United States
Environmental Protection
Agency
Environmental Research
Laboratory
Corvallis OR 97333
Research and Development
EPA-600/S3-83-051 Sept. 1983
4>EPA Project Summary
Interlaboratory Root Elongation
Testing of Toxic Substances on
Selected Plant Species
Hilman C. Ratsch
Four contract laboratories and three
EPA laboratories participated in the
Intel-laboratory testing of 10 toxic sub-
stances on a representative ptent species
from five families. Seeds were germi-
nated on filter paper saturated in a
solution of the toxic substance and
incubated for 115 houfs. The root lengths
were measured to evaluate the toxic
effects of the chemical concentrations
on the various species. The objective of
the testing was to estimate the con-
centration of chemical which reduced
root length to 50% of the control length.
This research attempts to determine
the precision of this bioassay used to
evaluate environmental effects under
the Toxic Substances Control Act
(TSCA). Although the method proved
to give a uniform plant-growth environ-
ment the species variability in relation-
ship to the chemical concentrations
that inhibit root growth makes it diffi-
cult to use this assay on more than one
species at a time.
This Project Summary was developed
by EPA's Environmental Research Lab-
oratory, Corvallis, OR, to announce key
findings of the research project that is
fully documented in a separate report
of the same title (see Project Report
ordering information at back).
Introduction
A rapid, simple, and precise bioassay is
needed to measure the adverse effects of
toxic substances on terrestrial vegetation
in the environment. Root elongation tests
were conducted on five plant species
including lettuce (Lactuca satrva L var.
Buttercrunch), radish (Raphanus sativus
L. var. Cherry Belle), wheat (Triticum
aestivum L. var. Stephens), cucumber
(Cucumis sativus L. var. Hybrid Spartan
Valor), and red clover (Trifolium pratense
L. var. Kenland). These species are im-
portant economically as agricultural crops.
They are also important ecologically in
terms of family size, distribution, and
abundance. Herbicide bioassays, heavy
metal screening tests, salinity and mineral
stress tests, and allelopathic studies show
these plants to be sensitive to many toxic
compounds. Each species was selected to .
represent a different crop rJirectly or in-
directly consumed in the human food
chain. These plants germinate quickly and
easily, root growth is rapid and relatively
uniform, and the cultivars are readily
available.
The chemicals used included silver
nitrate, sodium fluoride, cadmium chlor-
ide, methane arsonic acid (MAA), endo-
sulfan, 2.4-D acid, polyethylene glycol
20,000 linear (Carbowax®), pentachloro-
nitrobenzene (PCNB), cineole, and mon-
uron. These substances represent various
types of toxic substances such as inor-
ganic metals, inorganic nonmetals, organo-
metalhc compounds, herbicides, fungi-
cides, insecticides, osmoticum, and allelo-
pathic substances.
Test substances were prepared to ob-
tain nominal concentrations and were ad-
justed to pH 6.5. A controlled environ-
ment chamber was used to maintain a
uniform testing temperature at 25 ±*1°C.
A standardized technique was used to
germinate the seeds on filter paper sub-
strate in glass tanks for 11 5 hours. The
root of each plant was measured on a flat
surface and germination and seedling
morphology data was recorded.
The first test included a wide range of
concentrations to determine whether the
-------
test chemical was phytotoxic and to find
the appropriate sensitivity range for each
species. Subsequent tests included chem-
ical concentrations selected to cover the
sensitivity range previously estimated.
The chemical concentrations selected were
generally equally spaced on a log scale and
were used only after they satisfied certain
qualifying criteria. For example, one of the
following criteria had to be satisfied before
a chemical concentration was identified as
reducing the root length of a given species
by 50% of the control:
1. At least one test must include three
concentrations in which mean root
length is between 1 2-85% of the con-
trols with at least one greater than 50%
of control and at least one smaller than
50% of the control.
2. At least one test must include one
concentration in which the mean root
length is between 60-80% of the con-
trol and a second concentration in
which the mean root length is between
15-40% of the control.
If the mean root length was based on the
germination of fewer than 10 seeds per
plate for controls on test concentrations,
those data were not used under either
criteria above.
The EC50's were calculated by select-
ing data only from the linear portion of the
dose-response curve. A linear regression
was used to fit the data set selected. Using
the regression equation, the chemical con-
centration required to reduce the root
length to 50% of the control was calculated.
Results
The control root lengths for each species
at each laboratory are sjiown in Table 1.
The mean root length (x) represents 1 5
roots of each species measured in each of
(N) number of control tests. The variation
(s) in mean root length within laboratories
is a result of seedling variability, the varia-
tion in environmental controls, and the
experimental error of the testing Jabora-
tory. The coefficient of variation (V) was
computed to show the relative precision
for control root lengths.
An analysis of variance (Table 2) shows
a significant difference in control means
among laboratories. There is no significant
difference in control means among chem-
icals tested (Zero dose, chemical). Statis-
tically, there was a significant interaction
between laboratory and zero dose (chem-
ical) and between laboratory and species.
However, only the laboratory and species
interaction was considered large enough
to be important.
Except for Carbowax and cineole, seed
germination was not significantly inhibited
for those chemical concentrations that
inhibited root growth. Although reduced
in length, the seedlings were either normal
in appearance or showed necrosis, or were
weak and spindly. With 2,4-D, the roots
were often short, stubby, and thickened.
Figure 1 shows a typical dose-response
curve for test data in which root length was
plotted against the concentration (ex-
pressed as the log 10) of the treatment.
Data in the linear portion of the curve
where small changes in dose result in
large changes in root length best describe
the dose-response relationship. The tox-
icity parameter (EC50) in this assay is
defined as the chemical dose (mg/l) re-
quired to reduce root length of treated
seedlings to 50% of the control.
The EC50 values and the multiplicative
standard error were calculated from the
toxicity test data from each of the labora-
tories. Results for endosulfan and penta-
chloronitrobenzene (PCNB) gave no sig-
nificant inhibition of root growth for any
species or laboratory.
For some chemicals there were large
differences in species sensitivity, while in
others the EC50 estimates did not differ
greatly. For example, EC50's for methane
arsenic acid (MAA) range from 464-962
mg/l for wheat and from 9-27 mg/l for
red clover. For sodium fluoride, the ECBO's
for wheat and red clover range from 1 1 9-
286 mg/l and 217-425 mg/l, respec-
tively. Estimating an EC50 for cineole and
Carbowax was difficult because germina-
tion and root elongation were inhibited at
nearly the same chemical concentrations.
An analysis of variance of the log 10 of
the estimated EC50's shows that there is a
significant difference among six of the
chemicals evaluated for toxicity on the five
plant species. The differences due to
chemical are considerably larger than
those for species. There are no significant
differences among the laboratory esti-
mated ECBO's for chemicals and species.
This implies that all laboratories were
equally accurate in estimating EC50's for
the chemicals and species. The overall
EC50 means were calculated for all lab-
oratories to compare chemicals and species
(Table 3). Wheat is consistently the most
tolerant species (largest EC50) for the
chemicals tested except for sodium fluorida
Lettuce and red clover were generally the
most sensitive (smallest EC50) although
for sodium fluoride lettuce has the largest
EC50.
By comparing the responses of different
species to the chemical tested, an indica-
tion of the nature of the chemical toxicity
can be determined Sodium fluoride was
unique because all the species tested
were considerably more tolerant to it than
the other five chemicals tested. Also, the
response for all species occurred in a
similar concentration range. These results
indicated a nonselective mode of action for
sodium fluoride in inhibiting root growth.
The range of mean EC50's for silver
nitrate and cadmium chloride were 3-156
mg/l and 7-92 mg/l, respectively. These
results indicate a more selective mode of
action in which chemicals inhibit root
growth to a greater degree in some species
than in others.
For monuron and 2,4-D, the differences
in the mean EC50's are small if wheat is
excluded. These results indicate a less
distinctive selective mode of action.
There is a significant difference in
standard error of the EC50 estimate for
laboratories. Although the laboratories
were able to accurately estimate an EC50,
Table 1. Control Mean Root Length (x) in mm. Standard Deviation (s), and Coefficient of Variation (V) for Laboratories and Species
Red Clover
Laboratory
1
2
3
4
5
6
7
x
/V
63
44
50
54
74
56
41
x
25
23
27
25
22
32
28
26
s
4
3
5
6
5
6
4
V
17
14
20
24
25
17
15
N
64
35
49
54
76
54
41
Lettuce
x
69
60
62
59
50
67
64
62
s
5
6
8
12
12
7
8
V
7
10
12
20
23
10
13
N
64
44
50
54
76
56
41
Wheat
x
80
66
91
76
78
64
76
76
s
8
12
12
17
11
10
7
V
10
18
13
22
14
15
10
N
64
44
50
54
74
60
41
Cucumber
x
103
111
118
112
87
113
95
106
s
12
10
16
15
23
13
19
V
12
9
13
13
27
12
20
N
64
44
50
54
75
58
41
Radish
x
129
134
156
138
95
133
126
130
s
29
24
29
39
36
28
18
V
23
18
19
28
37
21
14
x
82
79
91
82
67
82
78
-------
Table 2.
Analysis of Variance Table for Control Mean Root Length by Laboratory, Species, and
Zero Dose (Chemical) with the Laboratory x Species x Zero Dose Interaction
Designated as Error
Source
df
MS
P-value
Laboratory
Zero Dose (Chemical)
Species
Laboratory x Zero Dose
Laboratory x Species
Zero Dose x Species
Laboratory x Zero Dose x Species
6
7
4
42
24
28
168
0.0622
0.002 1
4.0820
0.0035
0.0163
0.0017
0.0015
41.47
1.40
2,721.33
2.33
10.87
1.13
2.10
2.01
2.37
1.38
1.52
1.48
Totals
279
80
o
•g
70-
60-
50-
40-
20-
10 -
*
*
**:
-2.0
-1.0 0.0
Concentration (Log mg/IJ
\
1.0
2.0
doses were selected in terms of equal
spacing on a log scale and the number of
doses on the linear part of the response
curve, 2) the slope of the regression line,
3) the accuracy of the preparation of the
chemical dose, and 4) the variability of the
root length response for each species. The
critical factor appears to be the allocation
of an adequate number of chemical con-
centrations which give responses in the
linear portion of the dose-response curve.
Conclusions
Reduction in root length is a valid and
sensitive plant response to exposure to
chemical substances. It is a suitable test
for evaluating phytotoxic substances over
a wide range of concentrations.
The interlaboratory testing procedures
in this bioassay study provided a uniform
root growth environment The results with-
in and among laboratories showed this
uniformity. Most variation was due to
biological differences among species.
Using a single method to evaluate five
species simultaneously was cumbersome.
It is difficult to satisfy all criteria for all the
species at the same time. The precision of
the EC50 estimates varied significantly
indicating the need for more data points to
adequately define the dose-response rela-
tionships for all five species. A minimum
of four observations in the linear portion of
the dose-response curve should be used.
These points should also be equally spaced
on a log scale with the interval between
doses approximately 1.5 to 2.0.
Although this test is difficult to use on
five species at once, it might be used more
efficiently for one species at a time. The
results would give some indication of the
general response of most species but
would not reflect the response of all
species.
Figure 1. Dose-response curve for lettuce and cadmium chloride using data from laboratory
No. 1.
there was enough variability in technique
among laboratories that the precision of
the estimates differ significantly. The
standard error of the EC50 estimates are
also significantly different for species. The
interactions between laboratory, chemical.
and species do not appear to be an im-
portant factor, although the laboratory x
chemical and chemical x species inter-
actions are statistically significant
The precision in which the EC50 is
estimated depends on 1) how well the
-------
Table 3. EC50 (mg/l) Geometric Means for All Laboratories for Chemical and Species t
Species
AgNO3 CdCI2 Monuron MAA
NaF
2.4-D PCNB
Carbowax
Cineole
Endosulfan
Cucumber
Lettuce
Radish
Red Clover
Wheat
19(4)
3(5)
65(3)
96(2)
156 (1)
18(2)
7(5)
17(3)
16(4)
92(1)
83(3)
79(4)
91 (2)
60(5)
176(1)
87(2}
18(4)
25(3)
18(5)
655 (1)
165 (5)
489 (1)
345 (2)
303 (3)
236 (4)
0.03 (2)
0.03 (3)
0.04 (4)
0.04 (5)
2.54 (1)
*
*
*
*
*
74,000(5)
147,00(^(3)
99,000° (4)
157,000a(2)
207.000(1)
1695(2)
7 17° (5)
1257b(3)
2158(1)
1084(4)
*
*
*
*
*
(t) Species ranks within chemical where 1 is most tolerant and 5 is most sensitive are in parentheses.
* No significant inhibition of root growth.
(a) Geometric mean based on estimates from 4 laboratories.
(b) Geometric mean based on estimates from 5 laboratories.
(°) Geometric mean based on estimates from 6 laboratories.
The EPA author Hitman C. Ratsch is with the Environmental Research
Laboratory, CorvaHis, OR 97333.
The complete report, entitled "Interlaboratory Root Elongation Testing of Toxic
Substances on Selected Plant Species," (Order No. PB 83-226 126; Cost:
$8.50, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA author can be contacted at:
Environmental Research Laboratory
U.S. Environmental Protection Agency
200 S. W. 35th Street
Corvaflis, OR 97333
«US GOVERNMENT PRINTING OFFICE 1983-659-017/7199
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
0000329
------- |