Acute toxicity of organic chemicals to Gammarus pulex correlates with sensitivity of Daphnia magna across most modes of action.

Abstract

We investigated the sensitivity of the freshwater crustacean amphipod Gammarus pulex towards organic xenobiotic compounds in comparison to the sensitivity of the crustacean cladoceran Daphnia magna. In addition we studied the influence of the chemical's mode of action on the relationship between the sensitivity of G. pulex and that of D. magna. We tested the acute toxicity of twelve compounds (Malathion, Aldicarb, Carbofuran, 2,4-dichloroaniline, 2,4-dichlorophenol, 1,2,3-trichlorobenzene, 4,6-dinitro-o-cresol, 2,4,5-trichlorophenol, Ethylacrylate, 4-nitrobenzyl-chloride, Sea-nine, Imidacloprid) with different modes of action and physicochemical properties towards the freshwater amphipod G. pulex in laboratory experiments. Additional toxicity data was collected from the peer-reviewed literature and databases (data pairs for 44 chemicals in total). The chemicals were assigned to seven mode of action groups. The relationship between the sensitivity of G. pulex (48h-LC50s and 96h-LC50s) and that of D. magna (48h-EC50s) was investigated using regression analysis and correlation plots. G. pulex is two to three orders of magnitude more sensitive towards neonicotinoids than D. magna (P=0.0046, n=3). For organophosphates we found that D. magna is more sensitive than G. pulex by approximately a factor of six (P=0.0256, n=6). There was no significant difference between the sensitivity of D. magna and that of G. pulex in any of the other mode of action groups; however chemicals with the same mode of action grouped together in the same area of the correlation plot. Without the neonicotinoids 75% of all G. pulex toxicity data were within one order of magnitude of the D. magna data and 100% within two orders of magnitude. The regressions with all data and with all data minus neonicotinoids were both significant linear relationships with slopes around one and intercept around zero. Thus, G. pulex is generally equally sensitive towards organic xenobiotics as D. magna.

Full text

Page 1

Aquatic Toxicology 103 (2011) 38–45
Contents lists available at ScienceDirect
Aquatic Toxicology
journal homepage: www.elsevier.com/locate/aquatox
Acute toxicity of organic chemicals to Gammarus pulex correlates with sensitivity
of Daphnia magna across most modes of action
Roman Ashauera,b,∗, Anita Hintermeistera,b, Eva Potthoffa, Beate I. Eschera,b
aEawag, Swiss Federal Institute of Aquatic Science and Technology, 8600 Dübendorf, Switzerland
bThe University of Queensland, National Research Centre for Environmental Toxicology (Entox), 39 Kessels Rd, Brisbane, Qld 4108, Australia
a r t i c l ei n f o
Article history:
Received 2 December 2010
Received in revised form 1 February 2011
Accepted 8 February 2011
Keywords:
Interspecies correlation estimation
Species sensitivity distribution
Chemical stress
Risk assessment
Aquatic invertebrate
Pesticide
a b s t r a c t
We investigated the sensitivity of the freshwater crustacean amphipod Gammarus pulex towards organic
xenobiotic compounds in comparison to the sensitivity of the crustacean cladoceran Daphnia magna.
In addition we studied the influence of the chemical’s mode of action on the relationship between
the sensitivity of G. pulex and that of D. magna. We tested the acute toxicity of twelve compounds
(Malathion, Aldicarb, Carbofuran, 2,4-dichloroaniline, 2,4-dichlorophenol, 1,2,3-trichlorobenzene, 4,6-
dinitro-o-cresol, 2,4,5-trichlorophenol, Ethylacrylate, 4-nitrobenzyl-chloride, Sea-nine, Imidacloprid)
with different modes of action and physicochemical properties towards the freshwater amphipod G.
pulex in laboratory experiments. Additional toxicity data was collected from the peer-reviewed litera-
ture and databases (data pairs for 44 chemicals in total). The chemicals were assigned to seven mode of
action groups. The relationship between the sensitivity of G. pulex (48h-LC50s and 96h-LC50s) and that
of D. magna (48h-EC50s) was investigated using regression analysis and correlation plots.
G. pulex is two to three orders of magnitude more sensitive towards neonicotinoids than D. magna
(P=0.0046, n=3). For organophosphates we found that D. magna is more sensitive than G. pulex by
approximately a factor of six (P=0.0256, n=6). There was no significant difference between the sen-
sitivity of D. magna and that of G. pulex in any of the other mode of action groups; however chemicals
with the same mode of action grouped together in the same area of the correlation plot.
Without the neonicotinoids 75% of all G. pulex toxicity data were within one order of magnitude of the
D. magna data and 100% within two orders of magnitude. The regressions with all data and with all data
minus neonicotinoids were both significant linear relationships with slopes around one and intercept
around zero. Thus, G. pulex is generally equally sensitive towards organic xenobiotics as D. magna.
© 2011 Elsevier B.V. All rights reserved.
1. Introduction
1.1. Background
Environmental risk assessment of chemicals is generally based
on toxicity tests with standard test organisms (van Leeuwen and
Vermeire, 2007). One of these is the crustacean cladoceran Daph-
nia magna, which naturally occurs in lentic freshwater systems. In
lower tiers of risk assessments of chemicals toxicity data on D.
magna is combined with assessment factors to account, amongst
other factors, for interspecies variation (Chapman et al., 1998).
Thus, it is of interest how the sensitivity of D. magna and Gammarus
pulex, a non-standard test organism, varies over a wide range of
∗Corresponding author at: Swiss Federal Institute of Aquatic Science and Tech-
nology, Überlandstrasse 133, 8600 Dubendorf, Switzerland. Tel.: +41 448235233;
fax: +41 448235311.
E-mail address: roman.ashauer@eawag.ch (R. Ashauer).
chemicals. The freshwater crustacean amphipod G. pulex, naturally
occurring in lotic water bodies, has also been widely used in toxic-
itytesting(Kunzetal.,2010)andplaysanimportantroleindetritus
processing in streams (Maltby et al., 2002). More recently we have
usedG.pulexfordevelopingmodelsthatcanbeusedforriskassess-
ment of fluctuating concentrations of chemicals (Ashauer et al.,
2007a,b, 2010). Fluctuating concentrations and repeated pulses
of pollutants are more likely encountered by long-lived stream
dwelling organisms such as G. pulex compared to organisms that
live in ponds and lakes and have a shorter life such as D. magna.
For the integration of modeling approaches based on G. pulex into
current risk assessment schemes it is necessary to know if there
are systematic differences in sensitivity compared to the standard
testorganismofcurrentriskassessmentpractice.Consequentlythe
question has arisen whether G. pulex generally tends to be more,
less or equally sensitive to organic xenobiotic toxicants compared
to the closely related standard test organism D. magna.
Thesensitivityofaquaticorganismsandtheirrelationshipshave
been investigated by ranking species (Wogram and Liess, 2001;
0166-445X/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.aquatox.2011.02.002

Page 2

R. Ashauer et al. / Aquatic Toxicology 103 (2011) 38–45
39
Rubach et al., 2010) or with regression models (Dyer et al., 2006;
Raimondo et al., 2007, 2010). Both, the ranking approach (Rubach
et al., 2010) as well as the regression modeling studies (Raimondo
et al., 2010) have found a dependence of species sensitivity rela-
tionships on modes of toxic action.
1.2. Objective
The objective of this study was to investigate the sensitivity
of the freshwater crustacean amphipod G. pulex towards organic
xenobiotic compounds in comparison to the sensitivity of the crus-
tacean cladoceran D. magna. In addition we studied the influence
of the chemical’s mode of action on the relationship between the
sensitivity of G. pulex and that of D. magna.
1.3. Study outline
We testedtheacute
Carbofuran,
1,2,3-trichlorobenzene,
Ethylacrylate,
toxicity of
2,4-dichloroaniline,
4,6-dinitro-o-cresol,
4-nitrobenzyl-chloride,
twelve compounds
(Malathion,
dichlorophenol,
2,4,5-trichlorophenol,
Sea-nine, Imidacloprid) with different modes of action and
physicochemical properties towards the freshwater amphipod
G. pulex in laboratory experiments. Additional toxicity data for
organic chemicals was collected from peer-reviewed literature
and corresponding toxicity data for D. magna were collected from
databases and peer-reviewed literature. The relationship between
the sensitivity of G. pulex (48h-LC50s and 96h-LC50s) and that of
D. magna (48h-EC50s) was investigated using regression analysis
and correlation plots.
Aldicarb,2,4-
2. Material and methods
2.1. Acute toxicity tests with G. pulex
Adult G. pulex were collected several times during 2008/2009
from a small headwater stream in the Itziker Ried (coordinates: E
702150, N 2360850), ca. 20km southeast of Zürich, Switzerland.
After collection, the test organisms were acclimatized for a mini-
mum of five days to the experimental conditions (13◦C, 12h:12h
light:dark). Experiments were carried out in Pyrex®beakers, each
containing 500mL of pre-aerated artificial pond water (APW,
Naylor et al., 1989). Each beaker contained ten organisms at the
start of the experiment and they were fed ad libitum with a
minimum of three horse-chestnut leaf discs (diameter 20mm)
inoculated with the fungi Cladosporium herbarum (Naylor et al.,
1989).
A geometric dilution series of seven concentrations was used
in each toxicity test, with two replicate beakers per concentra-
tion (i.e. 20 organisms per concentration) and one solvent control
beaker and one blank control beaker. Dosing stocks were made by
dilution in acetone from a mixture of14C-labelled and unlabelled
materialexceptforMalathion,AldicarbandCarbofuranwhereonly
14C-labelled material was used. After spiking the dosing solution
directlyintothetestmedium,beakerswerecarefullystirred,sealed
withparafilmandkeptat13◦Cundera12h:12hlight:darkregime.
Live/dead organisms were counted after 24, 48, 72 and 96h
by gently prodding and observation of movement of appendages.
Organisms were counted as dead if none of the appendices were
moving after prodding for three times. Dead organisms were
removed. At the same times 1mL of the test solution was sampled
from each beaker and chemical concentrations were quantified
using liquid scintillation counting. Exposure concentrations were
measured for all compounds, except for Aldicarb and Carbofuran,
where nominal concentrations were used.
Table 1 lists the tested chemicals. The supporting information
contains more details on collection dates, purities and14C-label
position of chemicals, sample processing and quantification of
radioactivity.
A log-logistic dose-response model with variable slope was fit-
ted to the survival data with GraphPad Prism (v. 4.03, GraphPad
Software Inc., USA) using the averages of the measured exposure
concentrations from the different sampling times for each treat-
ment. The parameters top and bottom were fixed to 100% and 0%.
2.2. Collection of additional data from literature
Additional LC50 data for G. pulex were collected from the
peer-reviewed literature (Table 2). Corresponding 48h-EC50 data
for D. magna was collected from the FOOTPRINT database
(http://www.eu-footprint.org/ppdb.html) where available. The
FOOTPRINT database is based on quality controlled and reviewed
datausedforregistrationofpesticides(FOOTPRINT,2009)andcon-
tains data of the highest quality available. The remaining toxicity
data for D. magna were collected from the US-EPA ECOTOX AQUIRE
database (http://cfpub.epa.gov/ecotox/), where the median was
used in case of multiple entries, and from the peer-reviewed liter-
ature. Thus we collected all data pairs for acute toxicity of organic
xenobiotics for these two species that were available through our
study, peer-reviewed literature and the two electronic databases
(FOOTPRINT and ECOTOX AQUIRE).
2.3. Data analysis
All toxicity values were log transformed to obtain normally
distributed data and then analyzed with linear regression and by
means of correlation plots to investigate the relationship between
sensitivity of D. magna and G. pulex. As the toxicity data for G.
pulex consists of mainly 48h-LC50 and 96h-LC50 values, the data
with different test durations were first analyzed separately. The
regression of the 48h-EC50 D. magna vs. the 48h-LC50 G. pulex
(n=27) was compared with that of the 48h-EC50 D. magna vs. the
96h-LC50 G. pulex (n=34). As the slopes and intercepts of these
two regressions did not differ significantly (see Supplementary
information) all the G. pulex toxicity data with different test dura-
tions were pooled and analyzed together. We also added two data
pairs where only 24h-LC50s were available. The total sample size
is 44 data pairs (Table 2).
Each chemical was assigned a mode of action based on
its molecular structure and its classification in the FOOTPRINT
database (FOOTPRINT, 2009). Seven modes of action were assigned
(Table 2): baseline toxicity, pyrethroids, organophosphates, car-
bamates, neonicotinoids, effects on nervous system (other than
organophosphates, carbamates, neonicotinoids and pyrethroids),
effects on energy production (which combine uncoupling and inhi-
bition of energy transduction). Not all chemicals were assigned a
mode of action and only modes of action with at least three data
pairs were further analyzed.
Model II least squares regressions (Deming regression) were
applied, as there is random error in both variables and were fitted
to log transformed data using GraphPad Prism (v. 4.03, GraphPad
Software Inc., USA). Correlation plots of all data and for each mode
of action group were used to identify mode of actions, which dif-
fer from the whole dataset in that either G. pulex or D. magna are
clearly more sensitive than the other.
3. Results
3.1. Acute toxicity data for G. pulex
The acute toxicity data measured for the twelve compounds in
thisstudyandtheparametersofthefitteddose-responsemodelare

Page 3

40
R. Ashauer et al. / Aquatic Toxicology 103 (2011) 38–45
Table 1
Acute toxicity measured in this study.
CompoundParametera
24h 48h 72h96h
1,2,3-TrichlorobenzeneLC50 (nmol/L)
95% conf. interv. of LC50
Slope
4618
4434–4808
−25.27
>2930
4748
4633–4867
−64.89
2333
1954–2787
−4.33
52571
51016–54173
−17.18
15193
13997–16491
−6.61
1805
1610–2024
−6.19
25325
22392–28642
−5.60
1558
1267–1916
−2.90
27
25–30
−3.21
55519
47927–64312
−4.50
430
279–664
−1.05
3.45
3.15–3.78
−3.30
97
86–111
−5.20
4462
4349–4578
−19.26
1361
1217–1521
−3.57
46998
46186–47824
−92.18
12722
11595–13959
−4.37
1196
1109–1290
−4.73
15753
14612–16984
−4.24
1266
1087–1474
−2.69
20
17–24
−3.05
30821
27985–33944
−3.38
405
225–729
−1.08
1.68
1.46–1.95
−2.54
61
5–66
−4.94
4462
4349–4578
−19.26
836
728–960
−2.95
47189
45141–49329
−30.47
11186
9816–12747
−3.23
835
772–904
−4.06
10706
9947–11524
−3.99
1208
1059–1378
−2.56
17
15–19
−4.22
18607
16852–20544
−2.94
514
298–888
−1.22
1.01
0.93–1.09
−5.14
45
42–48
−4.61
2,4,5-TrichlorophenolLC50 (nmol/L)
95% conf. interv. of LC50
Slope
2,4-DichloroanilineLC50 (nmol/L)
95% conf. interv. of LC50
Slope
57896
53009–63234
−7.87
22112
18541–26370
−5.16
>2132
2,4-DichlorophenolLC50 (nmol/L)
95% conf. interv. of LC50
Slope
4,6-Dinitro-o-cresolLC50 (nmol/L)
95% conf. interv. of LC50
Slope
4-Nitrobenzyl-chloride LC50 (nmol/L)
95% conf. interv. of LC50
Slope
>30111
Aldicarbb
LC50 (nmol/L)
95% conf. interv. of LC50
Slope
3461
2937–4079
−3.646
82
49–140
−1.94
>70248
Carbofuranb
LC50 (nmol/L)
95% conf. interv. of LC50
Slope
Ethylacrylate LC50 (nmol/L)
95% conf. interv. of LC50
Slope
Imidaclopridc
LC50 (nmol/L)
95% conf. interv. of LC50
Slope
404
303–538
−1.57
>3.84MalathionLC50 (nmol/L)
95% conf. interv. of LC50
Slope
Sea-nineLC50 (nmol/L)
95% conf. interv. of LC50
Slope
>123
aThe sigmoidal dose response model is: lethality =
bExposure concentrations were measured, except for Aldicarb and Carbofuran, where nominal concentrations were used.
cIn case of Imidacloprid the endpoint immobility was measured instead of mortality.
1
1+10(logLC50−logConcentration)×slope.
given in Table 1. A systematic comparison with acute toxicity data
from other studies for these compounds and G. pulex is not possi-
ble due to lack of data. Additional acute toxicity data for G. pulex
was collected from the peer-reviewed literature for 32 compounds
(Table 2).
All acute toxicity data for G. pulex measured in this study refers
to the endpoint mortality, except for Imidacloprid. We observed
that Imidacloprid causes immobility (moving pleopods but no
swimming) at much lower concentrations than those required for
mortality (no moving appendices), thus we measured immobility
for Imidacloprid instead of mortality. Organisms were counted as
immobile if they did not swim after prodding. Due to the differ-
ent concentration ranges of mortality and immobility a LC50 for
mortality could not be calculated for Imidacloprid.
In the other toxicity tests immobility was not observed, organ-
isms were either dead or alive (and mobile), except for Malathion
where those beakers with partial mortality also contained a
few alive but immobile organisms. However, the occurrence of
immobile organisms was transient, at the same concentrations as
mortality and in only 14% of all observations. Thus, Imidacloprid
is the only chemical in our study with a large difference between
mortality and immobility.
Beketov and Liess (2008b) measured a 96h-LC50 of 1056
(confidence interval: 664–1760)nmol/L for Imidacloprid in G.
pulex, which is about twice the value measured in our study.
The actual difference is even larger as (Beketov and Liess,
2008b) measured mortality whereas we measured immobil-
ity and found a lower effect concentration. Slightly different
observation methods between our study and that of (Beketov
and Liess, 2008b) could have contributed to the different
toxicity, besides different source populations of G. pulex as
well as experimental conditions such as maintenance and test
media.
3.2. Acute toxicity data for D. magna
Toxicity data for D. magna was collected for all compounds
where G. pulex toxicity data were available, either measured in
this study or from literature (Table 2). Note that for D. magna
all toxicity data are values for immobility. Immobility is the
commonly measured endpoint for acute toxicity in daphnids
as mortality is more difficult to observe in such small organ-
isms.

Page 4

R. Ashauer et al. / Aquatic Toxicology 103 (2011) 38–45
41
Table 2
Acute toxicity data for Daphnia magna and Gammarus pulex used in regression and correlation analyses.
CompoundCASMoA/chemical
class groups for
data analysis
Source of D.
magna data
48h-EC50 D.
magna
LC50 G. pulex
Source of G.
pulex data
Time of G. pulex
toxicity data
nmol/L
223649
nmol/L
225
Acetamiprid 135410-20-7 NeonicotinoidFOOTPRINTa
(Beketov and
Liess, 2008b)
This study
(Beketov and
Liess, 2008b)
(Ashauer et al.,
2007b)
(Ashauer et al.,
2010)
(Bluzat and
Seuge, 1979)
This study
(Kuhn and
Streit, 1994)
(McLoughlin
et al., 2000)
(Stephenson,
1982)
(Cold and
Forbes, 2004)
(Beketov and
Liess, 2008b)
(Van
Wijngaarden
et al., 2009)
(Schroer et al.,
2004)
(McLoughlin
et al., 2000)
This study
(Bluzat and
Seuge, 1979)
This study
(Beketov and
Liess, 2008b)
(Aboul-Ela and
Khalil, 1987)
96h
Imidacloprid
Thiacloprid
138261-41-3
111988-49-9
Neonicotinoid
Neonicotinoid
FOOTPRINTa
FOOTPRINTa
332473
336736
430
1385
48h
96h
Chlorpyrifos 002921-88-2OrganophosphateFOOTPRINTa
0.2859.69 48h
Diazinon 000333-41-5OrganophosphateFOOTPRINTa
3.2927.848h
Fenthion55-38-9 Organophosphate FOOTPRINTa
20.550.3 48h
Malathion
Parathion
000121-75-5
56-38-2
Organophosphate
Organophosphate
FOOTPRINTa
(Guilhermino
et al., 2000)
FOOTPRINTa
2.12
7.42
3.45
11.0
48h
96h
Pirimiphos-methyl 29232-93-7 Organophosphate0.68818.0 48h
Cypermethrin52315-07-8Pyrethroid(Stephenson,
1982)
FOOTPRINTa
0.721 0.240 24hc
Esfenvalerate66230-04-4Pyrethroid2.14 0.338 48h
Fenvalerate51630-58-1 PyrethroidFOOTPRINTa
0.071 0.405 96h
Gamma-
cyhalothrin
76703-62-3PyrethroidFOOTPRINTa
0.0700.036
48h
Lambda-
cyhalothrin
Permethrin
91465-08-6Pyrethroid FOOTPRINTa
0.8000.07048h
52645-53-1 Pyrethroid FOOTPRINTa
15.31.12 96h
Aldicarb
Carbaryl
000116-06-3
000063-25-2
Carbamate
Carbamate
FOOTPRINTa
FOOTPRINTa
2207
31.8
1558
144
48h
48h
Carbofuran
Fenoxycarb
001563-66-2
79127-80-3
Carbamate
Carbamate
FOOTPRINTa
FOOTPRINTa
42.527.448h
96h 16595741
Methomyl 16752-77-5Carbamate (Pereira and
Goncalves,
2007)
ECOTOX
AQUIREb
1494685 96h
2,4,5-
Trichlorophenol
000095-95-4Effect on
energy
production
Effect on
energy
production
Effect on
energy
production
Effect on
energy
production
Effect on
nervous
systemd
Effect on
nervous
systemd
Effect on
nervous
systemd
Effect on
nervous
systemd
Effect on
nervous
systemd
Baseline
toxicant
Baseline
toxicant
Baseline
toxicant
Baseline
toxicant
140592333This study 48h
4,6-Dinitro-o-
cresol
000534-52-1
ECOTOX
AQUIREb
161501805This study48h
Pentachlorophenol000087-86-5
FOOTPRINTa
1689.621026(Ashauer et al.,
2007b)
48h
Tebufenpyrad119168-77-3
FOOTPRINTa
13872.2 (Beketov and
Liess, 2008b)
96h
Endosulfan115-29-7
FOOTPRINTa
108121.5 (Cengiz and
Unlu, 1999)
48h
Indoxacarb173584-44-6
FOOTPRINTa
11374774 (Beketov and
Liess, 2008b)
96h
Iprodione 36734-19-7
FOOTPRINTa
199910479(Beketov and
Liess, 2008b)
96h
Ivermectin 70288-86-7
FOOTPRINTa
3.483.72 (Alonso et al.,
2010)
48h
Lindane58-89-9
FOOTPRINTa
5502 103(Bluzat and
Seuge, 1979)
48h
1,2,3-
Trichlorobenzene
2,4-D
000087-61-6(Zhao et al.,
1998)
FOOTPRINTa
74134618 This study 24hc
94-75-745240710405 (Seuge et al.,
1978)
This study
96h
2,4-Dichloroaniline000554-00-7 ECOTOX
AQUIREb
ECOTOX
AQUIREb
3734 5257148h
2,4-Dichlorophenol000120-83-2
1595115193 This study 48h

Page 5

42
R. Ashauer et al. / Aquatic Toxicology 103 (2011) 38–45
Table 2 (Continued)
CompoundCAS MoA/chemical
class groups for
data analysis
Source of D.
magna data
48h-EC50 D.
magna
LC50 G. pulex
Source of G.
pulex data
Time of G. pulex
toxicity data
3,4-Dichloroaniline 95-76-1Baseline
toxicant
Baseline
toxicant
Baseline
toxicant
Respiration
inhibitore
Inhibition of
mitosis and cell
divisione
Inhibits protein
synthesise
Reactive
toxicantf
Not applicable,
fungicidee
(Adema and
Vink, 1981)
FOOTPRINTa
74065 104925(Girling et al.,
2000)
(Girling et al.,
2000)
(Stephenson,
1983)
(Beketov and
Liess, 2008b)
(Van
Wijngaarden
et al., 1998)
(Beketov and
Liess, 2008b)
This study
48h
Atrazine1912-24-9
39410269547 48h
Phenol108-95-2(Hermens et al.,
1984)
FOOTPRINTa
244395765062 48h
Azoxystrobin 131860-33-8
57066996h
Carbendazim10605-21-7
(Van
Wijngaarden
et al., 1998)
FOOTPRINTa
1674403
48h
Cyprodinil121552-61-2
146 3063 96h
Ethylacrylate000140-88-5
(Staples et al.,
2000)
(van
Wijngaarden
et al., 2010)
FOOTPRINTa
78905 5551948h
Fluazinam79622-59-6
333 757(van
Wijngaarden
et al., 2010)
(Beketov and
Liess, 2008b)
48h
Prochloraz 67747-09-5Disrupts
membrane
functione
Reactive
toxicant
Disrupts
membrane
structure and
functione
114155787 96h
Sea-nine 064359-81-5 ECOTOX
AQUIREb
FOOTPRINTa
26.697.5This study 48h
Tecnazene117-18-0
767 1115(Whale et al.,
1988)
96h
aFOOTPRINT PPDB database: The Pesticide Properties Database (PPDB) developed by the Agriculture & Environment Research Unit (AERU), University of Hertfordshire,
funded by UK national sources and the EU-funded FOOTPRINT project (FP6-SSP-022704). http://sitem.herts.ac.uk/aeru/footprint/index2.htm, last accessed 29 October 2010.
bECOTOX AQUIRE database: http://cfpub.epa.gov/ecotox/, last accessed 29 October 2010. The median of two or three values present in the database was used here.
cBoth toxicity data were for 24h test duration (24h-EC50 Daphnia magna and 24h-LC50 Gammarus pulex).
dEffects on nervous system other than those caused by pyrethroids, organophosphates, carbamates and neonicotinoids.
eThese compounds could not be assigned to any group or mode of action with at least three compounds. We list a mode of action based on the FOOTRPINT PPDB database
(for effects on target organisms). Note that these not necessarily correspond to the mode of action in G. pulex or D. magna.
fMode of action from (Harder et al., 2003).
3.3. Comparison of G. pulex with D. magna via correlation
analysis
The toxicity data of the 44 compounds in our dataset range
over seven orders of magnitude. When all toxicity data of D. magna
were plotted against data of G. pulex the neonicotinoids were iden-
tified as a clearly separate group in the correlation plot (Fig. 1A,
neonicotinoids are circled). G. pulex is two to three orders of mag-
nitude (factor of 100 to 1000, mean 571) more sensitive towards
neonicotinoidsthanD.magna(two-tailed,pair-wiset-testwithlog-
transformed toxicity data, P=0.0046, ˛=0.05, n=3; Fig. 1B). For
organophosphates we found that D. magna is more sensitive than
G. pulex by approximately a factor of six (two-tailed, pair-wise t-
test with log-transformed toxicity data, P=0.0256, ˛=0.05, n=6;
Fig. 1C).
By visual inspection of the correlation plots we observed that
G. pulex was slightly more sensitive towards pyrethroids (Fig. 1D),
however, that difference is not significant (two-tailed, pair-wise t-
test with log-transformed toxicity data, P=0.1376, ˛=0.05, n=6).
The plots for carbamates (Fig. 1E), effects on energy production
(Fig. 1F), effects on the nervous system (Fig. 1G) and baseline
toxicants (Fig. 1H) do not reveal any systematic deviation from
the 1:1 correlation. The two data points in the plot for effects
on the nervous system (Fig. 1G) where G. pulex is more than one
order of magnitude more sensitive than D. magna are Endosulfan
and Lindane. However, the difference between the sensitivity of
D. magna and G. pulex for this mode of action is not significant
(two-tailed, pair-wise t-test with log-transformed toxicity data,
P=0.4921, ˛=0.05, n=5). The data that deviate most from the 1:1
line in the plot for baseline toxicants (Fig. 1H) are a base (2,4-
dichloroaniline,2,4-DCA)andanacid(2,4-D).Forthesecompounds
pH dependent chemical speciation and uptake of the organisms
may contribute to inter-experimental variability in the data and
the observed deviations from the 1:1 line.
For the whole dataset, with the neonicotinoids included, 68%
and 93% of all G. pulex toxicity data are within one and two orders
of magnitude of the D. magna data, respectively. Without the neon-
icotinoids 75% of all G. pulex toxicity data are within one order of
magnitude of the D. magna data and 100% within two orders of
magnitude.
3.4. Comparison of G. pulex with D. magna via regression analysis
The slope of the regression of all data was not significantly dif-
ferentfromoneandtheslopeoftheregressionofalldataminusthe
neonicotinoids was even closer to one (Table 3). The slopes of the
regressions of the mode of action groups were not significantly dif-
ferent from zero (i.e. no clear linear relationship between D. magna
and G. pulex toxicity). The lack of regressions different from zero
for the mode of action groups is plausible, given that any mode
of action group covered only two to three orders of magnitude in
hydrophobicity, that variability ranged about one order of magni-
tude in each direction of the 1:1 line and that each mode of action
group consisted of a small number of data pairs (three to seven).
4. Discussion
G. pulex appears to be generally more sensitive towards neon-
icotinoids than D. magna by a factor of 100–1000. The large
difference between the endpoints mortality and immobility in the
G. pulex toxicity test with Imidacloprid is unusual. Both obser-
vations deserve further investigation and consideration in risk

Page 6

R. Ashauer et al. / Aquatic Toxicology 103 (2011) 38–45
43
Fig. 1. Correlation plots with 1:1 line (dashed) and lines indicating a deviation by one order of magnitude (dotted). The time point for the G. pulex toxicity data is 24h, 48h
or 96h depending on availability (see Table 2).

Page 7

44
R. Ashauer et al. / Aquatic Toxicology 103 (2011) 38–45
Table 3
Model II regressions of log transformed toxicity data.
All data
All data without neo-nicotinoids
Pyrethroids
Organo-phosphates
Neo-nicotinoids
Carbamates
Effects on energy
production
Effects on nervous
system
Baseline toxicants
Best-fit valuesa
Slope
0.88±0.09
0.98±0.08
0.47±0.29
0.37±0.36
4.90±3.80
1.23±0.65
1.08±1.12
1.31±0.96
0.68±0.74
Y-intercept when X=0.0
0.16±0.29
0.095±0.24
−0.59±0.24
1.02±0.26
−24.06±20.75
−0.074±1.62
−0.51±3.95
−1.26±2.85
1.38±3.55
X-intercept
−0.18
−0.10
1.26
−2.75
4.91
0.06
0.47
0.96
−2.01
95% confidence intervals
Slope
0.70–1.06
0.82–1.13
−0.33 to 1.28
−0.62 to 1.36
−43.33 to 53.13
−0.83 to 3.29
−3.73 to 5.90
−1.75 to 4.37
−1.21 to 2.59
Y-intercept when X=0.0
−0.43 to 0.74 −0.39 to 0.58
−1.26 to 0.08
0.30–1.74
−287.7 to 239.6
−5.23 to 5.08
−17.52 to 16.50
−10.32 to 7.80
−7.75 to 10.50
Is slope significantly non-zero?
P value
<0.0001
<0.0001
0.1794
0.3573
0.4197
0.1541
0.4356
0.2665
0.3966
Deviation from zero?
Significant
Significant
Not sign.
Not sign.
Not sign.
Not sign.
Not sign.
Not sign.
Not sign.
Data
Number of data pairs
44
41
6
6
3
5
4
5
7
alog LC50Gammaruspulex=slope×log LC50Daphniamagna+intercept
assessment. A study with the neonicotinoid thiacloprid and seven
freshwater arthropods, also including G. pulex, found that D. magna
was the least sensitive species (Beketov and Liess, 2008a). It
appears that D. magna may be exceptionally insensitive towards
neonicotinoids rather than G. pulex being exceptionally sensi-
tive.
In the sensitivity ranking of (Rubach et al., 2010) the fami-
lies Gammaridae and Daphniidae were ranked next to each other
for carbamates, but further apart for pyrethroids and organophos-
phates. We found mode of action-specific sensitivity differences
between G. pulex and D. magna for organophosphates (as also
observed by (Rubach et al., 2010)) and neonicotinoids. We did
not find such a relationship for pyrethroids (unlike Rubach et al.
(2010)), probably due to our small sample size (n=6). For the other
mode of action groups the correlation plots do not indicate any
systematic sensitivity differences between G. pulex and D. magna.
However,thecorrelationplotsshowthatcompoundswiththesame
mode of action tend to group together in one area of the plot. This
pattern is very clear for the pyrethroids, organophosphates, neon-
icotinoids and baseline toxicants, less so for the carbamates and
the two more heterogeneous, aggregated mode of action groups’
effects on energy production and effects on nervous system. Still,
there were no significant linear relationships between G. pulex and
D. magna toxicity data for the mode of action groups. The lack of
mode of action-specific significant linear relationships (partly due
to small sample size) does not allow us to generate mode of action-
specific interspecies correlation estimation models for G. pulex and
D.magna,ashasbeendoneformanyotherspeciespairs(Dyeretal.,
2006; Raimondo et al., 2007, 2010).
The regressions with all data and with all data minus neonicoti-
noids were both significant linear relationships with slopes around
one. When neonicotinoids are excluded the regression resembles
the 1:1 line very closely. As the slope is around one (confidence
interval: 0.82–1.14) and the intercept around zero (confidence
interval: −0.39 to 0.58) we conclude that G. pulex is generally
equally sensitive towards organic xenobiotics as D. magna. Such
a simple prediction rule would predict sensitivity of G. pulex based
on that of D. magna within one order of magnitude for 75% of all
compounds and within two orders of magnitude for 100% of the
compoundsinourdataset,ifneonicotinoidsareexcluded.Forcom-
pounds with modes of action that are not represented in our study
such a prediction would be more uncertain, because they might
belong to a group of chemicals which also exhibit a large difference
between the sensitivity of G. pulex and D. magna, just as with the
neonicotinoids.
We conclude from our study that the acute toxicity of organic
chemicals to G. pulex correlates with the sensitivity of D. magna
across most modes of action, with the clear exception of neonicoti-
noids and a less pronounced divergence for organophosphates. The
implications for risk assessment of chemicals are twofold. First, the
correlation of the sensitivity of the two organisms towards differ-
ent chemicals indicates that, assessments based on the one species
can be extrapolated to the other. Second, the variability between
the two species for the same chemical, and thus, extrapolation fac-
tors for interspecies variation, may be up to two or three orders of
magnitude.Thelimitednumberoftoxicitydatafororganismsother
than D. magna prevents a better understanding of mode of action
specific sensitivity differences between species.
Acknowledgements
Funding was provided by the Swiss National Science Foun-
dation (grant 200021-119795), the Swiss Federal Office for the
Environment (FOEN, grants 09.033.PJ/I362-1602 and 09.0012.PJ),
the SETAC-CEFIC-LRI Innovative Science Award and the Univer-

Page 8

R. Ashauer et al. / Aquatic Toxicology 103 (2011) 38–45
45
sity of Queensland Travel Award for international collaborative
research. We thank Kristin Schirmer for reviewing this manuscript.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.aquatox.2011.02.002.
References
Aboul-Ela, I.A., Khalil, M.T., 1987. The acute toxicity of three pesticides on organisms
of different trophic levels as parameters of pollution in lake Wadi El Rayan, El
Fayoum. Egypt. Proc. Zool. Soc. A. R. Egypt. 13, 31–36.
Adema, D.M.M., Vink, I.G.J., 1981. A comparative-study of the toxicity of 1,1,2-
trichloroethane,dieldrin,pentachlorophenoland3,4-dichloroanilineformarine
and freshwater organisms. Chemosphere 10 (6), 533–554.
Alonso, A., De Lange, H.J., Peeters, E., 2010. Contrasting sensitivities to toxicants of
the freshwater amphipods Gammarus pulex and G. fossarum. Ecotoxicology 19
(1), 133–140.
Ashauer, R., Boxall, A.B.A., Brown, C.D., 2007a. Modeling combined effects of pulsed
exposure to carbaryl and chlorpyrifos on Gammarus pulex. Environ. Sci. Technol.
41 (15), 5535–5541.
Ashauer, R., Boxall, A.B.A., Brown, C.D., 2007b. New ecotoxicological model to simu-
latesurvivalofaquaticinvertebratesafterexposuretofluctuatingandsequential
pulses of pesticides. Environ. Sci. Technol. 41 (4), 1480–1486.
Ashauer, R., Hintermeister, A., Caravatti, I., Kretschmann, A., Escher, B.I., 2010.
Toxicokinetic-toxicodynamic modeling explains carry-over toxicity from expo-
sure to diazinon by slow organism recovery. Environ. Sci. Technol. 44 (10),
3963–3971.
Beketov,M.A.,Liess,M.,2008a.Acuteanddelayedeffectsoftheneonicotinoidinsec-
ticide thiacloprid on seven freshwater arthropods. Environ. Toxicol. Chem. 27
(2), 461–470.
Beketov,M.A.,Liess,M.,2008b.Potentialof11pesticidestoinitiatedownstreamdrift
of stream macroinvertebrates. Arch. Environ. Contam. Toxicol. 55 (2), 247–253.
Bluzat, R., Seuge, J., 1979. Effects of 3 insecticides – carbaryl fenthion and lindane
– acute toxicity in 4 lymnaeid invertebrates – chronic toxicity in pulmonate
lymnea. Environ. Pollut. 18 (1), 51–70.
Cengiz, E.I., Unlu, E., 1999. The effects of the different concentrations of thiodan on
the mortality rates of Gambusia affinis and Gammarus pulex. Biochem. Arch. 15
(3), 251–254.
Chapman,P.M.,Fairbrother,A.,Brown,D.,1998.Acriticalevaluationofsafety(uncer-
tainty) factors for ecological risk assessment. Environ. Toxicol. Chem. 17 (1),
99–108.
Cold, A., Forbes, V.E., 2004. Consequences of a short pulse of pesticide exposure for
survival and reproduction of Gammarus pulex. Aquat. Toxicol. 67 (3), 287–299.
Dyer, S.D., Versteeg, D.J., Belanger, S.E., Chaney, J.G., Mayer, F.L., 2006. Interspecies
correlationestimatespredictprotectiveenvironmentalconcentrations.Environ.
Sci. Technol. 40 (9), 3102–3111.
FOOTPRINT, (2009). The Pesticide Properties Database (PPDB) developed by the
Agriculture & Environment Research Unit (AERU), University of Hertfordshire,
funded by UK national sources and the EU-funded FOOTPRINT project (FP6-SSP-
022704), http://www.eu-footprint.org/ppdb.html.
Girling, A.E., Pascoe, D., Janssen, C.R., Peither, A., Wenzel, A., Schafer, H., Neumeier,
B., Mitchell, G.C., Taylor, E.J., Maund, S.J., Lay, J.P., Juttner, I., Crossland, N.O.,
Stephenson, R.R., Personne, G., 2000. Development of methods for evaluating
toxicity to freshwater ecosystems. Ecotoxicol. Environ. Saf. 45 (2), 148–176.
Guilhermino, L., Diamantino, T., Carolina Silva, M., Soares, A.M.V.M., 2000. Acute
toxicitytestwithDaphniamagna:analternativetomammalsintheprescreening
of chemical toxicity? Ecotoxicol. Environ. Saf. 46 (3), 357–362.
Harder, A., Escher, B.I., Schwarzenbach, R.P., 2003. Applicability and limitation of
QSARs for the toxicity of electrophilic chemicals. Environ. Sci. Technol. 37 (21),
4955–4961.
Hermens,J.,Canton,H.,Steyger,N.,Wegman,R.,1984.Jointeffectsofamixtureof14
chemicals on mortality and inhibition of reproduction of Daphnia magna. Aquat.
Toxicol. 5 (4), 315–322.
Kuhn, K., Streit, B., 1994. Detecting sublethal effects of organophosphates by mea-
suringacetylcholinesteraseactivityinGammarus.Bull.Environ.Contam.Toxicol.
53 (3), 398–404.
Kunz, P.Y., Kienle, C., Gerhardt, A., 2010. Gammarus spp. in aquatic ecotoxicology
and water quality assessment: toward integrated multilevel tests. Rev. Environ.
Contam. Toxicol. 205, 1–76.
Maltby, L., Clayton, S.A., Wood, R.M., McLoughlin, N., 2002. Evaluation of the Gam-
marus pulex in situ feeding assay as a biomonitor of water quality: robustness,
responsiveness, and relevance. Environ. Toxicol. Chem. 21 (2), 361–368.
McLoughlin, N., Yin, D.Q., Maltby, L., Wood, R.M., Yu, H.X., 2000. Evaluation of
sensitivity and specificity of two crustacean biochemical biomarkers. Environ.
Toxicol. Chem. 19 (8), 2085–2092.
Naylor, C., Maltby, L., Calow, P., 1989. Scope for growth in Gammarus pulex, a fresh-
water benthic detritivore. Hydrobiologia 188–189 (1), 517–523.
Pereira, J.L., Goncalves, F., 2007. Effects of food availability on the acute and chronic
toxicityoftheinsecticidemethomyltoDaphniaspp.Sci.TotalEnviron.386(1–3),
9–20.
Raimondo, S., Jackson, C.R., Barron, M.G., 2010. Influence of taxonomic relatedness
andchemicalmodeofactioninacuteinterspeciesestimationmodelsforaquatic
species. Environ. Sci. Technol. 44 (19), 7711–7716.
Raimondo, S., Mineau, P., Barron, M.G., 2007. Estimation of chemical toxicity to
wildlife species using interspecies correlation models. Environ. Sci. Technol. 41
(16), 5888–5894.
Rubach, M.N., Baird, D.J., Van Den Brink, P., 2010. A new method for ranking mode-
specific sensitivity of freshwater arthropods to insecticides and its relationship
to biological traits. Environ. Toxicol. Chem. 29 (2), 476–487.
Schroer, A.F.W., Belgers, J.D.M., Brock, T.C.M., Matser, A.M., Maund, S.J., Van Den
Brink, P.J., 2004. Comparison of laboratory single species and field population-
level effects of the pyrethroid insecticide lambda-cyhalothrin on freshwater
invertebrates. Arch. Environ. Contam. Toxicol. 46 (3), 324–335.
Seuge, J., Bluzat, R., Rodriguez Ruiz, F.J., 1978. Effects of a herbicide mixture (2,4
D and 2,4,5 T): acute toxicity in four species of limnal invertebrates; chronic
toxicity in the valved mollusc lymnea. Environ. Pollut. 16 (2), 87–104.
Staples, C.A., Murphy, S.R., McLaughlin, J.E., Leung, H.W., Cascieri, T.C., Farr, C.H.,
2000. Determination of selected fate and aquatic toxicity characteristics of
acrylic acid and a series of acrylic esters. Chemosphere 40 (1), 29–38.
Stephenson, R.R., 1982. Aquatic toxicology of cypermethrin I. Acute toxicity to
some freshwater fish and invertebrates in laboratory tests. Aquat. Toxicol. 2
(3), 175–185.
Stephenson,R.R.,1983.Effectsofwaterhardness,watertemperature,andsizeofthe
test organism on the susceptibility of the freshwater shrimp, Gammarus pulex
(L.), to toxicants. Bull. Environ. Contam. Toxicol. 31 (4), 459–466.
van Leeuwen, C.J., Vermeire, T.G. (Eds.), 2007. Risk Assessment of Chemicals—An
Introduction. Springer, Dordrecht, The Netherlands.
van Wijngaarden, R.P.A., Arts, G.H.P., Belgers, J.D.M., Boonstra, H., Roessink,
I., Schroer, A.F.W., Brock, T.C.M., 2010. The species sensitivity distribution
approach compared to a microcosm study: a case study with the fungicide
fluazinam. Ecotoxicol. Environ. Saf. 73 (2), 109–122.
Van Wijngaarden, R.P.A., Barber, I., Brock, T.C.M., 2009. Effects of the pyrethroid
insecticide gamma-cyhalothrin on aquatic invertebrates in laboratory and out-
door microcosm tests. Ecotoxicology 18 (2), 211–224.
VanWijngaarden,R.P.A.,Crum,S.J.H.,Decraene,K.,Hattink,J.,VanKammen,A.,1998.
Toxicicity of derosal (active ingredient carbendazim) to aquatic invertebrates.
Chemosphere 37 (4), 673–683.
Whale, G., Sheahan, D., Matthiessen, P., 1988. The toxicity of tecnazene, a potato
sprouting inhibitor, to freshwater fauna. Chemosphere 17 (6), 1205–1217.
Wogram, J., Liess, M., 2001. Rank ordering of macroinvertebrate species sensitivity
to toxic compounds by comparison with that of Daphnia magna. Bull. Environ.
Contam. Toxicol. 67 (3), 360–367.
Zhao, Y.H., Ji, G.D., Cronin, M.T.D., Dearden, J.C., 1998. QSAR study of the toxicity of
benzoic acids to Vibrio fischeri, Daphnia magna and carp. Sci. Total Environ. 216
(3), 205–215.