Effects of the herbicide 2,4-D on the growth of nine aquatic macrophytes

Abstract

A study was conducted to determine the effect of the herbicide 2,4-dichlorophenoxyacetic acid (2,4-D) on nine submersed macrophyte species. The first objective of the study was to investigate the sensitivity of various endpoints in macrophyte toxicity tests. A second objective was to investigate the implications of hormesis in the risk assessment of 2,4-D. 2,4-D was applied in concentrations ranging from 10 to 3000 μg L−1. Endpoints determined 4 weeks after the start of the treatment were based on shoot and root growth in water. The EC50s were calculated using models excluding and including a parameter describing hormesis. Results indicated that the total length of the roots can be regarded as a sensitive endpoint for the response of a macrophyte to 2,4-D. For the tested rooted macrophyte species, the EC50 values for the length and number of the roots ranged from 92 to 997 and from 112 to 1807 μg L−1, respectively. At low concentrations (10 and 30 μg L−1), stimulation of some of the endpoints (hormesis) was found for several of the species. Although hormesis may have ecological implications, its importance for the ecological risk assessment of 2,4-D in this study was limited.

Full text

Effects of the herbicide 2,4-D on the growth of nine aquatic macrophytes
J. Dick M. Belgers
a,
*, Ruud J. Van Lieverloo
a
, Leo J.T. Van der Pas
a
,
Paul J. Van den Brink
a,b
a
Alterra, Wageningen University and Research Centre, P.O. Box 47, 6700 AA Wageningen, The Netherlands
b
Wageningen University, Department of Aquatic Ecology and Water Quality Management, Wageningen University and Research Centre,
P.O. Box 8080, 6700 DD Wageningen, The Netherlands
Received 25 January 2006; received in revised form 8 November 2006; accepted 9 November 2006
Abstract
A study was conducted to determine the effect of the herbicide 2,4-dichlorophenoxyacetic acid (2,4-D) on nine submersed macrophyte species.
The first objective of the study was to investigate the sensitivity of various endpoints in macrophyte toxicity tests. A second objective was to
investigate the implications of hormesis in the risk assessment of 2,4-D. 2,4-D was applied in concentrations ranging from 10 to 3000 mgL
1
.
Endpoints determined 4 weeks after the start of the treatment were based on shoot and root growth in water. The EC
50
s were calculated using
models excluding and including a parameter describing hormesis. Results indicated that the total length of the roots can be regarded as a sensitive
endpoint for the response of a macrophyte to 2,4-D. For the tested rooted macrophyte species, the EC
50
values for the length and number of the
roots ranged from 92 to 997 and from 112 to 1807 mgL
1
, respectively. At low concentrations (10 and 30 mgL
1
), stimulation of some of the
endpoints (hormesis) was found for several of the species. Although hormesis may have ecological implications, its importance for the ecological
risk assessment of 2,4-D in this study was limited.
#2006 Elsevier B.V. All rights reserved.
Keywords: Macrophytes; Herbicide; 2,4-D; EC
50
; Hormesis; Risk assessment
1. Introduction
In the current risk assessment procedure for pesticides,
plant testing has focused on a relatively small number of
species of aquatic primary producers, most being algae (EU,
1997). The most commonly used vascular plants in toxicity
tests are duckweeds, which belong to the Lemnaceae (Hoffman
et al., 1995; Lewis, 1995). Although these species are often
used as representative species for all other aquatic macro-
phytes, there is evidence that they are less sensitive to
pesticides than rooted submersed species (Turgut and Fomin,
2001; Swanson et al., 1991). Rooted submersed macrophytes
are rarely used in toxicity tests (Hoffman et al., 1995; Lewis,
1995), which is mainly due to the lack of established
standardised test methods. Rooted submersed macrophytes
are, however, of great ecological significance in aquatic
ecosystems. Macrophytes are very important for nutrient
cycling, for the provision of habitats for aquatic invertebrates
and periphyton, and for physical processes such as turbidity
and stratification (Scheffer, 1998). This makes it important to
use rooted macrophytes in risk assessment procedures,
especially when they are expected to be sensitive (e.g., in
the case of herbicides). The use of rooted macrophytes in test
procedures has a significant advantage over the use of
Lemnaceae, as they allow not only biomass and number of
fronds, but also more ecologically relevant endpoints to be
determined, like root number, total root length, side shoot
number and side shoot length (Turgut and Fomin, 2001).
The first aim of the study described in the present paper was
to investigate the sensitivity of various endpoints in macrophyte
toxicity tests. To this end, the study determined the relative
sensitivities of nine macrophytes, five dicot species and four
monocot species, to the herbicide 2,4-dichlorophenoxyacetic
acid (2,4-D). The study describes laboratory toxicity values
(EC
50
) for 2,4-D for various endpoints (shoot dry weight, shoot
growth, total shoot length, number of roots, total length of roots
and dry weight of roots) in nine macrophytes that typically
occur in Dutch drainage ditches.
www.elsevier.com/locate/aquabot
Aquatic Botany 86 (2007) 260–268
* Corresponding author. Tel.: +31 317 478735; fax: +31 317 419000.
E-mail address: Dick.Belgers@wur.nl (J.D.M. Belgers).
0304-3770/$ – see front matter #2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.aquabot.2006.11.002

Broad-leaf dicotyledonous macrophyte species are more
susceptible to 2,4-D than narrow leaf monocots, which makes
2,4-D a selective herbicide (Lembi, 1996; Madsen, 2000;
Parsons et al., 2001). Since four of the nine tested species are
monocots, we expected these species to be relatively insensitive
to 2,4-D.
A second aim of the present study was to investigate the
implications of hormesis (growth stimulation or other
beneficial effects at sub toxic concentrations) for the risk
assessment of macrophytes. Hormesis has been reported in a
large number and a wide range of toxicological studies,
including the effects of herbicides on plants (Forsyth et al.,
1997; Calabrese et al., 1999; Calabrese and Baldwin, 2000;
McCann et al., 2000; Christopher and Bird, 1992). Forsyth et al.
(1997) found an increase in the production of flowers in
Myriophyllum sibiricum as a result of the presence of 2,4-D at
10 mgL
1
, and an increase in tuber production in Potamogeton
pectinatus at 10 mgL
1
2,4-D. To evaluate the significance of
hormesis for the current study, calculations were made using
models excluding and including hormesis.
2. Materials and methods
2.1. Plant material
The single-species tests described in the present paper used
nine aquatic plant species. Lemna trisulca,Elodea nuttallii,
Ceratophyllum demersum and Potamogeton lucens were
obtained from a commercial source. Ranunculus circinatus,
Ranunculus aquatilis,Potamogeton crispus and Myriophyllum
spicatum were collected, by hand, from experimental ditches
located at the Sinderhoeve experimental station in Renkum,
The Netherlands (Drent and Kersting, 1993). Ranunculus
peltatus was obtained from field ditches in the vicinity of
Wageningen, The Netherlands. Four of the nine species tested
are monocots; L. trisulca,E. nuttallii,P. crispus and P. lucens,
the other species are dicots. Plants were obtained 2 days prior to
the start of the experiments. During the pre-treatment period,
plants were acclimatised in pond water in a temperature
controlled room (20 38C) using a 14 h light/10 h dark cycle
and a fluorescent illumination system producing 220 20
mmol m
2
s
1
(four Philips HPI-T, 400 W lamps).
2.2. Test procedures
The single-species tests were performed during the first 6
months of the year 2002. Testing took place in three series. L.
trisulca,E. nuttallii,M. spicatum and C. demersum were tested
in February and R. circinatus,R. aquatilis,and P. crispus in
March, while the last tests, with R. peltatus,P. lucens and again
with E. nuttallii, were performed in May. All plants were tested
using a 4-week static test method with a single application. All
tests were conducted under the same temperature and lighting
conditions used during the acclimation period. The macro-
phytes were tested in 1.5-L glass jars (water depth 12 cm) with
no sediment, containing filtered (45 mm) nutrient-poor water
(1200 mL) originating from experimental ditches located at the
Sinderhoeve experimental station (Drent and Kersting, 1993).
The water was enriched with (in mM): N (0.0307), P (0.0025),
C (0.0065) and with 0.1 mL L
1
Tropica Mastergrow (K: 0.79,
Mg: 0.39, S: 1.01, B: 0.004, Cu: 0.006 Fe: 0.07, Mn: 0.04 Mo:
0.002 and Zn: 0.002 (%, w/w)). These amounts of nutrients (N,
P and K), C and trace elements were added twice a week. No
aeration of the test media took place during the experiments. A
constant test volume was maintained during the test period by
adding filtered water (twice a week) to replace evaporation
losses. Water level, temperature, dissolved oxygen (DO) were
monitored several times a week.
In each test system, we placed a known amount of uniform
plant material, fresh weight between 0.30 and 3.00 g (consisting
of top shoots while, when present, roots were gently removed).
Because of plant phenotypical variation different numbers of
cuttings were used per plant species. Only one cutting was used
for P. lucens, two for M. spicatum and three cuttings were used
for E. nuttallii,R. circinatus,R. aquatilis,P. crispus and R.
peltatus. One multibranched clump per test system was used for
C. demersum and 10–12 clusters for L. trisulca. Before the plant
material was weighed, it was gently blotted dry with a tissue.
The initial amount of each macrophyte was determined by
weighing two extra portions and drying the plant material at
105 8C (24 h) to determine its dry weight.
At the end of the experiment, on day 28, total plant dry
weight was determined for each test unit (dried at 105 8C for
24 h). Where relevant, the total shoot length of the macrophyte,
dry weight of the roots and total length and number of roots,
were also determined.
All plants were exposed to 2,4-D at concentrations of 0, 10,
30, 100, 300, 1000 and 3000 mgL
1
. All treatments were
arranged in a randomised block design with two replicates.
Additionally, three extra test systems (0, 30 and 3000 mgL
1
)
per plant species were used to determine the fate (behaviour) of
the herbicide (fate systems). The herbicide 2,4-D was added on
day 0 of the single-species study.
2.3. Chemical analyses
For the 2,4-D analysis, depth-integrated water samples
(100 mL) were taken from the fate systems on days 0, 2, 7, 14
and 28. Samples were taken from the systems by means of a
glass pipette. The initial test concentrations of 2,4-D in all test
systems were checked by taking water samples 1 h after the 2,4-
D application. For the 2,4-D analysis, water samples (100 mL)
from the fate systems, with initial concentrations of 0 and
30 mgL
1
, were mixed with 0.2 mL HCl (37%). 2,4-D was
extracted using Waters 0.2 g Oasis
TM
HLB extraction columns,
conditioned with 12 mL methanol and 2 mL deionised water.
Elution of 2,4-D from the extraction column was carried out
with two aliquots of 1 mL methanol. The eluate was collected
in a 10 mL graduated tube and diluted with distilled water to a
volume of 5.0 mL. Water samples from the fate systems with an
initial concentration of 3000 mgL
1
were measured after 10-
fold dilution. The water samples were analysed with high
performance liquid chromatography (HPLC). The mobile
phase (water:acetonitrile:titrisolbuffer; pH 3; 35:61:4, v:v:v)
J.D.M. Belgers et al. / Aquatic Botany 86 (2007) 260–268 261

Fig. 1. Results of single-species tests evaluating the effects of 2,4-D on the dry weight of nine aquatic macrophyte species. Means and standard errors are given
(n= 2).
J.D.M. Belgers et al. / Aquatic Botany 86 (2007) 260–268262

was set at a flow rate of 1 mL min
1
. A Waters XTerra
TM
MS
c
18
column (length 150 mm, width 4.6 mm) was used together
with a Waters XTerra
TM
MS c
18
(length 20 mm, width 3.9 mm)
guard column. The oven temperature was adjusted to 40 8C. A
wavelength of 230 nm was used for the detection of the 2,4-D
samples.
2.4. Data analysis
Logistic regression, assuming a Poisson distribution of the
data, was used to calculate the EC
50
(for biomass and relative
growth) (Van den Brink et al., 1997). In the case of biomass, the
100% effect was set at a biomass of 0 g. In the case of relative
Fig. 2. Results of single-species tests evaluating the effects of 2,4-D on other endpoints than the dry weight of seven aquatic macrophyte species. Means and standard
errors are given (n= 2).
J.D.M. Belgers et al. / Aquatic Botany 86 (2007) 260–268 263

growth, the 100% effect was set at a growth of 0 g per 4 weeks.
This means that, based on the same data, EC
50
values for
biomass and relative growth can differ substantially (see Va n
den Brink et al., 1997 for a visual representation). In addition, a
model including a parameter f, describing hormesis, as
introduced by Van Ewijk and Hoekstra (1993), was fitted to
the data. The models were also compared by testing the
significance of the hormesis parameter f(Van Ewijk and
Hoekstra, 1993). Both models were programmed in GenStat for
Windows, 6th ed. (Payne, 2002).
3. Results
3.1. 2,4-D analysis and test water quality
Water samples from the fate systems taken 1 h post-
treatment showed that 89 9.5% of the intended concentration
of 2,4-D was present in the test systems. After 28 days, an
average of 87 16% of the intended concentration remained.
Concentrations of 2,4-D in the fate systems remained relatively
unchanged during the 28-day experimental period, indicating
that the chemical was very persistent in the water and glass jar
system. No 2,4-D was detected in the control fate systems.
There was an increase in pH in the test systems for all plant
species during the 28-day experimental period (pH day 4: 8.4–
9.2, day 28: 9.2–10.6). Relatively small differences were found
between the measured minimum and maximum water quality
conditions (DO and temperature) at 4 h and 28 days post-
treatment.
3.2. 2,4-D effects on plant growth
The effect of 2,4-D on the evaluated endpoints of the nine
aquatic plant species is presented in Figs. 1 and 2.Thecontrols
showed a growth of more than 65% in eight of the nine
plant species (Fig. 1). Only M. spicatum showed a negative
growth in the controls (29%). For this plant species, no EC
50
values were calculated. The toxicity, EC
50
values of 2,4-D for
each of the nine species of aquatic plants, are provided in
Tables 1 and 2.
With an EC
50
of 242 mgL
1
, for relative growth R. aquatilis
was the most sensitive plant species, followed by E. nuttallii
(B), R. circinatus and P. lucens. The data for L. trisulca,E.
nuttallii (A), R. peltatus and C. demersum did not allow an EC
50
for the growth of the plant to be calculated. When all endpoints
are considered, the lowest EC
50
s for all species lie within an
order of magnitude (between 60 and 574 mgL
1
), except for L.
trisulca and C. demersum, for which no effects could be
demonstrated on any endpoint. The length of the C. demersum
shoot, however, was greater at 100 and 300 mgL
1
than at the
other concentrations (Fig. 2), though no clear dose–response
relation could be established.
In contrast to the first experiment with E. nuttallii, the
second experiment with this species yielded a growth inhibition
at the two highest concentrations (Fig. 1). Whereas effects on
the shoot parts of the plants were very different between the two
experiments, effects on roots were comparable (Table 1).
Table 1
EC
50
(mgL
1
) calculations for six endpoints, including 95% confidence intervals (parentheses) and regression deviance values, based on a model without hormesis
Species Biomass
(g dry weight)
Mean
deviance
Relative
growth
Mean
deviance
Roots
(g dry weight)
Mean
deviance
Shoot length
(cm)
Mean
deviance
Root length
(cm)
Mean
deviance
No. of
roots
Mean
deviance
Lemna trisulca >3000 <0.1 >3000 <0.1 – – – – – – – –
Elodea nuttallii (A) >3000 <0.1 >3000 <0.1 898 (213–3781) <0.1 >3000 0.8 574 (368–894) 6.1 982 (600–1607) 0.9
Ceratophyllum
demersum
No fit – No fit – – – No fit – – – – –
Ranunculus
circinatus
2731 (1556–4796) <0.1 719 (329–1618) <0.1 111 (91–135) <0.1 1120 (186–6744) 14 100 (x–x) 1.6 112 (100–125) 0.2
R. aquatilis >3000 <0.1 242 (47–1237) <0.1 No fit – 683 (140–3339) 6.2 92 (55–155) 69 No fit –
Potamogeton
crispus
>3000 <0.1 >3000 <0.1 347 (155–780) <0.1 1988 (1431–2762) 2.1 290 (282–398) 1.5 326 (271–393) 0.7
R. peltatus No fit – No fit – 245 (x–x)<0.1 140 (18–1058) 6.4 263 (59–1167) 12 271 (x–x) 1.1
P. lucens >3000 <0.1 1300 (484-3493) <0.1 No fit – 1063 (344–3283) 1.7 181 (60–547) 8.7 299 (267–334) 0.9
E. nuttallii (B) 2243 (768–6548) <0.1 292 (44–1904) <0.1 1096 (852–1411) <0.1 785 (391–1580) 2.5 997 (x–x) 13 1807 (x–x)1
‘–’ indicates that no data were available for this endpoint, ‘no fit’ means that no convergence was obtained in the model. ‘x–x’ indicates that it was not possible to calculate a confidence interval.
J.D.M. Belgers et al. / Aquatic Botany 86 (2007) 260–268264

A comparison of all endpoints in the two E. nuttallii
experiments shows that only a small difference between the
two experiments was found for the lowest EC
50
(574 and
292 mgL
1
). In M. spicatum,noEC
50
for effects observed on
roots and new shoots were calculated because of the negative
growth in the controls. At the two highest treatment levels, the
growth of R. circinatus was clearly inhibited (Fig. 1). The
corresponding EC
50
of 100 mgL
1
indicates that this inhibition
is not substantial at lower concentrations (Table 1). Stimulation
of the total plant length in R. circinatus was observed at
30 mgL
1
(Fig. 2).
The EC
50
calculation of 242 mgL
1
on the growth of R.
aquatilis indicates that this growth inhibition was relatively
small at the lower concentrations. At the three highest treatment
levels, a reduction in most of the endpoints was observed for P.
crispus (EC
50
290 mgL
1
). A minor increase in the number
of new shoots was found at the 300 and 1000 mgL
1
treatment
levels (Fig. 2). The three highest treatment levels caused an
effect on most endpoints measured for R. peltatus, except for
biomass (EC
50
136 mgL
1
;Table 1). For P. lucens, effects
were observed on all endpoints, except root biomass
(EC
50
181 mgL
1
). The length of the roots was higher at
10 mgL
1
and lower at higher levels (Fig. 2).
Table 2 shows the results of the EC
50
calculations taking
hormesis into account. Due to hormetic effects the model was
able to calculate EC
50
values for 13 of the 39 fitted endpoints.
Only for two cases was this hormesis significant, viz., the length
of the shoot and root parts of R. aquatilis.TheEC
50
for shoot
length changed the most, as it decreased by a factor of 2
(Tables 1 and 2).
4. Discussion
4.1. Fate of 2,4-D
Boyle (1980) found a relatively fast dissipation of 2,4-D in
ponds: after approximately 10 days, the concentration in the
water phase was half of the nominal concentration. A review
report for the active substance 2,4-D by the European
Commission (EU, 2001) presented a DT50 of 29 days based
on a water/sediment study.
The present study found a very slow dissipation of 2,4-D
from the test systems. After 28 days, 89 2.3% of the nominal
concentration was still present at the highest treatment level.
The most likely explanation for this low degradation rate is the
lack of sediment in the test systems. Microorganisms present in
the sediment are probably responsible for the degradation of
2,4-D in water/sediment systems. Some microorganisms are
capable of using 2,4-D as their sole carbon source or of co-
metabolizing the herbicide 2,4-D (Fournier, 1980). The results
of the current experiments are therefore representative for
predicting the effects of a chronic exposure to 2,4-D.
4.2. Plant toxicity
Most published work on toxicity tests with 2,4-D on
macrophytes (Swanson et al., 1991; Fairchild et al., 1999; EU,
Table 2
EC
50
(mgL
1
) calculations for six endpoints, including 95% confidence intervals (parentheses) and regression deviance values, based on a model with hormesis
Species Biomass
(g dry weight)
Mean
deviance
Relative
growth
Mean
deviance
Roots
(g dry weight)
Mean
deviance
Shoot length
(cm)
Mean
deviance
Root length
(cm)
Mean
deviance
No.
of roots
Mean
deviance
L. trisulca No hor – No hor – – – – – – – – –
E. nuttallii (A) No hor – No hor – No hor – No hor – 334 (91–1228) 27 1073 (711–1618) 1
C. demersum No fit – No fit – – – No fit – – – – –
R. circinatus No hor – No hor – 51 (19–136) <0.1 No hor – No hor – No hor –
R. aquatilis No hor – No hor – No fit – 342 (185–633) (sf) 2.2 108 (85–139) (sf) 75 No fit –
P. crispus No hor – No hor – No hor – No hor – 124 (38–398) 35 191 (33–1101) 8.4
R. peltatus No fit – No fit – No hor – 76 (25–229) 8.2 499 (64–3907) 20 No hor –
P. lucens >3000 <0.1 1348 (636–2857) <0.1 No fit – No hor – No hor – 302 (214–428) 0.9
E. nuttallii (B) No hor – No hor – No hor – 1053 (237–4673) 5.7 No hor – No hor –
‘–’ indicates that no data were available for this endpoint, ‘no fit’ means that no convergencewas obtained in the model, ‘no hor’ means that no EC
50
was calculated due to the fact that no hormesis effect was found in the
model with hormesis (i.e., fis zero), ‘x–x’ indicates that it was not possible to calculate a confidence interval, ‘sf’ means that the model with hormesis fitted significantly better than the model without hormesis.
J.D.M. Belgers et al. / Aquatic Botany 86 (2007) 260–268 265

2001; Michel et al., 2004) has involved data on the sensitivity to
2,4-D of members of the duckweed family (Lemna spp.). All
these studies indicate that duckweeds are insensitive or only
moderately sensitive to 2,4-D (reported EC
50
values range from
500 to >6000 mgL
1
). Our study found an EC
50
>3000
mgL
1
for the relative growth of L. trisulca. Since L. trisulca
belongs to the monocots, this relative insensitivity to 2,4-D was
to be expected.
Published EC
50
data on the toxicity of 2,4-D for other
macrophytes than duckweeds are rare, though Myriophyllum
spp. have been used more frequently in 2,4-D toxicity
evaluations than other species. Swanson et al. (1991) found
that Myriophyllum was killed by 44 mgL
1
of 2,4-D after 30
days of exposure. Turgut and Fomin (2002) studied the effect of
2,4-D on M. aquaticum and determined 2-week EC
50
values of
20 and 158 mgL
1
on root number and root length,
respectively. A field experiment evaluating the effects of
100 mgL
1
2,4-D on M. sibiricum in a wetland showed a
reduction of shoot growth (weight) by about 40% and a
mortality rate of 52% in 60 days (Forsyth et al., 1997). Field
observations of M. spicatum beds during experiments with
2000 mgL
1
2,4-D indicated that after 1 week of exposure,
Myriophyllum plants turned brown and were partly uprooted,
and that some plants were floating (Kobraei and White, 1996).
Some of these observations are in accordance with the findings
of our study. However, we were unable to generate EC
50
values
for M. spicatum because of the negative growth in the controls.
The reason for this negative growth is unclear.
The results of the two tests with E. nuttallii (A and B) show
that results of different tests with the same macrophyte can
result in differences in growth among controls. Although the
intraspecific endpoint variation is relatively high for E. nuttallii,
it is noted that the EC
50
for the most sensitive endpoints for both
E. nuttallii tests are almost in de same order of magnitude (A;
length of roots 574 mgL
1
, B: relative growth 292 mgL
1
).
The intraspecific variation as reported for E. nuttallii is an
important observation that can not be ignored. Toxicity tests
with rooted macrophytes are mostly performed with plant
material collected from the field (streams, ponds and ditches)
(Fairchild et al., 1998;Cedergreen et al., 2004). The period and
place (growth circumstances) of collection could play an
important role in the growth of shoots and roots and therefore in
the sensitivity of the plants. The intraspecific variation of E.
nuttallii is probably due to differences in conditions related to
the period and/or place of collection. Elodea plant parts were
obtained from a commercial source. There is no information
available about the exact place of collection, the growth
circumstances and whether the Elodea (A) and Elodea (B) plant
parts were from the same origin. More research is needed on the
biology (macrophyte specific physiology and metabolic path-
ways) of macrophytes during growth season to understand the
reaction of plant (parts) growing under laboratory conditions.
Also prolonging the pre-treatment period could reduce the
variability in plant material used for the tests.
R. aquatilis was the most sensitive plant species in terms of
the EC
50
(92 mgL
1
) for the total length of the roots. Turgut
and Fomin (2002) found calculated EC
50
values in M.
aquaticum of 50 and 158 mgL
1
2,4-D for total root length
and root number, respectively; values which are very similar to
our values for most of the tested macrophytes (Table 1). Roshon
(1997) reported 14-day EC
50
values in M. sibiricum of only 13
and 18 mgL
1
2,4-D for inhibition of total root length and root
number, respectively. In most rooted macrophytes, roots play
an important role, not only as anchoring systems, but also for
the uptake of nutrients from the sediment and/or the water
column (Smart and Barko, 1985;Mantai and Newton, 1982).
Since contaminants like 2,4-D are absorbed by these roots
(WHO, 1989), root growth inhibition following 2,4-D exposure
would be most likely to be harmful to rooted macrophytes. The
total length of the roots can therefore be regarded as an
important endpoint for the sensitivity of a macrophyte to a
contaminant like 2,4-D. In Lemna toxicity tests (OECD, 2002),
roots growth is seldom used as an endpoint. An important
ecological endpoint like root growth is missed when only
Lemna spp. are used in toxicity tests.
Since L. trisulca belongs to the monocots we expected L.
trisulca, like L. minor, to be relatively insensitive to 2,4-D. With
an EC
50
>3000 mgL
1
for both biomass and relative growth
we indeed found a relative insensitivity of L. trisulca to 2,4-D.
The other three monocods; E nuttallii,P. crispus and P. lucens,
however, were sensitive to 2,4-D (Table 1). Floating macro-
phytes like L. minor are only exposed through their small root
system and lower leaf surface. The impact of 2,4-D on these
plants will be less then when the plant is totally submerged (like
E nuttallii,P. crispus and P. lucens). L. trisulca is in contrast to
other duckweeds a submerged macrophyte, and it is never-
theless not sensitive. In contrast to the other three monocots,
however, L. trisulca and the other insensitive macrophyte C.
demersum do not develop roots during the test. 2,4-D is
absorbed by these roots (WHO, 1989), so the lack of root
growth could explain the relative insensitivity of these two
macrophytes for 2,4-D. The common belief that 2,4-D is acting
only on broadleaf plants (dicots) is therefore not always true,
and whether the macrophytes develop roots during the test
period might be of larger importance.
As expected, at low 2,4-D levels (10 and 30 mgL
1
), the
root growth of several species was found to be stimulated
(Fig. 2). This increased growth at low concentrations has been
termed hormesis (Calabrese and Baldwin, 2000). Hormesis can
be defined as the stimulatory effect of a subtoxic concentration
of a toxin. Such an effect could be positive or negative in the
long run. Christopher and Bird (1992) reported a change in leaf
morphology of M. spicatum caused by low concentrations of
2,4-D (60 mgL
1
). The increase found in flower production in
M. sibiricum and in tuber production in P. pectinatus was due to
the presence of 10 mgL
1
2,4-D (Forsyth et al., 1997). Such
changes in plant morphology and/or increased root growth at
low 2,4-D concentrations do not necessarily imply a healthy
response by the plant. The longer-term consequences of this
non-natural response of the plant remain unclear.
To date, hormesis has usually been ignored in ecological
risk assessment, nor do standard statistical techniques
currently in use take hormesis into account (Cedergreen
et al., 2005; Calabrese and Baldwin, 2003; Chapman, 2001;
J.D.M. Belgers et al. / Aquatic Botany 86 (2007) 260–268266

Calabrese et al., 1999). It may, however, be appropriate to
include hormesis in EC calculations. A general concentration–
response model (Van Ewijk and Hoekstra, 1993)hasbeen
suggested for the analysis of hormetic concentration–response
curves. Tab le 2 shows the results of EC
50
calculations based on
a model including hormesis. Despite some significant
differences, there are no drastic changes in the EC
50
values
calculated for these endpoints. For R. aquatilis,theEC
50
calculated, excluding and including hormesis, for shoot length
was 683 and 342 mgL
1
, respectively. Although the hormesis
model yielded different EC
50
values for eleven endpoints than
the model with no hormesis, these differences were not
significant (Tab l e 2 ). For Plucensand R. aquatilis ahigh
stimulus was found at low concentrations (10 and 30 mgL
1
)
for shoot length and number of roots, respectively (Fig. 2).
Despite this stimulus, no hormesis could be established for
these endpoints due to the fact that the standard deviation for
these endpoints was too high. In our study, the use of the
hormesis model did not lead to considerable changes in the
toxicity values. Nevertheless, we feel that the use of a model
which does take hormesis into account to calculate EC
50
values
for pesticides would be more appropriate because hormesis is
unpredictable, even when we know the mode of action of the
toxin. So, ideally, models with and without hormesis should be
used in all cases and the significance of hormesis should be
tested.
Acknowledgements
This research was funded jointly by the UK Department of
Environment, Food and Rural Affairs (DEFRA) and the Dutch
Ministry of Agriculture, Nature and Food Safety (DLO/PO
research programme 416). We thank Michiel Daam for his help
during the experimental stage. Wies Akkermans for the
implementation of the hormesis model and Jan Klerkx, Rene
van Wijngaarden, Gertie Arts and Theo Brock for their
comments in preparing the manuscript.
References
Boyle, T.P., 1980. Effects of the aquatic herbicide 2,4-D DMA on the ecology of
experimental ponds. Environ. Pollut. 21, 35–49.
Calabrese, E.J., Baldwin, L.A., Holland, C.D., 1999. Hormesis: a highly
generalizable and reproducible phenomenom with important implications
for risk assessment. Risk Anal. 19 (2), 261–281.
Calabrese, E.J., Baldwin, L.A., 2000. Chemical hormesis: its historical founda-
tions as a biological hypothesis. Hum. Exp. Toxicol. 19, 2–31.
Calabrese, E.J., Baldwin, L.A., 2003. Toxicology rethinks its central belief—
HORMESIS demands a reappraisal of the way risks are assessed. Nature
421, 691–692.
Cedergreen, N., Streibig, J.C., Spliid, N.H., 2004. Sensitivity of aquatic plants
to the herbicide metsulfuron-methyl. Ecotox. Environ. Saf. 57, 153–161.
Cedergreen, N., Ritz, C., Streibig, J.C., 2005. Improved empirical models
describing hormesis. Environ. Toxicol. Chem. 24 (12), 3166–3172.
Chapman, P.M., 2001. The implications of hormesis to ecotoxicology and
ecological risk assessment (ERA). BELLE Newslett. 10 (1).
Christopher, S.V., Bird, K.T., 1992. The effects of herbicides on development of
Myriophyllum spicatum L. cultured in vitro. J. Environ. Qual. 21, 203–207.
Drent, J., Kersting, K., 1993. Experimental ditches for research under natural
conditions. Water Res. 27 (9), 1497–1500.
EU, 1997. Council Directive 97/57/EC of September 21, 1997; Establishing
annex VI to Directive 91/414/EEC Concerning the placing of plant protec-
tion products on the market. Official Journal of the European Communities
L265, pp. 87–109.
EU, 2001. Review report for the active substance 2,4-D, 2,4-D 7599/VI/97-final.
European Commission, Health & Consumer protection Directorate-Gen-
eral.
Fairchild, J.F., Ruessler, P.S., Carlson, A.R., 1998. Comparative sensitivity
of five species of macrophytes and six species of algae to atrazine,
metrubuzin, alachlor and metolachlor. Environ. Toxicol. Chem. 17,
1830–1834.
Fairchild, J.F., Sappington, L.C., Ruessler, D.S., 1999. An ecological risk
assessment of the potential for herbicide impacts on primary productivity
of the lower Missouri river. In: Morganwalp, D.W., and Buxton, H.T. (Eds.),
U.S. Geological Survey Toxic Substances Hydrology Program—Proceed-
ings of the Technical Meeting, vol. 2 of 3, Charleston, SC, March 8–12,
1999.Contamination of Hydrologic Systems and Related Ecosystems: U.S.
Geological Survey Water-Resources Investigations Report 99-4018B, pp.
323–330.
Forsyth, J.D., Martin, A.P., Shaw, G.G., 1997. Effects of herbicides on two
submersed aquatic macrophytes, Potamogeton pectinatus L. and Myrio-
phyllum sibiricum Komarov, in a prairie wetland. Environ. Pollut. 95, 259–
268.
Fournier, J.C., 1980. Enumeration of the soil micro-organisms able to degrade
2,4-D by metabolism or co-metabolism. Chemosphere 9, 169–174.
Hoffman, D.J., Rattner, B.A., Burton Jr., G.A., Cairns Jr., J., 1995. Handbook of
Ecotoxicology. Lewis Publishers, Boca Raton.
Kobraei, M.E., White, D.S., 1996. Effects of 2,4-dichlorphenoxyacetic acid on
Kentucky algea: simultaneous laboratory and field toxicity testing. Arch.
Environ. Cont. Tox. 31, 571–580.
Lembi, C.A., 1996. Assesment of 2,4-D use in aquatic systems in the United
States. In: Biologic and economics assesment of bebifits from use of
phenoxy herbicides in the United States. NAPIAP Report 1-PA-96, USDA,
pp. 179–184.
Lewis, M.A., 1995. Use of freshwater plants for pytotoxicity testing: a review.
Environ. Pollut. 87, 319–336.
Madsen, J.D., 2000. Advantages and Disadvantages of Aquatic Plant Manage-
ment Techniques, ERDC/EL MP-00-1. U.S. Army Engineer Research and
Development Center, Vicksburg, MS.
Mantai, K.E., Newton, M.E., 1982. Root growth in Myriophyllum: a specific
plant response to nutrient availability? Aquat. Bot. 13, 45–55.
McCann, J.H., Greenberg, B.M., Solomon, K.R., 2000. The effect of creosote
on the growth of an axenic culture of Myriophyllum spicatum L. Aquat.
Toxicol. 50, 265–274.
Michel, A., Johnson, R.D., Duke, O., Scheffler, B.E., 2004. Dose-response
relationships between herbicides with different modes of action and growth
of Lemna paucicostata: an improved ecotoxicological method. Environ.
Toxicol. Chem. 23, 1074–1079.
OECD, July 2002. Guidelines for Testing of Chemicals: Revised Proposal for a
New Guideline 221: Lemna Growth Inhibition Test, 22 pp.
Parsons, J.K., Hamel, K.S., Madsen, J.D., Getsinger, K.D., 2001. The use of
2,4-D for selective control of an early infestation of Eurasian Water-
milfoil in Loon Lake, Washington. J. Aquat. Plant Manage. 39, 117–
125.
Payne, R.W., 2002. GenStat
1
Release 6.1 Reference Manual. Lawes Agricul-
tural Trust. Rothamsted Experimental Station, Harpenden, Hertfordshire,
UK.
Roshan, R.D., 1997. Atoxicity test for the effects of chemicals on the nontarget
submersed aquatic macrophyte, Myriophyllum sibericum Komarov. Ph.D.
Thesis. University of Guelph. Guelph, Ont.
Scheffer, M., 1998. Ecology of Shallow Lakes. Chapman & Hall, London.
Smart, R., Barko, J.W., 1985. Laboratory culture of submersed freshwater
macropphytes on natural sediments. Aquat. Bot. 21, 251–263.
Swanson, S.M., Rickard, C.P., Freemark, K.E., Macquarrie, P., 1991. Testing for
pesticide toxicity to aquatic plants: recommendations for test species. In:
Gorsuch, J.W., Lower, W.R., Wang, W., Lewis, M.A. (Eds.), Plants for
Toxicity Assessment: ASTM STP 1115. American Society for Testing and
Materials, Pennsylvania, pp. 77–97.
J.D.M. Belgers et al. / Aquatic Botany 86 (2007) 260–268 267

Turgut, C., Fomin, A., 2001. Establishment of standardized growth conditions
of Myriophyllum aquaticum (Vell.) Verdcourt for testing sediment toxicol-
ogy. J. Appl. Bot. 75, 80–84.
Turgut, C., Fomin, A., 2002. Sensitivity of the rooted macrophyte Myriophyllum
aquaticum (Vell.) Verdcourt to seventeen pesticides determined on the basis
of EC
50
. Environ. Contam. Tox. B 69, 601–608.
Van den Brink, P.J., Hartgers, E.M., Fettweis, U., Crum, S.J.H., Van Donk, E.,
Brock, Th.C.M., 1997. Sensitivity of macrophyte-dominated freshwater
microcosms to chronic levels of the herbicide linuron. I. Primary producers.
Ecotox. Environ. Saf. 38, 13–24.
Van Ewijk, P.H., Hoekstra, J.A., 1993. Calculations of the EC
50
and its
confidence interval when subtoxic stimulus is present. Ecotox. Environ.
Saf. 25, 25–32.
World Health Organization, 1989. 2,4-Dichlorophenoxy-acetic acid
(2,4-D), Environmental Health Criteria 84. Environmental Aspects,
Geneva.
J.D.M. Belgers et al. / Aquatic Botany 86 (2007) 260–268268