High-temperature calcination products of diiron trioxide and amorphous silica resulting in a glassy silica matrix

Sediment toxicity

Administrative data

Endpoint:

sediment toxicity: long-term

Data waiving:

other justification

Justification for data waiving:

other:

Justification for type of information:

JUSTIFICATION FOR DATA WAIVING
According to Column 2 of Information Requirement 9.5.1, Annex X, Commission Regulation (EU) 1907/2006 ” Long-term toxicity testing shall be proposed by the registrant if the results of the chemical safety assessment indicates the need to investigate further the effects of the substance and/or relevant degradation products on sediment organisms. The choice of the appropriate test(s) depends on the results of the chemical safety assessment.”

According to Section 8.4.2 of ECHA Guidance on IR & CSA, Part B: Hazard assessment (Version 2.1; ECHA, 2011), “For substances which are classified as harmful, toxic or very toxic to aquatic life (i.e. H412, H411, H410 and H400), an aquatic PNEC can be derived. In these circumstances there are unclassified hazards to the sediment and soil compartments because toxicity to aquatic organisms is used as an indicator of concern for sediment and soil organisms, and a screening risk characterisation is undertaken using the equilibration partitioning method (EPM) to derive PNECs for sediment and soil. Hence quantitative exposure assessment, i.e. derivation of PECs, is mandatory for the water, sediment and soil environmental compartments.

Substances with the only environmental classification as ‘May cause long lasting harmful effects to aquatic life’ (i.e. H413) have been established as persistent in the aquatic environment and potentially bioaccumulative on the basis of test or other data. There are also potential hazards for these substances for the sediment and soil compartments, because these substances are potentially bioaccumulative in all organisms and are also potentially persistent in sediment and soil. Hence exposure assessment is mandatory for the water, sediment and soil environmental compartments, which may be quantitative or qualitative as appropriate. PBT and vPvB substances have been established as persistent and bioaccumulative (and the former also as toxic) in the environment as a whole. Hence qualitative exposure assessment is mandatory for the water, sediment and soil environmental compartments…

If there are ecotoxicity data showing effects in aquatic organisms, but the substance is not classified as dangerous for the aquatic environment, an aquatic PNEC can nevertheless be derived thus indicating a hazard to the aquatic environment. In these circumstances there are also unclassified hazards to the sediment and soil compartments because toxicity to aquatic organisms is used as an indicator of concern for sediment and soil organisms and a screening risk characterisation is undertaken using the equilibration partitioning method (EPM) to derive PNECs for sediment and soil. Hence quantitative exposure assessment, i.e. derivation of PECs, is mandatory for the water, sediment and soil environmental compartments.”

High-temperature calcination products of diiron trioxide and amorphous silica resulting in a glassy silica matrix can be considered environmentally and biologically inert due to the characteristics of the synthetic process (calcination at a high temperature of approximately 1000°C), rendering the substance to be of a unique, stable crystalline structure in which all atoms are tightly bound and not prone to dissolution in environmental and physiological media. This assumption is supported by available transformation/dissolution data (Grané, 2010) that indicate a very low release of pigment components. Transformation/dissolution of High-temperature calcination products of diiron trioxide and amorphous silica resulting in a glassy silica matrix (24-screening test according to OECD Series 29, loading of 100 mg/L, pH 6 and 8) resulted in mean dissolved iron concentrations of 0.131 µg/L Fe and 0.002 µg/L Fe, silicon concentrations of 215.7 µg/L Si and 204 µg/L Si at pH 6 and 8, respectively. Dissolution of iron and silicon is highest at pH 6, therefore pH 6 is considered as pH that maximised metal release. Metal release at the 1 mg/L loading and pH 6 remained below the respective LOD for iron and silicon (<0.22 µg/L Fe and <0.07 µg/L Si). After 28 days at the 1 mg/L loading and pH 6 iron concentrations of 0.29 µg/L were measured whereas silicon concentrations remained below the LOD (< 0.07 µg/L Si). Thus, the rate and extent to which High-temperature calcination products of diiron trioxide and amorphous silica resulting in a glassy silica matrix produces soluble (bio)available ionic and other silicon- and iron-bearing species in environmental media is limited. Hence, the pigment can be considered as environmentally and biologically inert during short- and long-term exposure. The poor solubility of High-temperature calcination products of diiron trioxide and amorphous silica resulting in a glassy silica matrix is expected to determine its behaviour and fate in the environment, and subsequently its potential for ecotoxicity.

Proprietary studies are not available for High-temperature calcination products of diiron trioxide and amorphous silica resulting in a glassy silica matrix. The poorly soluble substance High-temperature calcination products of diiron trioxide and amorphous silica resulting in a glassy silica matrix is evaluated by comparing the dissolved metal ion levels resulting from the transformation/dissolution test after 7 and 28 days at a loading rate of 1 mg/L with the lowest acute and chronic ecotoxicity reference values (ERVs) as determined for the (soluble) metal ions. The ERVs are based on the lowest EC50/LC50 or NOEC/EC10 values for algae, invertebrates and fish. Acute and chronic ERVs were obtained from the Metals classification tool (MeClas) database as follows: For iron ions, the acute and chronic ERVs are above 1 mg/L, respectively, and a concern for short-term (acute) and long-term (chronic) toxicity was not identified (no classification). An acute as well as chronic ERV for silicon has not been derived since a concern for short-term (acute) and long-term (chronic) toxicity of silicon ions was not identified (see also OECD, 2004). According to ECHA Guidance on the Application of the CLP Criteria (Version 5.0, July 2017), “Where the acute ERV for the metal ions of concern is greater than 1 mg/l the metals need not be considered further in the classification scheme for acute hazard.” Further, ” Where the chronic ERV for the metal ions of concern corrected for the molecular weight of the compound (further called as chronic ERV compound) is greater than 1 mg/l, the metal compounds need not to be considered further in the classification scheme for long-term hazard.” Due to the lack of an acute and chronic aquatic hazard potential for soluble silicon and iron ions and the fact that dissolved silicon and iron concentrations were below the LOD (<0.22 Fe and <0.07 µg/L Si) after 7 days and below 0.5µg/L (0.29 Fe and <0.07 µg/L Si) after 28 days at pH 6 in the T/D test, respectively, it can be concluded that the substance High-temperature calcination products of diiron trioxide and amorphous silica resulting in a glassy silica matrix is not sufficiently soluble to cause short- or long-term toxicity at the level of the acute or chronic ERVs (expressed as EC50/LC50 or NOEC/EC10, respectively).

In accordance with Figure IV.4 “Classification strategy for determining acute aquatic hazard for metal compounds” and Figure IV.5 „Classification strategy for determining long-term aquatic hazard for metal compounds “of ECHA Guidance on the Application of the CLP Criteria (Version 5.0, July 2017) and section 4.1.2.10.2. of Regulation (EC) No 1272/2008, the substance High-temperature calcination products of diiron trioxide and amorphous silica resulting in a glassy silica matrix is poorly soluble and does not meet classification criteria for acute (short-term) and chronic (long-term) aquatic hazard.

High-temperature calcination products of diiron trioxide and amorphous silica resulting in a glassy silica matrix is not classified as dangerous for the aquatic environment, an aquatic PNEC cannot be derived, thus not indicating a hazard to the aquatic environment. In these circumstances there are also not unclassified hazards to the sediment compartment because toxicity to aquatic organisms is used as an indicator of concern for sediment organisms and a screening risk characterisation (using the equilibration partitioning method to derive a PNEC for sediment) cannot be undertaken. Thus, High-temperature calcination products of diiron trioxide and amorphous silica resulting in a glassy silica matrix does not have a “non-classified hazard” potential.

Silicon is naturally abundant in sediments and a significant fraction of total silicon is associated with the solid phase (e.g., quartz). The major mineral phases of stream sediment are generally silicates or carbonates. Weathering of minerals is the primary source of Si in fresh water (Salminen et al. 2005).

Monitoring data for silicon background concentrations in stream water and sediments are provided by the FOREGS Geochemical Baseline Mapping Programme, which offers high quality, multi-purpose homogeneous environmental geochemical baseline data for Europe. A total of 743 stream sediment samples were processed in the FOREGS-program for the EU-27, UK and Norway to determine representative silicon stream sediment concentrations. Silicon sediment concentrations range from 11,691.1 to 439,586.6 mg/kg with 5th, 50th and 95th percentiles of 161,384.4, 290,875.4 and 385,246.2 mg/kg, respectively.

The mobility of iron in water is strongly controlled by the redox conditions and the abundance of complexing agents in the aqueous solution. Under most natural conditions in stream water, iron is relatively immobile and precipitates rapidly, mainly due to the very low solubility of iron (III) hydroxide in its various forms. Its solubility is strongly influenced by redox conditions (Salminen et al. 2005). Based on the FOREGS dataset, iron concentrations of sediments range from 0.6 to 192.4 g Fe/kg with 5th, 50th and 95th percentiles of 5.8, 20.1 and 45.3 g Fe/kg, respectively.

Taking into account the high quality and representativeness of the FOREGS data set, the 95th percentiles of 385,246.2 mg Si/kg and 45.3 g Fe/kg sediment can be considered as representative background concentration of silicon and iron in European stream sediments.

Regarding essentiality, “Silicon is considered necessary for various functions in some species, including diatom algae, gastropods and mammals. Silicon deficiency in animals may lead to delays in growth, bone deformations and abnormal skeletal development, and one of the symptoms of silicon deficiency is aberrant connective and bone tissue metabolism (Pérez-Granados and Vaquero, 2002).

Iron is an essential trace element for living organisms including animals, plants and microorganisms. It serves as a cofactor in a variety of enzymes involved in the electron transport processes of various metabolic pathways, i.e., photosynthesis, respiration, and nitrogen fixation. It is a structural component of the metalloenzyme nitrogenase, which is the key enzyme in the nitrogen fixation process performed by different microorganisms. Due to its involvement in the synthesis and maintenance of chlorophyll, iron is a vital factor for the photosynthetic performance of plants. As an essential component of the haemoglobin of red blood cells, iron functions as a carrier of oxygen in the blood and muscles of animals (e.g. US EPA, 2003; Colombo et al. 2014).

Regarding the potential of bioaccumulation the OECD (2004) states, “the bioavailable forms of silica (SiO2) are dissolved silica [Si(OH)4] almost all of which is of natural origin. The ocean contains a huge sink of silica and silicates where a variety of the marine habitat (diatoms, radiolarians, and sponges) is able to exploit this resource as a construction material to build up their skeletons”. Most organisms contain silicon at least at trace levels. Whereas silicon is essential for some organisms, including diatom algae, gastropods and mammals, and actively taken up, others take it up passively and excrete it.

“Due to the known inherent physico-chemical properties, absence of acute toxic effects as well as the ubiquitous presence of silica/silicates in the environment, there is no evidence of harmful long-term effects arising from exposure to synthetic amorphous silica/silicates (OECD, 2004).” Thus, given the ubiquitous presence of silica and silicates in the environment, silicon is regarded as element without or with a very low potential for bioconcentration and bioaccumulation.

The uptake of iron as essential element into cells is actively regulated by a strict homeostatic control system. The active regulation of iron uptake in combination with internal detoxification mechanism indicates a low potential for iron bioaccumulation. This assumption is supported by results of Bustamante et al. (2000) indicating that iron concentrations of digestive glands of cephalopods living in natural and in iron-enriched habitats are similar. Winterbourn et al. (2000) further demonstrate that iron does not biomagnify but rather “biodilutes” up the aquatic food chain. Thus, the potential for bioaccumulation of iron in aquatic environments can be expected to be low.

Considering abundance, bioavailability, essentiality and low bioaccumulation, the potential of silicon and iron ions for toxicity to sediment organisms can be expected to be low.

High-temperature calcination products of diiron trioxide and amorphous silica resulting in a glassy silica matrix is not classified as harmful, toxic or very toxic to aquatic life or may cause long lasting harmful effects to aquatic life. High-temperature calcination products of diiron trioxide and amorphous silica resulting in a glassy silica matrix is also not an unclassified hazard to the aquatic environment. Based on the poor solubility, bioavailability, lack of a potential for bioaccumulation and toxicity to aquatic organisms and considering ubiquitousness and bioavailability of silicon and iron in sediment as well as essentiality and low bioaccumulation of silicon and iron, High-temperature calcination products of diiron trioxide and amorphous silica resulting in a glassy silica matrix is also not considered an unclassified hazard to the sediment compartment. Results of the chemical safety assessment do not indicate the need to investigate further the effects of High-temperature calcination products of diiron trioxide and amorphous silica resulting in a glassy silica matrix on sediment organisms. Therefore, the study on the long-term toxicity to sediment organisms does not need to be conducted in accordance with Column 2 of Information Requirement 9.5.1., Annex X, Commission Regulation (EU) 1907/2006.

References:

Bustamante et al. (2000) Bioaccumulation of 12 trace elements in the tissues of the nautilus Nautilus macromphalus from New Caledonia. Marine Pollution Bulletin 40/8: 688-696.

Colombo et al. (2014) Review on iron availability in soil: interaction of Fe minerals, plants, and microbes. Journal of Soils and Sediments 14: 538–548.

US EPA (2003) Ecological Soil Screening Level for Iron, Interim Final, OSWER Directive 9285.7-69.

Salminen et al. (2005) Geochemical Atlas of Europe - Part 1: Background information, Methodology and Maps. EuroGeoSurveys.

Winterbourn et al. (2000) Aluminium and iron burdens of aquatic biota in New Zealand streams contaminated by acid mine drainage: effects of trophic level. The Science of The Total Environment 254, 45-54.

Pérez-Granados and Vaquero (2002) Silicon, aluminium, arsenic and lithium: Essentiality and human health implications. The Journal of Nutrition Health and Aging 6/2:154-62.

OECD (2004) SIDS Initial Assessment Profile Silicon dioxide, Silicic acid, aluminum sodium salt, Silicic acid, calcium salt. SIAM 19, 19-22 October 2004.

Cross-reference

Cross-reference 1

Reason / purpose for cross-reference:

data waiving: supporting information

Reference

Endpoint:

monitoring data

Type of information:

other: report

Adequacy of study:

key study

Reliability:

1 (reliable without restriction)

Rationale for reliability incl. deficiencies:

data from handbook or collection of data

Qualifier:

no guideline required

Principles of method if other than guideline:

Evaluation and summary of high quality environmental geochemical data for Europe, which is provided by the Forum of European Geological Surveys (FOREGS) and the European Geochemical Mapping of Agricultural and Grazing Land Soil (GEMAS), with respect to silicon concentrations in stream water, stream sediment and topsoil, as well as in agricultural soil and grazing land.

GLP compliance:

no

Type of measurement:

other: Geochemical background and ambient silicon concentrations in different environmental compartments across Europe

Media:

other: Natural stream water, stream sediment and topsoil, as well as agricultural and grazing land soils

Details on sampling:

FOREGS and GEMAS data for EU-27 countries plus UK and Norway were considered, i.e. data from non-EEA countries such as Albania, Bosnia and Switzerland were excluded from further analysis.

FOREGS:
- The FOREGS sampling grid was based on GTN grid cells developed for Global Geochemical Baseline mapping. This grid divides the entire land surface into 160 km x 160 km cells covering an area of 4,500,000 km2.
- Sampling methodology, preparation and analysis are described by Salminen et al. (2005).
- FOREGS data for EU-27 countries plus UK and Norway were considered, i.e. data from non-EEA countries such as Albania and Switzerland were excluded from further analysis.
- A total of 795 stream water samples of silicon dioxide and 839 sediment samples of silicon dioxide were processed in the FOREGS-program, including 743 paired samples, i.e. samples with the same coordinates for the sampling location of stream water and sediment.
- The FOREGS dataset reports silicon/silicon dioxide concentrations for 833 topsoil samples sampled on a grid across Europe. A topsoil sample was taken at each site from 0-25 cm (excluding material from the organic layer where present).
- Reported silicon dioxide concentrations were converted into silicon concentrations.
- High quality and consistency of the obtained data were ensured by using standardised sampling methods and by treating and analysing all samples in the same laboratory of each country.

GEMAS:
- Samples from 33 out of 38 European countries were analysed to develop a suitable harmonised geochemical data base for soils. The sampling started in the spring 2008 and the first four months of 2009.
- The whole GEMAS project area of 5,600,000 km2 was divided into a grid with 50 km x 50 km cells.
- To generate harmonised data sets, all project samples were processed by a central sample preparation facility in Slovakia.
- GEMAS data for EU-27 countries plus UK and Norway were considered, i.e. data from non-EEA countries such as Bosnia and Switzerland were excluded from further analysis.
- The GEMAS dataset reports silicon concentrations of 1,867 samples from the regularly ploughed layer (Ap-horizon) of agricultural land (arable land; 0 - 20 cm) and of 1,781 samples from the top layer of grazing land (soil under permanent grass cover; 0 - 10 cm) sampled on a grid across Europe.

FOREGS DATABASE STREAM WATER/SEDIMENT:

- Sampled stream water and sediments cover a wide range of environmental conditions. Water parameters such as pH, hardness and organic carbon concentrations extend over several magnitudes. Silicon water levels range from 0.05 to 33.7 mg/L with 5^th, 50^th and 95^th percentiles of 0.8, 3.7 and 8.7 mg/L, respectively.

- In the sediment, silicon concentrations range from 11,691.1 to 439,586.6 mg/kg with 5^th, 50^th and 95^th percentiles of 161,384.4, 290,875.4 and 385,246.2 mg/kg, respectively (Table 1).

- Taking into account the high quality and representativeness of the data set, the 95^th percentile of 8.7 mg/L can be regarded as representative background concentration for dissolved silicon in European surface waters and the 95^th percentile of 385,246.2 mg/kg as representative background concentration of silicon in European stream sediments.

- Regarding the partitioning of silicon in the water column, stream water/sediment partition coefficients range from 7,056 to 5,290,000 L/kg. Since FOREGS sampled on a grid aiming to equally represent geochemical baseline concentrations across Europe, a European median log K_p value of 4.87 is derived.

Table 1: Water parameters and silicon/silicon dioxide concentrations of stream sediment and stream water and respective partitioning.

	Parameter	#	Unit	Min.	Max.	5^thP	50^thP
water	pH ^1	735 ^2	-	9.80	4.50	8.50	7.70
water	Ca	743	mg/L	0.23	592.00	1.64	42.70
water	Cl	743	mg/L	0.14	4,560.00	0.49	9.32
water	HCO3	741 ^3	mg/L	0.69	1,804.42	5.36	131.67
water	K	743	mg/L	< 0.01	182.00	0.15	1.64
water	Mg	743	mg/L	0.05	230.00	0.46	6.22
water	Na	743	mg/L	0.23	4,030.00	1.00	6.76
water	NO3	743	mg/L	< 0.04	107.00	< 0.04	3.10
water	DOC	741 ^4	mg/L	< 0.50	57.94	0.60	4.79
water	SO4^2-	743	mg/L	< 0.30	2,420.00	1.18	17.10
water	SiO2	743	mg/L	0.10	72.00	1.66	8.00
water	Si ^5	743	mg/L	0.05	33.67	0.78	3.74
sediment	SiO2	743	%	2.50	94.00	34.51	62.20
sediment	Si ^5	743	mg/kg	11,691.13	439,586.58	161,384.39	290,875.37
Partitioning (Kp)	Si (sed/water)	743	L/kg	7,056	5,290,000	26,762	73,789
Log Kp	Si (sed/water)	743	-	3.85	6.72	4.43	4.87

¹ Statistics are based on H⁺ concentrations rather than pH.

² Removal of 2 outliers < pH 4.3 and 6 negative values.

³ Removal of 2 outliers < 0.01.

⁴ Removal of 1 outlier > 70 mg/L and 1 negative values.

⁵ Values converted from SiO₂.

FOREGS DATABASE Background soil concentrations

- Sampled soils cover a wide range of environmental conditions. Soil parameters, including pH and TOC, cover several magnitudes.

- Baseline silicon levels in topsoil range from 6,874.4 to 452,306.5 mg/kg with 5^th, 50^th and 95^th percentiles of 159,055.5, 317,531.2 and 415,680.6 mg/kg, respectively (see Table 2)

- Taking into account the high quality and representativeness of the data set, the 95^th percentile of 415,680.6 mg/kg can be regarded as representative background concentration of silicon in topsoil of EU countries.

Table 2: Concentrations of silicon/silicon dioxide in topsoil samples.

Parameter	Unit	#	Min.	Max.	5^thP	50^thP	95^thP
pH ^1	-	802	7.55	3.38	7.31	5.49	4.28
TOC	%	799	0.07	46.61	0.56	1.72	5.86
SiO2	%	833	1.47	96.72	34.01	67.90	88.89
Si ^2	mg/kg	833	6,874.4	452,306.5	159,055.5	317,531.2	415,680.6

¹ Statistics are based on H⁺ concentrations rather than pH.

² Values converted from SiO₂.

GEMAS DATABASE AGRICULTURAL AND GRAZING LAND SOIL CONCENTRATIONS:

- Silicon levels of agricultural soil range from 11,499.0 to 448,274.0 mg/kg with 5^th, 50^th and 95^th percentiles of 156,447.5, 311,829.0 and 417,708.0 mg/kg, respectively (see Table 3). In grazing land, soil concentrations of silicon range from 7,058.0 to 450,293.0 mg/kg with 5^th, 50^th and 95^th percentiles of 133,489.0, 299,650.0 and 409,396.0 mg/kg, respectively (see Table 4)

Table 3: Agricultural soil concentrations.

Parameter	Unit	Method	#	Min.	Max.	5^th P	50^th P	95^th P
CEC	meq/100g	AAS	1,867	1.80	48.30	6.10	15.80	33.30
pH (CaCl2)	pH	pH-meter	1,867	3.32	7.98	4.14	5.71	7.45
TOC	%	IR	1,854	0.40	46.00	0.70	1.70	5.67
Silicon	mg/kg	XRF	1,867	11,499.00	448,274.00	156,447.50	311,829.00	417,708.00

Table 4: Grazing land soil concentrations.

Parameter	Unit	Method	#	Min.	Max.	5^thP	50^thP	95^thP
CEC	meq/100g	AAS	1,781	2.54	49.88	8.27	17.96	37.74
pH (CaCl2)	pH	pH-meter	1,780	3.26	8.06	4.03	5.38	7.45
TOC	%	IR	1,780	0.41	49.00	0.94	2.80	11.05
Silicon	mg/kg	XRF	1,781	7,058.00	450,293.00	133,489.00	299,650.00	409,396.00

Conclusions:

Representative background or ambient concentrations of silicon/silicion dioxide in environmental compartments are tabulated below.

compartment, unit, concentration (50th P), concentration (95th P)
background stream water, mg/L SiO2, 8.0, 18.6
background stream water, mg/L Si, 3.7*, 8.7*
background stream water sediment, % SiO2, 62.2, 82.4
background stream water sediment, mg/kg Si, 290,875.4*, 385,246.2*
background topsoil, % SiO2, 67.9, 88.9
background topsoil, mg/kg Si, 317,531.2*, 415,680.6*
agricultural soil, mg/kg Si, 311,829.0, 417,708.0
grazing land soil, mg/kg Si, 299,650.0, 409,396.0
* based on measured SiO2.

Based on the FOREGS dataset, the 95th percentile of 8.7 mg/L can be regarded as representative background concentration of dissolved silicon in European surface waters and the 95th percentile of 385,246.2 mg/kg as representative background concentration of European stream sediments. Regarding the respective partitioning between sediment and water, a European median log Kp value of 4.87 is derived.

Based on the FOREGS dataset, the 95th percentile of 415,680.6 mg/kg can be regarded as representative background concentration of silicon in topsoil of EU countries. Representative silicon concentrations (95th percentile) of agricultural and grazing land soil (i.e. ambient levels) amount to 417,708.0 and 409,396.0 mg/kg, respectively, according to the GEMAS dataset.

Cross-reference 2

Reason / purpose for cross-reference:

data waiving: supporting information

Reference

Endpoint:

monitoring data

Type of information:

other: report

Adequacy of study:

key study

Reliability:

1 (reliable without restriction)

Rationale for reliability incl. deficiencies:

data from handbook or collection of data

Qualifier:

no guideline required

Principles of method if other than guideline:

Evaluation and summary of high quality environmental geochemical data for Europe, which is provided by the Forum of European Geological Surveys (FOREGS) and the European Geochemical Mapping of Agricultural and Grazing Land Soil (GEMAS), with respect to iron concentrations in stream water, stream sediment and topsoil, as well as in agricultural soil and grazing land.

GLP compliance:

no

Type of measurement:

other: Geochemical background and ambient iron concentrations in different environmental compartments across Europe

Media:

other: Natural stream water, stream sediment and topsoil, as well as agricultural and grazing land soils

Details on sampling:

FOREGS and GEMAS data for EU-27 countries plus UK and Norway were considered, i.e. data from non-EEA countries such as Albania, Bosnia and Switzerland were excluded from further analysis.

FOREGS:
- The FOREGS sampling grid was based on GTN grid cells developed for Global Geochemical Baseline mapping. This grid divides the entire land surface into 160 km x 160 km cells covering an area of 4,500,000 km2.
- Sampling methodology, preparation and analysis are described by Salminen et al. (2005).
- FOREGS data for EU-27 countries plus UK and Norway were considered, i.e. data from non-EEA countries such as Albania and Switzerland were excluded from further analysis.
- A total of 795 stream water samples and 832 sediment samples were processed in the FOREGS-program, including 737 paired samples, i.e. samples with the same coordinates for the sampling location of stream water and sediment.
- The FOREGS dataset reports iron concentrations for 825 topsoil samples. A topsoil sample was taken at each site from 0-25 cm (excluding material from the organic layer where present).
- High quality and consistency of the obtained data were ensured by using standardised sampling methods and by treating and analysing all samples in the same laboratory of each country.

GEMAS:
- Samples from 33 out of 38 European countries were analysed to develop a suitable harmonised geochemical data base for soils. The sampling started in the spring 2008 and the first four months of 2009.
- The whole GEMAS project area of 5,600,000 km2 was divided into a grid with 50 km x 50 km cells.
- To generate harmonised data sets, all project samples were processed by a central sample preparation facility in Slovakia.
- GEMAS data for EU-27 countries plus UK and Norway were considered, i.e. data from non-EEA countries such as Bosnia and Switzerland were excluded from further analysis.
- The GEMAS dataset reports iron concentrations of 1,867 samples from the regularly ploughed layer (Ap-horizon) of agricultural land (arable land; 0-20 cm) and of 1,781 samples from the top layer of grazing land (soil under permanent grass cover; 0-10 cm) sampled on a grid across Europe.

FOREGS DATABASE STREAM WATER/SEDIMENT:

- Sampled stream water and sediments cover a wide range of environmental conditions. Water parameters such as pH, hardness and organic carbon concentrations extend over several magnitudes. Iron water levels range from < 1.0 (< LOQ) to 4,820.0 µg/L with 5^th, 50^th and 95^th percentiles of 4.1, 65.4 and 1,120.0 µg/L, respectively.

- In the sediment, iron concentrations range from 0.6 to192.4 g/kg with 5^th, 50^th and 95^th percentiles of 5.8, 20.1 and 45.3 g/kg, respectively (Table 1).

- Taking into account the high quality and representativeness of the data set, the 95^th percentile of 1,120.0 µg/L can be regarded as representative background concentration for dissolved iron in European surface waters and the 95^th percentile of 45.3 g/kg as representative background concentration of iron in European stream sediments.

- Regarding the partitioning of iron in the water column, stream water/sediment partition coefficients range from 1,413 to 294,000,000 L/kg. Since FOREGS sampled on a grid aiming to equally represent geochemical baseline concentrations across Europe, a European median log K_p value of 5.48 is derived.

Table 1: Water parameters and iron concentrations of stream sediment and stream water and respective partitioning.

	Parameter	#	Unit	Min.	Max.	5^thP	50^thP	95^thP
water	pH ^1	729 ^2	-	9.80	4.50	8.50	7.70	6.10
water	Ca	737	mg/L	0.23	592.00	1.61	42.44	143.52
water	Cl	737	mg/L	0.14	4,560.00	0.49	8.97	68.37
water	HCO3	735 ^3	mg/L	0.69	1,804.42	5.35	128.02	371.20
water	K	737	mg/L	< 0.01	182.00	0.14	1.62	9.81
water	Mg	737	mg/L	0.05	230.00	0.46	6.15	38.11
water	Na	737	mg/L	0.23	4,030.00	1.00	6.66	48.28
water	NO3	737	mg/L	< 0.04	107.00	< 0.04	3.09	39.91
water	DOC	735 ^4	mg/L	< 0.50	57.94	0.60	4.80	23.10
water	SO4^2-	737	mg/L	< 0.30	2,420.00	1.18	16.80	164.01
water	Fe	737	µg/L	< 1.00	4,820.00	4.10	65.40	1,120.00
sediment	Fe	737	g/kg	0.60	192.40	5.78	20.10	45.28
Partitioning (Kp)	Fe (sed/water)	737	L/kg	1,413	294,000,000	14,317	303,282	6,805,391
Log Kp	Fe (sed/water)	737	-	3.15	8.47	4.16	5.48	6.83

¹ Statistics are based on H⁺ concentrations rather than pH

² Removal of 2 outliers < pH 4.3 and 6 negative values.

³ Removal of 2 outliers < 0.01.

⁴ Removal of 1 outlier > 70 mg/L and 1 negative values.

FOREGS DATABASE Background soil concentrations

- Sampled soils cover a wide range of environmental conditions. Soil parameters, including pH and TOC, cover several magnitudes.

- Baseline iron levels in topsoil range from 0.7 to 152.4 g/kg with5^th, 50^th and 95^th percentiles of 3.5, 19.6 and 44.3 g/kg, respectively (see Table 2).

- Taking into account the high quality and representativeness of the data set,the 95^th percentile of 44.3 g/kg can be regarded as representative background concentration of iron in topsoil of EU countries.

Table 2: Concentrations of iron in topsoil samples.

Parameter	Unit	#	Min.	Max.	5^thP	50^thP	95^thP
pH ^1	-	798	7.55	3.38	7.31	5.49	4.29
TOC	%	805	0.07	46.61	0.55	1.72	5.87
Fe	g/kg	825	0.70	152.40	3.52	19.60	44.30

¹ Statistics are based on H⁺ concentrations rather than pH.

GEMAS DATABASE AGRICULTURAL AND GRAZING LAND SOIL CONCENTRATIONS:

- Iron levels of agricultural soil range from 404.5 to 133,926.1 mg/kg with 5^th, 50^th and 95^th percentiles of 3,223.4, 17,165.1 and 36,998.8 mg/kg, respectively (see Table 3). In grazing land, soil concentrations of iron range from 510.3 to 94,759.1 mg/kg with 5^th, 50^th and 95^th percentiles of 3,448.0, 16,949.0 and 38,345.0 mg/kg, respectively (see Table 4).

Table 3: Agricultural soil concentrations.

Parameter	Unit	Method	#	Min.	Max.	5^thP	50^thP	95^thP
CEC	meq/100g	AAS	1,867	1.80	48.30	6.10	15.80	33.30
pH (CaCl2)	pH	pH-meter	1,867	3.32	7.98	4.14	5.71	7.45
TOC	%	IR	1,854	0.40	46.00	0.70	1.70	5.67
Iron	mg/kg	AR	1,867	404.50	133,926.10	3,223.35	17,165.05	36,998.83
Iron	mg/kg	XRF	1,867	909.00	154,715.00	5,595.00	24,620.00	48,499.00
Iron	mg/kg	MMI	1,867	0.50	2,200.00	7.00	35.00	249.00

Table 4: Grazing land soil concentrations.

Parameter	Unit	Method	#	Min.	Max.	5^thP	50^thP	95^thP
CEC	meq/100g	AAS	1,781	2.54	49.88	8.27	17.96	37.74
pH (CaCl2)	pH	pH-meter	1,780	3.26	8.06	4.03	5.38	7.45
TOC	%	IR	1,780	0.41	49.00	0.94	2.80	11.05
Iron	mg/kg	AR	1,781	510.31	94,759.13	3,447.99	16,948.95	38,344.67
Iron	mg/kg	XRF	1,781	979.00	114,422.00	5,665.00	24,129.00	49,727.00

Conclusions:

Representative background or ambient concentrations of iron in environmental compartments are tabulated below:

compartment, unit, concentration (50th P), concentration (95th P)
background stream water, µg/L Fe, 65.4, 1,120.0
background stream water sediment, g/kg Fe, 20.1, 45.3
background topsoil, g/kg Fe, 19.6, 44.3
agricultural soil, mg/kg Fe, 17,165.1, 36,998.8
grazing land soil, mg/kg Fe, 16,949.0, 38,344.7

Based on the FOREGS dataset, the 95th percentile of 1,120.0 µg/L can be regarded as representative background concentration for dissolved iron in European surface waters and the 95th percentile of 45.3 g/kg as representative background concentration of European stream sediments. Regarding the respective partitioning between sediment and water, a European median log Kp value of 5.48 is derived.

Based on the FOREGS dataset, the 95th percentile of 44.3 g/kg can be regarded as representative background concentration of iron in topsoil of EU countries. Representative iron concentrations (95th percentile) of agricultural and grazing land soil (i.e. ambient levels) amount to 36,998.8 and 38,344.7 mg/kg, respectively, according to the GEMAS dataset.