Euro-Mediterr J Environ Integr (2017) 2:5

DOI 10.1007/s41207-017-0015-z

ORIGINAL PAPER

The bait-lamina earthworm test: a possible addition to the chronic

earthworm toxicity test?
Stephan Jänsch1 • Adam Scheffczyk1 • Jörg Römbke1

Received: 3 November 2016 / Accepted: 2 February 2017 / Published online: 13 February 2017
Ó Springer International Publishing Switzerland 2017

Abstract The bait-lamina method is a functional test Keywords Ecotoxicology
Eisenia andrei
Feeding
method which is successfully used in field monitoring activity
Functional test
Chemicals
Soil quality
studies to assess the feeding activity of soil organisms. This
endpoint addresses an important soil function and service:
nutrient cycling. Therefore, the test has recently been Introduction
standardized by the International Organization for Stan-
dardization. In the approach presented here, the bait strips The bait-lamina method was originally developed for
were used in the laboratory as an additional endpoint in the measuring the biological activity of soils (Von Törne
standard earthworm reproduction test, regularly performed 1990). It is based on an optical evaluation of the feeding on
for the registration of chemicals as well as in soil assess- small portions of a thin-layered bait substrate exposed to
ment. The combined test was performed with the model edaphic soil processes (Kratz 1998). This method has the
chemicals zinc nitrate and tributyltin-oxide (TBT-O) in advantage of being able to detect effects on the soil fauna
Organization for Economic Co-operation and Development in a short time and with little effort. In ecotoxicology, the
artificial soil and eight natural soils. It was checked whe- bait-lamina method has mainly been used to monitor the
ther the results, available after 1 week, could be used to effects of chemicals on the feeding activity of the soil
predict the effects of the chemicals on earthworm repro- biocoenosis in the field (Federschmidt and Römbke 1994;
duction, available after 8 weeks. Out of 15 comparisons of Van Gestel et al. 2001; Simpson et al. 2012). At least four
EC50 values made in this study, in six of them there was no groups of 16 bait-lamina are inserted into the upper soil
difference. In four cases the feeding rate was more sensi- layer and, depending on the soil, evaluated for the feeding
tive and in five cases it was the other way around. The bait- activity (percentage of empty holes) after a period of 10–20
lamina earthworm test (BLET) could be performed in the days. The method has also already been applied in the
laboratory either as a screening tool for estimating the assessment of terrestrial model ecosystems (TME), where
range of chronic ecotoxicity of chemicals or for testing the four strips per TME were inserted and evaluated after
habitat quality of potentially contaminated soils (e.g. as 2 weeks (Förster et al. 2004). Except for some preliminary
part of a quick ‘‘on-site’’ analysis). Further research, work in the 1990s (Kampmann 1994), the method has not
mainly the preparation of a standard stand-alone BLET, its been used in ecotoxicological laboratory tests so far.
application on chemicals with different mode-of-actions, Earthworms are important in the soil due to their ability
and the definition of a reference substance as a positive to change their habitat or even create new ones for other
control, is needed. organisms through various activities: thus they are con-
sidered to be ‘‘ecosystem engineers’’ (Jones et al. 1994).
They fulfill various ecologically important functions in
& Jörg Römbke soil, thus finally lead to an improved soil structure, increase
[email protected]
of water infiltration (Edwards and Lofty 1977), often to the
1
ECT Oekotoxikologie GmbH, Böttgerstr. 2-14, formation of a humic layer close to the soil surface and to
65439 Flörsheim, Germany an increased yield in orchards or grassland (Van Groenigen

123
5 Page 2 of 11 Euro-Mediterr J Environ Integr (2017) 2:5

et al. 2014). Even the activity of the compost worms of the

genus Eisenia results in the formation of water-stable ag-
gregates (Schrader and Zhang 1997). The species E. andrei
is recommended in various international standard guideli-
nes (ISO 1998; OECD 1984, 2004), because it is easy to
culture and breed. E. andrei is reproducing hermaphrodi-
cally, but usually two partners are necessary (Dominguez
et al. 2003). For these reasons earthworms are very suit-
able ecotoxicological test organisms.
In contrast to the bait-lamina method, the earthworm
reproduction test is a common laboratory requirement for
assessing the effects of chemicals (pesticides, metals,
pharmaceuticals, etc.) (OECD 2004). The overall
approach of this work was to add bait-lamina strips to the
Fig. 1 Filling of bait-material into the bait-lamina strips
earthworm reproduction test in the laboratory. This
combination, called bait-lamina earthworm test (BLET),
was used to test two model chemicals (a metal and a tin- apart from each other (Fig. 1). The target soil depth of the
organic chemical) in OECD (Organization for Economic first aperture is 0.5 cm and that of the 16th aperture 8.0 cm.
Co-operation and Development) artificial soil (OECD
1984) and eight natural field soils (all of them originating Preparation of bait-lamina strips
from Germany). These two substances have been selected
since they represent two very different chemical classes The bait substrate comprised 70% cellulose powder, 25%
and thus, mode-of-actions. In addition, their effects on finely ground wheat bran and 5% activated carbon powder.
soil organisms (besides earthworms also springtails) and The components were homogeneously mixed and stirred with
plants any were already studied in a wide range of soils, deionized water to a viscid paste. The paste was filled into the
including those used in this study (e.g. Römbke et al. holes of the bait-lamina from both sides in various cycles
2006a, 2007). Especially the latter is known as being (Fig. 1). The strips were air-dried for at least 24 h prior to use.
toxic for annelids and, in particular, snails (Moore et al.
1991; Giusti et al. 2013). Earthworms
In particular, two key questions were addressed when
performing this combination of the two test methods: The earthworms used in these tests, Eisenia andrei (Lum-
bricidae), came from a synchronized culture, were adult,
1. Is it possible to use the results of the BLET, available had a fresh weight between 300 and 600 mg, and were at
after 1 week, to predict the range of effects of a test least 2 months but not more than 1 year old. The worms
chemical on earthworm reproduction, available after were fed with rolled oats at least every 2 weeks and
8 weeks? If so, the BLET test could be used as a range- occasionally with cow dung. Worms selected for the test
finding test for the earthworm reproduction test. were acclimatized in the respective soil under test condi-
2. Could the BLET be used for determining the effects of tions for at least 24 h before starting the test.
potentially contaminated field soils on the functional
endpoint ‘‘feeding rate’’? In this case, it could be used Soils
for the assessment of soil habitat quality.
The OECD artificial soil (OECD 1984) and eight natural
soils, being representative for Germany, were selected and
Materials and methods characterized for these tests (Jessen-Hesse et al. 2005). The
sampling locations of these soils are given Fig. 2. Indi-
Description of the bait-lamina strips vidual properties of these soils are compiled in Table 1.
Soil samples taken at the respective sites were air-dried,
The bait-lamina strips used are made of polyvinyl chloride sieved (5 mm mesh size) and stored in 25 l plastic buckets
(PVC) and have a length of 120 mm, width of 6 mm and at room temperature, for no longer than 3 months. In the
thickness of 1 mm. They have a pointed tip at the lower tests, they were moistened to 40–60% of the maximum
end. In the lower part (85 mm) of each strip, 16 bi-conical water-holding capacity (WHCmax). The amount of soil
apertures of 1.5 mm diameter are drilled that are 5 mm used in the tests per vessel was 500 g dry weight (DW).

123
Euro-Mediterr J Environ Integr (2017) 2:5 Page 3 of 11 5

Test chemicals was obtained from Merck (Darmstadt, Germany; CAS–

19154-63-3). In addition to the treatments with zinc nitrate,
Two model chemicals were selected to cover a broad range four vessels were spiked with 1900 mg nitrate/kg DW soil
of chemicals properties: Zinc nitrate-tetrahydrate (as potassium nitrate). This concentration was equivalent to
(Zn(NO3)2 9 4H2O) with a molar mass of 261.44 g/mol the nitrate concentration when testing the highest zinc
nitrate concentration.
Bis(tri-n-butyltin)oxide (TBT-O; C24H54OSn2) with a
molar mass of 596.07 g/mol was obtained from Crompton
(Bergkamen, Germany; CAS-No. 56-35-9). Aqueous
solutions of each test substance were mixed into the soils.
Each treatment (=concentration) was set up separately.

Test performance

In the chronic earthworm tests (ISO 1998), ten worms in

each of the four replicates were exposed to the spiked test
soil for 28 days at 18–22 °C and 400–800 lx (light regime
16:8 h). Test vessels were polystyrene cups with a base
area of 110 9 155 mm and a height of 60 mm (Bellaplast
590, Polarcup, Alf, Germany). Five concentrations of zinc
nitrate were selected according to literature data (Römbke
et al. 2006a): 181, 362, 725, 1449 and 2898 mg Zn(NO3)2/
kg DW soil, corresponding to 63, 125, 250, 500 and
1000 mg zinc/kg DW soil. For TBT-O, five concentrations
per soil were selected according to range-finding tests: for
OECD artificial soil 3.2, 10.0, 31.6, 100 and 316 mg TBT-
O/kg DW soil, for LUFA St. 2.2 soil 1.0, 3.2, 10.0, 31.6
and 100 mg TBT-O/kg DW soil, and for all other soils
0.32, 1.0, 3.2, 10 and 31.6 mg TBT-O/kg DW soil. The
adult worms were fed with finely ground cow manure (free
of growth promoters, nematicides or similar veterinary
pharmaceuticals, provided by Jesus Bruderschaft e.V.
Fig. 2 Locations in Germany where the test soils have been collected Gnadenthal, Hünfelden, Germany). Before starting the

Table 1 Physico-chemical characterization of the tested soils (eight natural field soils and OECD artificial soil)
Soila WHCmax (ml/kg) pH (CaCl2) OC (%) N (%) C/N CEC (cmol ?/kg) Sand (%) Silt (%) Clay (%)

OECD 631 6.0 4.7 0.07 67.1 8.9 75.4 16.6 8.04
BRG 601 4.9 2.34 0.29 8.1 14.1 13.6 56.7 29.7
BWZ 307 3.8 1.54 0.05 30.8 3.3 81.3 13.6 5.1
GGI 232 5.5 0.94 0.06 15.7 2.0 80.5 15.7 3.82
HAG 611 5.2 2.64 0.28 9.4 13.2 12.8 62.3 24.9
SBG 653 5.8 3.37 0.33 10.2 11.8 27.0 47.1 25.9
SHA 474 7.4 2.22 0.16 13.9 19.8 7.79 69.7 22.5
SOE 483 6.6 1.63 0.16 10.2 13.8 1.97 83.0 15.0
LUFA St. 2.2 500 6.1 2.7 0.19 14.2 7.9 76.9 16.3 6.84
WHCmax maximum water-holding capacity, OC organic carbon, CEC cation exchange capacity (Jessen-Hesse et al. 2005)
a
Abbreviations and locations of soils: OECD artificial soil according to OECD (1984), BRG Breddewarden, Lower Saxony, BWZ Weitzgrund,
Brandenburg, GGI Raddusch, Brandenburg, HAG Frankfurt-Harheim, Hesse, SBG Schmallenberg, North Rhine-Westphalia, SHA Schafstädt,
Saxony-Anhalt, SOE Soest, North Rhine-Westphalia, LUFA St. 2.2 Landwirtschaftliche Untersuchungs- und Forschungsanstalt 2.2 standard soil,
Speyer, Rhineland-Palatinate; all locations are in Germany

123
5 Page 4 of 11 Euro-Mediterr J Environ Integr (2017) 2:5

inserted so that all holes were beneath the soil surface

(Fig. 3). After 1 week, the strips were removed and the
number of empty holes were optically assessed (Fig. 4).

Data assessment

The statistical calculations for all tests were performed

with the program ToxRat, Version 2.09 (TOX-Rat 2003).
For the calculation of the NOEC (no observed effect con-
centration) and LOEC (lowest observed effect concentra-
tion) values (reproduction, feeding activity on bait-lamina),
data were checked for normality (R/S test, Kolmogorov–
Smirnov test) and variance homogeneity (Cochran’s test).
Depending on the results, either the Welch t test for
Fig. 3 Use of bait-lamina strips in test vessels of the standard inhomogeneous variances with Bonferroni adjustment, the
earthworm reproduction test Williams t test (monotonous dose–response) or Dunnett’s
t test (non-monotonous dose–response) were used (EC
2007). In case the data were not normally distributed, the
U test according to Mann–Whitney (significance level
corrected for multiple comparisons) was applied. The sig-
nificance level for the parametric tests (one-sided) was
p B 0.05. For the calculation of the LC50 (median lethal
concentration), EC10 (10% effective concentration) and
EC50 (median effective concentration) values (reproduc-
tion, feeding activity on bait-lamina), a linear regression
was performed via probit analysis.
As an indicator of the power of the hypothesis tests the
minimum detectable difference (MDD) in % of the control
value was calculated for the NOEC values (TOX-Rat
2003). For EC50-values the 95% confidence limits were
determined, as far as possible. To investigate the precision
and robustness of the different endpoints and compare the
Fig. 4 Visual examination of the feeding activity on bait-lamina results of the two test systems used, the arithmetic means
strips with standard deviation and the coefficient of variation
(CV) in % were calculated. Variance of the results is also
experiment an amount of 5 g food per 500 g soil DW was presented by the minimum and maximum values and the
mixed into the test soil. After the start of the experiment, factor between the minimum and maximum value. With
food was first provided one day after application of the test respect to the distribution of the single values in addition to
item and introduction of the adult worms. Thereafter, the the arithmetic mean, the geometric mean was calculated to
adult worms were fed weekly during the first 4 weeks of compare the endpoints.
the test. An amount of 5 g of food moistened with 10 ml
deionized water was spread on the soil surface of each test
container. If the food remained uneaten the ration was Results
reduced on demand. After removing of the adults on day
28, a further 5 g of food moistened with 10 ml deionized Validity criteria
water was mixed into the soil of each test vessel. Mortality,
biomass and morphological or behavioral changes of the According to the ISO 11268-2 (1998) guideline, test results
adult worms were recorded after 28 days and the number are valid in case the following criteria are fulfilled: adult
of juvenile earthworms after 56 days by a heat-extraction mortality B10%, number of juveniles [30 per test vessel,
method (water bath 60 °C). and the coefficient of variation B30%. These validity cri-
The bait-lamina were applied in parallel to the chronic teria were fulfilled in the controls of all tests with both test
earthworm tests. Each test vessel received two bait-lamina substances (Römbke et al. 2006a, 2007). In the case of the
strips (i.e. 32 holes) at test start that were transversally bait-lamina field test ISO 18311 (ISO 2016) requires that at

123
Euro-Mediterr J Environ Integr (2017) 2:5 Page 5 of 11 5

least 30% of the baits are pierced in the control vessels. between \181 and C2898 mg Zn(NO3)2/kg DW soil, i.e.
This criterion was fulfilled in all tests. differing at least by a factor of 16 apart. EC10 values
varied between 43.9 and C2898 mg Zn(NO3)2/kg DW soil,
Zinc-nitrate with an even higher factor between them (66). The EC50
values differed less, just by a factor of 9.7, covering a range
The test results for the reproduction of earthworms in zinc between 299 and [2898 mg Zn(NO3)2/kg DW soil. The
nitrate spiked soils are summarized in Table 2 (for further NOEC- and EC10-values were similar (i.e. at most factor
details see Scheffczyk et al. 2014). Depending on the soil two apart) except for soil SOE where the EC10 was by a
type the NOECs varied between 181 and 1449 mg factor of seven lower than the NOEC. When comparing the
Zn(NO3)2/kg DW soil, i.e. by a factor of eight. The EC10- NOEC-, EC10- and EC50-values only for the soils BWZ
values varied by a factor of 11, i.e. between 118 and and SOE (EC10 only) the former two were clearly lower
1289 mg Zn(NO3)2/kg DW soil. The NOEC- and EC10- than the latter. In all other cases, they were less than a
values were similar (i.e. at most by a factor of two apart). factor of two apart. For the GGI-soil the difference could
The EC50-values differed between 422 and 1903 mg not be determined because statistically significant effects
Zn(NO3)2/kg DW soil, i.e. there the factor between them is were already observed at the lowest test concentration of
only 4.5. When comparing the NOEC-, EC10-, and EC50- 181 mg Zn(NO3)2/kg soil DW. In SHA-soil even the
values determined in the nine tests performed, only in two NOEC was higher than the highest test concentration of
cases (BWZ, SHA) were both former values lower than the 2898 mg Zn(NO3)2/kg soil DW. The mean of the MDD
latter by a factor of two. The mean of the MDD was 29% was 33% and the CV was smallest for the LOEC (71%).
and the CV was smallest for the EC50 (38%). Based on the EC50 values (the most robust and thus
The test results for the feeding activity in soils spiked suitable endpoint), the experiences gained in this contri-
with zinc nitrate are summarized in Table 3. Due to tech- bution with zinc-nitrate are summarized as follows
nical reasons, no bait-lamina was inserted alongside the (Fig. 5). Results are classified as different as long as their
earthworm tests with zinc-nitrate in OECD artificial soil confidence intervals do not overlap. In six tests for which
and LUFA St. 2.2. In the remaining seven tests performed, toxicity values could be determined for both earthworm
it was found that the NOECs covered a very broad range, reproduction and feeding activity no difference in the

Table 2 Results of the

Soil EC10 EC50 (95% confidence limits) NOEC LOEC MDD (%)
earthworm reproduction test
(Römbke et al. 2006a); all OECD 1110 1731 (ND) 725a 1449a 26a
values in mg Zn(NO3)2/kg DW a a
soil and comparison with OECD BRG 1106 1361 (1346–1372) 725 1449 31a
a a
artificial soil (excluded from BWZ 118 422 (194–791) 181 362 29a
statistics) GGI 527 864 (844–885) 725 a
1449 a
40a
HAG 1090 1439 (ND) 725a 1449a 20a
a a
SBG 773 1903 (ND) 1449 2898 38a
b b
SHA 427 1297 (1130–1496) 362 725 21b
a a
SOE 1289 1804 (1788–1820) 1449 2898 24a
b b
LUFA St. 2.2 478 1083 (690–1727) 725 1449 31b
Statistics
Mean 726 1272 792 1585 29
Min 118 422 181 362 20
Max 1289 1903 1449 2898 40
Factor 10.9 4.5 8.0 8.0
SD 406 484 454 907
CV (%) 56 38 57 57
Geometric mean 593 1167 664 1329
EC10 10% effective concentration, EC50 median effective concentration, NOEC no observed effect
concentration, LOEC lowest observed effect concentration, MDD minimum detectable difference, ND not
determinable, SD standard deviation, CV coefficient of variation. See Table 1 for abbreviations and
locations of soils
a
Dunnett’s t test for homogenous variances
b
William’s t test for homogenous variances

123
5 Page 6 of 11 Euro-Mediterr J Environ Integr (2017) 2:5

Table 3 Results of the bait

Soil EC10 EC50 (95% confidence limits) NOEC LOEC MDD (%)
lamina test; all values in mg
Zn(NO3)2/kg DW soil and OECD NT NT NT NT NT
comparison with OECD
artificial soil (excluded from BRG 530 1113 (968–1281) 725a 1449a 38a
b b
statistics) BWZ 184 396 (352–445) 181 362 23b
a,c a
GGI 43.9 299 (ND) \181 181 25a
a a
HAG 817 1325 (1203–1460) 725 1449 32a
SBG 1412 2291 (2089–2511) 1449a 2898a 31a
a a
SHA [2898 [2898 (ND) C2898 [2898 40a
a a
SOE 103 711 (506–1000) 725 1449 39a
LUFA St. 2.2 NT NT NT NT NT
Statistics
Mean 855 1290 983 1527 33
Min 43.9 299 \181 181 23
Max [2898 [2898 C2898 [2898 40
Factor 66 9.7 16 16
SD 1022 978 946 1076
CV (%) 120 76 96 71
Geometric mean 394 973 656 1076
EC10 10% effective concentration, EC50 median effective concentration, NOEC no observed effect
concentration, LOEC lowest observed effect concentration, MDD minimum detectable difference, ND not
determinable, NT not tested, SD standard deviation, CV coefficient of variation. See Table 1 for abbrevi-
ations and locations of soils
a
Dunnett’s t test for homogenous variances
b
William’s t test for homogenous variances
c
Lowest concentration significantly different compared to the control; the lowest concentration was used
for the calculation of the mean values

Fig. 5 Comparison of the EC50 2600

(median effective
concentration) values for the 2400 Bait lamina
8-week earthworm reproduction 2200 Reproduction
EC50 [mg Zn(NO3)2/kg DW soil]

and the 1-week bait-lamina

earthworm tests in soils spiked 2000
with zinc nitrate
1800

1600

1400

1200

1000

800

600

400

200
n.t. n.d. n.t.
0
OECD BRG BWZ GGI HAG SBG SHA SOE St. 2.2

sensitivity of both test endpoints was found in two soils reproduction test was more sensitive. However, in only two
(BWZ, HAG). In three soils the BLET was more sensitive cases (GGI, SOE), did the EC50 values differ by more than
(BRG, GGI, SOE), and in one soil (SBG) the earthworm a factor of two.

123
Euro-Mediterr J Environ Integr (2017) 2:5 Page 7 of 11 5

TBT-O performed, it was found that the NOECs covered a range

between 0.3 and 3.2 mg TBT-O/kg DW soil, i.e. differing
Chronic endpoints for effects on the reproduction of by a factor of 10. EC10 values varied only slightly more,
earthworms in TBT-O spiked soils are summarized in i.e. 0.3 and 5.4, with a factor of 18 between them. The
Table 4 (for further details see Scheffczyk et al. 2014). EC50 values differed less, just by a factor of 6.1, covering
Depending on the soil type the NOECs varied between a range between 1.4 and 8.5 mg TBT-O/kg DW soil. In the
\0.3 and 1.0 mg TBT-O/kg DW soil, i.e. by a factor of nine tests performed, the NOEC- and EC10-values were
C3.3. The EC10-values varied by a factor of 83.3, i.e. similar (i.e. at most by a factor of two apart). Comparing
between 0.03 and 2.5 mg TBT-O/kg DW soil. The EC50- the NOEC-, EC10- and EC50-values in only three cases
values differed between 0.5 and 4.7 mg TBT-O/kg DW (OECD, GGI, SHA) the former two were clearly lower
soil, i.e. the factor between them is only 9.4. When com- than the latter. In all other soils (BRG, BWZ, HAG, SBG,
paring the NOEC-, EC10- and EC50-values determined in SOE, LUFA St. 2.2) the EC50 was less by a factor of two
the nine tests performed, in almost all cases (except HAG, higher than the corresponding EC10 and/or NOEC. In the
SHA and SOE), both former values were by a factor of two soil HAG, the EC50 was even slightly below the corre-
lower than the latter. NOEC- and EC10-values were at sponding NOEC due to the high MDD of 43%. For the
most a factor of two apart in five soils (OECD, BWZ, GGI, SHA-soil the difference could not be determined because
SBG, LUFA St. 2.2). In four soils (BRG, HAG, SHA, statistically significant effects were already observed at the
SOE) the EC10 was more than a factor of two higher than lowest test concentration of 0.3 mg TBT-O/kg soil DW.
the NOEC. In the test in GGI soil the NOEC was lower The mean of the MDD was 34% and the CV (%) was
than the lowest concentration tested. The mean of the similar for all endpoints (47–54%).
MDD was 19% and the CV was smallest for the EC50 Again, the EC50 values are used as basis for the
(45%). comparison between the results of the earthworm repro-
The test results for the feeding activity in soils spiked duction tests and in the BLET with TBT-O (Fig. 6).
with TBT-O are summarized in Table 5. In the nine tests Results are classified as different as long as their

Table 4 Results of the

Soil EC10 EC50 (95% confidence limits) NOEC LOEC MDD (%)
earthworm reproduction test
(Römbke et al. 2007); all values OECD 3.9 13.4 (10.1–17.8) 3.2 b
10.0 b
20b
in mg TBT-O/kg DW soil and b b
comparison with OECD BRG 2.1 4.7 (3.0–6.5) 1.0 3.2 21b
artificial soil (excluded from BWZ 0.3 2.0 (1.3–3.0) 0.3b 1.0b 26b
statistics) GGI 0.03 0.5 (0.01–1.3) \0.3 a,c
0.3 a
34a
a a
HAG 2.4 4.1 (ND) 1.0 3.2 16a
SBG 0.6 2.5 (1.5–4.0) 0.3b 1.0b 8b
SHA 2.4 3.9 (ND) 1.0b 3.2b 20b
SOE 2.5 4.1 (ND) 1.0b 3.2b 17b
b b
LUFA St. 2.2 1.0 3.1 (2.9–3.1) 1.0 3.2 11b
Statistics
Mean 1.4 3.1 0.74 2.3 19
Min 0.03 0.5 \0.3 0.3 8
Max 2.5 4.7 1.0 3.2 34
Factor 83.3 9.4 3.3 10.7
SD 1.04 1.4 0.36 1.3
CV (%) 73 45 49 57
Geometric mean 0.8 2.6 0.64 1.8
EC10 10% effective concentration, EC50 median effective concentration, NOEC no observed effect
concentration, LOEC lowest observed effect concentration, MDD minimum detectable difference, ND not
determinable, SD standard deviation, CV coefficient of variation. See Table 1 for abbreviations and
locations of soils
a
Dunnett’s t test for homogenous variances
b
William’s t test for homogenous variances
c
Lowest concentration significantly different compared to the control; the lowest concentration was used
for the calculation of the mean values

123
5 Page 8 of 11 Euro-Mediterr J Environ Integr (2017) 2:5

Table 5 Results of the bait

Soil EC10 EC50 (95% confidence limits) NOEC LOEC MDD (%)
lamina test; all values in mg
TBT-O/kg DW soil and OECD 5.3 16.0 (13.4–19.2) 3.2a 10.0a 13a
comparison with OECD b b
artificial soil (excluded from BRG 5.4 6.9 (ND) 3.2 10.0 33b
b b
statistics) BWZ 4.5 8.5 (ND) 3.2 10.0 28b
b b
GGI 3.0 7.9 (5.2–12.0) 3.2 10.0 35b
b b
HAG 2.1 3.1 (2.5–3.9) 3.2 10.0 43b
SBG 2.7 3.7 (2.5–5.4) 3.2b 10.0b 45b
b,c b
SHA 0.3 1.4 (1.0–1.9) \0.3 0.3 22b
b b
SOE 2.0 3.4 (3.0–3.9) 1.0 3.2 29b
b b
LUFA St. 2.2 4.1 7.4 (5.9–9.2) 3.2 10.0 36b
Statistics
Mean 3.0 5.3 2.6 7.9 34
Min 0.3 1.4 \0.3 0.3 22
Max 5.4 8.5 3.2 10.0 45
Factor 18 6.1 10 33
SD 1.6 2.7 1.2 3.9
CV (%) 54 51 47 49
Geometric mean 2.4 4.6 2.1 5.6
EC10 10% effective concentration, EC50 median effective concentration, NOEC no observed effect
concentration, LOEC lowest observed effect concentration, MDD minimum detectable difference, ND not
determinable, SD standard deviation, CV coefficient of variation. See Table 1 for abbreviations and
locations of soils
a
William’s t test for homogenous variances
b
Dunnett’s t test for homogenous variances
c
Lowest concentration significantly different compared to the control; the lowest concentration was used
for the calculation of the mean values

Fig. 6 Comparison of the EC50 20

(median effective
concentration) values for the 18 Bait lamina
8-week earthworm reproduction Reproduction
and 1-week bait-lamina 16
EC50 [mg TBT-O/kg DW soil]

earthworm tests in soils spiked

with tributyltin-oxide 14

12

10

0
OECD BRG BWZ GGI HAG SBG SHA SOE St. 2.2

confidence intervals do not overlap. Of the nine compa- and in four soils (BRG, BWZ, GGI, LUFA St. 2.2) the
rable tests, no difference in the sensitivity of both test earthworm reproduction test was more sensitive. In four
endpoints was found in four soils (OECD, HAG, SBG, cases (BWZ, GGI, SHA, LUFA St. 2.2), the EC50 values
SOE). In one soil the BLET was more sensitive (SHA), differed by more than a factor of two.

123
Euro-Mediterr J Environ Integr (2017) 2:5 Page 9 of 11 5

Discussion most important when tested in a range of artificial soils

with different pH values and OM contents (Spurgeon and
Test performance and validity criteria Hopkin 1996). In the earthworm tests presented here, the
correlation between zinc toxicity and soil factors like pH,
No technical difficulties were observed when performing OC content and CEC was not significant, which might be
the tests (e.g. earthworms were not stressed by the inclu- caused by the fact that the tested soils had very different
sion of the strips). Methodologically, the test design chosen properties, while in the work cited above only artificial
here was determined by the requirements of the earthworm soils with different adjusted pH values and OM contents
reproduction test according to the ISO guideline 11268-2 were tested (Römbke et al. 2006a). Therefore, it is not clear
(ISO 1998). A corresponding guideline was since also whether the availability of zinc indirectly played a role in
published as OECD guideline 222 (OECD 2004). the effects of zinc on the feeding activity.
Regarding validity, all criteria as defined in these guideli- Compared to heavy metals, few studies assessing the
nes have been fulfilled. The same is true for the BLET as influence of soil properties on the toxicity of organic
originally defined for field investigations according to ISO chemicals to soil invertebrates have been performed (e.g.
guideline 18131 (ISO 2016). Van Gestel 1992). As the pore water concentration of non-
ionizable substances depends on the organic matter content
Comparison of test sensitivity of the soil, effects on organisms should decrease with
increasing organic matter content of the soil (Lock et al.
In the context of the application of the BLET as a screening 2002). While there are examples that this relationship is not
tool, its sensitivity compared to the more complex and always true (e.g. for Lindane (log Kow = 3.85; Lock et al.
longer lasting earthworm reproduction test needs to be 2002), in the case of TBT-O (log Kow = 3.74) a clear and
discussed. There is no information in the scientific litera- significant correlation between EC50 values (i.e. a chronic
ture since the BLET has not been used in laboratory eco- endpoint) and organic matter content could be shown
toxicological tests so far. Here, the two EC50 values were (Römbke et al. 2007) for earthworms. When discussing the
considered to be different in case the respective 95% modeling of the uptake of polycyclic aromated hydrocar-
confidence limits did not overlap. bons (PAH) in soil, Jager (2003) also mentions that sorp-
So far, no distinct difference regarding sensitivity of one tion of ‘‘difficult’’ substances like (organo-)metals in soils
of the two endpoints (reproduction and feeding activity) cannot be predicted on the basis of organic matter content
has been identified. Out of 15 comparisons made in this alone, meaning that more measurements with a broad range
study, in six cases there was no difference. In four cases the of soils and, probably, more species are necessary.
feeding rate was more sensitive and in five cases it was
earthworm reproduction. In fact, the two values did only Test improvement
once differ by a factor of more than five; in almost all
comparisons they differed by a factor of less than three. In case the BLET is to be used as a stand-alone laboratory
Therefore, the use of the BLET as a screening tool and/or trial, several parameters of the test design may be adjusted,
range-finding test can be recommended since comparable e.g. in order to reduce variability or increase sensitivity and
effects can already be detected after 1 week (chronic statistical power. In particular, these parameters are
earthworm test: 8 weeks). However, since the results may – Test duration: what is the minimum test duration to
depend on the respective mode-of-action of the test achieve a sufficient control feeding activity and what
chemicals, a more profound assessment of the BLET is test duration is too long so that differences in feeding
needed on the basis of further testing experience with more activity are no longer detected or the advantage of a
substances belonging to different classes. short-term screening method gets lost? In the field, it is
recommended to run the test until at least 30% but
Influence of soil properties preferably not all bait slits are pierced. For example, at
tropical sites with an average air temperature of about
Soil properties clearly influence the outcome of ecotoxi- 26 °C a test duration of 1 week was chosen (Römbke
cological tests with soil organisms. This is either because et al. 2006b). This test duration was also sufficient in
they have a direct effect on the animals and their activities our laboratory tests which were performed at a
(e.g. Ma and Bonten 2011), or (probably even more) due to temperature of 18–22 °C.
their impact on the bioavailability of soil contaminants. In – Addition of food: before starting the test, it was
the case of zinc, this relationship has been substantiated: discussed whether—in contrast to the standard earth-
For the earthworm Eisenia fetida, pH and OM content are worm reproduction test—no additional food besides the

123
5 Page 10 of 11 Euro-Mediterr J Environ Integr (2017) 2:5

bait-lamina substrate should be incorporated into the control (i.e. reference substance) needs to be identified that
soil to avoid distracting the worms from feeding on causes a defined reduction of the feeding activity under
them. However, the results gained here show that standardized test conditions. For the time being, the mini-
despite the availability of food the earthworms fed on mum control feeding activity of 30% as defined for the
the bait-lamina. This is true in tests in which the field bait-lamina test could be used. Additionally, a maxi-
relationship between the endpoint feeding rate and mum permissible coefficient of variation for feeding
earthworm reproduction is studied, but in short-term activity in controls may be a useful validity criterion, but
tests feeding is not recommended to get clearer results. before this could be defined a meta-analysis of bait-lamina
– Number of animals and bait strips per test vessel/size of results is needed.
test vessels: it may be useful to expose the earthworms
individually or in smaller groups than 10 and use
smaller test vessels with less soil and possibly one only
bait strips. This would further reduce testing demands Conclusions
and might even reduce variability and increase statis-
tical power since the number of replicates may in turn The results presented here can be summarized as follows:
be increased.
– It is possible to use the results of a BLET to predict the
range of effects of a test chemical on earthworm
Regular test use reproduction, i.e. to use the BLET as a range-finding
test. Further tests (running with both endpoints for one
When using the BLET for the identification of appropriate and 8 weeks, respectively) with more chemicals and
concentrations in tests with chemicals, a problem could different mode-of-actions are needed to confirm the
occur in case the respective chemical does affect the relationship between both endpoints.
reproduction and the feeding activity of the earthworms at – The BLET, running for just a week, can be used for the
very different concentrations. However, this possible quick determination of the effects of potentially
discrepancy is mitigated by the fact that the 1-week BLET contaminated field soils on an important functional
would be designed as a range-finding test. When referring endpoint.
to standard guidelines, such tests are usually performed at
In case this test is going to be used in other regions than
concentrations with a spacing factor of ten (e.g. OECD
Central Europe, further testing with regional soils is needed
2004). The effect values (i.e. the EC50 values) found in
to provide a robust basis for the evaluation and applica-
this study for the two different endpoints were with one
bility of the results. The BLET is sensitive, delivers fast
exception (far) less than a factor of 10 apart, meaning that
results and requires little resources and training. Although
the effects observed in the BLET would mostly still result
some effort has to be put in getting more experience with
in a suitable range of concentrations for the earthworm
the BLET, in the future it might be used as an additional
reproduction test. However, in order to assure that the
method in soil ecotoxicology.
range-finding results of the BLET are robust, further
In any case, the BLET that has a very good ratio
studies with chemicals with different mode-of-actions
between practical efforts and meaningful results would be a
have to be performed.
good addition for ‘‘on-site’’ test batteries (i.e. short-lasting
When using BLET for the quick determination of the
and relatively simple test methods). In addition, the BLET
effects of potentially contaminated field soils it is clear that
would allow assessing the effects of a chemical on a soil
the endpoint feeding rate could be influenced by the
function related to nutrient cycling, i.e. an extremely
respective soil properties. Although so far only soils from
important function provided and/or in the field regulated
Germany have been used, this influence has probably been
partly by earthworms (e.g. Brussaard 2012; Keith and
well covered for Central Europe considering the quite large
Robinson 2012).
differences in soil properties such as pH, organic carbon or
clay content. However, this might not be true for other Acknowledgements This research was part of the ‘‘ERNTE’’-project
regions with very different soil properties—a difference funded by the German Federal Ministry for Education and Research
which has been highlighted in various ecotoxicological (R&D No. 0330300).
investigations, e.g. in the Mediterranean regions (Chelinho
Compliance with ethical standards
et al. 2011; Madani et al. 2015).
Regarding a standardization of this method the current Conflict of interest On behalf of all authors, the corresponding
lack of several validity criteria is problematic. A positive author states that there is no conflict of interest.

123
Euro-Mediterr J Environ Integr (2017) 2:5 Page 11 of 11 5

References Ma W-C, Bonten LTC (2011) Bioavailability pathways underlying

zinc-induced avoidance behavior and reproduction toxicity in
Brussaard L (2012) Ecosystem services provided by the soil biota. In: Lumbricus rubellus earthworms. Ecotoxicol Environ Saf
Wall DH et al (eds) Soil Ecology and Ecosystem Services, 1st 74:1721–1726
Madani S, Coors A, Haddioui A, Ksibi M, Pereira R, Sousa JP,
edn. Oxford University Press, Oxford, pp 45–58
Chelinho S, Domene X, Campana P, Natal-da-Luz T, Scheffczyk A, Römbke J (2015) Effects of contaminated soils from a former
Römbke J, Andrés P, Sousa JP (2011) Improving ecological risk iron mine (Ait Amar, Morocco) on enchytraeids and predatory
assessment in the mediterranean area: selection of reference soils mites in standard laboratory tests. Ecotoxicol Environ Saf
119:190–197
and evaluating the influence of soil properties on avoidance and
reproduction of two oligochaete species. Environ Toxicol Chem Moore DW, Dillon TM, Suedel BC (1991) Chronic toxicity of
30:1050–1058 tributyltin to the marine polychaete worm, Neanthes arenaceo-
Dominguez J, Velando A, Aira M, Monroy F (2003) Uniparental dentata. Aquat Toxicol 21:181–198
reproduction of Eisenia fetida and E. andrei (Oligochaeta: OECD (Organization for Economic Co-operation and Development)
Lumbricidae): evidence of self-insemination. Pedobiol (1984) OECD guideline for testing of chemicals No. 207.
47:530–534 Earthworm acute toxicity test. Paris
EC (Environment Canada) (2007) Guidance document on statistical OECD (Organization for Economic Co-operation and Development)
methods for environmental toxicity tests. Report EPS 1/RM/46 (2004) Guidelines for the testing of chemicals No. 222.
Edwards CA, Lofty JR (1977) Biology of earthworms, 2nd edn. Earthworm reproduction test (Eisenia fetida/Eisenia andrei).
Chapman and Hall, London Paris
Römbke J, Jänsch S, Junker T, Pohl B, Scheffczyk A, Schallnaß H-J
Federschmidt A, Römbke J (1994) Erfahrungen mit dem Köder-
streifen-Test auf zwei fungizidbelasteten Standorten. Braunschw (2006a) Improvement of the applicability of ecotoxicological
Naturkdl Schr 4:675–680 tests with earthworms, springtails and plants for the assessment
Förster B, Van Gestel CAM, Koolhaas JE, Nentwig G, Rodrigues of metals in natural soils. Environ Toxicol Chem 25:776–787
JML, Sousa JP, Jones SE, Knacker T (2004) Ring-testing and Römbke J, Höfer H, Garcia MVB, Martius C (2006b) Feeding
field-validation of a terrestrial model ecosystem (TME)— an activities of soil organisms at four different forest sites in
instrument for testing potentially harmful substances: effects of Amazonia using the bait-lamina method. J Trop Ecol
carbendazim on organic matter breakdown and soil fauna 22:313–320
feeding activity. Ecotoxicology 13:129–141 Römbke J, Jänsch S, Junker T, Pohl B, Scheffczyk A, Schallnaß H-J
Giusti A, Barsi A, Dugué M, Collinet M, Thomé JP, Joaquim-Justo C, (2007) The effect of tributyltin-oxide on earthworms, collem-
Roig B, Lagadic L, Ducrot V (2013) Reproductive impacts of bolans and plants in artificial and natural soils. Arch Environ
Contam Toxicol 52:525–534
tributyltin (TBT) and triphenyltin (TPT) in the hermaphroditic
freshwater gastropod Lymnaea stagnalis. Environ Toxicol Scheffczyk A, Frankenbach S, Jänsch S, Römbke J (2014) Compar-
Chem. 32:1552–1560 ison of the effects of zinc nitrate and tributyltin-oxide on the
ISO (International Organization for Standardization) (1998) Soil reproduction and avoidance behavior of the earthworm Eisenia
andrei in laboratory tests using ten soils. Appl Soil Ecol
quality—effects of pollutants on earthworms (Eisenia fetida).
Part 2: determination of effects on reproduction. ISO 11268-2. 83:253–257
Geneva Schrader S, Zhang H (1997) Earthworm casting: stabilisation or
ISO (International Organization for Standardization) (2016) Soil destabilisation of soil structure. Soil Biol Biochem 29:469–475
quality—method for testing effects of soil contaminants on the Simpson JE, Slade E, Riutta T, Taylor ME (2012) Factors affecting
feeding activity of soil dwelling organisms—bait-lamina test. soil fauna feeding activity in a fragmented lowland temperate
ISO 18311. Geneva deciduous woodland. PLoS One 7(1):e29616. doi:10.1371/
Jager T (2003) Worming your way into bioavailability. Modelling the journal.pone.0029616
uptake of organic chemicals in earthworms. Dissertation, Spurgeon DJ, Hopkin SP (1996) Effects of variations of the organic
University of Utrecht matter content and pH of soils on the availability and toxicity of
Jessen-Hesse V, Römbke J, Jänsch S, Hund-Rinke K, Terytze K zinc to the earthworm Eisenia fetida. Pedobiologia 40:80–96
ToxRatÒ (2003) Software for the statistical analysis of biotests.
(2005) Auswahl und Charakterisierung von Böden zur Opti-
mierung der Anwendung ökotoxikologischer Testmethoden bei Copyright: ToxRat Solutions GmbH, Alsdorf
der Beurteilung potentiell belasteter Standorte. Bodenschutz Van Gestel CAM (1992) Toxicity of chemicals for earthworms: a
4(05):110–116 review. In: Greig-Smith P, Becker H, Edwards PJ, Heimbach F
Jones CG, Lawton JH, Shachak M (1994) Organisms as ecosystem (eds) Ecotoxicology of earthworms. Intercept, Andover,
engineers. Oikos 69:373–386 pp 44–54
Kampmann T (1994) Entwicklung eines standardisierten Labortests Van Gestel CAM, Van der Warde JJ, Derksen A, Van der Hoek EE,
mit Köderstreifen für ökotoxikologische Prüfungen: erste Veul MFXW, Bouwens S, Rusch B, Kronenburg R, Stokamn
Vorversuche. Braunschw Naturkdl Schr 4:681–686 GNM (2001) The use of acute and chronic bioassays to
Keith AM, Robinson DA (2012) Earthworms as natural capital: determine the ecological risk and bioremediation efficiency of
ecosystem service providers in agricultural soils. Economology J oil-polluted soils. Environ Toxicol Chem 20:1438–1449
Van Groenigen JW, Lubbers IM, Vos HMJ, Brown GG, De Deyn GB,
II(Year II):91–99
Kratz W (1998) The bait-lamina test—general aspects, applications Van Groenigen KJ (2014) Earthworms increase plant produc-
and perspectives. Environ Sci Pollut Res 5:94–96 tion: a meta-analysis. Nat Sci Rep 4:6365
Lock K, De Schamphelaere KC, Janssen CR (2002) The effect of Von Törne E (1990) Assessing feeding activities of soil-living
animals. I. Bait-lamina tests. Pedobiologia 34:89–101
lindane on terrestrial invertebrates. Arch Environ Contam
Toxicol 42:217–221

123