Registration Dossier

Ecotoxicological information

Toxicity to terrestrial arthropods

Currently viewing:

Administrative data

Endpoint:
toxicity to terrestrial arthropods: long-term
Type of information:
experimental study
Adequacy of study:
weight of evidence
Reliability:
2 (reliable with restrictions)
Rationale for reliability incl. deficiencies:
other: Already evaluated by the Competent Authorities for Biocides and Existing Substance Regulations.
Cross-referenceopen allclose all
Reason / purpose:
reference to same study
Reason / purpose:
reference to other study

Data source

Reference
Title:
Unnamed
Year:
2005

Materials and methods

Test guidelineopen allclose all
Qualifier:
according to
Guideline:
ISO 11267 (Inhibition of Reproduction of Collembola by Soil Pollutants)
Qualifier:
according to
Guideline:
other: ISO 11268-2
GLP compliance:
no
Application method:
soil

Test material

Reference
Name:
Unnamed
Type:
Constituent
Details on test material:
IUCLID4 Test substance: other TS

TS-Freetext:
Test material
Copper was dosed to the soils as CuCl2.2H2O

Lot/Batch number
No data provided

Specification
Deviating from specification given in section 2 as follows;
Cu2+ was dosed to the soils as CuCl2.2H2O. No data on the
specification of the CuCl2.2H2O are provided.

Purity
No data provided

Composition of Product
No data provided

Further relevant properties
No further relevant properties

Method of analysis
This Research project was performed concurrent with two
other projects, referenced below and summarised in the
98/8/EC dossier submission. Analytical resource was
combined where possible, and data are reported in each
report as appropriate. Methodology was similar for each
project.

Natural soils.
These soils were analysed under this project and methodology
is as reported.

Leached and Aged soils:
For the leaching and ageing study, soils analysis is
reported in Smolders, E. and Oorts, K (2004) and Rooney, C.P
et al (2004).

Field validation soils:
Spiked soils were analysed under this project and are
reported herein, the field contaminated soils analysis is
reported in Smolders, E. and Oorts, K (2004) and Rooney, C.P
et al (2004).

Total soil Copper:
Total copper concentrations of all soils and treatments were
determined by aqua regia digestion in which 10 g dry wt of
soil and 10 mL of aqua regia (7.5 ml of HCl and 2.5 ml HNO3)
were kept at 50°C for 16h and subsequently heated at 160°C
for 1h. After cooling, the samples were diluted and analysed
using flame atomic absorption spectrometry.

Soil pore water total Copper:
Because of the strong variation in soil physical properties,
two different methods were used to isolate pore water from
all soils. For most soils the 'double chamber'
centrifugation method was used in which samples of soil are
put in a Nalgene polycarbonate tube with a little hole which
is filled with 0.5 cm quartz wool. Pore water is extracted
by centrifugation at 1320 g for 25 min and collected in a
Nalgene polypropylene open mouth centrifuge tube in which
the polycarbonate tube is put. For soils with a high clay
content the 'direct' centrifugation method was used. The
same open mouth tubes are filled with soil and centrifuged
at ± 22000 g for 15 min. The supernatant was then gently
decanted. Once pore water was extracted, it was filtered
over an Acrodisc Syringe Filter. Pore water copper
concentrations were analysed with Inductively Coupled
Plasma-Optical Emission Spectroscopy.

Soil pore water Copper activity (Cu2+):
Copper (Cu2+)activity was measured with a Thermo Orion
ionplus Cu-ion selective electrode. Prior to use the
electrode was polished with the polishing strip and soaked
for 5 min in 0.025 M H2SO4. Subsequently, the electrode was
rinsed thoroughly and stored in deionized water until
calibration. The electrode was calibrated using 1.00 10-4 M
Cu2+ solutions with which a range of 10-5 to 10-12 M Cu2+
was established using ethylenediamine. Ionic strength in
these solutions is adjusted to 0.01 M using a 1 M NaNO3
solution. For calibration, the copper activity in each of
these solutions was calculated with Visual Minteq version
2.23 (Gustafsson, 2004). The overall slope of the
calibration curve was approximately -28.3 mV/(log10aCu2+).
During calibration and measurements: light, stirring speed
and temperature are kept constant. The electrode gives an
equilibrium reading when the variation of potential was less
than 0.3 mV within 3 min. Since the amount of soil solution
obtained was relatively small, pH measurements were carried
out immediately after p(Cu) measurements. Measurements were
performed in 5 ml of soil solution.

Reference material
Yes - with every set of soil samples analysed during the
natural spiking phase of the experiment, a sample of
certified reference soil was digested and analysed
(calcerous loam soil, BCR CRM 141 R, Community Bureau of
Reference, European Commission).

Sampling and analysis

Details on sampling:
every 6 months

Test substrate

Details on preparation and application of test substrate:
Substrate type: 19 natural soils representing the range of volues for the soil parameters affecting bioavailability occuring in Europe + two standard soils (LUFA 2.2 and OECD soil). Properties of the soils: see table A7_5_2_1-4 in attached document.
Equilibration time: 7d

Test organisms

Test organisms (species):
other: Folsomia candida and Eisenia fetida
Details on test organisms:
F. candida: test organisms 10-12 days old
E. fetida: test organisms are worms with fully developed clitellum and wet weight between 400 and 700 mg.

Study design

Study type:
laboratory study
Total exposure duration:
28 d

Test conditions

Test temperature:
20+-1°C
pH (if soil or dung study):
depending on test soil.
Humidity:
between 50 and 60% of water holding capacity
Photoperiod and lighting:
light:dark cycle of 16:8
Nominal and measured concentrations:
6 to 9 copper concentrations (control plus 5 to 8 treatments)

Results and discussion

Effect concentrationsopen allclose all
Duration:
28 d
Dose descriptor:
NOEC
Effect conc.:
174 mg/kg soil dw
Nominal / measured:
meas. (not specified)
Basis for effect:
reproduction
Remarks:
- F. candida, sandy loam Nottingham
Remarks on result:
other: see attached table
Duration:
28 d
Dose descriptor:
NOEC
Effect conc.:
28.2 mg/kg soil dw
Nominal / measured:
meas. (not specified)
Basis for effect:
reproduction
Remarks:
- F. candida, loamy sand, Houthalen
Duration:
28 d
Dose descriptor:
NOEC
Effect conc.:
279 mg/kg soil dw
Nominal / measured:
meas. (not specified)
Basis for effect:
reproduction
Remarks:
- F. candida, loamy sand, Rhydtalog
Duration:
28 d
Dose descriptor:
EC10
Effect conc.:
1 390 mg/kg soil dw
Nominal / measured:
meas. (not specified)
Basis for effect:
reproduction
Remarks:
- F. candida, Sandy clay loam, Zegveld
Duration:
28 d
Dose descriptor:
EC10
Effect conc.:
55.5 mg/kg soil dw
Nominal / measured:
meas. (not specified)
Basis for effect:
reproduction
Remarks:
- F. candida, loamy sand, Kovlinge
Duration:
28 d
Dose descriptor:
NOEC
Effect conc.:
53.1 mg/kg soil dw
Nominal / measured:
meas. (not specified)
Basis for effect:
reproduction
Remarks:
- F. candida, sandy clay, Souli
Duration:
28 d
Dose descriptor:
NOEC
Effect conc.:
172 mg/kg soil dw
Nominal / measured:
meas. (not specified)
Basis for effect:
reproduction
Remarks:
- F. candida, loamy sand, Montpellier
Duration:
28 d
Dose descriptor:
NOEC
Effect conc.:
276 mg/kg soil dw
Nominal / measured:
meas. (not specified)
Basis for effect:
reproduction
Remarks:
- F. candida, clay, Aluminusa
Duration:
28 d
Dose descriptor:
NOEC
Effect conc.:
244 mg/kg soil dw
Nominal / measured:
meas. (not specified)
Basis for effect:
reproduction
Remarks:
sandy clay loam, Woburn
Duration:
28 d
Dose descriptor:
NOEC
Effect conc.:
237 mg/kg soil dw
Nominal / measured:
meas. (not specified)
Basis for effect:
reproduction
Remarks:
- F. candida, silt loam, Ter Munck
Duration:
28 d
Dose descriptor:
NOEC
Effect conc.:
534 mg/kg soil dw
Nominal / measured:
meas. (not specified)
Basis for effect:
reproduction
Remarks:
- F. candida, silty clay loam, Vault de lugny
Duration:
28 d
Dose descriptor:
NOEC
Effect conc.:
160 mg/kg soil dw
Nominal / measured:
meas. (not specified)
Basis for effect:
reproduction
Remarks:
- F. candida, silty clay loam, Rots
Duration:
28 d
Dose descriptor:
NOEC
Effect conc.:
887 mg/kg soil dw
Nominal / measured:
meas. (not specified)
Basis for effect:
reproduction
Remarks:
- F. candida, Clay, Souli
Duration:
28 d
Dose descriptor:
NOEC
Effect conc.:
453 mg/kg soil dw
Nominal / measured:
meas. (not specified)
Basis for effect:
reproduction
Remarks:
- F. candida, silt loam, Marknesse
Duration:
28 d
Dose descriptor:
NOEC
Effect conc.:
139 mg/kg soil dw
Nominal / measured:
meas. (not specified)
Basis for effect:
reproduction
Remarks:
- F. candida, loam, Barcelona
Duration:
28 d
Dose descriptor:
NOEC
Effect conc.:
632 mg/kg soil dw
Nominal / measured:
meas. (not specified)
Basis for effect:
reproduction
Remarks:
- F. candida, clay, Brécy
Duration:
28 d
Dose descriptor:
NOEC
Effect conc.:
538 mg/kg soil dw
Nominal / measured:
meas. (not specified)
Basis for effect:
reproduction
Remarks:
- F. candida, loam, Guadalajara
Duration:
28 d
Dose descriptor:
NOEC
Effect conc.:
493 mg/kg soil dw
Nominal / measured:
meas. (not specified)
Basis for effect:
reproduction
Remarks:
- F. candida, sandy clay, Hygum
Duration:
28 d
Dose descriptor:
NOEC
Effect conc.:
27.9 mg/kg soil dw
Nominal / measured:
meas. (not specified)
Basis for effect:
reproduction
Remarks:
- F. candida, loamy sand, Wageningen A
Duration:
28 d
Dose descriptor:
NOEC
Effect conc.:
48 mg/kg soil dw
Nominal / measured:
meas. (not specified)
Basis for effect:
reproduction
Remarks:
- F. candida, loamy sand, Wageningen D
Details on results:
part 2 and 3 of the results: see WOE, Criel 2005.part2/part 3

Any other information on results incl. tables

RS-Freetext:
Copper analysis - All Soil tests

Initial concentrations of test substance - Total copper
concentrations of all soils and treatments are presented in
tables A7_5_2_1-3 to A7_5_2_1-8.
Slight differences in experimental values reported during
the experimental phase (tables A7_5_2_1-5 to A7_5_2_1-8) and
the soil characterisation (tables A7_5_2_1-3 to A7_5_2_1-4)
were observed. These are probably attributable to sampling,
extraction and analytical variability, not considered to
have adversely affected the conclusions of the tests.
Soil pore water total Copper and Soil pore water Copper
activity (Cu2+) data are not reported, however effect data
based upon these parameters are presented in the report.

Concentrations of reference material - For this reference
material a value of 43.3 ± 1.6 (sd) Cu mg/kg dry wt was
found over all analyses, which is 92.4 ± 3.3 (sd) % of the
value given as certified aqua regia soluble fraction
(Quevauviller et al., 1996). In most cases the measured
value differed less than 10% from the nominal concentration.

Effects of soil characteristics

Concentration-response curves

E. fetida and F. candida were able to reproduce in all 19
soils selected for the bioavailability subproject. In case
of E. fetida the results for the reproduction assay in 4
soils (Houthalen, Rhydtalog, Barcelona and Guadalajara) were
omitted from further analyses due to low control
reproduction and/or high variability between replicate. For
F. candida one soil (Kövlinge II) was left out of analyses
for the same reason.

The fitted concentration-response curves for the remaining
soils are shown in Figures A7_5_2_1-1 and A7_5_2_1-2 for E.
fetida and F. candida, respectively. Hormesis was
significant (p = 0.001) in souli II soil for the 28d
reproduction assay with E. fetida (Figure 1). Curve fitting
was carried out with all the the hormetic responses
included, using a transformed formula of that given in
Material and Methods. In the Zegveld soil (in case of E.
fetida and F. candida), the Brécy soil (in case of E.
fetida) and the Kövlinge I soil (in case of F. candida) the
control response is somewhat higher than what is estimated
from curve fitting (Figures A7_5_2_1-1 and A7_5_2_1-2).

Toxicity thresholds

Table A7_5_2_1-9 gives an overview of 28d NOECs, EC10s,
EC20s and EC50s determined in the natural 19 soils for
reproduction of E. fetida and F. candida. Toxicity data for
both species are expressed as measured total copper
concentrations. Table A7_5_2_1-10 gives the same data
expressed as measured added copper concentrations.
In five cases no toxicity values are reported due to low
control reproduction and/or high variability between
replicates. In some cases no NOEC could be determined as a
significant effect was observed in the lowest concentrations
tested (unbounded LOEC). EC10 values for one soil in the E.
fetida reproduction assay (Aluminusa) and one soil in the F.
candida reproduction assay (Gudow) were less than the lowest
concentration tested.
Toxicity values vary largely among soils. For the earthworm
E. fetida 28d NOECs, EC10s, EC20s and EC50s range (on a
total copper basis) from 53.8 to 328 mg Cu/kg dry wt (factor
~ 6.1), 35.7 to 485 mg Cu/kg dry wt (factor ~ 13.6), 46.3 to
573 mg Cu/kg dry wt (factor ~ 12.4) and 72.0 to 781 mg Cu/kg
dry wt (factor ~ 10.8), respectively.
For the springtail F. candida these ranges are much larger:
(on a total copper basis) ranges from 30.1 to 922 mg Cu/kg
dry wt (factor ~ 30.6), 12.2 to 1450 mg Cu/kg dry wt (factor
~ 119), 20.2 to 1720 mg Cu/kg dry wt (factor ~ 85.1) and
45.4 to 2270 mg Cu/kg dry wt (factor ~ 50) are found for 28d
NOECs, EC10s, EC20s and EC50s, respectively.
Based on ECx values Montpellier is the most sensitive soil
(i.e. the soil that has lowest toxicity thresholds) for E.
fetida. For F. candida, Zegveld is the least sensitive and
Gudow and Houthalen are the most sensitive soils.

Effects of Leaching and Ageing

Concentration-response curves

Figures A7_5_2_1-3 and A7_5_2_1-4 give an overview of the
concentration-response curves obtained in the ageing assays
for E. fetida and F. candida, respectively.
Uniform scaling on both axes was employed to facilitate
comparison between the different treatments. Results on the
6 months ageing soils are not reported as tests were
performed one year after soil sampling. In the case of E.
fetida, no concentration-response curves are given of the
fresh spike assay for the Houthalen and Barcelona soil and
the 18 months ageing treatment for the Houthalen soil due to
low control reproduction and high variability between
replicates. For the 12 months ageing assay in the Houthalen
soil the estimated EC50 values for F. candida reproduction
was higher than the highest concentration tested so no
results are shown. In Woburn only one partial effect was
found after 18 months of ageing and the results were left
out of consideration in further analysis.

Toxicity thresholds

The results of the ageing experiment for the 3 soils studied
are summarised in Table A7_5_2_1-10. Toxicity data for both
species are expressed as measured total concentrations.
Table A7_5_2_1-12 gives the same data as measured added
concentrations.
In the case of E. fetida, two sets of results (Houthalen and
Barcelona soil) are incomplete. For F. candida, at least one
set of threshold values is missing for every soil. For the
Barcelona soil the ageing experiment remains inconclusive.
Some influence of leaching and/or ageing can be seen for the
Houthalen soil. No effect of leaching was found for the
Woburn soil for both test species used. Copper toxicity
decreased with ageing in the Houthalen and Woburn soil for
both invertebrates.

Field validation results

Concentration-response curves

Figure A7_5_2_1-5 presents the concentration-response curves
for the Hygum field transect and the lab spiked control soil
for E. fetida.
In contrast to the data of the lab spiked control soil, no
response curve could be fitted through the field transect
data.
Figure A7_5_2_1-6 presents the concentration-response curves
for the Hygum, Wageningen A, Wageningen D, Woburn MS and
Woburn SC soils and the lab spiked control soils for F.
candida. A concentration-response curve could be fitted to
none of the field transect data. It is possible that the
lower pH in the control soil (pH 5.0 vs. pH 5.3 in field
transect) resulted in the observed lower reproduction for
the control soil in the Wageningen D transect. The number of
juveniles found in the control treatment of the Woburn MS
soil is significantly different from the number observed in
the soil from the transect points. The reason for this could
not be ascertained. Figures A7_5_2_1-3 and A7_5_2_1-3 show
the concentration-response curves for both test species for
the field transects and their respective control soils,
based on soil solution copper concentration.

Toxicity thresholds

Table A7_5_2_1-13 gives an overview of 28d NOEC, EC10, EC20
and EC50 values determined for transects of field
contaminated soils and the lab spiked control soils for
reproduction of E. fetida and F. candida. Toxicity data for
both species are expressed as measured total concentrations.
Table A7_5_2_1-14 gives the same data for both species as
measured added concentrations.
In three cases (Hygum, Wageningen A and Wageningen D) no
significant effect was detected at the highest concentration
sampled in the field transect. In those cases the NOEC is
unbounded and no ECx values could be determined. In other
cases (Hygum for E. fetida, Woburn MS and Woburn SC) the
effect seen in the highest tested concentration was lower
than 50%, which resulted in a weak concentration response
fit. A significant adverse effect was found in the lowest
concentration tested in the Woburn MS field transect. This
is probably due to an unknown confounding factor. These
results were not used in further analysis. Due to an
insufficient effect at the highest concentration tested, no
ECx values could be determined in the lab spiked control
soil of Wageningen D and Woburn MS.

Summary and Conclusion

Materials and methods
To study the effect of soil characteristics on the toxicity
of copper to terrestrial invertebrates, chronic toxicity
tests with Eisenia fetida and Folsomia candida were
performed in 19 natural soils according to ISO testing
guidelines (ISO 11268-2 and ISO 11267). These uncontaminated
soils were selected to represent the range of values for the
soil parameters affecting bioavailability that can be found
in European soils. Soils were spiked in the laboratory with
CuCl2 solutions and 4 week reproduction assays with both
invertebrates were conducted.
In an additional testing programme, the same organism
testing systems were employed using various soil preparation
systems to study the effect of ageing.
Three of the soils were spiked at the Laboratory for Soil
and Water Management, K.U. Leuven, and leached or placed
outdoors for a period of 18 months. Soils were sampled every
6 months and chronic toxicity tests were performed. For
field validation experiments transects of copper
contaminated soils and their uncontaminated control soils
were sampled at different locations and used as exposure
soils in ecotoxicity tests.
The guidelines were followed in principle, the deviations
being in relation to the preparation of the soils. These
deviations were core to the research programme, and are not
considered to have affected the validity of the results.

Results and discussion
A significant correlation (R2 = 0.41, p = 0.014) was found
between toxicity on reproduction (EC50 values) for E. fetida
and F. candida, indicating that soil factors are affecting
the bioavailability of copper in the two invertebrates
species in a similar way. Below, 28d EC50 values for F.
candida are plotted against 28d EC50 values for E. fetida.
As most data points in the graph are situated above the 1:1
line, it can be concluded that for most soils the earthworm
E. fetida is more sensitive to copper toxicity than the
springtail F. candida.

For both species variation in copper toxicity thresholds was
best explained by changes in cation exchange capacity (CEC).
Using the obtained regression algorithms, NOECs, EC10s,
EC20s and EC50s could be predicted within a factor of 2 for
E. fetida and within a factor of 3 for F. candida. These
models were also validated using the independent results
obtained in the field study.
In the ageing study, leaching only altered copper toxicity
to F. candida for one (Houthalen soil) out of three soils.
For E. fetida the results of two soils (Houthalen and
Barcelona) remain inconclusive and no effect of leaching was
found in the remaining soil (Woburn). Copper toxicity
decreased with ageing in the Houthalen and Woburn soil for
both invertebrates, the effect of ageing remains
inconclusive in the Barcelona soil. In the the field
transect study, no concentration-response curves could be
fitted to the toxicity data of the field contaminated soils.
In three cases a lab-to-field factor could be calculated, in
those soils there was clear evidence for lower copper
toxicity in field contaminated soils. The other two cases
remained inconclusive as no toxicity values could be
determined for the lab spiked control soils.

ECx
Soil toxicity values varied largely among soils. For the
earthworm E. fetida 28d NOECs, EC10s, EC20s and EC50s ranged
from 53.8 to 328 mg Cu/kg dry wt , 35.7 to 485 mg Cu/kg dry
wt, 46.3 to 573 mg Cu/kg dry wt and 72,0 to 781 mg Cu/kg dry
wt, respectively.
For the springtail F. candida these ranges were much larger:
30.1 to 922 mg Cu/kg dry wt, 12.2 to 1450 mg Cu/kg dry wt,
20.2 to 1720 mg Cu/kg dry wt and 45.4 to 2270 mg Cu/kg dry
wt were found for 28d NOECs, EC10s, EC20s and EC50s,
respectively.

Conclusion


Soil characteristics clearly influence the toxicity of
copper to E. fetida and F. candida; this is demonstrated by
large variation in toxicity effect data found for both
invertebrate species tested in the 19 soils studied.

Linear regression analysis revealed that cation exchange
capacity (CEC) correlated well with the toxicity effect data
for the species tested in this study. 57% of the variation
in EC50s for both invertebrate species can be explained by
changes in cation exchange capacity. Based on the obtained
regression equations we were able to predict toxicity within
a factor of two for E. fetida and a factor of three for F.
candida. Using these normalization algorithms, variation in
28d EC50s among soils can be reduced from a factor of 11 to
a factor of 4 for E. fetida and from a factor 50 to a factor
9 for F. candida.

When toxicity is expressed as Cu2+ activity, variation in
toxicity thresholds is even larger. In that case highly
significant correlations (R2 ± 0.95) where found with soil
solution pH.

For the Barcelona soil the effect of leaching/ageing is
inconclusive. No effect of leaching could be found except
for the Houthalen soil. A clear effect of ageing s.s. (i.e.
ageing effect without the effect of leaching) appears for
both invertebrate species in the Houthalen and Woburn soil.
In the the field transect study, no concentration-response
relations could be established through the field transect
data. In three cases no significant effect was detected at
the highest concentration sampled and in the other cases the
observed effect was lower than 50%. In three cases a
lab-to-field factor could be calculated, in those soils
there was clear evidence for lower copper toxicity in field
contaminated soils. The other two cases remained
inconclusive as no toxicity values could be determined for
the lab spiked control soils.

Applicant's summary and conclusion

Validity criteria fulfilled:
yes
Conclusions:
Good quality study. NOEC data were used for the PNEC derivation. Reliable added NOEC values varied between 28 and 1390 mg/kg for F candida (reproduction) and between 48 and 303 mg/kg for E. fetida (reproduction). For aged soils, reliable added NOEC values varied between 85 and 804 mg/kg for F. candida (reproduction).
Executive summary:

Ecotoxicity tests on copper are carried out with E. fetida and F. Candida using a wide range of soils. wide Soil characteristics clearly influence the toxicity of copper to E. fetida and F. candida; this is demonstrated by
large variation in toxicity effect data found for both invertebrate species tested in the 19 soils studied.

Linear regression analysis revealed that cation exchange capacity (CEC) correlated well with the toxicity effect data
for the species tested in this study. 57% of the variation in EC50s for both invertebrate species can be explained by
changes in cation exchange capacity. Based on the obtained regression equations we were able to predict toxicity within
a factor of two for E. fetida and a factor of three for F. candida. Using these normalization algorithms, variation in
28d EC50s among soils can be reduced from a factor of 11 to a factor of 4 for E. fetida and from a factor 50 to a factor
9 for F. candida.

When toxicity is expressed as Cu2+ activity, variation in toxicity thresholds is even larger. In that case highly
significant correlations (R2 ± 0.95) where found with soil solution pH.

For the Barcelona soil the effect of leaching/ageing is inconclusive. No effect of leaching could be found except
for the Houthalen soil. A clear effect of ageing s.s. (i.e. ageing effect without the effect of leaching) appears for
both invertebrate species in the Houthalen and Woburn soil. In the the field transect study, no concentration-response
relations could be established through the field transect data. In three cases no significant effect was detected at
the highest concentration sampled and in the other cases the observed effect was lower than 50%. In three cases a
lab-to-field factor could be calculated, in those soils there was clear evidence for lower copper toxicity in field
contaminated soils. The other two cases remained inconclusive as no toxicity values could be determined for the lab spiked control soils.