Council Regulation (EC) No 440/2008 of 30 May 2008 laying down test methods pursuant to Regulation (EC) No 1907/2006 of the European Parliament and of the Council on the Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH) (Text with EEA relevance)

• 35.If the pressure reading is negative, do not open the valve. Moisture sometimes accumulates in the syringe needle and tubing, indicated by a small negative pressure reading. In this case remove the needle, shake the tubing, dry with a tissue and fit a new needle.

pH measurement

• 36.Measure and record the pH of the contents of each bottle after the final pressure measurement.

DATA AND REPORTING

Expression of results

• 37.Calculate the sum and average of the pressures recorded at each time interval for each set of replicate bottles and calculate the mean cumulative gross gas pressure at each time interval for each set of replicates. Plot curves of mean cumulative gas production (Pa) against time for control, test and reference bottles. Select a time on the linear part of the curve, usually 48 hours, and calculate the percentage inhibition (I) for each concentration from equation [1]:

I = (1 – Pt/PC) × 100	[1],

• where

I	=	percentage inhibition,in %;
---	---	---
Pt	=	the gas pressure produced with test material at selected time, in Pascal (Pa);
---	---	---
Pc	=	the gas pressure produced in the control at the same time, in Pascal (Pa).

• It would be advisable to draw both plots, i.e. Plot I against concentration and also against logarithm of the concentration so that the curve which is nearer to linearity may be selected. Assess the EC50 (mg/l) value visually or by regression analysis from that curve nearer to linearity. For comparative purposes it may be more useful to express the concentration of the chemical as mg chemical/g of total dry solids. To obtain this concentration, divide the volumetric concentration (mg/l) by the volumetric concentration of dry sludge solids (g/l) (paragraph 16).

• 38.Calculate either the percentage inhibition achieved by the single concentration of the reference chemical used or the EC50 if a sufficient number of concentrations have been investigated.

• 39.Convert the mean pressure of the gas produced in the control Pc(Pa) to the volume by reference to the pressure-meter calibration curve (Appendix 2) and from this calculate the yield of gas, expressed as the volume produced in 48 hours from 100 ml undiluted sludge at a solids concentration of 2 % (20 g/l) to 4 % (40 g/l).

Validity criteria

• 40.Results from the ISO inter-laboratory trial (5) showed the reference chemical (3,5-dichlorophenol) caused 50 % inhibition of gas production in a range of concentrations of 32 mg/l to 510 mg/l mean 153 mg/l (paragraph 10). This range is so wide that firm limits for inhibition cannot confidentially be set as validity criteria; this should be possible when developments have shown how to produce more consistent inocula. The volumes of gas produced in control bottles in 48 hour ranged from 21 ml/g sludge dry matter to 149 ml/g (mean 72 ml/g). There was no obvious relation between volume of gas produced and the corresponding EC50 value. The final pH varied between 6,1 and 7,5.

• 41.The test is considered to be valid when an inhibition of greater than 20 % is obtained in the reference control containing 150 mg/l of 3,5-dichlorophenol, more than 50 ml of gas per g of dry matter is produced in the blank control and the pH value is within the range of 6,2 to 7,5 at the end of the test.

Test Report

• 42.The test report must include the following information:

Test chemical

• — common name, chemical name, CAS number, structural formula and relevant physico-chemical properties;

• — purity (impurities) of test chemical.

Test conditions

• — volumes of liquid contents and of headspace in test vessels;

• — descriptions of the test vessels and gas measurement (e.g. type of pressure-meter);

• — application of test chemical and reference chemical to the test system, test concentrations used and use of any solvents;

• — details of the inoculum used: name of sewage treatment plant, description of the source of waste water treated (e.g. operating temperature, sludge retention time, predominantly domestic sewage or industrial waste, etc.), concentration of solids, gas production activity of anaerobic digester, previous exposure or possible pre-adaptation to toxic chemicals or site of collection of mud, sediment etc;

• — incubation temperature and range;

• — number of replicates.

Results

• — pH values at end of test;

• — all the measured data collected in the test, blank and reference chemical control vessels, as appropriate (e.g. pressure in Pa or millibars) in tabular form;

• — percentage inhibition in test and reference bottles, and the inhibition-concentration curves;

• — calculation of EC50 values, expressed as mg/l and mg/g;

• — gas production per g sludge in 48 hours;

• — reasons for any rejection of the test results;

• — discussion of results, including any deviations from the procedures in this test method and discuss any deviations in the test results due to interferences and errors from what would be expected;

• — address also whether the purpose of the test was to measure the toxicity to either pre-exposed or non pre-exposed microorganisms.

LITERATURE

• (1)Chapter C.11 of this Annex: Activated Sludge, Respiration Inhibition Test.

• (2)Chapter C.43 of this Annex: Anaerobic biodegradability of organic compounds in digested sludge: method by measurement of gas production.

• (3)International Organisation for Standardisation (2003) ISO 13 641-1 Water Quality — Determination of inhibition of gas production of anaerobic bacteria — Part 1: General Test.

• (4)International Organisation for Standardisation (2003) ISO 13 641-2 Water Quality — Determination of inhibition of gas production of anaerobic bacteria — Part 2: Test for low biomass concentrations.

• (5)ISO (2000) Ring test of ISO 13 641-1 and ISO 13 641-2. Determination of inhibition of activity of anaerobic bacteria. BL 6958/A. Evans MR, Painter HA. Brixham Environmental Laboratory, AstraZeneca UK Ltd., Brixham, TQ5 8BA UK.

• (6)Swanwick JD, Foulkes M (1971). Inhibition of anaerobic digestion of sewage sludge by chlorinated hydrocarbons. Wat. Pollut. Control, 70, 58-70.

• (7)HMSO (1986) Determination of the inhibitory effects of chemicals and waste waters on the anaerobic digestion of sewage sludge. ISBN 0 117519 43 X, In: Methods for the Examination of Waters and Associated Materials UK.

• (8)Shelton DR, Tiedje JM (1984). General method for determining anaerobic biodegradation potential. Appl. Env. Microbiol. 47 850-857.

• (9)Battersby NS and Wilson V (1988). Evaluation of a serum bottle technique for assessing the anaerobic biodegradability of organic compounds under methanogenic conditions. Chemosphere 17, 2441-2460.

• (10)Wilson V, Painter HA and Battersby NS (1992). A screening method for assessing the inhibition of the anaerobic gas production from sewage sludge. Proc. Int. Symp. on Ecotoxicology. Ecotoxicological Relevance of Test Methods, GSF Forschungzentrum, Neuherberg, Germany (1990). Eds. Steinberg C and Kettrup A, pp117-132 (1992).

• (11)Kawahara K, Yakabe Y, Chida T, and Kida K (1999). Evaluation of laboratory-made sludge for an anaerobic biodegradability test and its use for assessment of 13 chemicals. Chemosphere, 39 (12), 2007-2018.

• (12)International Organization for Standardization (1995) ISO 10 634 Water Quality — Guidance for the preparation and treatment of poorly water-soluble organic compounds for the subsequent evaluation of their biodegradability in an aqueous medium.

• (13)International Organization for Standardization (1997) ISO 11 923 Water Quality — Determination of suspended solids by filtration through glass-fibre filters.

Appendix 1

Example of an apparatus to measure biogas production by gas pressure

Image omitted; see official source

Key:

1	—	Pressure-meter
---	---	---
2	—	3-way gas-tight valve
---	---	---
3	—	Syringe needle
---	---	---
4	—	Gastight seal (crimp cap and septum)
---	---	---
5	—	Head space
---	---	---
6	—	Digested sludge inoculum

Test vessels in an environment of 35 °C ± 2 °C

Appendix 2

Conversion of the pressure-meter

The pressure-meter readings may be related to gas volumes by means of a standard curve and from this the volume of gas produced per g dry sludge per 48 hours may be calculated. This activity index is used as one of the criteria by which to assess the validity of test results. The calibration curve is produced by injecting known volumes of gas at 35 °C ± 2 °C in serum bottles containing a volume of water equal to that of the reaction mixture, VR;

• — Dispense VR ml aliquots of water, kept at 35 °C ± 2 °C into five serum bottles. Seal the bottles and place in a water bath at 35 °C ± 2 °C for 1 hour to equilibrate;

• — Switch on the pressure-meter, allow to stabilise, and adjust to zero;

• — Insert the syringe needle through the seal of one of the bottles, open the valve until the pressure-meter reads zero and close the valve;

• — Repeat the procedure with the remaining bottles;

• — Inject 1 ml of air at 35 °C ± 2 °C into each bottle. Insert the needle (on the meter) through the seal of one of the bottles and allow the pressure reading to stabilise. Record the pressure, open the valve until the pressure reads zero and then close the valve;

• — Repeat the procedure with the remaining bottles;

• — Repeat the total procedure using 2 ml, 3 ml, 4 ml, 5 ml, 6 ml, 8 ml, 10 ml, 12 ml, 16 ml, 20 ml, and 50 ml of air;

• — Plot a conversion curve of pressure (Pa) against gas volume injected (ml). The response of the instrument is linear over the range 0 Pa to 70 000 Pa, and 0 ml to 50 ml of gas production.

Appendix 3

Identified factors which can lead to false results

(a) Quality of the bottle-caps

Different types of septa for the serum bottles are available commercially; many of them, including butyl rubber, lose tightness when pierced with a needle under the conditions of this test. Sometimes the pressure falls very slowly once the septum has been pierced with the syringe needle. The use of gas-tight septa is recommended to overcome leaks (paragraph 12(b)).

(b) Moisture in the syringe needle

Moisture sometimes accumulates in the syringe needle and tubing, and is indicated by a small negative pressure reading. To rectify this remove the needle and shake the tubing, dry with a tissue and fit a new needle (paragraphs 12(c) and 35).

(c) Oxygen contamination

Anaerobic methods are subject to error from contamination by oxygen, which can cause lower gas production. In this method this possibility should be minimised by the use of strictly anaerobic techniques, including use of a glove box.

(d) Gross substrates in sludge

The anaerobic gas production and the sensitivity of the sludge are influenced by substrates which are transferred with the inoculum into the test bottles. Digested sludge from domestic anaerobic digesters still often contains recognisable matter like hair and plant residues of cellulose, which tend to make it difficult to take representative samples. By sieving the sludge gross insoluble matter can be removed, which makes representative sampling more likely (paragraph 16).

(e) Volatile test chemicals

Volatile test chemicals will be released into the headspace of the test bottles. This may result in the loss of some of the test material from the system during venting after pressure measurements, yielding falsely high EC50 values. By suitable choice of ratio of headspace volume to liquid volume and by not venting after taking pressure measurements, the error can be reduced (10).

(f) Non-linearity of gas production

If the plot of mean cumulative gas production against incubation time is not approximately linear over the 48h period, the accuracy of the test may be lowered. To overcome this, it may be advisable to use digesting sludge from a different source and/or to add an increased concentration of the test substrate-nutrient broth, yeast extract and glucose (paragraph 29).

Appendix 4

Application to environmental samples of low biomass concentration — anaerobic muds, sediments, etc.

INTRODUCTION

• A.1 In general, the specific microbial activity (volume of gas produced per g dry solids) of naturally occurring anaerobic muds, sediments, soils, etc, is much lower than that of anaerobic sludge derived from sewage. Because of this, when the inhibitory effects of chemicals on these less active samples are to be measured some of the experimental conditions have to be modified. For these less active samples there are two general course of action possible:

• (a) Carry out a modified preliminary test (paragraph 25) with the undiluted sample of mud, soil, etc at 35 °C ± 2 °C or at the temperature at the sample site of collection, for more accurate simulation (as in Part 1 of ISO 13 641);

• (b) Or make the test with a dilute (1 in 100) digester sludge to simulate the low activity expected from the environment sample, but maintain the temperature at 35 °C ± 2 °C (as in Part 2 of ISO 13 641).

• A.2 Option (a) may be achieved by following the method described here (equivalent to Part 1 of ISO 13 641), but it is essential to make a preliminary test (paragraph 25) to ascertain optimal conditions, unless these are already known from previous testing. The mud or sediment sample should be thoroughly mixed, e.g. in a blender, and, if necessary, diluted with a small proportion of de-aerated dilution water (paragraph 14) so that it is sufficiently mobile to be transferred by a coarse-tipped pipette or a measuring cylinder. If it is considered that nutrients may be lacking, the mud sample may be centrifuged (under anaerobic conditions) and re-suspended in the mineral medium containing yeast extract (A.11)

• A.3 Option (b). This reasonably mimics the low activity of environmental samples but lacks the high concentration of suspended solids present in these samples. The role of these solids in inhibition is not known, but it is possible that reaction between the test chemicals and constituents of the mud, as well as adsorption of the test chemicals onto the solids, could result in a lowering of toxicity of the test chemical.

• A.4 Temperature is another important factor: for strict simulation, tests should be made at the temperature of the sample site, since different groups of methane-producing consortia of bacteria are known to operate within different temperature ranges, namely thermophiles (~ 30-35 °C), mesophiles (20-25 °C) and psychrophiles (< 20 °C), which may display different inhibitory patterns.

• A.5 Duration. In the general test, Part 1, using undiluted sludge, the production of gas in the 2-4 days was always sufficient, while in Part 2 with one-hundred diluted sludge insufficient gas, if any, was produced in this period in the ring test. Madsen et al (1996), in describing this latter test, say at least 7 days should be allowed.

Testing with low biomass concentration (Option b)

The following changes and amendments should be made, adding to or replacing some existing paragraphs and sub-paragraphs of the main text.

• A.6 Add to Paragraph 6: Principle of the test;

‘This technique may be used with 1 in 100 diluted anaerobic sludge, partially to simulate the low activity of muds and sediments. The incubation temperature may be either 35 °C or that of the site from which the sample was collected. Since the bacterial activity is much less than in undiluted sludge, the incubation period should be extended to at least 7 days.’

• A.7 Add to paragraph 12 (a):

‘the incubator should be capable of operating down to temperatures of 15 °C.’

• A.8 Add an extra reagent after Paragraph 13:

‘Phosphoric acid (H3PO4), 85 % by mass in water.’

• A.9 Add at end of Paragraph 16:

‘Use a final concentration of 0,20 ± 0,05 g/l of total dry solids in the test.’

• A.10 Paragraph 17. Test substrate

This substrate is not to be used, but is replaced by yeast extract (see paragraphs 17; A.11, A.12, A.13).

• A.11 A mineral medium, including trace elements, for diluting anaerobic sludge, is required and for convenience the organic substrate, yeast extract, is added to this medium.

Add after Paragraph 17

‘(a) Test mineral medium, with yeast extract.

This is prepared from a 10-fold concentrated test medium (paragraph 17 (b); A.12) with a trace element solution (paragraph 17 (c); A.13). Use freshly supplied sodium sulphide nonahydrate (paragraph 17 (b); A.12) or wash and dry it before use, to ensure that it has sufficient reducing capacity. If the test is performed without using a glove box (paragraph 12 (j)), the concentration of sodium sulphide in the stock solution should be increased to 2 g/l (from 1 g/l). Sodium sulphide may also be added from an appropriate stock solution through the septum of the closed test bottles, as this procedure will decrease the risk of oxidation, to obtain a final concentration of 0,2 g/l. Alternatively titanium (III) citrate (paragraph 17 (b)) may be used. Add it through the septum of closed test bottles to obtain a concentration of 0,8 mmol/l to 1,0 mmol/l. Titanium (III) citrate is a highly effective and a low-toxicity reducing agent, which is prepared as follows: Dissolve 2,94 g of trisodium citrate dihydrate in 50 ml of oxygen-free dilution water (paragraph 14) (which results in a 200 mmol/l solution) and add 5 ml of a titanium (III) chloride solution (15 g/100 ml dilution water). Neutralise to pH 7 ± 0,5 with sodium carbonate and dispense to an appropriate serum bottle under a stream of nitrogen gas. The concentration of titanium (III) citrate in this stock solution is 164 mmol/l. Use the test medium immediately or store at 4 °C for no longer than 1 day.

A.12 - (b)Tenfold concentrated test medium, prepared with the following:

anhydrous potassium dihydrogenphosphate (KH2PO4)	2,7 g
Disodium hydrogen phosphate (Na2HPO4)	4,4 g
(or 11,2 g dodecahydrate) ammonium chloride (NH4Cl)	5,3 g
calcium chloride dihydrate (CaCl2·2H2O)	0,75 g
magnesium chloride hexahydrate (MgCl2·6H2O)	1,0 g
iron (II) chloride tetrahydrate (FeCl2·4H2O)	0,2 g
resazurin (redox indicator)	0,01 g
sodium sulphide nonahydrate (Na2S·9H2O)	1,0 g
(or titanium (III) citrate) final concentration	0,8 mmol/l to 1,0 mmol/l
trace element solution (see paragraph 17 (c); A.13)	10,0 ml
yeast extract	100 g
Dissolve in dilution water (paragraph 14) and make up to:	1 000 ml

A.13 - (c)Trace element solution, prepared with the following:

manganese (II) chloride tetrahydrate (MnCl2·4H2O)	0,5 g
ortho-boric acid (H3BO3)	0,05 g
zinc chloride (ZnCl2)	0,05 g
copper (II) chloride (CuCl2)	0,03 g
sodium molybdate dihydrate (Na2MoO4·2H2O)	0,01 g
cobalt (II) chloride hexahydrate (CoCl2·6H2O)	1,0 g
nickel (II) chloride hexahydrate (NiCl2·6H2O)	0,1 g
disodium selenite (Na2SeO3)	0,05 g
Dissolve in dilution water (paragraph 14) and make up to:	1 000 ml’

A.14 Paragraph 25: Preliminary test

It is essential that a preliminary test is made as described in paragraph 24, except that the concentration of sludge solids should be one hundredth of those given, that is 0,1 g/l, 0,2 g/l and 0,4 g/l. The duration of incubation should be at least 7 days.

Note: In the ring test (5) the headspace volume was much too high at 75 % total volume; it should be in the recommended range of 10 %-40 %. The relevant criterion is that the volume of gas produced at around 80 % inhibition should be measurable with acceptable precision (e.g. ± 5 % to ± 10 %).

A.15 Paragraph 26 to 30: Addition of test chemical, inoculum and substrate.

The additions are made in the same way as described in these paragraphs, but the substrate solution (paragraph 17) is replaced by the test medium plus yeast extract substrate (A.11).

Also, the final concentration of dry sludge solids is reduced from 2 g/l - 4 g/l to 0,2 ± 0,05 g/l (A.9). Two examples of the addition of components to the test mixture are given in Table A.1, which replaces the table in paragraph 29.

A.16 Paragraph 33: Incubation of bottles

Because of the expected lower rate of gas production, incubation is carried on for at least 7 days.

A.17 Paragraph 34: Pressure measurements

The same procedure for measuring the pressure in the headspace of the bottles is used as described in paragraph 34 if the amounts in the gaseous phase are required. If total amounts of CO2 plus CH4 are to be measured, the pH of the liquid phase is reduced to about pH 2 by the injection of H3PO4 into each relevant bottle and measuring the pressure after 30 minutes shaking at the temperature of the test. However, more information on the quality of the inoculum may be obtained by measuring the pressure in each bottle before and after acidification. For example when the rate of CO2 production is much higher than that of methane, the sensitivity of the fermentative bacteria may be altered and/or methanogenic bacteria are preferentially affected by the test chemical.

A.18 Paragraph 36: pH measurement

If H3PO4 is to be used some extra bottles, to which no H3PO4 is added, would have to be set up especially for the pH measurement.

REFERENCE:

Madsen, T, Rasmussen, HB; and Nilsson, L (1996), Methods for screening anaerobic biodegradability and toxicity of organic chemicals. Project No.336, Water Quality Institute, Danish Environment Protection Agency, Copenhagen.

Table A.1.

Examples of the test set-up for test batches

Reaction Mixture constituents	Example 1	Example 2	Normal order of addition
Concentration of prepared inoculum (g/l)	0,42	2,1	—
Volume of inoculum added (ml)	45	9	4
Concentration of inoculum in test bottles (g/l)	0,20	0,20	—
Volume of test medium added (ml)	9	9	2
Volume of dilution water added (ml)	36	72	3
Concentration of yeast extract in test bottles (g/l)	9,7	9,7	—
Volume of test chemical stock solution (ml)	3	3	1
Total liquid volume (ml)	93	93	—

Appendix 5

Definitions

For the purpose of this test method the following definitions are used:

Chemical means a substance or a mixture.

Test chemical means any substance or mixture tested using this test method.

C.35.   SEDIMENT-WATER LUMBRICULUS TOXICITY TEST USING SPIKED SEDIMENT

INTRODUCTION

• 1.This test method is equivalent to OECD test guideline (TG) 225 (2007). Sediment-ingesting endobenthic animals are subject to potentially high exposure to sediment bound chemicals and should therefore be given preferential attention, e.g. (1), (2), (3). Among these sediment-ingesters, the aquatic oligochaetes play an important role in the sediments of aquatic systems. By bioturbation of the sediment and by serving as prey these animals can have a strong influence on the bioavailability of such chemicals to other organisms, e.g. benthivorous fish. In contrast to epibenthic organisms, endobenthic aquatic oligochaetes (e.g. Lumbriculus variegatus) burrow in the sediment, and ingest sediment particles below the sediment surface. This ensures exposure of the test organisms to the test chemical via all possible uptake routes (e.g. contact with, and ingestion of contaminated sediment particles, but also via porewater and overlying water).

• 2.This test method is designed to assess the effects of prolonged exposure of the endobenthic oligochaete Lumbriculus variegatus (Müller) to sediment-associated chemicals. It is based on existing sediment toxicity and bioaccumulation test protocols, e.g. (3), (4), (5), (6), (7), (8), (9), (10). The method is described for static test conditions. The exposure scenario used in this test method is spiking of sediment with the test chemical. Using spiked sediment is intended to simulate a sediment contaminated with the test chemical.

• 3.Chemicals that need to be tested towards sediment-dwelling organisms usually persist in this compartment over long time periods. Sediment-dwelling organisms may be exposed via several routes. The relative importance of each exposure route, and the time taken for each to contribute to the overall toxic effects, depends on the physical-chemical properties of the chemical concerned and its ultimate fate in the animal. For strongly adsorbing chemicals (e.g. with log Kow > 5) or for chemicals covalently binding to sediment, ingestion of contaminated food may be a significant exposure route. In order not to underestimate the toxicity of such chemicals, the food necessary for reproduction and growth of the test organisms is added to the sediment before application of the test chemical (11). The test method described is sufficiently detailed so that the test can be carried out whilst allowing for adaptations in the experimental design depending on the conditions in particular laboratories and the varied characteristics of test chemicals.

• 4.The test method is aimed to determine effects of a test chemical on the reproduction and the biomass of the test organisms. The measured biological parameters are the total number of surviving worms and the biomass (dry weight) at the end of the exposure. These data are analysed either by using a regression model in order to estimate the concentration that would cause an effect of x % (e.g. EC50, EC25, and EC10), or by using statistical hypothesis testing to determine the No Observed Effect Concentration (NOEC) and the Lowest Observed Effect Concentration (LOEC).

• 5.Chapter C.27 of this Annex, ‘Sediment-water chironomid toxicity test using spiked sediment’ (6), provided many essential and useful details for the performance of the presented sediment toxicity test method. Hence, this document serves as a basis on which modifications necessary for conducting sediment toxicity tests with Lumbriculus variegatus were worked out. Further documents that are referred to are e.g. the ASTM Standard Guide for Determination of the Bioaccumulation of Sediment-Associated Contaminants by Benthic Invertebrates (3), the U.S. EPA Methods for Measuring the Toxicity and Bioaccumulation of Sediment-Associated Contaminants with Freshwater Invertebrates (7), and the ASTM Standard Guide for Collection, Storage, Characterization, and Manipulation of Sediments for Toxicological Testing and for selection of samplers used to collect benthic invertebrates (12). In addition, practical experience obtained during ring-testing the test method (13), ring-test report), and details from literature are major sources of information for drawing up this document.

PREREQUISITE AND GUIDANCE INFORMATION

• 6.Information on the test chemical such as safety precautions, proper storage conditions and analytical methods should be obtained before beginning the study. Guidance for testing chemicals with physical-chemical properties that make them difficult to perform the test is provided in (14).

• 7.Before carrying out a test, the following information about the test chemical should be known:

• — common name, chemical name (preferably IUPAC name), structural formula, CAS registry number, purity;

• — vapour pressure;

• — solubility in water.

• 8.The following additional information is considered useful before starting the test:

• — octanol-water partition coefficient, Kow;

• — organic carbon-water partitioning coefficient, expressed as Koc;

• — hydrolysis;

• — phototransformation in water;

• — biodegradability;

• — surface tension.

• 9.Information on certain characteristics of the sediment to be used should be acquired before the start of the test (7). For details see paragraphs 22 to 25.

PRINCIPLE OF THE TEST

• 10.Worms of similar physiological state (synchronised as described in Appendix 5) are exposed to a series of toxicant concentrations applied to the sediment phase of a sediment-water system. Artificial sediment and reconstituted water should be used as media. Test vessels without the addition of the test chemical serve as controls. The test chemical is spiked into the sediment in bulk for each concentration level in order to minimise variability between replicates of each concentration level, and the test organisms are subsequently introduced into the test vessels in which the sediment and water concentrations have been equilibrated (see paragraph 29). The test animals are exposed to the sediment-water systems for a period of 28 days. In view of the low nutrient content of the artificial sediment, the sediment should be amended with a food source (see paragraphs 22 to 23, and Appendix 4) to ensure that the worms will grow and reproduce under control conditions. In this way it is ensured that the test animals are exposed through the water and sediment as well as by their food.

• 11.The preferred endpoint of this type of study is the ECx (e.g. EC50, EC25, and EC10; effect concentration, affecting x % of the test organisms) for reproduction and biomass, respectively, compared to the control. It should however be noted, that considering the high uncertainty of low ECx (e.g. EC10, EC25) with extremely high 95 %-confidence limits (e.g. (15)) and the statistical power calculated during hypothesis testing, the EC50 is regarded the most robust endpoint. In addition, the No Observed Effect Concentration (NOEC), and the Lowest Observed Effect Concentration (LOEC) may be calculated for biomass, and reproduction, if the test design and the data support these calculations (see paragraphs 34 to 38). The purpose of the study, ECx or NOEC derivation, will determine the test design.

REFERENCE TESTING

• 12.Performance of the control organisms is expected to demonstrate sufficiently the ability of a laboratory to perform the test, and if historical data are available, the repeatability of the test. In addition, reference toxicity tests may be conducted in regular intervals using a reference toxicant to assess the sensitivity of the test organisms. 96 h reference toxicity tests in water only may satisfactorily demonstrate the sensitivity and condition of the test animals (4)(7). Information on the toxicity of pentachlorophenol (PCP) in complete tests (28 d exposure to spiked sediment) is included in Appendix 6, and in the report on the ring test of the Test Method (13). The acute, water-only toxicity of PCP is described e.g. in (16). This information can be used for comparison of test organism sensitivity in reference tests with PCP as reference toxicant. Potassium chloride (KCl) or copper sulphate (CuSO4) have been recommended as reference toxicants with L. variegatus (4)(7). To date, establishment of quality criteria based on toxicity data for KCl is difficult due to lack of literature data for L. variegatus. Information on the toxicity of copper towards L. variegatus can be found in (17) to (21).

VALIDITY OF THE TEST

• 13.For a test to be valid, the following requirements should be fulfilled:

• — A ring-test (13) has shown that for Lumbriculus variegatus, the average number of living worms per replicate in the controls should have increased by a factor of at least 1,8 at the end of exposure compared to the number of worms per replicate at the start of exposure.

• — The pH of the overlying water should be between 6 and 9 throughout the test.

• — The oxygen concentration in the overlying water should not be below 30 % of air saturation value (ASV) at test temperature during the test.

DESCRIPTION OF THE TEST METHOD

Test system

• 14.Static systems without renewal of the overlying water are recommended. If the sediment-to-water ratio (see paragraph 15) is appropriate, gentle aeration will normally suffice to keep the water quality at acceptable levels for the test organisms (e.g. maximise dissolved oxygen levels, minimise build-up of excretory products). Semi-static or flow-through systems with intermittent or continuous renewal of overlying water should only be used in exceptional cases, since regular renewal of overlying water is expected to affect chemical equilibrium (e.g. losses of test chemical from the test system).

Test vessels and apparatus

• 15.The exposure should be conducted in glass beakers of e.g. 250 ml measuring 6 cm in diameter. Other suitable glass vessels may be used, but they should guarantee a suitable depth of overlying water and sediment. Each vessel should receive a layer of approximately 1,5 – 3 cm of formulated sediment. The ratio of the depth of the sediment layer to the depth of the overlying water should be 1:4. The vessels should be of suitable capacity in compliance with the loading rate, i.e. the number of test worms added per weight unit of sediment, (see also paragraph 39).

• 16.Test vessels and other apparatus that will come into contact with the test chemical should be made entirely of glass or other chemically inert material. Care should be taken to avoid the use of materials, for all parts of the equipment that can dissolve, absorb test chemicals or leach other chemicals and have an adverse effect on the test animals. Polytetrafluoroethylene (PTFE), stainless steel and/or glass should be used for any equipment having contact with the test media. For organic chemicals known to adsorb to glass, silanised glass may be required. In these situations the equipment will have to be discarded after use.

Test species

• 17.The test species used in this type of study is the freshwater oligochaete Lumbriculus variegatus (Müller). This species is tolerant to a wide range of sediment types, and is widely used for sediment toxicity and bioaccumulation testing [e.g. (3), (5), (7), (9), (13), (15), (16), (22), (23), (24), (25), (26), (27), (28), (29), (30), (31), (32), (33), (34), (35)]. The origin of the test animals, the confirmation of species identity (e.g. (36)) as well as the culture conditions should be reported. Identification of species is not required prior to every test if the organisms come from an in-house culture.

Culturing of the test organisms

• 18.In order to have a sufficient number of worms for conducting sediment toxicity tests, it is useful to keep the worms in permanent laboratory culture. Guidance for laboratory culture methods for Lumbriculus variegatus, and sources of starter cultures are given in Appendix 5. For details on culturing this species see references (3), (7), (27).

• 19.To ensure that the tests are performed with animals of the same species, the establishment of single species cultures is strongly recommended. Ensure that the cultures and especially the worms used in the tests are free from observable diseases and abnormalities.

Water

• 20.Reconstituted water according to Chapter C.1 of this Annex (37) is recommended for use as overlying water in the tests; it can also be used for the laboratory cultures of the worms (see Appendix 2 for preparation). If required, natural water may be used. The chosen water must be of a quality that will allow the growth and reproduction of the test species for the duration of the acclimation and test periods without showing any abnormal appearance or behaviour. Lumbriculus variegatus has been demonstrated to survive, grow, and reproduce in this type of water (30), and maximum standardisation of test and culture conditions is provided. If a reconstituted water is used, its composition should be reported, and the water should be characterised prior to use at least by pH, oxygen content, and hardness (expressed as mg CaCO3/l). Analysis of the water for micropollutants prior to use might provide useful information (see, e.g., Appendix 3).

• 21.The pH of the overlying water should be in the range of 6,0 to 9,0 (see paragraph 13). If increased ammonia development is expected, it is considered useful to keep the pH between 6,0 and 8,0. For testing of e.g. weak organic acids, it is advisable to adjust the pH by buffering the water to be used in the test, as described e.g. by (16). The total hardness of the water to be used in the test should be between 90 and 300 mg CaCO3 per liter for natural water. Appendix 3 summarises additional criteria for acceptable dilution water according to OECD Guideline No. 210 (38).

Sediment

• 22.Since uncontaminated natural sediments from a particular source may not be available throughout the year, and indigenous organisms as well as the presence of micropollutants can influence the test, a formulated sediment (also called reconstituted, artificial or synthetic sediment) should preferably be used. Use of a formulated sediment minimises variability of test conditions as well as introduction of indigenous fauna. The following formulated sediment is based on the artificial sediment according to (6), (39) and (40). It is recommended for use in this type of test ((6), (10), (30), (41), (42), (43)):

• (a) 4-5 % (dry weight) sphagnum peat; it is important to use peat in powder form, degree of decomposition: ‘medium’, finely ground (particle size ≤ 0,5 mm), and only air-dried.

• (b) 20 ± 1 % (dry weight) kaolin clay (kaolinite content preferably above 30 %).

• (c) 75-76 % (dry weight) quartz sand (fine sand, grain size: ≤ 2 mm, but > 50 % of the particles should be in the range of 50-200 μm).

• (d) Deionised water, 30–50 % of sediment dry weight, in addition to the dry sediment components.

• (e) Calcium carbonate of chemically pure quality (CaCO3) is added to adjust the pH of the final mixture of the sediment.

• (f) The total organic carbon content (TOC) of the final mixture should be 2 % (± 0,5 %) of sediment dry weight and should be adjusted by the use of appropriate amounts of peat and sand, according to (a) and (c).

• (g) Food, e.g. powdered leaves of Stinging Nettle (Urtica sp., in accordance with pharmacy standards, for human consumption), or a mixture of powdered leaves of Urtica sp. with alpha-cellulose (1:1), at 0,4 - 0,5 % of sediment d.w., in addition to the dry sediment components; for details see Appendix 4.

• 23.The source of peat, kaolin clay, food material, and sand should be known. In addition to item g), Chapter C.27 of this Annex (6) lists alternative plant materials to be used as a source of nutrition: dehydrated leaves of mulberry (Morus alba), white clover (Trifolium repens), spinach (Spinacia oleracea), or cereal grass.

• 24.The chosen food source should be added prior to or during spiking the sediment with the test chemical. The chosen food source should allow for at least acceptable reproduction in the controls. Analysis of the artificial sediment or its constituents for micro-pollutants prior to use might provide useful information. An example for the preparation of the formulated sediment is described in Appendix 4. Mixing of dry constituents is also acceptable if it is demonstrated that after addition of overlying water a separation of sediment constituents (e.g. floating of peat particles) does not occur, and that the peat or the sediment is sufficiently conditioned (see also paragraph 25 and Appendix 4). The artificial sediment should be characterised at least by origin of the constituents, grain size distribution (percent sand, silt, and clay), total organic carbon content (TOC), water content, and pH. Measurement of redox potential is optional.

• 25.If required, e.g. for specific testing purposes, natural sediments from unpolluted sites may also serve as test and/or culture sediment (3). However, if natural sediment is used, it should be characterised at least by origin (collection site), pH and ammonia of the pore water, total organic carbon content (TOC) and nitrogen content, particle size distribution (percent sand, silt, and clay), and percent water content (7), and it should be free from any contamination and other organisms that might compete with, or prey on the test organisms. Measurement of redox potential and cation exchange capacity is optional. It is also recommended that, before it is spiked with the test chemical, the natural sediment be conditioned for seven days under the same conditions which prevail in the subsequent test. At the end of this conditioning period, the overlying water should be removed and discarded.

• 26.The sediment to be used must be of a quality that will allow the survival and reproduction of the control organisms for the duration of the exposure period without showing any abnormal appearance or behaviour. The control worms should burrow in the sediment, and they should ingest the sediment. Reproduction in the controls should at least be according to the validity criterion as described in paragraph 13. The presence or absence of fecal pellets on the sediment surface, which indicate sediment ingestion by the worms, should be recorded and can be helpful for the interpretation of the test results with respect to exposure pathways. Additional information on sediment ingestion can be obtained by using methods described in (24), (25), (44), and (45), which specify sediment ingestion or particle selection in the test organisms.

• 27.Manipulation procedures for natural sediments prior to use in the laboratory are described in (3), (7), and (12). The preparation and storage of the artificial sediment recommended to be used in the Lumbriculus test is described in Appendix 4.

Application of the test chemical

• 28.The test chemical is to be spiked to the sediment. As most test chemicals are expected to have low water solubility, they should be dissolved in a suitable organic solvent (e.g. acetone, n-hexane, cyclohexane) at a volume as small as possible in order to prepare the stock solution. The stock solution should be diluted with the same solvent to prepare the test solutions. Toxicity and volatility of the solvent, and the solubility of the test chemical in the chosen solvent should be the main criteria for the selection of a suitable solubilising agent. For each concentration level the same volume of the corresponding solution should be used. The sediment should be spiked in bulk for each concentration level in order to minimise between-replicate variability of the test chemical concentration. Each of the test solutions is then mixed with quartz sand as described in paragraph 22 (e.g. 10 g of quartz sand per test vessel). In order to soak the quartz sand completely, a volume of 0,20 - 0,25 ml per g of sand has been found sufficient. Thereafter, the solvent must be evaporated to dryness. In order to minimise losses of the test chemical through co-evaporation (e.g. depending on the chemical's vapour pressure), the coated sand should be used immediately after drying. The dry sand is mixed with the suitable amount of formulated sediment of the corresponding concentration level. The amount of sand provided by the test-chemical-and-sand mixture has to be taken into account when preparing the sediment (i.e. the sediment should thus be prepared with less sand). The major advantage of this procedure is that virtually no solvent is introduced to the sediment (7). Alternatively, e.g. for field sediment, the test chemical may be added by spiking a dried and finely ground portion of the sediment as described above for the quartz sand, or by stirring the test chemical into the wet sediment, with subsequent evaporating of any solubilising agent used. Care should be taken to ensure that the test chemical added to sediment is thoroughly and evenly distributed within the sediment. If necessary, subsamples may be analysed to confirm the target concentrations in the sediment, and to determine degree of homogeneity. It may also be useful to analyse subsamples of the test solutions to confirm the target concentrations in the sediment. Since a solvent is used for coating the test chemical on the quartz sand, a solvent control should be employed which is prepared with the same amount of the solvent as the test sediments. The method used for spiking, and the reasons for choosing a specific spiking procedure other than described above should be reported. The method of spiking may be adapted to the test chemical's physical-chemical properties, e.g. to avoid losses due to volatilisation during spiking or equilibration. Additional guidance on spiking procedures is given in Environment Canada (1995) (46).

• 29.Once the spiked sediment has been prepared, distributed to the replicate test vessels, and topped with the test water, it is desirable to allow partitioning of the test chemical from the sediment to the aqueous phase (e.g. (3)(7)(9)). This should preferably be done under the conditions of temperature and aeration used in the test. Appropriate equilibration time is sediment and chemicals specific, and can be in the order of hours to days and in rare cases up to several weeks (4-5 weeks) (e.g. (27)(47)). In this test, equilibrium is not awaited but an equilibration period of 48 hours to 7 days is recommended. Thus, time for degradation of the test chemical will be minimised. Depending on the purpose of the study, e.g., when environmental conditions are to be mimicked, the spiked sediment may be equilibrated or aged for a longer period.

• 30.At the end of this equilibration period, samples should be taken at least of the overlying water and the bulk sediment, at least at the highest concentration and a lower one, for analysis of the test chemical concentration. These analytical determinations of the test chemical should allow for calculation of mass balance and expression of results based on measured initial concentrations. In general, sampling disturbs or destroys the sediment water system. Therefore it is usually not possible to use the same replicates for sampling of sediment and worms. Additional ‘analytical’ vessels of appropriate dimensions have to be set up, which are treated in the same way (including the presence of test organisms) but not used for biological observations. The vessel dimensions should be selected to provide the sample amounts required by the analytical method. Details of sampling are described in paragraph 53.

PERFORMANCE OF THE TEST

Preliminary test

• 31.If no information is available on the toxicity of the test chemical towards Lumbriculus variegatus, it may be useful to conduct a preliminary experiment in order to determine the range of concentrations to be tested in the definitive test, and to optimise the test conditions of the definitive test. For this purpose a series of widely spaced concentrations of the test chemical are used. The worms are exposed to each concentration of the test chemical for a period (e.g. 28 d as in the definitive test) which allows estimation of appropriate test concentrations; no replicates are required. The behaviour of the worms, for example sediment avoidance, which may be caused by the test chemical and/or by the sediment, should be observed and recorded during a preliminary test. Concentrations higher than 1 000 mg/kg sediment dry weight should not be tested in the preliminary test.

Definitive test

• 32.In the definitive test, at least five concentrations should be used and selected e.g. based on the result of the preliminary range-finding test (paragraph 31), and as described in paragraphs 35, 36, 37 and 38.

• 33.A control (for replication see paragraphs 36, 37 and 38) containing all constituents, except for the test chemical, is run in addition to the test series. If any solubilising agent is used for application of the test chemical, it should have no significant effect on the test organisms as revealed by an additional solvent-only control.

Test design

• 34.The test design relates to the selection of the number and spacing of the test concentrations, the number of vessels at each concentration and the number of worms added per vessel. Designs for ECx estimation, for estimation of NOEC, and for conducting a limit test are described in paragraphs 35, 36, 37 and 38.

• 35.The effect concentration (e.g. EC50, EC25, EC10) and the concentration range, over which the effect of the test chemical is of interest, should be bracketed by the concentrations included in the test. Extrapolating much below the lowest concentration affecting the test organisms or above the highest tested concentration should be avoided. If — in exceptional cases — such an extrapolation is done, a full explanation must be given in the report.

• 36.If the ECx is to be estimated, at least five concentrations and a minimum of three replicates for each concentration should be tested; six replicates are recommended for the control or — if used — the solvent control in order to improve the estimation of control variability. In any case, it is advisable that sufficient test concentrations are used to allow a good model estimation. The factor between concentrations should not be greater than two (an exception can be made in cases when the concentration response curve has a shallow slope). The number of replicates at each treatment can be reduced if the number of test concentrations with responses in the range of 5 – 95 % are increased. Increasing the number of replicates or reducing the size of the test concentration intervals tends to lead to narrower confidence intervals for the test.

• 37.If the LOEC/NOEC values are to be estimated, at least five test concentrations with at least four replicates (six replicates are recommended for the control or — if used — the solvent control in order to improve the estimation of control variability) should be used, and the factor between concentrations should not be greater than two. Some information on the statistical power found during hypothesis testing in the ring test of the test method is given in Appendix 6.

• 38.A limit test may be performed (using one test concentration and controls) if no effects are expected up to 1 000 mg/kg sediment d.w. (e.g. from a preliminary range-finding test), or if testing at a single concentration will be adequate to confirm a NOEC value of interest. In the latter case, a detailed rationale for selection of limit concentration should be included in the test report. The purpose of the limit test is to perform a test at a concentration sufficiently high to enable decision makers to exclude possible toxic effects of the chemical, and the limit is set at a concentration which is not expected to appear in any situation. 1 000 mg/kg (dry weight) is recommended. Usually, at least six replicates for both the treatment and controls are necessary. Some information on the statistical power found during hypothesis testing in the ring test of the test method is given in Appendix 6.

Exposure conditions

Test organisms

• 39.The test is conducted with at least 10 worms for each replicate used for determination of biological parameters. This number of worms corresponds to approximately 50 - 100 mg of wet biomass. Assuming a dry content of 17,1 % (48), this results in approximately 9 - 17 mg of dry biomass per vessel. U.S. EPA (2000 (7)) recommends to use a loading rate not exceeding 1: 50 (dry biomass: TOC). For the formulated sediment described in paragraph 22, this corresponds to approximately 43 g sediment (dry weight) per 10 worms at a TOC content of 2,0 % of dry sediment. In cases where more than 10 worms are used per vessel, the amount of sediment and overlying water should be adjusted accordingly.

• 40.The worms used in a test should all come from the same source, and should be animals of similar physiological state (see Appendix 5). Worms of similar size should be selected (see paragraph 39). It is recommended that a sub-sample of the batch or stock of worms is weighed before the test in order to estimate the mean weight.

• 41.The worms to be used in a test are removed from the culture (see Appendix 5 for details). Large (adult) animals that do not show signs of recent fragmentation are transferred to glass dishes (e.g. petri dishes) containing clean water. They are subsequently synchronised as described in Appendix 5. After regenerating for a period of 10 to 14 d, intact complete worms of similar size, which are actively swimming or crawling after a gentle mechanical stimulus, should be used for the test. If the test conditions differ from the culture conditions (e.g. in temperature, light regime, and overlying water), an acclimation phase of e.g. 24 h at temperature, light regime, and using the same overlying water as in the test should be sufficient to adapt the worms to the test conditions. The adapted oligochaetes should be allocated randomly to the test vessels.

Feeding

• 42.Since food is added to the sediment prior to (or during) application of the test chemical, the worms are not fed additionally during the test.

Light and temperature

• 43.The photoperiod in the culture and the test is usually 16 hours (3), (7). Light intensity should be kept low (e.g. 100-500 lx) to imitate natural conditions at the sediment surface, and measured at least once during the exposure period. The temperature should be 20 °C ± 2 °C throughout the test. On one given measuring date the difference of temperature between test vessels should not be higher than ± 1 °C. The test vessels should be placed in the test incubator or the test area in a randomised way, e.g. in order to minimise bias of reproduction due to vessel location.

Aeration

• 44.The overlying water of the test vessels should be gently aerated (e.g. 2 - 4 bubbles per second) via a pasteur pipette positioned approx. 2 cm above the sediment surface so as to minimise perturbation of the sediment. Care should be taken that the dissolved oxygen concentration does not fall below 30 % of air saturation value (ASV). Air supply should be controlled and — if necessary — adjusted at least once daily on workdays.

Water quality measurements

• 45.The following water quality parameters should be measured in the overlying water:

Temperature:	at least in one test vessel of each concentration level and one test vessel of the controls once per week and at the start and the end of the exposure period; if possible, temperature in the surrounding medium (ambient air or water bath) may be recorded additionally e.g. at hourly intervals;
Dissolved oxygen content:	at least in one test vessel of each concentration level and one test vessel of the controls once per week and at the start and the end of the exposure period; expressed as mg/l and % ASV (air saturation value);
Air supply:	should be controlled at least once daily on workdays and — if necessary — adjusted;
pH:	at least in one test vessel of each concentration level and one test vessel of the controls once per week and at the start and the end of the exposure period;
Total water hardness:	at least in one replicate of the controls and one test vessel at the highest concentration at the start and the end of the exposure period; expressed as mg/l CaCO3;
Total ammonia content:	at least in one replicate of the controls and in one test vessel of each concentration level at the start of the exposure period, and subsequently 3 × per week; expressed as mg/l NH4 + or NH3 or total ammonia-N.

• If measurement of water quality parameters requires removal of significant water samples from the vessels, it may be advisable to set up separate vessels for water quality measurements so as not to alter the water-to-sediment volume ratio.

Biological observations

• 46.During the exposure, the test vessels should be observed in order to assess visually any behavioural differences in the worms (e.g. sediment avoidance, fecal pellets visible on the sediment surface) compared with the controls. Observations should be recorded.

• 47.At the end of the test, each replicate is examined (additional vessels designated for chemical analyses may be excluded from examination). An appropriate method should be used to recover all worms from the test vessel. Care should be taken that all worms are recovered uninjured. One possible method is sieving the worms from the sediment. A stainless steel mesh of appropriate mesh size can be used. Most of the overlying water is carefully decanted, and the remaining sediment and water is agitated to result in a slurry, which can be passed through the sieve. Using a 500 μm mesh, most of the sediment particles will pass the sieve very quickly; however, sieving should be done quickly, in order to prevent the worms from crawling into or through the mesh. Using a 250 μm mesh will prevent the worms from crawling into or through the mesh; however, care should be taken that as little as possible of the sediment particles is retained on the mesh. The sieved slurry of each replicate vessel may be passed through the sieve a second time in order to ensure that all worms are recovered. An alternative method could be warming of the sediment by placing the test vessels in a water bath at 50 – 60 °C; the worms will leave the sediment and can be collected from the sediment surface by use of a fire-polished wide-mouth pipette. Another alternative method could be to produce a sediment slurry and pour this slurry onto a shallow pan of suitable size. From the shallow layer of slurry the worms can be picked up by a steel needle or watchmakers' tweezers (to be used rather like a fork than forceps to avoid injuring the worms) and transferred to clean water. After separating the worms from the sediment slurry, these are rinsed in test medium and counted.

• 48.Independently of the method used, laboratories should demonstrate that their personnel are able to recover an average of at least 90 % of the organisms from whole sediment. For example, a certain number of test organisms could be added to control sediment or test sediments, and recovery could be determined after 1 h (7).

• 49.The total number of living and dead individuals per replicate should be recorded and assessed. The following groups of worms are considered to be dead:

• a) there is no reaction after a gentle mechanical stimulus

• b) there are signs of decomposition (in combination with ‘a’)

• c) number of missing worms

• Additionally, the living worms can be assigned to one of three groups:

• a) large complete worms (adults) without regenerated body regions

• b) complete worms with regenerated, lighter-coloured body regions (i.e., with new posterior part, with new anterior part, or with both new posterior and anterior parts)

• c) incomplete worms (i.e., recently fragmented worms with non-regenerated body regions)

• These additional observations are not mandatory, but can be used for additional interpretation of the biological results (for example, a high number of worms assigned to group c may indicate a delay of reproduction or regeneration in a given treatment). Additionally, if any differences in appearance (e.g. lesions of the integument, oedematous body sections) are observed between treated and control worms, these should be recorded.

• 50.Immediately after counting/assessment, the living worms found in each replicate are transferred to dried, pre-weighed and labelled weigh pans (one per replicate), and killed using a drop of ethanol per weigh pan. The weigh pans are placed in a drying oven at 100 ± 5 °C to dry overnight, after which they are weighed after cooling in a desiccator, and worm dry weight is determined (preferably in g, at least 4 post-decimal digits).

• 51.In addition to the total dry weight, the ash-free dry weight may be determined as described in (49) in order to account for inorganic components originating from ingested sediment present in the alimentary tract of the worms.

• 52.The biomass is determined as total biomass per replicate including adult and young worms. Dead worms should not be taken into account for the determination of biomass per replicate.

Verification of test chemical concentrations

Sampling

• 53.Samples for chemical analysis of the test chemical should be taken at least of the highest concentration and a lower one, at least at the end of the equilibration phase (before adding the test organisms), and at the end of the test. At least the bulk sediment and the overlying water should be sampled for analysis. At least two samples should be taken per matrix and treatment on each sampling date. One of the duplicate samples may be stored as a reserve (to be analysed e.g. in the event that initial analysis falls outside the ± 20 % range from the nominal concentration). In case of specific chemical properties, e.g. if rapid degradation of the test chemical is expected, the analytical schedule may be refined (e.g. more frequent sampling, analysis of more concentration levels) on the basis of expert judgment. Samples may then be taken on intermediate sampling dates (e.g. on day seven after start of exposure).

• 54.The overlying water should be sampled by carefully decanting or siphoning off the overlying water so as to minimise perturbation of the sediment. The volume of the samples should be recorded.

• 55.After the overlying water has been removed, the sediment should be homogenised and transferred to a suitable container. The weight of the wet sediment sample is recorded.

• 56.If analysis of the test chemical in the pore water is required additionally, the homogenised and weighed sediment samples should be centrifuged to obtain the pore water. For example, approximately 200 ml of wet sediment can be filled into 250 ml centrifugation beakers. Thereafter the samples should be centrifuged without filtration to isolate the porewater, e.g. at 10 000 ± 600 × g for 30 - 60 min at a temperature not exceeding the temperature used in the test. After centrifugation, the supernatant is decanted or pipetted taking care that no sediment particles are introduced, and the volume is recorded. The weight of the remaining sediment pellet is recorded. It may facilitate the estimation of the mass balance or recovery of the test chemical in the water-sediment system, if the sediment dry weight is determined at each sampling date. In some cases it might not be possible to analyse concentrations in the pore water as the sample size is too small.

• 57.Failing immediate analysis, all samples should be stored by an appropriate method, e.g. under the storage conditions recommended for minimum degradation of the particular test chemical (e.g., environmental samples are commonly stored at – 18 °C in the dark). Obtain information on the proper storage conditions for the particular test chemical — for example, duration and temperature of storage, extraction procedures, etc. — before beginning the study.

Analytical method

• 58.Since the whole procedure is governed essentially by the accuracy, precision and sensitivity of the analytical method used for the test chemical, check experimentally that the precision and reproducibility of the chemical analysis, as well as the recovery of the test chemical from water and sediment samples are satisfactory for the particular method at least at the lowest and highest test concentrations. Also, check that the test chemical is not detectable in the control chambers in concentrations higher than the limit of quantification. If necessary, correct the nominal concentrations for the recoveries of quality control spikes (e.g. where recovery is outside 80 - 120 % of spiked amount). Handle all samples throughout the test in such a manner so as to minimise contamination and loss (e.g. resulting from adsorption of the test chemical on the sampling device).

• 59.The recovery of test chemical, the limit of quantification, and the limit of detection in sediment and water should be recorded and reported.

DATA AND REPORTING

Treatment of results

• 60.The main mandatory response variables of the test to be evaluated statistically are the biomass and the total number of worms per replicate. Optionally, reproduction (as increase of worm numbers) and growth (as increase of dry biomass) could be also evaluated. In this case, an estimate of the dry weight of the worms at start of exposure should be obtained e.g. by measurement of the dry weight of a representative sub-sample of the batch of synchronised worms to be used for the test.

• 61.Although mortality is not an endpoint of this test, mortalities should be evaluated as far as possible. In order to estimate mortalities, the number of worms that do not react to a gentle mechanical stimulus or showed signs of decomposition, and the missing worms should be considered dead. Mortalities should at least be recorded and considered when interpreting the test results.

• 62.Effect concentrations should be expressed in mg/kg sediment dry weight. If the recovery of test chemical measured in the sediment, or in sediment and overlying water at start of exposure, is between 80 and 120 % of the nominal concentrations, the effect concentrations (ECx, NOEC, LOEC) may be expressed based on nominal concentrations. If recovery deviates from the nominal concentrations by more than ± 20 % of the nominal concentrations, the effect concentrations (ECx, NOEC, LOEC) should be based on the initially measured concentrations at the beginning of the exposure, e.g. taking into account the mass balance of the test chemical in the test system (see paragraph 30). In these cases, additional information can be obtained from analysis of stock and/or application solutions in order to confirm that the test sediments were prepared correctly.

ECx

• 63.ECx-values for the parameters described in paragraph 60 are calculated using appropriate statistical methods (e.g. probit analysis, logistic or Weibull function, trimmed Spearman-Karber method, or simple interpolation). Guidance on statistical evaluation is given in (15) and (50). An ECx is obtained by inserting a value corresponding to x % of the control mean into the equation found. To compute the EC50 or any other ECx, the per-treatment means (

Image omitted; see official source ) should be subjected to regression analysis.

NOEC/LOEC

• 64.If a statistical analysis is intended to determine the NOEC/LOEC, per-vessel statistics (individual vessels are considered replicates) are necessary. Appropriate statistical methods should be used. In general, adverse effects of the test item compared to the control are investigated using one-tailed (smaller) hypothesis testing at p ≤ 0,05. Examples are given in the following paragraphs. Guidance on selection of appropriate statistical methods is given in (15) and (50).

• 65.Normal distribution of data can be tested e.g. with the Kolmogorov-Smirnov goodness-of-fit test, the Range-to-standard-deviation ratio test (R/s-test) or the Shapiro-Wilk test, (two-sided, p ≤ 0,05). Cochran's test, Levene test or Bartlett's test, (two-sided, p ≤ 0,05) may be used to test variance homogeneity. If the prerequisites of parametric test procedures (normality, variance homogeneity) are fulfilled, One-way Analysis of Variance (ANOVA) and subsequent multi-comparison tests can be performed. Pairwise comparisons (e.g. Dunnett's t-test) or step-down trend tests (e.g. Williams' test) can be used to calculate whether there are significant differences (p ≤ 0,05) between the controls and the various test item concentrations. Otherwise, non-parametric methods (e.g. Bonferroni-U-test according to Holm or Jonckheere-Terpstra trend test) should be used to determine the NOEC and the LOEC.

Limit test

• 66.If a limit test (comparison of control and one treatment only) has been performed and the prerequisites of parametric test procedures (normality, homogeneity) are fulfilled, metric responses (total worm number, and biomass as worm dry weight) can be evaluated by the Student test (t-test). The unequal-variance t-test (Welch t-test) or a non parametric test, such as the Mann-Whitney-U-test may be used, if these requirements are not fulfilled. Some information on the statistical power found during hypothesis testing in the ring test of the method is given in Appendix 6.

• 67.To determine significant differences between the controls (control and solvent control), the replicates of each control can be tested as described for the limit test. If these tests do not detect significant differences, all control and solvent control replicates may be pooled. Otherwise all treatments should be compared with the solvent control.

Interpretation of results

• 68.The results should be interpreted with caution if there were deviations from this test method, and where measured concentrations of test concentrations occur at levels close to the detection limit of the analytical method used. Any deviations from this test method must be noted.

Test report

• 69.The test report should include at least the following information:

• — Test chemical: — chemical identification data (common name, chemical name, structural formula, CAS number, etc.) including purity and analytical method for quantification of test chemical; source of the test chemical, identity and concentration of any solvent used. — any information available on the physical nature and physical-chemical properties as obtained prior to start of the test, (e.g. water solubility, vapour pressure, partition coefficient in soil (or in sediment if available), log Kow, stability in water, etc.);

• — Test species: — scientific name, source, any pre-treatment, acclimation, culture conditions, etc..