (Burton , 1991).
Bacterial bioassays can be grouped in five major categories : 1) Population growth , 2) Substrate consumption , 3) respiration , 4) ATP luminescence and 5) Bioluminescence (Parvez et al, 2007). Since bioassays based on bioluminescence are rapid , sensitive , reproducible and cost effective and more over they provide an easy record of the effects produced on a living organisms , they are often chosen as the first screening method in a test battery encouraging their extensive use in aquatic toxicity tests. The most suitable species for bioluminescence tests are vibrio fischeri (v. fischeri) , vibrio harvey (v.harvey) , p. leiognathi and pseudomonas fluoresence (Girotti et al , 2007).
Bioluminescence assay based on v. fischeri has been reported as one of the most sensitive across a wide range of chemicals , compared to other bacterial assays such as Nitrification Inhibition , Respirometry , ATP lulminescnece and enzyme inhibition(Girotti et al , 2007). In this assay a suspension of v.fischeri bacteria in saline water is exposed to chemical of interest and the decrease in light output of its natural luminescnece is measured to estimate the toxic effects of chemical(Kaiser , 1998) .
Several commercial test kits such as MicroTox (Azure Environmental) , Lumistox (Dr. Lange GmbH, Berlin , Germany) and biotox (Bioorbit , Turku , Finland) are available (Kaiser , 1998) . Moreover a version of v.fischeri test called Deltatox has been also developed for field testing(Wadhia and Thompson , 2007).
2.15 Biochemical mechanism of Luminescence in Vibrio Fischeri :
In luminescent organisms, light emission usually results from an interaction between the enzyme luciferase, reduced flavin and a long chain aldehyde in the presence of oxygen and constitutes part of the cell’s electron transport system. The emission of light depends upon on this flow of electrons and therefore the level of light output reflects any changes in the metabolic activity and health of the organisms (Ribo and Kaiser , 1987) .
Reduced flavin mononucleuotide (FMNH2) plays a key role in the bioluminescence reaction. Flavin mononucleotide(FMN) is reduced to FMNH2 upon reaction with the reduced form of nicotinamide adenine dinucleotide phosphate ( NAD(P)H ) in presence of flavin reductase enzyme.
NAD(P)H + H + FMN NAD(P) + FMNH2
Reduced FMNH2 gets oxidized into FMN and H20 upon reaction with molecular oxygen in the presence of aldehyde and luciferase enzyme . In this reaction blue green light of wavelength 490nm is emitted.
FMNH2 + 02 + R- CHO FMN + H20 + R-COOH + light
2.16 MicroTox Test System :
Since its development by Beckman Instruments , Microtox has emerged as the most popular aquatic bio assay due to its advantages as mentioned previously. The test uses a non pathogenic naturally luminescent marine baterium v. fischeri (Strain NRRL B -11117). It is a short term acute toxicity test which measures the reduction in bioluminescence of the bacteria upon exposure to toxic substances and express the toxicity as EC50 (Effective Concentration : concentration which causes a 50% reduction in the level of bioluminescence) values measured at 5 , 15 , 30 minutes . (Qureshi et al , 1998).
The microtox test system (appendix 2) includes four toxicity tests : 1)The microtox acute toxicity tests , 2) The microtox solid phase toxicity test , 3) The microtox chronic toxicity test and 4) the Mutatox Genotoxicity test(Johnson et al ,1998).
2.17 Comparison of MicroTox with other bioassays :
The microtox test has been used and compared with other toxicity bioassays in a number of studies. A description of all studies which have compared the microtox test with at least one other acute toxicity bioassay is the scope of this work. A summary of correlation co-efficient of microtox test results with three common acute bioassays (e.g. Fathead minnows , Rainbow Trout and Daphnids) has been given in the table below.
As the correlation coefficient indicates the degree of similarity between the two datasets , the good correlations of microtox test results with other test indicates same or increased sensitivity of microtox compared the three bioassays.
MATERIALS AND METHODS
3.1 Study Area and Sample Collection :
The study area is the River Brent which flows through north-west London. It is a minor tributary to the River Thames and is 17.9 miles long. It is a highly urbanised catchment and has undergone many periodic modifications for the prevention of flood. After second worldwar it was channelized in U-shaped concrete channel and thus had lost almost all of its wild life and the characteristics of a natural river. A river restoration project has been initiated in 1999 to restore the river in 2 km section of the river within Tokyngton Park (Wembley , North London).
The river upstream of the park is surrounded by heavy vegetation and industrial estates. The North circular road is located just down stream of the sampling location and further up-stream is the Great Central Way(major London road) and the Northern Line tube line. Mitchell Brook is the upstream tributary to the site which drains water from nearby residential estates (St. Raphael’s). The River receives variety of pollution loads due to treated and untreated sewage, urban road runoff in it which carries a wide variety of pollutants within it(see section 2.3). Surface sediment samples were collected from an area of deposited sediment (sediment bar) located within the restored section of the river. Samples were collected using a plastic scoop and transferred to plastic bags and were frozen until analysed.
3.2 Sediment Drying :
The frozen samples were defrosted overnight at room temperature and dried in oven at 50C for 24 hours period or until the cracks appeared in the samples. After drying the crucibles containing sediment samples were allowed to cool in the dessicator.
3.3 Sediment Sieving and sample storage:
The dried sediments were ground using a pestle and mortar and any large surface debris were removed from the sample. Sediments were sieved to collect the <1 mm sediment fraction. All sieved sediment samples were stored in plastic bag at 4C.
3.4 Chemicals and Reagents:
Analytical grade chemicals and reagents (supplied by Fisher Ltd.) were used for the extraction of metals. The required concentrations of chemicals (see Table 3.1) were prepared on a daily basis. Deionised water (obtained from Milli Q filtration system) was used for dilution and the preparations of all solutions.
3.5 Laboratory Glass ware and Equipments:
All glassware and equipment used in the extraction of metals were washed in a 10% nitric acid bath, rinsed with deionised water. The equipments were dried in an oven at 30C.
3.6 Nitric Acid digestion:
To determine the environmentally available metals, a strong acid digestion method described as below was used for the metal release.
Two replicates of 10g sediment were divided in to subsamples of 5 g sediment .The subsamples were transferred to 100ml teflon beakers to which 50 ml of concentrated nitric acid was added. Beakers were left on a sand bath at 80-110C overnight. Following digestion 1% nitric acid solution was added and the samples were filtered using Whatman filter No.41 and the filtrate was collected in 100ml volumetric flask. The volume of the filtrate was made up to 100 ml using 1% nitric acid and stored at 4C prior to analyze by ICP-OES.
3.6.1 Preparation of Sediment Residue samples for Microtox test :
The sediment residues from the subsamples of each replicate were then transferred to centrifuge tube and washed with 64 ml of deionised water. The samples were centrifuged at 3000 rpm for 30 min. After the centrifuge the samples were collected in crucible and were dried at 50C for 24 hours period. The dried samples were stored in plastic bags at 4C for microtox solid phase test analysis.
3.7 Metal Speciation using single extractions: For the first three fractions of the Tessier Scheme , single extractions were carried out on separate subsamples using the methodology described in Tessier’s Scheme. For organic matter bound metals extractions the two step method described by Tack et al was employed. In this method the sediment samples were first treated for Fe-Mn oxides bound metal extraction and the residue of the samples were then treated for organic matter bound metal extraction using HNO3/H2O2 step.
Except for exchangeable and organic matter bound metal, the metal content corresponding to carbonate and Fe-Mn Oxides bound metals were calculated by subtracting the results obtained in the consecutive steps.
The extractions were carried out in 75 ml polyethylene tubes. As the volume of the tubes was not sufficient to accumulate the amount of reagents required, two replicates of 10g samples were divided in subsamples of 5g samples. Morever , during the collection of sediment residue after extraction , some amount of samples loss was observed. Thus to compensate the amount of sediment loss during collection , for Fe-Mn Oxides and organic matter bound fractions two replicates of 12 g sample were subdivided in aliquots of 3 g sample.
After single extractions the subsamples were centrifuged at 3000 rpm for 30 min. The supernatant liquid was separated from the solid phase and for the adjacent subsample of each replicate, it was collected in a single volumetric flask of either 100 ml or 250 ml size.
3.8 Inductively Coupled Plasma -Optical Emission Spectrometry (ICP-OES):
Samples were analysed for eight heavy metals (Cd, Cr, Cu, Pb, Fe, Mn, Ni, Zn) using Perkin Elmer Plasma 40 ICP-OES instrument. The details of the procedure are given as below:
3.8.1 Stock Solutions and Standards Preparations:
To prepare standards for each metal, from 1000ppm stock solutions of all 8 metals , 10 ml of the stock solution was pipetted out into 100ml flask to prepare a stock solution of 100ppm concentration. From these 100 ppm stock solutions 1ml , 0.5 ml and 0.1 ml solution of each metal were transferred into 100ml flask and made up to the required mark to obtain the multi element standards of 1000ppb , 500ppb and 100 ppb for the metals.
To obtain matrix matched calibration curves, standard solutions were prepared using the same chemical/reagents present in the analyte (e.g. for exchangeable metal analytes , standards were prepared using MgCl2). Calibration blanks were also prepared using the same chemicals/reagents as the analyte.
3.8.2 Calibration of instrument:
To calibrate the instrument for measurement of the eight heavy metals , the elements were selected and by running 1000ppb standard solution , the wavelength of each element was calibrated and the measurement of the emission was adjusted at the peak of the emission line . Using an artificial intelligence algorithm , the background corrections were calculated by the computer software automatically and these background corrections were subtracted from the total emission at the wavelength of measurement for each element. The wavelengths and background correction details are summarised in
Table 3.2. Once the wavelength calibration of all metals was completed, the standards including blank were run to obtain the calibration curve and the emissions for each element were recorded by the computer.
3.8.3 Analysis of Samples: Samples were run in the instrument to determine the concentrations of elements in it. Between samples deionised water blank was run to reduce the chance of carry over from the previous sample.Where concentrations exceeded the highest standard, appropriate dilutions were made.The concentrations in the analytes were obtained in g/l .
From the concentrations obtained in analytes using ICP-AES, the final concentrations of the metals per gram of sediment(dry weight) were calculated as follows :
Final concentration in analyte (g/l) = ICP conc. In sample (g/l) X DF
Where DF ( Dilution Factor) = final volume of dilution sample analysed in ICP(ml)/ volume of sample taken for dilution(ml)
Conc. In sediment sample (g/g) = (final conc. X total volume of analyte) / (sediment weight X 1000)
3.9Toxicity Analysis of Sediments:
The toxicity analysis of the bulk sediment and residue sediment samples from each metal extraction step was carried out using Microtox Solid Phase Test(SPT) protocol. Microtox analyzer (model 500) connected to Microtox data collection and reduction system through an IBM compatible computer was used to generate and process the data.
3.9.1 Reagent , Solutions and Accessories :
The Microtox SPT toxicity tests were carried out reconstituting a freeze dried strain of marine bacterium v. fischeri ( NRRL number B-11177). This reagent approximately contains 108 bacteria and 2% NaCl in it. The microtox diluent, a non toxic solution to test organisms which contains 2% NaCl was used to reconstitute the bacteria and the same solution was used to prepare serial dilutions for phenol standard test. For solid phase test , solid phase diluent was used for serial dilutions of the sediment suspension solutions. The test were carried out using cuvettes supplied by Microbic Ltd.(Carlsbad , CA) , Solid phase tubes and filter columns. Micro pipetters of 0-20l , 0-1000l and 1-5 ml were used to prepare the dilutions and a water bath at 15C was used for the incubation of the bacteria with solutions.
3.9.2. Microtox Analyzer :
The SDI model 500 analyzer is a dual purpose instrument which serves both as an incubator and luminometer. The incubation is carried out at two temperatures : 1) the thirty cuvettes : located on the luminometer as rows A through F and columns 1 to 5 (used for test samples) are incubated at 15C and 2) the reagent well : used for storing one stock culture cuvette of luminous bateria is incubated at 5C (Johnson et al , 2005).
The luminometer contains a photomultiplier tube that measure the light emission from bioluminescence bacteria. The analyser was operated under standard working conditions using a PC containing Microtox Omni software package.
3.9.3. Phenol Standard Test :
A reference test was conducted using phenol as a reference toxicant. To prepare the phenol standard approximately 0.050 g of crystalline phenol was added to volumetric flask of 500 ml and diluent was added up to the mark. The solution was mixed well by inverting the flask several times and it was covered with aluminium foil to protect the phenol standard from light. This phenol standard has a (EC50)5min of 13-26 mg/lit . A phenol standard was run prior each day’s analysis.
3.9.4 Solid Phase Test :
The sediment residues from the single extractions were washed with deionised water and centrifuged at 3000 rpm for 30 min. The supernatant was then discarded and the sediment residue for each replicate was collected in crucible and dried at 50C for 24 hours. After drying the sediment residue samples were stored in plastic bags at 4C for MicroTox analysis.
To analyze the toxicity of sediment samples using solid phase test, 7g of sediment sample was weighed carefully. To this sample 35 ml of solid phase diluent was added and a sediment suspension was prepared in a disposable beaker. This sediment was stirred using magnetic stirrer for 10 min to allow homogenized mixing .1500L of sample suspension was transferred to a series of solid phase tubes and twelve 1:2 serial dilutions of the suspension were prepared including two controls. Both the controls and serial dilutions were prepared in replicates .Dilutions and controls were prepared in solid phase tubes placed in a water bath at 15C. 20L of reconstituted bacteria were transferred to each solid phase tube and bacteria were incubated in the tubes for 20 minutes. Filter columns were inserted and the bacteria along with solution were filtered out. From this filtered solution 500L of solution was transferred to cuvettes placed in the microtox analyzer and luminescence readings were obtained to generate EC50 values.
Al most all metals were extracted above the detection limit of the ICP-OES instrument. The concentrations of metals in the both sediment samples ranged from 3.01 g/g to 3.9 g/g for Cd , 9.0 g/g to 23.00 g/g for Cr , 102g/g to 130 g/g for Cu , 11850 g/g to 23900 g/g for Fe , 328 g/g to 397g/g for Mn , 9900 g/g to 12400 g/g for Ni , 167 g/g to 332 g/g for Pb and for Zn 465 g/g to 574 g/g. In statistical analysis of data , standard deviation of the finite population is used to measure the variability of the data from the mean of the population and along with mean it is reported .It provides useful information about the degree of variability of two data sets with similar means However , in case of variables which are measured on incomparable scales relative standard deviation which is the ratio of standard deviation to mean , is calculated to examine the variability of the data set(Moore and Cobby , 1998).
The relative standard deviations (RSD) for Cd , Cu , Mn , Ni , Zn were 10.61% , 9.32 % , 7.68 % , 7.26% and 8.39% respectively . While for Cr, Fe and Pb the relative standard deviations were 37.06% , 30.90% and 25.10% respectively. The relative standard deviation for the five metals ( Cd, Cu, Mn, Ni,Zn) were found within acceptable range while for Cr , Fe and Pb RSD values indicates high variability in the measurements of the metals in the sediment samples. This variability can be attributed to many factors which include a) difficulties in obtaining representative samples, b) contamination of instrument and apparatus used in the analysis due to presence of elements in the atmosphere and c) impartial digestion of particular elements from the sediment matrices due to certain forms of metals which are difficult to put in solution.
As the sediments are collected from highly polluted urbanised area which has varying sources and input of pollutants to the river course, there are chances of wide variations in sediment metal concentrations and thus obtaining a representative sample might be a principal factor leading towards high variability in the results.
4.1.1 Relative abundance to metals :
The relative abundance of metals in increasing order is Cd< Cr< Cu < Pb < Mn < Zn < Ni < Fe with Cd the least abundant metal and Fe the most.The range of concentrations and sequence of relative abundance of metals in the sediment of Brent river at the sampling location reveals similar pattern observed in the urban rivers in European Union and UK(see Table 2.1). However, direct comparisons with different studies may be complicated due to the dissimilarity in digestion procedures and intensity of pollution discharges into the rivers (De Miguel et al 2005). Moreover, the sediment typology, river hydrodynamic conditions and geographical conditions of the river catchments highly influence the concentration of metals in the river sediments.
4.1.2 Comparison with Sediment Quality Guide lines(SQGs) : As sediment quality guidelines can provide scientific benchmarks , or reference points for evaluating the potential for observing adverse biological effects in aquatic systems(CCME ,2001) , a relative comparison of the chemical concentration of the pollutant with the guidelines is recommended in screening level risk assessment. However for metals in sediment, various guidelines are available which differ in their method of deriving sediment quality assessment values. To make a good comparison with selected guidelines, a comparative analysis of metal concentrations in the sediment samples with selected guidelines values has been presented in table 4.2. While comparing the metal concentrations with Sediment Quality Guidelines (SQGs) set by US DOE , it is found that with the exception of Cr and Mn, all other metal concentrations exceeded the Threshold Effect concentrations (TEC) . The concentration of Cu and Ni exceeded Probable Effect Level Concentration (PEL) and High No Effect Concentrations(NEC) which indicate that adverse effects are likely to occur on the aquatic ecosystem of river sediments due to these metals. The concentration of Pb was also found higher than NEC concentration indicating a risk of adverse effects in the sediments.
The comparison of sediment metal concentrations with Ontario’s guidelines also followed similar pattern. Except for Mn and Cr for all metals Lowest Effect Level concentrations(Low) were exceeded while for Cu and Ni Severe Effect Level concentration(Severe) were also exceeded . This comparison indicates that toxic effects might become apparent and might have affected the benthic organisms in the sediments due to Cu and Ni.
When comparing with Dutch Intervention values , it was found that all metal concentrations were higher than Dutch Target Values except for Cr metal indicating a risk of metal pollution . The concentration of Ni was found far higher than the Dutch Intervention values (11050 g/g in sediments compared to 210 g/g Intervention values). For Fe also similar trend was observed while comparing with Dutch Intervention values but Iron is not considered as potentially toxic element. The concentrations of Cu and Pb in the sediments were close to the Dutch Intervention values indicating possible pollution of sediments due these metals. Thus , according to Dutch intervention values sediments were found to be polluted due to high concentrations of Cu ,Fe , Ni and Pb. For Mn and Zn the comparison with Dutch guidelines could not be made as no target and intervention values are available for these metals.
The Interim Sediment Quality Guideline Level (ISQGL) set in Canadian SQGs were exceeded for Cd , Cu , Zn , Pb and for Cd , Zn and Pb even the Probable Effect Level(PEL) were also exceeded indicating possible adverse effect on ecosystem life might occur in sediments due to these elements.
Thus, comparison with various sediment quality guidelines indicates that for Cu ,Ni ,Pb ,Cd and Zn the higher threshold levels are exceeded for one of the guidelines and thus there are chance that adverse effects are likely to occur on aquatic ecosystems associated with river sediments due to these metals. While the concentrations of Cr and Mn were within the guideline limits posing no possible or severe threat due to these metals and Fe is not considered as toxic metal thus the higher concentration of Fe might not pose any threat to the aquatic ecosystems.
4.1.3 Association of metals and Source Identification:
In order to understand behaviour, origin and transport of metals within riverine environment, correlation statistical analysis is applied to the total metal concentrations (Farkas et al , 2007).Various researchers ( Farkas et al 2007 ; Camusso et al 2002 ; Zheng et al 2008; Yalcin et al 2008 ) have used correlation analysis to identify the associations of metals and their relations in the sediments.
Correlation co-efficient is the measure of strength of relationship between two or more variables (Ott , 1988). The value of correlation coefficient lies between -1 to +1. The positive values indicate that one variable tends to increase while the other increases. On the other hand negative values indicate that one variable tends to decrease while the other variable increases. In statistical analysis for environmental data sets, Pearson’s correlation coefficient and Spearman’s Rank coefficient are widely used. Pearson’s correlation coefficient is used for normal data . However when the data is non-normal , the approach is to rank the data set and then on the ranked data set Pearson’s correlation coefficient is calculated which is then called Spearman’s Rank Correlation Coefficient . To estimate the normality of data , a probability test was conducted on the data sets and the probability plot of the total metal concentrations is plotted in the following figure (Fig. 4.1).
Looking at the P – values , it is evident that the data is loosely normal for Cd, Cr, Cu, Pb and certainly non-normal for Fe(p Value < 0.005). While for Nickel the data has been emerged as normal . The p values for Managanese (0.062) and Zinc (0.083) also provide weak evidence against the data to be considered as normal. Thus , considering the data to be non-normal , to evaluate the relationship between the metal concentrations in the sediment spearman’s rank correlation on the metal concentrations was performed and the correlation coefficient matrix is presented in Table 4.3. Depending upon the calculated values of correlation coefficients , Moore and Cobby (1998) suggested that a correlation coefficient value < 0.6021 provides no meaningful evidence of any association. The authors further suggested that a coefficient between 0.6021 to 0.7348 would provide some evidence of association while a coefficient in the range between 0.7348 and above would suggest a strong association.
Thus , based upon above range of classification , three distinct groups of metals having strong associations can be identifies as :1) Cu-Fe-Mn-Zn (r2 range 0.796-0.979) , 2) Ni-Cu-Mn (r2 range 0.754- 0.979) and 3) Pb-Fe-Zn (r2 range from 0.723-0.802)and thus suggesting similar sources and behaviour patterns for the associations of these metals. The correlation coefficient of 0.636 between Cd and Pb is significant suggesting some relationship between the two metals indicating similar sources for these two metals. With this exception poor correlation coefficients of Cd and Cr were found with other metals indicating these metals were derived in the sediments from different sources compared to other metals.
As it is well recognised that metal pollution has a diffuse (non-point source) nature and due to the complex nature of association of metals, it is difficult to characterize the sources of individual metals or group of metals from the above correlation analysis. The distinct groups of associations indicates that metals in the sediments might have been derived from multiple sources within the urban environment which include CSOs, un treated waste water discharges , urban road run-off, roof run-off , combined and separate residential sewage flows and industrial waste water released in to the river (Thevenot et al , 2007). Moreover , automotive pollution is considered as one of the major source of Pb, Zn,Cu and Cd pollution in the urban aquatic environment (Rose and Shea , 2007). The correlations between Pb,Zn and Cu and between Pb and Cd also support that the automotive pollution might be a major source of pollution in the river sediments. The usage of River Brent as a receiver for treated and untreated discharges and the proximity of the sampling location to motorway , residential and industrial areas also support the source identification analysis made above.
4.2 Metal Fractionation using single extractions : All metals were detected in each fraction except Chromium in carbonate fraction. Certain difficulties were observed while quantifying the metals extracted with NaOAc. Each time the analyte sample was introduced into ICP instrument , the plasma torch was shifted to ‘Switch-OFF’ mode which made the analysis difficult.
This problem of ‘Switching-OFF’ of plasma was associated with the matrix effect of NaOAc which caused change in the plasma operating conditions. As sodium has low ionization potential , analytes containing sodium can originate matrix effect inside the plasma and/or in the liquid sample introduction system . The presence of sodium can change the plasma local temperatures and electronic density as well as the spatial distribution of the emitting species (Maestre etl al , 2002). Similar types of problems are thought to have arosed during the current analysis due to sodium metal and subsequent malfunctioning nebulizer.
To address this issue, samples were analysed using Flame Atomic Absorption Spectrometry(FAAS) but the results obtained with AAS were not compatible due to non-linearity of the calibration curves. However , after one month when the ICP instrument was serviced , the samples were again analysed in the ICP and an attempt was made to quantify the metals. But during the trial and error runs of the samples in ICP and AAS much amount of sample was lost and in diluted samples Chromium was found below detection limits.
In the above table, the amount of metal extracted with MgCl2 represents the ‘exchangeable’ metals . To calculate the amount of ‘carbonate bound’ bound metals , the metals extracted with MgCl2 were subtracted from the metals extracted with NaOAc . Similarly to obtain the amount of metal associated with ‘Fe-Mn Oxides(defined as reducible fraction)’ , the metal concentrations obtained using NaOAc extractions were subtracted from the NH2OH.HCl extracted metal concentrations. For the H2O2/HNO3 step, the extraction was carried out on the residue sediments of the NH2OH.HCl extraction step , therefore the amount of metal obtained using this two step extraction procedure is taken to represent the ‘organic matter bound’ or ‘oxidisable’ fraction of metals.
4.2.1 Partitioning Patterns of Metals in different fractions: Partitioning of the eight metals in all four operationally defined fractions is given in the fig.4.2 below. Each fraction is presented as the percentage of the sum total of all fractions. From the partitioning pattern, it is evident that Cd is the only metal associated with exchangeable and carbonate fractions in higher amount compared to other metals. As these fractions are considered as weakly bound and thus might become bioavailable rapidly (Jain , 2004). But as the amount of Cd available as in these fractions is below the guideline values given in the table 4.2 , it can be concluded that Cd might not pose any harm to the aquatic life in the river sediments.
The fractionation profile of Cr indicates that the metal is mainly partitioned between Fe-Mn Oxides and Organic matter bound phase. Cu shows the highest association with Organic matter with 92% of metal extracted in this fraction.
The association of Cr with organic matter can be attributed to the sewage outfalls and industrial discharge. In a speciation study for Thames river estuary O’ Reilly Weise et al (1997) has found similar association of Cr with organic matter bound fraction and the pollution of estuary through sewage outfalls and various industrial sewage sources. The higher proportion of Cu with organic matter also can be explained with the sewage discharges in the River Brent which carries organic matter in it favouring the intake of Copper into organic matter bound fraction through formation of organic complexes of this element. Baruah et al (1996) and Morillo et al (2002) have found similar results for Copper association with organic matter bound fraction in river and estuary sediments receiving high amount of sewage discharge.
From the partitioning pattern it is evident that the Fe-Mn oxide fraction is the dominant fraction carrying maximum amount of metals in it except for Chromium and Copper .Bird et al (2005) suggested that metals originating from anthropogenic sources are mainly partitioned in non-residual phases in the sediments and thus the associations of the metals with Fe-Mn oxides bound and organic matter (for Cr, Cu , Pb) fractions indicates anthropogenic pollution of sediment. These results are in accordance with the similar research carried out in various European rivers polluted with heavy metals (e.g. Farkas et al 2007 ; Klavins et al 2000 , Filgueiras et al 2004 and Relic et al 2005).The speciation pattern of metals strongly indicates that Fe-Mn Oxides are acting as major sinks of metals and thus contain most of the metal within this fraction. However , depending upon the redox potential and pH changes this fraction might become mobile thus bio available to aquatic biota(Jain ,2004).
4.2.2 Comparison of sum total of fractions with total metal digestion :
While comparing the sum total of metals extracted in the four stages of extractions with nitric acid digestion, it was found that all most all metals except Mn, Pb and Zn were extracted in significant higher amount in nitric acid digestion(see Table 4.1 and 4.5).These differences can be accounted to the fact that the nitric acid digestion is not a complete digestion procedure. Similar results for Pb and Cu were observed by Tack et al(1996) for aqua regia digestion and sum total of single extraction obtained using Tessier Scheme. In another experiment Sastre et al (2002) compared aqua regia and nitric acid digestion for total metal analysis of Cd, Cu, Zn and Pb and he observed that nitric acid digestion could led to underestimation of Zn in the samples. However , he concluded that for samples containing higher organic matter nitric acid digestion can be an alternative for aqua regia digestion but for samples containing lower organic matter and carbonate content there are chances of underestimation or over estimation of metal contents in the samples. Thus , the low organic matter and carbonate content of the sediments might have caused the discrepancies in the metal results obtained using sequential extractions and nitric acid digestion.
4.3 Sediment Toxicity Results: The results of all reference toxicity tests conducted using Phenol as a reference toxicant were found within the limits of IC50 5min 13-26 mg/l which indicates that correct test protocol was followed and the system was working satisfactorily. As Doe et al (2005) recommended R2 value of ? 0.9 for Solid Phase Test(SPT) results , We can conclude that the test results of both the replicates were satisfactory.
The mean EC50 value of the two sediment replicate was 12.84 g/l with a standard deviation of 8.84. The high standard deviation indicates significant variation in the toxicity results of sediment sample obtained using SPT.
For toxicity analysis of sediments using SPT , sediment composition has been found to be the most influential factor affecting the SPT results and many researchers (Benton et al 1995; Ringwood et al 1997) have reported false positive and negative results for SPT due to variation in sediment particle composition. Ringwood et al (1997) has observed loss of light output due to adherence of bacteria to silt particles indicating higher toxicity in the sediments. Therefore there are chances that the difference in the toxicity results might have been originated from the difference in the sediment composition of the two replicates and thus problem of obtaining a representative sample might have caused the possible variations in the toxicity results of the sediment samples.
4.3.1 Sediment Classification on the basis of Toxicity Results : To classify the sediment samples on the basis of toxicity results obtained in Solid Phase Test (SPT) , Kwan and Dutka (1995) suggested classification of the sediments as presented in Table 4.7.
Comparing the unprocessed sediment samples at mean EC50 values of 12.84 g/l (which is 1.284% of the sediment sample) with the above scheme, the sediment sample can be categorised as non-toxic. In comparing the individual EC50 values of the sediment replicates , replicate1 can be categorised as moderately toxic (an EC50 value of 6.585 g/l (0.658%) of sediment sample) and replicate2 can be categorised as non toxic (an EC50 value of 19.090 g/l (1.9 %) of sediment sample).
Another classification method used by Environment Canada as described by Doe et al(2005) , suggest that if EC50 values are <1000mg/l then the sample should be considered as toxic . While for the samples having EC50 values ?1000 mg/l the guidelines suggest that it should be compared with a clean reference sample and if the test sample EC50 values are 50% less than the clean reference sample then the test sample can be considered as toxic sample. However , as a reference sediment sample was not available , the second guideline could not be applied to this sediment toxicity study.
Using the various approaches described above, it can be concluded that the unprocessed sediment samples assessed within this study can be considered moderately toxic to non toxic.
4.4 Toxicity Results of the Sediment Residue of Single Extractions: The results of Solid Phase Test(SPT) carried out on the sediment residue of single extractions are presented in Table 4.7. Except for sediment residue replicate1 of oxidisable metal extracted using H2O2/HNO3, for all other sediment residue samples, the R2 values obtained were reasonable though not meeting the criteria of R2?0.90 for all replicates.The R2 value obtained for ‘oxidizable’ metal extracted sediment residue was 0.2207 which cannot be accepted as lower R2 value represents manual errors in conducting the test and thus the EC50 values obtained can not be considered as a valid estimation of toxicity of the sample.
4.4.1 Evaluation of change in the Toxicity after Extraction of Metals: In sequential extraction schemes, reagents applied at each stage extract out metals associated with particular metal binding fraction which may impose toxicity on the aquatic environment. Thus, after each single extraction step as metals associated with each fraction were removed, a reduction in the toxicity of the sediment residue could be anticipated.
However, comparing the mean EC50 values of sediment residue with the mean EC50 values of unprocessed sediments samples, a reverse trend was observed. The comparison of mean EC50 value of MgCl2 treated sediment residue (3.83 g/l) with unprocessed sediment EC50 values (12.84 g/l) indicated an increase in toxicity of the residue sediment.The comparison of mean EC50 values of NaOAc treated sediment residue(8.57 g/l) also revealed an increase in the toxicity of sediment residue after extraction process. Similarly the comparison of EC50 values of NH2OH.HCl (2.4735 g/l) , H2O2/HNO3(3.463 g/l) and HNO3 (7.553 g/l) treated sediment residue also showed an increase in the toxicity of sediment residue after the extraction process. The box plot (Fig.4.3) of the EC50 values of unprocessed sediment and residue sediment samples after each extraction step showed the same trend found while comparing the mean EC50 values of sediment residues with the EC50 values of unprocessed sediment sample.The lower locations of mean lines of EC50 values for sediment residues compared to EC50 values of unprocessed sediment indicates an increase in the toxicity of the sediment residue samples. The box plot also reveals that the increase in the toxicity of MgCl2 and NH2OH.HCl treated sediment residue is quite higher compared to other sediment residues treated with NaOAc and HNO3.
Furthermore , as there was a large difference between the EC50 value of two replicates of unprocessed sediment sample ,it would worthwhile to compare the toxicity of sediment residue samples after single extractions with the replicates of unprocessed sediment individually using individual value plot (Fig.4.4) .
Looking at the individual value plot , it is evident that toxicity of all sediment residue is increased compared to rep2 of unprocessed sediment. The comparison of toxicity values of MgCl2,NH2OH.HCl and NH2OH.HCl+H2O2 treated sediment residue to toxicity value of rep1 of unprocessed sediment indicates an increase in the toxicity. While HNO3 and NaOAc treated sediment residue shows a marginal decrease in the toxicity compared to toxicity of rep1.
In statistical analysis in order to assess the significance of the difference between the three or more samples, analysis of variance (ANOVA) is used. It is a single test of significance which helps to minimize the Type-I error rate which otherwise might be high in case of increasing number of two sample t-tests while comparing more than three sample means (Le Blanc , 2004). The randomized data, normality assumption and equal variance assumption are fundamental assumptions for ANOVA. However, in case of non parametric data sets a non-parametric Kruskal-Wallis(KW) test can be performed and this test is less sensitive to non equal variances than F-test used for ANOVA. This test procedure tests the hypothesis that the population medians are equal versus not equal (Le Blanc , 2004).
The probability plot of EC50 value data of all sediment samples is presented in Fig.4.5.
The p-value (0.037) of normality test of the data set indicates that the data is non-normal as the p -value is less than 0.05 . Thus, in order to assess the significance of difference between the population means of EC50 values of sediment samples, KW test was performed and the results of the test are presented in table 4.9.
The test procedure of KW test is similar to Mann-Whitney test , the data are first ranked together and then the calculations are carried out on the ranked data to produce necessary statistical results ( Siegel and Morgan , 1996).
The P-value (0.146) of the test suggests that there is insufficient evidence that the population medians of the EC50 values of different sediment samples differ statistically. Though the statistical test results are not significant the observed increase or decrease in the toxicity of sediment residue samples can be contributed to many factors.
As during the sequential extraction schemes , due to rigorous extraction conditions (e.g. pH, temperature) the equilibrium within the sediment is modified releasing toxic substances which might become bioavailable causing toxicity to the test organism and thus might have increased the toxicity of the sediment residues after MgCl2,NH2OH.HCl and H2O2/HNO3 extractions.
Moreover, sequential extractions are criticized for re adsorption and redistribution of some metals due to their incomplete dissolution and pH changes but it would not be important enough to question the results of the sequential extraction (Gleyzes et al 2002). However , as perfect knowledge on the effects of reagents on each phase during single extraction is not available (Gleyzes et al ,2002) , there are chances of re adsorption and redistribution of metals in the sediment residues after single extractions which require further investigation on these(re adsorption and redistribution) phenomena in single extraction schemes.
Furthermore, sediments are a complex matrix which varies in its physico-chemical properties with depths (Chapman , 1995) and distance from one location to another location. Thus, there are chances that the sediment samples might have a wide variation in the composition of toxic substances in it which might become bio available after the single extraction procedures and thus causing increase or decrease in the toxicity of the sediment residues.
CONCLUSION AND RECOMMENDATIONS FOR FURTHER WORK:
5.1 Metal Concentrations:
The relative abundance of metals in the sediment in increasing order found to be: Cd< Cr< Cu < Pb < Mn < Zn < Ni < Fe. The comparison of total metal concentrations with various sediment quality guidelines suggests that the threshold effect levels set by various guidelines above which adverse effects are likely to occur in the sediment are exceeded for Cu , Ni , Pb , Cd and Zn. This comparison demonstrates that the sediments are polluted due to these metals and raises concerns about their adverse effects on aquatic ecosystems of this part of the river. But as these guidelines are based on the total metal concentrations rather than the concentrations of most bio available fractions of metals and more over the bioavailability of sediment contaminants is influenced by various factors, there are chances of false positive and negative conclusions (Burton , 2002). Thus, the evaluation of biological effects on aquatic biota (e.g. benthic community characterization) is required to confirm whether adverse effects have been occurred on aquatic biota or not .
The correlation analysis identified three distinct groups of metals 1) Cu-Fe-Mn-Zn , 2 )Ni-Cu-Mn and 3) Pb-Fe-Zn . As the correlation between the metal concentrations indicates similar behaviour and origin, these three associations of metals indicates that instead of single source contributing to metals in the sediments, there might be many sources which influx the metals in the sediments supporting the hypothesis that metals have diffuse source of pollution in the urban aquatic environment and could be originated from many point and non point sources of pollution. The metals could have been derived from CSOs, un treated waste water discharges, urban road run off, combined and separate residential sewage flows and industrial waste water releases.
5.2 Metal Fractionation:
The partitioning pattern of the metals obtained using single extractions indicates that except Cr and Cu all other metals are contained within the Fe-Mn Oxides phase in the range of 70-94% of total metals extracted using the single extraction steps of Tessier’s sequential extraction scheme.This fraction is considered as less mobile compared to exchangeable and carbonate phase and act as a sink for the metals. 56 % of extracted Cr and 92 % of extracted Cu are contained within the organic matter bound phase indicating sewage outfalls as their major sources in the sediments. Except Cd the amount of metals contained within exchangeable and carbonate phase is less than 10%. However 12% of extracted Cd contained within Exchangeable phase and 14% within carbonate bound phase. As metals associated with exchangeable and carbonate fractions are considered as rapidly bioavailable and the Fe-Mn Oxides and organic matter have a scavenging effect on metals(Jain ,2004) , the less amount of metals associated with exchangeable and carbonate fractions indicates that the metals are less susceptible bioavailability while the higher concentrations of metals in Fe-Mn oxides and organic matter fractions indicates scavenging effects reducing the bioavailability of the metals. Thus , though the total metal concentrations are exceeding the guidelines for some of the metals , the metal fractionation patterns indicates that sediment might be less susceptible to metal toxicity due to their less availability in the most available fraction and scavenging effects of Fe-Mn Oxides and organic matter.
5.3 Toxicity Results for unprocessed sediments and change in toxicity of sediment residues:
The comparison of the results of Microtox SPT with various sediment classification methods indicates that the sediments are moderately toxic to non toxic. The toxicity rating suggests that the sediments may or may not pose harm to the aquatic ecosystems. However , the test results could not help to identify particular cause of toxicity in the sediments. The integrated analysis of total metal concentration, fractionation studies and toxicity testing indicates that though total concentrations are exceeding in the sediments, they are not indicative of adverse effects as the toxicity tests suggest moderate to low toxicity of the sediments. Furthermore the results of fractionation studies also indicates that due to scavenging of metals in relatively less available fractions , metals might be minor contributors to sediment toxicity and there are chances that some other pollutants might be contributing to the toxicity. As toxicity is a trophic level property , a battery of toxicity tests representing multiple trophic levels is further recommended to evaluate the adverse effects on aquatic biota in the sediments.
Though the Kruskal-Wallis test results of EC50 values of sediment are not statistically significant to assess the difference in sediment toxicity due to extraction of metals, the comparison of toxicity value of replicate1 of unprocessed sediment with the toxicity values of HNO3 and NaOAc treated sediment residues indicates a reduction in toxicity while toxicity of MgCl2 , NH2OH.HCl and NH2OH.HCl+H2O2 treated sediments indicates increase in the sediment toxicity. The comparison of toxicity value of replicate 2 of unprocessed sediment with the toxicity values of all sediment residues obtained from the single extraction indicates an increase in the sediment toxicity after metal extraction. However, it was identified that problem of obtaining a representative sample might be affecting the overall trend of the test results; the experiment can be carried out with larger sample sizes which helps in deriving statistical conclusions for the reduction/increase in the toxicity of the sediment residues after single extractions.
5.4 Recommendations for further research work:
The experiment could be performed with larger sample populations for toxicity test results so that statistical inferences can be made from the test results and toxicity data representative of the sediments can be obtained.
One possible approach is instead of cleaning the sediments for one particular group of pollutants (e.g. metals or PAHs) , using various chemical extraction techniques (e.g. metal extractions, solvent extraction using solvent of increasing polarity) the sediment can be cleaned for all possible pollutants. Using these chemical extraction techniques all possible pollutants can be extracted simultaneously from the sediment retaining the basic properties of the sediments which are exclusive of these pollutants. The sediment residue after these chemical extractions can be tested for toxicity and the results of these toxicity tests could be used as reference toxicity value for comparison in the toxicity tests. However, in order to assess the effect of these extractions on sediment properties, an analysis of sediment properties which include particle size characterization, pH, redox potential, CEC, organic matter content could be performed on the sediments before and after extraction.
A professional writer will make a clear, mistake-free paper for you!Get help with your assigment
Please check your inbox
I'm Chatbot Amy :)
I can help you save hours on your homework. Let's start by finding a writer.Find Writer