Gal Et Al 2008.pdf

Available online at www.sciencedirect.com

Environment International 34 (2008) 821 – 838 www.elsevier.com/locate/envint

Review article

Cobalt and secondary poisoning in the terrestrial food chain: Data review and research gaps to support risk assessment ☆ Judit Gál a , Andrew Hursthouse a,⁎, Paul Tatner a , Fran Stewart a , Ryan Welton b a

School of Engineering & Science, University of Paisley, Paisley PA1 2BE, United Kingdom b Cobalt Development Institute 167 High Street, Guildford, GU1 3AJ, United Kingdom Received 23 May 2007; accepted 19 October 2007 Available online 3 December 2007

Abstract Cobalt is a naturally occurring element found in rocks, soil, water, plants, and animals and has diverse industrial importance. It is cycled in surface environments through many natural processes (e.g. volcanic eruptions, weathering) and can be introduced through numerous anthropogenic activities (e.g. burning of coal or oil, or the production of cobalt alloys). The environmental behaviour of cobalt in terrestrial environment is relatively poorly studied and in particular where Co is used in industrial processes, the baseline information to support wider and long-term environmental impacts is widely dispersed. To support the adoption of new EU regulations on the risk assessment of chemicals, we review here the various aspects of the environmental chemistry, fate and transport of Co across environmental interfaces and discuss the toxicology and potential for bio magnification and food chain accumulation. The soil-to-plant transfer of Co appears to be viable route to expose lower trophic levels to biologically significant concentrations and Co is potentially accumulated in biomass and top soil. Evidence for further accumulation through soil-invertebrate transfer and to higher trophic levels is suggested by some studies but this is obscured by the relatively high variability of published transfer data. This variation is not due to one particular aspect of the transfer of Co in terrestrial environments. Influences are from the variability of geological sources within soil systems; the sensitivity of Co mobility to environmental factors (e.g. pH) and the variety of life strategies for metal elimination/use within biological species. Toxic effects of Co have been suggested for some soil–plant animal studies however, uncertainty in the extrapolation from laboratory to field is a major limitation. © 2007 Elsevier Ltd. All rights reserved. Keywords: Cobalt; Bioaccumulation; Secondary poisoning; Environmental variability; Bioavailability; Risk

Contents 1.

2.

Introduction . . . . . . . . . . . . . . . . . . 1.1. Industrial and domestic use . . . . . . 1.2. Cobalt as an essential element . . . . 1.3. Geochemical sources and forms . . . . Terrestrial environment . . . . . . . . . . . . 2.1. Environmental behaviour and mobility 2.2. Soil microorganisms . . . . . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

☆

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

Presented at the 24th European Meeting of the Society for Environmental Geochemistry & Health, Bratislava, Slovak Republic, April 2006. ⁎ Corresponding author. Tel.: +44 141 848 3213; fax: +44 141 848 3204. E-mail address: [email protected] (A. Hursthouse).

0160-4120/$ - see front matter © 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.envint.2007.10.006

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

822 822 822 823 824 824 825

822

J. Gál et al. / Environment International 34 (2008) 821–838

2.3.

Terrestrial plants . . . . . . . . . . . 2.3.1. Uptake . . . . . . . . . . . 2.3.2. Bioconcentration . . . . . . 2.3.3. Toxicity . . . . . . . . . . . 2.4. Aquatic and soil invertebrates . . . . 2.4.1. Uptake and bioaccumulation 2.4.2. Toxicity . . . . . . . . . . . 2.5. Vertebrates . . . . . . . . . . . . . . 2.5.1. Uptake and bioaccumulation 2.5.2. Toxicity . . . . . . . . . . . 3. Cobalt impacts on human diet . . . . . . . 4. Discussion and conclusions . . . . . . . . . Acknowledgement . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

1. Introduction 1.1. Industrial and domestic use Cobalt has many strategic industrial uses in cutting tools, super alloys, surface coatings, high speed steels, cement carbides, diamond tooling, magnets, ceramics and pigments (CDI, 2006). It is the 31st most abundant element in the crust (Krauskopf & Bird, 1995) and is mined in 17 countries, the most significant deposits being associated with Ni/Cu production in Africa and from arsenide ores in Canada and Morocco. Deep sea nodules and crusts provide 2.5–10 million tonnes of reserves and currently (2007) globally production levels are ∼54,000 tonnes/year (CDI, 2007). The behaviour of Co in the environment has previously been reviewed (Hamilton, 1994) and with the increased capability of analytical techniques and improved understanding of metal behaviour in the environment (Hursthouse, 2001), it is timely to consider further developments, particularly in light of changes to the legislative environment. Within the European Union this has recently become of more interest with the adoption of Regulation (EC) No 1907/2006 EU REACH legislation (Registration, Evaluation, Authorisation and restriction of Chemicals) requires producers to establish chemical safety assessments with the aim to ensure that potential hazards are both understood and suitably managed. This requires the holistic assessment of environmental fate and effects of substances (Huijbregts, 1999) and has also recently extended globally (e.g. the OECD's Strategic Approach to International Chemicals Management (SAICM) project, OECD, 2007). Technical guidance (EC, 2003) has been provided to provide a generic approach to risk assessment, focussing on human health and the wider environment. Ecological testing being incorporated with environmental fate and dispersion data to demonstrate likely exposure based on predicted exposure concentration (PEC) and predicted no effect concentration (PNEC). This approach works well for some substances which are volatile and are liable to diffuse through the environment (e.g. persistent organic pollutants) and where bio-concentration, -accumulation and -magnification factors can be used to predict PECoral and compare to PNECoral for wider risk assessment.

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

. . . . . . . . . . . . . .

826 826 827 829 830 830 831 831 832 832 833 834 835 835

However, for the metallic elements, the situation is further complicated by the knowledge that biological availability, regulation and accumulation are metal and biological speciesdependent (Gál et al., in press) and relationships are often nonlinear. Similar tissue concentrations may be derived from very different exposure and accumulation conditions (Hendrickx et al, 2004). Recently, it has been shown that, for many naturally occurring elements (e.g. Cu, W), the data available to establish suitable assessment is lacking (Sadhra et al., 2007; Koutsospyros et al., 2006). This review is based on the assessment of the terrestrial bioaccumulation and poisoning potential of Co and considers the key compartments of the assessment process highlighting data gaps and conflicts. A summary assessment framework or conceptual model is presented in the discussion section, highlighting the major areas of uncertainty. 1.2. Cobalt as an essential element Cobalt is naturally occurring element and it is widely distributed in rocks, soils, water and vegetation. It is usually found in association with nickel. Cobalt occurs in two oxidation states (Co2+ and Co3+). However, with the exception of certain complexes, Co3+ is thermodynamically unstable under typical redox and pH conditions (Nagpal, 2004; Palit et al., 1994). Cobalt is essential in trace amounts for humans and other mammals as it is an integral component of the vitamin B12 complex (Smith et al., 1981). This form of Co is obtained from microorganisms or from animal sources (USEPA, 2005). Vegetable sources of Co are important to ruminant animals (sheep and cattle). Cobalt deficiency in humans is similar to vitamin B12 deficiency, with symptoms of anaemia and with problems of the nervous system. As little as 0.1 μm Co as vitamin B12 per day is needed by adults, and total Co intake in the range 10 to 1800 μm per day (ATSDR, 2004) can be consumed on a daily basis without it being a hazard to human health (Atta-Aly, 2003). Cobalt in a different chemical form (i.e. not as part of vitamin B12) will stimulate blood formation, but this is unlikely to be a normal action. Although its essentiality in higher, non-leguminous plants is not clearly proven, there is some evidence of a favourable effect

J. Gál et al. / Environment International 34 (2008) 821–838

of Co on plant growth (Kabata-Pendias et al., 1992). Cobalt is reportedly an essential element for the growth of many marine algal species, including diatoms, chrysophytes, and dinoflagellates. It is also a micronutrient essential for some blue-green algae (Nagpal, 2004) taken up through photosynthetic pathways, and is required by microorganisms for nitrogen fixation in legumes. No biological use of Co is known other than its presence in vitamin B12 (HSDB, 2000). Although Co is an essential nutrient, excessive oral doses result in a variety of adverse responses. In higher concentrations, Co is toxic to humans and to terrestrial and aquatic animals and plants (Nagpal, 2004). The best characterised toxic responses are increases in red blood cell counts (polycythemia), cardiomyopathy, and effects on male reproductive system (Haga et al., 1996; USEPA, 2005). However, Co can be used safely in monitored medication to treat non-iron anaemia (Barceloux, 1999). Exposure to high levels of Co may also cause asthma, pneumonia, and wheezing (ATSDR, 2004). At present, the mechanisms underlying Co toxicity are poorly understood (USEPA, 2005).

Table 1 Summary of main Co mineral occurrences and geological settings Mineral name

Formula

Occurrence

Geological environment

Sulphides Cobaltite

CoAsS

Canada, Norway, U.S, Australia

High-temperature hydrothermal deposits and contact metamorphic rocks

CuS– Co2S3 Co3S4

Zaire, Zambia. U.S Sweden U.S, Commonwealth of Independent States, Zaire, Germany Zaire, Zambia Sweden

Carolite Linnaeite

Cattiertie Glaucodot

Oxides Erythrite

Asbolite

1.3. Geochemical sources and forms Cobalt is a naturally occurring element which is widely distributed in rocks, and soils where it is usually found as Co2+ (USEPA, 2005; Seigel, 2001). Total Co contents of soils vary depending on the parent material; there are also differences with depth in the soil profile and between different soil types derived from a common parent material due to natural pedological processes. Cobalt abundance can thus vary widely: from 0.05– 300 mg kg− 1, with an average content in range 0.1–15.0 mg kg− 1 (Alloway, 1997; Bowen, 1979; VROM et al., 1983; Young, 1979) from which the available part (the proportion of Co which is taken up by vegetation) is between 0.1 and 2.0 mg kg− 1 (Hamilton, 1994). In some cases, lack of Co in soils results in a deficiency of vitamin B12 in ruminants (Frank et al, 2004). Soils which contain b 10 mg kg− 1 Co are often classed as deficient (Hamilton, 1994), they are mainly derived from sandstones, limestones, some types of shale and from acid igneous rocks (i.e. granite). Co often occurs in association with Ni, Ag, Pb, Cu, and Fe ores, from which it is commonly obtained as a by-product. Soils near ore deposits, phosphate rocks, or ore smelting facilities, and soils contaminated by airport traffic, highway traffic, or other industrial pollution may contain relatively high concentrations of Co (ATSDR, 2004). The major ore deposits of Co occur as sulfides (Adriano, 2001) but Co also occurs in mineral forms as arsenides, and oxides and carbonates (Smith and Carson, 1981). A summary of main Co minerals and their geological setting is presented in Table 1. The FOREGS programme (FOREGS et al., 2006) provides online access to European baseline conditions, providing a median for top soil of 7.00 mg kg− 1 and with a range between b 1.0 and 255 mg kg− 1 and stream waters with a median of 0.16 μg l− 1 and range of 0.01–15.7 μg l− 1 (see: http://www.gtk.fi/ publ/foregsatlas/). Most Co (up to 79%, on average) in soils is contained in, or associated with Mn in various mineral forms (Nagpal, 2004).

823

Heterogenite

CoS2 (Co,Fe) AsS

3CoO– As2O5– 8H2O CoO– 2MnO2– 4H2O

E. Germany, Canada, U.S, Morocco New Caledonia, Australia???

CoO– Zair, U.S 2Co2O3– 6H2O

Arsenides Smaltite

CoAs2

Skutterudite

CoAs3

Safflorite

(Co,Fe) As2

Carbonates Sphaerocobaltite CoCO3

Canada, Morocco, U.S Canada, Morocco, U.S, Norway

Germany

Zair, Germany

In hydrothermal veins with other cobalt and nickel sulphides

Found in deep-seated deposits of hightemperature origin

Weathering of primary Coand As-bearing sulphide assemblages Secondary mineral occurring in joints and shears in the oxidised zone of dykes containing gold and base metal sulphides. Secondary mineral of smaltite (minor Co ore)

Hydrothermal ore mineral found in high-temperature veins with other Ni–Co minerals. Hydrothermal veins of moderate temperature and pressure.

Cobalt bearing veins that have been affected by carbonated waters

Data compiled from Downs and Hall-Wallace (2003); http://www.mineralsweb. com/; http://webmineral.com/chem/Chem-Co.shtml.

Cobalt co-precipitation or adsorption with Mn and Fe may affect Co mobility. The adsorption of Co on Mn minerals (birnessite MnO2) studied by X-ray absorption near-edge structure (XANES) (Kay et al., 2001), measurements indicated that most of the Co(II) is oxidised to Co(III) and that the Co(III) appears to be structurally incorporated (i.e., absorbed or coprecipitated) into the Mn minerals (Kay et al., 2001). Cobalt can also substitute for other trace metals (e.g., Cu, Pb, Zn, Cd) in a wide variety of minerals due to its similar geochemical properties. In such cases, formation of mixed Co-metal solids (e.g., sulfates, carbonates, hydroxides), may limit dissolved Co concentrations as well. The formation of Co complexes should be considered in the determination of solubility limits (Nagpal, 2004).

824

J. Gál et al. / Environment International 34 (2008) 821–838

2. Terrestrial environment The fluxes of Co through terrestrial environments can be due to natural or anthropogenic processes. Direct natural sources of Co to the surface environment include volcanic eruptions, seawater spray and forest fires, in addition to natural weathering related to soil derived dusts. Anthropogenic sources of Co to the atmosphere include coal-fired power plants and incinerators, as well as exhaust from vehicles. Cobalt released by human activities is mainly derived from nickel, copper, silver, lead, and iron mining and refining; metal production facilities; industrial boilers that burn coal and oil; vehicles that burn gasoline; and incinerators that burn refuse and sewage sludge. Industries reporting to USEPA released 7336 tons of Co into the environment in 1998 (ATSDR, 2004), most being released to land. Cobalt mining and processing activities, the production of alloys and chemicals containing Co, sewage effluents, urban run-off, and agricultural run-off are major anthropogenic contributors which increase Co levels in aquatic environments (ASTDR, 2004). Additional inputs of Co occur through the radionuclides 60 Co and 58Co which are moderately short-lived, artificial isotopes that are produced during the routine operation of nuclear reactors by neutron interaction with structural materials of the reactor. Small amounts may be released to the environment as contaminants in cooling water or in radioactive waste, particularly from discharges at nuclear sites undertaking fuel reprocessing (CEFAS, 2005). Radiologically and toxicologically, the routine discharges are not as significant compared to other nuclides, but can provide useful tracer information for Co mobility as they become dispersed through the ecosystem (Lux et al., 1995) or are used in experimental studies. In the environment, radioactive isotopes of Co will behave chemically like stable Co. However, 60Co and 58Co will also undergo radioactive decay according to their respective half-lives, 5.27 years and 71 days (ATSDR, 2004), which restricts the time period over which monitoring can take place (Djingova and Kuleff, 2002). Application of cobalt isotopes to specific aspects of terrestrial poisoning potential is included in appropriate points in this review. Potential contaminant exposure pathways for biota in the terrestrial environment are presented in Table 2. Bioconcentration (BCF), bioaccumulation (BAF) and biomagnification (BMF) factors are terms frequently used in the literature in a hazard identification/risk assessment context. They have been regarded as one of the key indicators to

Table 2 Exposure pathways considered for biota in terrestrial environments (USEPA, 2003) Birds and mammals

Ingestion of soils during grooming, feeding, and preening Ingestion of food contaminated as a result of uptake of soil contaminant

Plants Soil invertebrates

Direct contact Direct contact soil ingestion

understanding and identifying the potential to produce adverse effects in biota (EURAS, 2005). The concept that BCFs or BAFs can be used as an indicator of long-term or chronic toxicity to organisms stems from the assumption that larger BCFs are indicative of higher tissue concentrations, which in turn result in direct or secondary poisoning. The use of terms BCF and BAF varies widely in the literature, particularly for terrestrial studies, most commonly the models relate to aquatic systems where exposure assumptions can be more easily assumed in relation to measurements of the surrounding media. The majority of results reported as BCF (strictly ratio between body mass of chemical substance and concentration in surrounding water) are in fact BAF (ratio of organism content to surrounding environment including soil and water). In the soil system this introduces significant variability in derivation of relationships. Some studies have shown that accumulated metal (whole body concentrations) may be poorly, or even negatively, correlated with toxicity (EURAS, 2005). Organisms that tend to bioaccumulate metals to high levels do so because they are able to store the metals in a non-toxic forms (i.e., in granules, or bound to metallothioneins) (Rainbow, 2002). 2.1. Environmental behaviour and mobility The only significant sources of Co in soils are the parent material from which the soils are derived, which are enhanced only by deliberate applications of Co salts or Co-treated phosphate fertilizers to top soils, often undertaken to overcome deficiencies which can cause problems with ruminant nutrition or the cultivation of certain types of vegetation (Alloway, 1997). Cobalt is typically found at higher concentrations in ultrabasic rocks, where it is associated with olivine minerals, consequently soil developed from ultrabasic rocks is usually enriched with Co (Nagpal, 2004). Soils derived from basic igneous rocks or argillaceous sediments contain higher Co concentrations than soils derived from sandstones, limestones, and acid igneous rocks. Cobalt is relatively easily mobilized during weathering processes. The resulting distribution in soils is dependent on clay content, and the distribution of iron and manganese oxides (Paveley, 1988). In clay-rich soils, adsorption may be due to ion exchange at the cationic sites on clay, with either simple ionic Co or hydrolyzed ionic species such as CoOH+ (ATSDR, 2004). The possibility that Co can be complexed with biotic and abiotic constituents of environmental matrices means that it is critical that bioavailability is incorporated in the determination of definitive risk assessment. For these reasons, there is a need for a sophisticated interpretation of European risk assessment guidance and available data, and a solid understanding of the behaviour of metals in the environment (CDI, 2003a). In the soil–plant systems the behaviour of Co follows that of the Fe–Mn (Alloway, 1997, Li et al., 2004), modified by presence of organic matter and clay minerals. The presence of Mn and Fe are very likely to have the greatest influence on the bioavailability of Co, these elements bind with Co forming less biologically active species.

J. Gál et al. / Environment International 34 (2008) 821–838

The detailed chemical speciation (the precise molecular form) determines the availability of Co for biological uptake. It is commonly approached through sequential chemical extraction (SE) which has been criticised extensively for inconsistency in approach and wide range of methods applied (Hursthouse, 2001; Young et al, 2006). A number of other free ion approaches are also available (Zhang and Young, 2006), but provided operational considerations are taken, useful information regarding both the partitioning and relative reactivity of Co within soils and sediments and trends in behaviour can be elucidated (Tessier et al., 1979; Mossop and Davidson, 2003; Quevauviller, 2002, Gault et al., 2003). It has been suggested that soil Co is absorbed on Mn and organic matter complexes; the components adsorbed by soil oxides are not easily desorbed and do not exchange with the pool of dissolved Co, while Co which is sorbed onto humic substances is readily exchangeable (McLaren et al., 1986). In a comprehensive study of a soil data base for England and Wales (Suttle et al., 2003), assessment of 54 soil series for bioavailable (acetic acid extract) Co, showed that the total soil Co (5–24 mg kg− 1) had between 3 and 13% in the available fraction. The main finding of this survey was the strong negative relationships of available Co with soil pH and Mn content. In a survey of road impacts on urban soils (Woodard et al., 2007) a similarly modest portion of total Co was associated with more bioavailable fractions with Mn/Fe oxides dominating the association. The relative immobility of Co from incineration solids was also found to be high and a significant portion associated with Mn/Fe phases (Sočo & Kalembkiewicz, 2007). The ecotoxicity of Co to any soil organism is dependent on the physico-chemical properties of the soil, dictating the properties/response of the source material. As indicated above, parameters such as organic matter, clay mineral content, presence of complexing ions, adsorption, Eh/pH as well as moisture content of the soil material all play a defining role in Co bioavailability (translocation via root systems and foliage into plants) and hence its toxicity. However the dominant parameter in the terrestrial environment, regulating the availability of Co is pH. Fig. 1 shows the effect of pH (with Eh — the oxidising/reducing nature of the system) on the soluble Co species. Simple divalent Co dominates at lower pH, in most situations. Adsorption of Co2+ on soil colloids is high between pH 6 and 7, while leaching and plant uptake of Co are enhanced by lower pH (Nagpal, 2004; Watmough et al., 2005), with more acidic soils sorbing Co less strongly, allowing plant uptake (Chaney, 1983; USEPA, 2005). Very often Co reactions are similar to those of Ni in soil and plants with the exception being, when there is a strong component of soil MnO2 (Chaney, 1983; Li et al., 2004). Partition between solid and solution phases (often referred to as Kd when in equilibrium) is often used as an indicator of potential bioavailability. The sensitivity of Kd to soil properties has been demonstrated to be a good predictor of plant uptake (Watmough et al., 2005) but comparing field observations and experimental partitioning studies (e.g. Albrecht et al., 2003) is problematic, with Kds under similar soil pH conditions differing by 3–4 orders of magnitude. Evaluation of Kd in aquatic systems, particularly to validate use

825

Fig. 1. Eh–pH diagram showing the dominant aqueous Co species and solid phases. [Diagram was calculated at 25 °C and a concentration of 10− 14.8mol dm− 3 total dissolved Co, using Geochemists Workbench software and thermodynamic libraries].

in pollutant transport models (Osunaa et al., 2004; Krupka and Serne, 2002; Carrol & Harms, 1999) has shown wide enough variation to require case by case validation of distribution. Cobalt deposited from diffuse sources on soil is strongly incorporated in soil/solid phase particles and therefore does not easily migrate to depth. Cobalt released from power plants and other combustion processes usually become attached to very small particles (fine/accumulation mode) and Co contained in windborne soil is generally found in larger particles (coarse mode) (ATSDR, 2004; Hubbard et al., 1966; MOE, 1979). Studies of radioactive tracers in ploughed soil layers confirm relatively slow dispersal down soil profiles (Shinonaga et al., 2005). However, the form of the Co and the nature of the soil at a particular site will affect how far Co will penetrate into the soil profile (ATSDR, 2004). Strong retention in structured and intact field soil columns (Albrecht et al, 2003) of Co2+ during longterm irrigation experiments was also supported by observations from laboratory experiments using two Co EDTA chelates (Gwo et al., 2007). In this work, higher resolution of migration experiments suggested that oxidation state differences do influence migration and can be explained by competition between Fe and Mn oxide phases within the soil for each complex. Both in soils and sediment, the amount of Co that is mobile will increase under more acidic conditions (Watmough et al., 2005). Ultimately, most Co resides in the soil or sediment (Carr and Turenkian, 1961) and redistribution follows transport processes affecting these compartments. Examples might include water logging, which can release major redox sensitive Fe/Mn phases (Tagami and Uchida, 1998), along with associated Co. 2.2. Soil microorganisms Most microorganisms, including bacteria and algae, synthesise vitamin B12 (Chanarin, 1979) but data relating to the

826

J. Gál et al. / Environment International 34 (2008) 821–838

toxicity of Co to soil microorganisms is limited (Perez-Espinosa et al., 2002). As identified in the introduction, Co is required by microorganisms for nitrogen fixation in legumes. Cobalt is an important component of Rhizobium (common soil bacterium)– legume symbiosis. It constitutes the central atom in the porphyrin ring structure of the coenzyme cobalamin (vitamin B12 — is a general term to define a group of essential biological compounds known as cobalamins) that is essential for nodulation and bacterioid development (Bakken et al., 2004). It is known from previous reports that the vigour, nodule development and N content of some legumes might be positively affected by additional supply of Co when soil content of these elements is low (Marschner, 1995; O'Hara, 2001). The possibility that Co may be implicated in symbiotic nitrogen fixation, has been discussed (Reisenauer, 1960) and data presented which demonstrated the essential role of Co in the functioning of the Rhizobium-alfalfa (Medicago sativa L.) system in nitrogen fixation (Delwiche et al., 1961). It is concluded that Co ion plays a major role in the edicago– Rhizobiuni symbiosis (Delwiche et al., 1961). Sewage sludge applied to the land and water from sewage processing can contain Co but most is removed in the various stages of treatment. Reactions in sewage sludge between Co and microorganisms can form cobalamins, which are toxic to many forms of bacteria. The discharge of activated sludge may constitute an important source of cobalamins to natural waters (Hamilton, 1994), but it is not clear if Co reduces sludge breakdown. The influence of Co on microorganisms has been reviewed and it has been suggested (Reisenauer, 1960) that a protective influence of Co can account for the increased production of riboflavin by some organisms in the presence of optimal concentrations of Co+ 2 ions. Microorganisms in rumen fluids are capable of synthesizing a range of cobalamins, including vitamin B12, which is absorbed by the animal and provides levels of vitamins which are required for nutrition. Data for ecotoxicological impacts on microbial activity in soils suggests that loading of metals can have a significant impact and varies from metal to metal and species to species (Giller et al, 1998, 1999). Information on Co related impacts is absent. Biosorption of Co is controlled by different mechanisms in different biological species, for example, it involves the ionexchange mechanism in algae, but for fungi both metabolismindependent and metabolism-dependent processes operate. Physical conditions such as salinity, temperature, pH of the medium, and presence of other metals influence the process of Co uptake and accumulation in algae, fungi, and mosses (Palit et al., 1994). Cobalt has been shown to decrease the photoreversible absorbance of phytochrome in pea epicotyl and interferes with heme biosynthesis in fungi (Palit et al., 1994). In a study by Efroymson et al. (1997) a screening benchmark concentration for the toxicity of Co to soil microorganisms and microbial processes was calculated giving a value of 1000 mg kg− 1 Co in soils, based on a study by Lighthart et al. (1997). As a postscript to this aspect, it has been considered that the origin of organic matter/biomass may have a specific influence

on Co mobility. The observation that the Co-resistant fungi in serpentine outcrops in India were mainly representatives of the Moniliaceae fungi family (Pal et al., 2005) was used to study Co removal from solution by dried biomass from the site, identifying rapid and high capacity uptake and the competitive influence from other ions. These biosorption properties were considered useful in developing extraction systems to treat Co contaminated groundwater. However, this phenomenon is not unusual and many other biomass systems have been tested for metal removal. Indeed a recent study of Co uptake by algal biomass (Ozdemir et al., 2005) suggests that the fungal biomass produced on Co-rich substrates did not provide significantly different uptake capacity to that of common algal material or other sources of biomass. 2.3. Terrestrial plants 2.3.1. Uptake It has long been established that Co like a number of other elements is relatively toxic to plants when given in supranormal doses (Dekock, 1956). Plants can accumulate small (BCF bb0.5) amounts of Co from the soil, especially in the parts of the plant that are more routinely consumed, such as the fruit, grain, and seeds (ATSDR, 2004). The distribution of Co in plants is entirely species-dependent (Palit et al., 1994) and uptake is controlled by different mechanisms in different species. The availability of soil Co to plants is controlled by two major factors, MnO2 in soil (Co is strongly sorbed and coprecipitated with MnO2) and soil pH (Chaney, 1983; McKenzie, 1972). More acidic soils sorbing Co less strongly (Alloway, 1997; Hamilton, 1994; Palit et al., 1994; Sorokin et al., 2002; USEPA, 2005). Cobalt uptake is limited by the presence of humus and the presence of high concentrations of manganese in soil. Cobalt is taken up as a divalent cation, which is the only oxidation state commonly found in soil minerals. In a study by Kukiera et al. (2004) it was shown that the differences in uptake pattern Co by Alyssum (Ni and Co hyperaccumulator) from different soils and pH were probably related to the differences in organic matter and iron contents of the soils (Kukiera et al., 2004). Soil water status has a major influence on the amount of Co available for plant uptake (Hamilton, 1994; Robinson et al., 1999a,b). In poorly drained soils, the amount of extractable Co is greater than in areas which are well drained (Alloway, 1997) resulting in significantly increased plant uptake. In forest soils, it is relatively straightforward to estimate available soil Co from pH (Watmough et al., 2005), as metal concentrations in foliage and litter do not show dependence on other factors. This effect is also seen for lichens subject to varying degrees of atmospheric Co deposition from a range of natural and polluted atmospheres (Herzig et al., 1990). To obtain a response to applied Co it appears necessary for legumes to be growing in soil containing Rhizobia capable of symbiotic N fixation; but the soil must also be very low in available Co and N (Ozanne et al., 1963). An experimental study and a survey of organically farmed leys were undertaken to investigate whether N fixation in red

J. Gál et al. / Environment International 34 (2008) 821–838

clover pastures in Norway was limited by a low supply of Co (Bakken et al., 2004). There were indications that the supply of Co was insufficient to sustain the actual potential for symbiotic N fixation. As many of the investigated clover-soil systems were those previously known to be very low in Co, and the gain in N yield obtained by extra Co supply was marginal, this study suggested that Co deficiency is not a problem that merits any significant concern in the further development of legume based forage production systems in Norway (Bakken et al., 2004). Any stress that reduces plant activity will reduce N fixation. Factors like temperature and water may not be under the farmer control. But nutritional stress (especially P, K, Zn, Fe, Mo and Co) can be corrected with fertilizers. When a nutritional stress is corrected, the legume responds directly to the nutrient and indirectly to the increased N nutrition resulting from enhanced nitrogen fixation. Poor N fixation in the field can be easily corrected by inoculation, fertilization, irrigation or other management practices (Lindemann and Glover, 2003). 2.3.2. Bioconcentration With the exception of a few hyperaccumulator species (e.g. Berkheyda coddii, Alyssum corsicum, Lamiaceae, Scropulariaceae, Asteraceae, Cyperaceae), most terrestrial plant species do not bioconcentrate Co (Dzantor, 2004; Kukiera et al., 2004; Robinson et al., 1999a,b). Uptake of Co by plants is a function of the concentration of the element in the ionic form in the soil solution and the concentration present on the exchange sites of the cation exchangeable soil surfaces (Mollah and Begum, 2001). The highest accumulation of Co in plants occurs in the roots, although other parts of the plant also accumulate it to varying degrees (EURAS, 2005; Sasmaz and Yaman, 2006). This accumulation/biomagnification may potentially result in concentrations in predatory birds or mammals that are high enough to cause toxic effects, an effect known as secondary poisoning. In spiking experiments, bioconcentration factors (plant/soil) for 60Co and stable Co fertilizer vary from crop to crop. A number of different tropical crops produced factors which varied between 0.087 and 1.03 (Mollah and Begum, 2001) and these were found to diminish in the crops over subsequent years (see Fig. 2). In the case of tomato (undefined species) uptake

Fig. 2. Transfer factors for soil-to-plant uptake of 60Co spikes for consecutive years growth. Data from Mollah and Begum (2001).

827

from irrigated experiments exhibited strongest accumulation in the above ground components, with leaves showing higher levels than fruits (Sabbarese et al., 2002). This differentiation was also seen in experiments where tomato (Lycopersicon esculentum) was stressed by increasing Co concentrations and associated P and Fe (Chatterjee and Chatterjee, 2005). A similar experiment (Chatterjee and Chatterjee, 2000) was also undertaken using cauliflower (Brassica oleracea) in which plants were subjected to excess supply of Co in refined sand. The leaves of control plants showed higher accumulation of Co (10.1 mg kg− 1) in the tops than in roots (5.1 mg kg− 1). Those plants treated with Co showed more accumulation of the metal in the roots (1274 mg kg− 1) than the leaves (507 mg kg− 1). Excess supply of Co caused an increase in the P concentration (Chatterjee and Chatterjee, 2000). These spiking experiments provide the analytical security needed to systematically study uptake. However, their impact on field observations and verification is always problematic due to the dynamic nature of the soil–plant system (Hursthouse, 2001). Hysteresis and loss due to time dependent mechanisms on sorption and release processes (Gao et al, 2003; Mustafa et al., 2006) provide experimental artefacts which make interpretation more complex, particularly between studies. The mechanisms and details are beyond the review here. Indeed very little information on the effects relating to Cospecific transfer is available. One study has provided a field based evaluation of seasonal Co uptake into “foliage” (undefined plant species) as a food source for cattle (Khan et al., 2006). Whilst soil Co did not show much seasonal variation (unlike a number of other elements), plant content did vary between seasons — showing higher winter content and associated depletion with mobility of Mn during the season. A number of studies of plant communities and specific plant species indicate that simple species (e.g. liverwort) show significant (× 10) bioaccumulation factors (Samecka-Cymermann and Kempers, 1997) and these bioaccumulation factors are also found for rapidly developing parts of plants (e.g. pine needles, shoots) (Pyatt, 1999; Wagner, 1990). The shoots, leaves, fruits of a variety of plants surveyed within catchments in Northern Europe exhibited accumulation characteristics for Co which allowed regional scale biomonitoring (Reimann et al., 2001). Significant effects in the soil–plant transfer process occur in the accumulation of soil litter. Enrichment of the top soil through litter decay processes, adds organic matter to the surface horizons and presents an important step in the poisoning potential through the food chain (Tyler, 2005). This is particularly evident in waterlogged conditions, where soil redox conditions favour Co mobilization from the solid phase (Ye et al., 2001). Plants accumulating above 50–100 mg kg− 1 dry weight Co from soil are exposed to severe phytotoxicity (Chaney, 1983) but certain plants species have the ability to extract metals (such as Co) from soils, showing potential in phytoremediation technologies. However, only a few species have this ability. As defined by Baker and Brookes (1989) a plant containing

828

J. Gál et al. / Environment International 34 (2008) 821–838

N 1000 mg kg− 1 of metal in its tissues is considered to be a metal hyperaccumulator (Baker and Brookes, 1989; Chaney, 1983; Malaisse et al., 1999). In the case of Co, the existence of hyperaccumulators (Aeollanthus subacaulis var. ericoides, Senecio cf. coronatus (Asteraceae)) was noted during a study undertaken on the leaf metal content of plant communities in the Luiswishi Cu/Co mineralization zone, African Congo (Malaisse et al., 1999). Other plant Co concentrations from the same area are presented in Table 3. The A. subacaulis var. ericoides (mint family) Co hyperaccumulator plants were already known for their adaptation on metallic soils (Brooks and Herman, 1998). High Co contents of the daisy, Senecio cf. coronatus (Asteraceae) should also be noted. The levels measured in these plants represent a wide range Co concentration: 0–1429 mg kg− 1 dry weight Co. Baker and Brooks (1989) listed 9 hyperaccumulators of Co. The reason for the accumulation phenomenon is still being debated (Meharg, 2005; Kramer et al., 2007), relating physiological processes and ecological strategy, with advances in the detail of Table 3 Cobalt contents of plants (mg kg− 1 dry weight) at Luiswishi, African Congo, from Malaisse et al. (1999) Plant species

Co content (mg kg− 1)

Aeollanthus subacaulis var. ericoides — Lamiaceae Senecio cf. coronatus — Asteraceae Phyllanthus reticulatus — Euphorbiaceae Bulbostylis fusiformis — Cyperaceae Commelina bequaertii — Commelinaceae Hibiscus rhodanthus — Malvaceae Bulbostylis cupricola — Cyperaceae Lelya prostrata var. prostrata — Rubiaceae Triumfetta digitata — Tiliaceae Pandiaka carsonii var. linearifolia — Amaranthaceae Cyperus kibweanus — Cyperaceae Pimpinella acutidentata — Apiaceae Becium obovatum subsp. obovatum — Lamiaceae Euphorbia cyparissioides — Euphorbiaceae Ascolepis metallorum — Cyperaceae Aeschynomene pygmaea var. hebecarpa — Fabaceae Aristolochia heppii — Aristolochiaceae Gnidia hockii — Thymeleaceae Vernonia suprafastigiata — Asteraceae Chlorophytum sphacelatum — Liliaceae Becium centrali-africanum — Lamiaceae Eriospermum abyssinicum — Eriospermaceae Justicia cf. lolioides — Acanthaceae Fadogia cienkowskii — Rubiaceae Eriosema shirense — Fabaceae Thesium quarrei — Santalaceae Pleotaxis pulcherrima — Asteraceae Pteridium aquilinum subsp. centraliafricanum — Hypolepidiaceae Thunbergia aff. oblongifolia — Acanthaceae Fadogiella cf. stigmatoloba — Rubiaceae Ascolepis cf. protea var. bellidiflora — Cyperaceae Adenodolichos rhomboideus — Fabaceae Cryptosepalum maraviense — Caesalpiniaceae Protea welwitschii — Proteaceae Thesium quarrei — Santalaceae Annona stenophylla subsp. nana — Annonaceae Aristolochia hockii subsp. hockii — Aristolochiaceae Alectra sessiliflora — Scrophulariaceae

1,429 718 504 353 314 212 156 144 107 106 106 85 75 64 64 57 54 45 41 38 34 32 26 25 24 21 19 17 17 14 14 11 10 6 3 2 1 0

in plant transfer processes. Simplistically metal hyperaccumulation is an evolutionary adaptation of some plants to soil formed from parent materials that are rich in specific minerals. This characteristic gives such plants an ecological advantage over other plants to which elevated metal concentrations are unfavourable to growth. In studies of a Ni hyperaccumulator B. coddii, found in South African, it was noted that accumulation of Co was also high (Keeling et al., 2003). An investigation of the potential of the high-biomass productivity of Berkheya coddii to phytoextract Co and/or Ni from artificial metalliferous media, provided bioaccumulation coefficient for Co, for single element substrates (total metal concentration of 1000 mg kg− 1) of 50 (Fig. 3). Elevated Co concentrations significantly decreased the biomass production of B. coddii without affecting the bioaccumulation coefficients. This may indicate competition for binding sites in the root zone. The interference relationship between Ni and Co uptake demonstrated by B. coddii suggests a significant limitation to phytoextraction where both metals are present. It was finally concluded that B. coddii readily accumulates Co. However, in the presence of Ni, hyperaccumulation of both elements appears at Co concentrations that are phytotoxic (Fig. 3). Cobalt concentrations above 111 mg kg− 1 have a pronounced phytotoxic effect on biomass production in B. coddii without affecting the bioaccumulation coefficients of Co (Keeling et al., 2003). Pot trials were used to investigate the effects of MgCO3, CaCO3, sulphur, chelating agents (NTA, DTPA, EDTA) and acid mine tailings on Ni and Co uptake on the same hyperaccumulator B. coddii (Robinson et al., 1999a,b). Since CaCO3 was found not to have a significant effect on soil pH, it suggested that elevated Ca levels restrains the uptake of both Co and Ni. In general the addition of lime to soils reduces the availability of Co via pH change (Hamilton, 1994). Sulphur, and the addition of acid mine tailings, caused a highly significant increase uptake of both elements, relative to the controls. It was demonstrated that S could be used as a low-cost soil amendment to enhance the Co uptake of crops grown on ultramafic soils. Thus, land management procedures would enhance phytoremediation and phytomining operations for Ni and Co (Robinson et al., 1999a,b). These experiments indicate that neither soil amendments with Ca and Mg carbonates, nor the addition of chelating agents are effective in increasing metal uptake by B. coddii on serpentine soils. Instead these soil amendments caused a significant decrease in metal uptake, the opposite response to the impact in soils supporting nonhyperaccumulator plants, where an increase in metal uptake was observed (Robinson et al., 1999a,b). Patterns of Co accumulation and storage were investigated in tomato (L. esculentum) and wheat (Triticum aestivum) species (Bakkus et al., 2005), two types of higher plants which grow in very different ways and that differ in their strategies to absorb Fe. Plants were grown in nutrient solutions, with different Co treatments, using controlled environmental conditions (Bakkaus et al., 2005). The spatial distributions of K, Ca, Fe and Co in whole plants, and in leaf and stem sections examined by micro-

J. Gál et al. / Environment International 34 (2008) 821–838

829

Fig. 3. Ni and co uptake by hyperaccumulator Berkheya coddii. Data from Keeling et al (2003). a. Biomass concentration v soil concentration data for Ni and Co separately; b. BAF data with soil concentration for Ni and Co separately; c. biomass concentration v soil concentration data for Ni and Co combined; b. BAF data with soil concentration for Ni and Co combined [Ni — black, Co — red]. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

PIXE, revealed that in wheat leaves (monocotyledon), Co was distributed in exactly the same pattern as K and Ca, following the network of parallel veins. By contrast, in tomato leaves (dicotyledon), elemental maps showed that Co is not distributed in the same way as K, Ca and Fe. It needs to be further confirmed whether this may be a result of a detoxification mechanism, through which Co is sequestered from other elements. A study of 97 leaf samples taken from three Co-accumulating species of the tree Nyssa (tupelo) growing at various localities in New Zealand found maximum Co concentrations of 95 mg kg− 1 (dry mass) in leaves of N. sylvatica with a geometric mean of 7.1mg kg− 1 (Robinson et al., 1999a,b). Comparable values were obtained for N. aquatica and N. sinensis. Bioaccumulation coefficients were high and showed a disproportionately high uptake of Co (max 200) compared to other published values. BAC values increased exponentially with decreasing Co concentrations in the soil, highlighting the possibility of using Nyssa species to recover Co from soils impoverished in this element (Robinson et al., 1999a,b). For poisoning potential, it is also potentially significant and provides a curious route for enrichment of Co in soil organic matter/litter with implications for the transfer at the interface between primary productivity and soil invertebrates. 2.3.3. Toxicity It has long been established that Co like a number of other elements is relatively toxic to plants when given in supranormal

doses (Dekock, 1956). To date, Co phytotoxicity has not been of practical importance (Chaney, 1983). It was suggested by Pinkerton et al (1981) that soils naturally rich in Co are richer in Ni as well, and Ni phytotoxicity is usually the effect that dominates (Pinkerton et al., 1981; Robinson et al., 1999a,b). Cobalt doses have been shown to affect the growth and metabolism of plants, to different degrees, depending on the concentration and status of Co in rhizosphere and soil (Palit et al., 1994). Cobalt interacts with other elements to form complexes and the cytotoxic and phytotoxic activities of Co and its compounds depend on the physico-chemical properties of these complexes, including their electronic structure, ionic properties (charge-size relations) and coordination (Smith and Carson, 1981). Thus, the competitive absorption and mutual activation of associated metals influence the action of Co on various phytochemical reactions (Palit et al., 1994). Plant uptake of specific metal ions is a balance between regulation of metal uptake and detoxification and storage mechanism (Krämer et al., 2007). This includes link to nutrient uptake and regulation with a relatively narrow concentration range between deficiency and toxicity. The transport processes at the solution:plant interface involve transport by organic complexes and sorption to non-specific sites on the plant cell wall. The details of recent advances in the mechanistic understanding of the basis of resistance/tolerance, mutation and transport have been reviewed for metals in general (Meharg, 2005). Specific mechanistic details for a number of toxic elements highlight the

830

J. Gál et al. / Environment International 34 (2008) 821–838

role of a range of oxygen, nitrogen and sulphur containing molecules and the presence of free metal ions within the plant. A few reports exist of Co related studies, particularly detailing the interaction of Co2+ within the plant. The interaction of Co with other metals mainly depends on the concentration of the metals used. For example, high levels of Co2+ induce Fe deficiency in plants and suppress uptake of Cd by roots. It also interacts synergistically with Zn, Cr, and Sn. Ni overcomes the inhibitory effect of Co on protonemal growth of moss, thus indicating an antagonistic relationship (Palit et al., 1994). The mechanism of resistance to toxic concentrations of Co may be due to intracellular detoxification rather than defective transport. In higher plants, only a few advanced Cu-tolerant families showed cotolerance to Co2+. Tolerance towards Co2+ may sometimes determine the taxonomic shifting of several members of Nyssaceae. Due to the high Co content in serpentine soils, essential element uptake by plants is reduced, a phenomenon known as the “serpentine problem,” for New Caledonian families like Flacourtiaceae (tropical trees and shrubs). Large amounts of calcium in the soil may compensate for the toxic effects of metals in adaptable genera grown in this type of soil (Palit et al., 1994). Toxic concentrations inhibit active ion transport. In higher plants, absorption of Co2+ into the roots involves active transport. Subsequent transport through the cortical cells then involves both passive diffusion and an active uptake process. In the xylem, the metal is mainly transported by the transpirational flow. Distribution through the sieve tubes is accomplished by complexing with organic compounds (Meharg, 2005; Page and Feller, 2005). The lower mobility of Co2+ in plants, compared to other divalent cations restricts its transport to leaves from stems (Palit et al., 1994), and in a number of studies, active Co localisation in plant roots (Liu et al., 2000; Page and Feller, 2005) suggests that toxicity may relate to competitive reactions disrupting nutrient transport (e.g. of Fe and S in mung beans) and that comparisons between Co and Ni show very different transport velocities in the plant roots (e.g. in wheat and lupin) (Page and Feller, 2005; Page et al., 2006). Table 4 presents previously published plant toxicity data for Co. Low concentration of Co2+ in substrates stimulates growth across the range of simple algae to complex higher plants. Relatively higher concentrations are toxic. A similar relationship is seen with crop yield when the metal is used in various forms of fertilizer, pre-seeding, and pre-sowing chemicals. The toxic effect of Co on a plants morphology include; stimulating leaf fall, an inhibition of greening, discolouration of the veins, chloritic mottling (e.g. in cauliflower) (Chatterjee and Chatterjee, 2000), premature leaf closure, and reduced shoot weight (Palit et al., 1994). In the case of B. coddii elevated Co (above 111 mg kg− 1 ) concentrations significantly decreased the biomass production, yellowing the foliage leading to necrosis (dark green leaves with dead spots) (Keeling et al., 2003). Cobalt also has been shown to alter the sex of plants like Cannabis sativa, Lemna acquinoclatis, and Melon cultivars (Palit et al., 1994). There are some reports of beneficial effects of Co on for example drought resistance in melon (Atta-Aly, 2003). Plants can accumulate very small amounts of Co from the soil, and although animals that consume these plants will

Table 4 Plant toxicity data for Co (after USEPA, 2005) Reference

Test organism

Soil pH

OM Bioav. Tox Tox % score parameter soil (mg kg− 1)

TN & Associates, Inc, 2000 TN & Associates TN & Associates TN & Associates TN & Associates TN & Associates Rehab and Wallace, 1978

Alfalfa Medicago sativa

5.0

5.0

2

EC20

7

Barley Hordeum vulgare Radish Raphanus sativa Alfalfa M. sativa

5.0

5.0

2

EC20

40

5.0

5.0

2

EC20

23

6.3

0.1

2

EC20

18

Barley H. vulgare 6.3

0.1

2

EC20

102

Radish R. sativa

6.3

0.1

2

EC20

86

Cotton Gossypium 6.6 spp.

2.4

1

LOAEC

100

EC20 = Effect concentration for 20% of test population. LOAEC = Lowest-observed adverse effect concentration (Rehab and Wallace, 1978; TN & Associates, 2000). OM% = soil organic matter content. Bioav. score = Scoring of relative bioavailability is based on a three-point scale (i.e., 0, 1, or 2) with a score of “2” being the highest score (USEPA, 2003). Tox soil = Tox value soil concentration mg kg− 1 dry weight).

accumulate Co, it is not known to biomagnify up the food chain (ATSDR, 2004; Baudin et al., 1990; Lux et al., 1995). The level of Co in vegetable animal foodstuffs (VAF) fluctuates widely; thus, for the VAF richest in Co (clover, alfalfa and grass) the Co content ranges typically from 0.200– 0.600 mg kg− 1 (dry) and for those poorest in Co like cereal stems and seeds from 0.010–0.030 mg kg− 1 (Giroussi et al., 1995); and for forage samples in a semi-arid region of Pakistan 0.036–0.126 mg kg− 1 (dry). Common plant foods such as cabbage, green onions, mushrooms, spinach, tomato and pears contain about 0.2 mg kg− 1 Co while foods containing low amounts of Co (0.05 mg kg− 1) are apples, apricots, carrots, potatoes, oats, wheat and rice; barley, rye and peas contain intermediate amounts. For animal products, the highest concentrations of Co are found in liver (0.1–0.4 mg kg− 1 Co). 2.4. Aquatic and soil invertebrates 2.4.1. Uptake and bioaccumulation Amongst terrestrial invertebrates, earthworms are the most frequently studied. They are attractive organisms for the analysis of bio transfer of chemical constituents from soil due to their feeding behaviour and their size, ubiquity and ease of collection. Earthworms may accumulate Co in situations where that element is found in low concentrations in the environment (Crossley et al., 1995). According to experimental results it was suggested that Co is only slightly accumulated by the earthworm Aporrectodea tuberculata (Helmke et al., 1979) despite being an essential element (vitamin B12) (Neuhauser et al., 1984). Despite its influence on environmental mobility, pH is not considered to be a key factor in the uptake of Co and other

J. Gál et al. / Environment International 34 (2008) 821–838

inorganic constituents into earthworms. A study of log transformed concentrations was used to obtain significant regression relationships for the uptake of inorganic elements by earthworms, small mammals and plants (Blasland and Lee, 2003). A range of BCF values were reported for earthworm exposure studies on contaminated soils (Hankard et al., 2004). The range of BCF values 0.72–13.9 did not appear to be related to pH or other basic site characteristics, but may show influence from co-contamination. In a study of a naturally Ni (and Co) rich environment, containing significant plant cover by the Ni hyperaccumulator A. pintodasilvae (mustard), whilst Co accumulation factors (AF) were modest (0.06–0.002), little accumulation within the food chain could be discerned. In a spruce forest ecosystem (Southern Germany) litter decomposers, carnivores (soil, trunk), bark eaters all showed AF values considerably lower than unity for Co (Roth-Holzapfel, 1990). Implying that AF values are greatest when Co levels in the soils are lower. Biologically this could mean that the elimination rates for Co by the consumers increase as soil Co levels increase. A study on Portuguese Serpentinites (Peterson et al., 2003) showed whilst Co AF for plants including Ni hyperaccumulators were low (0.002–0.06) there was some evidence that site invertebrates were able to transfer enhanced plant metal concentrations into the food chain. Experimental results suggest that the earthworm E. foetida is unlikely to concentrate Co above levels found in soil or manure (Crossley et al., 1995). This study showed that Co is assimilated by E. foetida from manure at a low rate, less than 10% of ingestion, and excreted slowly. Low assimilation and slow turnover are consistent with reports for other sludge elements in earthworms (Ireland, 1983) and for terrestrial saprophytic invertebrates in general. Assimilation of Co from mineral soil itself appears to be irrelevant; rich organic matter should be the major, if not the only, source of available Co for earthworms. These results could be affected by environmental factors such as soil type (Crossley et al., 1995), thus soils with high organic matter content might pose a risk for bioaccumulation of Co in earthworms. Soil water status, seasonality and other biotic regimes may influence earthworm uptake and retention of pollutants in the soil. Studies of the integration of ligand binding models for prediction of metal toxicity to worms in soils are under development (Lock et al., 2006). Results for Co suggest that dissolved Co2+ exposure was the dominant route, but that the observed toxicity to Co was influenced by the solid phase pH and Ca/Mg content, indicating secondary solid-solution exchange processes mediated the effect. Plants are consumed by a diverse array of herbivores, and plant–herbivore interactions can be influenced by herbivore feeding mode (Gavloski and Lamb, 2000; Jhee et al., 2005). Studies that have compared damage caused by herbivores representing different feeding modes on the same plant species have often found differing impacts (e.g. Gavloski and Lamb, 2000). In a study using an artificial insect diet amended with eight metals including Co hyperaccumulated by plants (Coleman et al., 2005), the results showed that all tested metals were toxic at hyperaccumulator concentrations (N100 mg kg− 1) to larvae of an insect folivore, the diamondback moth Plutella

831

xylostella (Lepidoptera: Plutellidae). Furthermore, toxicity extended far below hyperaccumulator concentrations for all the metals tested, suggesting that plants other than hyperaccumulators may benefit from elemental defences and again provides a suggestion that poisoning impacts for Co could occur at relatively low environmental concentrations. From the aquatic invertebrates a study of marine ecosystems presented bioaccumulation data of Co in trophosome and vestimentum of the tube worm Riftia pachyptila from Guaymas basin, Gulf of California (Ruelas-Inzunza et al., 2005). The chemicals are absorbed by the plume of a worm and then transported through the vestimentum to the trophosome (where the microbes live inside the worm). Highest mean concentrations of Co (1.3–3.0 mg kg− 1) were detected in the trophosome, perhaps as a consequence of blood transport. Statistical tests (t test) revealed significant differences p b 0.05 in the case of Co concentrations in vestimentum increased accordingly with the size of specimens. 2.4.2. Toxicity Little is known about Co toxicity in terrestrial invertebrates. Considerable amount of information is available for aquatic invertebrates, considered to be the most sensitive group of organisms to cobalt exposure, followed by fish and plants (Nagpal, 2004). A report with limited data on Co ecotoxicty to springtail (Folsomia candida) (Lock et al., 2004) produced EC50 values of 1480 mg kg− 1 Co for artificial soil and 490 mg kg− 1 Co for a field soil. The difference being attributed to the higher “availability” of the Co in the field soil evidenced by the cation exchange capacity and lower pH. This emphasised the sensitivity of Co bioaccumulation to environmental conditions and the need to be careful when comparing filed and experimentally derived factors. A limited study of LC50 effects for Co in spiked field soils on the pot worm (Enchtraeus albidus) (Lock et al., 2006) produced a 57 fold range of effects (3.7–210 μM Co2+. The EC50 value established for freshwater invertebrates include 12 μg/l (Nagpal, 2004). 2.5. Vertebrates Much of our knowledge of Co toxicity is based on animal studies. Cobalt is essential for the growth and development of certain animals, such as cows and sheep (ATSDR, 2004). In general, soils containing b 5 mg kg− 1 Co, or about 0.3 mg kg− 1 of available Co may not provide sufficient Co to vegetation for the maintenance of health in cattle and sheep (Hamilton, 1994). Predators of small mammals are often excluded in sitespecific risk assessments as data is lacking for most contaminated sites. Limited datasets are available for metal concentrations in small mammal tissues in relation to environmental contamination (Wijnhoven et al, 2007) but limited research has been undertaken to develop models relating soil concentrations to whole body concentrations (Blasland and Lee, 2003). Determining the exposure of small mammals to metals can be problematic as they tend to be exposed to large areas exhibiting heterogeneous soil concentrations (CDI, 2003b; Wijnhoven et al, 2007).

832

J. Gál et al. / Environment International 34 (2008) 821–838

2.5.1. Uptake and bioaccumulation Ecological Soil Screening Levels (Eco-SSLs) for Co were estimated in a report by US EPA (2005). Eco-SSLs are concentrations of contaminants in soil that are protective of potential ecological receptors that commonly come into contact with soil or ingest biota that live in or on soil. The Ecological Soil Screening Level (Eco-SSL) for birds and mammals were derived by determining the TRV (Toxicity Reference Value) and using this in conjunction with a wildlife exposure model to back-calculate the Eco-SSL for each of three surrogate species. The Eco-SSL for birds was calculated for different trophic levels: herbivores (dove) were found to have an Eco-SSL of 200 mg kg− 1 dry weight; ground insectivores (woodcock) were found to have an Eco-SSL of 190 mg kg− 1 dw; and carnivores (hawk) was estimated to be 500 mg kg− 1 dw, using a log-linear uptake model (USEPA, 2005). The EcoSSL for mammals was also calculated for different trophic levels: herbivores (vole) were found to have an Eco-SSL of 720 mg kg− 1 dw; ground insectivores (shrew) were found to have an Eco-SSL of 240 mg kg− 1 dw; and carnivores (weasel) was estimated to be 600 mg kg− 1 dw, using a log-linear uptake model (USEPA, 2005). Bird feathers (occasionally faeces and eggs) can be used as non-destructive biomonitors for metal pollution. Metals are excreted into growing feathers during moult and once fully formed, the feathers become isolated from the rest of the body. A number of studies (Howarth et al., 2002; Janssens et al., 2001; Kavun, 2004; Veerle et al., 2004) suggest that the feathers reflect circulating metal (e.g. Cd, Co, Cu, Fe, Mn, Pb, Zn) concentrations in the blood at the time of their development. Exogenous contamination may increase the metal concentrations of exposed feathers. The concentrations of twelve metals, including Co, were measured in the feathers of free-living adult great tits (Parus major) (Veerle et al., 2004), in order to determine the extent of exogenous contamination. With the exception of Hg, the concentration of all metals was significantly different for feathers that were exposed to exogenous contamination over different periods of time. The results suggested that concentrations of most metals (with the exception of Hg and Zn) build up with increasing age of the feather, indicating that exogenous contamination may be an important source of metals in feathers. It is not known whether exogenous contamination onto the feather surface is caused by atmospheric deposition, or from secretion products of the uropygial gland smeared onto the feathers during preening, or both. Washing does not remove all exogenous contamination and methodologies are required to eliminate exogenous contamination completely (Janssens et al., 2001). Feathers may be useful as biomonitors of aerial or other forms of Co pollution, however it should be noted that if exogenous contamination is primarily due to atmospheric deposition the feathers may not represent the Co concentration in internal tissues during feather growth. Studies in two Antarctic penguin species show effects of diet on concentration levels of a wide range of elements including Co. Levels for Co are generally comparable to those seen in W Europe (Metcheva et al., 2006; Veerle et al, 2004).

2.5.2. Toxicity Cobalt toxicity in livestock was not reported under field conditions according to experimental studies Co toxicity can be caused by adding ∼ 10 mg kg− 1 Co to the diet of ruminants (Chaney, 1983; NRC, 1980). Cobalt, which is extremely deficient in the Amazon region, is the element, which is most lacking in Brazilian grassland. Young animals are observed to be meagre, lethargic, without appetite but gnawing at tree barks; they loose the hair from their tails and have a scrubby hide (Primavesi and Primavesi, 2002). In ruminants, Co is distributed throughout the body with high concentrations in liver, bone and kidney (EMFEMA, 2002; Underwood and NiFa, 1999). The metabolism and toxicity of Co has been reviewed (Taylor and Marks, 1978, Elinder, 1984; Elinder and Friberg, 1986) and the determination of Co levels in blood and urine is common practice in order to identify exposure and thresholds for health risks (Alessio and Dell'Orto, 1988). Longer-term exposure of rats, guinea pigs, hamsters, and pigs to lower levels of Co in the air results in lung damage and an increase in red blood cells (ATSDR, 2004). Short-term exposure of rats to high levels of Co in the food or drinking water results in effects on the blood, liver, kidneys, and heart (ATSDR, 2004). Parenteral administered (i.e. non-oral, such as injection) Co in rats is cleared rapidly from the body via urine (e.g. 63% within 4 h). For a single intravenous injection of Co chloride (to rats), most of the Co is excreted in urine and faeces over a period of a few days (Hamilton, 1994). Following the intravenous injection of Co chloride to pregnant mice, a high transfer to liver, kidney, pancreas, cartilage, foetal skeleton and other tissues was observed (Soremark et al., 1979). For the rat, the oral LD-50 for Co (Co-II compounds) is between 150 and 500 mg kg− 1 body wt. of anhydrous compound. Inhalation of Co compounds by laboratory animals produces a number of effects, e.g. emphysema and hyperplastic and hypertrophic changes. Myocardial Co toxicity has been reported for a number of animals (Soremark et al., 1979). Animal studies have shown that cobalt accumulates in the myocardium and causes myocytolysis. Detrimental effects on sperm production and fertility have been observed in rats (FSA, 2003). In both chick primary cultures and rodent fibroblast cell lines, dissolved Co is cytotoxic. Cobalt in various dose levels and chemical forms affects reproduction and induces genetic disruption in a number of experimental animals (Leonard and Lauwerys, 1990). Administered Co compounds are relatively ineffective in prokaryotic systems (Hamilton, 1994). No appropriate data have been found for the study of Co bioavailability for poultry and for pigs. The avian (dove, woodcock, hawk) and mammalian (vole, shrew, weasel) wildlife toxicity reference value for Co is equal to 7.61 mg Co kg− 1 bw (body weight) day− 1 and 7.33 mg Co kg− 1 bw day− 1, respectively (USEPA, 2005). It would therefore appear that from published literature on soil–plant–animal exposure, Co could cause toxicity to ruminants grazing healthy, Co-rich forage. Monogastric animals should not be affected by higher Co concentrations in their food

J. Gál et al. / Environment International 34 (2008) 821–838

because grain and fruit tissues are only slightly enriched in Co when Co phytotoxicity causes visual crop damage to leaves and shoots (Chaney, 1983). In contrast, animals consuming large amounts of leafy vegetables originating from phytotoxic soils may be at risk. The data in Table 5 present BCFs and BAFs, obtained during a review of bioaccumulation (SRS, 1999). A default value of 1 was used for the bioconcentration factors when no published data was available, emphasising the knowledge gaps for Co mobility in different trophic levels. In order to evaluate risks to predatory wildlife whole body contaminant concentrations in small mammals are needed. Because these data are lacking for most contaminated sites, predators of small mammals are frequently not considered in ecological risk assessments (Sample et al., 1998). While considerable data are available concerning chemical concentrations in small mammal tissues in relation to environmental contamination, models relating soil concentrations to whole body concentrations lack suitable information. The development of a database of chemical concentrations in soil and whole bodies of small mammals for 14 inorganic elements including Co (Sample et al., 1998), identified differences according to small mammal diet preferences, each species was assigned to one of the three trophic groups: insectivore (diet consisting primarily of insects and other invertebrates), herbivore (diet consisting primarily of plant material), and omnivore (diet consisting of both animal and plant material). Values for Co are shown in Table 6. Additional parameters that may influence uptake of contaminants by small mammals include age, sex, and season of year (Talmage and Walton, 1991). 3. Cobalt impacts on human diet Cobalt is a trace element found widely in the environment. Important human food sources of Co include fish (0.01 mg kg− 1), nuts (0.09 mg kg− 1), green leafy vegetables (such as broccoli and spinach: 0.009 mg kg− 1) and fresh cereals (such as oats 0.01 mg kg− 1) (FSA, 2003). The vegetable components agree with the variation seen in wider vegetation levels (Section 2.3.3) and emphasises the range of environmental and species differences in Co content. Most of the Co ingested is inorganic. Table 5 Bio-uptake factors (SRS, 1999) Element

Cobalt

Soil-to-plant a

Soil-toanimal a

SPv

SPr

BAFinv c

BAFmamm

ST

BCF

0.004

0.00 4

1.00

1.00

1.00

300

Animal-toanimal a

Soil-totissue b

Water-tofish a

SP = Soil-to-plant transfer; v = vegetative parts, r = reproductive parts. BAF = Bioaccumulation Factor; inv = invertebrate (unit less), mamm = mammal (d/kg). BCF = Bioconcentration factor for transfer from water to fish and other aquatic biota (L/kg). a HAZWRAP (1994). b Default value. c BAFinv also used for sediment-to-invertebrate transfer.

833

Table 6 Small mammal bioaccumulation of Co from locations in N. America (Sample et al., 1998) Location

Species

Trophic group

Co tissue mg kg− 1

Co soil mg kg− 1 dry

Uptake factor

Ontario a

Microtus pennsylvanicus M. pennsylvanicus M. pennsylvanicus Sigmodon hispidus Peromyscus leucopus

Herbivore

2.33

50

0.047

Herbivore

1.8

10

0.18

Herbivore

2.5

25

0.10

Herbivore

0.17– 0.25 0.15– 0.24

7.55–15.6

0.013– 0.038 0.010– 0.025

Ontario a Ontario a Oklahoma b Oklahoma b a b

Omnivore

9.6–10.15

Cloutier et al. (1985). PTI (1995, 1998).

Total dietary intake of Co for man varies with geography, geology and types of diet; the mean population intake of cobalt is 0.012 mg day− 1 (FSA, 2003). A value of 0.23 mg day− 1 was determined for Tarapur, India (average value for fishermen and farmers) (Bhat et al., 1973); 0.323 mg day− 1 for the USA (Schroeder et al., 1967) and 0.036 mg day− 1 for Japan (Yamagata et al., 1963). The variability of Co concentrations in foodstuffs is dependent on the geographical origin and its detailed ingredients. Cobalt contents in various foodstuffs have been compiled (Leblanc et al, 2005). This reveals that the top food groups for Co uptake in the human diet are milk and dairy products (∼ 32% of total Co intake). A rare instance of Co toxicity in man occurred in North America and Europe in the 1960s when Co sulfate was added to beer as a form of foam stabilizer (Imray and Landley, 1999) which produced severe cardiomyopathy, haematologic, neurologic and thyroid abnormalities (Hamilton, 1994). High soil Co content has also been implicated with the incidence of goitre (iodine deficiency disease in humans and animals). It has been previously identified that most microorganisms, including bacteria and algae, synthesise vitamin B12, and they constitute the only source of the vitamin (Chanarin, 1979). The vitamin B12 synthesised in microorganisms enters the human food chain through incorporation into food of animal origin. In many animals gastrointestinal fermentation supports the growth of these vitamin B12-synthesizing microorganisms, and subsequently the vitamin is absorbed and incorporated into the animal tissues. This is particularly true for the liver, where vitamin B12 is stored in large concentrations. Products from these herbivorous animals, such as milk, meat, and eggs, constitute important dietary sources of the vitamin unless the animal is subsisting in one of the many regions known to be geochemically deficient in Co (Sample et al., 1998). Milk from cows and humans contains binders with very high affinity for vitamin B12, whether they hinder or promote intestinal absorption is not entirely clear. Omnivores and carnivores, including humans, derive dietary vitamin B12 from animal tissues or products (i.e., milk, butter, cheese, eggs, meat, poultry, etc.). It appears that an insignificant amount of vitamin B12 required by humans is derived from

834

J. Gál et al. / Environment International 34 (2008) 821–838

Fig. 4. Schematic summary of Cobalt transfer in the evaluation of bioaccumulation in terrestrial compartments.

microflora (FAO/WHO, 2001), although vegetable fermentation preparations have also been reported as being possible sources of vitamin B12 (Van den Berg, 1998). Cobalt is mainly excreted in the urine but also in the faeces. Independent of the route of exposure, most Co is eliminated rapidly, with a small proportion being eliminated slowly and having a half-life of the order of years (FSA, 2003). This reflecting the development of regulation seen in all levels of the ecosystem (Sections 2.3.3). Overall the FAO (Food and Agricultural Organisation of the United Nations) and WHO (World Health Organisation) recommend a 2.4 μg day− 1 of vitamin B12 (equivalent to 0.1 µg day− 1 of Co) in the adult diet (FAO/WHO, 2001). The discharge of Co wastes into the environment can cause neurotoxicological disorders and genotoxicity in human beings and in chronic cases can even cause cancer (Lison et al., 2001; Pal et al., 2006). The data relating to genotoxicity and carcinogenicity of Co have been reviewed (De Boeck et al., 2003). The majority of the data in this area relates specifically to inhalation pathway often under occupational exposures and fall outside the remit of this review. 4. Discussion and conclusions Assessment of the data reviewed here reveals that the food chain is not protected from Co toxicity by the soil–plant-barrier.

Indeed, there is a suggestion that plant accumulation from the soil substrate enhances Co levels in biomass and top soil through litter fall (Tyler, 2005). This emphasises the potential for the enhancement of Co transfer into soil invertebrates and further biomagnification and potential for low level toxicity are suggested (Section 2.3.1). The high variability of both baseline soil Co and the environmental variables responsible for biomass production, coupled with the wide range of transfer factors requires fuller evaluation. Suggestions that soil pH is a good indicator of soil Co availability does not simply extrapolate to bioavailability. Species and site-specific conditions provide a range of natural variability of over an order of magnitude for typical transfer factors. A greater awareness of relationship between measurable components and impact for a series of key scenarios is needed. Very little information exists relating the toxicity of Co to soil microorganisms. Too little is known about plant tolerance, plant uptake in relation to tolerance, and/or bioavailability of essential plant Co to animals. Due to the gaps in data availability for Co concentrations (mainly field data) in animal tissues only limited research has been undertaken to develop models relating soil concentrations to whole body concentrations, which is crucial in order to evaluate risks to predatory wildlife. Bioavailability information is needed for high levels of Co in vegetable crops. There is a lack of information about different

J. Gál et al. / Environment International 34 (2008) 821–838

bioavailability tests for Co (i.e. selective-sequential extraction experiments). Whilst recent evaluation of Co-containing compounds, highlighted pH induced solubility effects (Stopford et al., 2003), these focussed on scenarios related to occupational exposure/deposition. Little is known or has been shown relating field data to mesocosm-scale release and uptake. Fig. 4 provides a schematic summary of poisoning potential and the issues/questions surrounding the definition of any ecological impact. Considering the central sequence (solid boxes) of transfers from environment through the food chain, the availability of Co for uptake is a dynamic consequence of the time integrated environmental influences on Co source terms. Despite this inherent variability it has been demonstrated that transfer to the biosphere through primary productivity is feasible and can provide situations where low trophic level poisoning issues may occur. Further biomagnification is less clear and steps between primary productivity and trophic levels need elaborating. In light of the terrestrial poisoning issue, the current literature does not allow a good assessment of the poisoning risk to be made. A number of issues need further investigation: 1. The available literature concentrates to a small extent on direct experimental or field data concerning soil–worm– worm-eating predator (bird/mammal) relationship. The order of magnitude variations in field observations at the lower end of the food chain (soil-to-invertebrate transfer in particular) provide considerable inherent variability for the assessment of biomagnification. 2. Predator risk = ratio between food (PEC oral pred) and no effect concentration for oral intake (PNEC oral). The availability of data to test this relationship is particularly poor, based on the “food basket” concept the variation in source Co content adds further to questions regarding environmental variability. 3. Data must meet quality assurance criteria which should include enough detail of methodology to assess the reliability of measurements. Most published information does not allow the detailed assessment of experiments to follow this and most data are produced focused on artificial and single step investigations of ecosystem transfer. The disparity between experimental and field observations is clear. 4. Diet content can be assessed using soil content and BAF/BCF for food type. Bioconcentration and bioaccumulation factor data sets are both incomplete and partial in the published literature. There is some information on vegetation contents but details of steps between trophic levels are missing. 5. Biomagnification (BMF) considers ratio between predator and prey concentrations (assuming single food and source). Whilst it is relatively straightforward to provide some data, comparable data sets do not exist and cannot allow calculation without large uncertainty levels. They do not account for food basket variations, in addition to biological species-specific Co management strategies. Acknowledgement We gratefully acknowledge the support of the Cobalt Development Institute, Guilford, UK in conducting this work.

835

References Adriano DC. Trace elements in terrestrial environments. 2nd. New York: Springer; 2001. Albrecht A, Schultze U, Bello Bugallo P, Wydler H, Frossard E, Flühler H. Behaviour of a surface applied radionuclide and a dye tracer in structured and repacked soil monoliths. J Environ Radioact 2003;68:47–64. Alessio L, Dell'Orto A. Biological monitoring of cobalt. Biological monitoring of toxic metals. New York: Plenum Press; 1988. p. 407–17. Alloway BJ. Heavy metals in soils. London: Blackie Academic and Professional; 1997. ATSDR. Toxicological profile for cobalt. Agency for toxic substances and disease registry. Atlanta, GA: U.S. Department of Health and Human Services, Public Health Service; 2004. Atta-Aly MA. Effect of galia melon seed soaking in cobalt solution on plant growth, fruit yield, quality and sudden wilt disease infection; 2003. http:// www.aaaid.org/pdf/magazine/2003/Cobalt%2067-72.pdf. Baker AJM, Brooks RR. Terrestrial plants which hyperaccumulate metallic elements A review of their distribution, ecology and phytochemistry. Biorecovery 1989;1:81–126. Bakkaus E, Collins R, Gallien JP, Khodja H, Carrot F, Gouget B. Concentration and distribution of cobalt in higher plants: the use of micro-PIXE spectroscopy. Nucl Instrum Meth B 2005;231:350–6. Bakken AK, Synnes OM, Hansen S. Nitrogen fixation by red clover as related to the supply of cobalt and molybdenum from some Norwegian soils. Acta Agriculturae Scandinavica, B. 54(2), 97-101(5). 2004. Barceloux DG. Cobalt. J Toxicol: Clin Toxicol 1999;37:201–16. Baudin J, Fritsch A, Jay G. Influence of labeled food type on the accumulation and retention of 60 Co by a freshwater fish, Cyprinus carpio L. Water Air Soil Pollut 1990;51:261–70. Bhat IS, Hedge AG, Chandramouli S. Evaluation of internal exposure toradionuclides of I, Cs, and Co, during maintenance operations on primary steam leak in a nuclear power station. Health Phys 1973;25: 135–9. Blasland B, Lee I. Collection of environmental fate data for cobalt and cobalt compounds in global soils and sediments. NY: BBL Sciences; 2003. Bowen HJM. Environmental chemistry of the elements. London: Academic Press; 1979. Brooks SC, Herman JS. Rate and extent of cobalt sorption to representative aquifer minerals in the presence of a moderately strong organic ligand. Appl Geochem 1998;13(1):77–88. Carr MH, Turekian KK. The geochemistry of cobalt. Geochim Cosmochim Acta 1961;23(1–2):9–60. Carroll J, Harms IG. Uncertainty analysis of partition coefficients in a radionuclide transport model. Water Res 1999;33:2617–26. CDI. Review of existing data on natural background and ambient concentrations of cobalt and cobalt compounds in soils and sediment. Corvallis, Oregon: Cobalt Development Institute; 2003a. CDI. Collection of environmental fate data for cobalt and cobalt compounds in global soils and sediments. BBL Sciences; 2003b. CDI. Cobalt facts. Guilford, UK: Cobalt Development Institute; 2006. http:// www.thecdi.com/ [accessed April 2007]. CDI. Cobalt News, April 2007 (07/2). Guilford, UK: Cobalt Development Institute; 2007. ISSN 1353-5587. CEFAS. Radioactivity in food and the environment, 10, 2004. Environmental Agency, Environment and Heritage Service, Food Standards Agency and Scottish Environment Protection Agency; 2005. ISSN 1365-6414. Chanarin I. The megaloblastic anaemia. London: Blackwell Scientific Oxford; 1979. Chaney RL. Potential effects of waste constituents on the food chain. Park Ridge, NJ: Noyes Data Corp; 1983. Chatterjee J, Chatterjee C. Phytotoxicity of cobalt, chromium and copper in cauliflower. Environ Pollut 2000;109(1):69–74. Chatterjee J, Chatterjee C. Deterioration of fruit quality for tomato by excess cobalt and its amelioration. Commun Soil Sci Plant Anal 2005;36: 1931–45. Cloutier N, Clulow F, Lim T, Dave N. Metal (Cu, Ni, Fe, Co, Zn, Pb) and Ra226 levels in meadow voles Microtus pennsylvanicus living on nickel and

836

J. Gál et al. / Environment International 34 (2008) 821–838

uranium tailings in Ontario, Canada: environmental and tissue levels. Environ Pollut 1985;10:19–46. Coleman CM, Boyd RS, Eubanks MD. Extending the elemental defense hypothesis: dietary metal concentrations below hyperaccumulator levels could harm herbivores. J Chem Ecol 2005;31:1669–81. Crossley DA, Blood ER, Hendrix PF, Seastedt TR. Turnover of cobalt-60 by earthworms (Eisenia foetida) (Lumbricidae, Oligochaeta). Appl Soil Ecol 1995;2(2):71–5. De Boeck M, Kirsch-Volders M, Lison D. Cobalt and antimony: genotoxicity and carcinogenicity. Mutat. Res. 2003;533:135–52. Dekock PC. Heavy metal toxicity and iron chlorosis. Ann Bot 1956;20:133–41. Delwiche CC, Johnson CM, Reisenauer HM. Influence of cobalt on nitrogen fixation by medicago. Plant Physiol 1961;36(1):73–8. Djingova R, Kuleff I. Concentration of caesium-137, cobalt-60 and potassium40 in some wild and edible plants around a nuclear power plant in Bulgaria. J Environmn Radioact 2002;59:61–73. Downs RT, Hall-Wallace M. The American Mineralogist Crystal Structure Database. Am Mineral 2003;88:247–50. Dzantor EK. Phytoremediation-Plant-Based Strategies for Cleaning Up Contaminated Soils, vol. 768. University of Maryland Extension Services Fact Sheet; 2004. p. 1–4. EC, Revision of the 1996 Technical Guidance Document (TGD) in Support of the Commission Directive 93/67/EEC on Risk Assessment for New Notified Substances and the Commission Regulation (EC) No. 1488/94 on Risk Assessment for Existing Substances and Directive of the European Parliament and of the Council Concerning the Placing of Biocidal Products on the Market. 2003. Efroymson RA, Will ME, Suter GW. Toxicological benchmarks for contaminants of potential concern for effects on soil and litter invertebrates and heterotrophic process: 1997. Revision. ES/ER/TM-126/R2; 1997. p. 1–151. Elinder CG. Health hazards from exposure to cobalt, with special reference to carcinogenic, mutagenic and teratogenic effects. Toxicol Environ Chem 1984;7:251–6. Elinder CG, Friberg L. In: Friberg L, Nordberg GF, Vouk VB, editors. Cobalt in handbook on the toxicology of metals, vol. 2. Amsterdam: Elsevier Science; 1986. p. 211–32. EMFEMA. Bioavailability of major and trace minerals. International Association of the European (EU) Manufacturers of Major, Trace and Specific Feed Mineral Materials. Brussels; 2001. EURAS. Fact Sheet 8-MERAG Program-Building Block: Bioconcentration, bioaccumulation and biomagnification of metals. Ghent, Belgium: EURAS; 2005. FAO/WHO. Human vitamin and mineral requirements. Rome: FAO and WHO; 2001. Frank A, McPartlin J, Danielsson R. Nova-Scotia moose mystery — a moose sickness related to cobalt and vitamin B12 deficiency. Sci Total Environ 2004;318:89–100. FOREGS De Vos W, Tarvainen T, (Chief-editors), Salminen R, Reeder S, De Vivo B, et al., 2006. Geochemical Atlas of Europe, Part 2: Interpretation of Geochemical Maps, Additional Tables, Figures, Maps, and Related Publications. Geological Survey of Finland, Espoo, 690 pp. [Printed version ISBN-951-690-956-6; electronic version ISBN-951-690-960-4]. 2006. FSA. Risk Assessment: Cobalt. UK: Food Standards Agency; 2003. Gál J, Markiewicz-Patkowska J, Hursthouse A, Tatner P, in press. Metal uptake by woodlice in urban soils, Ecotoxicol Environm Saf. doi:10.1016/j.ecoenv.2007.01.002. ____________________________. Gao Y, Kan AT, Tomson MB. Critical evaluation of desorption phenomena of heavy metals from natural sediments. Environ Sci Technol 2003;37:5566–73. Gault AJ, Polya DA, Lythgoe PR, F'Arquhar ML, Charnock JM, Wogelius RA. Arsenic speciation in surface waters and sediments in a contaminated waterway: an IC-ICP-MS and XAS based study. Appl Geochem 2003;18(9):1387–97. Gavloski JE, Lamb RJ. Specific impacts of herbivores: comparing diverse insect species on young plants. Environ Entomol 2000;29(1):1–7. Giller KE, Witter E, McGrath SP. Toxicity of heavy metals to microorganisms and microbial processes in agricultural soils: a review. Soil Biol Biochem 1998;30:1389–414. Giller KE, Witter E, McGrath SP. Assessing risks of heavy metal toxicity in agricultural soils: do microbes matter? Hum Ecol Risk assess 1999;5:683–9.

Giroussi S, Voulgaropoulos A, Ayiannidis AK, Golimowski J, Janicki M. Determination of cobalt in vegetable animal foodstuffs by differential pulse adsorptive voltammetry using alpha-benzil dioxime. Sci Total Environ 1995;176(1–3):135–9. Gwo J-P, Mayes MA, Jardine PM. Quantifying the physical and chemical mass transfer processes for the fate and transport of Co(II)EDTA in a partially weathered limestone–shale saprolite. J. Contam Hydrol 2007;90: 184–202. Haga Y, Clyne N, Hatroi N, Hoffman-Bang C, Pehrsson SK, Ryden L. Impaired myocardial function following chronic cobalt exposure in an isolated rat heart model. Trace Elem Electrolytes 1996;13(2):69–74. Hamilton EI. The geobiochemistry of cobalt. Sci Total Environ 1994;150(1–3):7–39. Hankard PK, Svendsen C, Wright J, Wienberg C, Fishwick SK, Spurgeon DJ, Weeks JM. Biological assessment of contaminated land using earthworm biomarkers in support of chemical analysis. Sci Total Environ 2004;330(1–3):9–20. Helmke PM, Robarge WP, Korotev RL, Schomberg PJ. Effects of soil-applied sewage sludge on concentrations of elements in earthworms. J Environ Qual 1979;8:322–7. Hendrickx F, Maelfait J-P, Bogaert N, Tojal C, Du Laing G, Tack FM, Verloo MG. The importance of biological factors affecting trace metal concentration as revealed from accumulation patterns in co-occurring terrestrial invertebrates. Environ Pollut 2004;127:335–41. Herzig R, Urech M, Liebendorfer L, Ammann K, Guecheva M, Landolt W. Lichens as biological indicators of air pollution in Switzerland. In: Leith H, Markert B, editors. Element Concentration Cadasters in Ecosystems. Weinheim: VCH; 1990. p. 317–32. Howarth D, Hulbert A, Horning D. A comparative study of heavy metal accumulation in tissues of the crested tern, Sterna Bergii, breeding near industrialized and non-industrialized areas. Aust Wildl Res 2002;8(3):665–72. HSDB. Cobaltous sulfate. Hazardous substances data bank. Bethesda, MD: National Library of Medicine; 2000. Hubbard DM, Creech FM, Cholak J. Determination of cobalt in air and biological material. Arch Environ Health 1966;13(2):190–4. Huijbregts MAJ. Priority assessment of toxic substances in the frame of LCA: development and application of the multi-media fate, exposure and effect model USES-LCA; 1999. Hursthouse AS. The relevance of speciation in the remediation of soils and sediments contaminated by metallic elements — an overview and examples from Central Scotland, UK. J Environ Monit 2001;3(1):49–60. Imray P, Langley A. Health-based soil investigation levels. National Environmental Health Forum Monographs, Soil 1999(1):45. Ireland MP. Heavy metal uptake and metal distribution in earthworms. Earthworm ecology from Darwin to vermiculture. Satchell JE London Chapman & Hall; 1983. p. 247–65. Janssens E, Dauwe T, Bervoets L, Eens M. Heavy metals and selenium in feathers of great tits (Parus major) along a pollution gradient. Environ Toxicol Chem 2001;20(12):2815. Jhee EM, Boyd RS, Eubanks MD. Nickel hyperaccumulation as an elemental defense of Streptanthus polygaloides (Brassicaceae): influence of herbivore feeding mode. New Phytol 2005;168:331–44. Kabata-Pendias A, Pendias H. Trace Elements in Soils and Plants. Boca Raton: CRS Press Inc.; 1992. Kavun VY. Heavy metals in organs and tissues of the European black vulture (Aegypius monachus): dependence on living conditions. Russ J Ecol 2004;35(1):51–4 (4). Kay JT, Conklin MH, Fuller CC, O'Day PA. Processes of nickel and cobalt uptake by a manganese oxide forming sediment in Pinal Creek, Globe Mining, Arizona. Environ Sci Technol 2001;35:4719–25. Keeling SM, Stewart RB, Anderson CWN, Robinson BH. Nickel and Cobalt phytoextraction by the hyperaccumulator Berkheyda coddii: implications for polymetallic phytomining and phytoremediation. Int J Pytoremediat 2003;5:235–44. Khan ZI, Ashraf M, Hussain A, McDowell LR. Seasonal variation of trace elements in semiarid veldt pasture. Commun Soil Sci Plant Anal 2006;37:1471–84. Koutsospyros A, Braida W, Christodoulatos C, Dermatas D, Strigul N. A review of tungsten: from environmental obscurity to scrutiny. J Hazard Mater 2006;136:1–19.

J. Gál et al. / Environment International 34 (2008) 821–838 Krämer U, Talke IN, Hanikenne M. Transition metal transport. FEBS Lett 2007;581:2263–72. Krauskopf KB, Bird DK. Introduction to geochemistry. New York: McGraw Hill; 1995. Krupka KM, Serne RJ. Geochemical factors affecting the behavior of antimony, cobalt, europium, technetium, and uranium in vadose sediments. Washington: Pacific Northwest National Laboratory; 2002. Kukiera U, Petersb CA, Chaneya RL, Angleb JS, Rosebergc RJ. The effect of pH on metal accumulation in two Alyssum species. J Environ Qual 2004;33: 2090–102. Leblanc JC, Guerin T, Noel L, Calamassi-Tran G, Volatier JL, Verger P. Dietary exposure estimates of 18 elements from the 1st French Total Diet Study. Food Addit. Contam. 2005;22(7):247–65. Leonard A, Lauwerys R. Mutagenicity, carcinogenicity and teratogenicity of cobalt metal and cobalt compounds. Mutat Res 1990;239:17–27. Li Z, McLaren RG, Metherell AK. The availability of native and applied soil cobalt to ryegrass in relation to soil cobalt and manganese status and other soil properties. N Z J Agric Res 2004;47:33–43. Lighthart B, Bond H, Ricard M. Trace elements research using coniferous forest soil/litter microcosms; 1977. USEPA-600/3-77-091. Lindemann WC, Glover CR. Nitrogen fixation by legumes. Cooperative Extension Service, College of Agriculture and Home Economics. New Mexico State University; 2003. Lison D, De Boeck M, Verougstraete V, Kirsch-Volders M. Update on genotoxicity and carcinogenicity of cobalt compounds. Occup Environ Med 2001;58:619–25. Liu J, Reid RJ, Smith FA. The mechanism of cobalt toxicity in mung beans. Physiol Plant 2000;110:104–10. Lock K, Because S, Criel P, Van Eeckhout H, Janssen CR. Exotoxicity of cobalt to the springtail Folsomia candida. Comparative Biochemistry & Physiology, Part C 2004;139(4):195–9. Lock K, De Schamphelaere KAC, Because S, Criel P, Van Eeckhout H, Janssen CR. Development and validation of an acute biotic ligand model (BLM) predicting cobalt toxicity in soil to the potworm Enchytraeus albidus. Soil Biol Biochem 2006;38:1924–1932. Loring Air Force Base Ecological Risk Methodology. Martin Marietta Energy Systems, Inc. 1994 Draft. Lux D, Kammerer L, Ruhm W. Cycling of Pu, Sr, Cs, and other long living radionuclides in forest ecosystems of the 30-km zone around Chernobyl. Sci Total Environ 1995;173/174:375–84. Meharg AA. Mechanisms of plant resistance to metal and metalloid ions and potential biotechnical applications. Plant Soil 2005;274:163–74. Malaisse F, Baker AJM, Ruelle S. Diversity of plant communities and leaf heavy metal content at Luiswishi copper/cobalt mineralization, Upper Katanga, Dem. Rep. Congo. Biotechnol Agron Soc Environ 1999;3(2):104–14. Marschner H. Mineral nutrition of higher plants. 2nd Edition. London: Academic Press; 1995. Metcheva R, Yurukova L, Teodorova S, Nikolova E. The penguin feathers as bioindicator of Antarctica Environmental State. Sci Total Environ 2006;362:259–65. McKenzie RM. The manganese oxides in soils. A review. Z Pflanzenernahr Bodenkd 1972;131:221–42. McLaren RG, Lawson DM, Swift RS. Sorption and desorption of cobalt by soils and soil components. J Soil Sci 1986;37:223. MOE. Air quality monitoring and vegetation sampling studies — cobalt area 1976–1978. Air quality assessment, Northeastern Region. Ontario Ministry of the Environment; 1979. Mollah AS, Begum A. A study on transfer factors of Co-60 and Zn-65 from soil to plants in the tropical environment of Bangladesh. Environ Mon Assess 2001;68:91–7. Mossop KF, Davidson CM. Comparison of original and modified BCR sequential extraction procedures for the fractionation of copper, iron, lead, manganese and zinc in soils and sediments. Anal Chim Acta 2003;478(1): 111–8. Mustafa G, Kookana RS, Singh B. Desorption of cadmium from goethite: effects of pH, temperature and aging. Chemosphere 2006;64:856–65. Nagpal NK. Water quality guidelines for cobalt. Ministry of Water, Land and Air Protection, Water Protection Section, Water, Air and Climate Change Branch, Victoria; 2004.

837

Neuhauser EF, Meyer JA, Malencki MR, Thomas JM. Dietary cobalt supplements and the growth and reproduction of the earthworm Eisenia foetida. Soil Biol Biochem 1984;16:521–3. NRC. Mineral tolerance of domestic animals. National Research Council. Washington, D.C.: National Academy of Sciences; 1980. OECD, Statement by the Joint Meeting of the Chemicals Committee and Working Party on Chemicals, Pesticides and Biotechnology, endorsed February 2007. http://www.oecd.org/document/52/0,2340,en_2649_ 201185_37999156_1_1_1_1,00.html [accessed: Tuesday, 17 April 2007]. O'Hara GW. Nutritional constraints on root nodule bacteria affecting symbiotic nitrogen fixation: a review. Aust J Exp Agric 2001;41(3):417–33. Osunaa BM, van Hullebuscha ED, Zandvoorta MH, Izab J, Lensa PNL. Effect of cobalt sorption on metal fractionation in anaerobic granular sludge. J Environ Qual 2004;33:1256–70. Ozanne PG, Greenwood EAN, Shaw TC. The cobalt requirement of subterranean clover in the field. Aust J Agric Res 1963;14(1):39–50. Ozdemir G, Ceyhan N, Manav E. Utilization of an exopolysaccharide produced by Chryseomonas luteola TEM05 in alginate beads for adsorption of cadmium and cobalt ions. Bioresour Technol 2005;96(15):1677–82. Page V, Feller U. Selective transport of zinc, manganese, nickel, cobalt and cadmium in the root system and transfer to the leaves in young wheat plants. Ann Bot 2005;96:425–34. Page V, Weisskopf L, Feller U. Heavy metals in white lupin: uptake, root-toshoot transfer and redistribution within plant. New Phytol 2006;171:329–41. Pal A, Ghosh S, Paul AK. Biosorption of cobalt by fungi from serpentine soil of Andama. Bioresource Technol 2006;97:1253–8. Palit S, Sharma A, Talukder G. Effects of cobalt on plants. Bot Rev 1994;60(2):149–81. Paveley CF. Heavy metal sources and distribution in soils, with special reference to Wales. Bradford UK: University of Bradford; 1988. Perez-Espinosa A, Moreno-Caselles J, Moral R, Perez-Murcia MD, Gomez I. Effect of increased cobalt treatments on sewage sludge amended soil: nitrogen species in soil and tranference to tomato plants. Arch Agron Soil Sci 2002;48(3):273–8. Peterson LR, Trivett V, Baker AJM, Aguiar C, Pollard AJ. Spread of metals through an invertebrate food chain as influenced by a plant that hyperaccumulated nickel. Chemoecology 2003;13:103–8. Pinkerton BP, Brown KW, Thomas JC, Zabcik D. Plant uptake of cobalt from a petroleum sludge affected soil. Agron Abstr 1981;32. Primavesi A, Primavesi O. Optimising climate–soil–pasture–cattle interactions in Brazil. Leisa Magazine; April, 12–13. 2002. PTI. Remedial Investigation and Feasibility Study at the National Zinc Site. Ecological Risk Assessment. Operable Unit 2. PTI Environmental Services, Bellevue, Wash, vol. IV; 1995. PTI. National Zinc Site Remedial Investigation Feasibility Study. Ecological Risk Assessment-Operable Unit 2. PTI Environmental Services, In Bechtel, vol. IV; 1998. Pyatt FB. Comparison of foliar and stem bioaccumulation of heavy metals by Corsican pines in the Mount Olympus area of Cyprus. Ecotoxicol Environ Saf 1999;42(1):57–61. Quevauviller P. Operationally-defined extraction procedures for soil and sediment analysis. Part 3: New CRMs for trace-element extractable contents. Trac-Trends Anal Chem 2002;21(11):774–85. Rainbow PS. Trace metal concentrations in aquatic invertebrates: why and so what? Environ Pollut 2002;120:497–507. Rehab FI, Wallace A. Excess trace metal effects on cotton: 2. Copper, zinc, cobalt and manganese in Yolo Loam Soil. Commun Soil Sci Plant Anal 1978;9(6):519–27. Reimann C, Koller F, Kashulina G, Niskavaara H, Englmaier P. Influence of extreme pollution on the inorganic chemical composition of some plants. Environ Pollut 2001;115(2):239–52. Reisenauer HM. Cobalt in nitrogen fixation by a legume. Nature 1960;186: 375–6. Robinson BH, Brooks RR, Clothier BE. Soil amendments affecting nickel and cobalt uptake by Berkheya coddii: potential use for phytomining and phytoremediation. Ann Bot 1999a;84:689–94. Robinson BH, Brooks RR, Hedley MJ. Cobalt and nickel accumulation in Nyssa (tupelo) species and its significance for New Zealand agriculture. N Z J Agric Res 1999b;42:235–40.

838

J. Gál et al. / Environment International 34 (2008) 821–838

Roth-Holzapfel M. Multi-element analysis of invertebrate animals in a forest ecosystem (Picea abies L.). In: Leith H, Markert B, editors. Element Concentration Cadasters in Ecosystems. VCH Weinheim; 1990. p. 280–95. Ruelas-Inzunza J, Paez-Osuna F, Soto LA. Bioaccumulation of Cd, Co, Cr, Cu, Fe, Hg, Mn, Ni, Pb and Zn in trophosome and vestimentum of the tube worm Riftia pachyptila from Guaymas basin, Gulf of California. Deep-sea Res, Part 1, Oceanogr Res Pap 2005;52(7):1319–23. Sabbarese C, Stellato L, Cotrufo MF, D'Onofrio A, Ermice A, Terrasi F, Alfieri S. Transfer of Cs-137 and Co-60 from irrigation water to soil–tomato plant system. J Environ Radioact 2002;61:21–31. Sadhra S, Wheatley A, Cross H. Dietary exposure to copper in the European Union and its assessment for EU regulatory risk assessment. Sci Total Environ 2007;374:223–34. Samecka-Cymermann A, A M, Kempers AJ. Bioindicators of heavy metals in soil by liverworts. Arch. Environ. Contam. Toxicol 1997;33:162–71. Sample BE, Beauchamp JJ, Efroymson RA, Suter GW. Development and validation of bioaccumulation models for small mammals. Lockheed Martin energy Systems Inc., ES/ER/TM-219. Tennessee: Oak Ridge; 1998. Sasmaz A, Yaman M. Distribution of chromium, nickel and cobalt in different parts of plant species and soil in mining area of Keban, Turkey. Commun Soil Sci Plant Anal 2006;37:1845–57. Schroeder HA, Nason AP, Tipton IH. Essential trace metals in man: cobalt. J Chronic Dis 1967;11:869–90. Shinonaga T, Schimmack W, Gerzabek MH. Vertical migration of Co-60, Cs137 & Ra-226 in agricultural soils as observed in lysimeters under crop rotation. J Environ Radioact 2005;79:93–106. Siegel FR. Environmental geochemistry of potentially toxic metals. New York: Springer-Verlag; 2001. Smith IC, Carson BL. Trace metals in the environment. Cobalt, vol. 6. Ann Arbor: Ann Arbor Science Publ. Inc.; 1981. Sočo E, Kalembkiewicz J. Investigations of chemical fractionation of Co and Ni in industrial fly ash and mobility of metals in environmental conditions. Chemosphere 2007;67:359–64. Soremark R, Diab M, Ardvidson K. Autoradiographic study of distribution patterns of metals which occur as corrosion products from dental restorations Scand. J Dent Res 1979;87:450–8. Sorokin N, Weeks JM, Comber S, Gardner M, Mitchell R, Pinot de Moira A, Whitehouse P. Assessment of existing data for the effect of cobalt exposure in the sediment and terrestrial compartments. Bucks: Cobalt Development Institute; 2002. SRS. Bioaccumulation and bioconcentration screening. Environmental Restoration Division; 1999. ERD-AG-003 Savannah River Site. Stopford W, Turner J, Cappellini D, Brock T. Bioaccessibility testing of cobalt compounds. J Environ Monit 2003;5:675–80. Suttle NF, Bell J, Thornton I, Agyriaki A. Predicting the risk of cobalt deprivation in grazing livestock from soil composition data. Environ Geochem Health 2003;25(1):33–9. Tagami K, Uchida S. Aging effect on bioavailability of Mn, Co, Zn & Tc in Japanese agricultural soils under waterlogged conditions. Geoderma 1998;84:3–13. Talmage SS, Walton BT. Small mammals as monitors of environmental contaminants. Rev. Environ. Contam. Toxicol. 1991;119:47–145.

Taylor A, Marks V. Cobalt: a review. J Hum Nutr 1978;32:165–77. Tessier A, Campbell PG, Bisson M. Sequential extraction procedure for the speciation of particulate trace metals. Anal Chem 1979;51:844–51. TN & Associates, I. Plant toxicity testing to support development of ecological soil screening levels. Washington, D.C.: National Center for Environ. Assess.; 2000. Tyler G. Changes in the concentrations of major, minor and rare-earth elements during leaf senescence and decomposition in a Fagus sylvatica forest. For Ecol & Manag 2005;206:167–77. Underwood EJ, NiFa S. The mineral nutrition of livestock. Wallingford: CABI Publishing; 1999. USEPA. Guidance for developing ecological soil screening levels. Washington: U.S. Environmental Protection Agency; 2003. USEPA. Ecological soil screening levels for cobalt. Washington, DC: US EPA; 2005. Van den Berg H, Dagnelie H, van Staveren WA. Vitamin B12 and seaweed 242–243. Lancet 1998;1:242–3. Veerle J, Tom D, Rianne P, Lieven B, Ronny B, Marcel E. The importance of exogenous contamination on heavy metal levels in bird feathers. A field experiment with free-living great tits, Parus major. J Enviro Monit 2004;6:356–60. VROM, Bodemsanering L. Guidelines for soil clean up. Netherlands Ministry of Housing, Planning and Environment, Soil, Water and Chemical Substances Department. The Hague; 1983. Wagner G. Variability of element concentrations in tree leaves depending on sampling parameters. In: Leith H, Markert B, editors. Element Concentration Cadasters in Ecosystems. VCH Weinheim; 1990. p. 41–54. Watmough SA, Dillon PJ, Epova EN. Metal partitioning and uptake in central Ontario forests. Environ Pollut 2005;134(3):493–502. Wijnhoven S, Leuven RSEW, van der Velde G, Jungheim G, Koelemij EI, de Vries FT, Eijsackers HJP, Smits AJM. Heavy-metal concentrations in small mammals from a diffusely polluted floodplain: importance of species- and locationspecific characteristics. Arch Environ Contam Toxicol 2007;52:603–13. Woodard TL, Amarasiriwardena D, Shrout K, Xing B. Roadside accumulation of heavy metals in soils in Franklin County MA and surrounding towns. Communi Soil Sci Plant Anal 2007;38:1087–104. Yamagata N, Kurioka W, Shimizu T. Balance of cobalt in Japanese people and diet. J Radiat Res (Tokyo) 1963;28:8–15. Ye ZH, Whiting SN, Lin ZQ, Lytle CM, Qian JH, Terry N. Removal and redistribution of iron, manganese, cobalt and nickel within a Pennsylvania constructed wetland treating coal combustion by-product. J Environ Qual 2001;30:1464–73. Young RS. Cobalt in biology and biochemistry. New York: Academic Press; 1979. Young SD, Zhang H, Tye AM, Maxted A, Thums C, Thornton I. Characterising the availability of metals in contaminated soils I: the solid phase: sequential extraction and isotope dilution. Soil Use Man 2006;21:450–8. Zhang H, Young SD. Characterising the availability of metals in contaminated soils II: the soil solution. Soil Use Man 2006;21:459–67.