Project Topics
Get Complete Project Material File(s) Now! »
Chemical assessments of soil metal availability
Finding which chemical method better re ect available metal pools for earthworm has been the subject of many studies. One of the main approaches was to correlate internal metal concentrations with available metal contents. In this review, we collected the results obtained in the literature with such an approach (Table 1.1). We indicated for each article whether a correlation could be nd, the strength of the correlation, the earthworm species under study, whether the correlation was observed in situ or in the laboratory, on how many sites/plots and a brief description of the chemical method conducted.
Metals readily available in the soil solution (dissolved metals, pore water, water-soluble) or weakly bound to soil particles (exchangeable) were often found correlated to metal concentra-tions in earthworms (Table 1.1). Water and weak salt extractions were therefore recognized as pertinent indicators of environmental availability relatively to earthworm. However, Table 1.1 also shows that the correlations between earthworm metal concentrations are often stronger with soil total metal concentrations than with easily extractable ones. Moreover, within a same study, the relationships are often speci c of a given metal, or only signi cant at a given site. Finally, several publications have not successfully related easily extractable metal con-centrations with internal metal contents in earthworms. These contrasted results point to the fact that the chemical assessments of weakly bound metal pools are not fully generic to explain internal metal concentrations in earthworms. No clear pattern can be distinguished either according to the species or metal considered. However, total metal concentrations are very often better descriptors of internal concentrations.
Several studies reported relationships between internal concentrations and DTPA- or EDTA-extractable amounts of metals (Kamitani and Kaneko, 2007; Dai et al., 2004; Lee et al., 2009). Using a sequential extraction procedure, Becquer et al. (2005) showed that internal Cd, Pb and Zn were related to organically-bound metals and bound to Mn and Fe hydroxides. Moreover, recent studies have developed a procedure of extraction that simu-late the enzymatic gut content of earthworm (SEG: simulated earthworm gut test). They showed good relationships between SEG-extractable metals and internal concentrations in earthworms (Ma et al., 2009; Gaw et al., 2012; Smith et al., 2010). The extent of the uptake of metals through the intestinal pathway could be the reason why internal concentrations are so often better correlated to total concentrations in soil than to easily extractable amounts of metals (Table 1.1).
Harmsen, 2007 recommended to separate an ’actual’ available fraction corresponding to the dissolved fraction, a ’potentially’ available fraction corresponding to the maximum amount that can be released and a non-available fraction. In the case of earthworms, the ’actual’ available fraction could be assessed by pore water or weak salt extractions while for the potentially available fraction there is no consensual method.
Relationships between soil metal availability and earthworm response to metal exposure
One the one hand, soil total metal content do not re ect toxicological e ects but on the other hand, earthworm tissue metal content is better predicted by soil total metal concentration. Several authors suggested that total internal concentrations were not pertinent to predict toxicity (Van Straalen et al., 2005; Luoma and Rainbow, 2005). Earthworms can bioaccu-mulate large amounts of metals in their tissue but only a portion of the total body burden causes toxicological e ects. Like other organisms, earthworms can sequester metals under inert forms that does not interact with sites of toxic action (Vijver et al., 2004). Other bio-logical endpoints than total internal content may be more appropriate to identify pertinent indicators for environmental availability.
The free ion concentration is often considered as the best predictor of metal toxicity to organisms (Qiu et al., 2014; Oste et al., 2001). This assumption historically emerged from scientists working on aquatic environments and gave rise to the free ion approach model (FIAM) and biotic ligand model (BLM) in which the concentration of a contaminant at a given target site in the organism (biotic ligand) governs metal toxicity (Paquin et al., 2002). The binding to the biotic ligand is modeled based on free ion concentration and takes into account cation competition for the target site (Di Toro et al., 2001). Terrestrial BLM (t-BLM) were developped for soil microorganisms, plants and invertebrates (Thakali et al., 2006). Based on the rationale that earthworms are primarily exposed to the soil solution, terrestrial BLM were applied to earthworms (Steenbergen et al., 2005; Thakali et al., 2006), and free ion approaches were used to predict earthworm survival rate (Qiu et al., 2014). Such an approach assumes that free ion metals in the soil solution are directly available to earthworms, but neglects other potentially available metal pools that can be available after soil has passed through the earthworm gut. In addition, free ion approaches and t-BLM were mostly applied after unrealistic earthworm exposure (sand, arti cially contaminated soils with narrow ranges of soil properties (Steenbergen et al., 2005; Qiu et al., 2014)). Their pertinence in the context of eld-contaminated soils remain to be addressed.
Context-dependent relationships between toxicological endpoints and chemical extrac-tions are found in the literature, consistently with the approach with internal contents. In highly contaminated soils, earthworms growth and reproduction were not signi cantly re-lated to metal availability assessed by CaCl2 extractions (Arnold et al., 2003; Smith et al., 2010). However, other authors demonstrated the opposite (Daoust et al., 2006; Owojori et al., 2010). Using DTPA, Owojori et al. (2010) observed that the extractable metal contents were related to earthworm biomass, survival and reproduction, Daoust et al. (2006) however found otherwise.
The relationships between available metal concentrations and biological endpoints in earthworms remain unclear. A number of chemical extractions (e.g. CaCl2, NH4NO3, EDTA) are however routinely used in the literature. The identi cation of a proper method that best explains variations in earthworm biological endpoints is further challenged by the correlations between metal concentrations obtained after di erent extractions or modeling procedures. It is noteworthy that the ability of chemical methods to explain variations in other endpoints than the total body burden was rarely addressed; even though body burdens are not consid-ered to properly re ect metal bioavailablity.
Biomarkers: de nition and position within the concept of metal bioavailability
Biomarkers are biological responses related to the exposure to or the toxic e ect of an environ-mental chemical (Scott-Fordsmand and Weeks, 2000). They can cover a number of biological levels of organization, even if the term commonly implies responses at the subindividual level (Lagadic et al., 1994; Spurgeon et al., 2005). Biomarkers were proposed as interesting tools within the framework of environmental risk assessment because they represent early warning signals. Indeed, as they precede the e ects at the life cycle level, biomarkers may o er the potential to be able treat a contaminated site before any adverse e ect on the populations oc-cur (Svendsen et al., 2004). Within ISO 17402, 2008, di erent measurements of toxicological bioavailability were proposed (mortality, reproduction, etc.).
Two categories of biomarkers are distinguished: biomarkers of exposure and of biomarkers of e ect. Biomarkers of e ect are directly related to the risk of adverse health e ects, while biomarkers of exposure precede any adverse health e ects and are directly related to the exposure to chemicals (Forbes et al., 2006). According to this de nition, biomarkers of exposure can be considered as direct measures of metal bioavailability (Lanno et al., 2004; Eason and O’Halloran, 2002). They quantify the bioactive fraction of the pollutants, i.e. toxicological bioavailability (Scott-Fordsmand and Weeks, 2000).
A number of biomarkers have been shown to respond to metal contamination in earth- worms: expression of the gene coding metallothionein (MT) (Bernard et al., 2010; Spurgeon et al., 2004), stability of lysosomal membranes (NRRT : neutral red retention time, Svend- sen and Weeks, 1997), genotoxicity assessed by the Comet assay (Reinecke and Reinecke, 2004), enzymes activites (e.g. implied in the response to oxidative stress (Laszczyca et al., 2004)), energy reserves (Holmstrup et al., 2011), surface cast production (Leveque et al., 2013). Several reviews already exist on earthworm biomarkers (Scott-Fordsmand and Weeks, 2000; Lionetto et al., 2012). The purpose of the present synthesis is not to give a catalogue of biomarkers but to highlight the gaps of knowledge that remain to be addressed in order to con rm their per-tinence as indicators of metal bioavailability. We identi ed four main gaps of knowledge related to: (i) the relationships between metal exposure and biomarker responses (ii) the assumption that metal exposure has a greater impact on biomarkers response than the e ect of confounding factors, notably soil characteristics, (iii) the nature of biomarkers responses in eld-contaminated soils that combine low levels of contamination and the presence of multiple metals.
Link between biomarkers responses and metal exposure
Most studies on biomarkers in earthworms exposed to metal contamination focused on the dose-response relationships considering total metal concentrations and not body loads or metal availability measurements (e.g. Reinecke and Reinecke, 2004).
A number of studies demonstrated that the expression of MT gene (coding for a protein involved in detoxi cation mechanisms of Cd, Cu, and Zn) increased with metal contamination (e.g. Brulle et al., 2007). Only two studies, however, demonstrated signi cant correlation be-tween internal metal concentrations and the expression of MT. Spurgeon et al. (2004) showed a relationship with Cu body loads. Galay-Burgos et al. (2005) showed MT expression was correlated to internal Cu and Cd concentrations. Both these studies were spiking experiment with elevated concentrations of Cd (up to 180 ppm in Galay-Burgos et al. (2005) and to 800 ppm in Spurgeon et al. (2004)) and Cu (up to 180 ppm (Galay-Burgos et al., 2005) and 640 ppm (Spurgeon et al., 2004)). These two studies considered the earthworm species Lumbricus rubellus (Ho meister, 1843). The stability of lysosomal membranes (NRRT) was shown to decrease with metal contamination in several earthworm species (Spurgeon et al., 2000; Svendsen et al., 2004). Several studies tested if NRRT was correlated to internal metal contents. Van Gestel et al. (2009) found signi cant relationship with internal Cu concentra-tion in earthworms while Berthelot et al. (2009) found the opposite in multi-contaminated soils. The Comet Assay that measures genotoxicity was correlated to internal As contents in Lumbricus terrestris (Linnee) (Button et al., 2010). Fourie et al. (2007) have however shown no relationship with Cd body loads in several species (but Cd is not highly genotoxic). Holmstrup et al. (2011) correlated glycogen contents with internal metal concentrations and demonstrated signi cant relationships with Al and Ni but not with Cu, Cd or Pb. Concern-ing the activity of enzymes involved in the response to oxidative stress, Lukkari et al. (2004), for example, demonstrated that GST (glutathione-s-transferase) activity was not correlated to internal metal concentration in earthworm collected from eld contaminated soils, while EROD (Ethoxyresoru n-O-deethylase) activity was positively correlated to Cu, Zn, Al and Fe contents.
Together, these various results are in agreement with the previous conclusions and with the rationale that internal metal contents do not re ect the e ects of metal (Luoma and Rainbow, 2005). Overall, it seems that only the studies that considered soils spiked with elevated concentrations of metals were able to demonstrate signi cant correlations. More importantly, considering a few number of soils or treatments makes it di cult to detect signi cant e ects of internal concentrations on biomarkers.
E ect of confounding factors: focus on soil characteristics
To be considered as a valid measurement of metal bioavailability, a biomarker needs to be weakly a ected by confounding factors (such as temperature, season, weight etc.). According to Svendsen et al. (2004) the in uence of confounding factors on biomarkers response needs to be minimal or well-characterized. Field studies demonstrated that metals were not always the most important factors a ecting biomarkers in earthworms. Peres et al. (2011) showed that the expression of the gene coding for MT responded to other factors than metals. Laszczyca et al. (2004) showed that the activities of several enzymes involved in the response to oxidative stress were sensitive to the season. Energy reserves levels were also demonstrated to be sensitive to temperature and fertilization (Overgaard et al., 2009; Bednarska et al., 2013). The e ect of confounding factors can greatly adverse the conclusions withdrawn from studying the responses of biomarkers. It is therefore very important to determine which are the most in uent and how much they change biomarkers response. Among confounding factors, soil properties are of crucial interest with regards to research on metal bioavailability because they in uence both metal availability and earthworm biology and ecology. Soil pH, OM, metal (hydr)oxides and clay are the most important determinants for metal speciation, and hence metal availability (Sauve et al., 2000). Soil characteristics can thus exert indirect e ects on biomarker responses via their in uence on metal specia-tion. Physico-chemical characteristics are also key factors controlling metal uptake by soil organisms (Thakali et al., 2006; Van Gestel, 2008). Peijnenburg et al. (1999) and Giska et al. (2014) showed that the uptake rates of metals in earthworm were determined by the pH. Nahmani et al. (2009) however found no clear pattern between soil properties and uptake rate constants.
Soil parameters can also a ect biomarkers response more directly via changing earthworms physiology and behavior. However, this topic is rarely addressed. It is known that soil properties a ect earthworms abundance in situ. The most important parameters are pH, SOM contents and soil texture (Curry, 2004). Earthworms are rarely present beneath a pH of 3.5. SOM are the food base for the earthworm and are thus vitally important. Soil texture play an important role. Equilibrated soil textures are more favorable than sandy soils and higher energy expenditure can be expected in compact medium (Lavelle, 1988).
Several studies demonstrated that these soil properties a ected metal toxicity for earth-worms. For example, Owojori et al. (2010) showed that increased clay content was associated to decreased mortality of Eisenia fetida (Savigny 1826) in Cu contaminated soils. Daoust et al. (2006) demonstrated the in uence of pH, SOM and clay content over Cu toxicity (mor-tality) on E. fetida. Irizar et al. (2014) found Cd toxicity was modulated by SOM in E. fetida.
Biomarker responses in eld-contaminated soils
In the case of di use pollution, soils are often moderately or lowly contaminated. Moreover, the process of ageing lead to decrease bioavailability of metals over time in eld-contaminated soils (Lock et al., 2006). In addition, eld contaminated soils are most of the time contam-inated by mixtures of metals. The question arises whether biomarkers can indicate metal bioavailability in such a context.
First, their response to low doses is often hormetic. Calabrese, 2008 de ned hormesis as \a biphasic dose-response phenomenon characterized by a low-dose stimulation and a high-dose inhibition ». In the case of earthworms and metal contamination, several studies reported such hormetic-like response of biomarkers. Earthworms growth and reproduction was shown to be stimulated by low levels of metals in both soil and worms (Svendsen and Weeks, 1997; Spurgeon et al., 2005; Spurgeon et al., 2004; Ma, 2005). Galay-Burgos et al., 2005 reported biphasic hormetic-like response of MT expression in earthworms exposed to Cd and Cu. In addition, the activity of enzymes involved in the response to oxidative stress (CAT, GST, SOD (superoxide dismutase)) was demonstrated to exhibit such hormetic dose-response (Laszczyca et al., 2004; Zhang et al., 2009). Overall, there is little information on the hormetic e ects of metals on earthworm biomarkers (Zhang et al., 2009). The fact that a biomarker increases at low doses means that the contaminant is both environmentally and toxicologically available. Within a bioavailability framework, the low-dose stimulation can thus be indicative while this is not the case for the assessment of harmful e ects or risk.
Biomarkers may further respond to multiple metals in eld-polluted soils. This char-acteristic of the eld context is again leading to the expectation of a complex response of biomarkers. The toxic e ects of mixture of metals are more and more addressed in ecotoxi-cology studies (Jonker et al., 2005). The e ect of di erent metals interacting in an additive, synergistic or antagonistic manner is generally investigated using arti cial exposure experi-ments (e.g Khalil et al., 1996; Lock and Janssen, 2002). Such studies are valuable to address the mechanisms underlying the complex e ects of mixtures. However the prevalence of those interactions within more complex systems such as eld-contaminated soils also need to be investigated.
Several reasons may thus prevent the use of biomarkers to assess toxicological bioavail-ability in eld-contaminated soils. An approach commonly used to study the response of biomarkers is to consider one single reference soil, several treatments with high doses of total metals, and often arti cially contaminated soils. If this approach is needed to understand the underlying mechanisms, it is equally important to study the response of biomarkers in more complex gradients of soils. In eld-contaminated soils, low doses, multiple contamination and soil parameters may interact at a broader scale of observation and lead to very di erent results than when considering simpli ed systems.
Environmental bioavailability, a key step in the causal model
Environmental bioavailability represents the mediator between soil metal availability and tox-icological bioavailability. Within a causal framework, it is thus a key component. Bioavail-ability is a dynamic process (Peijnenburg and Jager, 2003). Toxicokinetics approaches quantify uptake and elimination rates, bioaccumulation and time to reach steady-state internal concentration. A number of authors used the uptake rate constants to assess metal bioavail-ability to earthworm (Peijnenburg et al., 1999; Van Straalen et al., 2005). Toxicokineics approaches were mostly conducted in the laboratory, following exposure to arti cial soils, arti cial contamination and using the model earthworm species (E. fetida) (Giska et al., 2014). But more and more studies have considered eld-polluted soils (Nahmani et al., 2009; Peijnenburg et al., 1999) and more representative earthworm species (Giska et al., 2014). Nahmani et al. (2009) and Peijnenburg et al. (1999) found variable uptake rate constants in E.andrei and E.fetida. Nahmani et al. (2009) concluded that it was di cult to extrapolate uptake rates from one soil to the other, and the uptake kinetics parameters were soil-speci c.
Bioaccumulation of metals in organism is an integrative assessment of chemical exposure in contaminated environments. Luoma and Rainbow, 2005 however highlighted its variability and its complexity. Indeed, bioaccumulation is the resultant of both uptake and excretion processes. Van Straalen et al. (2005) showed bioaccumulation was less pertinent than uptake rate to assess metal bioavailability to earthworm. Two main drawbacks are usually pointed out: (1) bioaccumulation does not address the dynamic nature of bioavailability and (2) it is not a good predictor for the e ects of metals on organisms (Luoma and Rainbow, 2005). Nevertheless, bioaccumulation is routinely reported in papers addressing metal bioavailability and e ects on earthworms. Moreover, measuring metal tissue concentration is crucial for risk assessment of metal biomagni cation (Vandecasteele et al., 2004).
Within the de nition of bioavailability, it is unclear if metal tissue concentration is a re-sultant of environmental bioavailability or of toxicological bioavailability. In ISO 17402, 2008, \accumulation » is positioned within the toxicological bioavailability step, but it is unclear if accumulation designates the internal metal content or the fact that the metal is accumu-lated at higher concentrations than in an organism unexposed to metal pollutants. Lanno et al., 2004 indicated that metal concentration in the organism was intermediary between environmental and toxicological bioavailability. Indeed, increasing metal concentrations in earthworm is the result of uptake processes. Thus, within the causal de nition of bioavail-ability, internal metal concentration is caused by uptake processes but is not necessarily the cause of the organism toxic response.
One of the reasons why internal metal content is not a good predictor of the e ects of metals on organism is that organism protect themselves from the side-e ects of metals by sequestration in certain tissue of their body (Vijver et al., 2004). Similarly to metals in soil, only a fraction of total internal metals can lead to an e ect. Earthworms compart-mentalize metals, for example, in the chloragogenous tissue and in the posterior alimentary canal (Morgan et al., 2004; Andre et al., 2009). At the subcellular level, two major path- ways of detoxi cation are metal binding to proteins (e.g. MT Sturzenbaum et al., 2004) and precipitation into insoluble metal concretions (metal-rich granules) (Vijver et al., 2004). Subcellular fractionation procedures partition metal body burden into operationally de ned fractions notably cytosol, metal-rich granules and a fraction containing tissue, intact cells and cell membranes (called debris hereafter) following di erential centrifugations. They were developed on aquatic organisms (Wallace et al., 1998) and applied to earthworms (Vijver et al., 2007; Jones et al., 2009). If metal concentrations in certain subcellular compartments are considered to re ect toxicologically bioavailable metal pools, it was rarely proved in the case of earthworms. Moreover, the subcellular distribution is metal-speci c and several fractions combine both toxic and detoxi ed forms of metals (Vijver et al., 2004).
Several studies considered the time-variation of metal concentrations in subcellular com-partments in earthworms (Jones et al., 2009; Li et al., 2009; Arnold et al., 2008). The relationship between subcellular partitioning and soil availability was only reported for Cu in earthworm (Vijver et al., 2007). Moreover, the subcellular partitioning was rarely related to biological endpoints. The added value of metal subcellular partitioning to assess metal bioavailability could be addressed, notably by determining its variations in ranges of eld-contaminated soils (Jones et al., 2009). Metal concentrations in subcellular fractions are the result of uptake processes such as the total internal contents. However, within a causal de nition of bioavailability, metal concentrations in certain fractions can be considered to have causal e ects on toxicological bioavailability. Similarly to soil available concentrations, metal contents in certain subcellular fractions could be de ned as toxicologically available, while toxicological bioavailability is the resultant of the interaction of such internal pools with biological processes.
Table of contents :
1 Synthesis
1.1 Graphical modeling of metal bioavailability
1.2 How to assess soil metal availability?
1.3 Biomarkers as indicators of metal bioavailability to earthworm?
1.4 Environmental bioavailability, a key step in the causal model
1.5 Concluding remarks
2 Experimental design
2.1 Specication and assumptions of the structural equation model
2.2 Choice of the target earthworm species
2.3 Selection of the series of eld-contaminated soils
2.4 Choice of experimental conditions of the earthworm exposure
3 Subcellular partitioning of metals in earthworm and metal speciation in soils
3.1 Introduction
3.2 Material and Methods
3.3 Results and Discussion
3.4 Conclusions
4 Relationship between earthworm energy reserves and soil metal availability
4.1 Introduction
4.2 Materials and Methods
4.3 Results
4.4 Discussion
4.5 Conclusion
5 Internal metal compartmentalization and biomarkers in earthworms
5.1 Introduction
5.2 Methods
5.3 Results and Discussion
6 Testing the structural equation model of metal bioavailability to earthworm
6.1 Introduction
6.2 Model description
6.3 Methods
6.4 Results and discussion
6.5 Conclusions
Discussion & Perspectives
Bibliography
